An Evaluation of Factors Related to the Disproportionate Representation of Children of Color in Santa Clara County's Child Welfare System: Child Welfare Practices and Ethnic/Racial Disproportionality in the Child Welfare System, Phase 3, Final Report
An Evaluation of Factors Related to the Disproportionate
Representation of Children of Color in
Santa Clara County's Child Welfare System:
Child Welfare Practices and
Ethnic/Racial Disproportionality in the Child Welfare System
Phase 3
Final Report
Submitted to the County of Santa Clara
Social Services Agency
Department of Family and Children's Services
By
The Child Welfare Research Team
College of Social Work
San Jose State University
September 2, 2003
(revised November 17, 2003)
Principal Contributors
Child Welfare Research Team
Alice M Hines, Ph.D.
Principal Investigator
Peter Allen Lee, Ph.D.
Co-Investigator and Primary
Data Analyst
Laurie Drabble, Ph.D.
Co-Investigator and Primary
Researcher on Qualitative
Component
Kathy Lemon, MS. W.
Senior Research Assistant
Julian Chow, Ph.D.
School of Social Welfare,
University of California at
Berkeley, Consultant
Alfred Perez, MS. W.
The Pew Commission on
Children in Foster Care
Georgetown University,
Consultant
Lonnie R. Snowden, Ph.D.
School of Social Welfare,
University of California at
Berkeley, Consultant
TABLE OF CONTENTS
PAGE
EXECUTIVE SUMMARY
ACKNOWLEDGEMENTS
I.
Introduction
i
xi
II.
Overview ofStudy Objectives and Approach for Phase 3
6
III.
Background and Literature Review
8
Impact of Child Welfare Practices on Children and Families of Color
10
Beginning Choice Points
10
Beginning Choice Points: Impact on Children and Families of Color
13
Decision-Making Practices at Beginning Choice Points
13
Decision-Making Practices: Risk Assessment Tools
Decision-Making Practices: Multidisciplinary Teams/Family Group
Conferencing
16
16
Decision-Making Practices at Beginning Choice Points: Impact on
Children and Families of Color
18
Voluntary Family Maintenance Practices/Family Preservation Services
Voluntary Family Maintenance Practices/Family Preservation Services:
20
20
Impact on Children and Families of Color
Home-Based Prevention Efforts
21
Home-Based Prevention Efforts: Impact on Children and Families of Color
Continuing Care and Ending Choice Points
22
23
Factors Associated with Family Reunification
23
Family Reunification Services
Family Reunification Services: Impact on Children and Families of Color
Parenting Education Services
Parenting Education Services: Impact on Children and Families of Color
Substance Abuse Treatment
Substance Abuse Treatment: Impact on Children and Families of Color
Domestic Violence Treatment
Domestic Violence Treatment: Impact on Children and Families of Color
24
26
26
27
28
29
29
31
TABLE OF CONTENTS
PAGE
Social Network Interventions/Family to Family Initiative
31
Family to Family/Social Network Interventions: Impact on Children and
Families of Color
33
Mental Health Services for Children in Out-of-Home Care
33
Mental Health Services: Impact on Children of Color in Out-of-Home Care
Foster Parent and Kinship Support Services
34
35
Foster and Kinship Support Services: Impact on Children and Families of
Color
36
Permanency Planning Practices
37
Permanency Planning Practices: Impact on Children and Families of Color
39
Impact of Contextual Factors on Child Welfare Practices for Children
40
and Families of Color
The Role of Cultural Competency in Child Welfare Practice
40
Cultural Competency: Impact on Children and Families of Color
Rural Child Welfare Practices
Rural Child Welfare Practices: Impact on Children and Families of Color
41
42
43
The Role of the Court System in Child Welfare Practice
The Role of the Court System in Child Welfare Practice: Impact on
43
44
Children and Families of Color
Summary of Key Findings and Implications
IV.
Santa Clara County Child Welfare Practices
45
47
Successful Cases Analysis
48
Methods
Safety Outcomes
Permanency Outcomes
Family and Child Well-Being Outcomes
Summary of Key Findings and Implications
Qualitative Analysis of Santa Clara County Child Welfare Practices
48
48
52
56
58
61
Methods
Beginning Choice Points: Practices that Negatively Impact Children and
61
66
Families of Color
TABLE OF CONTENTS
Beginning Choice Points: Practices that Positively Impact Children and
PAGE
69
Families of Color
Beginning Choice Points: Recommendations to Improve Practices
70
Continuing Care/Ending Choice Points:
71
Practices that Negatively Impact Children and Families of Color
Continuing Care/Ending Choice Points: Practices that Positively Impact
74
Children and Families of Color
Continuing Care/Ending Choice Points: Recommendations to Improve
75
Practices
Contextual Factors that Negatively Impact Practices for Children and
76
Families of Color
Contextual Factors that Positively Impact Practices for Children and
80
Families of Color
Contextual Factors: Recommendations to Improve Practices for Children
81
and Families of Color
Summary of Key Findings and Implications
84
Comparison of the Main County Offices and South County Offices
87
Quantitative Comparison
87
Quantitative Methods
87
County Location and Child Characteristics
88
County Location and System-Related Factors
90
Qualitative Comparison
97
Qualitative Methods
97
Context of Service Provision in South County Offices
98
Practices in Service Provision in South County Offices
99
Summary of Key Findings and Implications: Quantitative and Qualitative
101
Comparison of Main County Offices and South County Offices
Multivariate Analysis of Factors and Predicting Reunification:
103
Preliminary Models
Results: Closed Case Sample
I 03
Summary of Key Findings and Implications: Closed Case Sample
104
Results: Case Record Review Sample
I 05
Summary of Key Findings and Implications: Case Record Review Sample
108
TABLE OF CONTENTS
PAGE
The Role of the Court System in Child Welfare Practice
109
Quantitative and Qualitative Methods: Analysis of Court-Ordered Changes
109
Quantitative Results: Analysis of Court Ordered Changes
111
Qualitative Results: Analysis of Court Ordered Changes
113
Qualitative Methods: Relationship Between the Child Welfare and Court
114
Systems
Qualitative Results: Relationship Between the Child Welfare and Court
114
Systems
Summary of Key Findings and Implications: The Role of the Court System
116
in Child Welfare Practice
V.
Statewide Comparative Analysis
118
Methods
118
Contextual Tables: County CJiild Populations and CWS Involvement
119
Descriptive Results: Effective Practices for Children and Families of
124
Color
Beginning Choice Point Practices
124
Continuing Care/Ending Choice Point Practices
135
Both Beginning and Continuing Care/Ending Choice Point Practices
146
Organizational or Structural Practices
154
Summary of Key Findings and Implications: Descriptive Results
165
Multivariate Analysis: Impact of Statewide Practices on Differential
170
Representation
VI.
Multivariate Methods
171
Latino Overrepresentation
174
African American Overrepresentation
174
Asian American/Pacific Islander Underrepresentation
175
Summary of Key Findings and Implications: Multivariate Analysis
177
Phase 3 Summary and Conclusions
VII. Summary of Phases 1,2 and 3: Overall Conclusions and Implications for Santa
180
194
Clara County's Child Welfare System
References
203
TABLE OF CONTENTS
Attachments
1. Social Worker Focus Group Interview Guide
2. Summary Table of Focus Group Participant Demographics
3. Family Focus Group Interview Guide
4. Supervisor Individual Interview Guide
5. Court Case Record Review Data Extraction Form
6. Court Officer Unit Focus Group Interview Guide
7. Statewide Survey on Effective Practices for Children and Families of Color
8. Summary Chart of Phase 2 Findings
Executive Summary
Introduction
While child abuse and neglect appears to affect children of all racial and ethnic
origins (US Department of Health and Human Services, 1998; Sedlack & Broadhurst,
1996), an analysis of national, California and Santa Clara County data on the ethnicities
of children in out of home placement reveals that, compared to their presence in the
general population, there is a disproportionate involvement of children of color in the
public child welfare system (CWS). In Santa Clara County, in particular, when compared
to the general population, African American, Hispanic/Latino and Native American
children are overrepresented in Santa Clara County's CWS, while Asian
American/Pacific Islander and White children are underrepresented. African Americans
represent 4 percent of the general child population in the county, and are 13.6 percent of
children in supervised foster care. Hispanic/Latino children represent 30 percent of the
general child population in Santa Clara County and constitute 55.3 percent of the child
welfare cases. Native Americans are approximately 0.5 percent of Santa Clara County's
population and represent 0.6 percent of children in the CWS. Asian American/Pacific
Islander children represent 21 percent of the general county child population and 5.4
percent of children in the CWS; Whites constitute 45 percent of the general child
population and 25.0 percent of the child welfare population (Needell et al., 2003; U.S.
Bureau of the Census, 2000).
The disproportionate involvement of children of color in the CWS has long been
an issue of concern for CWS workers, clients, researchers and government and
community groups; yet few studies have systematically investigated the factors
associated with this disproportionality. In an effort to better understand the factors related
to the disproportionate number of children of color in the CWS in Santa Clara County,
the Department of Family and Children Services (DFCS) contracted with the Child
Welfare Research Team (CWRT) in the College of Social Work at San Jose State
University to conduct a three-year study on this topic.
The overall question posed by DFCS was: What are the primary reasons why
children of color are disproportionately represented in Santa Clara County's Child
Welfare System? In order to address the complexity of this question, the CWRT elected
to employ a multiphase/multimethod approach, beginning with an initial exploratory
phase that was completed in May 2001. Data for Phase I were gathered from three
sources: research literature at the national, state and county levels; Santa Clara County's
management information system (CWS/CMS); and focus group discussions with
professionals in Santa Clara County who provide child welfare services, as well as
parents, caregivers and youth who are recipients of CWS services. Four central themes
emerged from Phase I findings (please see Phase 1 Final Report for a complete
description of the findings from which these themes were generated).
Conclusions from Phase 1 indicated that more research was needed that focused
on the actual experiences of children in care and the individual and family-related
characteristics that are associated with these experiences. Phase 1 results also indicated
that CWS/CMS data alone would not provide the information needed to address these
research aims. Thus, Phase 2, which concluded in August 2002 focused on characteristics
of children and families of color in the CWS and experiences they had as they wended
their way through the system. The methodology for Phase 2 included extensive, in-depth
case record reviews with a sample of 403 closed child welfare case records, enabling us
to collect information on the individual child and family, as well as system-related
characteristics not included in the CWS/CMS database. The case record review data
provided extensive information on decisions at key points in the system and children's
progress through the system, as well as services ordered for children at the time of case
disposition and at subsequent hearings. In addition, a parallel descriptive analysis of 1720
closed cases within the CWS/CMS database allowed us to examine the entire histories of
children as they progressed through the CWS and to conduct analyses that were not
possible with the CWS/CMS cross-sectional dataset that we used in Phase 1. Key
informant interviews with managers and supervisors in the county's DFCS at key
decision points in the system provided additional information pertaining to cultural and
environmental aspects of the agency. (Please see Phase 2 Final Report for a complete
description of the study methods and findings).
The current report provides findings from Phase 3 that was originally to be
conducted from September 2002 to August 31, 2003. However, due to contractual
changes involving the Scope of Work, the CWRTwas not able to begin work until the
end of January 2003. Thus, the research involved in Phase 3 was conducted over a
shorter seven-month time period. This third phase of the research project focused on
agency practice and ways in which current practice interacts with child, family and
cultural characteristics of different racial/ethnic groups. Phase 3 of the study addressed
three overall aims: 1) to assess identified key practices for children and families of color
in Santa Clara County's CWS, 2) to examine the influence of the court system on service
recommendations for children and families and, 3) to conduct a statewide comparative
analysis of identified key practices that may affect the disproportionate representation of
children of color in the CWS.
The primary methodology for Phase 3 included semi-structured in-depth
interviews and focus groups with DFCS supervisors and managers, social workers and
parents involved in the CWS; a qualitative and quantitative comparison of child welfare
practices and outcomes between the DFCS South County and Main County offices; a
reanalysis of cases from Phase 2 with successful outcomes and an analysis of factors that
contributed to those positive outcomes; a supplemental data collection for our case record
review sample so as to include information on court ordered changes to the social worker
recommendations at the jurisdictional/dispositional hearing; and, a statewide survey of
county child welfare directors, managers or supervisors in California counties and
quantitative analyses using a statewide database consisting of county-level
characteristics.
ii
Overall Conclusions
Results from Phase 2 indicated that different racial/ethnic groups experience
unique pathways through the CWS (Hines et al., 2002). At every point in the system,
various child welfare practices and services impact children and families. Phase 2 results
indicated that children and families of color in Santa Clara County's CWS tend to be
recommended services that are traditional and formal in nature and that do not appear to
meet the wide range of needs experienced by these highly diverse racial/ethnic family
groups. Indeed, children and families of color in the CWS represent a high risk group
who are often impacted by myriad psychosocial challenges yet services recommended for
these families tend to be limited to a one-size-fits all approach.
Moreover, there exists relatively little empirical evidence that traditional child
welfare services are effective in maintaining or reunifying children and improving family
functioning. Even less research has been conducted on the impact of child welfare
practices for children and families of color. Yet because children and families of color
are disproportionately represented in the CWS, the effectiveness of child welfare
practices for these groups is of particular interest. Research suggests that children and
families of color tend to have longer stays in the CWS, that they receive fewer and less
comprehensive services, and have poorer case outcomes than White children and families
(Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996). This bleak
outlook for children and families of color in the CWS creates an urgent need to
understand child welfare practices and services that result in enhanced outcomes for these
families. It is especially important to identify the types of individual and contextual
factors that hinder or support the effectiveness of these practices.
Phases I and 2 of this study focused on examining reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where it
occurs. The primary goal ofPhase 3 was to assess key practices for children and
families ofcolor in the CWS at both the county and state levels in order to provide
information that might help generate strategies aimed at reducing ethnic/racial
disparities and develop more effective services and supports for children and families of
color.
In order to develop a context for examining practices in Santa Clara County's
CWS, we first conducted analyses to assess types ofpractices and case characteristics
associated with successful outcomes for children and families across different
racial/ethnic groups. Using data from our Phase 2 Case Record Review sample and
defining successful outcomes according to US DHHS guidelines (i.e., safety,
permanency, and family and child well-being), results indicated that there were minimal
ethnic differences in successful outcomes. It is important to note that major ethnic
differences do exist in Santa Clara County's CWS. Phase 1 and 2 confirmed that certain
ethnic groups are over-represented (i.e., Latinos, African Americans) while others are
under-represented (i.e., Asian American/Pacific Islanders) in the county's CWS. Phase 2
results also indicated that there are significant ethnic differences in demographic, systemrelated, and psychosocial characteristics and that many of these variations appear at early
iii
choice points in the system. Phase 3 results suggest that when success is defined
according to DHHS guidelines, ethnic differences at the conclusion of the case are not
evident. One interpretation for this finding might be that while there is a clear
disproportion of ethnic groups in the system, once in the system, children as a whole,
regardless of race/ethnicity, are generally faring the same. Phase 2 results found that
children and families of color in Santa Clara County's CWS are receiving a "one-size fits
all" approach to services (Hines et al., 2002). Although this may seem equitable and
logistically efficient, it appears ultimately ineffective in serving a diverse group of
children and families.
While little research exists that empirically supports policies, programs and
practices that attempt to reduce ethnic/racial disproportionality, some programs and
practices appear promising and merit further exploration. In general, these practices aim
to reduce the need for out-of-home placement and increase family and community
participation in decision making.
In our analysis of Santa Clara County child welfare practices, results based on
focus group and interview data indicated that in-home services, family conferencing, and
team decision making (TDM) were viewed by participants as ways to improve decisionmaking withfamiliesfrom diverse cultural and ethnic groups. Participants noted that both
formal and informal collaborative relationships with other services delivery systems, such
as domestic violence and substance abuse treatment systems, are critical for successful
interventions with children and families. Furthermore, participants recognized the
importance of training and cross-training to enhance the capacity of helping professionals
in child welfare and other systems to better address the needs of children and families of
color.
It is interesting that most practices and contextual factors mentioned by social
workers had both positive and negative aspects. Many of the practices and contextual
factors that negatively impact children and _families of color were actually barriers to
successful implementation of the practices that social workers felt have a positive impact
on children and families of color. Although practices such as orientation of new clients;
family and team decision-making; preventive services; substance abuse services; cultural
competency; collaboration with other agencies and systems; a strength-based approach;
and the use of relative placements were described as having a positive impact on children
and families of color, certain negative practices and contexts impeded their effective
implementation. For instance, having to screen out large numbers of inappropriate
referrals; inconsistency in decision-making practices; a shortage of services, particularly·
substance abuse and preventive services; lack of client access to services; difficult
protocols for placing children with kin; gaps in cultural competency; time limits; and
agency-level factors such as heavy caseloads, staff shortages, substantial amounts of
paperwork, lack of access to information about resources; and confusion about the
agency's overall mission and key policies all impeded implementation of the best
practices. These findings suggest that in addition to building on current positive practices
for children and families of color, CWS stakeholders should also actively work both
internally and in collaboration with partners to reduce barriers to these best practices.
iv
Our analysis comparing South County and the Main Office enabled us to examine
contextual factors that may have an impact on child welfare practices, in this case, small
vs. large and urban vs. rural. Results indicate that children in South County appear to
have better outcomes, i.e., they have fewer number of unique placement homes in the
current episode, have a shorter average stay per placement, a shorter length of time in
out-of-home placement, and are younger at time of case closure. Study participants
described South County as "a small county or rural area." This geographic difference,
and factors related to an organizational culture that emphasize a commitment to
maintaining and reunifying families, the practice of providing an orientation to the child
welfare system, as well as the finding that fewer workers are assigned during the course
of the case may all be related to these positive outcomes for children in South County. It
is interesting to note, however that our multivariate analyses that examined case
characteristics and system-related factors related to reunification indicated that when
taking other factors into account, geographical location was not a significant predictor of
reunification. Rather, fewer workers, shorter case duration, and fewer placements were
the strongest predictors. More research on factors related to agency culture and
geographical location as they impact child welfare practice and outcomes for children and
families of color is clearly warranted.
Research indicates that the court system is likely to have an impact on the
practices that are provided to children and families in the CWS. In our analysis
conducted to determine the extent to which court orders changed social worker
recommendations at the jurisdictional/dispositional hearing, results indicated that the
court made changes to the initial social worker in approximately halfofthe cases.
Results further indicated that that ethnicity was not related to whether or not a change
occurred These results do not necessarily indicate that the system and courts treat
children and families from different ethnicities the same, but rather that when the judge
decides to make a change from the social worker recommendations, these changes appear
to occur equally acros~ ethnic groups. It was the opinion of members of the Court
Officer Unit focus group that ethnicity does not contribute to decisions in court, but
rather it is circumstances that drive the case.
Findings from qualitative analysis exploring the relationship between the child
welfare and court system painted suggested that the collaboration between the child
welfare system and courts is problematic. Agency-wide focus group participants felt that
the child welfare and court system have different perspectives on the needs and
circumstances of children and families in the CWS and that the court system may have
unrealistic expectations of families. More research is this area is needed, as is further
investigation of ways in which relationships between court personnel and child welfare
workers have an impact on the experiences of children and families of color involved in
the CWS.
Results from our statewide survey indicate that certain types ofpractices are
considered by county child welfare personnel to be effective with divers_e populations,
and that these practices are related to disproportionate representations ofchildren of
V
color in county child welfare systems. Descriptive results revealed four overall types of
practices considered to be most effective with children and families of color, as well as
various specific practices within these types of practices including: 1) beginning choice
point practices, which included family group conferencing or other team decision-making
practices, voluntary family maintenance or other prevention/family preservation services,
and risk assessment tools or systems, 2) continuing care/ending choice point practices,
which included, placement related practices, the Family to Family Initiative, and
intensive and collaborative interventions for high-risk children, 3) both beginning and
continuing/ending choice point practices, which included, community-based services, and
strength based assessments, and 4) organizational or structural practices which included,
cultural competency of staff, collaboration with American Indian tribes and vertical case
management.
Similar to findings on practices identified within Santa Clara County's DFCS,
practices at the statewide level were described as effective with children and families of
color because they are inclus_ive, collaborative, culturally appropriate and involve client's
communities. Although the research literature does not provide convincing evidence of
the effectiveness of these practices, the fact that respondents reported these practices as
those that are most effective with children and families of color suggests that they may
serve as promising models for children and families of color and that more rigorous
evaluations of these practices are needed.
Phase 3 also attempted to assess how the practices identified through the
statewide survey might affect differential representation ofchildren of color in the CWS.
Our analyses uncovered significant associations between practice types, practices and
practice characteristics and disproportionate representation. It is interesting that different
patterns of association were found for Latinos, African Americans and Asian
American/Pacific Islanders.
Specifically, beginning choice point practices were found to have a beneficial
effect on Latino overrepresentation. In particular, family group conferencing or other
team decision-making interventions predicted lower levels of Latino overrepresentation.
However, Latino overrepresentation was greater, not less, when interventions were
implemented targeting areas with high concentrations of distressed families. Latino
overrepresentation was also greater when clients, foster families, out-of-home placement
providers, and other service systems were involved in the development of the practice. As
such, it appears that practices may be developed because stakeholders perceive Latino
overrepresentation and initiate programs in response.
In contrast to Latinos, targeting practices to at-risk areas for African American
populations was beneficial in reducing overrepresentation, as well as targeting of very
young children. Less overrepresentation was linked also to whether foster families and
clients and contract agencies initiated an intervention, and whether it focused on
improving psychosocial outcomes of birth families. The findings suggest a picture almost
the reverse of that found for Latinos. Additionally, there appear to be lower levels of
African American overrepresentation than expected when key community stakeholders
vi
help to initiate the interventions that focus on high-risk areas and populations. When
interventions have a goal of strengthening the well being of birth families, then
implementation is associated with less African American overrepresentation. Yet when
counties implemented Family-to-Family initiatives greater African American
overrepresentation resulted. Again, the implementation of Family to Family programs
may be used reactively, in response to high perceived levels of overrepresentation.
As with Latinos, targeting of high need areas was less beneficial for Asian
American/Pacific Islander children than not engaging in this kind of intervention.
Involving clients and service systems in initiation, again more common for interventions
targeting high need areas, was also relatively unsuccessful. On the other hand,
management and worker support for a practice was associated with greater success.
While researchers have described the existence of ethnic/racial disparities in every
aspect of the CWS, little attention has been given to ways to address the existing
disproportinality. It is our hope that the information contained in Phase 3 of the Children
of Color Study will help in the development of more effective programs and practices for
families and children of color at both the county and state level. Overall, our findings
from Phase 3 suggest that efforts to address racial and ethnic disparities in the child
welfare system should involve a more concentrated focus on child and family well-being,
involving the creation of family-centered and community-based services that are
inclusive and collaborative and specifically designed to meet the needs of the diverse
cultural and ethnic groups in the child welfare population.
Recommendations for the CWS
Based on Phase 3 findings, we make recommendations/suggestions in the following
areas:
1) Tracking successful case outcomes for families and children of color:
•
In order to better assess whether safety, permanency, and family and child
well-being are maintained and enhanced, records should include more
complete and measurable information. We acknowledge that child
welfare case records were not originally intended for research. Yet, if the
agencies responsible for providing effective services wish to evaluate the
performance of their programs more reliably and validly, improving
information collection and tracking should be considered.
•
The creation of a case summary check-out form used at case closure
would be one way to track child outcomes. The check-out form would
indicate the child and family's status in major areas of psychosocial wellbeing (i.e., health, education, and finances), ·in addition to the existing case
outcomes. Given the amount of missing data assessing these domains
describing case and child status, and the relative scarcity of information
vii
recorded in the last two court hearings, this form would enable
measurement and evaluation of "successful cases and outcomes" in a more
consistent objective manner.
•
The amount of missing data and the manner in which information is
assessed also implies the need for more reliable and valid methods of
recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their
families or of those adopted high or low? And compared to what?
Implementing better record keeping and data collection over time on key
-characteristics, based on an agreed-upon operational definition would
better address the question of effectiveness and performance.
Implementing these practices across counties and across states would also
facilitate evaluation.
·
2) Identifying promising practices within Santa Clara's CWS for families and children of
color:
•
The findings from the qualitative component of this study affirm the value
of many efforts that are already in place, such as culturally specific ER
response units and family resource centers, and other that are in the
process of implementation including efforts to reduce caseloads, initiation
of team decision making, and participation in the Family-to-Family
initiative.
•
Practices that are inclusive, collaborative, culturally appropriate and
involve client's communities appear to be best suited to meeting the needs
of children and families of color. Although the research literature does not
provide convincing evidence of the effectiveness of these practices, the
fact that respondents reported these practices as those that are most
effective with children and families of color suggests that they may serve
as promising models for children and families of color and that more
rigorous evaluations of these practices are needed.
•
The findings also point to opportunities to further strengthen services for
children and families of color. Recommendations related to training,
institutionalizing group orientation for clients, and other suggestions from
study participants could be used to inform practice and planning.
•
It would be beneficial for a team of managers, line staff, family
representatives, and community members to review the recommendations
suggested by focus group and interview participants to identify and
prioritize possible practices for adoption based on their feasibility and
utility for children and families of color.
viii
3) Identifying promising practices for families and children of color based on a
comparison between South County and the Main Office:
•
Future research should include a larger subsample from South County and
include methods to understand the reasons behind these location
differences. Also, a larger South County Office location sample may
enable more detailed comparisons by ethnicity.
•
Practices, such as providing a group orientation to families that may assist
them in navigating through the system may be easily strengthened in
South County Offices and adapted to other regions of Santa Clara County,
as recommended by study participants. This practice may be particularly
helpful to low-income families and families of color that may be
intimidated by or unfamiliar with child welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for
other regions of Santa Clara County, though the success of this practice
may be linked to the "small county" dynamic described by South County
study participants. Some participants in focus groups that took place at the
main office mentioned this model as promising while others asserted that
specialized expertise in different areas of social work was an advantage in
providing quality services that are informed by "the most current
information."
•
Further examination of ways in which the culture of commitment to
maintaining and reunifying families is created, communicated, and
continued merits attention. In addition, an exploration of how this culture
might be adapted to other parts of the county would be beneficial.
4) Identifying factors related to the Court System that may have an impact on practices
for families and children of color in the CWS:
•
Although the quantitative and qualitative information described many of
the circumstances related to whether or not recommendations may be
upheld by the court, it is very difficult to ascertain the degree to which
each circumstances influences a court ordered change. More examination
of this area is needed, as is further investigation of ways in which
relationships between court personnel and child welfare workers have an
impact on the experiences of children and families of color involved in the
cws.
ix
•
Better and more exhaustive record keeping may not be the solution, as
many informal and undocumented actions in the negotiation process occur
regularly, and are confidential, thus making much of the data unavailable
for research and evaluation purposes.
•
Assessing other factors such as courtroom environment, stakeholder
characteristics, and other judicial issues would be advantageous in future
research.
5) Identifying promising practices for families and children of color based on statewide
data:
•
County directors and child welfare administrators identified practices with
similar characteristics as those identified within Santa Clara County, i.e.,
those that are inclusive, collaborative, culturally appropriate and involve
client's communities appear to be best suited to meeting the needs of
children and families of color. However, further research and evaluation is
necessary to determine which practices most effectively meet the needs of
individual racial/ethnic groups.
•
While multivariate analyses suggest that certain practices do predict levels
.of over and underrepresentation of children of color in the child welfare
system, additional research to determine the mechanisms through which
these practices impact this disproportionate representation is needed.
•
Additional explorations into the relationship between practices,
disproportionate representation of children of color in the child welfare
system and county characteristics would also help shed light on the
process through which certain practices impact children and families of
color.
X
Acknowledgements
This report on Phase 3 of the Children of Color study could not have been
completed without the important contributions of so many people.
We wish to thank the former Director of Santa Clara County Social Services
Agency, Yolanda Rinaldo, the initiator of an important discussion with Dr. Sylvia R.
Andrew, the Dean of the College of Social Work at San Jose State University, about the
possibility of a joint research project with the Agency and the College.
Thank you to Leroy Martin, former Director of the Department of Family and
Children's Services, who provided valuable oversight for the research project until his
retirement and to Norma Doctor Sparks, current Director of the Department of Family
and Children's Services. Many thanks also to Will Lightbourne, Director, County of
Santa Clara, Social Services Agency.
We would also like to thank Zonia Sandoval Waldon, Deputy Director of the
Department of Family and Children's Services for her assistance throughout the third
year of the project. Thank you also to Ken Borelli, Sue Farr and Connie Vega.
Judi Boring and her staff, David Heinen and Yvonne Zing, were especially
helpful with providing access to the CWS/CMS system and arranging for the data set that
was used in this report. Thanks also to Diana Weir and her staff who were responsible
for gathering the additional CWS/CMS data for Phase 3. Judy Bushey was extremely
helpful in guiding us through the court report sections of the child welfare case records.
Thank you to Stan Lee and his staff who provided valuable training on the court report
section of the child welfare case records and to Maria Gonzalez who made the case
records available for coding.
Our special thanks to Dean Sylvia R. Andrew who provided members of the
research team with the precious resource of matched release time. Thank you also to
Simon Dominguez for his valuable feedback and encouragement.
The completion of Phase 3 was also made possible by our talented and dedicated
student research assistants: Lisa Navarrette, Priscilla Ribeiro, and Graciela Escobar. Their
work on the qualitative focus groups and supervisor interviews was truly invaluable.
Bernard Voon, who worked as a student research assistant for Phase 2. of the research
project, generously returned to help with supplemental case record reviews in Phase 3 as
a volunteer. Each of our student assistants brought noteworthy enthusiasm, dedication
and insight to the project. Their substantial contributions are deeply appreciated.
In addition, this research could not have been carried out without the cooperation
of the staff of the agency, and we would like to acknowledge the importance of the
xi
contributions of everyone who took the time to attend focus groups, participate in
individual interviews, or assist in organizing staff and family focus groups.
A special thank you to all the child welfare directors and administrators who took
the time to participate in our statewide survey on effective practices for families and
children of color. Their contributions were extremely valuable.
Thanks to Shaaron Gilson for her assistance as Project Manager and for her
thoughtful comments on an initial draft of the report and to Carol Pearce for her
assistance with the statewide survey and the case record review data collection.
Thank you to Jerri Carmo, Juliana Brooks, Bill Yabumoto and the staff at the San
Jose State Foundation for their help in administering this important project.
Finally, we would like to thank the Santa Clara County Board of Supervisors,
especially Blanca Alvarado and James Beall for their ongoing support of the Children of
Color study and their commitment to improving child welfare practices for children and
families in the system.
xii
I.
INTRODUCTION
An analysis of national data on the ethnic composition of children in out-of-home-placement
reveals that, compared to their presence in the general population, there is a disproportionate
involvement of children of color in the public child welfare system (CWS). Statistics indicate that
African American and Native American children are overrepresented in the CWS. African
American children represent 15 percent of the general child population, yet comprise approximately
42 percent of children in the CWS and Native American children constitute approximately 1 percent
of the child population and represent approximately 2 percent of the CWS. Though not all states
provide data on Hispanic/Latino children in the CWS, aggregate statistics from those that do reveal
that while Hispanic/Latino children make up approximately 16 percent of the national child
population, they comprise 15 percent of the CWS population. Asian American/Pacific Islanders and
Whites also tend to be underrepresented at the national level. Asian American/Pacific Islander
children comprise approximately 4 percent of the general child population and 1 percent of the
CWS, while White children who are approximately 64 percent of the general child population,
constitute 36% of children in the CWS (Federal Interagency Forum on Child and Family Statistics,
2001; U.S. Census Bureau, 2001; U.S. Department of Health and Human Services, 1999; U.S.
Department ofHealth and Human Services, 2001).
Further analysis at the California level provides important information on differences and
trends that may go undetected at the national level. In California, a large and ethnically diverse
state, African Americans constitute 6 percent of the general population (32 percent of whom are
below the age of 18), but represent approximately 32.9 percent of children in supervised foster care.
Hispanics/Latinos comprise approximately 32 percent of the general population in California with
43 percent being below the age of 18, and represent 36.2 percent of children in the state CWS.
Native Americans represent 1.0 percent of California's population and constitute approximately 1.2
percent of the children in the CWS. Asian American/Pacific Islanders are under-represented in
California's CWS, as they constitute approximately 11 percent of California's general population,
with approximately 30 percent being younger than 18, but represent just 2.1 percent of the children
in the CWS. While Whites comprise approximately 47 percent of the population in the state, with
only 20 percent being below the age of 18, they constitute 27.5 percent of the children in
California's CWS (Needell, Webster, Cuccaro-Alamin, Armijo, Lee, Brookhart et al., 2003; Reyes,
2001; U.S. Census Bureau, 2000).
In Santa Clara County, there also exists a disproportionate representation of children of
color in the County's CWS. When compared to the general population, African American,
Hispanic/Latino and Native American children are overrepresented in Santa Clara County's CWS,
while Asian American/Pacific Islander and White children are underrepresented. African Americans
represent 4 percent of the general child population in the county, and are 13.6 percent of children in
supervised foster care. Hispanic/Latino children represent 30 percent of the general child population
in Santa Clara County and constitute 55.3 percent of the child welfare cases. Native Americans are
approximately 0.5 percent of Santa Clara County's population and represent 0.6 percent of children
in the CWS. Asian American/Pacific Islander children represent 21 percent of the general county
child population and 5.4 percent of children in the CWS; Whites constitute 45 percent of the general
child population and 25.0 percent of the child welfare population (Needell et al., 2003; U.S. Bureau
of the Census, 2000).
2
Project Description
The disproportionate involvement of children of color in the CWS has long been an issue of
concern for CWS workers, clients, researchers and government and community groups; yet few
studies have systematically investigated the factors associated with this disproportionality. In an
effort to better understand the factors related to the disproportionate .number of children of color in
the CWS in Santa Clara County, the Department of Family and Children Services (DFCS)
contracted with the Child Welfare Research Team (CWRT) in the College of Social Work at San
Jose State University to conduct a three-year study on this topic.
The overall question posed by DFCS was: What are the primary reasons why children of
color are disproportionately represented in Santa Clara County's Child Welfare System? In order to
address the complexity of this question, the CWRT elected to employ a multiphase/multimethod
approach, beginning with an initial exploratory phase that was completed in May 2001. Data for
Phase 1 were gathered from three sources: research literature at the national, state and county levels;
Santa Clara County's management information system (CWS/CMS); and focus group discussions
with professionals in Santa Clara County who provide child welfare services, as well as parents,
caregivers and youth who are recipients of CWS services. Four central themes emerged from Phase
1 findings (please see Phase 1 Final Report for a complete description of the findings from which
these themes were generated). The four emerging themes from Phase 1 included:
1). Little is known about specific pathways through the CWS and ways in which these
pathways differ for various racial/ethnic groups. Much of the research on children in the
CWS focuses on factors relates to movement in and out of the system. The specific
pathways through the system for children of various ethnicities is largely unknown, and may
provide important insights into the factors related to the disproportionate involvement of
children of color. Focus group participants in Phase I frequently stated that once a child of
color enters the system, it is very hard for that child to exit.
2). Various racial/ethnic groups may receive different treatment at key decision-making
points in the system. Differential treatment of children of color in both the initial reporting of
child maltreatment and throughout the various stages within the CWS may exist. CWS/CMS
data for Santa Clara County corroborate the possible differential treatment of children of
color at various points in the system. Phase 1 analysis revealed that African American
children spent significantly more time in placement than their White, Hispanic/Latino, and
Asian American/Pacific Islander peers. Differences in placement type were also foundalthough kincare (placement with a relative) was the most frequent out of home placement
(OHP) type across ethnicities, the second most frequent OHP type varied among
racial/ethnic groups. For African American, Native American, White, and Hispanic/Latino
children, the second most common OHP type was a Foster Family Agency. For Asian
American/Pacific Islander children, the second most frequent placement was a Foster
Family Home. Asian American/Pacific Islander (18%) children were also placed at the
Children's Shelter at higher percentages than children of other racial/ethnic groups. In
addition, focus group participants in Phase 1 expressed concern for possible racial/ethnic
differences in ways in which services and resources were allocated and ways in which the
judicial system handled cases.
3
3). System level changes recently initiated at the federal, state and local level will
undoubtedly have an impact on children of color in the CWS. Recent federal policies,
including the Multi-Ethnic Placement Act-Interethnic Adoption Provision (MEPA-IEP,
1996); the Adoption and Safe Families Act (AFSA, 1997); the Personal Responsibility and
Work Opportunity Reconciliation Act (PRWORA, 1996); as well as state and local county
policies all impact the ways in which families and children of color enter and stay in the
CWS. Recent policy shifts toward expedited permanent placements for children in out-ofhome care and shortened timelines for reunification, as well as new restrictions for
Temporary Assistance to Needy Families (TANF) recipients (more than half of children in
foster care are from welfare eligible families) may increase the likelihood of children of
color entering and staying in the CWS-although much more research is still needed to
determine this association.
4). Factors related to the disproportionate representation of children ofcolor in the CWS
are multiple and complex. No clear consensus from the research literature exists on how
families and children of color become and stay involved in the CWS. Available evidence
suggests that, rather than one primary cause, there appear to be numerous and interrelated
factors associated with disproportionate rates of children of color in the CWS. Factors found
to be associated with CWS involvement for children of all ethnicities, include but may not
be limited to: parental mental illness substance abuse, and incarceration; spousal abuse;
living in poverty; living in neighborhoods with concentrated poverty, and living in
communities with low levels of social organization. For children of color specifically,
research also indicates that race and class biases in initial reporting and subsequent CWS
service delivery do exist.
Conclusions from Phase I indicated that more research was needed that focused on the
actual experiences of children in care and the individual and family-related characteristics that are
associated with these experiences. Phase 1 results also indicated that CWS/CMS data alone would
not provide the information needed to address these research aims. Thus, Phase 2, which concluded
in August 2002 focused on characteristics of children and families of color in the CWS and
experiences they had as they wended their way through the system. The methodology for Phase 2
included extensive, in-depth case record reviews with a sample of 403 closed child welfare case
records, enabling us to collect information on the individual child and family, as well as systemrelated characteristics not included in the CWS/CMS database. The case record review data
provided extensive information on decisions at key points in the system and children's progress
through the system, as well as services ordered for children at the time of case disposition and at
subsequent hearings. In addition, a parallel descriptive analysis of I 720 closed cases within the
CWS/CMS database allowed us to examine the entire histories of children as they progressed
through the CWS and to conduct analyses that were not possible with the CWS/CMS crosssectional dataset that we used in Phase 1. Key informant interviews with managers and supervisors
in the county's DFCS at key decision points in the system provided additional information
pertaining to cultural and environmental aspects of the agency. (Please see Phase 2 Final Report for
a complete description of the study methods and findings). Based on analyses conducted during
Phase 2, five overall conclusions were drawn:
1). Families belonging to each of the four racial/ethnic groups studied, including: African
Americans, Hispanic/Latinos, Asian American/Pacific Islanders and Whites are
characterized by unique constellations of risk factors.
4
2). The characteristics that emerged in our analyses as key characteristics ofeach of the
four racial/ethnic groups have been associated in the research literature as risk/actors/or
child abuse and neglect. These characteristics have also been associated with increased
serious behavioral and adjustment problems in children and adolescents.
3). Once in the child welfare system, children in each ofthe four racial/ethnic groups follow
different pathways and experience different outcomes.
4). The services orderedfor families ofcolor are generally limited to a one-size-fits all
approach and to a small array ofavailable services. The reliance on traditional formal
services does not appear to meet the needs of these culturally diverse families and children.
5). There is a need to involve multiple social service systems in a comprehensive and
coordinated effort to meet the needs ofchildren and families ofcolor.
The current report provides findings from Phase 3 that was originally to be conducted from
September 2002 to August 31, 2003. However, due to contractual changes involving the Scope of
Work, the CWRTwas not able to begin work until the end of January 2003. Thus, the research
involved in Phase 3 was conducted over a shorter seven-month time period. This third phase of the
research project focused on agency practice and ways in which current practice interacts with child,
family and cultural characteristics of different racial/ethnic groups. Phase 3 of the study addressed
three overall aims: 1) to assess identified key practices for children and families of color in Santa
Clara County's CWS, 2) to examine the influence of the court system on service recommendations
for children and families and, 3) to conduct a statewide comparative analysis of identified key
practices that may affect the disproportionate representation of children of color in the CWS.
The primary methodology for Phase 3 included semi-structured in-depth interviews and
focus groups with DFCS supervisors and managers, social workers and parents involved in the
CWS; a qualitative and quantitative comparison of child welfare practices and outcomes between
the DFCS South County and Main County offices; a reanalysis of cases from Phase 2 with
successful outcomes and an analysis of factors that contributed to those positive outcomes; a
supplemental data collection for our case record review sample to gather information on the role of
the court system in child welfare practice, including court ordered changes to the social worker
recommendations at the jurisdictional/dispositional hearing; and, a statewide survey of county child
welfare directors, managers or supervisors in California counties and quantitative analyses using a
statewide database consisting of county-level characteristics.
Overview of the Report
The next section of this report provides an overview of the study objectives and approach for
Phase 3, as well as the specific research questions that guide the analyses presented in each section.
The third section reviews the recent research on evidence-based child welfare practices and their
application to children and families of color. Section four focuses on the impact of child welfare
practices on children and families of color in the Santa Clara County's Department of Family and
Children Services (DFCS). Several related analyses were conducted to understand the types of
practices and case characteristics that are associated with more successful outcomes for children and
families from various racial/ethnic groups; the types of child welfare practices used in Santa Clara
County; how practices in the Main County Offices differ from South County Office practices; how
5
practices interact with child, family and cultural characteristics of differing racial/ethnic groups to
produce various outcomes; and the role of the court system on child welfare practices. Section five
describes results from our statewide comparative analysis of effective practices for children and
families of color in the child welfare system. A summary of the report, as well as recommendations
and conclusions are contained in section six.
Definition of Terms used in this Report
Racial/ethnic group refers to cultural heritage and country of origin. Racial/ethnic group is used
interchangeably with the terms, children of color and ethnic group.
Disproportionate representation ofchildren ofa particular racial/ethnic group refers to the
difference between the proportion of children of a particular racial or ethnic group in the CWS and
the proportion of children of a particular racial or ethnic group in the general population.
Overrepresentation ofchildren ofa particular _racial/ethnic group exists when the proportion of
children of a certain racial or ethnic group in the CWS exceeds its proportion in the general
population.
Underrepresentation ofchildren of a particular racial/ethnic group exists when the proportion of
children of a certain racial or ethnic group in the CWS is less than its proportion in the general
population.
African American refers to children of African American heritage. The term Black is included in
this category.
Asian American/Pacific Islander refers to children of Asian American and Pacific Islander heritage.
The following population groups are included: Asian Indian, Cambodian, Chinese, Filipino,
Guamanian, Hawaiian, Hmong, Japanese, Korean, Laotian, Polynesian, Samoan and Vietnamese
and other Asian American/Pacific Islanders.
Hispanic/Latino/a refers to children of Mexican American and Latin American heritage. The terms
Hispanic and Latino/a are included in this category.
Native American refers to children of Native American heritage and includes those designated as
Alaska Natives/Aleuts. The term American Indian is included in this category.
White refers to children of European heritage. The term Caucasian is included in this category.
Other category - Those in the Other ethnic subgroup included Native Americans, Ethiopians,
White-Armenians, and White-Middle Easterners
6
II.
OVERVIEW OF STUDY OBJECTIVES AND APPROACH
FORPHASE3
While Phases I and 2 of this study focused on investigating reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where they occur, the
overarching goal ofPhase 3 was to focus on practices that take place within the CWS that either
promote or reduce disproportionality. More information about child welfare practices including
programs, services, strategies, policies, and/or tools can help address racial/ethnic disproportionality
by guiding the development of more effective services and supports for children and families of
color in the CWS.
In an effort to examine ways to address disproportionality within the CWS, the CWRT explored
three overall aims during Phase 3. The first aim included an assessment of identified key practices
for children andfami/ies ofcolor in the Santa Clara County's CWS. Specific research questions
related to this aim include:
•
•
•
•
•
•
What are the key practices in Santa Clara County's CWS that may have an impact on
children and families of color?
What empirical evidence exists for the effectiveness of these practices?
Have these practices been evaluated regarding their efficacy with varying racial/ethnic
groups?
How do these practices interact with child, family and cultural characteristics of differing
racial/ethnic groups involved in Santa Clara County's CWS?
How do practices in South County vs. the Main Office differ? (i.e., are there differences in
worker socialization, training, values/attitudes, stated and unstated expectations norms that
contribute to differences in practice and client outcomes?)
What practice characteristics are necessary in order to promote more positive outcomes for
children and families from different racial/ethnic groups involved in Santa Clara county's
CWS?
The methodology used to explore these research questions included 1) a review of the current
research literature on evidence-based practice for children in the CWS and the application of these
practices to children and families of color, 2) semi-structured in-depth interviews with DFCS
supervisors and managers, 3) semi-structured focus groups with DFCS social workers and parents
in the CWS, 4) a qualitative and quantitative comparison of child welfare practices and outcomes
between the DFCS South County and Main County offices, and 5) a review of cases from Phase 2
of the research with successful outcomes and an analysis of factors that contributed to positive
outcomes.
The second aim of Phase 3 included an exploration ofthe influence of the court system on
child welfare practice, including court ordered changes to the social worker service
recommendations at the jurisdictional/dispositional hearing. We were interested in whether the
court system imposed any changes to the social workers' recommendations offered at the
jurisdictional/dispositional hearing in a child's case, and if these changes differed by the child's
7
ethnicity, as well as the nature of the relationship between the child welfare and court systems.
Specific research questions related to this aim include:
•
•
•
•
To what degree does the court system impose changes to the social workers' case plan
recommendations?
What types of changes to the social workers recommendations does the court system
order?
Do court ordered changes to the social workers' recommended case plan differ by child's
ethnicity?
How does the relationship between the child welfare and court system positively or
negatively impact child welfare practices and outcomes for children and families of color?
The methodology used to explore these research questions included case record reviews of the 403
closed child welfare cases used in Phase 2 of the research. Cases were reviewed again and
information on court ordered changes to the social worker recommendations at the
jurisdictional/dispositional hearing was gathered. Additionally, agency-wide focus groups provided
information on the how the relationship between the child welfare and the court system has an
impact on practices and outcomes for children and families of color.
The third aim of the research involved a statewide comparative analysis of identified key
practices that may affect the disproportionate representation ofchildren ofcolor in the CWS.
Specific research questions associated with this aim include:
•
•
•
What are key practices in California counties that may have an impact on children and
families of color?
What practices are considered most promising by individual counties for children and
families of color?
Comparing counties with large racial/ethnic representation, what factors predict
over/underrepresentation of various racial/ethnic groups?
The methodology used to explore these research questions included 1) a statewide survey of county
child welfare directors, managers or supervisors in California counties and 2) quantitative analyses
using a statewide database consisting of county-level characteristics.
Phase 3 focused on exploring ways to address the disproportionality of children and families
of color within the CWS by examining practices that may either contribute to or reduce ethnic/racial
disparities. In order to accomplish the aims outlined above, we combined qualitative and
quantitative research methods and examined both practices within the system, as well as contextual
factors that can potentially have an impact on what transpires within the system. The report begins
with a review of the literature on effective practices for children and families in the CWS, followed
by a close up view of practices within Santa Clara County and then broadens to include a wider
perspective on statewide child welfare practices aimed at children and families of color. The
following section will describe the background and literature related to effective practices for
children and families of color in the CWS.
8
III. BACKGROUND AND LITERATURE REVIEW
Different racial/ethnic groups experience unique pathways through the child welfare system
(CWS) (Hines et al., 2002). At every point in the system, various child welfare practices and
services impact children and families. Phase 2 results indicated that children and families of color in
Santa Clara County's CWS tend to be recommended services that are traditional and formal in
nature and that do not appear to meet the wide range of needs experienced by these highly diverse
racial/ethnic family groups. Indeed, children and families of color in the CWS represent a high risk
group who are often impacted by myriad psychosocial challenges (see Phase 2 report for a full
description), yet services recommended for these families tend to be limited to a one-size-fits all
approach.
Moreover, there exists relatively little empirical evidence that traditional child welfare
services are effective in maintaining or reunifying children and improving family functioning. Even
less research has been conducted on the impact of child welfare practices for children and families
of color. Yet because children and families of color are disproportionately represented in the CWS,
the effectiveness of child welfare practices for these groups is of particular interest. Research
suggests that children and families of color tend to have longer stays in the CWS, they receive
fewer and less comprehensive services, and have poorer case outcomes than White chHdren and
families (Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996). This bleak
outlook for children and families of color in the CWS creates an urgent need to understand child
welfare practices and services that result in enhanced outcomes for these families. It is also crucial
to identify the types of individual and contextual factors that hinder or support the effectiveness of
these practices.
This section will review the current research on evidence-based child welfare practices and
their application to children and families of color. Practices will be discussed in relation to key
choice points in the CWS including, beginning, continuing care, and ending choice points. Please
see Figure 1 for a flow chart of possible pathways and key choice points in Santa Clara County's
CWS. Additionally, contextual factors including cultural competency of workers and services; the
unique needs of rural child welfare populations; and the role of the court system will be discussed.
Across all stages of the CWS from beginning to end, highlights of practices that appear to be most
effective will be presented and the ways in which these best practices may impact children and
families of color will be discussed.
9
Figure 1: Possible Pathways and Key Choice Points In the Child Welfare System
Referral
(Report of Abuse or Neglect)
Emergency Response
(Screening)
Case Closed
Case Opened:
Further
Investigation
Jurisdictional Hearing:
Decision Regarding Substantiation
of Abuse/Neglect
Case Dismissed:
Unsubstantiated
Allegations
Family
Maintenance
Case Opened:
Voluntary Family
Maintenance
Case Closed:
Short-Term Services
Provided and Family
Stabilized Within 6
Months
Dispositional Hearing:
Substantiated Allegations,
Further Action and Services
Family Reunification:
Placement in FFH, FF A, Group
Home, Kincare, or Other Facility
Bypass
Hearings: Including 6-Month Review
Case Closed:
Family Stabilized
Through Family
Maintenance
Case Closed:
Family Stabilized
Through Family
Reunification
Case Closed:
Permanent Placement into
Adoption, Guardianship,
or Long Term Care
Impact of Child Welfare Practices on Children and Families of Color
Beginning Choice Points
When a child is reported to the CWS, there are various pathways that she or he may take
through the system (see figure I). At the screening stage, the case may be closed altogether with no
further action; it may be opened for six months of voluntary family maintenance services; or the
case may be opened for further investigation. If the case is opened for further investigation, a
jurisdictional/dispositional hearing typically occurs in which there is a judicial finding on whether
the allegation of maltreatment is substantiated. If allegations are found to be unsubstantiated, the
case may be dismissed, however, if substantiation occurs, children may be placed in out-of-home
care, with family reunification services being ordered, or children may be maintained in the home
with orders for family maintenance services. In some cases, families may be bypassed for services
and have their parental rights terminated at this early stage in the system.
Following the pathway that children may take through the CWS, the screening process is the
first entry point into the CWS. On a national level, the U.S. Department of Health and Human
services ([U.S. DHHS], 2001) reports that approximately 2.7 million referrals were received by the
CWS in 2001; 32.7 percent of these referrals were screened out and closed with no further action.
Other national studies have found similar rates of screening out referrals, ranging from 36 percent to
42 percent (Tumlin & Geen, 2000; Wells, Fluke, & Brown, 1995).
Research suggests that certain case characteristics are associated with the decision to open a
case for further investigation or to refer to voluntary family maintenance services. For instance,
reports involving more than one child, children under age two, severe injury, a minor or other type
of injury or physical condition, allegation of sexual abuse, previous reports to the CWS, (Wells et
al., 1995); as well as child welfare cases involving AFDC receipt; alleged parental drug use; a
female primary victim; reports made by a family member; and reports with direct evidence (Karski,
1999) all have a higher likelihood of investigation than cases not involving these characteristics.
Additionally, cases involving children of color are opened for investigation at a higher rate
than cases involving White children. In Wells et al. 's (1995) analysis of 12 child welfare sites,
African American children were significantly more likely to have reports investigated than children
of other ethnicities, however when the data was dis-aggregated and examined by sites, it was found
that almost the entire sample of African American children reported to the CWS came from two
sites that exhibited particularly high investigation rates. Yet, in an analysis of data from five states,
Fluke, Yuan, Hedderson, and Curtis (2003), found that African American children were
significantly overrepresented at the stage of investigation in each of the five states, as well as in all
counties within the five states, while White children tended to be underrepresented at the stage of
investigation. Rates of disproportionality at the investigation stage for other racial/ethnic groups,
including Asian American/Pacific Islander, Hispanic and Native American children showed much
wider variation. Depending on the state, these groups were both underrespresented and
overrepresented at the investigation stage.
Factors associated with the provision of voluntary services are somewhat less well
understood. In a statewide survey in Missouri, Drake ( 1996) analyzed factors associated with the
provision of preventive voluntary services to 3,442 children and families in the CWS, in
11
comparison to cases that were substantiated by the court. Cases were divided into three types of
maltreatment: sexual abuse, physical abuse and neglect. Across all three types of maltreatment,
cases from rural areas were significantly more likely to receive preventive services than cases from
urban or metro areas. Additionally, predictors of preventive service provision in sexual abuse cases
and neglect cases included those that were referred by a non-anonymous source. Physical abuse
cases were more likely to receive preventive services when a mandated reporter reported them.
Neglect cases were more likely to receive preventive services when they involved White children
and families, parents who were unemployed and children and families from non-poor
neighborhoods (Drake, 1996).
For children who are not offered voluntary family maintenance services, investigations
result in a decision regarding whether to substantiate the maltreatment and take further action or to
not substantiate and dismiss the case. Nationally, 27.5 percent of investigated referrals are
substantiated (U.S. DHSS, 2001). State data suggest similar substantiation rates. One study
conducted in New York state revealed a 30 percent substantiation rate across types of maltreatment,
with a 48 percent substantiation rate for physical abuse, 39 percent for sexual abuse and 28 percent
for neglect (Eckenrode, Powers, Doris, Munsch, & Bolger, 1988). Interestingly, there appears to be
little relationship between screening rates and substantiation rates. States with higher screening in
rates do not appear to have higher rates of substantiation of child maltreatment (Tumlin & Geen,
2000).
In California, the overall rate of substantiation of all referrals is 21.2 percent (Needell et al.,
2003). Statewide rates of substantiation between various racial/ethnic groups appear similar; Black
children have a 22 percent substantiation rate, White children 21.4 percent, Hispanic children 23.4
percent, Asian children 24.5 percent, and Native American children 25.3 percent. Across all ethnic
groups, younger children have higher rates of substantiation than do older children (Needell et al.,
2003).
Santa Clara County's overall rate of substantiation in 2002 was 17 .1 percent--a rate
somewhat lower than the state average. Santa Clara County also appears to have greater disparities
between racial/ethnic groups in substantiation rates. For instance, Black children in Santa Clara
County have an overall substantiation rate of21 percent, White children 15.4 percent, Hispanic
children 18 percent, Asian children 15.l percent, and Native American children 15.l percent. Rates
of substantiation between racial/ethnic groups in Santa Clara County also depend on the age of the
child. As is the case within the state, younger children have a higher likelihood of substantiation,
however, in Santa Clara County it appears that Black children between Oand 2 years of age have a
relatively high rate of substantiation (76.8%), whereas, White children 0-2 have a substantiation rate
of 62.6 percent, Hispanic children 52.6 percent; Asian children 46 percent, and Native American
children 66.6 percent (Needell et al., 2003).
Certain factors appear to be associated with the decision to substantiate an allegation. For
instance, reports from professional reporters tend to be substantiated at a higher rate than other
reports (Eckenrode et al, 1988; English, Marshall, Coghlan, Brummel, & Orme, 2002), as well as
cases referred by law enforcement (English et al., 2002). Poverty and early child bearing have been
identified as strong predictors of substantiation (Lee & George 1999). A history of prior referrals or
involvement in the CWS are also associated with higher rates of substantiation (Eckenrode et al.,
1988; English et al., 2002), and one statewide survey in Washington state found that referrals from
rural areas were more likely to be substantiated, although this effect was small (English et al.,
12
2002). Additionally, cooperation with child welfare agencies has been found to be associated with a
lower likelihood of substantiation (English et al., 2002). Factors associated with substantiation also
depend to some degree on the type of maltreatment alleged. For instance, sexual abuse reports that
involve older, female children in large households tend to have a higher substantiation rate than
cases not involving these characteristics,·and the same is true for neglect reports that involve
younger children with previous child welfare referrals (Eckenrode et al., 1988).
Cases involving children of color appear to have an elevated likelihood of substantiation.
One large study conducted in New York state found that for allegations of physical abuse or
neglect, African American and Hispanic cases were substantiated at a higher rate than White cases
(Eckenrode, et al., 1988). In a more recent study conducted in Minnesota all types of maltreatment
were substantiated at a higher rate for non-White groups than for White groups; this effect remained
even after factors such as type of reporter, type of maltreatment, geographic location and
characteristics of the child and perpetrator were statistically controlled (Ards, Myers, Malkis,
Sugrue, & Zhou, 2003). Similarly, in Drake's (1996) statewide analysis of child welfare cases in
Missouri, non-White racial status was a significant predictor of substantiation even when other case
characteristics were statistically controlled. In contrast, Fluke et al.'s (2003) analysis of
disproportionate representation of race/ethnicity in investigation and victimization rates in five
states found little disproportionaility in rates of substantiation of maltreatment.
In addition to substantiation of an allegation, decisions at the beginning stages of the CWS
are also focused on whether to place a child in out-of-home care. Studies have consistently found
certain factors to be associated with the decision to remove a child from home. Children of color, in
particular African American children have an elevated likelihood of out-of-home placement
(Needell, Brookhart, & Lee, 2003; U.S. DHHS, 2001). Other factors associated with the decision to
remove a child include inadequate or low income (Berry, 1991; Lindsey, 1991); cases involving
prior maltreatment reports (Dalgleish & Drew, 1989; U.S. DHHS, 2001); cases involving families
with many children (Berry, 1991); lack of cooperation from the family (Dalgleish & Drew, 1989;
Davidson-Arad, 2001; Karski, 1999); cases involving young children, mothers as perpetrators, and
reports made by educational personnel are all associated with court intervention and placement is
substitute care (U.S. DHHS, 2001).
Few studies have examined factors associated with the decision to bypass family reunification
services and terminate parental rights at the early stages of the CWS. In California, any one of 13
bypass criteria can be used to deny reunification services and pursue an alterative permanent
placement. These criteria include:
1. Parents' whereabouts are unknown for six months.
2. Parental mental disability that prevents the utilization of services.
3. A sibling has been removed from the home, returned and then removed from parental custody
again.
4. Parent has caused the death of a child.
5. Parent caused severe emotional damage.
6. Parent committed severe physical or sexual abuse to the child, sibling or half-sibling.
7. Reunification has been terminated on a sibling or half-sibling due to indicators# 3, 5 or 6.
8. Parent has been found guilty of rape that conceived the child.
9. Willful abandonment constituting serious danger to the child.
10. Parent failed to reunify with a sibling or half-sibling.
11. Parent convicted of a violent felony.
13
12. Parent has extensive history of substance abuse and resisted treatment for three
years prior, or failed to benefit twice.
13. Parent does not want family maintenance or family reunification services, or does not want
the child removed.
In the Bay Area Social Services Consortium's ([BASSC], 2000) report on the use of bypass
criteria within San Mateo and Santa Clara Counties, approximately half of the cases reviewed had
bypass criteria present. Twenty ·six percent of these bypass cases were due to parental substance
abuse, 18 percent were due to a sibling with parental rights terminated or with a permanent plan,
and 13 percent had a parental mental disability that prevented the utilization of services (BASSC,
2000). However, not all cases with a bypass criterion were actually bypassed. Of the cases with a
bypass criterion present, 73 percent of cases involving parental mental disability; 38 percent of
cases with a sibling who had parental rights terminated or who has a permanent plan; and 13 percent
of cases involving parental substance abuse were actually bypassed. Overall 25 percent of the entire
sample were bypassed and denied reunification services. Additionally, BASSC (2000) reported that
the use of bypass criteria is not empirically or clinically valid and that bypass criteria are not
predictive of a lowered likelihood for reunification. Moreover, the report suggested that that since
many bypass indicators are common within the child welfare population, their use in denying family
reunification services may not be warranted.
Beginning Choice Points: Impact on Children and Families of Color
Research clearly indicates that certain/actors are associated with various decisions at
beginning choice points and these factors disproportionately impact children and families of color.
In general, the most consistent factors found to increase the likelihood of investigation,
substantiation or out-of-home placement include cases involving more than one child, young
children, previous reports to the child welfare system, and families with low incomes (Eckenrode et
al., 1988; English et al., 2002; Karski, 1999; Lee & George, 1999; Wells et al., 1995). Children and
families of color are at an increased risk for all of these factors, but in addition, studies have
indicated that even when these factors are statistically controlled, children of color, and in particular
African American children are more likely to have an allegation investigated and substantiated and
are also more likely to be placed out of the home (Ards et al., 2003; Drake, 1996; Wells et al.,
1995). Additionally, there is evidence to suggest that in cases of neglect, children and families of
color may receive voluntary family maintenance services less frequently than White children and
families (Drake, 1996). Taken together these findings suggest that beginning choice point decisions
may be biased against children and families ofcolor. Indeed it appears from the research that once
a child ofcolor is reported, they are more likely than White children to be set on a pathway through
the child welfare system that involves court intervention and out of home removal.
Decision-Making Practices at Beginning Choice Points
The types of decision-making practices that child welfare workers use at these early stages
of the CWS are crucial in determining the trajectory of the pathway that children and families take
through the system. However, decisions on whether to investigate a referral, provide voluntary
services, substantiate an allegation, remove a child or bypass a case are often complex, ambiguous
and multifaceted. Child welfare workers must simultaneously consider numerous factors, including
child safety, the best interests of the child, parental rights and preserving family integrity. These
14
factors must be considered in a complicated context; children and families in the CWS often face
multiple and complex challenges and child welfare policies and mandates fluctuate frequently.
Given this difficult context, it is not surprising that studies have found a significant lack of
consistency in child welfare decision-making practices across various stages of the system (Rossi,
Schuerman, & Budde, 1999; Tumlin & Geen, 2000; Westat, 2001a). For instance, disparities in
screening out rates between states have been found to range from a low of 5 percent to a high of 78
percent (Tumlin & Geen, 2000), while disparities between counties have been found to range from a
low of 1 percent to a high of 71 percent (Wells, et al., 1995). There is some evidence to suggest that
child welfare workers may use more restrictive screening practices when resources or staff are
reduced or when the number of referrals is high (Curtis, Boyd, Liepold, & Petit, 1995 as cited in
Tumlin & Geen, 2000; Hutchinson, 1989, as cited in Jones, 1993). Moreover, screening decisionmaking practices tend to be influenced by the average severity level of cases within the agency as a
whole and characteristics of the surrounding community. In one survey, workers in 11 child welfare
agencies in the state of New Jersey rated vignettes of child maltreatment reports as less severe in
agencies that handled the more severe cases of child maltreatment. These agencies were also located
in areas that were more socially and economically disadvantaged-suggesting that characteristics
specific to agency sites or communities influence screening and substantiation decision-making
practices (Wolock, 1982).
Although virtually all states and counties screen child maltreatment reports to some degree,
screening practices and policies vary greatly between states and counties. National surveys suggest
that a majority of states do have some form of written agency policies to guide screening and
investigation decisions (Downing, Wells & Fluke, 1990); yet few states have formal or explicit
procedures or instruments delineating reports that should be screened out {Tumlin & Geen, 2000).
Approximately 65 percent of states have a single review screening process in which the decision to
investigate is made by one worker (Tumlin & Geen, 2000). Moreover, the training of child welfare
workers who screen referrals has also been found to vary widely from state to state; with some
states having educational or work experience requirements for screening workers, while others do
not {Tumlin & Geen, 2000).
Studies investigating consistency between child welfare workers in their decision-making
practices indicate wide variability in these practices. For instance, Rossi et al. (1999) explored
decision-making practices among 130 child welfare workers and experts in three states. Results
indicated that although certain factors were associated with the decision to maintain a child in the
home, including families with no prior referrals and cases involving employed caretakers who
showed signs of engagement in services, the thresholds used to trigger an out-of-home placement
decision were not consistent. For instance, some participants decided on out-of-home care in cases
involving just one prior referral, while others recommended out-of-home placements in cases
involving two or more prior referrals. Using the same data set, Schuerman, Rossi and Budde (1999)
calculated the level of agreement between workers on decisions about services, as ranging from a
high of 64 percent to a low of 48 percent, indicating a considerable lack of consensus among
workers regarding out-of-home placement decisions.
Studies also suggest that child welfare workers consider numerous factors in their decisionmaking practices and that these factors are generally not considered in isolation from one another
(English et al., 2002). For instance, a recent study by the U.S. Department of Health and Human
15
Services (2003a) identified four overall factors that influence decisions regarding whether to
substantiate an allegation, including:
1) Case factors, such as child characteristics, severity of the maltreatment, chronicity of
maltreatment, caretaker characteristics, caretaker/child relationship, social and economic
factors, and perpetrator access to the child.
2) Decision-maker factors, such as caseworkers' individual tendencies, experience level,
relationships with co-workers, self-assessments of skills and degree of importance
placed on state policy.
3) Organizational factors, such as workload stress, and degree of supervisor support.
4) External factors, such as state policies.
All of these factors were found to impact the decision-making process of child welfare workers in
complex ways that may differ between individual workers, counties and states (U.S. DHHS, 2003).
Yet there is some evidence to suggest that decision-making practices within a particular
agency may be more consistent than between agencies. Alter ( 1985) has suggested that child
welfare workers tend to use heuristics or assumptions to guide decision-making and that workers
within a particular agency often develop a mutual, although unspoken understanding of the kinds of
cases that should be investigated. Indeed there does appear to be greater consistency in initial
decision-making practices within agencies than between agencies. For instance, Wells et al.'s
(1995) study of 12 child welfare sites, found wide variability between the sites in their investigation
decision-making practices, yet when all other factors were held constant, the site of the report was
more significant than any other factor in predicting an investigation. These results suggest that the
type of decision-making practices common to a particular child welfare agency play an important
role in decision-making practices related at the beginning stages of the CWS.
In addition to the importance of agency norms in decision-making practices, the labels used
to describe families may also trigger certain decisions. For instance, Martin, Peters, and Glisson
(1998) surveyed 100 child welfare workers in five states and found that workers did not make
service and placement recommendations for children based on a uniform psychosocial assessment.
Instead, decisions were influenced more by the labels attached to children upon entry into care and
judicial decisions. Children whose files contained language such as "substance abuser, child of
substance abuser, sex offender, victim of sexual abuse and victim of physical abuse" (Martin et al.,
1998, p. 3), were more likely to be recommended mental health services and more restrictive
placements-regardless of their actual need as indicated in the psychosocial assessment. Similarly,
children who were adjudicated delinquent or who were placed in out-of-home care due to behavior
problems rather than maltreatment were also recommended mental health services and more
restrictive placements, regardless of their actual need.
The influence of agency norms, the use of implicit assumptions and the complexity inherent
in child welfare decisions has prompted many practioners and researchers to call for more objective
or collaborative decision-making practices. Structured risk assessment tools, multidisciplinary
teams, and family group conferencing are considered by many to offer a more objective, balanced
and inclusive method of decision-making at beginning choice points.
16
Decision-Making Practices: Risk Assessment Tools
Structured risk assessment tools are designed to assist child welfare workers in making
accurate assessments at various stages in the CWS. Risk assessment tools generally evaluate factors
that are considered to be associated with various outcomes including such items as prior referrals
and prior substantiated cases; the presence of substance abuse, mental health or economic problems,
family constellation, the type, severity and chronicity of the maltreatment and circumstances
surrounding the current incident (Camasso & Jagannathan, 2000) Risk assessment tools fall into
two general categories: 1) consensus based systems in which risk factors in the instrument are based
on the consensus ofjudgments made by experts in the field, and 2) actuarial systems in which risk
factors are identified based on empirical evidence of factors associated with future maltreatment
(Baird & Wagner 2000). There is some evidence to suggest that actuarial systems more accurately
capture levels of risk in families referred to the CWS (Baird & Wagner, 2000).
The use of risk assessment tools varies greatly from state to state and county to county, and
these tools can be used at various decision points in the CWS from intake to case closing (Wald &
Woolverton, 1990). Although risk assessment tools are widely used in CWS agencies, their ability
to accurately assess levels of risk is an area of considerable debate in the child welfare field (Wald
& Woolverton, 1990). Several researchers have noted that risk assessment instruments have only
minimal reliability and validity in predicting case outcomes (Camasso & Jagannathan 2000;
DePanfilis & Scannapieco, 1994; Doueck, English, DePanfilis, & Moote, 1993; Wald &
Wooverton, 1990).
In 1998, certain California counties, including Santa Clara County began using a
comprehensive risk assessment system, known as Structured Decision-Making (SOM). SDM has
several components including response priority, safety assessment, risk assessment, family needs
and strengths assessment, case planning and management, case reassessment, workload based
resource allocation, and the use of management information systems (Children's Research Center,
2003). These components are designed to assist workers in assessing the severity of allegations,
making structured decisions, managing workloads, minimizing the trauma associated with
maltreatment and preventing the recurrence of maltreatment (California Department of Social
Services, 2003). The risk assessment tool used in this system is an actuarial instrument that is
intended to estimate the likelihood that maltreatment will recur. In addition to the risk assessment,
SOM also calls for a family needs assessment in which family's strengths and resources are to be
considered in conjunction with the risk assessment. Although peer reviewed evaluations of the
SDM project are not available, the Children's Research Center (as cited by the California
Department of Social Services, 2003) reports that a 12-month evaluation on the SOM project was
conducted in Michigan. Results indicated that cases receiving SDM had 27 percent fewer new
referrals, 54 percent fewer new substantiated allegations, 40 percent fewer children removed to
foster care, and 42 percent fewer child injuries than children not receiving SDM. Although these
results are promising, more evaluations are needed to discern the effectiveness of SOM, specifically
for children and families of color.
Decision-Making Practices: Multidisciplinary Teams/Family Group Conferencing
The use of multidisciplinary teams to assist in child welfare decision-making has gained
popularity in recent years. Multidisciplinary teams are designed to better coordinate efforts between
child welfare, law enforcement, legal professionals and medical professionals in making decisions
17
about child welfare cases. These efforts are intended to improve the efficiency and effectiveness of
decision-making activities and also in some cases-to reduce the potential for traumatizing children
through repeated investigatory interviews (Sheppard & Zangrillo, 1996). Approximately two thirds
of states have been found to use multidisciplinary teams in making child welfare decisions (Kolbo
& Strong, 1997).
Research on the effectiveness of multidisciplinary teams in improving decision-making in
child welfare is lacking. In one of the few studies to address this issue, Hochstadt and Harwicke
(1985) evaluated the use of multidisciplinary teams for 180 children referred to the CWS for
suspected child maltreatment. The multidisciplinary team included child welfare workers, medical
professionals and psychological personnel who completed a multidisciplinary evaluation and
participated in a case conference in which specific service recommendations and treatment plans
were made. Although no comparison group was used in the study, findings suggested that the use of
multidisciplinary teams was associated with a high percentage of children and families obtaining
recommended services on their treatment plans. One hundred percent of children with a
recommendation to return home did so, and 92 percent with a recommendation to foster care
followed that recommendation. Other service recommendations that were received included visiting
nurse (76%), additional medical assessment (64%), infant program (64%) and individual
psychotherapy (59%). The authors suggest that the use of multidisciplinary teams can positively
influence the receipt of recommended services.
In an effort to discern how multidisciplinary teams in child welfare work together,
Lewandowski and Glen Maye (2002), conducted a survey of 165 child welfare and community
professionals who were a part of a multidisciplinary child welfare team in an urban Midwest
county. The sample included social workers, counselors, community members, education
professionals, law enforcement, legal professionals and medical professionals. Findings revealed
that 72 percent of respondents had team meetings at regularly scheduled times with support staff
assigned to the team, and 53 percent had a designated team chairperson. Forty eight percent of the
sample indicated that decision-making was most likely to occur around case planning, 32 percent
indicated that decision-making was least likely to occur at intake, and 28 percent reported that
decision-making occurs at all stages. Inadequate communication was identified as the greatest
barrier to effective team decision-making, followed by inadequate resources and inadequate
professional respect. Lastly, although most respondents (63%) agreed that family members should
be active and equal participants in the team decision-making process, only 35.5 percent reported
that family members attended team meetings.
Recently, more concerted attempts to involve family members and extended support systems
in decision-making have occurred. Many child welfare systems are moving toward the use of family
group conferences to improve decision-making. The use of family group conferences began in New
Zealand in 1989, when legislation mandating their use was implemented (Sieppert, Hudson, &
Unrau, 2000). Family group conferences, also referred to as family group decision-making, are
designed to bring together family members, relatives, and other support systems in order to make
decisions about a case. The family group conference is intended to identify the family's strengths
and resources; to develop a plan to ensure child safety and improve family functioning; and to foster
cooperation, collaboration and communication between families and professionals (American
Humane Association, 2003: Pennell, & Buford, 2000). These methods are based on the principle
that families themselves possess the most information about what decisions should be made; the
approach is intended to be family centered, strengths based, and takes into consideration issues of
18
culture and community (American Humane Association, 2003).
In addition to family group conferencing, recent efforts have also been aimed at "team
decision-making," a closely related concept that appears to widen the scope of participants in the
decision-making process. Team decision-making is one component of the Casey Foundation's
Family to Family Initiative, a broad child welfare effort aimed at improving outcomes for children
in the CWS (see page 31 for a discussion of the Family to Family Initiative). The team decisionmaking component of the initiative involves convening a broad range of participants to share
information about the family and develop a plan for child safety, family maintenance or
reunification. The team decision-making meeting can involve any participants who can contribute to
the decision-making process, including family members, neighbors, foster parents, service
providers, child welfare workers and supervisors, other resource staff or other community
representations (Casey Foundation, 2003a).
Research on the effectiveness of family group conferencing and team decision-making on
outcomes for children and families is lacking, although some studies have investigated the process
of implementing family group conferences. For instance Sieppert, et al. (2000) investigated the
process of implementing a family group conferencing pilot project in Canada. The sample included
23 child welfare families with court involvement and one family group conference was conducted
for each family. Mothers were the most frequent family members to attend the conferences (91 %),
while 52 percent involved a father, 70 percent involved grandmothers, 57 percent involved aunts,
and 35 percent involved grandfathers. Other members of the conferences included child welfare
workers (83%), and foster parents (30%). Findings revealed that family group conferences tended to
be resource-intensive. Workers performed a wide range of activities in preparing for and
implementing the family group conferences. An average of 17 telephone contacts, 2.4 in-person
contacts, and 3 letter contacts were made prior to each family conference. Workers also spent time
traveling to meet with prospective conference attendees. On average, workers spent 7 hours
preparing for each conference. Family conferences lasted an average of2.75 hours. At the end of
the conference, families were asked to rate their experience of the group. Seventy percent reported
being highly satisfied with conference preparation and members in attendance; 85 percent felt a
high degree of freedom to speak out in the group; 76 percent reported feeling involved in the
decision-making process and 65 percent felt highly satisfied with the decisions that resulted from
the conference. Unfortunately the ethnicity the families involved in the family group conferences
was not provided.
Decision-Making Practices at Beginning Choice Points:
Impact on Children and Families of Color
In the beginning stages of the child welfare system, the decision-making practices ofchild
welfare workers can have a substantial impact on children and families ofcolor. However in
general there appears to be considerable inconsistency in the types of decision-making practices
used in the child welfare field (Rossi et al., 1999; Schuerman et al., 1999; Tumlin & Geen, 2000;
Westat, 2001a). Although many child welfare systems are moving toward more objective or
multidisciplinary approaches to decision-making, a majority of states still use a single review
process in which decisions to investigate are made by one worker (Tumlin & Geen, 2000); and few
states have formal policies and procedures for decision-making at early stages of the CWS (Tumlin
& Geen, 2000).
19
However, making child welfare decisions in isolation can have deleterious effects on
children and families of color. Research suggests that child welfare workers often use implicit
assumptions to guide decision-making. Ways in which each worker thinks about diverse families,
and the overall spoken and unspoken expectations and norms of a particular child welfare agency
can have an enormous impact on decision-making practices (Cohen, 2003). As Cohen (2003) notes,
The cultural and racial background of families influences the specific factors that workers
consider in assessing the severity of risk and level of intervention. Decisions are more likely
to be made on the basis of deficits in available resources, accepted· agency practice, personal
values and biases, and notions of an ideal family, than by application of consistent case rules
(Chohen, 2003, p. 145).
It is speculated that in some cases, race/ethnicity may impact decision-making by resulting in a
higher level of intervention, while in other cases race/ethnicity may result in less intervention effort.
Much of this variation appears to·depend on the worker and the context in which the worker makes
decisions (Cohen, 2003).
The use of risk assessment tools and multidisciplinary teams/family group conferencing may
have the potential to improve decision-making practices affecting children and families of color.
Although research on the effectiveness of risk assessment tools in assessing risk for future
maltreatment has not been consistently demonstrated in the research literature, the move toward
more objective ways of assessing risk at beginning stages of the child welfare system may be
promising for children of color. For instance, in one study conducted by the California Department
of Social Services (1999), 2,500 substantiated cases from counties that used the SDM system, were
evaluated to assess whether the risk assessment instrument used in SDM had any differential impact
on African American, Hispanic, or White children and families. Findings indicated that Whites had
slightly higher scores on the risk assessment instrument than did African Americans or Hispanics
and the authors concluded that the risk assessment instrument did not disparately affect African
Americans or Hispanics. While this finding is encouraging, other studies have suggested that risk
assessment instruments in general are not always able to accurately predict case outcomes Camasso
& Jagannathan 2000; DePanfilis & Scannapieco, 1994; Doueck et al., 1993; Wald & Wooverton,
1990). More research is needed on the impact of risk assessment tools in general, and their impact
on children and families of color in particular.
Additionally, more research on the effectiveness ofmultidisciplinary teams and family group
conferencing is needed The family centered, strength based and inclusive philosophy offamily
group conferences may be a particularly useful strategy for children and families of color, many of
whom are experiencing numerous risk/actors in a highly stressed environment. The active
engagement offamilies and support systems in decision-making at early stages of the child welfare
system may potentially result in enhanced outcomes for children and families of color.
Yet there is some evidence to suggest that these: approaches can be labor, time and resource
intensive (Sieppert et al., 2000). As such, the effective implementation of these practices may
require the use of increased support staff to assist in scheduling and arranging meetings, as well as
identifying persons to be included in the meeting. Additionally, because beginning choice point
decision-making practices are often made quickly and in the context of crisis situations, the effective
use of multidisciplinary teams and family group conferences requires coordinated efforts and
appropriate support staff. Clearly, more research is needed to determine how group decision-
20
making can be effectively integrated into child welfare agencies and how these practices impact
outcomes for children and families ofcolor.
Voluntary Family Maintenance Practices/Family Preservation Services
In cases where maltreatment is not substantiated, families may be referred to voluntary
family maintenance services. Because the overall goal of voluntary family maintenance services is
to prevent an out-of-home placement and further involvement with the CWS, these services are
considered beginning choice point practices. Much of the research literature on services and
practices aimed at preventing out-of-home placement refer to these services as "family preservation
services," (FPS) or "intensive family preservation services" (IFPS). Many researchers and
practioners have noted that services falling under the rubric of "family preservation services" can be
quite diverse, although they do tend to share the common goal of preventing child maltreatment and
out-of-home placements (United States General Accounting Office, 1997). The model of FPS that is
most often described in the research literature includes short-term, intensive and home-based
services that are aimed at family skill-building, while also including concrete services, such as
public assistance, child care, employment training or transportation assistance (Littell, 1997).
In general, there are numerous shortcomings in the research on the effectiveness of FPS;
many studies use small sample sizes, no comparison groups and often fail to adequately describe the
nature of the intervention and the characteristics of the clients. As a result, evaluations of FPS often
yield conflicting results. While some studies report that FPS do not result in lower placement rates
or improved family functioning (Chaffin, Bonner, & Hill, 2001; Littell & Schuerman, 1995; Littell,
1997; Westat, Inc. 2001b); other studies suggest that FPS do improve outcomes for children and
families (Pecora, Fraser, Bennett, & Haapala, 1991; Scannapieco, 1994; Spaid, Fraser & Lewis,
1991; Wells & Whittington, 1993).
There is also evidence to suggest that FPS outcomes vary depending on case characteristics.
A number of different studies have noted that FPS clients who experience an out-of-home
placement typically have more risk factors and less resources than families who avoid placement.
The particular risk factors affecting FPS clients who experience out-of-home placements have been
found to vary wideJy. For instance, FPS cases involving substance abuse, mental illness, criminal
behavior, disability, low income, or previous CWS involvement have all been linked to an increased
likelihood of an out-of-home placement (Fraser, et al., 1991; Thieman & Dail, 1997; Thieman et al.,
1990, as cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman,
1995).
Voluntary Family Maintenance Practices/Family Preservation Services:
Impact on Children and Families of Color
Research on voluntary family maintenance practices and FPS has not systemically
evaluated the efficacy of these programs for diverse children and families ofcolor. Many studies
fail to perform analyses examining the possible differential outcomes of these services for children
and families of color. In general, research on FPS has not consistently demonstrated that these
services are effective in preventing out-of-home placement or improving family functioning
(Chaffin et al., 2001; Littell & Schuerman, 1995; Littell, 1997; Westat, Inc. 2001b), and some
researchers have concluded that there is no reliable research that supports the effectiveness of these
services (Littell & Schuerman, 1995). However, some studies do suggest that FPS may be most
21
effective with lower-risk families (Fraser et al., 1991; Thieman & Dail, 1997; Thieman et al., 1990,
as cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman, 1995).
Yet because children and families of color referred to the child welfare system often possess
numerous risk factors (Hines et al., 2002), the use of traditional FPS for these diverse clients may
not be effective. Research does suggest that FPS that are most effective with high-risk clients are
those that are more intensive, and those that involve high client participation and the delivery of
concrete services in the home (Berry, 1992; Cash & Berry, 2002; Nelson & Landsman, 1992, as
cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman, 1995).
Home-Based Prevention Efforts
Research suggests that the use of home-based services may be effective in preventing
maltreatment and further involvement in the CWS (Olds, Eckenrode, Henderson, Kitzman, Powers,
Cole et al., 1997). Most research on the use of in-home services has focused on their efficacy for
pregnant women or mothers of young children. While the nature of in-home services may vary, in
general these services are aimed at improving health behaviors of pregnant and parenting women,
facilitating parenting skills, increasing support systems and encouraging healthy parent child
attachments (Olds et al., 1997). Many practioners and researchers have suggested that the perinatal
period is an important window of opportunity and that services offered during this time have
increased potential to offset a number of negative outcomes, including child maltreatment
(Guterman, 1999).
For instance, in one study involving pregnant women in rural New York state, 324 women
participated in a home nurse visitation program that sought to improve health related behaviors
during pregnancy, parenting behaviors and the personal life course development of the mother
(Olds et al., 1997). Specific services included linking families with health and human services;
involving family members and friends in the care of the child; educational services to increase
parenting abilities; and assisting mothers in completing education, obtaining employment and
family planning. Participants received services for two years; outcomes assessed at a 15-year
follow-up indicated that compared to pregnant women who received no home visitation services,
mothers in the program had significantly fewer verified reports of child maltreatment in which they
were the identified perpetrator. The program appeared to have even more of a protective effect on
unmarried women and women experiencing income problems. These groups were the least likely to
have a verified child maltreatment report. Women in the program were also less likely to have
substance abuse problems, had fewer contacts with the criminal justice system, had fewer
subsequent pregnancies and were less likely to use public assistance (Olds et al., 1997).
Interestingly, a separate analysis of the same data set indicated that the protective effects of
home visitation in reducing the likelihood of child maltreatment did not hold true for mothers
experiencing domestic violence. Mothers in the home visitation program who reported more than 28
incidents of domestic violence during the 15-year follow-up period did not experience a reduced
likelihood of verified child maltreatment (Eckenrode, Ganzel, Henderson, Smith, Olds, Powers et
al., 2000). These results suggest that different risk factors may have varying degrees of influence on
outcomes for home visitation program participants.
In addition to child maltreatment prevention, in-home services have also been identified as a
useful public health strategy to improve health related behaviors and outcomes for mothers and
young children. Kitzman, Olds, Henderson, Hanks, Cole, Tatelbaum et al. (1997) evaluated a home-
22
visiting program that involved in-home services delivered by nurses to 1,139 low-income, primarily
African American women during pregnancy and the first two years of the child's life. Results
indicated that, compared to women who had no home-visiting intervention, women who received
in-home services experienced fewer injuries and fewer hospitalizations for their children, and also
experienced fewer pregnancies and a lower rate of pregnancy induced hypertension.
Not surprisingly, research also suggests that the nature of the interventions provided through
in-home programs plays a role in their effectiveness. For instance, some studies on in-home services
report no increased benefit from their use (Barth, 1991; Fraser, Armstrong, Morris, & Dadds, 2000).
However it appears from the literature that in-home programs are most effective when health or
social services professionals, as opposed to paraprofessionals, deliver the services. In-home services
that are long term, generally lasting longer than 6 months are also considered to be more effective
(Barth, 1991; Olds et al., 1997).
Home-Based Prevention Efforts:
Impact on Children and Families of Color
Although differential outcomes for various racial/ethnic groups receiving home-based
prevention services are not often cited in the research, studies do generally support the effectiveness
of these services for vulnerable parenting women. Participation in home-based services has been
linked to a reduced likelihood ofchild welfare system involvement, substance abuse problems,
contact with the criminal justice system, public assistance usage, subsequent pregnancies (Olds et
al., 1997), as well as fewer child injuries and hospitalizations, and better health during pregnancy
(Kitzman et al., 1997).
Additionally, there is evidence to suggest that home-based services are effective with highrisk populations and families of color. For instance Olds et al. (1997) found that home-based
interventions were associated with even better outcomes among unmarried and low SES mothers,
and Fraser et al. (2000) reported that mothers with poor attachment to their child and who had a low
sense of competence were more likely to remain in a home-based program. Kitzman et al. ( 1997)
also noted that in-home services were associated with better outcomes for a sample of primarily
African American mothers. Furthermore, McGuigan, Katzev, and Pratt (2003) reported that older
Hispanic mothers had the greatest likelihood of remaining in a home-based program, suggesting
that in-home services may be particularly useful with families of color.
However research a,so suggests that the effectiveness of home-based interventions depends
to some degree on the types of risk factors present and the components of the interventions. For
instance, the presence of domestic violence has been found to eliminate the protective effects of
home-based interventions (Ekenrode et al., 2000); and unmarried mothers with less than a high
school education; mothers Who received no prenatal care; and mothers who have had multiple births
may be less likely to benefit from home-based programs than mothers without these challenges
(Murphey & Braner, 2000). Furthermore, mothers living in communities with a high rate of
violence have been found to have an increased likelihood of dropping out of home-based programs
(McGuigan et al., 2003); and teenage mothers and those with insecure housing also appear to have
an increased risk of dropping out of home-based programs (Fraser et al., 2000). It also appears
from the literature that home-based programs may be most effective when they last more than 6
months; when they are delivered by health or social service professionals (Barth, 1991; Olds et al.,
23
1997); and when there is a high level of supervision for the workers delivering the services
(McGuigan et al., 2003).
In general, home-based prevention efforts appear to be a promising best practice for
vulnerable child welfare populations. More research is needed to discern what the necessary
components of intervention are and to what degree these interventions differentially impact diverse
groups. Additionally, it is important to note that home-based interventions may be time, labor and
resource intensive and as such, child welfare agencies may have difficulty implementing these
services. However, clients of these services clearly appear to benefit and home-based programs do
tend to result in enhanced outcomes for parents and children.
Continuing Care and Ending Choice Points
If child maltreatment is substantiated, children typically enter the CWS under some form of
court supervision. Generally families who are not bypassed for services are assigned to either family
maintenance or family reunification services (see figure 1). Under the Adoption and Safe Families
Act of 1997, families generally receive up to 18 months of family reunification services, at which
point they are either reunified or parental rights are terminated and the child enters post permanency
planning services. As such, continuing care and ending choice point practices generally include
family reunification services, including specific treatments for parents; as well as services for
children in out of home care, services for alternate caregivers, and permanency planning services
for children who are not reunified with their birth parents.
Factors Associated with Family Reunification
Studies have clearly identified certain factors that are predictive of reunification or nonreunification. Research consistently demonstrates that White children are reunified at a higher rate
than are Black children (Courtney, 1994; Jones, 1998; McMurty & Lie, 1992; Wells & Guo, 1999;
Westat, Inc., 2001c); and young children tend to be reunified more often than older children
(Courtney, 1994; Goerge, 1990; Westat, Inc., 2001c). Harris and Courtney (2003) report that single
parent African American families are particularly unlikely to reunify, while two-parent Hispanic
families have a significantly higher likelihood of reunification than two-parent Caucasian or African
American families. Socioeconomic status also plays a role in reunification; receipt of AFDC and
lower socioeconomic status is associated with a deceased likelihood of reunification (Barth,
Snowden, Broeck, Clancey, Jordan, & Barusch, 1987; Courtney, 1994; Jones, 1998), and Westat,
Inc. (2001c) found that parents with a high school education who were currently employed were
more likely to be reunified than parents who do not have these attributes. Interestingly, Wells and
Guo (2003) report that mothers who move from welfare to work actually reunify _with their children
at a slower rate than other groups. Some studies have also found that children with disabilities or
health problems are less likely to be reunified than children without these problems (Courtney,
1994; McMurty & Lie, 1992; Wells & Guo, 1999). Additionally, parents who are non-compliant
with court orders and treatment are significantly more likely to lose custody of their children than
are court compliant parents (Atkinson & Butler, 1996; Jellinek, Murphy, Poitrast, Quinn, Bishop &
Goshko, 1992; Smith, 2003).
Demographic information indicates that most children in family reunification services come
from single parent homes, often headed by the mother (Fein & Staff, 1993; Gillespie, Byrne, &
24
Workman, 1995; Lewandowski & Pierce, 2002; Walton, Fraser, Lewis, Pecora, & Walton, 1993).
Financial problems are common among family reunification clients, (Fein & Staff 1993; Gillespie et
al., 1995; Walton et al., 1993); as are previous out-of-home placements (Lewandowski & Pierce,
2002; Walton et al., 1993). In California, 16.l percent ofBlack children are in family reunification
services; 25.l percent of White children; 24.8 percent of Hispanic children, 20.3 percent of Asian
children, and 22 percent ofNative American children are in family reunification services (Needell
et al., 2003). Children in family reunification services in California also tend to be young; 43.8
percent of children under the age of one are in family reunification services; 32.9 percent between
the ages of one and two; and 27.6 percent between the ages of three and five are in family
reunification services. In Santa Clara County, 15.7 percent of Black children are in family
reunification services, 24.3 percent of White children, 25 percent of Hispanic children, 20 percent
of Asian children and 13.3 percent of Native American children are in family reunification services.
In Santa Clara County, 33.3 percent of children under the age of one are in family reunification
services, 32.8 percent between the ages of one and two; and 30.7 percent between the ages of three
and five are in family reunification services (Needell et al., 2003).
Family Reunification Services
Family reunification services are generally focused on addressing the broad range of issues
that have contributed to the removal of the child, or that prevent the child from returning home
safely. These services are grounded in the notion that the parent child attachment is of primary
importance and the continuity of these attachments is critical to the well-being and development of
the child (Maluccio & Ainsworth, 2003). Family reunification services are generally offered for a
period lasting 6-18 months and they may include a wide range of practices that vary greatly from
county to county and state to state. As is the case with family preservation practices, a large number
of different interventions fall under the definition of "family reunification services;'1 some family
reunification interventions are modeled after FPS and are intensive, short term and family-centered;
while others focus on the delivery of specific services designed to ameliorate particular problems.
As with research on the effectiveness of family preservation services, research on the
effectiveness ofFRS is lacking; most family reunification studies have small sample sizes, no
comparison groups, no clear definition of the intervention and often fail to measure family and child
well-being in addition to reunification rates (Littell & Schuerman, 1995). Studies evaluating the
impact ofFRS demonstrate mixed results, with many researchers concluding that there is no solid
evidence that FRS are effective in returning children to their birth families (Littell & Schuerman,
1995). Indeed rates of reunification for participants in FRS have been found to range all the way
from 25 percent to 100 percent (Littell & Schuerman, 1995).
For instance, Fein and Staff (1993) evaluated the effectiveness of a fairly intensive family
reunification program in which services were delivered by both a social worker and a family
support worker; workers carried small caseloads and provided a broad range of both concrete and
clinical services, such as parenting training, budgeting assistance, transportation, help with job
training, as well as counseling, or support for substance abuse treatment. Although no comparison
group was used in the study, results indicated that 38 percent of children were reunified with their
families after one year of participation in services-and of these 10 percent were returned to OHC
at the end of the second year.
25
Some researchers have concluded that family reunification services that are intensive in
nature, sometimes following a family preservation model, may be most effective. Walton et al.
(1993), evaluated the effectiveness of family reunification services that were limited to 90 days and
were delivered in the home of the birth parents by workers carrying small caseloads. Interventions
averaged 3.1 hours a week and were focused on concrete services, as well as psychosocial training.
Findings indicated that 93 percent of families receiving FRS reunified after the 90-day treatment
period, compared to 28 percent of a comparison group receiving regular child welfare services. At a
12-month follow-up, the FRS children group (75%) was still more likely to be living in the home,
than the comparison group (49%).
Similarly, Gillespie et al. (1995) also reported promising results from an intensive
reunification program that was modeled after family preservation services. Services were provided
at the birth parents home, for approximately 8 to 10 hours a week and included a combination of
concrete and clinical services. In addition, this program also provided support and training to the
foster parents caring for the children, increased contact between the foster parents and social
workers, as well as the facilitation of joint meetings between foster parents and birth parents to
improve consistency of parenting and to share information. Birth parents were also allowed
increasing numbers of visits based on their participation in services. Although no comparison group
was used, findings indicated that 79 percent of the children were reunified, and at 12-month followup 71 percent of the sample was sill living at home.
Less encouraging results were noted by Lewandowski and Pierce (2002) in their evaluation
of the effects of Missouri's Family Centered Out Of Home Care (FCOHC) Program, a program that
incorporates family-centered principles aimed at involving families in reunification case plans as
soon as the child enters out of home care. Services were intensive, time-limited and strengths-based.
Progress was monitored through team meetings involving the families, social workers, and other
representatives from social service systems. Differences between families receiving the FCOHC
and comparison groups receiving standard child welfare services were significant although not in
the expected direction. The FCOHC group actually experienced fewer reunifications and more
returns to out of home care, than the comparison group. Yet results did indicate that, on average,
children in the FCOHC group who spent more than seven days in out of home care, also spent fewer
days overall in out of home care than the comparison group.
In addition to the impact of family reunification services themselves, the decision-making
practices of workers also play a role in whether or not families reunify. In Westat's (2001a) analysis
of reunification decision-making practices among nine workers in three public child welfare
agencies in Washington DC, workers reported they could generally tell within the first few months
of a reunification case whether reunification would actually occur. Factors identified by workers as
important in their decision to return a child home included, the parents' compliance with services
and visitations, the strong desire of parents to have their children returned, the safety of the child
and the child's wishes and well-being. Workers also indicated that parents who were noncompliant
with services or non-responsive to efforts at reunification and cases involving young children were
most likely to lead to a decision to terminate parental rights. In general workers in this study felt
that time limits for reunification were a positive factor in the decision-making process because time
restrictions allow them to terminate parental rights more quickly in cases where reunification
appears unlikely; however workers also noted that some families with more serious problems could
be reunified if they received services beyond federal time limits. However since the sample size in
this study was small, these results should be interpreted with some caution.
26
Family Reunification Services:
Impact on Children and Families of Color
The impact offamily reunification services for children and families ofcolor is not
adequately addressed in the research literature. Although we know that children and families of
color tend to reunify less frequently than White children, the types ofservices that increase the
likelihood of reunification for children and families ofcolor are not well understood. In general
FRS research does not clearly support the effectiveness of these interventions, either on family
reunification or on family functioning (Littell & Schuerman, 1995; Lewandowski & Pierce, 2002).
Furthermore, the actual interventions that are associated with "family reunification services" are
often not clearly defined.
Research does suggest that families who fail to reunify often have serious social and
economic risk factors that do not appear to be adequately addressed by current family reunification
services. For instance, families with a decreased likelihood of reunification are those that have
income problems (Barth et al., 1987; Courtney, 1994; Jones, 1998); parents with less than a high
school education; unemployed parents (Westat, Inc., 2001c); and children with health problems or
disabilities (Courtney, 1994; McMurty & Lie, 1992; Wells & Guo, 1999). Yet these characteristics
also describe the general child welfare population. Clearly more research is needed to discern the
types of FRS interventions that are effective with these high-risk groups, as well as the possible
differential outcomes for children and families ofcolor.
Specific Family Reunification Services: Parenting Education, Substance Abuse, Domestic
Violence, and Family to Family/Social Network Interventions
In addition to research on the effectiveness of family reunification services as a whole, some
studies have focused on the effects of specific services aimed at improving particular problems such
as parenting deficits, substance abuse, domestic violence and poor social networks. As with family
reunification services, studies on specific treatments to reunify parents with their children are
limited and often have methodological problems such as small sample sizes, lack of comparison
groups, lack of treatment consistency and lack of adequate long-term follow-up.
Parenting Education Services
Programs aimed at teaching parents new parenting skills are frequently included in family
reunification case plans. For instance, Phase 2 case record review data indicated that in Santa Clara
County, 70.2 percent of mothers were recommended parenting education services at the
jurisdictional/dispositional hearing (Hines et al., 2002). In general parenting education services are
didactic in nature and focus on increasing participants' knowledge of effective parenting skills.
Research on the effectiveness of parenting education programs for families in the child welfare
system is lacking. Most parenting education evaluations have examined the impact of these
programs on parents whose children have specific behavioral problems-not for families who have
had children removed as a result of maltreatment. Research on parenting education for nonmaltreating parents has yielded some promising results (Dore & Lee, 1999). However, only a
handful of studies have evaluated parenting education services for child welfare parents. Results
from these studies suggest that child welfare parents may experience increased benefits from
parenting education programs that combine a didactic approach with another parenting education
27
service (Brunk, Henggeler, & Whelan, 1987; Carlo, 1993; Dore & Lee, 1999; Wolfe, Edwards,
Manion, & Koverola, 1988).
More rigorous studies of traditional didactic parenting education services have been
performed with samples involving parents whose children have behavioral problems. Most of these
studies suggest that low risk parents may experience significant short term benefits from parenting
education services (Petsch, Schultz, & Wahler, 1999; Nicholson, Brenner, & Fox, 1999; Patterson,
Chamberlain, & Reid, 1982). However, the research also suggests that high-risk parents typically
experience little benefit from parenting education services (Dore & Lee, 1999).
Parenting Education Services:
Impact on Children and Families of Color
Research on the impact ofparenting education services for children and families of color is
lacking. Studies fail to mention the possible impact ofcultural differences in parenting behaviors on
the success or failure ofparenting education services (Forehand & Kotchick, 1996). Parenting
education appears to be most effective for lower-risk families, however, numerous stressors such as
poverty, substance abuse, domestic violence, low education levels, and community violence often
impact parents involved in the child welfare system. Dore and Lee (1999) note these risk factors
may make cognitive behavioral and didactic approaches to parenting education particularly unsuited
to these populations:
For those who are highly stressed by poverty and its accompanying pressures, even
sustaining a commitment to the training process can be overwhelming. Those parents who
lack the cognitive foundation needed for more didactic and learning based approaches seem
unable to benefit fully from behaviorally based training models (Dore & Lee, 1999, p. 322).
More promising approaches to parenting education for maltreating parents appear to involve a
multifaceted approach that includes not only instructional training in parenting behaviors, but also
experiential training and interventions that address the emotional and psychosocial stressors
affecting these high-risk parents (Carlo, 1993; Brunk et al., 1987; Dore & Lee, 1999; Wolfe et al.,
1988).
Indeed, there is some limited evidence to suggest that parents in the child welfare system
may experience increased benefits from a parenting education program than combines didactic
parenting education with another parenting related service such as behavior training, experiential
training or multistystemic therapy (Brunk et al., 1987; Carlo, 1993; Wolfe et al., 1988). However,
more research is needed in this area. For instance, it is unclear which parenting related services are
most effective when paired with didactic training and it is also unclear how well parents retain
information learned. Long-term follow-up studies that track the retention of information in
parenting education services for child welfare parents, as well as child and family outcomes, are
needed.
Additionally, few studies have addressed the role of culture and ethnicity in the effectiveness
ofparenting education services/or parents in the child welfare system. Forehand and Kotchick
(1996) note, " ... different cultures have different histories that shape cultural theories about
parenting. Without an awareness of the values ofa particular culture, attempts to implement
parenting programs with its members may fail" (Forehand & Kotchick, 1996, p.197). As such,
28
research needs to focus on how parenting education outcomes may differ for different cultural and
ethnic groups and what types ofparenting education approaches might be most successful/or
diverse families.
Substance Abuse Treatment
Parental substance use is common among the child welfare population. In 1998, the United
States General Accounting Office [U.S. GAO] conducted a study of two state CWS systems in
California and Illinois, and reported that 65 percent of open foster care cases in California and 74
percent in Illinois, involved a family where one or both parents had a substance abuse problem for
which treatment was required. Research indicates that cases involving parental substance abuse
reunify at a significantly lower rate than do cases not involving substance abuse (Smith, 2003), and
in general, children from substance abusing households experience poorer outcomes than children
from non-substance abusing households (U.S. GAO, 1997).
As a result of the widespread su~stance use among child w~lfare parents and the poor
outcomes for cases involving parental substance abuse, many family reunification case plans
include requirements for some form of substance abuse treatment. These interventions can include
various components such as drug testing, 12-step groups, outpatient treatment or inpatient
treatment. For instance, Phase 2 case record review results indicated that 40.9 percent of mothers in
Santa Clara County are ordered random drug or alcohol testing, 40.0 percent are ordered substance
abuse or alcohol treatment and 26.8 percent are ordered 12-step groups (Hines et al., 2002).
Unfortunately few studies have evaluated the impact of substance abuse services for child welfare
parents (Gregoire & Schultz, 2001). More research efforts have focused on the effectiveness of
these services for pregnant and parenting women who may not necessarily be involved in the CWS.
In one of the few studies that has evaluated substance abuse treatment among child welfare
parents, Gregoire and Schultz (2001) analyzed the treatment and placement outcomes for 167 child
welfare parents in Lancaster County Pennsylvania. Findings revealed that those participants who
had support from significant others were more likely to complete substance abuse treatment and
also experienced better sobriety outcomes. Moreover, parents who were sober at nine months
follow-up were more likely than parents who were not sober to have child custody or to have
maintained parental rights.
Ways in which child welfare workers make decisions about cases involving substanceaffected parents also impacts the likelihood of reunification. Karoll and Poertner (2002) identified
certain factors considered to be associated with readiness for reunification in their qualitative study
of judge's, caseworker's and substance abuse counselor's indicators ofreunification. Factors
associated with professional' s indication of readiness for reunification included: 1) parental
motivation, 2) evidence of substance abuse recovery, 3) evidence of competence and reliability, 4)
the presence of social support, 5) evidence of improved parenting skills and 6) legal aspects of child
loss and prospects for reunification. However there was considerable variability among respondents
in how many clean drug tests are necessary for reunification. Although the median response was 8
months of clean tests, approximately 30 percent felt 6 months was sufficient, while 32 percent felt
that 12 months of clean tests were needed before reunification should occur.
Although research on the effectiveness of substance abuse treatment for child welfare
parents is scant, more research efforts have focused on the effectiveness of these services for
29
pregnant and parenting women who may not necessarily be involved in the CWS. However, since
most child welfare agencies respond to referrals for prenatal drug exposure efforts to offset drug use
among pregnant and parenting women is an important concern for the CWS (Ondersma, Malcoe, &
Simpson, 2001).
Overall, research on the effectiveness of substance abuse services for pregnant and parenting
women suggests these services may be beneficial, however the degree of benefit appears to depend
on the type and intensity of the intervention, as well as the case characteristics. For instance, there is
some evidence that suggests that residential treatment may be more beneficial than outpatient
treatment for women with substance abuse problems. (Haller et al., 1997). Moreover, parenting
women experience better outcomes in substance abuse treatment when they are able to reside in
residential treatment with their children (Clark, 2001; Metsch et al., 2001 ). Studies also indicate that
women are more likely to complete treatment and experience better outcomes when they have the
support of significant others (Gregoire & Schultz, 2001 ), and when family and significant others are
involved in treatment (Carten, 1996). Parenting women have also identified support and
encouragement from non-judgmental staff as particularly important in their ability to successfully
complete treatment (Akin & Gregoire, 1997; Carten, 1996).
Substance Abuse Treatment:
Impact on Children and Families of Color
Research on the effectiveness ofsubstance abuse treatment for parents color in the child
welfare system is lacking. In general, studies do not describe differential outcomes for various
racial/ethnic groups in these treatments and few studies have actually examined the impact of
substance abuse treatment on reunification outcomes for children and families ofcolor in the child
welfare system. However, research does suggest that parenting women involved in substance abuse
treatment are a high-risk group. Most participants of these services are young single women of
color, with low educational levels and income problems (Carten 1996; Clark, 2001; Knight et al.,
2001; Metsch et al., 2001 ). In general, research on the effectiveness of substance abuse treatment
services for these women has yielded promising results; however the degree of treatment impact
appears to depend on the type and intensity of the intervention, as_ well as case characteristics.
Studies suggest that participants in substance abuse treatment services are more likely to maintain
child custody (Gregoire & Schultz, 2001); and maintain sobriety (Clark, 2001; Metsch et al., 2001).
However, women with less than a high school education, criminal justice system involvement and
poor social supports are less likely to complete treatment (Knight et al., 2001 ).
Overall, research suggests that high-risk parenting women benefit from substance abuse
treatment that is comprehensive; allows them to involve their children and other family members or
supports; and is delivered by supportive, nonjudgmental staff. Although research has not
specifically evaluated the role ofethnic diversity in substance abuse treatment services for
parenting women in the child welfare system, the fact that many of the participants of these services
are women ofcolor does suggest that these services may be beneficial for a diverse population.
However, more research is needed to determine the role of culture in treatment, as well as possible
differential outcomes for various ethnic/racial groups.
Domestic Violence Treatment
Child welfare professionals are becoming increasingly aware of the importance of
30
addressing issues of domestic violence for families attempting to reunify with their children. Rates
of domestic violence among families in the CWS vary from 38 percent to 54 percent (Bowen, 2000;
McGuigan & Pratt 2001). Phase 2 results from case record review data indicated that in Santa Clara
County, 45.4 percent of cases involve domestic violence (Hines et al., 2002). Research from the
child welfare field on the effectiveness of domestic violence treatments in reunifying families and
improving outcomes for families is extremely lacking. Instead, child welfare research efforts related
to domestic violence have primarily focused on describing barriers to effective collaboration
between domestic violence and child welfare professionals and attempts to overcome these barriers.
Indeed, practioners and researchers have noted that child maltreatment and spousal abuse
service systems have historically operated in isolation from one another and typically have different
service delivery approaches (Beeman, Hagemeister, & Edleson, 1999; McKay, 1994). Recent
efforts to increase collaboration between the child welfare and domestic violence fields has focused
on cross-training between the two professions, as well as simplified processes for referring families
to child welfare or domestic violence services (Findlater & Kelly, 1999; Whitney & Davis, 1999).
Additionally, the creation of specific child welfare policies for families experiencing domestic
violence and the use of multidisciplinary task forces to address ongoing issues related to
collaboration between fields has also been described as helpful in increasing coordination and
collaboration (Findlater & Kelly, 1999; Whitney & Davis, 1999).
Santa Clara County has implemented efforts to merge child welfare and domestic violence
fields in order to better serve families affected by both child maltreatment and domestic violence.
Santa Clara County is implementing the Greenbook Project, a multidisciplinary effort involving
representatives from child welfare, the domestic violence community, law enforcement and the
juvenile court. The overall goal of the project is to change the way these groups assist families who
experience both domestic violence and child maltreatment. During the first year of implementation,
the Greenbook Project placed domestic violence advocates from Next Door Solutions to Domestic
Violence at the County Department of Social Services. These advocates provide consultation to
child welfare workers on issues of domestic violence at every level of the child welfare system.
Additionally, the Greenbook Project also visited the Miami-Dade County Dependency Court
Intervention Program for Family Violence to learn key aspects of a domestic violence advocacy
program. Other activities of the Greenbook Project include cross-training between child welfare,
domestic violence, law enforcement and legal personnel, improving cultural competency,
developing guiding principles, changing agency policy and practice, and integrated parenting
programs (Lightbourne, 2002).
The collaboration of child welfare and domestic violence fields to better meet the needs of
families affected by both issues is promising. However, in general these efforts are not focused on
specific treatments for the batterer. Indeed, child welfare research has not addressed the
effectiveness of domestic violence services for child welfare parents. Yet studies from related fields
have evaluated effectiveness of domestic violence treatment for batterers who may not necessarily
be involved in the CWS. Domestic violence treatment for batters typically involves a group
treatment that includes psycho-educational and cognitive behavioral techniques designed to change
faulty beliefs, reduce anger, improve communication and increase non-violent coping skills
(Morrel, Elliot, Murphy, & Taft, 2003). Less common are psychodynamic or support groups.
Findings from studies on the effectiveness of these treatments are mixed and much of the research is
limited due to small sample sizes, lack of comparison groups and findings that rely on self-reports.
Some studies have reported that targeted domestic violence services are effective in reducing
31
violent beh~viors (Dutton, 1986; Palmer, Brown, & Barrera, 1992), yet other studies report no
benefits from these treatments (Dunford, 2000). As with many other treatments, domestic violence
treatment tends to be most effective for participants with relatively few other risk factors (Morrel et
al., 2003; Taft, Murphy, Elliot, & Keaser, 2001).
Domestic Violence Treatment: Impact on Children and Families of Color
Research on the impact of cross-system collaboration and specific domestic violence
treatment for children and families ofcolor is lacking. Although child welfare research describing
cross-system collaboration between domestic violence professionals and child welfare professionals
is promising, these studies do not address the role ofcultural issues on domestic violence, or the
treatment of batterers and battered women. Similarly, research from related fields describing
outcomes of didactic domestic violence treatment also rarely consider the role of culture or
ethnicity.
Research suggests that the domestic violence field and the child welfare field have
historically had different approaches to treatment that have hindered collaborative efforts (Beeman
et al., 1999). Recently, more concerted efforts to link the two fields have occurred. Typically these
efforts involve the use of multidisciplinary task forces to address issues related to collaboration
(Findlater & Kelly, 1999; Lighbourne, 2002); cross-training (Findlater & Kelly, 1999; Lighbourne,
2002; Whitney & Davis, 1999); simplified processes to refer battered women to child maltreatment
prevention programs and to refer battered child welfare mothers to domestic violence services
(Findlater & Kelly, 1999; Whitney & Davis, 1999); and, the creation ofspecific child welfare
policies stating procedures for families experiencing both child maltreatment and domestic violence
(Findlater & Kelly, 1999; Whitney & Davis, 1999). Although these efforts have great potential to
impact children and families of color in the child welfare system, research has not yet addressed this
topic.
The collaborative efforts between child welfare and domestic violence tend to focus on
services and assistance to women and children, with less of an emphasis on treatments for the
perpetrators ofdomestic violence. As with other didactic interventions, such as parenting education,
there is evidence to suggest that domestic violence treatment for batterers is more effective with
low-risk groups than with high-risk groups (Morrel et al., 2003). However, research also suggests
that most men who are participating in domestic violence treatment are a high risk group who often
experience multiple stressors (Brown, Werk, Caplan, & Seraganian, 1999). Moreover, the presence
of certain risk factors appears to reduce the likelihood of treatment completion (Taft et al., 2001).
Additionally, one study found that African American men were most likely to drop out of domestic
violence treatment, even after other factors were statistically controlled, suggesting that these
treatments may not be as effective for African Americans (Taft et al., 2001 ). However, in general,
more research is needed to discern the impact of domestic violence treatments for various
racial/ethnic groups in the child welfare system.
Social Network Interventions/Family to Family Initiative
Some child welfare services have focused on the role of social network interventions on
improving outcomes for children and families. Social network interventions are designed to enhance
both formal and informal social networks, including increased networking among family members,
32
friends, neighbors, as well as more structured networking involving support groups (Gaudin,
Wodarski, Arkinson, & Avery, 1990). These interventions are based on research indicating that
people living in neighborhoods in which there are high child maltreatment rates tend to describe
these communities as lacking in formal and informal support systems (Garbarino & Kostelny,
1992). Additionally, DePanfilis ( 1996) has noted that in addition to a lack of social support systems,
neglectful families may be particularly distrustful of any support systems that are available.
In an effort to address these issues, social network interventions, such as the Family to
Family Initiative have been implemented. Although the goals of Family to Family are extensive and
go beyond just social network interventions, one of the initiative's main emphases is on
"strengthening the network of families available to care for abused and neglected children in their
own communities" and "building partnerships with at-risk neighborhoods toward that end" (Annie
E. Casey Foundation, 2003b, p. 1). Indeed, many interventions provided through Family to Family
are designed to strengthen neighborhood networks for birth and foster families in the child welfare
system. These strengthened networks are considered to have the potential to assist families in
learning new skills and becoming more self-sufficient. The tactics used to increase social networks
for families, and also increase collaboration between the child welfare system and local
communities are varied. In general there is an emphasis on identifying families' naturally occurring
support network~ and involving these networks in services. Interventions can be focused on
identifying and utilizing "natural helpers" in the communities and neighborhoods where birth and
foster families reside. Natural helpers are considered to understand the culture of a particular
community better than professional helpers. Efforts are also aimed at increasing overall
collaboration between child welfare agencies and the communities in which child welfare clients
reside.
Outcome studies on the effectiveness of Family to Family interventions are not available.
However, Gaudin et al. (1990) evaluated the effectiveness of the Social Network Intervention
Project (SNIP) for 52 neglectful families in the child welfare system. The intervention, which lasted
approximately 10 months, included first an assessment of existing formal and informal support
networks in the lives of the families, and then a series of interventions designed to enhance these
networks. Direct interventions were made with informal supports, including family members,
friends or neighbors. Efforts were also made to identify and utilize informal helpers existing in the
families' neighborhoods. Formal support networks were also encouraged through the use of support
groups, parent aides and social skills training. Post-test results from the evaluation indicated that
compared to a comparison group who received normal child welfare interventions, the families
receiving the social network services fared better on a number of outcomes assessed by their
caseworkers. Intervention parents were rated as having improved parenting skills and attitudes, and
they experienced significant increases in the size of their formal and informal support networks. At
the end of the intervention, 59 percent of the families had their child welfare cases closed, while
23 .5 percent of the comparison group has their cases closed.
Although the use of social network interventions may be a promising child welfare practice,
Gaudin et al. (1990) notes that there are organizational and professional barriers to the
implementation of such interventions. For instance, results from the SNIP evaluation indicated that
it was difficult for social workers implementing the project to maintain a focus on social network
interventions within the context of a public child welfare agency. The predominate model of
services in the agency tended to focus on case management services, and barriers such as high
33
caseloads, heavy paperwork, frequent staff turnover, and lack of collaboration between human
service agencies all contributed to difficulties in implementing the social network intervention.
Family to Family/Social Network Interventions:
Impact on Children and Families of Color
Researchers have not evaluated the impact ofsocial network interventions, such as the
Family to Family Initiative on children and families ofcolor. Given that children andfamilies of
color involved in the child welfare system are likely to reside in impoverished communities, the use
ofsocial network interventions with this population is ofparticular interest. Research exploring the
assets and informal supports imbedded in impoverished communities is limited. It certainly appears
likely that strengthening support networks for vulnerable families would be beneficial, however
research on the mechanisms through which to increase support networks for families is lacking.
Indeed, the exact interventions associated with utilizing informal helpers, strengthening
neighborhood networks and increasing overall collaboration between child welfare agencies and
the communities in which child welfare clients reside are not clearly defined. As Gaudin et al.
(1990) has suggested, there may be significant barriers to effective implementation of these
interventions within the crisis-driven, resource strapped child welfare field. More research is
needed to first discern the types of interventions associated with social network practices, as well as
the ways in which outcomes from these services may differ for various racial/ethnic groups.
Mental Health Services for Children in Out-of-Home Care
Children in the child welfare system are at an increased risk for a variety of psychological
and behavioral problems (Stevenson, 1999). Studies have documented the long-term negative
consequences of both child maltreatment and lengthy stays in out-of-home care (Courtney, Piliavin,
Grogan-Kaylor, & Nesmith, 200 I; Stevenson, 1999). In order to ameliorate these negative
consequences, many children in out-of-home care are referred to mental health services. Research
suggests that mental health services are the most commonly recommended treatment for children in
the child welfare system (Kinard, 2002). Studies have also found that children in foster care are
more likely to have a mental health diagnosis and to use mental health services than comparison
groups. For instance, children in foster care are more likely than children receiving public assistance
or children receiving SSI to have a mental health diagnosis (Harman, 200 I; dosReis, 2001 );
children in foster care tend to have more mental health hospitalizations; and use more mental health
services overall than children receiving public assistance (Reiff, 2001 ).
Yet despite the relatively high rate of mental health service usage among children in the
CWS, the nature and impact of these services are not well understood. Indeed mental health services
can include a broad range of interventions, including support groups, individual counseling, family
counseling, residential services or inpatient services. One model of services, Treatment Foster Care
(TFC), has been the topic of child welfare research. TFC, also referred to as therapeutic foster care
is an intensive, family centered intervention that seeks to place youth with behavioral and emotional
disorders in community and family settings. Children are typically placed with foster parents and
the placement is intended to be both therapeutic and family-centered. Intensive services are
delivered to parents, alternate caregivers and the child. Some studies have suggested that TFC can
be an effective transitional placement for children moving from restrictive settings into less
restrictive placements (Chamberlain, 1994, as cited in Farmer, Wagner, Burns, & Richards 2003;
Clarke, Hawkins, Murphy, & Sheeber, 1993, as cited in Farmer et al., 2003). However, not all
34
studies have found TFC to be effective. For instance, Fanner et al. (2003) used administrative data
to analyze the impact ofTFC on foster youth with emotional or behavioral disorders. Findings
indicated that use ofTFC did not result in reduced use of group homes.
In addition to TFC, some child welfare interventions for children focus on a "wraparound"
process that is targeted to high risk children who may have emotional or behavioral problems and
who are currently placed in a level 12 or higher group home, or are at risk for entry into a more
restrictive placement (California Department of Social Services [COSS], 2001). Wraparound
services are described as a "family-centered, strength-based, needs-driven planning process for
creating individualized services and supports for ·children and their families" (COSS, 2001, p. I).
Wraparound interventions are collaborative in nature, often involving multiple social service
systems in the planning and delivery of a wide range of services. Few studies have systematically
investigated the efficacy of wraparound services for children and families in the child welfare
system. However, some research suggests that children in wraparound are less likely to be removed
from their community and may experience a reduction in behavioral problems (Bruns, Burchard, &
Yoe, 1995, as cited in Toffalo, 2000; Clark, Lee, Prange, & McDonald, I 996, as cited in Toffalo,
2000; VanDenBerg, 1993, as cited in Toffalo, 2000). Yet other research has reported no significant
benefit from wraparound services (Clarke, Schafer, Burchard, & Welkowitz, 1992, as cited in
Toffalo, 2000).
Mental Health Services:
Impact on Children of Color in Out-of-Home Care
Although research is lacking in identifying effective mental health practices for children in
foster care, studies have demonstrated that children ofcolor are less likely to be recommended for
these services and to use these services. In Santa Clara County, Phase 2 case record review results
indicated that 36.2 percent of children were recommended individual counseling and 19 .6 percent
were recommended other counseling services. Differences between racial/ethnic groups in
recommended counseling services were also found. For instance, 18 percent of Black children were
recommended individual counseling, 46.3 percent of White children, 31.9 percent of Latino
children, and 35.5 percent of Asian American/Pacific Islander children were recommended
individual counseling, suggesting that children of color, and in particular African American children
have a decreased likelihood of being recommended individual counseling (Hines et al., 2002).
Differences in service recommendation and provision were also noted by Garland and
Besinger (1997) in their analysis of racial/ethnic differences in court referred pathways to mental
health services for children in foster care. Results suggested a trend for White children to be more
likely than African American or Hispanic children to be court ordered counseling/psychotherapy.
White children were also significantly more likely than African American or Hispanic children to
receive counseling during their first eight months in out-of-home care. And although not significant,
White children had higher rates of service use in general than did African American or Hispanic
children. Racial/ethnic differences in mental health service recommendations and service usage
have been found to remain even after the effects of other demographic factors are controlled
(Garland, Hough, Landsverk, McCabe, Yeh, Ganger et al., 2000).
Research on the effectiveness of mental health practices for children in out-of-home care is
lacking, and even less is known about the impact of these services for children of color.
Comprehensive treatments for high-risk children, such as treatment foster care and wraparound
35
services have not been adequately researched and have in general not consistently demonstrated
positive effects. Studies on other mental health interventions for children ofcolor in the child
welfare system are lacking. However, research does suggest that children of color are less likely
than. White children to be referred to mental health services (Garland & Besinger, 1997; Hines et al.,
2002). The reduced likelihood for children ofcolor to be referred to mental health services is of
particular concern because children of color experience long stays in the child welfare system and
are at an increased risk/or a variety ofpoor outcomes (Courtney et al., 1996). As such, more
efforts should be aimed at ensuring that children of color are referred to and receive appropriate
mental health support, and more research is needed to evaluate the impact ofthese services on
outcomes for children ofcolor.
Foster Parent and Kinship Support Services
Although the majority of research on child welfare practices has focused on interventions
with birth families and children, some research has emphasized the im.portance of services and
training for alternative caregivers, including foster parents and kin caregivers (Cuddeback & Orme,
2002; Gordon, McKinley, Satterfield, & Curtis, 2003). Some studies have suggested that kin
caregivers may receive fewer services than non- kin caregivers (Gordon et al., 2003; Scannapieco,
Hegar, & McAlpine, 1997, as cited in Cuddeback, 2002). However, other research has reported no
differences between services delivered to kin and non-kin caregivers (Cuddeback & Orme, 2002).
Research suggests that the use of kin placements is common in California and in Santa Clara
County. In 2002, 36.4 percent of children in supervised foster care in California were in a kin
placement and in Santa Clara, 40.l percent were placed with kin (Needell et al., 2003).
Additionally, children and families of color are especially likely to be placed with kin. For instance,
point in time reports indicate that in California, 40.5 percent of Black children in supervised foster
care are placed with kin; 37.0 percent ofNative American; 35.8 percent ofHispanic, 29.5 percent of
White and 29.4 percent of Asian children are placed with kin. In Santa Clara County, 66.7 percent
of Native American children are placed with kin; 41.9 percent of Hispanic; 41.4 percent of Black;
35.6 percent of Asian; and 32.5 percent of White children (Needell et al., 2003).
The prevalence of kin placements has prompted some researchers to explore the service and
support needs of kin caregivers. Gordon et al. (2003) noted the unique needs of kin caregivers in
their analysis of focus groups with 37 kin caregivers. Findings revealed that unlike non-kin
caregivers, many kin caregivers did not expect to be caring for the children in their care; they had to
adjust their family roles and many described this adjustment as sudden and significant. Kin
caregivers also described having to adjust to the involvement of the child welfare agency in their
lives; many reported feeling unappreciated by the child welfare agency and excluded from the
decision-making process for their kin children. They also described a great need for additional
support services as they took on the unexpected responsibility of caring for their kin children.
Other studies have noted that non-kin caregivers also experience unique challenges. For
instance, although non-kin caregivers are expected to support reunification efforts, many may have
conflictual feelings about supporting contact between the children in their care and the birth
families (Sanchirico & Jalonka, 2000). As a result, some child welfare practices have focused on
providing specialized training and support to non-kin caregivers to assist in the process of
reunification. For instance, Sanchirico and Jalonka (2000) surveyed 650 non-kin caregivers in New
York State and noted that only 19 percent received both specialized training to be a foster parent
36
and support and guidance from the agency. Those who received both training and support also
performed the most activities related to encouraging contact between children and their birth
families. Foster parents who received either training or support but not both also exhibited some
tendency to encourage parent child contact, however those who received both training and support
were significantly more likely to support reunification efforts. These results suggest that specialized
services delivered to foster parents can have a beneficial impact on reunification.
Additionally, other research has focused on evaluating more concerted efforts to involve
foster families in the reunification process. For instance, Landy and Munro (1998) evaluated the
impact of the Shared Parenting program. The Shared Parenting program encourages foster families
to work as a team with birth parents in order to improve the parenting skills of the parents. Foster
families were encouraged to provide support and guidance to birth parents through on-going contact
and involvement. An improved relationship between foster and birth families was expected to
increase birth parents motivation and interest in learning new skills, thereby increasing
reunification. Thirteen families participated in the evaluation. Results revealed difficulties in
retaining parents in the program. Over half of the participants did not complete the program, either
because of lack of compliance or because workers discharged families due to changes in status that
decreased the possibility of reunification. Outcomes for parents who remained in the program
revealed that those with higher socioeconomic status, families who moved less, families with no
health problems and families with less risk factors were more likely to reunify.
Although research suggests that kin and non-kin caregivers have different service needs,
studies have demonstrated that overall, both kin and non-kin caregivers have numerous unmet
service needs. For instance, Cuddeback and Orme's (2002) conducted a national evaluation of the
service experiences and needs of kin and non-kin caregivers, and found that both groups reported
numerous unmet service needs. Most alternate caregivers expressed a need for day care, respite
care, liability insurance, heath care, and transportation. Additionally, the majority of caregivers in
both groups reported feeling unprepared to care for their foster children. The authors concluded "the
levels of training and services received were inadequate for both groups of caregivers" (Cuddeback
& Orme, 2002, p. 900).
Foster and Kinship Support Services:
Impact on Children and Families of Color
The impact offoster and kinship support services ofchildren and families ofcolor has not been
adequately researched. However, research clearly indicates that kin and non-kin caregivers are in
need of increased training and supportive services. Kin caregivers may need increased support as
they take on the unexpected responsibility and role of a caregiver (Gordon et al., 2003), while nonkin caregivers may need additional support and training to facilitate the reunification process.
(Sanchirico & Jalonka, 2000). In general, both kin and non-kin caregivers describe the need for
increased services, such as day care, respite care, liability insurance, heath care, transportation,
specialized training and increased participation in the child welfare decision making process
(Cuddeback & Orme, 2002; Gordon et al., 2003; Sanchirico & Jalonka, 2000).
Research on the impact offoster and kinship support services on child welfare outcomes is
lacking. However, there is some evidence to suggest that interventions with foster parents are
associated with better reunification outcomes (Sanchirico & Jalonka, 2000). Yet more research is
clearly needed to discern the impact of services to both kin and non-kin caregiver, especially for
37
different racial/ethnic groups.
Permanency Planning Practices
The federal Adoption and Safe Families Act of 1997 shortened reunification time frames
and mandated the use of concurrent planning-a process in which both a reunification plan and an
alternative permanent placement plan are created. For children who are not reunified, the permanent
placement plan with its goal of finding a suitable alternative permanent placement for the child
becomes a major focus of services. As such, permanency planning and placement practices often
impact children at the ending stages of the CWS.
Very few studies have investigated the types of practices that are used at the ending stages in
the system and how these practices impact children of color. However, the impact of child welfare
practices on children who are not reunified is particularly important for children and families of
color. The U.S. Department of Health and Human Services (2002) reports that approximately 23
percent of children in foster care have had parental rights terminated and children of color are
disproportionately represented in this group. Nationally, 43 percent of legally free children are
African American, 34 percent are White, 13 percent are Hispanic, 2 percent are Native American
and 1 percent are Asian Pacific Islander (U.S. DHHS, 2002). In California, 68.2 percent of Black
children in foster care are in permanent placement services, 48.4 percent of White children; 41.5
percent of Hispanic children; 32.5 percent of Asian children, and 53.9 percent ofNative American
children are in permanent placement services (Needell et al., 2003). Santa Clara County has similar
racial/ethnic rates of permanent placement clients; 64.4 percent of Black children are in permanent
placement services; 44.7 percent of White children; 41.3 percent of Hispanic children; 30.6 percent
of Asian children and 43.3 percent ofNative American children are in permanent placement
services (Needell et al., 2003).
Federal policy mandates that these children have safe and permanent alternative homes,
however, in reality many wait months and years for a permanent placement and some will reside in
long term foster care until they emancipate from the system at age 18. In Kemp and Bodonyi's
(2002) analysis of legally free children in the state of Washington, the median length of stay in outof-home care for these children was 50. 7 months and demographic factors were significantly related
to permanency outcomes. For instance, older children, boys, African Americans, and Asian Pacific
Islanders were all less likely to achieve a permanent placement than others groups. Interestingly,
Latino children were significantly more like than White children to achieve a permanent placement.
The interaction of demographic characteristics was also significant in predicting permanency. 0 Ider
African American boys had an even higher chance of not achieving a permanent placement. Other
research corroborates these findings. For instance, African American children have been found to be
less likely to be adopted than White or Latino children and African Americans who are freed for
adoption wait significantly longer periods of time for a permanent placement than do children from
other racial/ethnic groups (Barth, Courtney & Berry, 1994, as cited in Kemp & Bodonyi, 2002;
Kapp, McDonald, & Diamond, 2001).
The federal Multi Ethnic Placement Act (MEPA) of 1994 and the Interethnic Adoption
Provisions (IEP) of 1996 were enacted to increase the number of children who were adopted by
preventing discrimination in the placement of children on the basis of ethnicity and allowing for
increased use oftransracial adoption. The impetus behind MEPA-IEP was a belief that children of
color, especially African Americans were languishing in care, while workers searched for same race
38
foster or adoptive homes. MEPA-IEP also called on child welfare agencies to increase efforts to
recruit foster and adoptive parents who reflect the ethnic and cultural backgrounds of their children
in care. Although MEPA-IEP may have the potential to reduce the time legally free children spend
in substitute care, rates of adoption for children of color in the CWS do not appear to have improved
since its enactment. Indeed, Courtney (1997) suggests that transracial adoption policies and
practices may not actually have a large impact on children of color. Because many legally free
children of color in out-of-home care are older and may have disabilities or health problems, many
potential adoptive parents may choose not to adopt these children. Thus, even though child welfare
agencies may seek transracial adoptions, the actual pool of available adoptive parents may be too
limited to have an impact on children of color waiting for a permanent placement (Courtney, 1997).
In an effort to understand the factors that contribute to adoptive parents willingness to adopt
Black children, Brooks and James (2003) surveyed 541 participants of the California Long-Range
Adoption Study, a longitudinal study of adopters in California. Results indicated significant
differences between participants who were willing to adopt Black children and those who were not. .
Of the sample 36 percent were willing to adopt a Black child. In general, parents who were willing
to adopt Black children were also more willing to adopt children of varying ages and special needs
children; they were more likely to be motivated to adopt due to religious or humanitarian reasons
and many eventually did adopt Black children from foster care. Interestingly, approximately 50
percent of parents who indicated they were unwilling to adopt a Black child also reported that
adoption workers or agencies discouraged them from adopting Black children, while only 11
percent of those parents who were willing to adopt Black children were encouraged by their worker
or agency to do so.
Additionally, parents who adopt transracially may experience unique challenges and have
special service needs. In an effort to identify the special needs of transracial adoptive parents Vidal
de Haymes and Simon (2003) conducted interviews with 20 youth and their adoptive parents. Youth
in the sample were either African American (70%), Latino (15%) or biracial (15%), while 95
percent of adoptive parents were White and 5 percent were Latino. Findings revealed that parents
often felt that others misjudged their intentions for adopting a child of color and that their families
"deserved pity" (Vidal de Haynes & Simon, 2003, p. 257). Many families reported that child
welfare workers, especially workers of color did not adequately support their decision to adopt a
child of color. Parents felt that workers viewed their desire to adopt transracially as curious, and
many felt that they received only minimal assistance in their efforts to adopt a child of color.
Parents identified numerous needed pre and post adoptive supports and resources. The most
frequently noted needed services were formal linkages to cultural activities, products, or services
that could educate parents on cultural issues and also provide a link for the child to their own
cultural group. Other needed support services included establishing a network of parents who
adopted transracially, educating workers to support transracial adoption, providing parenting classes
to address issues of cross cultural parenting, and developing a mechanism to help White parents
address their own possible racism, as well as racism from their family and friends.
In an effort to understand how worker perceptions may impact the likelihood of transracial
adoption for children of color, Carter-Black (2002) conducted an ethnographic study often Black
child welfare workers. Although the sample size makes generalizing the results problematic,
findings revealed that workers felt Black children need to develop an ethnic identity and an
understanding of their cultural heritage and that the CWS in general does not meet the needs of
Black clients and poor clients. The recruitment, support and maintenance of Black adoptive and
39
foster parents by the agency were seen as lacking. Many also felt that MEPA-IEP could actually be
harmful to Black children because it may result in agencies being less culturally aware or sensitive
to diverse populations.
A very (2000) also explored the impact of worker perceptions on permanent placement
practices in her study on the "hardest to place children" in New York State. Eighty case files were
examined and children's workers completed a questionnaire on their perceptions and practices
related to these children. Demographics of the sample are consistent with other studies on children
who have lengthy stays in out-of-home care. Seventy three percent of the sample were male, 60
percent were Black, 30 percent White and 10 percent of another ethnicity, many had substantial
disabilities; approximately 80 percent had siblings in the CWS and they spent approximately 11. 79
years in substitute care. Responses to survey questions revealed that 41.4 percent of workers
believed that children in the sample were ultimately not adoptable. These beliefs appeared to result
in reduced efforts on behalf of the workers to recruit potential adoptive parents. Worker attitudes
toward appropriate placements revealed that 44 percent felt placement with gay or lesbian parents
was not appropriate, and 42 percent felt transracial adoption was not appropriate. When asked to
identify resources that would be helpful in placing the children, workers most often identified
respite services, intensive adoptive parent training, intensive post-placement services, more
effective recruitment techniques and higher adoption subsidies. These findings suggest that workers
negative perceptions of hard to place children appear to result in reduced efforts to recruit adoptive
parents.
Permanency Planning Practices:
Im pact on Children and Families of Color
In general, child welfare research has focused relatively little attention on practices at the
later stages ofthe child welfare system. Although child welfare services have the goal of ensuring a
permanent placement for children, research indicates that many children who have had parental
rights terminated may wait years in out-of-home care before a permanent placement is found (Kemp
& Bodonyi, 2002), and many children remain in out-of-home care until they emancipate at age 18
(Courtney et al., 2001 ). Yet the impact of permanency planning practices on outcomes for children
of color is of great importance. Research indicates that African American children are
overrepresented among children who have had parental rights terminated and are waiting/or a
permanent placement. African American and Asian Pacific Islander children both have a reduced
likelihood ofachieving a permanent placement, and older African American boys have an even
lower likelihood ofa permanent placement (Kemp & Bodonyi 2002).
Some researchers have noted that MEPA-IEP and the issue oftransracial adoption may
actually have little impact on children of color in permanent placement services. Because the pool
of adoptive parents willing to adopt legally free children of color (who are often older, and may
have disabilities or health problems) may be relatively small, policies aimed at transracial adoption
may not have their intended impact (Courtney, 1997). Yet other research suggests that the
perceptions and practices of workers may have an important influence on the likelihood of
achieving a permanent placement. Indeed, negative perceptions of hard to place children may result
in reduced efforts to recruit adoptive parents (Avery, 2000). Additionally, parents who have adopted
transracially report that workers often did not adequately support their decision to adopt
transracially (Vidal de Haymes & Simon, 2003), and Black child welfare workers have expressed
some hesitancy regarding MEPA-IEP, believing that it could be harmful to Black children because
40
it may result in agencies being less culturally aware or sensitive to diverse populations (CarterBlack, 2002). However, in general more research is needed to discern the types of permanent
placement practices that impact children and families of color.
Impact of Contextual Factors on Child Welfare Practices
for Children and Families of Color
Contextual factors can impact both the delivery of child welfare practices and the outcomes
of these practices. Although numerous contextual factors exist, this section will focus on three
areas: 1) cultural competency of child welfare workers and practices, 2) child welfare practice in
rural areas, and 3) the role of the court system. The potential impact of these contextual factors on
practices and outcomes for children and families of color will be discussed.
The Role of Cultural Competency in Child Welfare Practice
Child welfare workers often work with children and families from a wide range of cultures
other than their own. Many practioners and researchers have noted that effective child welfare
practices are those that acknowledge and incorporate the importance of culture in the delivery of
services (Miller & Gaston, 2003). Indeed, Miller and Gaston (2003) note that inherent assumptions
within the child welfare system are grounded in Anglo-Saxon values and cultural norms about child
rearing and family. Child welfare legislation and policies often follow European standards of
culture and White, middle class family values are the standard through which ethnically diverse
parents and children are compared. As such, children and families exhibiting alternative values may
be seen as deviant by the system. These conflicts in attitudes regarding acceptable parenting
behavior may contribute to ineffective or harmful child welfare practices (Miller & Gatson, 2003).
In an effort to combat ethnocentrism in the child welfare system, many agencies have placed
increasing importance on ensuring that workers, programs, policies and practices are "culturally
competent." In general, the term cultural competence refers to an ability to recognize and respect
similarities and differences in beliefs, interpersonal styles, values, norms, and behaviors of various
ethnic and cultural groups (Roberts, 1990, as cited in Schriver, 1998).
Although cultural competency is widely viewed as an important aspect of effective child
welfare practice, research on the cultural competency of child welfare systems and the impact of
this competency on children and families is lacking. Some studies describe the problems diverse
clients experience in trying to access culturally and linguistically competent services, other studies
have described ways to improve cultural competency in child welfare agencies, while other research
has addressed mechanisms for child welfare agencies to work more collaboratively with ethnically
diverse communities.
The issue of access to culturally and linguistically appropriate services for children and
families of color is of tremendous importance. Suleiman (2003) notes that for Latino clients in
particular, language barriers can exclude them from necessary services and produce significant
delays in service delivery. Indeed language barriers are implicated as a factor in the underutilization
of services among Hispanic clients (Gomez, Ruiz, & Rumbaut, 1985, as cited in Suleiman, 2003),
and many researchers have discussed the importance of language in accurate assessment and
treat~ent of clients (Suleiman, 2003). Suleiman (2003) further notes that denial of access to
41
services and programs for people with limited English proficiency is in violation of the Civil Rights
Act of 1964; yet federally funded agencies have wide discretion in the degree to which language
needs of clients are addressed.
As such, the responsibility to provide and monitor culturally competent and linguistically
appropriate services has largely fallen on child welfare systems themselves. Some studies have
proposed various models of achieving cultural competence in child welfare agencies. For instance,
Mederos and Woldeguiorguis (2003) discussed the role of child welfare managers in producing
culturally competent child welfare agencies. Recommendations for increasing cultural competence
included first and foremost, the managers' recognition that disproportionality in services and
outcomes for children of color does exist. Indeed, other researchers have noted that a prerequisite to
true cultural competence within an agency is a commitment to cultural competence principles and
practices among top level staff (McPhatter & Ganaway, 2003). Additionally, to produce agency
wide cultural competence, managers must prioritize staff diversity through recruitment and
maintenance of bilingual and bicultural staff. Child welfare managers also need to reach out to
diverse communities to form collaborations with community representatives. This collaboration can
help the child welfare system better understand cultural differences in parenting behaviors, as well
as assist in the creation of policies and practices that are relevant to those communities.
Indeed, other studies have also documented a need for improved collaboration between child
welfare agencies and culturally diverse communities. For instance, Hosley, Gensheimer, and Yang
(2003), described the implementation process of two programs aimed at improving collaborations
between the child welfare system and two Southeast Asian communities in Minnesota. Both
programs sought to improve practices and outcomes for children and parents from these
communities. The collaborations required committed involvement from numerous agencies, as well
as parents and youth from the communities. Language barriers made the use of bilingual and
bicultural staff at collaborative meetings extremely important. Effective collaboration depended on
a number of different factors, including the need for a trusting relationship between all parties, the
incorporation of different cultural styles in the structure of meetings, decision-making and
communication styles, the need to have bilingual and bicultural staff involved in the process and
that active involvement from th~ youth and parents is crucial to successful collaboration.
Cultural Competency:
Impact on Children and Families of Color
Research has not yet investigated the impact ofculturally competent child welfare practices
on children and families of color, however there is evidence to suggest that practices that are not
culturally or linguistically appropriate result in underutilization of services and ineffective
treatments and are in violation of clients civil rights (Gomez et al., 1985, as cited in Suleiman,
2003; Suleiman, 2003). Some researchers have described processes for achieving cultural
competence within child welfare agencies. A wide range of activities aimed at increasing cultural
competence have been identified, including support from top level staff, the prioritization of staff
diversity, and forming collaborations between child welfare and ethnically and culturally diverse
communities. These activities are intended to combat the child welfare system's legacy of using
European standards of culture and White, middle class family values in assessing diverse children
and families (Miller & Gatson, 2003).
42
However achieving cultural competency in child welfare can be a difficult task. For instance
it may be challenging for child welfare agencies to recruit and retain bilingual/bicultural staff and
the process of building collaborations with diverse communities may be time consuming and labor
intensive. Yet the impact of cultural competency on children and families of color is of tremendous
importance. More research is needed to discern how child welfare agencies recruit and retain
bilingual/bicultural staff, as well as how collaborations are built between child welfare and
ethnically diverse communities. Most importantly, research is needed to systematically evaluate the
impact ofcultural competence on practices and outcomes for children and families ofcolor.
Rural Child Welfare Practices
Much of the research literature on the effectiveness of child welfare practices is conducted
in urban settings. Researchers have suggested that urban child welfare research is often not relevant
to rural communities who may face uniquely different challenges than their urban counterparts
(Templeman & Mitchell, 2002). Indeed several studies have suggested that rural communities are
particularly vulnerable to a number of different risk factors. Rural communities in comparison to
urban communities have an elevated rate of poverty, there are fewer employment opportunities in
rural communities, and rural communities have been found to have lower levels of social well-being
than urban communities (Menanteau-Horta, & Yigzaw, 2002).
In addition to community differences, the characteristics of urban and rural child welfare
workers also differ. For instance, Landsman (2002) surveyed child welfare workers in 115 rural and
urban communities in Missouri and found several differences between the two groups. Findings
indicated that agency size played an important role in practice, many of the differences between
rural and urban child welfare practice were also noted between smaller and larger agencies. Rural
and small agencies possessed workers with a greater number of years spent in their current position
compared to urban and large agencies, they also experienced greater autonomy, decision-making,
growth opportunities, and agency supportiveness, as well as more job satisfaction and
organizational commitment. In contrast, urban and large agencies had more workers with a social
work degree, more work overload, degree of specialization and adequacy of resources, they also had
higher community stress and availability of jobs. These findings suggest that child welfare workers
in rural communities are generalist practioners who may be more satisfied with their job than their
urban counterparts, while urban workers tend to have more resources and a higher degree of
specialization.
Other studies have suggested that rural child welfare agencies provide more services and
intervene more often than urban child welfare systems. For instance, research suggests that rural
child welfare agencies have a higher likelihood of offering preventive services to children and
families that do urban agencies (Drake, 1996); they tend to have higher levels of substantiation of
maltreatment (English et al., 2002); and rural child welfare agencies are more likely to intervene in
cases involving drug exposed infants (Ondersma et al., 2001 ).
The differences between urban and rural communities and chil~ welfare agencies have
important implications for child welfare practice. Templeman and Mitchell (2002) have noted that
the application of urban child welfare practices to rural communities is inappropriate. In their
qualitative study of child welfare workers, they report the need for rural specific child welfare
practices. For instance, participants noted that families in rural communities are often
geographically isolated, they often have incomplete knowledge of available resources, lack of
43
access to specialized services, and restricted mobility. Thus, traditional child welfare services that
call on families to travel to and from a variety of specialized interventions are not generally feasible
with this population. However respondents also noted that rural communities possess certain assets
that should be utilized for more effective rural child welfare practice. Most importantly, rural
communities were seen by social workers as having a strong network of informal and at times,
formal supports, including linkages between neighborhoods, and the use of faith based
organizations. More effective child welfare practice for rural communities was described as using
these natural supports to build better collaboration between child welfare agencies and rural
communities. This collaboration was seen as a mechanism through which child welfare could
perform outreach to the community.
Rural Child Welfare Practices:
Impact on Children and Families of Color
The. impact of rural child welfare practices on children andfamilies ofcolor is not
sufficiently addressed in the research literature. In general, rural communities have an increased
rate of poverty, fewer employment opportunities and lower levels of social well-being than do urban
communities (Menanteau-Horta, & Yigzaw, 2002). Differences between child welfare workers in
rural and urban areas have also been found (Landsman, 2002), and there is some evidence to
suggest that rural child welfare agencies intervene at a higher rate than do urban child welfare
agencies (Drake, 1996; English et al., 2002; Ondersma et al., 2001). These differences between
urban and rural communities and child welfare agencies suggest that applying an urban model of
child welfare practice to rural communities is not justified (Templeman & Mitchell, 2002). The
unique needs of these communities and their child welfare populations call for increased research
efforts to identify effective practices for children and families of color in rural areas.
The Role of the Court System in Child Welfare Practice
For children who become dependents of the court, the judicial system may significantly
impact practices and outcomes. In child maltreatment proceedings, judges typically have authority
over decisions about services and certain case outcomes. Indeed, studies have noted that
noncompliance with court orders is a strong predictor of termination of parental rights (Atkinson &
Butler, 1996; Jellinek et al., 1992; Smith, 2003). Other studies report that the courts can have an
important influence over the services to which children are referred (Garland & Besinger, 1997).
As such, the ways in which court systems and child welfare systems interact is an important
contextual issue affecting practices and outcomes. In general there is little research on the ways in
which child welfare and court personnel collaborate and how this collaboration, or lack of
collaboration impacts children and families of color. However the few studies that have addressed
this issue report that the two professions experience numerous barriers to effective collaboration.
For instance, in Russell's (1988) survey of98 caseworkers and 69 attorneys, significant differences
in role perception were found between the two groups. Child welfare workers and attorneys
disagreed about who should perform a number of different tasks. A majority of both groups felt that
they should have primary responsibility for the same tasks, including requesting authorization for a
dependency hearing, deciding on the allegations, explaining and interpreting court proceedings to
parents, deciding if children should testify, entering into agreements with parents regarding case
disposition, and recommending dispositions to the court. Both groups expressed a need to have
attorney and child welfare worker roles clarified and a need for cross training between the two
44
groups. Additionally, ·in open-ended responses to survey questions, both attorney and child welfare
respondents expressed some irritation over their interactions with one another. Attorneys felt that
child welfare workers should "stop trying to be attorneys" (Russell, 1988, p. 212), and child welfare
workers expressed that "we know about what lawyers do, but they don't know about what we do.
Therefore training of lawyers to casework responsibilities would be helpful" (Russell, 1988, p. 212).
In an attempt to address barriers to collaboration between child welfare and the court
system, Johnson and Cahan (1992) reported on the "Children Can't Wait Project," a training
program aimed at cross training for attorneys and social workers, and overall system modifications.
The program convened a series of seminars for child welfare workers and attorneys. Seminars
included cross training and open explorations of one another's expectations. Social workers in the
seminars expressed a need for attorneys and court personnel to increase respect and understanding
of their profession, while court personnel described a need for social workers to provide more
timely and accurate case preparation and to be knowledgeable of legal definitions. Other seminar
activities included creating action plans to expedite court proceedings by improving communication
between attorneys and workers, and modifying court proceedings themselves. Although both
attorneys and social workers had good attendance at the initial seminars of the project, attorneys had
much less involvement in multidisciplinary groups working to implement the project. Outcomes of
these collaborative efforts were not described.
Additionally, Knepper and Barton (1997) explored courtroom dynamics of court personnel
and child welfare professionals during court proceedings. District judges, circuit judges, circuit
clerks, Foster Care Review Board chairs, Court Appointed Special Advocates, and attorneys from
five sites in one state completed surveys. Interviews with court personnel and caseworkers were
also conducted, as well as focus groups with foster parents and caseworkers, court observations, and
case record reviews. Findings revealed that in general, caseworkers felt that judges respected their
opinions and judgments about cases. Yet judges also demonstrated great influence on certain
decisions and actions. When caseworkers adhered to group norms, judges often rewarded them by
not ordering "unrealistic" practices (Knepper & Barton, 1997, p. 298), and providing more
reasonable time frames for mental health evaluations to be completed. Many respondents also noted
that court proceedings were often used to finalize agreements, rather than negotiate; they lasted an
average of 10 minutes with prosecutors and other attorneys doing most of the talking. Survey results
indicated that court personnel generally do not receive training on child maltreatment issues prior to
beginning their job. Approximately 77 percent of district judges and 79 percent of circuit judges
reported that there were no training requirements for attorneys working in child welfare.
The Role of the Court System in Child Welfare Practice:
Impact on Children and Families of Color
Few studies have evaluated the potential impact of the court system on children and families
ofcolor in the child welfare system. However the court system clearly does play a significant role in
case processes and outcomes (Garland & Besinger 1997; Knepper & Barton, 1997). For instance,
Garland and Besinger ( 1997) noted that White children in the child welfare system tended to be
more likely than African American or Hispanic children to be court ordered
counseling/psychotherapy, however no racial/ethnic differences were found in social worker
recommendations in counseling/psychotherapy.
45
Research also indicates that there are numerous barriers to effective collaboration between
court personnel and child welfare professionals, including lack of clarity in their different roles
(Russell, 1988). Studies suggest that collaboration between these two groups might be improved
through cross training and clarification of roles (Cahan, 1992; Russell, 1988). More research is
needed to discern how the court system impacts child welfare practices as well as outcomes for
children and families of color.
Summary of Key Findings and Implications
This section reviewed research on evidence-based child welfare practices and their
application to children and families of color, as well as the impact of contextual factors on child
welfare practices for children and families of color. Overall, research on the effectiveness of child
welfare practices across the various choice points in the CWS is somewhat limited and
unfortunately many studies fail to assess the possible differential impact of child welfare practices
on diverse ethnic groups. Nevertheless, certain practices evaluated in the literature are emerging as
promising practices for children and families of color and there is a need to continue to promote
evidence-based practice in child welfare.
Following the pathway that children may take through the child welfare system, research on
child welfare practices at the beginning choice points. including screening, investigation and
provision of voluntary services reveals mixed results on the effectiveness ofthese practices.
Research does indicate that cases involving children of color are investigated and substantiated at a
higher rate than cases involving White children (Ards, et al., 2003; Drake, 1996; Eckenrode et al.,
1988; Wells, et al., 1995). Children of color are also more likely than White children to be placed in
out-of-home care (Needell et al., 2003; U.S. DHHS, 2001).
Studies also demonstrate that there is little overall consistency in the decision-making
practices of individual child welfare workers (Rossi et al., 1999; Schuerman et al., 1999). Practices
designed to improve decision-making including the use of structured risk assessment tools, or
multidisciplinary teams/family group conferencing may have the potential to improve decisionmaking practices, yet research has not yet evaluated the effectiveness of these interventions with
ethnically diverse populations.
Perhaps the most promising beginning choice point practice to emerge from the literature is
the use of home-based prevention services. Research does demonstrate that home-based prevention
efforts are effective in preventing maltreatment and further involvement in the CWS, as well as
improving health related behaviors and outcomes for mothers and young children (Kitzman et al.,
1997; Olds et al., 1997).
Research on continuing care and ending choice point practices provide limited evidence
that these practices are successful in reunifying families or ensuring a permanent placement for
children. However, research does consistently demonstrate that Black children are reunified at a
lower rate than are White children (Courtney, 1994; Jones, 1998; McMurty & Lie, 1992; Wells &
Guo, 1999; Westat, Inc., 2001 c). Evaluations of traditional family reunification services. and
specific treatments, such as parenting education, substance abuse, domestic violence or Family to
Family/social network interventions appear to suggest that these services have limited overall
success, and tend to be most successful with lower-risk parents (Dore & Lee, 1999; Morre/ et al.,
46
2003). However, research does suggest that substance abuse services for parenting women in
particular, are most effective when they are residential, fairly intensive and allow mothers to live
with their children while in treatment (Clark, 2001; Haller et al., 1997; Metsch et al., 2001).
Research on mental health se-rvices for children in out-of-home care suggests that children
ofcolor are less likely to be recommended for and to use child welfare se-rvices while in out-ofhome care (Garland & Besinger, 1997; Garland et al., 2000). Studies on intensive interventions for
children in out-of-home care, including wraparound services and treatment foster care, have yielded
mixed results (Clarke et al., 1992: Farmer et al., 2003). Additionally, although research does
indicate that foster and kinship alternate caregivers are in need of a variety of services, research on
the effectiveness of these services is not available (Cuddeback & Orme, 2002). Similarly, research
on permanency planning practices is limited There is evidence that children of color are more
likely than White children to have lengthier stays in the CWS once parental rights have been
terminated (Barth et al., 1994; Kapp et al., 2001 ). Research also suggests that some child welfare
workers may fail to encourage transracial adoption or placement (Avery, 2000; Brooks & James,
2003).
Additionally, contextual factors such as cultural competency of child welfare workers and
practices; child welfare practice in rural areas; and the role of the court system all impact the
effectiveness of child welfare practices. Cultural competence in child welfare settings, although
widely acknowledged as important, has not been adequately researched. Studies do suggest that the
lack of linguistically appropriate se-rvices may result in the under-utilization of se-rvices, or
inaccurate or incomplete assessments and inte-rventions (Gomez et al., 1985; Suleiman, 2003).
Moreover, research also suggests a need to differentiate between rural and urban areas when
researching and discussing child welfare practices (Templeman & Mitchell, 2002). Rural areas in
comparison to 1!,rban areas have been found to have unique risk/actors and child welfare agencies
in rural and urban areas have been found to differ in a number of ways (Drake, 1996; English et
al., 2002; Menanteau-Horta & Yigzaw, 2002; Ondersma et al., 2001). As such, there is a need to
define and evaluate rural-specific child welfare practices (Templeman & Mitchell, 2002).
Although more research is needed to discern how the court system affects child welfare
practice, research does suggest that there are numerous barriers to effective collaboration between
court and child welfare systems. Cross-training and clarification of roles have been identified as
important factors in improving the relationship between the child welfare and court systems (Cahan,
1992; Russell, 1988).
In general, much of the research literature on the impact of child welfare practices is
inconclusive. Yet very few studies describe differences in outcomes by ethnic/racial groups and
most of the research fails to even mention the effect of culture and ethnicity on the delivery and
impact of child welfare practices. This gap in the research literature is striking and has important
implications for children and families of color in the CWS. The lack of adequate knowledge of the
types of practices, services, programs, policies, tools, or strategies that are effective for children and
families of color hinders our ability to remedy the disproportionate representation of children of
color in the CWS and may leave many children and families of color without adequate or
appropriate services. The next section focuses on child welfare practices in Santa Clara County's
child welfare system.
47
IV. SANTA CLARA COUNTY CHILD WELFARE PRACTICES
In Phase 3, one area of investigation focused on the impact of child welfare practices on
children and families of color in the Santa Clara County's Department of Family and Children
Services (DFCS). Several related analyses were conducted to understand the types of practices and
case characteristics that are associated with more successful outcomes for children and families
from various racial/ethnic groups; the types of child welfare practices used in Santa Clara County;
how practices in the Main County Offices differ from South County Office practices; how practices
interact with child, family and cultural characteristics of differing racial/ethnic groups to produce
various outcomes; and the role of the court system on child welfare practices. As such, this section
contains five overall analyses:
1) A successful case analysis. This analysis examined case record review data collected in
Phase 2 to determine the types of practices and case characteristics associated with
successful case outcomes.
2) A qualitative analysis of Santa Clara County child welfare practices. Focus group and
interview data from Santa Clara County workers were collected to obtain information
about current practices, how these specific practices might positively or negatively
impact children and families of color, and what practices might be created, enhanced or
maintained to better serve children and families of color.
3) A comparison of Main County Offices versus South County Offices. This analysis used
both quantitative data from CWS/CMS, as well as qualitative data from focus groups in
South County to compare case characteristics and practices between the two areas of the
county.
4) A multivariate analysis of factors associated with reunification. This analysis
examined system-related factors and case characteristics predictive of reunification of
children with their families.
5) An analysis of the role of the court system in child welfare practice. This analysis
used case record review data collected in Phase 3 to determine whether and what type of
court-ordered changes are made to the social worker case plan recommendations at the
jurisdictional/dispositional hearing. Results from agency-wide focus groups related to
the relationship between the child welfare and court systems are also provided.
Taken together, these five sections provide extensive and important information on Santa Clara
County child welfare practices and factors that are associated with successful outcomes for children
and families of color.
48
Successful Cases Analysis
In Phase 3, one aim was to define and assess successful cases and outcomes in Santa Clara
County's Child Welfare System. An analysis of factors associated with successful case outcomes
and how these factors and outcomes differ by racial/ethnic group helps shed light on the types of
practices and conditions necessary to promote enhanced outcomes for children and families of
color.
Methods
First, our task was to operationally define "successful cases and outcomes" for this
quantitative portion, which we based on the United States Department of Health and Human
Services (US DHHS) Children's Bureau standardized outcomes. These standardized outcomes are
part of the U.S. DHHS' child and family services reviews and the Title IV-E foster care eligibility
reviews (U.S. DHHS, 2003b). These levels of success and outcomes are:
Safety
• Children are first and foremost protected from abuse and neglect
• Children are safely maintained in their homes whenever possible
Permanency
• Children have permanency and stability in their living situations
• The continuity of family relationships and connections is preserved for families
Family and Child Well-Being
• Families have enhanced capacity to provide for their children's needs
• Children receive appropriate services to meet their educational needs
• Children receive adequate services to meet their physical and mental health needs
Second, from our in-depth reviews of the 403 cases sampled for Phase 2, we matched basic
information and additional data collected from the court reports during Phase 3 with the US DHHS
guidelines. Although information from every court hearing available within our sample was
recorded, we relied primarily on data from the jurisdictional/dispositional hearing and the last two
hearings of each case to evaluate "successful cases and outcomes." We then expanded our analyses
started in Phase 2 to evaluate "successful cases and outcomes" and differences by ethnicity for this
current phase.
Results
Safety
Regarding safety, referring to the US DHHS criteria mentioned above, we used three
variables to indicate "children are first and foremost protected from abuse and neglect." Number of
prior referrals and previous number of times in the CWS indicated a risk of harm affecting these
children. The number of times removed from the family indicated a family environment that was
assessed as harmful and thus warranted additional removal.
49
We also used the initial case assignment from the jurisdictional/dispositional hearing to
denote whether "children are safely maintained in their homes whenever possible." A similar
variable, case closure type, which also refers to maintenance of the child in the home, will be
discussed in the permanency section.
Number of Prior Referrals
Of the 403 cases, 48 had missing data. Of the remaining 355 cases, 118 (33.2%) had no
prior referrals, 111 (31.2%) had 1 or 2 prior referrals, and the remaining cases (35.6%) had more
than 2 prior referrals. This indicated that more than half of the cases had prior referrals, which are
used by social workers in establishing potential risk to child safety. One-third of the cases had more
than two prior referrals. Please see Table 1a.
. Retierralsa
Table 1a: Freauency of Pnor
Number of Prior
Frequency
Referrals
0
118
1
75
2
36
3
35
4
24
5
18
6
14
7
6
8
7
4
9
10
3
11
5
12
3
13
5
14
1
15
0
16
0
17
0
18
1
a
Based on 355 cases with valid information
Percentage
33.2%
21.1%
10.1%
9.9%
6.8%
5.1%
3.9%
1.7%
2.0%
1.1%
.8%
1.4%
.8%
1.4%
0.3%
0.0%
0.0%
0.0%
0.3%
There was not a statistically significant relationship between number of prior referrals and
ethnicity. An average of2.46 prior referrals (sd = 3.14) were made before the case was opened for
the current episode. Please see Table I b.
50
Table lb Num berofp nor
. Re erra1s B:y Eh
. Grouoa
t me
Total
White
African
American
Prior Referrals
2.64
Avg.
2.46
2.14
(sd)
(3.14)
(3.27)
(2.71)
Latino
Asian
American/PI
Other
2.46
(3.25)
1.62
(2.14)
2.81
(3.37)
a Based on 355 cases with valid information
Previous Times in the CWS
Of the 403 cases, 93 had missing data. Of the remaining 310 cases, 204 (65.8%) had no
prior experience with the CWS while 106 (34.2%) did have prior experience with the CWS. This
indicates that one-third of cases had an elevated risk for abuse or neglect as evidenced by prior
substantiated abuse or neglect. Please see Table 2a.
Table 2a: Previous Times in the CWS8
Previous Times
Frequency
In the CWS
0
204
1
85
2
10
3
6
4
4
5
0
6
1
a Based on 310 cases with valid information
Percentage
65.8%
27.4%
3.2%
1.9%
1.3%
0.0%
0.3%
There was not a significant relationship between the number of previous times in the CWS
and ethnicity. On average, children across ethnic groups were previously in the CWS 0.47 times (sd
= 0.83). Please see Table 2b.
Table 2b: Previous Number of Times in CWS bv Ethnicity a
White
Latino
Total
African
American
Average
(sd)
0.47
(0.83)
0.25
(0.49)
0.50
(0.92)
0.46
(0.78)
Asian
American/PI
Other
0.63
(1.06)
0.56
(0.82)
a Based on 310 cases with valid information
Number of Times Removed from Family in Current Episode
Of the 403 cases, 15 had missing data. Of the remaining 388 cases, 257 (66.2%) were
removed once and 131 (33.8%) were removed more than once, indicating that in one-third of the
cases, the safety level in the home regarding the protection from abuse and neglect was not
consistent. Please see Table 3a.
51
·1 .m Current E,p1so
Tabl e 3a: T1mes Removed firom Fam11y
· de8
Number of
Frequency
Percentage
Times Removed
1
257
66.2%
2
98
25.3%
3
24
6.2%
4
7
1.8%
5
2
0.5%
a Based on 388 cases with valid information
There was a significant relationship between the number of times removed from family in
current episode and ethnicity. Across all ethnic groups, children were removed an average of 1.45
times (sd = 0.74). African American children had the highest average number of times removed
from family in current episode with 1.67 times (sd = 0.90), and Asian American/Pacific Islander
children had the lowest average with 1.26 times (sd = 0.51 ). Please see Table 3b.
Table 3b: Number of Times Removed from Family in Current Episode by Ethnicity'a
White
Asian
Total
African
Latino
American/Pl
American
Average
(sd)
a
1.45
(0.74)
1.67
(0.90)
1.36
(0.68)
1.44
(0.68)
1.26
(0.51)
Other
1.64
(0.92)
Based on 388 cases with valid information
Case Assignment After Jurisdictional/Dispositional Hearing
Of the 403 cases, 12 had missing data. Of the remaining 391 valid cases, 88 (22.5%) were
assigned to family maintenance. This indicates that less than one-quarter of cases are safely
maintained in the home whenever possible. There was not a significant relationship between case
assignment after the jurisdictional/dispositional hearing and ethnicity. Overall, most cases were
assigned to family reunification (71.1%, 278 of391 cases). Please see Table 4.
• a
. d'1ct1ona
. 1/0"1spos1t1onaIHearmg·
Table 4 Case A"
ssuinment aft er Juns
White
Latino
Total
African
American
34
Family
27
88
7
(19.4%)
(25.8%)
(22.5%)
Maintenance
(14.9%)
Other
Asian
American/PI
9
(29.0%)
11
(26.2%)
Family
Reunification
278
(71.1%)
38
(80.9%)
91
(68.9%)
101
(72.7%)
19
(61.3%)
29
(69.0%)
Other
25
(6.4%)
2
(4.3%)
7
(5.3%)
11
(7.9%)
3
(9.7%)
2
(4.8%)
a Based
on 391 cases with valid information
52
Permanency
Regardingpermanency, referring to the US DHHS criteria mentioned above, we used three
variables to indicate whether "children have permanency and stability in their living situations."
The number of unique placement homes and the number of placements in the current episode ·
indicated the consistency of the children's living situation while in the CWS. Since permanency and
stability also imply a solution of consistency within particular settings, case closure type was also
examined. To evaluate the outcome of whether "the continuity of family relationships and
connections is preserved for families" we also examined the proportion of reunified families within
the case closure type.
Number of Unique Placement Homes in Current Episode
Of the 403 cases, 15 had missing data. Of the remaining 388 valid cases,-162 (41.8%) had
only one or two unique placement homes. 224 (58.2%) had more than two unique homes, which
indicated that less than half of the children in the CWS are in a relatively permanent living situation.
Please see Table Sa.
Table 5a: Num berofU mque
.
Pl acement Homesa
Percentage
Number of
Frequency
Placement Homes
14.2%
1
55
27.6%
2
107
16.0%
62
3
11.9%
4
46
7.0%
5
27
6.2%
24
6
2.1%
7
8
2.8%
8
11
1.8%
9
7
2.3%
10
9
0.5%
2
11
1.0%
12
4
2.1%
13
8
0.8%
14
3
0.5%
2
15
0.0%
16
0
0.8%
17
3
0.3%
18
1
0.3%
19
1
0.3%
20
1
0.0%
21
0
0.3%
22
1
0.3%
1
23
0.0%
0
24
0.3%
25
1
53
26
27
28
29
30
31
32
33
34
35
a
0
2
0
0
0
0
0
1
0
1
Based on 388 cases with valid information
0.0%
0.5%
0.0%
0.0%
0.0%
0.0%
0.0%
0.3%
0.0%
0.3%
There was not a significant relationship between number of unique placement homes in
current episode and ethnicity. Across all ethnic groups, there was an average of 4.54 (sd = 4.67)
unique placement homes in the child's current episode. Please see Table Sb.
Table 5b: Number of Unique Placement Homes by Ethnicity•a
Total
African
White
Latino
American
Average
(sd)
a
4.54
(4.67)
5.35
(6.48)
4.70
(4.60)
4.44
(4.26)
Asian
American/PI
Other
3.29
(3.22)
4.36
(4.59)
Based on 388 cases with valid information
Number of Placements in Current Episode
Of the 403 cases, 16 had missing data. Of the remaining 387 valid cases, 222 (57.4%) had
only one or two placements. However, 165 cases had more than two placements indicating that
42.6% of children in CWS are in a relatively inconsistent living situation. Please see Table 6a.
Table 6a: Num berof Umque
.
Placement Homesa
Percentage
Number of
Frequency
Placements
20.4%
1
79
37.0%
2
143
14.5%
56
3
42
10.9%
4
5.2%
5
20
2.8%
6
11
2.1%
7
8
0.8%
8
3
2.1%
8
9
1.0%
4
10
1
0.3%
11
12
0
0.0%
54
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
a
4
2
1
0
0
1
0
1
1
0
1
0
0
0
0
0
0
0
0
0
1
Based on 387 cases with valid information
1.0%
0.5%
0.3%
0.0%
0.0%
0.3%
0.0%
0.3%
0.3%
0.0%
0.3%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.3%
There was not a significant relationship between the number of placements in current
episode and ethnicity. Across ethnic groups the average number of placements in current episode
was 3.31 (sd = 3.34). Please see Table 6b.
Table 6b: Number of Placements in Current Episode By Ethnicity'a
White
Latino
Total
African
American
Average
(sd)
3.31
(3.34)
3.27
(3.23)
3.77
(4.06)
2.99
(2.78)
Asian
American/PI
Other
3.00
(3.16)
3.20
(2.83)
a Based on 387 cases with valid information
Case Closure Type
Of the 403 cases, 386 had valid information for the case closure type. Including those
adopted, reunified with family, or placed with a relative or guardian, 183 (47.4%) of the cases have
permanency and stability in their living situation. There was not a significant relationship between
case closure type and ethnicity. Please see Table 7.
55
Tb
a le 7: Case Closure Tvoe By Ethnicitv'a
Total
African
American
White
Latino
Asian
American/PI
Other
Family
Stabilized
98
(25.4%)
(27.1%)
13
40
(30.5%)
31
(23.5%)
5
(16.1%)
9
(20.5%)
Guardianship
established or
placement
with relative
44
(11.4%)
5
(10.4%)
13
(9.9%)
21
(15.9%)
2
(6.5%)
3
(6.8%)
Reunified
with parent or
guardian,
court or noncourt specified
25
(6.5%)
3
(6.3%)
11
(8.4%)
6
(4.5%)
0
(0.0%)
5
(11.4%)
Adoption
16
(4.1%)
4
(8.3%)
6
(4.6%)
2
(1.5%)
2
(6.5%)
2
(4.5%)
Emancipation
57
(14.8%)
6
(12.5%)
11
(8.4%)
25
(18.9%)
7
(22.6%)
8
(18.2%)
27
(7.0%)
6
(12.5%)
7
(5.3%)
10
(7.6%)
1
(3.2%)
3
(6.8%)
119
(30.8%)
11
(22.9%)
43
(32.8%)
37
(28.0%)
14
(45.2%)
14
(31.8%)
(FM)
Incarceration,
runaway or
medical
services
Court ordered
termination
and other
a Based on 386 cases with valid information
Reunification
Of the 403 cases, 386 had valid information to determine whether children were reunified
with their family at case closure. Of these 386 cases, 263 (68.1 %) were not reunified with family
56
and 123 (31.9%) were. This indicates the continuity of family relationships and connections are
preserved for less than one-third of these families. Please see Table 8a.
· h F amity
·1 a
T able 8a: Propo rf10n of Ch"ld
1 ren Reum"fi1ed wit
Frequency
Percentage
Not Reunified
with Family
263
68.1%
Reunified with
Family
123
31.9%
a Based
on 386 cases with valid information
There was not a significant relationship between reunification with family and ethnicity.
Please see Table 8b. However, it is important to note that in the multivariate analyses that appears
later in this report (please see page 103), significant variations in reunification by ethnicity were
found even after controlling for other demographic and system-related factors.
Table 8b Cht"Id ren Reum"fi1ed By Ethnicitya
Total
African
American
White
Latino
Asian
American/PI
Other
Not Reunified
with Family
263
(68.1%)
32
(66.7%)
80
(61.1%)
95
(72.0%)
26
(83.9%)
30
(68.2%)
Reunified
with Family
123
(31.9%)
16
(33.3%)
51
(38.9%)
37
(28.0%)
(16.1%)
5
14
(31.8%)
a Based on 386 cases with valid information
Family and Child Well-Being
Regardingfamily and child well-being, referring to the US DHHS criteria mentioned above,
we created a summary score based on 16 variables categorized in 4 domains (overall well-being,
economic status, health status, and substance abuse related to the mother and the father) evaluated
from the last two hearings to indicate if "families have enhanced capacity to provide for their
children's needs." Although we originally wanted to examine each domain separately, the
extremely sparse information available from the last two hearings of each case prohibited these
comparisons at such a detailed level.
We then evaluated whether the child's education status was positive from the last two
hearings to indicate if "children receive appropriate services to meet their educational needs.
57
Finally, we created a summary score to evaluate the status of the child. This was based on 8
variables within similar domains above (child's overall well-being, economic status, health status,
and substance abuse) to indicate whether "children receive adequate services to meet their physical
and mental health needs."
Families Enhanced Capacity to Provide For Their Children's Needs
Of the 403 cases, 156 had valid information to determine if families had an enhanced
capability to provide for their children's needs. A positive score on this quantified information
meant that this capacity was generally enhanced, and a negative score meant the capacity was
generally diminished. The possible range for this summary score was -2.0 to +1.0. The average
score was 0.01 (sd = 0.71), which indicated that on the whole, case outcomes seemed neutral in
terms of family's capacities to provide for their children's needs. There was a marginally significant
relationship with ethnicity. African American families appear to have a more enhanced capacity
while Asian American/Pacific Islander families seem to have a more diminished capacity. Please
see Table 9.
Table 9: Family's Capacity to Provide for Their Children's Needs By Ethnicity'a
White
Asian
Total
African
Latino
American/PI
American
Average
(sd)
0.01
(0.71)
0.35
(0.59)
0.00
(0.77)
-0.04
(0.69)
-0.50
(0.53)
Other
0.07
(0.62)
a Based on 156 cases with valid information
Children Receiving Appropriate Educational Services
Of the 403 cases, 76 had valid information to determine if children received appropriate
services to meet their educational needs. A positive score meant that their educational needs were
generally met, and a negative score meant that their educational needs were generally unmet. The
possible range for this summary score was -1.0 to +1.0. The average score was 0.30 (sd = 0.71)
which indicated that educational needs were being somewhat met. There was not a significant
relationship with ethnicity. Please see Table 10.
Table 10: Children Receiving Appropriate Services to Meet Educational Needs By Ethnicitya
Other
Asian
White
Latino
Total
African
American/PI
American
Average
(sd)
a Based
0.30
(0.92)
0.75
(0.71)
on 76 cases with valid information
0.32
(0.89)
0.03
(0.98)
0.50
(0.93)
1.00
(0.00)
58
Children Receiving Adequate Services to Meet Physical and Mental Health Needs
Of the 403 cases, 156 had valid information to determine if children received adequate
services to meet their physical and mental health needs. A positive score in the summary variable
meant that their needs were generally met and a negative score that their needs were generally
unmet. The possible range for this score was -1.0 to+ 1.0. The average score was 0.11 (sd = 0.71)
indicating that on the whole, the case outcomes seemed neutral in regards to meeting children's
needs. There was not a significant relationship with ethnicity. Please see Table 11.
T a ble 11 Ch.Id
I ren R ece1vmg Adequate serv1ces to Meet Needsa
Total
African
White
Latino
American
Average
(sd)
0.11
(0.71)
-0.05
(0._76)
0.21
(0.69)
0.07
(0.71)
Asian
American/PI
Other
0.13
(0.64)
0.07
(0.83)
a Based on 156 cases with valid information
Summary of Key Findings and Implications
In Phase 2, we analyzed major case characteristics on the main sample of 1720 cases
(primarily from CWS/CMS data) and additional psychosocial factors measuring child and family
well-being for the 403 cases relying on data gathered directly from the case files. We conducted
ethnic comparisons for both samples. In Phase 3, our aim was to evaluate successful cases and
outcomes using the US DHHS guidelines and assess ethnic differences. Results indicate that there
were minimal ethnic differences
Variations of Safety, Permanency, and
Family and Child Well-Being by Ethnicity
In regards to safety, there was a significant relationship between the number of times
removed from family in current episode and ethnicity where African American children had the
highest average number of times removed from family in current episode and Asian
American/Pacific Islander children had the lowest average.
In terms offamily and child well-being, there was a marginally significant relationship
between a families capacity to provide for their children's needs and ethnicity, with African
American families appearing to have a more enhanced capacity while Asian American/Pacific
Islander families seem to have a more diminished capacity. However, from this portion of our
analysis, the general absence of significant differences by ethnic group assessed from the end of the
case implies that certain aspects leading to a successful case and outcome are not related to
ethnicity.
It is important to note that major ethnic differences do exist in Santa Clara County's CWS,
but many of these variations appear at different choice points in the system (such as psychosocial
status at the beginning of the case, and assignment to voluntary family maintenance). In addition, a
combination of factors experienced through the course of the case may be related to success or
59
positive outcomes. However, our results suggest that ethnic differences at the conclusion of the
case are not evident, if we define success according to DHHS guidelines.
Implications Related to the Disproportionate Representation of
Children of Color in the CWS
Given the relatively few ethnic differences in evaluating successful cases and outcomes,
what does this imply about ethnicity and the CWS? The absence of ethnic differences in this
segment does not negate the previous findings in Phase 1 and 2 confirming that certain ethnic
groups are over-represented (i.e., Latinos, African Americans) while others are under-represented
(i.e., Asian American/Pacific Islanders) in the county's CWS. Phase 2 results also indicate that
there are significant ethnic differences in many demographic, system-related, and psychosocial
characteristics. Yet, what does the absence of ethnic differences in successful cases and outcomes
mean in context? While results suggest there is a clear disproportion of ethnic groups in the system,
once in the system, the children and families are generally faring the same. Their relatively
homogeneous experience may be a result ofthe "one-size-fits all" limited array ofservices offered.
Although this may seem equitable and logistically efficient, it appears ultimately ineffective in
serving a diverse group of children and families. Also, to determine better how ''well" these
children are actually doing while in the system, the availability ofcomparative data, clear
operational definitions for "success," and more reliable and valid data are needed.
Limitations and Recommendations
The major limitations for this successful cases and outcomes component include the lack of
information available (especially at case closure), to establish an operational definition of a
"successful case and outcome," and the need for baseline and comparative data.
Data Issues
To assess better whether safety, permanency, and family and child well-being are
maintained and enhanced, child welfare records should include more complete and measurable
information. We acknowledge that child welfare case records were not originally intended for
research. Yet, if the agencies responsible for providing effective services wish to evaluate the
performance of their programs more reliably and validly, improving information collection and
tracking should be considered.
One step would be to create a case summary check-out form used at case closure or in an
exit-interview type situation. The check-out form would indicate the child and family's status in
major areas of psychosocial well-being (i.e., health, education, and finances), in addition to the
existing case outcomes. Given the amount of missing data assessing these domains describing case
and child status, and the relative scarcity of information recorded in the last two hearings
(presumably the place where outcomes of a case would be recorded) this form would enable
measurement and evaluation of "successful cases and outcomes" in a more consistent objective
manner. The amount of missing data and the manner in which information is assessed also implies
the need for more reliable and valid methods of recording of information.
Operational Definition of Successful Case and Outcome
60
Having a clear operational definition of a "successful case and outcome" is vital to program
evaluation and research. When we used the US DHHS criteria, the initial impression maybe that a
relatively low proportion of"successful cases and outcomes" seems to exist in regards to safety,
permanency, and well-being as for example only 32% of the children are reunified with their
families. However, again availability of comparative data along with the consistent use of a welldefined operational definition of "success" is needed to understand the actual success rates for
children and youth in the CWS. Two questions arise from our results and interpretation. First, what
actually is "success?" For example, adoption for one case may be a failure if a family has improved
its ability to support and nurture a child, but was denied reunification because the system still
deemed those parents unfit. Adoption for another case may be a success if the system permanently
removes a child from a persistently dangerous or unhealthy family environment. The same
difficulty in evaluation occurs when a family is reunified (presumably a success) but the child later
falls victim to reabuse (failure). Finding an agreed-upon operational definition may be first a
theoretical or philosophical question rather than a research one, but is important for the County to
consider if their programs are to determine their effectiveness
The second question concerns a judgment of Santa Clara County's CWS performance
according to the proportion of "successful cases and outcomes." Given the above limitations and
issues, Phase 3 cannot give a final answer to how well the county's CWS is serving children and
families. This brings us to our third limitation and set of recommendations, the need for baseline
and comparative data.
Baseline and Comparative Data
It was difficult to assess the successful cases and outcomes in Santa Clara County's CWS.
Are the proportions of children reunified with their families or of those adopted high or low? And
compared to what? Implementing better record keeping and data collection over time on key
characteristics, based on an agreed-upon operational definition would better answer the question of
effectiveness and performance. Implementing these practices across counties and across states
would also facilitate evaluation.
The recent Little Hoover Commission report (2003) urges reform in the CWS, especially for
the county and state to improve the leadership and management structure guiding the system (Little
Hoover Commission, 2003). As part of this reform, an accurate evaluation of success for children
and families involved in the CWS is needed.
61
Qualitative Analysis of Santa Clara County Child Welfare Practices
The qualitative component of this study consisted of both focus groups and interviews and
was designed to obtain insights about current child welfare practices used within the DFCS, how
these specific practices might positively or negatively impact children and families of color, and
what practices might be created, enhanced or maintained to better serve children and families of
color.
Methods
For the purposes of both the focus groups and the interviews, "practices" were defined
broadly to include programs, services, strategies, policies, or tools. Qualitative methods, including
focus groups and interviews, are ideally suited for capturing and documenting common experiences
about what "works" and what might be improved from the perspective of individuals involved in
different parts of the child welfare system, including Social Workers, supervisors and clients. These
methods may also help to illuminate quantitative findings, offer information about formal and
informal local practices, and provide a context for understanding a variety of factors that might
influence practices. The methods employed for gathering qualitative data through focus groups and
through interviews are described below.
A total of 16 structured focus group interviews were conducted during the course of this
study. Thirteen of the focus groups were composed of Social Workers, specifically workers with
titles of Social Worker II or III. The focus groups were structured to explore practices at different
choice points in the child welfare system. Specific topics for focus groups included the following:
• Emergency Response/Dependent Intake (early choice points related to entry into the system and
initial assessment and investigation).
• Diversion (provision of voluntary family maintenance, informal supervision, early intervention,
and other efforts to divert families to services without court intervention).
• Case Assignment (early choice points related to cases that are substantiated and opened).
• Mandated Services (including court ordered parenting classes, substance abuse treatment,
domestic violence services, mental health counseling or other services).
• Time Limits (including local, state and federal time limits and the impact on practices with
children and families of color).
In order to ensure that concerns of clients were adequately represented in the project, three
focus groups were conducted with parents in the child welfare system. Individual interviews were
also conducted with supervisors with experience in each of the choice points described above.
Individual interviews were used to supplement and affirm findings from the focus groups.
Social Worker Focus Group Sampling and Recruitment
Prospective participants for the Social Worker focus groups were randomly selected from a
list of workers with title of Social Worker II or Social Worker III in different units/division. Lists of
randomly selected Social Workers were generated using SPSS (Statistical Package for Social
Sciences). This random selection was conducted ''without replacement," so that if Social Workers
were randomly selected for a specific focus group topic, they would be excluded from random
samples for other focus groups. Table 12 below provides a list of the 13 Social Worker focus
62
groups, the child welfare units from which focus group participants were selected, and the number
of focus group participants.
Approximately 15 to 20 Social Workers were randomly selected to receive an invitation to
attend each of the 13 Social Worker focus groups. In recognition of possible scheduling conflicts,
emergencies, or vacations of invitees, we invited more Social Workers than were expected to attend
each focus group. Prospective focus group participants were sent an initial invitation by email two
weeks to one month in advance of the focus group. Non-responding Social Workers were sent a
second invitation to participate in the focus group. Finally, prospective focus group participants who
agreed to attend a focus group were sent a confirmation email and/or phone call two to three days
prior to the scheduled group. Each focus group was comprised of approximately 5 - 8 members,
though some were smaller. In particular, the two case assignment focus groups were small, with two
and three participants in each of the groups respectively. A total of 67 Social Workers participated
in the focus group interviews.
Table 12 Emo ovee Focus Group Comoos1t1on
Focus Group
Units Included in Selection Process
Diversion
ER/DI
Case Assignment
Mandated
Services
Time Limits
Court Services
• Early Intervention/Intensive Family Services/
Home Supervision
• Informal Supervision
Family Resource Centers
• Asian Pacific Family Resource Center
• Nuestra Casa Family Resource Center
• Uiirani Familv Resource Center
Emergency Response
• Screening
• Emergency Response
Court Services
• Deoendent Intake
Emergency Response
• Emergency Response
Court Services
• Dependent Intake
Continuing and Child Welfare Services Bureau
• Court Supervision
• Domestic Violence Unit - Court Supervision
• Continuing
• Spanish Speaking Services
• Continuing Multilingual
Adoptions Family Assessment and Development
• Adoptions and Family Drug Court
Continuing and Child Welfare Services Bureau
• Court Supervision
• Domestic Violence Unit - Court Supervision
• Continuing
• Spanish Speaking Services
• Continuing Multilingual
Adoptions Family Assessment and Development
• Pre-Adoption and Drug Court
• Family Assessment and Development
Number of
Groups
Number of
Participants
2
14
2
5
2
14
2
2
11
63
South County
TOTALS
•
Family Linka~e
South County Region
• Gilroy Family Center
• Rosanna Child Welfare Unit
• Monterey Road Child Welfare Unit
• Highland Child Welfare
3
13
13
67
Social Worker Focus Group Procedures
The Social Worker focus groups were designed to examine practices that impact children
and families of color at different choice points or units within the child welfare system. To this end,
the focus group questions explored the context in which services are provided, specific practices
that might positively or negatively impact children and families of color, and recommendations
about practices that might be maintained, enhanced or initiated to better serve children and families
of color (please see Attachment 1 for the Social Worker Focus Group Interview Guide). Each of the
focus groups took place in a room located at a DFCS office. Ten focus groups were conducted at the
main administrative office in downtown San Jose and three were conducted at Gilroy Family
Resource Center in South County. The focus groups were approximately 1½ hours in length.
All focus group participants received and signed a consent form that explained the study and
confirmed their voluntary participation in the focus group. All focus group participants also
completed a brief demographic questionnaire. The summary of focus group participant
demographics is provided in Attachment 2.
Family Focus Group Sampling and Recruitment
Three focus groups were conducted with families involved in the child welfare system. Two
focus groups were conducted with parents of children from populations that are over-represented in
the child welfare system in Santa Clara County, specifically Latino and African American children
and families. Each of these focus groups was organized in partnership with one of the family
resource centers and non-random, purposive sampling methods were employed. Participants for the
Latino parents' focus group were recruited among recent graduates of a monolingual Spanish
speaking parenting class at the Nuestra Casa Resource Center. Four parents agreed to participate in
the focus group and two were able to attend. Staff ofUjirani Family Resource Center recruited
participants for the African American parents focus group. Five parents and four staff attended the
focus group. Comments from parents and staff were documented separately. Finally, a focus group
of foster parents was convened to explore the needs of children of color in out of home placement.
The six focus group participants in the foster family focus group were all women; 2 participants
identified as Latina, 1 identified as Latina/Native American, 2 identified as Caucasian/European
American, and 1 identified as African American.
Family Focus Group Procedures
The three parent focus groups were approximately I hour in length. Questions explored in
these group interviews were focused on what has worked well for clients in getting services from
Child Welfare; what has not worked well and what services should be kept, improved or created to
better meet the needs of families. Please see Attachment 3 for the Family Focus Group Interview
64
Guide. Participating parents signed a consent form and each parent received a $10 gift certificate
from Target as a token of appreciation for contributing to the study. The family focus groups are
described in Table 13 below.
Table 13 F amity
·1 F OCUS Group Compos1t1on
Focus Group
Sponsoring Organization
Latino/ Mono-Lingual
Spanish Speaking Parents
African American Parents
Foster Families
Nuestra Casa
TOTALS
Ujirani Family Resource
Foster Home Licensing,
Recruitment & Intake - Resource
Families Support Team
Number
of
Groups
I
Number of
Participants
I
I
5
6
3
13
2
Individual Interviews Sampling and Recruitment
Fourteen in-depth interviews, of approximately one hour in length, were conducted with
supervisors in different positions in the Department of Family and Children's Services. Prospective
supervisor interviewees were randomly selected using the same sampling procedures described
earlier. In several cases, because the total number of supervisors was small, all supervisors in some
practice areas (such as Emergency Response or Dependent Intake) were included on the list of
prospective invitees. Based on the assumption that some supervisors would decline or be ·unable to
complete interviews, more participants were invited than target number of interviews (15). A total
of23 supervisors were invited to participate in interviews and 14 completed interviews.
Individual Interview Procedures
The interviews used a critical incident approach that is designed to elicit descriptions of
specific behaviors that help formulate the critical requirements of a specific activity (Flanigan,
1954), such as effective child welfare practice with children and families of color. The critical
incident technique has been used to research decision-making in child welfare (Banach, 1998) and
is a valuable tool for obtaining information about specific behaviors and practices from
practitioners, who are in the best position to evaluate which of these practices "work" or "don't
work." Rather than gathering general opinions and hunches, this method allows interviewees to
recall specific experiences related to the provision of services and to delineate factors related to
success or failure in outcomes, such as outcomes related to child welfare intervention for children
and families of color.
Interviewees responded to three (3) open ended questions and several "probe" questions that
solicited detail about specific critical incidents in practice with children and families of color. In
general, the questions explored: 1) factors and practices associated with a case perceived by the
interviewee to represent a successful outcome with an African American, Latino or Asian
American/Pacific Islander family, 2) factors and practices associated with a case considered
unsuccessful, and 3) specific practices that result in or could potentially result in enhanced
65
outcomes for children and families of color (please see Attachment 4 for the Supervisor Individual
Interview Guide). Demographic information was collected at the close of the interview.
The supervisors interviewed for this study represented different practice areas of DFCS
including Emergency Response (5), Court Services (3), Continuing and Child Welfare Services
Bureau (2), Adoptions Family Assessment and Development (1), and South County (3).
Interviewees all had extensive experience working in the Department of Social Services and
possessed an average of 17.5 years working with County DFCS (range of5.5 years to 33 years). All
fourteen interviewees possessed an MSW degree and some also were LCSW's. Half of the
interviewees were men and half were women. Four of the interviewees were Latino/a, five were
Caucasian/Euro American, three were African American, and two were Asian American/Pacific
Islander. Ten of the key informants are fluent in speaking another language, and seven use these
language skills in their work. A summary of demographic information for the 14 interviewees is
provided in Table 14.
Table 14 Superv1sor Interv1ewee Demograp 1cs
Bureau/Unit
Gender
n(¾)
Emergency Response= 5 (35.7%)
Court Services
=3 (21.4%)
Continuing
= 2 (14.3%)
Adoptions
= 1 (7.1%)
South County
= 3 (21.4%)
n(¾)
Males= 7 (50.%)
Females= 7 (50.%)
Ethnicity
n (%)
= 3 (21.4%)
African American
Asian/Pacific Islander= 2 (14.3%)
Latino/Hispanic
= 4 (28.6%)
White
= 5 (35.7%)
YearsCurrent
Position
mean
(std. dev.)
3.7
(2.4)
Years County
mean
(std. dev.)
17.5
(8.9)
Qualitative Analysis Procedures
Two members of the Child Welfare Research Team, a trained focus group leader and a
research assistant, were present during each of the focus groups with the exception of three of the
smaller focus groups that were facilitated and documented by one focus group leader. Notes were
taken during both focus group and individual supervisor interviews. Audio tape recordings were
made during social worker focus groups and individual supervisor interviews to ensure accuracy of
the transcribed notes. Data for the qualitative portion of this study consist of the transcribed notes
taken during each of the focus group and individual interviews. Qualitative methods of data analysis
were employed to analyze the focus group and individual supervisor interview data. First, major
themes were summarized for each of the choice points in the social worker focus group and
supervisor interview data. Second, themes from the family focus groups were analyzed separately.
Third, a separate analysis of themes in the qualitative data from South County Offices was
conducted to provide insights and supplement quantitative analysis of possible differences between
South County Offices and Main County Offices within DFCS. Finally, common themes across all
focus groups and interviews were identified and documented. The findings of the analysis of themes
from focus groups and interviews are presented below and are followed by specific
recommendations that emerged across focus groups and interviews.
66
Qualitative Results: Themes from Focus Groups and Interviews
Participants in focus groups and interviews were asked to describe practices that may
negatively or positively impact children and families of color in the Santa Clara County's CWS, as
well as contextual factors that may affect the delivery ·ofthese services or client outcomes. This
section describes the main themes that emerged from these focus groups and interviews. Following
the pathway that children may take through the CWS, both negative and positive practices are
described at beginning choice points as well as continuing care/ending choice points. Lastly,
important contextual factors will also be discussed in relation to the delivery and impact of these
practices. 1
Beginning Choice Points:
Practices that Negatively Impact Children and Families of Color
The main themes related to practices that negatively impact children and families of color at
beginning choice points included, 1) inappropriate referrals to the CWS, 2) inconsistency in
decision-making practices, and 3) inadequate prevention of child maltreatment or entry into the
cws.
Large numbers of inappropriate referrals were noted as having a negative impact on children
and families of color. Study participants reported that, "Many mandated reporters have no clear
definition or evidence of abuse," and that low-income children and families are frequently referred
into the system because of issues more closely associated with poverty than abuse or neglect. Focus
group participants stressed the high number of referrals from schools "based on their observations
of children wearing old or dirty clothes," "because children had lice," or in one case "because a
child brought an expired can of food to a food drive, the child was thought to therefore be
neglected." Some participants commented that low-income children from wealthier school districts
were disproportionately reported to child protective services. "A child who is at a marginal
economic level will be seen differently at a school in Los Gatos as opposed to a school anywhere
else."
Although study participants stressed that inappropriate referrals are generally screened out
by the Screening Unit, repeated referrals can be damaging because they create a perception of
increased risk that may or may not be accurate. As the number of inappropriate referrals increases
for a particular family the likelihood ofDFCS opening a case may also increase. This situation was
described as having a potentially negative impact on children and families of color in particular
because they are often more likely to have socioeconomic problems. Some respondents also
suggested that children of color are more likely to be scrutinized by law enforcement. "Police are
more harsh on African American, API and Latino Families, especially if they don't speak English,"
commented another participant.
These inappropriate referrals were described as negatively impacting screening practices
because social workers often have to spend a great deal of time screening out inappropriate .
1
Please note that focus group and interview participants discussed contextual factors related to the court system. These
findings are presented in the section entitled "The Role of the Court System in Child Welfare Practice" which begins on
page 109.
67
referrals. This situation then often reduces the amount of time social workers have to identify and
serve families who are in need of voluntary services.
Additionally, study participants felt that, in some cases, inconsistency in decision-making
practices during beginning choice points negatively impact children and families of color. Although
decision-making practices occur at all stages of the CWS, many participants.discussed these
practices in relation to clients' entry into the CWS. Although staff generally noted the utility of
structured risk assessment tools used at the entry point into the CWS, they also stressed that these
tools are limited and that social workers need to develop recommendations using both assessment
tools and their own judgment. In relation to the latter, participants expressed concern that decisionmaking is often determined by the biases and perceptions of individual social workers. Social
workers named specific risk assessment tools that are used at beginning stages of the CWS as
useful, such as the decision tree used during screening and a written protocol used in Early
Intervention. At the same time, social workers suggested that application of assessment tools might
vary among social workers based on a number of factors including individual worker bias, size of
the unit and unit level, shared expectations about assessment, and level of supervisor support and
oversight.
Moreover, study participants noted that, "Decision-making all depends on the worker and
his or her attitudes and values," and "We can so easily sway the picture so it is positive or negative
by how we write up our narratives for the court reports; the report has a lot of influence and
sometimes peoples' personal issues get in the way of making informed decisions." Furthermore,
interviewees noted that social workers are not immune to stereotypes of different groups in the
dominant culture and in the media and that social workers often judge families based on their own
personal experiences and perceptions about what is "good enough" for children. Some social
workers pointed out that, for example, workers from other cultures might misinterpret
communication norms in African American families as abusive. "Sometimes African American
families have norms about speaking loudly, yelling, and threatening (like I am going to pop
you) ... but kids are rarely hurt ... an African American social worker might be less likely to
[negatively] interpret these kind of verbalizations than a social worker from a different
background."
According to study participants, decision-making practices are also influenced by
socioeconomic factors. Families experiencing unemployment, lack of access to resources such as
daycare, or problems obtaining health care for children may be initially assessed as having a higher
risk for maltreatment. For example, "older siblings are taking care of younger siblings because
parents can't afford childcare," noted one participant. "If there were more good daycare there
wouldn't be as many neglect cases, even once in the system we [workers] can't get them [clients]
child care," added another. Social workers also underscored the impact of local high housing costs
on low-income families. This problem was encapsulated by the comments of one worker: "Many
families partner up with other individuals to maintain housing and just because there are numerous
people living in the home, relatives sharing beds, doesn't mean that this is abuse or neglect. It's
poverty." The case example below typified the illustrations provided by study participants:
A Social Worker 1 was covering a case and she was concerned about the family because
there were four children living in a basement with their family. This was a Chicano family.
She reported that the children might need to be removed I went out to this social worker
and saw completely different things. The basement was an apartment. The place was small,
68
but had couches and a crib for the baby; it was cluttered but not dirty ... Maybe another
worker might have justified what this worker saw and removed the kids for perhaps no good
reason at all.
The issue of social worker attitudes toward clients was particularly salient to family focus
group participants. Several parents commented on the notable variation in how they are treated by
different social workers. A remark that typified this theme was made by one parent who said, "some
social workers do their job because they really want to help other and some just do it for the money
and don't seem to care about others." Parents also described feeling undermined by social workers
who appeared to have a negative attitude about the client. One client summarized, "some of the
social workers are disrespectful, unfriendly, they look down on the clients and make you feel
humiliated." They elaborated, "It gets to the point where it makes the client feel like 'I don't want to
go in there,' but I have to go in there for my child and when you come out you're in tears."
Inadequate prevention of child maltreatment or entry into the CWS was also identified as
having a negative impact on children and families of color at beginning choice points. Participants
commented on the need for services in the community that would help to prevent entry into the
child welfare system. Focus group members stated that there is a need for education about child
welfare laws and definitions of child abuse, particularly for cultures in which norms related to
discipline of children differ from expectations in the dominant culture. Participants suggested that
such programs could be offered through Social Service agencies but also in community based social
institutions, such as churches, and that some of the communities that have benefited from this type
of preventive education in the past include African American, Asian American/Pacific Islander
( e.g., Vietnamese, Chinese, and Filipino), Latino, and immigrant communities.
Study participants also emphasized the problems in obtaining services without opening a
case. For example, one worker stated that "to get into a wrap-around service from an outside
agency, such as schools or probation, without the child being a dependent takes months, by this
point the problem will have escalated." Many social workers stated that opening a case had both
benefits and drawbacks: families had greater access to services and support through resources such
as Family Preservation Funds but were at greater risk for having a child removed because of time
limits or establishing a history with DFCS.
Several social workers also noted that it is actually more work for them to maintain a child
in the home with Voluntary Family Maintenance (VFM) services than it is to remove a child from
the home and refer to court intervention. Additionally, for families who do make it into VFM or
Informal Supervision (IS) services, focus group participants pointed out that resources were not
available to provide follow-up support to families to prevent re-entry into the system once their
cases are closed. "The family may still have needs but we have to close it. We can only keep it in IS
for 6 months (or 3 months in VFM); it relates to funding because you lose federal and state money
when a case is closed and you are on County money-and there is none," exemplified comments
from interviews. Consequently, "clients may end up back in services." Ideally "workers would be
able to provide some additional help to them [families] to help themselves."
Time lags between case assignment to Voluntary Family Maintenance and social workers
actually receiving the case were also noted as an issue. "By the time the social workers gets the
case, three weeks have gone by and the social worker has only 2 ½ months left [out of three months
of services under Voluntary Family Maintenance]," elaborated one social worke~.
69
Beginning Choice Points:
Practices that Positively Impact Children and Families of Color
The main themes related to practices that positively impact children and families of color at
beginning choice points included, 1) orientation of clients to the CWS, 2) team decision-making, 3)
family involvement in decision-making, and 4) the use of child care and in-home services.
Providing an orientation for parents to help them understand the components of the child
welfare systems, the court process, current laws and policies related to time limits, and how to work
with social workers was named as a best practice for families first entering the CWS. Social
workers in South County portrayed this as a best practice that was formalized into an orientation
group while social workers in other areas of the county described conducting orientations with
clients individually. Social workers and supervisors advocated for strengthening and
institutionalizing the practice of providing an orientation for clients throughout the county. Some
participants suggested that an orientation workshop could be provided or co-lead by someone
"outside the system" that might be perceived by clients as more neutral.
It would be helpful for parents to go to a workshop to educate them about time lines.
Parents don't get it immediately. Sometimes by the time the client "gets it, " it may too late.
Also sometimes social workers don't explain it well. Sometimes clients are just traumatized
about what all this means and you have to walk them through the process for the first 2
months and you have 6 months time limit - it's very difficult. The workshop should also
teach the clients how to read a court report and what the recommendations mean.
Team decision-making <TOM), which is currently being initiated in Santa Clara County,
was viewed by participants as a promising tool for mediating individual bias and cultural
misunderstanding by social workers. "Even with the decision tree in place we know it hasn't been
working because there are still so many children of color and there is still a lot of the worker's
subjective perspective put into the report; hopefully TDM can reduce this." TDM was also
perceived as a vehicle to "help get everyone on the same page" with professionals in other systems,
such as schools and service providers. Staffing to discuss cases within units and with
interdisciplinary teams was identified as a positive practice for children and families of color that
could be enhanced by TDM. Some social workers expressed concern about the amount of time that
might be required by TDM's and that "it may be hard to get everyone together in a limited time."
Study participants identified family involvement in decision-making. including the use of
family conferences as a positive practice, particularly for children and families of color. Participants
suggested this was particularly important for cultural groups that place high value on extended
family. "In the Latino culture it's appropriate to extend an invitation to grandparents, godparents,
and all parental figures," illustrated one social worker. Study participants stressed the importance of
extending family involvement to include both biological and non-biological family members. For
example, some participants noted that African American families often have non-biological aunts,
uncles, or other extended family members that should be included in family conferencing. Many
social workers and supervisors viewed the developing family-to-family initiative as a promising
strategy for expanding and institutionalizing family participation in decision-making.
70
Last, the use of childcare and in-home services were also noted as a positive practice. For
instance, in one social worker's opinion, families with adequate childcare have about an 80 percent
participation rate in diversion services, compared to 50 percent participation rate for parents without
adequate childcare. Social workers, supervisors, and families discussed the importance of childcare,
in relation to the ability of families to succeed in services. "It's really helpful to have child care
through Calworks; We need to make sure that is not cut because people can't work or train if there
is no one to watch their kids -This is a huge thing."
Social workers also noted that the provision of in-home services allows for "working with
clients in their own environment" and providing opportunities for one-on-one mentoring and role
modeling new behavior. Social workers identified agencies and programs that provide in-home
parenting classes as more effective than traditional didactic intervention practices alone. Social
workers also stated that providing in-home visits and follow up services, with the support of Social
Worker l's, was an effective practice that is underemployed because of heavy caseloads and scarcity
of resources.
Beginning Choice Points:
Recommendations to Improve Practices
Recommendations to improve practices at beginning choice points in the CWS included 1)
reducing bias and increasing accountability in decision-making, 2) training professionals in other
systems, 3) providing client orientations to families first entering the CWS, and 4) investing in
prevention and early intervention. Specifically,
Reduce Bias and Increase Accountability in Decision Making
•
•
•
•
Implement team decision-making in a manner that includes consideration of client culture.
Examine ways to facilitate accountability for team decisions.
Explore mechanisms to ensure accountability in decision-making and minimize the possible
influence of individual or supervisor bias in decision-making.
Continue the practice of facilitating cultural consultations between staff, with the support of
supervisors, including consultations across units.
Continue and expand practices related to family conferencing.
Training Professionals in Other Systems
•
•
•
Provide or collaborate in providing training for mandated reporters, including schoolteachers
and health care professionals.
Provide or collaborate in cross training with other systems and agencies "so we can understand
one another's needs and views-what we need from each other." Cross training may also serve
"to increase communication" or even "help with preventing some families from coming into the
system, including families that are already being seen by the agencies."
Provide or facilitate more training for police regarding removal, working with social workers to
assess risk before removal, coordinating with social workers regarding child and family
interviews, and ensuring sufficient training for law enforcement officers assigned to conducting
sexual abuse interviews.
71
Provide Client Orientation
•
•
•
Educate parents referred into the system about child abuse and neglect laws. Provide workshops
in their own language on topics such as court processes, how to work with social workers,
timelines, and how to navigate the system.
Consider having such workshops lead or co-lead by someone outside of the system that may be
perceived by clients as more neutral.
Hire and Train Social Worker l's to assist with family education and orientation.
Invest in Prevention and Early Intervention
•
•
•
Explore mechanisms to continue providing prevention services, including education of diverse
communities about child abuse and elder abuse.
Explore mechanisms to fund counseling and intervention services for families at risk that are
less restrictive than federally funded programs such as Family Preservation, which frequently
require formally opening a case to allow families to access resources.
Continue support and explore opportunities to expand voluntary and diversion services,
including early intervention and voluntary services at the family resource centers should
continue. These efforts are critical to "determine if families should be in the system" and ensure
opportunities for less restrictive placements for children.
Continuing Care/Ending Choice Points:
Practices that Negatively Impact Children and Families of Color
Study participants noted three overall practices that negatively impact children and families
of color at continuing care/ending choice points, including 1) problems with service plan
recommendations, 2) inconsistency in assessment of substance abusing families and need for
stronger linkages with substance abuse services, 3) barriers and service gaps related to relative and
foster placements.
Problems with service plan recommendations, including a "cookie cutter approach" to
services, and unrealistic expectations of families were noted as practices that have a negative impact
on children and families of color. For instance, several study participants suggested that using a
"cookie cutter approach" or "template for recommendations" is a practice among both social
workers and district attorneys that negatively impacts children and families of color. Participants
offered a number of explanations for this dynamic including "social workers being jaded, believing
that since they had worked with a similar family before, that the family is the same and the
interventions should be the same." Similar observations from social workers in other focus groups
suggested that "sometimes workers will get challenged if they don't take the cookie-cutter
approach," and sometimes "DA's think all parents need counseling, parenting class, etc." Several
social workers called for greater flexibility in working with families and "more creativity regarding
case plans needs. For example, some parents may just need more time to spend with their children."
Others cited examples where clients were mandated to services that were not appropriate including
a mother mandated to services before an assessment was conducted and a father "who is in a 52week batters' program-who had one incident of losing his temper and hitting his child-and
72
though hitting her was not okay he would be better served in family therapy and an advanced
parenting class."
Additionally, some service plan recommendations were also described as placing unrealistic
expectations on families that can set them up for eventual failure. These practices tend to impact
low-income families and were perceived to disproportionately impact children and families of color.
First, service plans frequently require "a laundry list of mandated services" that are unrealistic for
families. "I have a lot of clients who when I sit down with them and go over the service plan with
them for the first time they say 'I'm overwhelmed, I can't do this." "Families are in a panic to get
everything done on the service plan from ER to continuation," explained one social worker focus
group participant. Other participants noted that a family might be mandated to substance abuse
services, parenting classes, and mental health services while seeking employment and new housing.
Some families "just give up," observed participants. Moreover, clients are often required to attend
services that conflict with their job schedules, particularly clients whose schedules change from
week to week. "The expectation for clients to reschedule their work schedule is difficult, especially
those that receive hourly wages, and causes even more stress for the family."
Many participants expressed concerns that high expectations imposed on families are
reflective of and "ideal" rather than "good enough" standard for families. One study participant
summarized this dynamic: "We are not required to have 'perfect parents,' we need minimal
standards of parenting ... Just because you or I would not parent that way does not mean there is a
risk to the child." "From a client's side, they are being told by a court system and a social system
that the way that they were brought up and the way that things were done in their [clients'] family is
wrong and not acceptable," noted a social worker focus group participant.
Finally, study participants cited insufficient resources to support the success of families
attending mandated services as a problem. For example, participants in several focus groups stated
that the shortage of bus passes and increased paperwork required to obtain the limited passes that
were available negatively impact families, particularly children and families of color. "Most of
these clients don't have a vehicle so for the low income families it is challenging for them to get
their children back ... how are they supposed to fulfill their required programs without
transportation?"
Inconsistency in assessment of substance abusing families and need for stronger linkages
with substance abuse services was also noted as negatively impacting children and families of color
at the continuing care/ending choice point. Participants stressed the importance of effectively
addressing the needs of families with substance abuse problems who they perceived to comprise a
majority of caseloads. Study participants described a number of success stories in which families
were reunified when parents completed substance abuse treatment and also highlighted the fact that
when appropriate substance abuse services are provided they can be an effective best practice.
At the same time, study participants suggested that protocols for decision-making with
substance abusing families and formal agreements with substance abuse treatment agencies were
inadequate. Participants also espoused different standards about when it might be appropriate to
reunify a child with a parent or place a child with a relative in recovery. Furthermore, participants
suggested that there was a need for developing protocols for addressing relapse with substance
abusers in conjunction with both treatment programs and the courts. "We [social workers] pay lip
service, we say relapse is a part of recovery, but the system does not reflect that." Finally, several
73
study participants noted a lack of clarity and consistency in standards for assessing risk in the case
of drug exposed infants, suggesting that race and class bias may influence both initial referral and
assessment of parents of drug exposed infants in child welfare.
Study participants also commented on the scarcity of substance abuse services, particularly
residential programs for women and children. Although a variety of substance abuse treatment
services are offered throughout Santa Clara County, participants pointed out that the need for
services was greater than availability, particularly for clients whose primary language is not
English. Several participants, particularly in South County, noted that a centralized intake process
for entree into treatment was a barrier for many families because of time and transportation issues.
Additionally, social workers felt that barriers and service gaps related to relative and foster
placements also negatively impacted children and families of color at continuing care/ending choice
point practices. While the use of relative placements was described as an effective practice for
children and families of color, study respondents proposed that poverty issues and timelines
converged with other factors to create substantial obstacles to relative placement. First, pressure to
move children out of the children's shelter, in the context of little staff support to locate prospective
relative placements, often results in lost opportunities to place a child with a family member. Some
social workers described feeling "in a double bind" and, in some cases, deeply frustrated because of
having to place a child in a foster care situation that was perceived as less positive for the child.
Second, even if a relative placement is identified, it is often "impossible to complete the relative
placement paperwork within 24 hours, with the required background checks and fingerprinting."
Third, although social workers can seek a Director's Exemption to place children with families
where members may have had a criminal background, many social workers expressed concern
about the fact that even a distant criminal history can create barriers to appropriate relative
placements. Finally, new standards for relative placement "are the same as a licensed foster home
and are unrealistic for many families; this causes children to not be placed with relatives."
In general, although study participants expressed commitment to ensuring appropriate
relative placements for children, the study participants suggested that the requirements represented
by the relative packet "doesn't take into account economics or cultural practices" and
"disproportionately impacts families of color." These concerns were repeated across focus groups
and interviews and were epitomized by the following observations and examples:
The only people that can flt into it [relative packet] are those with money ... the packet
doesn't take into account the families we deal with. Many families are able to put four
children in a bedroom and they 're able to do fine, each child has their own space. It
[relative packet] does not allow for any arrangement like that at all. "
There was a case with 10 and 12 year old Hispanic/Native American girls who were
abandoned and went to live with a relative. The relative contacted CPS and stated that the
mother had not returned and wanted to know what to do with the children. The worker was
unable to assess the relative 's home immediately; therefore, the children were then placed
in a foster home. The licensed foster home was horrible, they ffoster parent] placed the
older girl on the floor on a crib mattress. It was terrible and all because they couldn't
continue to stay where they had been, because I [workerJ couldn't get the clearances fast
enough.
74
Three young Mexican American girls needed a place to live. The aunt was willing to take
them in and the children were comfortable with her. The uncle had a criminal history and
was a recovering heroin addict, who had been clean/or approximately 20 years. On paper
the uncle looked really bad, the worker was unable to clear his past. The children were
placed in foster care and may be adopted ifthe mother doesn't comply with the s~rvice plan.
In addition to barriers to relative placements, social workers pointed out a need to develop
strategies for working with foster care and adoptive families. Respondents noted that these alternate
caregivers are often deeply disappointed when the children in their care are reunified. Some social
workers suggested working more closely with fost-adopt families about the probabilities for
reunification. Foster family and social worker focus groups also expressed a need to provide
training to foster families in practical strategies for caring for children across cultural groups and a
need to develop mechanisms for experienced foster parents to mentor new foster parents.
Continuing Care/Ending Choice Points:
Practices that Positively Impact Children and Families of Color
Social workers described three overall practices that positively impact children and families
of color at the continuing care/ending choice points including I) parenting education services that
involve both parents and children in services, 2) cultural matching for children in foster care, and 3)
recruitment, training and coaching for foster families.
Parenting education services that involve both parents and children in services were
described as practices that positively impact children and families of color. Social workers and
supervisors noted that parenting classes that involved children and provided opportunities for
parent-child interaction were also a promising practice in comparison to traditional parenting
classes. Some social workers suggested that it might be helpful to ''wait until children have been
returned to order parenting classes" or to arrange for children's participation in classes.
Additionally, social workers described cultural matching for children in foster care in which
workers attempt to place children in a home with foster parents of a similar background as a
preferred practice. This was perceived to be critical in relation to matching the primary language of
the child and particularly important in the context of federal and state time lines and concurrent
planning. At the same time, both social workers and foster parents stressed the importance of a safe
and caring placement for children, whether a cultural match was possible or not. Social workers
expressed concern that the particular issues and needs of children, including but not exclusive to
culture, are sacrificed to the expedience of"finding any placement." Foster parents echoed a
concern about considering both the needs of the child and the experience of the foster parents when
arranging a "match." One foster parent caring for a special needs child of a similar background
noted that she would have been overwhelmed and tempted to abandon the idea of foster parenting if
her first foster child had been particularly challenging. Foster parents stressed that they and other
foster parents often have successful experiences with children of different backgrounds and that
they would "never want to see a child wait for a permanent home or left in a shelter because of a
cultural difference."
Social workers and foster parents suggested that recruitment. training and coaching for
foster parents is critical to working effectively with children and families of color. Specifically,
foster families need to know about how to care for children placed in their home. For example,
75
foster parents caring for African American children for the first time required basic information
about hair care. Foster parents suggested that best practices for working with children of color
included obtaining coaching and training from other foster parents, learning about the child's
culture and the individual child, and collaborating with social workers to interface with birth
parents. Study participants noted that tensions between foster families and birth families were
common but that positive practices for children required minimizing these potential problems.
Foster families described "success stories" in which foster parents maintained supportive
relationships with children who were returned to their family of origin.
Social workers and foster parents both noted the need to recruit diverse foster families,
particularly Latino and African American families in Santa Clara County. One foster mother who
identified as Latina suggested that some families in her community might be reluctant to become
foster parents because they are more likely to already have children than Anglo families. Study
participants identified outreach to prospective foster families in underrepresented communities as an
important continuing practice.
Continuing Care/Ending Choice Points:
Recommendations to Improve Practices
Recommendations to improve practices at continuing care/ending choice points included, I)
providing or facilitating access to concrete support for families, 2) facilitating increased availability
and access to services, and 3) supporting and developing foster families. Specifically,
Provide or Facilitate Access to Concrete Support for Families
•
•
•
Examine options for restoring access to bus passes for clients, both in terms of increasing
number of available passes and decreasing paperwork required from social workers to obtain
passes for clients.
Explore ways to increase access to childcare for parents participating in voluntary or mandated
services.
Facilitate access to services related to transitional housing, jobs, financial assistance (e.g.,
CalWORKS), and for families as well as educational interventions for children in the child
welfare system.
Facilitate Increased Availability/Access to Services
•
•
•
Work with policy makers, courts, communities, and other service delivery systems to identify
and address issues related to insufficient services for families. Areas where services are lacking
or limited in relation to the perceived need include substance abuse treatment (particularly
residential treatment for women and children), transitional and drug free living, mental health
services for children, services for young children exposed to trauma or substance abuse, after
school programs and other resources for children.
Collaborate with other systems to protect funding of key services in the context of state budget
problems, such as services for Oto 5 programs.
Work with contract agencies to facilitate inclusion of children in parenting classes to allow for
both didactic and experiential learning opportunities.
76
•
Sustain and increase in-home services, such as intensive wrap around services and one-on-one
mentoring during home visits.
Support and Develop Foster Families
•
•
•
Continue and expand efforts to recruit foster families that are representative of diverse
populations in the child welfare system and that match the languages of different children and
families (e.g., additional Spanish speaking and Vietnamese speaking foster families).
Provide training for foster families in practical strategies for caring for children across cultural
groups.
Continue to develop mechanisms for experienced foster parents to mentor new foster parents.
Contextual Factors that Negatively Impact Practices for Children and Families of Color
Study participants noted three overall contextual factors that negatively impact child welfare
practices for children and families of color 1) agency level factors that detract from social work
practice, 2) a need for more bilingual and bicultural social workers and services, and 3) time limit
constraints.
Certain agency level factors that detract from social work practice were mentioned by many
study participants as negatively impacting practices for children and families of color. For instance,
heavy caseloads and staff shortages were identified as critical factors impeding effective practices
for children and families of color. Children and families of color, particularly families who may be
less trusting or more intimidated by the system, may require language specific services, or who may
have multiple vulnerabilities were deemed very likely to be impacted by the heavy caseloads of
social workers. One social worker who works primarily with Spanish speaking families observed
that many of her clients require "a lot of hand holding" in navigating through the system. Given
heavy caseloads, even dedicated social workers that invest substantial time with some families will
have other families "slip through the cracks." One participant explained:
The weight ofthe caseload is so heavy-there's no time to do social work. I always feel I
have to rush. Every minute ofyour day is booked. There is no time to explain court
processes to clients.
Many participants commented positively on agency plans to reduce caseloads. This was
recognized as a critical move toward improving overall effectiveness in day-to-day social work
practice. Yet some social workers commented on some of the weaknesses in the current method for
determining caseloads. "There's no equity in the way the agency does the point system for
determining workers' caseloads .... it's based on a 100 point system and you could have 70 points
that could be 35 children in permanent placement, which is a lot of children to manage, whereas
someone else could also have 70 points in Family Reunification (FR) cases, which is not the same
workload." Others suggested that there appeared to be "no rhyme or reason to how cases get
assigned-workers with a lot of points will still get assigned cases-even when a different worker
has a lower number of points." Social workers noted that the point system accounts for the number
of children in a caseload and also assigns a higher number of points to FR cases than to Permanent
Placement (PP) and Family Maintenance (FM) cases. However, they point out that the point system
77
does not necessarily account for the severity of the case and for individual differences between
cases. "Even a case with low points can be very complicated and take up a lot of your time."
Social workers also highlighted the lack of adequate support staff to work effectively with
families, including help with transportation, investigating relative placements, and follow up with
clients at home as negatively impacting children and families of color. Social workers also
commented on the emotional strain on the job and risk for burnout that is exacerbated by too little
support.
We don't have enough support staff-not enough Social Worker J's. It's a lot harder to
manage a good FR case without a SW I to document things, help with transport, etc ... it
becomes more challenging.
This leads you on the road to burnout-I always want to do my best-I want to work harder
to get the parents to be successful. There are some workers who are tremendously burned
out.
Additionally, social workers noted that substantial amounts of required paperwork can
detract from building relationships with clients. Several social workers suggested that, as an
example, writing interim court reports in addition to required reports was duplicative and used time
that would be better spent with clients. Social workers also pointed out that not having easy access
to centralized and updated information about resources and services for families impedes effective
practice. The need for centralized and current resource information was noted as important to
avoiding "reinventing the wheel" in relation to researching resources that other workers already
knew about.
Moreover, confusion about the agency's overall mission, as well as confusion about key
agency and child welfare policies was also described as negatively impacting child welfare practices
for children and families of color. A need for greater clarity about the mission of DFCS was
expressed. "Who is the client, the child or the parent?" asked one participant. Participants varied in
their perspectives about the balance between child protection and family preservation. Social
workers also expressed confusion over certain key policies including concurrent planning, home
assessment protocols and protocols for voluntary versus court placements. Social workers suggested
that clearer guidelines related to implementation of concurrent planning were needed. Additionally,
participants reported that a more detailed written protocol on home assessments would be
beneficial, "the policy is not in writing, there is no written protocol and the most recent memo has
DRAFT written on it." Consequently, it is "difficult completing assessments and explaining the
policy to relatives, when the workers aren't clear about it themselves." Last, uncertainty about
protocols for voluntary versus court placements were noted. "There are misconceptions about
voluntarily placing a child and when that child is placed into protective custody... about whether
this protocol is required for all placements, it seems to go back and forth," commented one social
worker. Another noted "I don't have the understanding about what is a voluntary placement versus
a child taken into protective custody; it seems that ifwe get involved in any kind of removal that it
generates a protective custody."
A need for more bilingual and bicultural social workers and services was also noted as a
contextual factor that negatively impacts children and families of color. Although study participants
generally commented on the diversity of DFCS staff as a strength, there was a perceived need for
78
continued growth in this area. "There are not enough workers who speak the different languages of
the families," noted focus group participants. Some social workers suggested that "the clients are
left in limbo when a worker who speaks their language is not found" and "when a worker is found
that does speak the language of the client then two months have passed by; due to time line
constraints the worker is forced to make decisions without knowing much about the family."
"English speaking workers assigned to Spanish-speaking clients is a problem-even if the parents
and child speak English, a lot of the extended family may not," noted one worker. Reliance on
interpreters was identified as problematic because it can be difficult to find interpreters within a
limited amount of time which can lead to an inability of clients to obtain a timely answer to a
question from a worker because it may take days to find an interpreters. Moreover, there can be
problems with interpreters who are culturally inappropriate, who do not translate appropriately, or
who fail to read all of the court report to the clients.
In spite of these limitations, additional qualified interpreters and translators are needed to
facilitate communication with clients and to provide materials in client's primary language. Study
participants stressed the need for increasing language capacity both through social work staff and
through access to qualified interpreters. Access to qualified interpreters remains critical to working
in a region with representation of a growing number of cultural and language groups. "Lack of
interpreters causes continuance of cases," noted one social worker. Study participants offered a
number of illustrations of the impact that language barriers had on clients. One example is offered
below.
There are only two Korean interpreters in the South Bay. In one case with a Korean family,
Korean interpreter could not be found for a court hearing. This father had to take offfrom
work five days in a row to go to court, because each day an interpreter wasn't there. This
impacted the father because of the time he had to take offofwork. Each day he became
more and more frustrated. By the time the judge saw the case, the judge saw him at his
most frustrated point.
In addition to the need for bilingual and bicultural social workers, study participants also
expressed a need for more culturally competent services in general. "There is a need for more
culturally diverse counselors and therapists," noted study participants. Barriers to successful
compliance with mandated services include lack of counselors who have knowledge of different
cultures and long waiting lists for the few culturally diverse counselors in the area, according to
study participants. Furthermore, social worker participants observed that White families appeared to
attach less stigma and experience less resistance to accepting mental health or counseling services
than families of color. "Lack of services designed for the needs of a specific ethnicity or family can
cause Dependent Intake to hold onto cases longer due to being unable to find the necessary
resources." Some social workers suggested that existing services for English speaking clients varied
in quality and that contract agencies also differed in the degree to which they prioritized access to
services for DFCS clients.
A shortage of services for monolingual non-English speaking clients was identified as a
substantial problem for many families. "Families stay in the system longer because of lack of
Spanish speaking and other language services ... this can be frustrating for families, they want to
give up." Study participants described struggling tenaciously to find appropriate services for
families. Examples included searching for domestic violence and substance abuse services in
Tagolog, finding a Vietnamese-speaking counselor, and locating a group home for a deaf child.
79
Several participants expressed concerns that services provided in English for clients whose primary
language is non-English were ineffectual and, on occasion, harmful to families such as when
"English-speaking counselors who work with monolingual families will use the children as
translators." Spanish speaking services were perceived to be an issue throughout the county and
were particularly described as pressing by South County study participants. During one focus group
discussion of this issue a participant provided an example:
A Hispanic monolingual mother living in Sunnyvale does not have transportation [to
mandated services]. There are no Spanish resources in her area. They are all in the San
Jose/Santa Clara area, which causes the family to commute farther. It is difficult finding
resources in her area and with no bus passes transportation is a big issue.
Study participants also described time limit constraints as having a negative impact on
children and families of color. Focus group participants noted that it was often difficult to work
toward family reunification in the context of federal and state time limits. Participants uniformly
affirmed the importance of expediting placement of a child into a permanent, safe and stable home
and to seeking the least restrictive placement options for children. At the same time, they suggested
that timelines made it difficult to have enough time to work with families that require services in
order to safely reunify. "Fifteen months is not enough time to work with the client's
issues/problems. The reality of what people are dealing with is not there in the law," was a typical
observation by social workers. Several social workers and interviewees offered examples of
children who were placed into adoption who would otherwise have been able to remain with their
families if time limits were less rigid: "Court time limits played a large role in the adoption of the
child ... the time limits affected the outcome of the family and did not take into account how difficult
it is to change the way one had lead their life in a matter of a few short months."
The issue of access to timely services in the context of time limits was also perceived as an
important issue. An interviewee described a successful case in which a Latina client with a
substance abuse problem was able to reunify largely because the social worker was able to get the
mother into substance abuse treatment despite waiting lists: "There are waiting lists for services and
not enough time to successfully complete treatment with the waiting lists and the time limits ...This
family was lucky; the worker was able to get the mother into treatment within 2-3 weeks, so being
at the right place at the right time accounts for this [success]." Conflicts between child welfare time
lines were perceived as particularly common in relation to addressing the needs of substance
abusing parents that represent a majority of cases. "The Federal government only gives families in
permanent planning 12 months to complete services and parents only have 6 months for children
under 3 years of age; this is especially difficult for parents who abuse substances."
Study participants suggested that time limits disproportionately impacted families of color:
"Children of color are brought into the system because parents are on drugs,.they aren't being
watched and are in an unsafe environment at that point. The children stay in the system because the
expectation is for the parents to all of a sudden to get better because their children are in the system,
'so if you [parents] don't get better we [social workers] are going to keep your children. This is
unrealistic, especially with the court timelines," exemplified comments made by many study
participants. Families whose primary language is other than English were also perceived as being at
greater risk for failing to reunify: "It is difficult to meet timelines with certain services that are not
offered in different languages; clients are placed on waiting lists or those services don't exist."
80
Contextual Factors that Positively Impact Practices for Children and Families of Color
Respondents noted four overall contextual factors that positively impact child welfare
practices for children and families of color including I) cultural matching and cultural consultation,
2) supportive supervision 3) collaboration with other agencies and systems, and 4) a strength-based
approach to services.
Cultural matching and cultural consultation was noted as a best practice for children and
families of color. Although study participants did describe a need for more bilingual and bicultural
staff, they also commended the agency's efforts to provide a cultural match between clients and
social workers. Cultural matching was viewed as an effective strategy for facilitating development
of rapport with clients and minimizing the potential for cultural bias in assessment. Specifically,
social workers and supervisors valued the culturally specific Emergency Response (ER) units and
acknowledged efforts to match clients and social workers as much possible in Dependent Intake
(DI). At the same time, respondents asserted that it is not always possible to assign to clients
workers of a similar cultural background. In this context, social workers described obtaining
cultural consultation from other child welfare professionals as a best practice.
At the same time, study participants acknowledged that clients might prefer to interface with
social workers of a different culture than their own. For example, several social workers that
specialize in working with Asian American/Pacific Islander groups commented that some families
prefer to be assigned to a social worker of a different background out of a sense of privacy, pride, or
concern about judgment. Both social worker and client focus group participants noted that cultural
matching was not a guarantee against bias. Social workers are not immune to beingjudgmental
about clients of their own culture or biased in relation to other issues such as socioeconomic status.
In some cases, clients described their experiences with social workers from a different background
as positive and that the quality of respect was more important that sharing a common background.
To this end, continuing training on cultural competence and working effectively with specific
population groups was also considered a key practice.
High quality, supportive supervision that integrates coaching in working effectively with
diverse families was viewed as critical to practice with children and families of color. "Some
supervisors want to hear about cases and are supportive-others just want to know if you got the
reports done on time," summarized one participant. In addition, supervisors were identified as key
players in the process of brokering cultural consultations with other social workers, particularly
between social workers from different units.
Collaboration with other agencies and systems concerned with the safety and welfare of
children and families of color was also highlighted as a beneficial practice for children and families
of color. For instance, developing interagency formal agreements and connections with liaisons in
other agencies was described as beneficial because it facilitates collaboration and minimizes the
potential for agencies to give conflicting messages and mandates to clients. Social workers are "able
to divert cases or get a better picture of the family when two agencies are working together to work
with the families, which helps not only us but also the client." Other agencies mentioned in this
context include probation, hospital, medical providers, family and mental health agencies,
CalWORKS, substance abuse treatment agencies, domestic violence agencies and other entities that
may be involved with clients.
81
Additionally, collaboration with other agencies to address systemic issues was described as a
best practice. Social workers noted that community meetings are held on a monthly basis for all
community agencies and are attended by program managers, but workers are encouraged to attend.
Participation in policy level discussion between systems was considered to be an important part of
collaboration. Several participants mentioned the Greenbook project, an initiative to better
coordinate domestic violence and child welfare services, as a model for interdisciplinary problem
solving and policy development. The Greenbook project successes could help inform evolving
collaboration with other key systems, such as tqe substance abuse treatment system.
Other collaborative efforts such as outstationed social workers, co-location of services and
Family Drug Court were noted as best practices for children and families of color. Participants
described having ER social workers at different sites such as the Family Violence Center and police
stations as a valued practice. Moreover, South County social workers noted "we are located at the
resource centers where there are community providers here that are not child welfare ... we've got
AA meetings, parenting classes-everything is centralized here and that makes it comfortable for
the families." Although social workers mentioned that co-location of services have been reduced
because of budget changes and other factors, it remains a positive practice for children and families,
particularly low-income families and families of color that may have less time and fewer resources
to access services. Lastly, social workers and supervisors described the family drug court as a model
for effectively working with substance abusing families in the child welfare system. Several
examples of "successful" cases involved families assigned to family drug court.
A strength-based approach to services was also mentioned as an important best practice that can
be used at all points in the CWS. Indeed, strength-based practice was lauded as a fundamental tenet
of social work that is crucial for working with children and families of color. Specific elements of
this practice named by study participants include, looking at the unique characteristics of each
family; building solutions with families and showing that they [families] do have some control;
being willing to try new things, especially when clients make the suggestions; involving the parent
in the removal of the children, a strategy that is empowering to parents and less traumatic for the
children; demonstrating a commitment to invest time with families on the "front-end" to build
rapport, facilitate client understanding of the child welfare system, and optimize chances for
maintaining or reunifying the family; working with families to identify relative placements when
out of home placements are required and coaching relatives through the application process; and
lastly, demonstrating a willingness to advocate for the child and family with the court and other
service delivery systems.
Contextual Factors:
Recommendations to Improve Practices for Children and Families of Color
Study participants described five overall recommendations to improve practices for children
and families of color as they relate to contextual factors, including I) reducing caseloads, increasing
support and streamlining service delivery, 2) maintaining and expanding multifaceted approaches to
providing culturally competent practice, 3) increasing linguistic capacity of systems and services for
families, 4) fostering an organizational culture that is strength-based and community based, 5)
enhancing collaboration, and 6) training social workers. Specifically,
Reduce Caseloads, Increase Support and Streamline Service Delivery
82
•
•
•
•
•
•
•
•
•
•
Continue efforts to reduce social worker caseloads. Reduction in caseloads was uniformly
identified as key to providing quality services to clients, particularly children and families of
color that may have complex needs and that may initially have less trust and/or less familiarity
with child welfare and court systems. "Social workers need time to do their job and to help
clients access culturally appropriate services and programs."
Provide additional Social Worker I support for in-home assessment/support, follow up with
families, and overall case support.
Provide dedicated staff, and explore other options, to expedite requirements to qualify relative
placements. For example, policies requiring relative placement standards similar to foster
families might be reevaluated (e.g., illustrations of this as a problem included an example of two
young girls were denied relative placement because they shared a queen size bed rather than
having separate beds). In addition, develop strategies for overcoming barriers to relative
placement, such as an expedited process for Director's Exemption in the case of a distant
criminal history of a family member in the prospective home.
Examine opportunities to facilitate quicker access for social workers to criminal background
checks.
Develop procedure or protocols to facilitate communication between units, particularly in
relation to transferring ER cases to voluntary family maintenance informal supervision.
Examine options for maximizing the skills of social workers with specific cultural or language
expertise. For example, social workers in ER and DI stated that support across units that
involved overtime was recently prohibited. However, social workers suggested that this practice
failed to reduce the total amount of overtime expended by social workers while limiting
flexibility in obtaining culturally and linguistically skilled consultation and assistance.
Develop centralized, and updated resources lists for a wide range of services, including service
providers with expertise in serving diverse cultural and language groups. This practice was
identified as key to avoiding duplication of effort and ensuring access to quality services for
clients. "Sometimes we make a lot of phone calls just to find out what someone two aisles from
you already knew." In addition, update Court Services Handbook, which was perceived by
several social workers to be out of date.
Examine opportunities to provide flexibility in office hours, alternate social workers to cover
cases when social workers are on vacation, or other mechanisms to reduce potential burnout.
Provide social workers access to adequate transportation (e.g., "better cars") for travel related to
working with families.
Develop protocols that reduce paperwork so "there is more time to do social work." For
example, interim court reports could be eliminated or reduced, templates for psychosocial
assessments or other guidelines for court reports could be provided, and social workers could
"get assistance especially with filing."
Maintain and Expand Multifaceted Approach to Providing Culturally Competent Practice
•
•
Continue practice of attempting to provide a cultural match for clients as a strategy for
facilitating rapport and, in some cases, minimizing cultural or class bias.
At the same time, maintain recognition that 1) some clients may prefer to see a Social worker
who is not of their background, 2) many clients value "being treated with respect" above being
assigned a social worker of the same background, and 3) the diversity of the population in Santa
Clara County may not always allow for "matching" clients and social workers in relation to
race/ethnicity or other demographics. Ensure training and support for all social workers in
83
•
•
cultural competence and addressing the needs of different population groups among all social
workers
Continue working to ensure diverse staff in units throughout the agency.
Continue using the family resource centers, including provision of services for mandated clients.
Increase Linguistic Capacity of Systems and Services for Families
•
•
•
•
•
Facilitate development of services specifically designed for families whose primary language is
not English in key areas including mental health, domestic violence, substance abuse,
counseling services, parenting without violence classes, and other services.
Continue to hire bilingual/bicultural staff. Need for greater linguistic capacity was reported for
a number of different language groups including for Spanish speaking, Vietnamese, and Filipino
clients.
Minimize need for use of translators. When needed, provide high quality translators who are
able to communicate concepts related to the law, the social work system, timelines, and court
· processes.
Provide forms and written materials in multiple languages.
Examine opportunities for establishing guidelines for contracting or collaborating agencies in
relation to linguistic services, such as procedures disallowing use of children as translators for
monolingual speaking parents.
Foster an Organizational Culture that is Strength-Based and Community-Based
•
•
•
•
Continue to identify mechanisms, such as the Family-to-Family Initiative, to build community
involvement in decision-making.
Communicate a commitment to assuring the least restrictive placement for children from
management, through supervision, and to line staff. Provide training and support for social
workers to advocate for clients in court and other settings.
Examine strategies to minimize what study participants described as internal disincentives to
ensuring the least restrictive placement for children including "County Council whose role it is
to protect the agency sometimes telling social workers to change recommendations" and the
greater paperwork and advocacy required for less restrictive options.
Ensure client access to information about complaint procedures and access to the Ombudsman.
Enhance Collaboration
•
•
Continue to develop collaborative case planning, linkages, information sharing and policy
development with other service delivery systems, including schools, service providers,
probation, CalWORKS, housing, and other systems. Collaborative planning and problem
solving was perceived as particularly crucial in an environment of sever fiscal restraints.
Develop protocols to facilitate access to mandated services for clients in the context of federal
and state timelines, such as formal agreements and protocols to expedite access to substance
abuse treatment.
Train Social Workers
84
•
•
•
•
•
•
•
•
Require cultural competence training for all new staff and mandate continuing education for all
social workers. Include, as part of the training, opportunities for self-reflection and examination
of personal bias related to culture of other issues frequently confronted in families, such as
substance abuse. In these areas, provide training that is high quality, dynamic, and delivered by
skilled trainers who are knowledgeable about both their subject area and child welfare.
Provide education within the agency about different cultural norms in parenting and family
strengths.
Provide training on specific issues that impact all families and that may differentially impact
children and families of color including policies and practices for addressing drug exposed
infants, substance abuse, domestic violence, sexual abuse, and other topics. In addition, provide
training on procedures and practical considerations for working with diverse families, such as
training on guardianship.
Provide training on immigration law and practical strategies for working with families, such as
identification of immigration status of the child at an early stage and working with international
relative placement.
Training on these topic areas should integrate consideration of cultural differences and
evidence-based practices for different populations.
Provide training for all staff including clerical staff and eligibility workers.
Facilitate forums for social workers to share resources and effective practice strategies.
Create opportunities for social worker access to training. "We can't go to trainings because we
have too much else to do-so we're denied professional development opportunities."
Summary of Key Findings and Implications
Study participants described numerous practices and contextual factors that have either a
negative or positive impact on children and families ofcolor in Santa Clara County's child welfare
system. There is some evidence in the research literature that supports certain best practices
mentioned by study participants. For instance, general research supports the effectiveness of inhome services (see literature review), however these services appear to be most effective when they
are intensive, long-term and delivered by health or social service professionals and when the
workers delivering these services receive a high level of supervision (Barth, 1991; McGuigan et al.
2003; Olds et al., 1997). In addition, some research has suggested that parenting classes for child
welfare parents are more effective when they pair a didactic approach with a more experiential
approach, such as parent and child interaction (Carlo, 1993).
Participants also identified family conferencing as an effective practice. These study
findings are consistent with limited research in this area that suggests that inclusive practices, in
which parents are engaged in placement and other decisions, may result in more placement stability
for children, although possible differential effects for diverse populations has not been researched
(Palmer, 1996). Participants generally expressed optimism about the adoption of team decision
making (l'DM) as a vehicle for improving decision-making with diverse families, although some
social workers expressed concern about time required for implementation. Research does suggest
that TOM can be labor intensive and can be difficult to implement effectively within the context of
the child welfare system (Sieppert et al., 2000). Finally, participants noted that both formal and
informal collaborative relationships with other services delivery systems, such as domestic violence
and substance abuse treatment systems, are critical for successful interventions with children and
families. Furthermore, participants recognized the importance of training and cross-training to
85
enhance the capacity ofhelping professionals in child welfare and other systems to better address
the needs ofchildren and families of color.
Interestingly, most practices and contextual factors mentioned by social workers had both
positive and negative aspects. Many of the practices and contextual that negatively impact children
and families of color were actually barriers to successful implementation of the practices that social
workers felt have a positive impact on children and families of color. Although practices such as
orientation ofnew clients; family and team decision-making; preventive services; substance abuse
services; cultural competency; collaboration with other agencies and systems; a strength-based
approach; and the use of relative placements were described as having a positive impact on
children and families ofcolor, certain negative practices and contexts impeded their effective
implementation. For instance, having to screen out large numbers of inappropriate referrals;
inconsistency in decision-making practices; a shortage ofservices, particularly substance abuse
and preventive services; lack ofclient access to services; difficult protocols for placing children
with kin; gaps in cultural competency; time limits; and agency-level factors such as heavy
caseloads, staff shortages, substantial amounts ofpaperwork, lack ofaccess to information about
resources; and confusion about the agency's overall mission and key policies all impeded
implementation ofthe best practices. These findings suggest that in addition to building on current
positive practices for children and families of color, CWS stakeholders should also actively work
both internally and in collaboration with partners to reduce barriers to these best practices.
Indeed, in order to address these barriers to effective implementation of best practices, study
participants mentioned numerous recommendations that centered around certain key themes. In
general, there was an emphasis on improving decision-making through increased accountability,
training and cross-training to reduce bias, and increased use of group decision-making or family
involvement in decision-making. Study participants also stressed a need to improve and expand
prevention, diversion and concrete services, as well as an overall need to expand culturally and
linguistically competent services, and develop ways to improve availability and access to services.
Once a case is opened, social workers also felt it valuable to provide an orientation to clients so that
they are more aware of the CWS and court processes and can better navigate the system.
Recommendations related to organizational factors were also noted, including reducing caseloads;
increasing support and streamlining service delivery; clarification ofthe mission ofDFCS, as well
as key child welfare policies; fostering an organizational culture that is strength-based and
community-based, and more overall training for social workers.
Liniitations and Conclusions
Qualitative data, such as focus groups and interviews, provide rich insights into practices
that may positively or negatively impact children and families of color. In any such endeavor, there
are possible biases that may impact the results. For example, given the likelihood of time conflicts
and emergencies, we invited more social workers to participate in the focus groups than we
expected to attend. It is not possible to determine if other factors influenced participation. Workers
with a particular interest in this topic may have made a special effort to attend. In addition, the
perspectives and opinions of the social workers, supervisors and families who participated in the
study may not be entirely reflective of the populations they represented. At the same time,
participants were representative of different units and facets of the child welfare system, different
race/ethnic groups, and varied levels of experience. The findings of this study affirm the value of
many efforts that are already in place, such as culturally specific ER response units and family
86
resource centers, and other that are in the process of implementation including efforts to reduce
caseloads, initiation ofteam decision making, and participation in the Family-to-Family initiative.
The findings of the study also point to opportunities to further strengthen services for children and
families ofcolor. Recommendations related to training, institutionalizing group orientation for
clients, and other suggestions from study participants could be used to inform practice and planning.
It would be beneficial for a team ofmanagers, line staff, family representatives, and community
members to review these findings to identify and prioritize possible practices for adoption based on
their feasibility and utility for children and families ofcolor.
87
Comparison of the Main County Offices and South County Offices
In an effort to better understand the influence of contextual factors on child welfare practices
and outcomes for children and families of color, a comparison of Santa Clara County's Main
Offices and South County Offices was conducted. This comparison included both quantitative and
qualitative analyses. Quantitative data from CWS/CMS were used to examine differences in case
characteristics between the Main County Offices and South County Offices. A qualitative
comparison of focus group data from the Offices and South County Offices was also conducted to
discern how workers perceive practices in these two geographical locations. These analyses help
shed light on the potential impact of agency and community context on case characteristics, child
welfare practices and outcomes for children and families of color in Santa Clara County's CWS.
Quantitative Methods
In Phase 2, the CWRT continued its investigation of the disproportionate representation of
children of color in the child welfare system by analyzing a sample of cases closed during an 18month period (January 2000 through June 2001). Key case characteristics were analyzed, including
ethnic comparisons. Please see the Phase 2 report for methodology details and case characteristics
by ethnicity (Hines et al., 2002).
One of the recommendations that emerged from Phase 2 of the study was to explore possible
differences in case characteristics and practices between the Main County Offices and the South
County Offices. In Phase 3, we used the basic characteristics from the 1720 cases analyzed in Phase
2 while adding the Santa Clara County location information provided via the Child Welfare System
/ Computer Management System (CWS/CMS).
Analyses
A series of exploratory, bivariate analyses were conducted with county location as the
distinguishing variable, Main County (MC) versus South County (SC). The purpose of this initial
set of analyses was to identify differences in the CWS related to county location in order to help
describe variations not only by ethnicity, but by other case and service characteristics. The
following sections will describe (1) the proportion of cases with each county location, (2) general
characteristics by county location, and (3) significant differences related to county location.
Definition of Terms
County Location - Two overall geographical areas within Santa Clara County designated as either
Main County Offices (centered primarily in the Northern part of the County, including the main
office on Julian Street in San Jose) or South County Offices (based in Gilroy).
Quantitative Results
County Location
Of the 1720 cases, 1633 (94.9%) had valid county location information with 87 missing
(5.1%). Of the 1633 valid cases, 1487 (91.1%) were served in MC, 50 in SC (3.1%) and 96 (5.9%)
88
in both county locations. Please see Table 15.
Table 15: Proportion of Cases by County Locationa
Total
Main (MC)
8
1633
1487
8
(91.1%)
South (SC)
Both
96
(5.9%)
50
(3.1%)
Based on 1633 valid cases, excluding 87 cases with missing location
For the primary ethnic comparison, we retained three county location designations, MC, SC,
and Both (cases served in MC and SC). However, given the range in the length of time children
designated as Both were served in MC and SC, and the focus on comparing the two stated
geographical areas, subsequent analyses excluded this third category.
County Location and Child Characteristics
Child's Ethnicity
There was a significant relationship between the child's ethnicity and county location. As
noted in Phase 1 and 2, Latinos compose the largest ethnic group served in the county's CWS.
When examined by location, those served in MC were more likely to be Latinos (44. 7%, 661 of
1480 MC cases). This proportion of Latinos was higher in SC (68.0%, 34 of 50 SC cases) and Both
(62.l %, 59 of95 Both cases). African American, Asian American/Pacific Islander, and Other
children's cases were least likely located in SC and Both locations. Please see Table 16.
T able 16 Eh
. Group B;y C ounty Locat1on
t me
Total
Child's Ethnicitya
South
Both
1625
1480
50
95
•
N size
African American
•
177
(10.9%)
170
(11.5%)
1
(2.0%)
6
(6.3%)
White
•
476
(29.3%)
434
(29.3%)
29
(30.5%)
Latino
754
(46.4%)
661
(44.7%)
13
(26.0%)
34
(68.0%)
Asian American
/Pacific Islander
156
(9.6%)
155
(10.5%)
1
(2.0%)
0
(0.0%)
Other ethnicity
62
(3.8%)
60
(4.1 %)
1
(2.0%)
1
(1.1%)
•
•
•
a
Main
59
(62.1%)
Based on 1625 cases with valid MC and SC county locations and county location information
Unfortunately, the South County Office S1Jbsample was too small to allow more detailed analyses by
ethnicity. However, we continued to pursue our comparison of the two primary county locations by
other key variables.
89
Child Demographics
There was not a significant relationship between gender and county location. Within MC,
52.2% was female (776 of 1486 cases) compared to SC where 50.0% was female (25 of 50 cases).
Within SC, 47.8% was male (710 of 1486 cases) compared to SC where 50.0% was male (25 of50
cases).
There was not a significant relationship between age and county location. The average age
of the MC sample was 5.95 years (sd = 5.16), while the average age of the SC sample was 6.57
years (sd = 4.19).
There was not a significant relationship between child's language and county location. The
majority of the MC and SC samples spoke English. However, in SC, Spanish was the only other
language other than English spoken. Please see Table 17.
T able 17 Ch.ld
1 D emograp h.1cs
Gender
N Size
• Female
Total
Main
South
1536
1486
50
801
(52.1%)
776
(52.2%)
25
(50.0%)
735
(47.9%)
710
(47.8%)
25
(50.0%)
5.97
(5.13)
5.95
(5.16)
6.57
(4.19)
8
•
•
Male
Average age in years (sdt
Child's Languagec
•
•
N Size
1524
1474
50
English
•
•
1308
(85.8%)
1262
(85.6%)
46
(92.0%)
Asian American/
Pacific Islander
62
(4.1%)
62
(4.2%)
0
(0.0%)
Spanish
•
144
(9.4%)
140
(9.5%)
4
(8.0%)
Other
IO
10
(0.7%)
0
(0.0%)
•
(0.7%)
a Based on 1536 cases with valid MC and SC county location and gender information
b Based on 1537 cases with valid MC and SC county location and age information
c Based
on 1524 cases with valid MC and SC county location and language information.
90
County Location and System-Related Factors
Voluntary Status and Reason for Removal
There was a significant relationship between voluntary status and county location. The
proportion of cases in voluntary services was higher in SC at 88.0% (44 of 50 cases) compared to
41 % in MC (600 of 1462cases).
There was not a significant relationship between the reason for removal and county location.
Please see Table 18.
T able 18 Vo1untary status an dReason fior Remova
Total
8
Voluntary Status
1512
• N Size
Main
South
1462
50
•
Voluntary
•
644
(42.6%)
600
(41.0%)
44
(88.0%)
Not Voluntary
868
(57.4%)
862
(59.0%)
6
(12.0%)
1206
1196
10
Reason for Removal b
•
N Size
General Neglect
•
161
(13.3%)
158
(13.2%)
3
(30.0%)
Severe Neglect
•
172
(14.3%)
172
(14.4%)
0
(0.0%)
Physical Abuse
•
300
(24.9%)
297
(24.8%)
3
(30.0%)
Sexual Abuse
•
73
(6.1%)
73
(6.1%)
0
(0.0%)
Emotional Abuse
•
40
(3.3%)
40
(3.3%)
0
(0.0%)
Other Maltreatment
460
(38.1%)
456
(38.1%)
4
(40.0%)
•
a Based on 1512 cases with valid MC and SC county location and voluntary service information
bBased on 1206 cases with valid MC and SC county location and removal information
91
Initial Out-of-Home Placement Facility Type
There was not a significant relationship between initial out-of-home placement facility type
and county location. Please see Table 19.
Table 19: Initial Out-of-Home Placement Facility Type
Total
Main
South
Initial Out of Home Placement
Facility Type8
• N Size
1206
1196
10
•
Foster Family Home
•
258
(21.6%)
258
(21.6%)
0
(0.0%)
Group Home
•
66
(5.5%)
66
(5.5%)
0
(0.0%)
Children's Shelter/
359
(29.8%)
353
(29.5%)
6
(60.0%)
Relative Home
•
343
(28.4%)
341
(28.5%)
2
(20.0%)
Foster Family Agency
•
124
(10.3%)
123
(10.3%)
1
(10.0%)
Guardian Home/Court
Specified Home
56
(4.6%)
55
(4.6%)
I
•
a
Receiving Home/ Non-EA
/AFDC
(10.0%)
Based on 1206 cases with valid MC and SC county location, and placement infonnation
Number of Times removed from Family and
Number of Unique Placement Homes in Current Episode
There was not a significant relationship between the number of times a child was removed
from their family in current episode and county location. On average, children in the sample were
removed 1.35 times (sd = 0.70).
There was not a significant relationship between the number of unique placement homes in
current episode and county location. The sample had an average of3.56 (sd = 3.86) unique
placement homes in current episode. Although MC cases had a higher average of placements (3.57,
sd = 3.88) compared to SC (1.90, sd = 1.10), this difference did not reach statistical significance,
most likely due to the small number of SC cases available for comparison. Please see Table 20.
92
Table 20: Number ofTimes removed from Family and Number of Unique Placement Homes in
Current E.
;p1sode
Average Number of Times
Removed from Family in
South
Total
1.35
(0.70)
Main
1.35
(0.70)
(0.32)
3.56
(3.86)
3.57
(3.88)
1.90
(1.10)
LIO
Current Episode (sd?
Average Number of Unique
Placement Homes in Current
Episode (sdl
a Based 1228 cases with valid MC and SC county location and removal information
b
Based on 1228 cases with valid MC and SC county location and placement information
Average Stay (in days) per Placement Facility, Number of Placements, and Total Length of
Time in Out of Home placement in Current Episode
There was not a significant relationship between the average stay (in days) per placement
facility in current episode and county location. The average number of days a child stayed in one
placement in their current episode was 244.06 (sd = 480.43). Although MC cases had a higher
average of245.73 days in placement (sd = 482.03) compared to 41.54 days (sd = 44.71) in SC, this
difference did not reach statistical significance, again most likely due to the small number of SC
cases available for comparison.
There was not a significant relationship between the number of placements in episode and
county location. On average, children in the sample had 2.84 placements (sd = 2.95). Although MC
cases had a higher average of2.85 placements (sd = 2.95) compared to 1.80 (sd = 1.32) for SC, this
difference did not reach statistical significance, again most likely due to the small number of SC
cases available for comparison.
There was not a significant relationship between the total length of time (in months) in outof-home placement in current episode and county location. On average, children in the sample spent
13.16 months (sd = 20.40) in out-of-home placement. Although MC cases had a higher average of
13.24 months in placement (sd = 20.47) compared to 3.28 months (sd = 4.24) for SC, this difference
did not reach statistical significance, again most likely due to the small number of SC cases
available for comparison. Please see Table 21.
93
Table 21: Average Stay (in days) per Placement Facility, Number of Placements, and Total Length
. 0 Ut 0 fH ome p1acement .m Current E,p1so
. de
0 fT1mem
Average stay ( in days) per
Placement Facility Type in
Current Episode (sd)8
Number of placements in
Current Episode (sd/
Total length of time (in months)
in Out of Home Placement in
Total
Main
South
244.06
(480.43)
245.73
(482.03)
41.54
(44.71)
2.84
(2.95)
2.85
(2.95)
1.80
(1.32)
13.16
(20.40)
13.24
3.28
(4.24)
(20.47)
Current Episode (sd/
a Based on 1228 cases with valid MC and SC county locations and placement information
b
Based on 1206 cases with valid MC and SC county locations and placement information
on 1200 cases with valid MC and SC county locations and time information
c Based
Number of Episodes, Total Length of Case (in years), and Number of Workers Assigned to
Case over Time in the Current Episode
There was not a significant relationship between the number of episodes and county
location. Children in the sample had an average of 1.19 (sd = 0.47) episodes.
There was a significant relationship between the total length of the case (in years) and
county location. On average, children in the sample had a case length of 1.87 years (sd = 2.78). MC
cases lasted longer with an average of 1.93 years (sd = 2.80) compared to 0.18 years (sd = 0.39) for
SC.
There was a significant relationship between the total number ofworkers assigned to case
over time in current episode and county location. Children in the sample had an average of 4.91
workers (sd = 2.75). MC cases were assigned a higher average of 5.03 different workers (sd = 2.72)
compared to 1.48 workers (sd = 0.84) for SC. Please see Table 22.
Table 22: Number of Episodes, Total Length of Case (in years), and Number of Workers Assigned
t0 Case over T1mem
. th e CurrentE,PlSO
. de
South
Total
Main
1.16
1.19
1.19
Average Number of Episodes (sd)8
(0.48)
(0.37)
(0.47)
Average Total Length of Case (in
years) (sdl
1.87
(2.78)
1.93
(2.80)
0.18
(0.39)
Average Number of Workers
Assigned to Case over Time in the
Current Episode (sdt
4.91
(2.75)
5.03
(2.72)
1.48
(0.84)
Based on 1512 cases with valid MC and SC county locations and episode information
Based on 1512 cases with valid MC and SC county locations and case time information
c Based on 1532 cases with valid MC and SC county locations and worker information
a
b
94
Last Out-of-Home Placement Type
There was not a significant relationship between last placement type and county location.
Please see Table 23.
T a ble 23 Last Out-o f.H
- ome Placement Tvoe, and Age at Time of Case Closure
Total
Main
Last Out of Home
Placement Type8
• N Size
South
1228
1218
10
•
Foster Family Home
•
260
(21.2%)
260
(21.3%)
0
(0.0%)
Group Home
•
77
(6.3%)
77
(6.3%)
0
(0.0%)
Children's Shelter
•
354
(28.8%)
348
(28.6%)
6
(60.0%)
Relative Home
•
346
(28.2%)
344
(28.2%)
2
(20.0%)
Foster Family Agency
•
133
(10.8%)
132
(10.8%)
1
(10.0%)
Guardian Home/
Court Specified Home
58
(4.7%)
57
(4.7%)
1
(10.0%)
a Based on 1228 cases with.valid MC and SC county locations and placement information
Age at Time of Case Closure and Service Type at Case Closure
There was not a significant relationship between age at time of case closure for current
episode and county location. On average, children in the sample were 8.19 years-of-age (sd = 5. 72)
at the time of case closure. Although MC cases closed with an older child's average age of 8.23
years (sd = 5.76) compared to 7.15 years (sd = 4.22) for SC children, this difference did not reach
statistical significance.
There was a significant relationship between service type at closure ofcase and county
location. Overall, the majority of cases ended in family maintenance (59.9%, 920 of 1537 cases).
However, SC had a higher proportion of cases in family maintenance (92.0%, 46 of 50 cases)
compared to MC (58.8%, 874 of 1487 cases). Please see Table 24.
95
. T ype at Case Closure
T a ble 24 A.ge at T.1meofC ase Closure an dS erv1ce
Total
Main
Age at Time of Case
Closure for the Current
8.19
South
(5.72)
8.23
(5.76)
7.15
(4.22)
1537
1487
50
Episode (sd)8
Service Type at Case
Closureb
•
•
N Size
Emergency Response
•
39
(2.5%)
29
(2.6%)
0
(0.0%)
Family Maintenance
920
46
•
874
(58.8%)
Family Reunification
•
Permanent Placement
(59.9%)
64
(92.0%)
67
(4.4%)
(4.3%)
3
(6.0%)
511
(33.2%)
510
(34.3%)
1
(2.0%)
a Based on 1482 cases with valid MC and SC county locations and child's age at case closure information
b
Based on 1537 cases with valid MC and SC county locations. and service type information
Case Closure Type
There was a significant relationship between case closure type and county location. The
most common case closure type was court ordered termination and other (32.1 %, 482 of 1502
cases), followed by family maintenance (31.6%, 475 of 1502 cases). However, 62% (31 of 50
cases) of SC closed with family maintenance compared to 30.6% (444 of 1452 cases) in MC.
Please see Table 25.
96
T a ble 25 C ase Closure T ype
Case Closure Typea
Main
1452
South
50
•
•
N Size
Adoption
•
246
(16.4%)
246
(16.9%)
0
(0.0%)
Emancipation
93
(6.2%)
93
(6.4%)
0
(0.0%)
475
(31.6%)
444
(30.6%)
31
(62.0%)
99
(6.6%)
97
(6.7%)
2
(4.0%)
Reunified with Parent
or Guardian, Court or
non-Court Specified
61
(4.1%)
59
(4.1%)
2
(4.0%)
Incarceration,
Runaway or Medical
services
46
(3.1%)
45
(3.1%)
1
(2.0%)
Court Ordered
Termination and Other
482
(32.1%)
468
(32.2%)
14
(28.0%)
•
•
•
•
•
a
Total
1502
Family Stabilized
(FM)
Guardianship
Established or
Placement with
Relative
Based on 1502 cases with valid MC and SC county locations and case closure information
97
Qualitative Comparison of the Main County Offices and South County Offices
In addition to the quantitative comparison of the Main County Offices and South County
Offices, qualitative focus group data were collected to obtain information on how workers perceive
practices in the two geographical areas, and how these practices may differ.
Qualitative Methods
Prospective participants for the Social Worker focus groups were randomly selected from a
list of workers with title of Social Worker II or Social Worker III in different units/divisions. The
random selection and recruitment process is described in detail in the section describing qualitative
methods (please see page 61 for a full description of focus group qualitative methods). As described
above, a total of 67 Social Workers participated in the 13 focus group interviews. Of this number,
13 Social Workers from various units in South County Offices participated in three South County
focus groups. All the focus group participants in the South County sample were female. Of these, 2
(15.4%) identified as African Ameri_can, 1 identified as Asian American/Pacific Islander (7.7%), 4
identified as Latina (26.9%), and 6 identified as white/Caucasian (32.8%). A majority of the
participants (n=12, 92.3%) had masters level education, primarily MSW degrees (Please see
Attachment 2 for additional information about focus group demographics)
Procedures
Procedures for all focus groups are described in detail in the Qualitative Methods section of
this report. Procedures for all focus groups, including South County Office focus groups were
identical. As described above, ten focus groups were conducted at the main administrative office in
downtown San Jose and three were conducted at Gilroy Family Resource Center in South County.
Analysis
The process for analysis of the South County Offices data compared to other focus groups
was parallel to that described above in the Qualitative Methods section. The data for the qualitative
portion of this study consist of the transcribed notes taken during each of the focus groups and
individual interviews. Qualitative methods of data analysis were employed to analyze the focus
· group and interview data. First, major themes were summarized for each of the choice points or
units in the Social Worker focus group and supervisor interview data. A separate analysis of the
qualitative data from South County Offices was conducted to identify themes that may have been
different or more prominent in these series of 3 focus groups compared to the other 10 social worker
focus groups. The purpose of this additional analysis of the qualitative data was to obtain insights
and supplement quantitative analysis of possible differences between South County Offices and
Main County Offices within DFCS. In addition to gathering information about practices that may
differ between the Main County Offices and the South County Offices, findings from the qualitative
data also provide insights that may help explain quantitative differences in the relationship between
location and other variables, such as differences in the proportion of voluntary family maintenance
cases.
98
Qualitative Results
Many of the contextual and practice issues described by social workers and supervisors in
South County Offices paralleled comments by study participants in other regions of Santa Clara
County, such as concerns related to client poverty, housing shortages, and insufficient substance
abuse and other treatment resources, particularly for non-English speaking language groups. At the
same time, a few themes emerged that were either stronger or unique to South County Offices.
These are described below.
Context of Service Provision in South County Offices
Demographics of South County
According to study participants, South County has a large Latino community that is strongly
represented among child welfare clients. In relation to these demographics, participants emphasized
the need for bilingual and bicultural child welfare professionals. They also reported a paucity of
culturally and language specific counseling services for monolingual Spanish speaking participants.
In addition, access to counseling and services for undocumented families in the child welfare
system were perceived as a problem. Participants noted that one strength of the Family Resource
Center is the bilingual and bicultural social work staff. At the same time, participants saw a need to
expand capacity to work with other communities. "Our resource center is reflective of the majority
community ofLatinos, but some clients may not feel represented," explained one South County
focus group member. Specifically, several social workers suggested that the current and evolving
demographic changes in the region called for greater representation of African American and Asian
American/Pacific Islander social workers.
Small County Dynamic
South County Offices were perceived as different from "the main" office in part because
dynamics related to the smaller size of the region. Participants described feeling more connected to
the community, other service providers and clients because of the "a small-town atmosphere."
Participants noted that, "The community is smaller and there is more interaction within the
community and with individuals from other agencies; this tends to make people work with clients
differently." For example, one social worker commented that she often sees clients and colleagues
at the grocery store and others mentioned that clients freely come into the child welfare offices and
Family Resource Centers asking for assistance.
Furthermore, study participants in South County consistently reported a high level of both
formal and informal consultation with one another. In addition to relying on other workers and
supervisors to assess families and develop case plans, social workers working in close physical
proximity are aware of one another's cases and routinely exchange informal support and advice
about working with clients and accessing resources. Some participants noted that they might be
"less likely to connect" in larger facilities or offices "built like a maze." The absence of security for
entering the agency building "changes the atmosphere because clients can just walk in, there is an
open door policy; this changes agency attitude in working with clients and removes barriers for
clients."
99
Culture of Commitment to Maintaining and Reunifying Families
South County study participants described a shared philosophy that the vast majority of
children referred to the child welfare system can and should be with their birth parents, whether it
be through family maintenance or through family reunification. Specific values and norms
embedded in the philosophy and practice of the region include emphasis on using family strengths,
including families in the process of making decisions, and aggressively seeking alternatives to
removal. Study participants underscored the importance of this approach in working with children
and families of color. Although these principles are valued throughout the child welfare system,
participants in South County stated that these principles were infused throughout the system and
were used consistently to guide practice and social workers' accountability in assessment and
decision-making. For example, social workers described substantial involvement of supervisors and
coworkers when removal of children into protective custody is required. Study participant
comments suggested that the smaller size of the agency helps to solidify a sense of shared
philosophy and reinforce practice norms.
When asked to articulate how these values and norms are sustained and passed on, workers
cited several factors. First, study participants emphasized the importance of consistency in
messages, policies, and direction from the Program Manager and supervisors. Second, social
workers suggested that these expectations related to seeking the least restrictive placement for
children are communicated beginning with the initial hiring and orientation process through
everyday practice. "People who get hired have the same philosophy as South County ... and
supervisors ingrain values into workers," noted participants. Third, social workers stated that it was
normative for workers to obtain both formal and informal support when they are making decisions.
"Workers do a lot of staffing and hash things out; we don't make decisions alone," noted one
participant. "By using co-workers and supervisors in decision-making, you gain another
perspective," observed another. Integrating the perspectives of other workers in decision-making
was identified as crucial to mediating potential bias or misunderstanding related to cultural or
socioeconomic class differences. Finally, social workers suggested that they received substantial
support and consultation from supervisors in decision-making and that supervisor values and
practices were perceived as critical to the practices of line staff. For example, supervisor
involvement in decision-making was described as a positive practice, particularly in relation to
taking children into protective custody and out of home placement. "My supervisor was on the
phone the whole time I had to remove a child," was offered as an example from one social worker.
Such oversight was perceived as a sign of support rather than solely a mechanism to assure social
worker accountability.
Practices in Service Provision in South County Offices
Vertical Case Management
Focus group participants and supervisor interviewees named the vertical case management
model, in which social workers assigned to a case work with the family throughout their case rather
than transferring cases to a different unit, as an effective practice with clients, particularly with
children and families of color. Some study participants suggested that the vertical case management
model works well in a small community. "It works because it puts social workers in the community
with the family, making social workers accountable .. .it forces social workers to have to deal with
100
that family not just in a [case] process sense but in a real social work sense of developing
relationships," summarized one supervisor interviewee. Social Workers suggested that maintaining
continuity through a specific case enhanced trust with clients and created a greater sense of
accountability. The opportunity to develop and sustain rapport with specific clients was deemed to
be particularly important for marginalized groups, including children and families of color. Families
that return to the system are also generally reassigned to the same social worker. "In this case a new
worker may not have rapport or know family strengths ... A worker who has worked with a family in
the past knows them well enough to call them on their stuff." At the same time, respondents also
described vertical case management as work intensive and requiring additional support to ensure its
success, such as the help of Social Worker I staff.
Study participants in the focus groups that took place at the main administrative office in
San Jose location appeared to be familiar with the vertical case management model. Some
mentioned that it might be a possibility as a positive practice for children and families of color in
other parts of the county while others stressed the importance of maintaining specialized units with
staff who have expertise and an opportunity to stay current with changing laws, policies and
practices in their area.
Orientation of Clients
Social Workers in South County frequently refer parents to a client orientation that helps
parents understand the child welfare system and policies related to timelines for reunification. The
orientation is offered in English and in Spanish. It is offered at different times and provides 6 hours
of information over three sessions. Three areas are addressed in the orientation: timelines and facets
of the child welfare system (ER, DI, FR, FM, adoption), how to work with the social worker, and
resources for children and families. The orientation is not mandatory and is not offered at all sites.
One focus group participant examined outcomes (identified as either reunification or closure of
case) and found that the orientation appeared to be particularly helpful for families in the Spanish
speaking orientation. In sum, the orientation for clients was recognized as an important resource for
families, particularly families of color, who might feel confused or intimidated by the child welfare
system. At the same time, workers appeared to feel that there were opportunities to further
strengthen, institutionalize, and evaluate this practice.
Case Conferencing
Study participants stressed as a best practice the formal and informal process of meeting
with co-workers and supervisors to review cases and participate in the decision-making process.
Accountability and support in decision-making was viewed as an important factor in minimizing
potential cultural or other bias about clients and in maximizing creative thinking about how best to
work with children and families. Inclusion of families in decision-making was also noted as a
positive practice that was considered particularly important for children and families from cultures
that place a high value on extended family, including both biological and fictive kin. "If you have a
voluntary family maintenance case, you have a family conference," observed one social worker.
101
Summary of Key Findings and Implications:
Quantitative and Qualitative Comparison
of Main County Offices and South County Offices
Unfortunately, the South County location subsample was too small to allow analyses by
ethnic group. However, some key case characteristics are significantly related to county location.
Specifically, in South County Offices, a higher proportion ofcases are in voluntary services, cases
are shorter in duration, fewer workers are assigned during the course of the case, and a higher
proportion of cases are in family maintenance at case closure. These findings are also statistically
noteworthy given the small subsample size available from SC. There were also other differences in
case characteristics by county location that did not attain statistical significance but are noteworthy
for further investigation. Children served in the South County Offices appear to have a fewer
number of unique placement homes in the current episode, have a shorter average stay per
placement, a shorter length of time in out-ofhome placement, and are younger at time ofcase
closure.
These findings imply that there is a difference in the style and quality of services provided in
the South County Office location. Qualitative findings are congruent with the quantitative data and
provide some insight into the dynamics that may contribute to this difference. The higher number of
cases in voluntary services, the larger proportion ofcases in family maintenance at case closure,
and the shorter duration ofcases is consistent with what South County study participants described
as a culture of commitment to maintaining and reunifying families. This commitment was described
as integral to the philosophy of managers and social workers and reflected in the expectations
communicated to social workers by most supervisors and by peers. The practice ofproviding an
orientation to the child welfare system, time limits, the courts, and how to work with social workers
may also contribute to the differences found in the quantitative analysis. Study participants
frequently described South County as similar to a small county or rural area. This geographic
difference, and the resulting sense of "connection to the community, " was identified as a factor in
the greater sense of shared philosophy and emphasis on prevention ofout-of-home placements.
Other studies suggest that region may influence practice. For example, one study (Drake, 1996) also
found that rural areas are more likely to offer preventive services than are urban areas. The finding
that fewer workers are assigned during the course ofthe case is also consistent with descriptions of
the vertical case management model. This model calls for social workers to carry the same case
from case opening to case closure. assigned to a case works with that family from the initiation to
closure of the case.
Limitations and Recommendations
Only about 3% of cases were designated from SC, and the third location designation of
"both" was dropped. A key informant from South County explained that many of the cases
categorized as "both" may have been opened by ER workers from the Main County Office but
carried in South County prior to assignment of South County staff to a specific region of the county.
Similarities in demographics between the South County Office cases and the "both" categories
would appear to be congruent with this hypothesis. However, since this explanation could not be
verified in the data that was available, the cases designated as both were not combined with other
South County data. Future research should include a larger subsample from SC and add methods to
102
understand the reasons behind these location differences. Also, a larger South County Office
location sample may enable more detailed comparisons by ethnicity.
Some practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County Offices and adapted to
other regions ofSanta Clara County, as recommended by study participants. This practice may be
particularly helpful to low-income families and families of color that may be intimidated by or
unfamiliar with child welfare and related systems. Other practices, such as vertical case
management, may hold promise for other regions ofSanta Clara County, though the success ofthis
practice may be linked to the "small county" dynamic described by South County study
participants. Some participants in focus groups that took place at the main office mentioned this
model as promising while others asserted that specialized expertise in different areas of social work
was an advantage in providing quality services that are informed by "the most current information."
Further examination ofhow the culture ofcommitment to maintaining and reunifying families is
created, communicated, and continued and how this might be adapted to other parts of the county
warrants attention.
103
Multivariate Analysis of Factors Predicting Reunification
Preliminary Models
In order to identify system-related factors and case characteristics predictive of reunification
of children with their families, two exploratory models were tested using multivariate analyses. The
first was with a larger sample of 1720 closed cases, and the second with 403 cases from a case
record review sample. Both samples were collected during Phase 2 and additional data was
collected and analyzed during Phase 3.
Results: Closed Case Sample
We chose 8 demographic and system-related variables available through CWS/CMS to
predict reunification: child's gender, ethnicity, age at time of case opening, the number of workers
assigned across the history of the case, the length of the case, the number of unique placement
homes assigned, the number of times removed from the family, and the county location (Main
County Offices versus South County Offices).
For our analysis, due to missing data across our set of variables, 1190 cases out of 1720
were available for logistic regression. Our model was significant in predicting the reunification of a
child with the family (X2 = 176.27, df= 11, p = .001, Cox & Snell K = 0.14). Specifically, Asian
American/Pacific Islanders were less likely than Whites, African Americans, and Latinos to be
reunified, and the fewer number of workers assigned, shorter length of a case, and fewer number of
unique placement homes assigned were related to reunification (while controlling for each variable
within the model). Please see Table 26.
104
. Larger sampe
. Reum'fi1cation m
Table 26 Log1s
. f1c Regress ion Pre d'1ctmg
o e1
l Md
Model
Cox & Snell Ri
xi,
0.14
176.27
11
.001
df
p
Predictors
B and Significance
Gender (female baseline)
Ethnicity
African American vs White
Latino vs White
Asian vs White
Other vs White
Latino vs African American
Asian vs African American
Other vs African American
Asian vs Latino
Other vs Latino
Other vs Asian
Age at case opening
Worker Number
Time Length of Case
Number of Unique Placement Homes
Number of Removals
County Location (Main vs South)
Using two-tailed tests* p < .05
** p < .01
Model Coding: 1 = reunified, 0 = not reunified
Odds Ratio
0.15
1.16
0.19
0.22
-0.59*
-0.33
0.02
-0.78**
-0.53
-0.81 ***
-0.55
0.26
-0.01
-0.19***
-0.02***
-0.l 0**
-0.02
0.40
*** p < .001
1.21
1.24
0.55
0.72
1.02
0.46
0.59
0.45
0~58
1.30
0.99
0.82
0.98
0.91
0.98
1.49
Summary of Key Findings and Implications:
Closed Case Sample
Our model composed of primarily demographic and system-related variables to predict
reunification (child's gender, ethnicity, age at time of case opening, the number of workers assigned
across the history of the case, the length of the case, the number of unique placement homes
assigned, the number of times removed from the family, and the county location) was significant.
Specifically, four of these variables predicted reunification: child's ethnicity, number ofworkers
assigned throughout the case, length ofthe case, and number of unique placement homes.
Asian American/Pacific Islanders were less likely than Whites, African Americans, and
Latinos to be reunified with their families. Given that Asian American/Pacific Islanders are one of
the minority groups least acculturated to mainstream society, many of their cases might be
perceived as more severe because of the differences between traditional Asian and mainstream
American cultural attitudes toward child rearing and discipline. Culturally, many traditional Asian
105
families have a high level ofparental control and tend to use corporal punishment as part oftheir
parenting (Chao &Tseng, 2002). Many Asian parents may also be unaware about the laws defining
abuse in this society. In these cases, cultural sensitivity by investigators and caseworkers would
help determine if these Asian family cases are indeed severe or ifthey are over-interpreted as such
because of a philosophical difference between the professional and the family in parenting and
discipline. Future empirical studies could help determine the extent to which abuse severity or
cultural differences explain this relatively low reunification rate among Asian American families.
However, as Phase 2 results indicated Asian American/Pacific Islanders are also the most likely to
be placed in Voluntary Family Maintenance (Hines et al., 2002), the CWS appears to value the
preservation of Asian American/Pacific Islander families when first referred to the system (perhaps
recognizing the challenges of culturally appropriate services, language issues, and other
circumstances related to the Asian American/Pacific Islander community) but once in the system,
Asian American/Paci.fie Islander family cases are not easily resolved and circumstances may persist
that place the child at risk/or further abuse of neglect.
The fewer number ofworkers assigned was also related to reunification. This finding
indicates consistency in service and the relationship between the social worker and the family may
improve the chances of reunification. Perhaps this is due to greater familiarity and thus more
effective advocacy for the child and family. This finding could also be a function of particularly
challenging family cases already less likely to be reunified needing a greater set of workers over
time (e.g., due to worker burnout in the case, or specialization needed across the case).
Shorter case length was also predictive ofreunification. Presumably, less severe cases
could be resolved in a timely fashion, usually concluding with reunification. Following a similar
logic above regarding the number of workers, a longer case may imply more problems, thus
involving more time for resolution, but also possibly a case already protracted into a situation where
reunification is unhealthy for the child.
The fewer number of unique placement homes assigned was related to reunification.
Change in placements may be due to systemic conditions (i.e. logistics and short term availability of
space) but multiple placements can also indicate a persistent problem with a child adapting to a
placement and accepting care offered. As this may again indicate a more severe case, reunification
could also become less likely.
Due to missing data across our set of variables, 1190 cases out of 1720 were available for
this multivariate analysis. More complete information would have improved confidence in our
findings. Also, as we have made references to the possibility ofseverity ofa case influencing
reunification, better measurement and a more comprehensive evaluation offamily history, abuse
and neglect circumstances, and other indicators could produce a stronger predictive model. In the
future, we may be able to include more psychosocial and case variables in a larger sample to build
upon our current findings.
Results: Case Record Review Sample
To identify better the influence of case characteristics and indicators of success (discussed in
the Successful Cases Analysis, beginning on page 48) we added 6 variables to the model with the
larger sample collected through our in-depth case record reviews and utilized the additional
106
information coded for Phase 3: a family's enhanced capacity to provide for their children's needs,
children receiving appropriate educational services, children receiving adequate services to meet
physical and mental health needs, number of referrals, number of previous times in the CWS, and
assignment to Family Maintenance or Family Reunification services. The location variable was not
included in this model as only one case coded from the South County Office area.
For our analysis, again due to missing data across our set of variables, 246 out of 403 cases
were available for logistic regression. Our expanded model was also significant in predicting the
2
reunification of a child with the family (x2 = 63.87, df= 16, p = .001, Cox & Snell R = 0.23).
Similar to the larger sample's model, Asian American/Pacific Islanders were still less likely than
Whites, African Americans, and Latinos to be reunified. Also, shorter time length of a case again
predicted reunification. However, in this expanded model with success indicators, the number of
workers assigned and number of unique placement homes assigned were not statistically related to
reunification. The child's age at time of case opening and the number of unique placement homes
did approach significance. Interestingly, the variables related to success cases and outcomes
( discussed in the Successful Cases Analysis, beginning on page 48) were not significant. Please see
Table 27.
107
. Reum"fl1cat1on
. .m C ase R ecord R ev1ew
.
Sampe
T able 27 Log1st1c
. . Regress ion P red"1ctmg
1 Model
Model
Cox& Snell If
x2
0.23
63.87
16
.001
df
p
Predictors
B and Significance
Gender (female baseline)
Ethnicity
African American vs White
Latino vs White
Asian vs White
Other vs White
Latino vs African American
Asian vs African American
Other vs African American
Asian vs Latino
Other vs Latino
Other vs Asian
Age at case opening
Worker Number
Time Length of Case
Number of Unique Placement Homes
Number of Removals
Number of Referrals
Previous Times in the CWS
Case Assignment (FM vs FR)
Child Receiving Adequate Educational
Services
Child Receiving Adequate Services to
Meet Physical and Mental Health Needs
Family's Enhanced Capacity to Provide
for Child's Needs
Using two-tailed tests* p < .05
** p < .01
+ approached significance at p < .10
Model Coding: 1 = reunified, 0 = not reunified
Odds Ratio
0.29
1.34
0.43
-0.19
-1.63*
-0.36
-0.61
-2.06**
-0.79
-1.44*
-0.17
1.27
-0.06+
-0.06
-0.03***
-0.13+
-0.05
-0.04
0.12
-0.50
0.10
1.21
1.53
0.83
0.20
0.54
0.13
0.46
0.24
0.84
3.56
0.95
0.95
0.98
0.88
1.05
0.96
1.13
0.61
1.10
-0.18
0.84
-0.13
0.88
*** p < .001
108
Summary of Key Findings and Implications:
Case Record Review Sample
These findings indicate that particular factors are important to consider when predicting
reunification, specifically ethnicity and the length oftime a case remains open. Similar to the model
in the larger closed case sample, this implies that more severe cases (those needing more time for
resolution) may be more problematic, thus reducing the chances of reunification. We saw in the
larger sample model that number of workers and unique placement homes were related to
reunification, but the number of workers was no longer significant and number of placement homes
approached significance. In addition, the younger the child at case opening the more likely
reunification would occur. This could be that an effort is made to reunify younger children with
their families given that they are in an especially crucial stage of psychological and emotional
development related to the attachment with parents. These variables should be explored in more
detail.
The finding again that Asian American/Pacific Islanders are less likely to be reunified than
Whites, African Americans and Latinos, highlights a main ethnic difference. In Phase 2 we
discovered that Asian American/Pacific Islanders were more likely than the other ethnic groups to
be enrolled in Voluntary Family Maintenance services (Hines et al., 2002). However, our results
here imply that when Asian American/Pacific Islander children are removed from the home, their
children are less likely to be reunified This may also indicate extreme outcomes where Asian
American/Pacific Islander children are either initially diverted from the system or once in the
mainstream ofthe system are less likely to be reunited with their families.
It was interesting that the successful case and outcome variables were not predictive of
reunification. This could be due to the significant impact of other variables overshadowing the
influence of success indicators. It could also be due to our use of reunification as a dependent
variable. As we discussed in the Successful Cases Analysis (see page 48) reunification is not
necessarily synonymous with success, and thus these variables may indeed be unrelated.
Again, due to missing data across our set of variables, 246 cases out of 403 were available
for this multivariate analysis. Although only a preliminary set of models, these findings are useful
for identifying areas in the CWS and family circumstances that can be studied further, especially in
conjunction with the specific goal of reunifying children with their families. For future research,
reasons as to why major system-related factors are not predictive would also be valuable.
109
The Role of the Court System in Child Welfare Practice
Few studies have examined the potential influence of the court system on child welfare
practices and outcomes, however as noted in the literature review (Section III), the court system is
likely to have a significant impact on the trajectory of child welfare cases. In order to examine this
issue further, this analysis examines the role of the court system in child welfare practice.
Specifically, two overall issues are addressed. First, an analysis of quantitative case record review
data, and qualitative data from one focus group conducted with the Court Officer's Unit will
describe court ordered changes to social worker recommendations at the initial
jurisdictional/dispositional hearing; the types of changes the court orders; if these changes differ by
child's ethnicity; and possible explanations for these changes. Secondly, a qualitative analysis from
the agency-wide focus groups (described on page 61) will focus on themes related to the
relationship between the child welfare and court systems.
Quantitative and Qualitative Methods:
Analysis of Court-Ordered Changes
Quantitative Analysis
The quantitative analysis of the role of the court system in child welfare practice included a
new case record review of the same sample of 403 closed child welfare cases that were included in
Phase 2 of the research.
Sampling
Case record reviews include the same sample of 403 closed child welfare cases that was
analyzed in Phase 2 of the research. This sample was created from 6761 total case closures over an
18-month period, and a data file containing 1753 unique cases representing one child per family and
one case opening was constructed. This data file was used to obtain our target sample of 403 cases
selected randomly, guided by stratification according to ethnicity, age and service type. Of the 403
cases reviewed for this court analysis, 12 cases were closed, but not reviewed because they were
with a worker and not at the retention center; and 13 were not reviewed because they had been reopened.
Data Collection
Data collection took place at the Santa Clara County's Social Services Record Retention
Center. A Senior Research Assistant worked with the Retention Center Manager to ensure that the
sample of case records from Phase 2 of the research were re-pulled for review. Cases were reviewed
by three research assistants, two of these research assistants participated in the Phase 2 case record
review data collection and so had already been trained in case record review methods and one
research assistant was newly trained to complete the case record reviews.
Instrumentation
A data extraction form was created and pilot tested (please see Attachment 5 for a copy of
the Court Case Record Data Extraction Form). The form contained questions related to court
110
ordered service or visitation changes for the child, mother, father, alternate caregiver, or the whole
case.
Analysis
We addressed this aim through a 3-part series of exploratory descriptive and bivariate
analyses. First, we identified whether there were any court ordered changes and if court ordered
changes in general were related to ethnicity. A single change in recommended services for a case,
regardless of magn~tude, would classify the case as one with a "court ordered change."
Second, we examined major changes and if these were related to ethnicity. We considered
any court ordered change in the three main service areas (those concerning bypass, family
maintenance, and family reunification) as a major change.
Third, we developed 4 summary scores consisting of the number of changes imposed
separately concerning the child, mother, and father, and then for the entire case. Across each case
there were 85 variables over 9 potential areas where a change could be ordered in addition to the
three main service areas. These areas were psychological evaluation, counseling services, health
services, drug/alcohol testing, substance abuse/alcohol treatment, parenting education, support
groups, domestic violence treatment, and visitation. Changes in these areas were assessed and
summed into the 4 scores for court ordered changes aforementioned: child changes, mother
changes, father changes, and total changes. These scores were also examined for ethnic differences.
Definition of Terms
Court Ordered Change - Major changes were defined as court orders to initiate bypass,
family maintenance, or family reunification services for the mother, father, and/or the family. The
remaining changes included those across 9 areas represented by 85 variables: psychological
evaluation, counseling services, health services, drug/alcohol testing, substance abuse/alcohol
treatment, parenting education, support groups, domestic violence treatment, and visitation. A court
ordered change could also include placing more restrictions or conditions on service
recommendations.
Qualitative Analysis
A separate focus group with representatives from the Court Office Unit was conducted
toward the end of the study in order to obtain feedback on the implications and meanings of the
quantitative findings. The focus group with the Court Office Unit contained 9 participants,
including two Social Workers (22.2%) Ill's, five social work supervisors (55.6%), one Program
Manager (11.1 %) and one Legal Clerk (11.1 %). The mean number of years working in child
welfare among the Court Officer's focus group was 11.6 years and ranged from 4 to 22. Focus
group participants had a mean of 3 .2 years in their current positions, with a range of .5 to 7 years.
Seven participants had MSW's (77.8%), one had a BSW (11.1%), and one reported unspecified
postgraduate education (11. 1%). Two of the participants identified as Latino (22.2%), two identified
as Asian American/Pacific Islander (22.2%), four identified as White (44.4%), and one identified as
mixed race (11.1%). Eight of the participants were female (88.9%) and one was male (11.1%).
111
Participants in the Court Officer Unit focus group were asked a series of questions designed
to provide insights into the "story behind the statistics." Specifically, Court Officers were invited to
share their experiences in relation to overall factors that contribute to court ordered changes and
comment on specific findings from quantitative data. Please see Attachment 6 for the specific
questions used to guide the Court Officer Unit focus group discussion.
Notes from the Court Officer Unit focus group were transcribed and analyzed for key
themes related to court changes. In addition, focus group respondent interpretations and insights
about results of the quantitative data were examined and summarized.
Quantitative Results:
Analysis of Court Ordered Changes
Frequency of Court Ordered Changes
Ofthe 403 cases, 178 (44.2%) recorded at least one court ordered change from the original
social worker recommendations from the jurisdictional/dispositional hearing, 200 (49 .6%) abided
by the original recommendations, and 25 (6.2%) were not available for evaluating the additional
court ordered changes.
Court Ordered Changes and Child's Ethnicity
We eliminated the cases with missing information to do our comparison by ethnicity,
leaving us with 378 valid cases. There was no significant relationship between the existence of a
court ordered change and ethnicity. Please see Table 28.
t me Groupa
Table 28 Proport1ons ofC ourt 0 rderedCh ane;es b,y Eh.
White
Latino
Total
African
American
Court Change
Asian
American/PI
Other
No
200
(52.9%)
25
(53.2%)
65
(51.6%)
68
(51.9%)
18
(62.1%)
24
(53.3%)
Yes
178
(47.1%)
22
(46.8%)
61
(48.4%)
63
(48.1%)
11
(37.9%)
21
(46.7%)
a Based
on 378 cases with valid information
Types of Service Changes
We examined the types of court changes and distinguished three to be relatively major given
their impact on the case: bypass, family maintenance, and family reunification. Ofthe 178 cases,
the most common change with these three types was the implementation of family reunification
services: 27 cases (15.2%) involved family reunification, 2 (1.1%) cases regarded bypass, and 1
case (0.6%) regarded family maintenance. Ethnic comparisons were conducted, but no relationships
were found with these major services. Please see Table 29.
112
Table 29 Bypass, Family Maintenance, and Family Reunification Court Changes by Ethnic Groupa
Total
African
White
Latino
Asian
Other
American
American/PI
Bypass
No
176
(98.9%)
22
(100.0%)
61
(100.0%)
2
0
(0.0%)
(0.0%)
No
177
(99.4%)
22
(100.0%)
60
(98.4%)
Yes
1
(0.6%)
0
(0.0%)
(1.6%)
(0.0%)
(0.0%)
(84.8%)
151
16
(72.7%)
55
(90.2%)
52
(82.5%)
(90.9%)
27
(15.2%)
(27.3%)
Yes
(1.1%)
0
62
(98.4%)
11
(100.0%)
20
(95.2%)
(1.6%)
1
0
(0.0%)
(4.8%)
63
(100.0%)
11
(100.0%)
21
(100.0%)
0
0
(0.0%)
10
18
(85.7%)
I
3
(14.3%)
1
Family
Maintenance
1
0
Family
Reunification
No
Yes
a Based
6
6
(9.8%)
11
(17.5%)
(9.1%)
on 178 cases with court change information
Court Changes Regarding Children, Mothers, and Father
Four scores were calculated to describe the number of changes imposed by the court, each
for the child, mother, and father, and finally a total score representing the entire case. The number
of changes related to the child's circumstances ranged from 0 to 4 with an average of 0.31 (sd =
0.65). Court ordered changes related to the mother were most common, ranging from 0 to 8
changes with an average of 0. 77 (sd = 1.28). The number of changes related to the father's
circumstances ranged from Oto 7 with an average of 0.67 (sd = 1.26). The number of total court
ordered changes related to a case ranged from 1 to 16 with an average of2.53 (sd = 2.22). No
significant ethnic differences were found regarding the number of court-imposed changes. Please
see Table 30.
113
Table 30: Number of Court Changes bv Ethnicity'a
White
Total
African
American
Child Related
Changes
Avg.
0.31
0.38
0.23
(0.78)
(sd)
(0.65)
(0.53)
Mother
Related
Changes
Avg.
(sd)
Latino
Asian
American/PI
Other
0.27
(0.54)
0.36
(0.50)
0.33
(0.73)
0.77
(1.28)
1.32
(1.94)
0.69
(0.99)
0.83
(1.30)
0.18
(0.40)
0.57
(1.29)
Father Related
Changes
Avg.
(sd)
0.67
(1.26)
0.82
(1.92)
0.61
(1.14)
0.62
(1.10)
0.55
(1.21)
0.90
(1.30)
Total Case
Changes
Avg.
(sd)
2.53
(2.22)
3.36
(3.36)
2.39
(1.63)
2.54
(2.39)
1.82
(1.33)
2.43
(2.01)
a Based
on 178 cases with court change information
Qualitative Results:
Analysis of Court Ordered Changes
Quantitative findings were presented to a focus group made up of 9 representatives of the
Agency's Court Officers Unit for discussion and feedback. Quantitative results suggested that
approximately halfofthe cases had indicated a court ordered change from the recommendations of
the social worker described from the jurisdictional/dispositional hearing. From our focus group
with members from the Court Officer Unit, the consensus was that the interpretation of the
measured proportion of court ordered changes found in the quantitative analysis, depended on what
we were actually measuring. The proportion is dependent on a variety of circumstances as to why
there was a change from the social worker recommendations including clerical errors needing
correction, changes due to modifications in the law, simple (albeit common) adjustments to services
such as visitation terms in addition to actual substantial changes due to conflicting opinions
concerning the direction of the case could all be related to a change between social worker
recommendations and court ordered.
Findings from the focus group with the Court Officer Unit provide additional information
regarding changes made to social worker recommendations at the jurisdictional/dispositional
hearing. First, capturing the details needed to conclude why social worker recommendations may
or may not be followed is extremely difficult. This is because variations by judge, courtroom
114
environment, social worker credibility and reputation, assigned lawyer (i.e., court appointed vs.
privately retained), as well as formal and informal case plan negotiation tactics frequently occurring
would complicate research. Second, the personal dynamics among the stakeholders (socials
workers, lawyers, judges, and families), including the strength of their positions and philosophical
standpoint can be such an influence that systemic conditions may be poor predictors of case
outcomes and success as they are related to the judicial process. Third, that despite the complex
circumstances involved, judges, guided by social workers, determine the best possible service plan,
while being aware of interpersonal dynamics and individual characteristics that may obscure
reasonable decisions. According to the focus group, social workers have a significant influence in
the court process and usually receive a majority of services they request for their clients.
Quantitative data indicated that few changes involved bypass, family maintenance, or family
reunification. The proportion of changes in these main service areas seemed reasonable to the focus
group. The low percentage of change in bypass, FM, and FR assignment should be low given the
legal criteria to qualify for such a change, as well as the intricate legal procedures.
When presented with quantitative findings that indicated no relationship between ethnicity
and court ordered changes in major service areas, or the overall number ofcourt-ordered changes,
it was the opinion of members of the focus group that ethnicity does not contribute to decisions in
court, but rather it is circumstances that drive the case.
Qualitative Methods:
Relationship Between the Child Welfare and Court Systems
Qualitative results describing the relationship between the child welfare and court systems
are gleanedfrom the agency-wide focus groups. These focus groups were structured to explore
practices at different choice points in the child welfare system. Please see page 61 for a full
description of agency-wide focus group methods and results.
Qualitative Results:
Relationship Between the Child Welfare and Court Systems
Results from agency-wide focus groups (please see page 61 for a.full description) suggest
that the relationship between DFCS workers and district attorneys and judges is problematic at
times. Many participants described difficulties in effective collaboration between the court system
and the child welfare system. Social workers suggested that problems between the court system and
the child welfare system may reflect fundamentally different perspectives on the needs and
circumstances of children and families in the CWS, and perhaps unrealistic expectations regarding
service plans. For instance, one social worker noted, "From a client's side, they are being told by a
court system and a social system that the way that they were brought up and the way that things
were done in their [clients'] family is wrong and not acceptable," noted a social worker focus group
participant. Several participants suggested that courts directly and indirectly set an unrealistic
standard for families, exemplified by one participant's comments:
"The court system just makes these orders and that's it. They don't consider anything else
except for this narrow focus on child and parent. The social worker has to look at resources
for the child, for the family, and language barriers. The court doesn't consider the impact it
115
has on the family, who has to pay for the services, shrinking or expanding resources. The
court is in an isolated bubble anything outside that bubble is not in their framework. "
Regarding the services ordered, another social worker commented, the "Court wants it all
done at once ...the court doesn't understand that for whatever reason the client may be only capable
of doing so much within the first 6 months." Additionally once the court does order a service or
case plan, social workers report that it is difficult to adjust mandated services "because the court
requires justification for removing the order; also it is difficult convincing the DA of [the need for]
new recommendations."
In general, focus group and interview participants expressed a need to bridge these gaps
between the court system and the child welfare system. Although some social workers described
having strong collaborative relationships with the court system, many described tensions with
attorneys and other court representatives in relation to recommendations for mandated services and
disposition of cases. Social workers also conveyed that it was common for court representatives to
dismiss their psychosocial family assessments and push for recommendations that the social
workers deemed disconnected from the reality and needs of families. "Legal aids rarely visit the
children or the parents, they need to be more reality-based," commented on worker. "Social workers
and the court system are both looking at the case from different views," noted another. Some
workers depicted the district attorneys as having more of a focus on "winning" a case than on
ensuring the best outcomes for the children and family.
Working with the DA is a challenge. Their role is to advocate for the child-but they often
do only one interview and they don't know the particulars of the case. There can be
antagonism between the social worker and the DA-there should be less antagonism
between the DA and the social worker and more cooperation.
Social workers also expressed concern about the fairness of the system. For example, "In court
cases, private attorneys are more likely to have their client's case dismissed compared to court
appointed attorneys." Many participants conveyed that many court representatives and social
workers exhibit a "gap between standards that common people have and that educated professional
people have," which may disadvantage low-income families and families of color. "Attorneys look
at the information provided by the social worker and make sweeping generalizations and judgments
about families," summarized one social worker. One supervisor captured concerns about failing to
recognize the context of client's lives:
When there is child abuse or neglect we raise our eyebrows to it, we go in there and want
everything to change but we 're not providing a basic foundation for change. We 're not
developing the community, we 're not providing better housing. We tell people to go out and
get suitable housing, but we 're not providing it.
Language issues also impact the experience offamilies in the court. "Families are unable to
communicate with their lawyers because of language barriers; English only attorneys can't
communicate with a monolingual Spanish speaking client," summarized one worker. Other workers
observed that "Families may not like to leave a message for an attorney because of the language
factor" and "the court process can be confusing for English speaking families, but more so to
Spanish speaking families." Furthermore, "the court reports are all in English and the clients don't
have a translated version that they can look at."
116
Additionally, social workers expressed concern that client outcomes are often dependent on
working relationships or attitudes ofDA 's. "There are some DA's that I have a great rapport
with-but a few are very judgmental and I feel like they are not supportive of the family." The
DA's are perceived as having substantial power. "The DA's word is very strong in court because
they are supposed to be speaking on behalf of the child," stated one social worker. Other social
workers suggested that some DA' s pressured social workers to change recommendations or would
attempt to discredit them in the courtroom, particularly when the DA's case was weak.
Agency-wide focus group results also provided recommendations regarding ways to enhance
collaborative relationships between the court system and the child welfare system. These
recommendations included,
•
•
•
•
•
Explore opportunities to develop standards for practice between the courts and the child welfare
system including development of shared policies, practices and procedures. Such efforts could
help address common concerns of social workers that generally called for a "switch from courtcentered practice to family/community-centered practice."
Develop a systems level mechanism for problem solving to address conflicts between social
workers and DA' s and "set a minimum standard or tone for what is appropriate behavior in the
court."
Facilitate development of role clarity between child welfare and the courts to address
widespread concern that decision-making is often not based on the social worker assessments
and, instead, is often based on perceptions ofDA's who may not have spent sufficient time with
children and who may not have evaluated family strengths.
Facilitate a dialog with both court and child welfare workers about standards for "good enough"
parenting and the dynamics ofpoverty in decision-making.
Explore opportunities for support of social workers in the court process through training,
enhanced supervision, or coaching from experienced colleagues.
Summary of Key Findings and Implications:
The Role of the Court System in Child Welfare Practice
In Phase 3, we examined the role of the court system in child welfare practice by exploring
the types of court ordered changes that are made to social worker recommendations at the initial
jurisdictional/dispositional hearing, as well as the relationship between the child welfare and court
systems. Quantitative results indicated that the court made changes to the initial social worker
recommendations at the jurisdictional/dispositional hearing in approximately halfofthe cases.
Ethnicity was not related to whether or not a change occurred. These results do not necessarily
indicate that the system and courts treat children and families from different ethnicities the same,
but rather that when the judge decides to make a change from the social worker recommendations,
these changes appear to occur equally across ethnic groups. It was the opinion of members of the
Court Officer Unit focus group that ethnicity does not contribute to decisions in court, but rather it
is circumstances that drive the case.
Findings from the focus group with the Court Officer Unit suggest that many factors
complicate an accurate measurement ofwhy social worker recommendations may or may not be
117
followed. Factors that may impede accurate assessment of the agreement between child welfare and
court personnel on service plan recommendations include the strong influence of individual
stakeholders involved, and formal and informal case plan negotiation tactics all of which paint a
much more complicated picture.
Qualitative findings from the agency-wide focus groups suggested that collaboration
between the child welfare and court systems is problematic. Agency-wide focus group participants
felt that the child welfare and court system have different perspectives on the needs and
circumstances of children and families in the CWS and that the court system may have unrealistic
expectations of families.
Additionally, many agency-wide focus group participants described the ways in which child
welfare and district attorneys and judges interact as ineffective. Some social workers felt that some
district attorneys and judges would dismiss their assessments and recommendations and may try to
pressure social workers to change their recommendations. A somewhat similar finding was noted by
Knepper and Barton, (1997) who found that although judges tended to accept social worker
recommendations, the relationship between social workers and the court system plays a significant
role in their decisions. Their study found that when social workers adhered to the group norms of
the court, judges rewarded them by not ordering "unrealistic" practices and allowing reasonable
time frames for mental health evaluations to be completed. Similarly, agency-wide focus group
findings seem to suggest that if social workers go against the implicit rules ofthe court system, then
judges may override their recommendations, but if they adhere to expectations ofthe court system
than recommendations are accepted.
Limitations and Recommendations
From these exploratory findings, issues arose regarding limitations and recommendations.
First, although quantitative and qualitative information described many of the circumstances behind
why social worker recommendations are not upheld by the court, it is very difficult to ascertain the
degree to which each circumstance influences a court ordered change. Second, better and more
exhaustive record keeping may not be the solution, as many informal and undocumented actions in
the negotiation process occur regularly, and are confidential, thus making much of the data
unavailable for research and evaluation purposes. Third, assessing other factors such as courtroom
environment, stakeholder characteristics, and other judicial issues would be advantageous in later
research. Although obtaining this information is a formidable task, these factors should be included
in the future, as measuring primary systemic factors no longer seem adequate to explain outcomes
in the CWS.
118
V.
STATEWIDE COMPARATIVE ANALYSIS
The statewide comparative analysis of effective practices for children and families of color
in the child welfare system included a statewide survey of child welfare directors, managers or
supervisors in California counties who were asked to describe effective practices for children and
families of color in their child welfare systems. In addition, this statewide comparative analysis,
included a quantitative analysis that identified the link between these practices and levels of
disproportionate representation of children of color in county child welfare systems. As such, two
overall analyses were included in the statewide comparative analysis of effective practices for
children and families of color in the child welfare system, 1) descriptive information from the
statewide survey on the types of practices, practices and practice characteristics identified by survey
respondents, and 2) multivariate results that link types of practices, practices and practice
characteristics to over or underrepresentation of children of color within counties. Additionally,
contextual information on the estimated proportions of children of color in county populations and
county child welfare systems within selected regions of California is provided, as well as
information on the estimated proportion of children within the overall county population who are in
the CWS.
Methods
The statewide survey on effective practices for children and families of color in the· child
welfare system was designed to gather information on 1) key practices in California counties that
may have an impact on children and families of color, and 2) practices considered most promising
for children and families of color. For the purposes of the survey, "practices" were defined broadly
to include programs, services, strategies, policies, or tools.
Data Collection
Child welfare directors, managers or supervisors were mailed an invitation to participate in
the survey, and a copy of the survey. Two weeks later, a follow-up email that contained the survey
as an attachment and telephone calls were made to encourage participation. Counties that failed to
respond to the survey received approximately 3 additional follow-up emails or telephone calls
requesting their participation or a referral to someone else in the agency who might be able to
participate.
Respondents were offered four options for returning the survey to the CWRT, either by 1)
completing the survey on their computer and emailing it as an attachment, 2) completing the survey
over the telephone with a member of the CWRT, 3) faxing the completed survey, or 4) mailing the
completed survey. Surveys that were completed over the telephone were tape recorded in order to
ensure the accuracy of information and once information was verified, the tapes were erased. Fortyfive counties responded to the survey, resulting in a 77.6% response rate. Of the 45 counties who
completed the survey, 23 returned it via email, 10 by telephone, 7 by fax, and 5 returned the survey
by mail.
119
Sample
The sample included child welfare directors, managers or supervisors from California
counties. Participants for the statewide survey were identified through the 2003 County Welfare and
Social Service Directory. An invitation to participate and a copy of the survey were mailed to each
child welfare director, manager or supervisor in each of the 58 counties in California. Of the 13
counties that declined to participate, 4 indicated that their child welfare population is not diverse
and they did not have any special programs for children and families of color; 3 replied that the
survey was too time consuming, 3 did not reply to the survey, 2 indicated that they would complete
the survey, but did not return it and one county indicated that they were going through a reorganization and could not complete the survey because of these changes.
Instrumentation
A survey entitled "Statewide Survey on Effective Practices for Children and Families of
Color in the Child Welfare System" was developed and tested in a pilot study involving five
counties (please see Attachment 7 for a copy of the Statewide Survey). Feedback provided from the
pilot counties resulted in a reduction in the number of questions and minor changes in wording. The
survey contains questions related to caseload characteristics of counties and the most effective
practices for children and families of color in each county's child welfare system. The survey asks
respondents to provide demographic information on children in various components of the county's
child welfare system. Respondents are also asked to identify up to five of the most effective
practices for children and families of color in their county's child welfare system, ranking them
from #1 (most effective), #2 (second most effective) and so on. Multiple choice questions on the
most effective practice (#1) are also included in the survey.
Analysis
Quantitative analyses were conducted using the Statistical Package for the Social Sciences
(SPSS) software. Descriptive results were generated by running frequencies and percentages for
variables of interest. Multivariate results were generated by conducting controlled tests of empirical
associations between types of practices, practices, and practice characteristics and levels of
disproportionate representation of children of color in county CWS's.
Qualitative data describing the most effective practices were analyzed through content
analysis methods. The types of practices within each of the four categories were identified and
descriptive information about their target population, description and reasons why these practices
are effective with children and families of color were delineated.
Contextual Tables:
County Child Populations and CWS Involvement
The following tables provide contextual information on selected county child populations
and CWS involvement by ethnic group. Please note that the counties presented in these tables are
not necessarily the counties who responded to the statewide survey. Also please note that because
the most recent information available on county child populations is for the year 2001, information
on CWS populations for the year 2001 is also provided. As such, current 2003 proportions of
120
children of color in California county CWS may vary slightly from the 2001 proportions provided
in the following tables.
This descriptive information is presented in order to provide a context through which to
view the results from the statewide survey. Because compiling information on all 58 California
counties would be cumbersome, information on selected counties is provided. Table 3 la provides
information on the estimated proportions of children of color in county populations and the CWS
within 13 Bay Area counties. Table 31 b described estimated proportions of children of color in
county populations and the CWS within selected counties by California regions. Last, Table 31 c
provides information on the estimated proportion of children within the population in the CWS
among selected counties by region.
121
. and Ch.Id
l . Popu Iat1on
We lfiare Siystem It lil ay
Table 31 a: Est1mate
.
l reno fC oorm
l
dP roport1ons o fCh.ld
Ethnic Group
Asian
Latino
County
African American
White
%in
%in
%in
%in
%in
%in
%in
Population
Population
Population
Population
24.6%
11.9%
22.3%
66.1%
33.7%
18.0%
19.0%
Alameda
20.2%
12.7%
45.2%
13.9%
38.5%
11.4%
54.1%
Contra Costa
18.7%
9.4%
6.6%
71.3%
3.2%
19.4%
Marin
35.7%
1.4%
18.5%
67.3%
62.1%
0.7%
4.9%
24.7%
Mendocino
14.2%
52.3%
5.8%
4.3%
57.6%
Monterey
32.0%
30.1%
11.8%
32.3%
17.4%
4.3%
Napa
61.2%
1.5%
67.4%
53.1%
60.7%
2.1%
32.1%
1.2%
43.9%
0.5%
San Benito
22.2%
12.3%
37.0%
27.6%
11.6%
12.9%
70.4%
San Francisco
32.0%
29.8%
33.0%
23.9%
38.5%
26.7%
4.3%
San Mateo
12.2%
32.8%
52.0%
28.1%
3.5%
Santa Clara
35.4%
27.4%
41.0%
49.5%
3.8%
53.8%
46.1%
1.1%
3.4%
Santa Cruz
42.4%
9.1%
16.0%
15.0%
18.9%
49.5%
41.2%
Solano
4.5%
2.0%
21.8%
28.8%
70.6%
57.4%
9.0%
Sonoma
cws
cws
State of
California
37.7%
30.3%
7.1%
31.3%
cws
42.9%
33.8%
11.7%
rea ount1es
%in
cws
2.9%
2.0%
3.3%
0.2%
1.7%
2.8%
1.2%
4.8%
9.4%
7.4%
0.2%
3.7%
1.0%
2.8%
Native American
% in
%in
Population
0.4%
0.7%
0.4%
0.5%
0.2%
0.3%
6.0%
12.9%
0.2%
0.4%
0.0%
0.6%
0.4%
1.2%
0.2%
0.8%
0.3%
0.1%
0.2%
0.8%
0.3%
0.2%
0.6%
0.6%
1.0%
3.2%
cws
0.5%
1.1%
Notes:
1. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in Januacy 2001 according to Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A.,
Lecy, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. & Kim, H. (2003). Child Welfare Services Reports for California. Retrieved [8/29/2003], from University of California at Berkeley Center for Social
Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>.
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org.
3. The Bay Area Counties are the 13 identified by the CWDA, and are also categorized as "big" counties in California except San Benito which is "small."
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California County Data Book summarized information on children aged 0-17
whereas the CWS data may include children as old as 19. Some totals may not equal 100% due to missing data or rounding.
122
• and Ch"ld
.
. Popu Iat1on
Ta ble 31b Est1mate
.
We lfiare Siystem w·Ith"m SI
I reno fC o lorm
I
e ectedC ount1es B,y Reg1on
dP roport1ons o fCh"ld
County
Bay Area
Alameda
San Francisco
Santa Clara
Mountain
Calaveras 5
Placer
Sacramento
Northern
Butte
Humboldt
Lassen5
Southern
Los Angeles
Orange
San Diego
Santa Barbara
Vallev
Fresno
Merced
Stanislaus
State of California
White
% in
African American
% in
% in
Population
Ethnic Group
Latino
%in
%in
Population
Asian
%in
%in
Population
Native American
%in
% inCWS
Population
%in
Population
cws
33.7%
27.6%
35.4%
18.0%
11.6%
27.4%
19.0%
12.9%
3.5%
66.1%
70.4%
12.2%
24.6%
22.2%
32.8%
11.9%
12.3%
52.0%
22.3%
37.0%
28.1%
2.9%
4.8%
7.4%
0.4%
0.2%
·0.2%
0.7%
0.8%
0.8%
86.3%
83.3%
53.7%
91.4%
79.0%
43.4%
0.4%
0.7%
12.5%
0.8%
3.6%
35.2%
10.1%
12.0%
17.9%
4.7%
12.0%
17.1%
0.8%
3.2%
14.9%
0.0%
1.8%
3.1%
2.3%
0.8%
1.0%
0.8%
2~8%
1.0%
75.1%
78.4%
79.3%
73.9%
65.6%
61.0%
1.8%
1.0%
1.1%
8.9%
2.3%
1.2%
13.8%
8.0%
15.0%
7.9%
5.1%
26.8%
7.5%
4.3%
1.1%
1.3%
0.3%
0.0%
1.7%
8.2%
3.5%
6.8%
22.4%
8.5%
19.8%
40.9%
46.0%
42.2%
15.2%
39.1%
35.1%
41.5%
9.0%
1.8%
7.1%
2.4%
42.3%
6.8%
24.5%
15.1%
60.1%
43.3%
36.6%
50.2%
38.9%
48.0%
35.0%
42.2%
10.8%
13.7%
9.9%
4.8%
2.9%
5.3%
3.0%
1.0%
0.2%
0.2%
0.5%
0.4%
0.4%
0.2%
2.2%
0.2%
31.1%
35.9%
53.4%
28.0%
35.0%
50.3%
5.7%
3.9%
2.3%
22.3%
13.9%
13.0%
48.5%
44.9%
35.1%
45.9%
44.0%
32.4%
13.8%
15.0%
8.4%
2.1%
6.4%
2.6%
0.8%
0.4%
0.8%
1.7%
0.1%
0.5%
37.7%
30.3%
7.1%
31.3%
42.9%
33.8%
11.7%
2.8%
0.5%
1.1%
cws
cws
cws
Notes:
I. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in January 2001 according to Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A.,
Lery, 8., Shaw, T., Dawson, W., Piccus, W., Magruder, J. & Kim, H. (2003). Child Welfare Services Reports for California. Retrieved (8/29/2003), from University of California at Berkeley Center for Social
Services Research website. URL: <http:/fcssr.berkeley.edu/CWSCMSreports/>.
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org.
3. Superscript "s" denotes a CWDA categorized "small" county as opposed to "big" county.
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California County Data Book summarized information on children aged 0-17
whereas the CWS data might include children as old as 19. Some totals may not equal I00% due to missing data or rounding.
123
Table 31c:
Estimated Proportion of Children Within the Population in the Child Welfare System Among
.
Select ed Counf1es B,y Reg1on
Region and County
Number of Children
Percentage of Child
inCWS
Population in the CWS
Bay Area
San Mateo
875
0.47%
Monterey
641
0.52%
San Benitos
84
0.54%
Napa
178
0.60%
Marin
361
0.70%
Solano
821
0.71%
. Sonoma
941
0.83%
Santa Cruz
588
0.88%
Santa Clara
4,401
0.94%
Contra Costa
3,061
1.29%
Alameda
6,033
1.54%
San Francisco
3,050
1.97%
Mendocino
628
2.77%
Mountain
Placer
1.28%
843
Calaverass
1.33%
128
Sacramento
8,436
2.45%
Northern
Lassens
82
1.11%
Humboldt
389
1.27%
Butte
827
1.62%
Southern
Santa Barbara
0.48%
516
Orange
7,316
0.89%
San Diego
9,935
1.19%
Los Angeles
1.80%
52,630
Valley
Stanislaus
1280
0.89%
Merced
822
1.11%
Fresno
4757
1.80%
State of California
139,823
1.41%
Notes:
1.
Children in Child Welfare System refer to those in a service component of ER. FR. FM or PP in January 2001 according to
Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A., Lery, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. &
Kim, H. (2003). Child Welfare Services Reports for California. Retrieved [8/29/2003], from University of California at Berkeley Center for
Social Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org
3. Superscript "s" denotes a CWDA categorized "small" county as opposed to "big" county
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California
County Data Book summarized information on children aged 0-17 whereas the CWS data might include children as old as t 9.
124
Descriptive Results:
Effective Practices for Children and Families of Color
Two counties did not provide information on effective practices; as such, a total of 43
effective practices for children and families of color in the child welfare system were identified.
Practices identified by the respondents as the most effective (# 1) for children and families of color
were divided into one of four categories: 1) beginning choice point practices (N = 13), 2) continuing
or ending choice point practices (N = 12), 3) organizational or structural practices (N = 14) and 4)
both beginning and continuing choice point practices (N = 4). Descriptive results for each of these
four categories are presented separately. In addition, descriptive statistics on the location, target
population, goals, history, factors related to implementation, factors that support the practice and
barriers to successful implementation of the practice are presented.
It should also be noted that many counties reported that they do not target services to any
specific racial/ethnic group and that the practices they identified as being most effective with
children and families of color, were also in general, their most effective practices for their overall
child welfare population.
Beginning· Choice Point Practices (N = 13)
Thirteen counties identified practices that take place at early choice points as their most
effective practices for children and families of color in the CWS. The 13 beginning choice point
practices identified by the counties fell into three categories of practices: 1) family group
conferencing, or other team decision-making services (n = 7, 53.8%) 2) voluntary family ·
maintenance or other prevention/family preservation services (n = 4, 30.8%), and 3) risk assessment
tools or systems to guide decision-making (n = 2, 15.4%). Please see Table 32.
Family group conference or other team decision making
ractice
Voluntary family maintenance or other prevention/family
reservation service
Risk assessment tool ors stem
7
53.8%
4
2
Table 33 provides qualitative data on the beginning choice point practices that are related to
family group conferencing or other team decision-making practices. Most respondents described the
target population as children and families who have been detained or are at risk for court-ordered
services, although some respondents described using family group conferencing for all families
involved with the CWS, not just those at risk for court ordered out-of-home placement.
Most respondents described family group conferencing or other team decision-making
practices as involving the family, extended family or the support system in the group conference.
Some respondents described staff meetings or multidisciplinary team meetings that did not involve
the family. Survey respondents described the practice as a collaborative effort with the goal of
producing a plan for the child and family. Many respondents described the process as strengths
based and focused on how best to help children and families.
125
In general, respondents felt that the practice was effective with children and families of color
because it involves clients in the decision-making process. This inclusive model of practice was
described as particularly helpful for children and families of color because they may be "more
distrustful of the system," and also because inclusive practice may be more culturally appropriate
with some families. Additionally, team meetings not involving family members were described as
being beneficial to children and families of color because more accurate decisions are expected
from bringing numerous perspectives together.
Table 34 provides qualitative data on voluntary maintenance or other prevention/family
preservation practices. The target population for these services included families at risk for courtordered out-of-home placements, as well as one program aimed at expectant mothers and parenting
women. These practices were described as intensive in nature and focused on preventing child
maltreatment and out-of-home placements.
Voluntary family maintenance or other prevention/family preservation services were
described as effective with children and families of color because their goal is to keep children with
their birth families. Additionally, these services were viewed as effective for children and families
of color because they are collaborative in nature, involving other systems or organizations such as,
Native American tribes, mental health and probation systems.
Table 35 provides qualitative data on risk assessment tools or systems. These practices are
targeted to all children reported to the child welfare system for suspected maltreatment. Risk
assessment practices were described as tools to help make decisions at critical points by assessing a
number of different risk factors.
Risk assessment tools or systems were described as effective with children and families of
color because they help to assess risk in an objective way, thereby minimizing potential bias based
on race/ethnicity.
Table 36 provides quantitative results on all of the beginning choice point practices
identified from the statewide survey. In general, beginning choice point practices are most often
delivered at the county child welfare office (n = 7, 53.8%); they tend to be targeted to children of all
ages (n = 12, 92.3%); as well as families of any ethnicity (n = 13, 100.0%). Families with substance
abuse problems (n = 9, 69.2%) are also targeted for these beginning choice point practices, as are
families with domestic violence, (n = 9, 69.2%), families with a mental health problem (n = 9,
69.2%), families with income problems (n = 8, 61.5%), and families with a disability (n = 8,
61.5%).
The most frequently identified goal of the practices was to prevent out-of-home placement
(n = 11, 84.6%); other goals included to assess level ofrisk (n = 10, 76.9%); improve psychosocial
functioning of family (n = 10, 76.9%); address substance abuse problems of family (n = 8, 61.5%);
address socioeconomic problems of family (n = 8, 61.5%); improve parenting skills, (n = 7, 53.8%);
address mental health problems of family (n = 7, 53.8%); ensure a permanent placement for the
child (n = 7, 53.8%), and reunify children with their birth families (n = 6, 46.2%)
Table 37 provides data on the history of the beginning choice point practices. Most were
first implemented in response to a particular problem (n = 7, 53.8%) and most were first initiated by
126
the director of the child welfare system (n = 8, 61.5%). The year the practices were first
implemented ranged from 1992 to 2002, and 15.4% (n = 2) of respondents reported that the practice
has changed significantly between now and the time it was first implemented.
Table 38 provides information on factors related to implementation ofbeginning choice
point practices. The majority of these practices are permanent (n = 10, 76.9%); most are funded
through state funds (n = 7, 53.8%) and 38.5% (n = 5) are mandated. Most social workers are
provided with training in how to implement the practice (n = 12, 92.3%); 15.4% (n = 2) of counties
contract with another agency or system to deliver the practice and 69.2% (n = 9) indicated that the
practice is widespread throughout the child welfare system. No respondents reported that the
practice is delivered in a specific geographic area or community. The majority of respondents (n =
8, 61.5%) reported the practice has been evaluated to determine its effectiveness and of those, 87 .5
% (n = 7) indicated that the practice is effective with the target population. Of those counties where
the practice had not been evaluated, 80.0% (n = 4) indicated that an evaluation is not needed.
Factors that support the practices and barrier to successful implementation are presented in
Table 39. The majority of respondents felt that families and children help support the successful
implementation of the practice (n = 10, 76.9%). Support from the mental heath system was also
frequently noted (n = 9, 69.2%); as was support from the substance abuse system (n = 8, 61.5%),
contract agencies (n = 7, 53.8%), county government officials (n = 7, 53.8%), and evaluations that
show practice is successful (n = 7, 53.8%). Few barriers to successful implementation were noted,
38.5% (n = 5) ofrespondents indicated that there were no barriers to successful implementation, and
38.5% (n = 5) also reported unstable funding is a barrier.
127
Table 33: Beginning choice point practices: Family group conference or other team decision-making practice (N = 7)
Name of
Target
Description of practice
Why effective with children and families of color
population
practice
Family
Available to all
The Family Group Conference is a tool for solving
I believe that this practice works well for families
Group
consumers who
family problems that is based on a simple, traditional
of color because it is based on principles shared
by those families. It involves much extended
Conferencing have family,
belief: The combination of family strengths and
extended family, community support can keep children safe and well
family, and is centered around a meal (family
friends, support cared for. A Family Conference is a gathering of family alone time). It is inclusive of all family members,
members, friends, and community specialists who join
including elders. Meetings can be held at the
together to improve the care and protection of a specific location that best fits the needs of the family.
child or children.
Family
Children of all
Family Conferencing is a program that uses strengthThis program allows families to assist in the
based methodology by allowing families to participate
Conferencing ages
decision process as to what is best for the family.
in the decision-making process as to how best help their
children. Family Conferencing is a gathering of
extended family members, other intimate support
people, and service providers in order to share
information and give the family an opportunity to
create a plan for their children.
Involving the extended family, non-related extended
Family
All families
These families are often even more distrustful of
Group
involved with
family and all service providers in a conference in
the "system" than most families; this practice
Conferencing CWS at any
which the family takes responsibility for making a plan allows the family to exert more control over the
for the child/
stage of the
direction of services and the planning for the
process can use
children.
this program
Youth Pilot
Children at risk This program targets all children (including children of Team decision-making/goal setting/evaluation
Project
of placement, or color) at risk of placement, or at risk of being moved to process which includes the family and all support
(YPP)
at risk of more
a more restrictive placement. The program uses monies persons they wish to include, as well as the
restrictive
that would be spent on foster care to provide services to various service providers working with the
family.
placement
maintain the child in the home or in a less restrictive
placement. The program uses a team decision-making
and goal-setting model, which includes the family and
supports.
128
Table 33 continued: Beginning choice point practices: Family group conference or other team decision-making practice (N = 7)
Target
Name of
Description of practice
Why effective with children and families of color
practice
population
Placement
All detained
After a child is detained, a placement meeting is held as Families and relatives are able to participate in
Meetings
children and
soon as possible to determin the "next best placement"
the decision making and are able to provide input
their families
for that child. In addition to Social Workers and other
in to the "next best placement" for a child, which
staff, parents and family members are included
can include placement with relatives or in a
whenever possible in this process, so that their input
culturally relevant home if the child cannot be
can be provided in this decision. The safety of the child safely returned to his or her parents.
and culturally relevant placements are goals ..
Detention
All children,
No child is detained by CPS or remains detained by law Many perspectives and opinions are brought to
Case Staffing CPS may
enforcement without being discussed as soon as
bear at inception, promoting more objectivity and
consider
possible in a staffing. The participants include the line balance regarding FACTS and EVIDENCE of
detaining or who ER worker in the matter, ER Supervisor, On-going
risk and best interests of the child. Less
have been
Case Supervisor, and CPS Program Manager. The ER
possibility of stereotyping, subjectivity, ignorance
detained by law worker presents the issues of protection, risk, current
of cultural factors.
enforcement in
situation, the court worker/supervisor, lends legal
an emergency
documentation evidence perspective, all others ask
questions, add comments with the ultimate goal of
consensus on whether to proceed with intervention and
what the intervention looks like.
Team
All ER referrals Consensus decision making with workers, key program This decision-making forum is used for ALL
considered for
Assessment
supervisors, mental health clinician, PHN, family
cases moving to on-going services. The core
Planning
voluntary or
decision meeting coordinator, and often other service
group attends pretty consistently and represents
court-ordered
providers. Type & level of CWS intervention
multiple perspectives, resulting in more consistent
services
determined after exploring safety/risk, family
and well thought out decision making regarding
strengths/needs, and available
case handling and service needs.
resources/services/strategies. This is a collaborative
process which may be utilized for other key case
decisions.
129
Table 34: Beginning choice point practices Voluntary family maintenance or other prevention/family preservation service (N = 4)
Target
Description of practice
Name of
Why effective with children and families of color
population
practice
Voluntary
Families at high Social worker and social work assistant are dedicated to American Indian Tribe has facilitated family
program. Services are based on strength-based and
conferencing for tribal families.
Family
risk of court
Maintenance intervention due family-focused approaches, including family
Program
to abuse and
conferencing and family decision-making. Promoting
neglect issues.
Safe and Stable Families funding used for contracting
services to include psych assessments, intensive case
mana~ement services, and a half-time family advocate.
Expectant
Program in which pregnant mothers with a CPS history Missing
High Risk
Mothers and
or involvement coupled with drug or alcohol use are
Infant
children under
targeted for intensive services and treatment to avoid
Program
age 1
placing their children into the foster care system.
Mothers of newborns are also a target population for
this program.
Family Preservation practice is an integrated,
Family
Abused and
Built a community consensus around the effective
comprehensive approach to strengthening and
service delivery model for Family Preservation in
Preservation neglected
the County; Supported public/private
families at
preserving families who are at risk of or already
collaborative planning and service delivery
imminent risk of experiencing problems in family functioning with the
efforts; Preventive-oriented; Collaboration with
placement
goal of assuring the physical, emotional, social,
educational, cultural and spiritual development of
other County Departments such as DPSS, Mental
children in a safe and nurturing environment.
Health and Probation; Strong and continued
leaderhip within the Department of Children and
Families Services.
Missing
Provide intervention and make referrals to other
We do not have foster homes in our county. We
Keep
agencies so that family can remain intact, if possible.
have seen children go through foster care system
families
together
and the majority do not do well.
130
Table 35: Beginning choice point practices Risk assessment tool or system (N = 2)
Description of practice
Target
Name of
population
practice
Research-based & validated tools for decision-making
All new intake
Structured
at critical points in child welfare and Foster Care ·
to CPS
Decision
systems developed by the Children's Research Center
(currently) will
Making
in collaboration with COSS & California Counties.
System
soon extend to
(SOM)
all active cases
The risk assessment process looks at the child factors,
Best practice All children
including age, disability, illnesses as well as the
reported
risk
parent's ability to care-take, the history of abuse, also
assessment
speaking with the child, the parent and collateral
during the
contacts and then making a decision. We do not have a
ER phase
universal risk assessment tool, the workers are trained
in this process. We have an ER team that does this
practice.
Why effective with children and families of color
Objective nature of factors assessed.
Children are assessed in an objective wayconsidering their factors that place them at risk,
rather than being targeted for their ethnicity.
131
T a ble 36 : Begmnmg c hoice point practices: Location, taniet population and goals (N = 13)
Frequency (Percentage)
Where does practice #1 take place?
7 (53.8%)
• County child welfare office
4 (30.8%)
• Throughout the child welfare system
2 (15.4%)
• Birth family's home
Contract
agency
site
2
(15.4%)
•
Foster family's home
1
(7.7%)
(7.7%)
1
• Another out-of-home placement site
Other
2 (15.4%)
Who is the target population for practice # 1?
12 (92.3%)
• Children of all ages
(7.7%)
1
• Very young children
13 (100%)
• Families of any ethnicity
9 (69.2%)
• Families with substance abuse problems
(69.2%)
9
Families
with
domestic
violence
•
9 (69.2%)
• Families with a mental health problem
8 (61.5%)
Families with income problems
8 (61.5%)
• Families with a disability
6 (46.2%)
• Other
What are the purposes and goals of practice # 1?
11 (84.6%)
• Prevent out-of-home placement
10 (76.9%)
Assess level of risk
10 (76.9%)
• Improve psychosocial functionimz of family
(61.5%)
8
• Address substance abuse problems of familv
(61.5%)
8
• Address socioeconomic problems of family
(61.5%)
8
Address housing problems of family
(53.8%)
7
Improve parenting skills
(53.8%)
7
• Address mental health problems of familv
(53.8%)
7
• Ensure a permanent placement for the child
(46.2%)
6
• Reunify children with their birth families
(7.7%)
1
• Other
•
•
•
•
•
•
132
Table 37: Beginning choice point practices: History of practice (N = 13)
Frequency (Percentage)
Why was practice # 1 first implemented?
7 (53.8%)
• In response to a particular problem
4 (30.8%)
• Funds became available
(30.8%)
New
policies
were
implemented
4
•
Who first initiated practice # 1?
8 (61.5%)
• Director of the child welfare system
4 (30.8%)
• Child welfare workers
3 (23.1 %)
Other
•
When was practice #1 first implemented?
1
(7.7%)
• 2002
3 (23.l %)
2001
1
(7.7%)
2000
•
2 (15.4%)
• 1999
1 (7.7%)
• 1998
1
(7.7%)
1995
(7.7%)
1
1992
Has practice # 1 changed significantly between the time it
was first implemented and now?
2 (15.4%)
• Yes
•
•
•
133
Table 38: Beginning choice point practices: Factors related to implementation (N = 13)
Frequency (Percentae;e)
Current implementation status of practice # 1
IO (76.9%)
• Permanent
(23.1%)
3
• Dependent on re-funding
What is the primary funding source for practice # 1?
(53.8%)
7
• State funds
5
(38.5%)
• Federal funds
(23.1%)
3
• County funds
2
(15.4%)
Other
•
Is practice # 1 mandated?
5 (38.5%)
• Yes
Are social workers provided with training in how to
implement practice #1?
12 (92.3%)
• Yes
Do you contract with another agency or system to deliver
practice # 1?
2 (15.4%)
• Yes
Is practice # 1?
9 (69.2%)
• Widespread throughout the child welfare system
4
(30.8%)
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
(0.0%)
0
• Yes
Has practice #1 been evaluated to determine its
effectiveness?
(61.5%)
8
• Yes
If yes, what were the results of the evaluation?
(87.5%)
7
• Effective with target population
(12.5%)
1
• Other
If practice # 1 has not been evaluated, why not?
(80.0%)
4
• An evaluation was not considered to be needed
(20.0%)
1
• Other
134
Table 39: Beginning choice point practices: Factors that support practice and barriers to successful
implementation (N = 13)
Frequency (Percentage)
What are the maior factors that support practice # 1
10 (76.9%)
• Suooort from families and children
(69.2%)
9
Support from the mental health system
8
(61.5%)
Support from the substance abuse system
(53.8%)
Support
from
contract
agencies
7
•
7 (53.8%)
• Support from county government officials
(53.8%)
7
• Evaluations show that practice is successful
(46.2%)
6
Support
from
the
community
•
(46.2%)
6
• Support from state government officials
(46.2%)
6
• Stable funding
(46.2%)
6
• Support from the health system
(30.8%)
4
• Support from the criminal justice system
(30.8%)
4
• Support from foster families
(23.1%)
3
• Support from other out-of-home placement providers
What have been the barriers to successful implementation of
this practice?
5 (38.5%)
• No barriers
5 (38.5%)
• Funding is not stable
(15.4%)
2
Insufficient
support
from
child
welfare
workers
•
(7.7%)
1
• Insufficient support from management
(7.7%)
1
• Insufficient support from families and children
• Evaluations are inconclusive as to the success of the
(7.7%)
1
practice
•
•
135
Continuing Care/Ending Choice Point Practices (N = 12)
Twelve counties identified continuing care/ending choice point practices as their most
effective with children and families of color in the CWS. The 12 continuing care/ending choice
point practices identified by the counties fell into three categories of practices: I) placement related
practices (n = 6, 50.0%); 2) the Family to Family Initiative (n = 4; 33.3%); and 3) intensive and
collaborative interventions for high-risk children (n = 2; 16.7%). Please see Table 40.
Table 40: Continuin care/endin
Intensive and collaborative interventions for high risk
children
4
2
16.7%
Table 41 contains qualitative data on placement related continuing care/ending choice point
practices. The target populations of these services varies, some are targeted to all children in out-ofhome care, others are targeted specifically to Native American children, one is targeted to children
in group homes {The Village Project) or at risk for entry into group homes and one is targeted to
parents with young children likely to enter out-of-home care (Shared Family Care).
Descriptions of the practices also varied. For instance, many participants noted that the first
priority is to place children with relatives; some discussed the role of culture and ethnicity in
making placement decisions. One practice (The Village Project) is described as a multidisciplinary
effort aimed at avoiding higher levels of care for children or working toward a lower level of care.
Lastly, one practice (Shared Family Care) is aimed at providing a placement for the whole family in
a mentor, foster home.
Respondents provided a variety of opinions regarding why the practice is effective with
children and families of color. Specifically, many mentioned the importance of placing children in
culturally appropriate homes-whether with relatives or with foster parents. The Shared Family
Care program is thought to be effective with children and families of color because it allows
families with few resources to be placed together and The Village Project was considered to be
effective with children and families of color because of a multicultural staff and flexibility of
funding sources.
Table 42 provides descriptive information on the Family to Family Initiative. The target
population for this practice varies somewhat; some respondents identified children and families in
foster care as the target population, others identified care providers and resource homes or all clients
and care providers. The Family to Family Initiative is described as engaging and involving
communities in preventing out-of-home placements, reunification, and providing resource homes
within children's neighborhoods. Many also described the practice as encompassing several aspects
of the child welfare system with the goal of improving the overall child welfare system.
136
Family to Family was described as effective with children and families of color because it
involves community and family support and is aimed at reducing disparities associated with gender,
and ethnicity and increasing the number of children placed in their own communities.
Table 43 provides qualitative data on continuing/ending choice point practices that are
aimed at intensive and collaborative interventions for high-risk children. The target population for
these services is children who are at risk for a more restrictive placement. These practices were
described as collaborative in nature-involving systems such as public health, community based
organizations, probation, behavioral health and faith based organizations. Interventions are
comprehensive, intensive and culturally competent.
Intensive and collaborative interventions for high-risk children were described as being
effective with children and families of color because the services are individualized to each family's
needs and are intensive in nature These practices were also described as "engaging families in a
non-threatening and respectful manner," and assessing the types of strengths and resources already
being used by the family. Children and families of color were described as benefiting from these
services because they are respectful of culture and involve a family-driven process
Quantitative results from survey responses are provided in Table 44. Most respondents
indicated that these continuing care/ending choice point practices take place throughout the child
welfare system (n = 7, 58.3%) and are targeted to children of all ages (n = 9, 75.0%), as well as
older children (11 to 18 years) (n = 5, 41.7%), young children (4 to 10 years), (n = 3, 35.0%), and
very young children (0 to 3 years) (n= 3, 25.0%). Most are targeted to families of any ethnicity (n =
10, 83.3%), with far fewer being targeted to African American families (n = 2, 16.7%), or American
Indian families (n = 2, 16.7%). Families with substance abuse problems were also identified as a
target population (n = 9, 75.0%) as were families with a mental health problem (n = 9, 75.0%),
families with domestic violence (n = 9, 75.0%), families with income problems (n = 7, 58.3%), and
families with a disability (n = 7, 58.3%).
The most frequently noted goal of the practice is to reunify children with their birth families
(n = 10, 83.3%), followed by ensuring a permanent placement for children (n = 8, 66.7%), and
preventing an out-of-home.placement (n = 8, 66.7%). Other commonly occurring goals included
improving parenting skills (n =7, 58.3%); improving psychosocial functioning of family (n = 6,
50.0%); addressing socioeconomic problems of family (n = 6, 50.0%); and addressing housing
problems of family (n = 6, 50.0%).
Table 45 provides information on the history of the continuing care/ending choice point
practices. The majority of practices were first implemented due to new policies (n = 8, 66.7%); the
director of the child welfare system was most often identified as the person who first initiated the
practice (n = 11, 91.7%), followed by child welfare workers (n = 4, 33.3%). Many practices were
first implemented in 2001 (n = 4, 33.3%), and 33.3% (n = 4) had changed significantly since first
being implemented.
Factors related to implementation are presented in Table 46. The majority of practices are
permanent (n = 11, 91.7%), and are funded by state funds (n = 7; 58.3%). Of the practices, 41.7% (n
= 5) are mandated and 91.7% (n = 11) of counties provide social workers with training on how to
implement the practices. Of the counties, 33.3% (n = 4) contract with another agency to implement
the practices, and 58.3% (n = 7) of respondents indicated that the practice is widespread throughout
137
the child welfare system. Very few respondents indicated that there is a specific geographic
community in which the practice is targeted (n = 2, 16.7%), and only 16.7% (n = 2) indicated that
the practice has been evaluated to determine its effectiveness. Of those practices that had been
evaluated, 100.0% (n = 2) reported that the practice is effective with the target population. For
practices that had not been evaluated, 33.3% (n = 4) have not been evaluated because of insufficient
resources.
Table 4 7 provides information on factors that support the practice and barriers to successful
implementation of the practice. The most commonly mentioned supportive factors included support
from child welfare workers (n = 11, 91.7%), support from management (n = 11, 91.7%), and
support from families and children (n = 11, 91.7%). Other factors that support the practice include
support from foster families (n = 9, 75.0%); contract agencies (n = 8, 66.7%); other out-of-home
placement providers (n = 7, 58.3%), the mental health system (n = 7, 58.3%), state government
officials (n = 7, 58.3%); and county government officials (n = 7, 58.3%). Many respondents did not
note barriers to successful implementation of the practice, no barriers were reported by 16.7% (n =
2) of respondents, and 16. 7% (n = 2) reported that unstable funding is a barrier to successful
implementation.
138
Table 41: Continuing care/ending choice point practices: Placement related practices (n = 6)
Description of practice
Name of practice Target population
Culturally
sensitive
placements
All children who
require out-ofhome placement
Shared Family
Care
Mothers/Fathers
with young
children likely to
enter out of home
care
Place children
with relatives
All Children
Native American
children placed
with families
through ICWA
Native Am/ICWA
In all cases, we must first consider any willing,
able and appropriate relatives for placement. If
this is not possible, we then try to place in a home
of the same ethnicity, religion, etc. If this is not
possible, then we let the foster parents know that
their culture is an important issue, and we are
expecting them to continue to allow the children
to live in a culturally sensitive environment.
This program provides placement for the family
in a mentor home. Usually one parent (mother in
most cases) is subject to intervention by child
welfare. Rather than place the child in foster
care, it is assessed that the mother is willing to
participate in Shared Family Care and will abide
with the conditions of working and living with a
mentor family.
Identify as quickly as possible family members
for placement
Attempt to place all children with families.
Counseling/assessment/service orientated.
Why effective with children and families
of color
Our county is predominately Caucasian,
so children of color are already truly a
minority. Placing them in a home that is
sensitive to that is very important.
For families with few resources it allows
for placement of child and parent
together. Most of our families with few
resources are those living in communities
impacted by poverty, drugs, and crimes.
Families of color primarily live in these
areas.
Children are kept in their communities
with relatives of same culture and
practices promote stability for children
They are treated the same as all other
cases---dignity--compassion-provided
services- acknowledgement of their
concerns and problems.
139
Table 41 continued: Continuing care/ending choice point practices: Placement related practices (n = 6)
Why effective with children and families
Name of practice Target population Description of practice
of color
SB163
Kids in group
Multidisciplinary effort to work with families and
A multicultural staff in the county and
program-"The
homes level 13 or
kids to avoid higher levels of care or get them
in the community-based providers and
flexibility of funding sources-money
14 or at risk for
village project"
into a lower level of care-like a foster home or
going into a group back into their home. The staff are bilingual and
can be used for activities outside the
home
bi-cultural and they are able to make connections group home, so it can be used for things
with multi-cultural families. They also work hand to prevent a group home placement and
and hand with a parent advocacy group-and they not just for kids already in a group home.
advocate for children and families especially kids
with educational difficulties.
ICWANative Americans When a Native American child is removed from
We are able to place many of our Native
placement of
the home of parents, we coritact any relatives in
American children in homes that comply
Native American
order to place the child in compliance with
with ICWA.
children
ICWA.
140
Table 42: Continuing care/ending choice point practices: Family to Family Initiative (n = 4)
Name of practice Target population
Description of practice
Family to Family
All children in out
ofhome
placements
Family to Family
Initiative
All clients and
care providers
Family to Family
Initiative
Children &
families in foster
care
Family to family
Resource homes
and the
community
Regional initiatives underway to work with local
communities to prevent removal and/or maintain
children in their community/school if they must
be removed. Community collaboration on child
safety and well being.
The use of the following Family to Family
strategies are applied to all aspects of child
welfare placement practices in this County: Team
decision making, Recruitment, Retention, and
Support of Care Providers, Self-Evaluation and
Community Partnership.
Community-based, family focused and designed
to meet 9 key outcomes toward improving our
foster care system including reducing disparities
associated with ethnicity, J2;ender & age.
This is an effort to engage the community in
providing resource homes for foster children who
are being placed in homes from their
neighborhoods in which they are being removed.
The community is being engaged and as well as
community based organizations.
Why effective with children and families
of color
Community and family
involvement/support.
Missing
Outcomes include: reduction in
disparities associated with sex, gender &
ethnicity, increasing kids placed in their
own neighborhoods & communities,
Missing
141
Table 43: Continuing care/ending choice point practices: Intensive and collaborative interventions for high-risk children (n = 2)
Name of practice Target population
Description of practice
Why effective with children and families
of color
Children who are
Children's
To develop a child 'who is emotionally
As with all families, intensive services,
in danger of being disturbed/behaviorally disturbed' through a
System of Care
individualized plans
placed in a higher collaborative interagency system of care with
level of care i.e.
services that are "youth-centered;" family
Group homes
involved & focused; culturally competent;
community based and comprehensive.
This practice really values families individually
Wraparound
Children who are
The practice engages with families in a
services
placed or at risk of and values their culture. It provides intensive and respectful and non-threatening manner
placement into
supportive in-home services to keep the child at
and asking them what supports, what
level 14 group
people, in your life and in your
home or return them home. It's also been a
home care
collaborative effort between probation, behavioral community that are important to you and
health, public health, CBO's, and faith-based
how can we incorporate those aspects of
organizations.
your life into this child and family plan to
help your child stay safely at home. And
by asking those questions and evaluating
what the family uses to thrive and asking
them to make that a part of the service
planning for their child is kind of a
foreign way of working with the families.
It is a family-driven, needs based process,
rather than a professionally-driven
process. There is respect for the family's
culture.
142
Table 44: Continuing care and ending choice point practices:
Location, target population and goals (N = 12)
Where does practice #1 take place?
•
Throughout the child welfare svstem
• County child welfare office
• Birth family's home
• Foster family's home
• Court
• Another out-of-home placement site
• Other
Who is the target population for practice # 1?
• Children of all ages
• Very young children (0 to 3 years)
• Young children (4 to 10 years)
• Older children (11 to 19 years)
• Families of any ethnicity
• African American families
• American Indian families
• Asian American/Pacific Islander families
• Latino families
• White families
• Families of another ethnicity
Families with substance abuse problems
Families with a mental health problem
• Families with domestic violence
• Families with income problems
• Families with a disability
• Other
What are the purposes and goals of practice # 1?
• Reunify children with their birth families
• Ensure a permanent placement for the child
• Prevent out-of-home placement
• Improve parenting skills
• Improve psychosocial functionin~ of familv
• Address socioeconomic problems of family
• Address housing problems of familv
• Address substance abuse problems of familv
• Address mental health problems of family
Assess level of risk
• Other
•
•
•
Frequency (Percentage)
7
4
4
4
4
4
5
(58.3%)
(33.3%)
(33.3%)
(33.3%)
(33.3%)
(33.3%)
(41.7%)
9
(75.0%)
(25.0%)
(35.0%)
(41.7%)
(83.3%)
(16.7%)
(16.7%)
(8.3%)
(8.3%)
(8.3%)
(16.7%)
(75.0%)
(75.0%)
(75.0%)
(58.3%)
(58.3%)
(50.0%)
3
3
5
10
2
2
1
1
1
2
9
9
9
7
7
6
10
8
8
7
6
6
6
5
5
4
4
(83.3%)
(66.7%)
(66.7%)
(58.3%)
(50.0%)
(50.0%)
(50.0%)
(41.7%)
(41.7%)
(33.3%)
(33.3%)
143
Table 45: Continuing care and ending choice point practices: History of practice (N = 12)
Frequency (Percentage)
Why was practice # I first implemented?
(66.7%)
8
• New policies were implemented
(25.0%)
3
• Funds became available
(16.7%)
2
In
response
to
a
particular
problem
•
(50.0%)
6
• Other
Who first initiated practice # 1?
11 (91.7%)
• Director of the child welfare system
4 (33.3%)
• Child welfare workers
2 (16.7%)
• Mental health system
(8.3%)
1
• Foster families
(8.3%)
1
Substance
abuse
system
•
(8.3%)
1
• Contract agencies
(8.3%)
1
• Criminal justice system
4 (33.3%)
• Other
When was practice # 1 first implemented?
•
•
•
•
2003
2002
2001
2000
Has practice # 1 changed significantly between the time it
was first implemented and now?
• Yes
1
2
4
2
(8.3%)
(16.7%)
(33.3%)
(16.7%)
4
(33.3%)
144
Table 46: Continuing care and ending choice point practices:
Factors related to implementation (N = 12)
Current implementation status of practice # 1
• Permanent
• Demonstration
What is the primary funding source for practice # 1?
• State funds
• County funds
• Federal funds
• Nongovernmental grant
• Other
Is practice # 1 mandated?
• Yes
Are social workers provided with training in how to
implement practice #1?
• Yes
Do you contract with another agency or system to deliver
practice # 1?
• Yes
• No
• Combination of internal and external resources
Is practice # 1?
• Widespread throughout the child welfare system
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
• Yes
Has practice #1 been evaluated to determine its
effectiveness?
• Yes
If yes, what were the results of the evaluation?
• Effective with target population
If practice #1 has not been evaluated, why not?
• Insufficient resources
• Insufficient time
• An evaluation was not considered to be needed
• Other
Frequency (Percentage)
11
1
(91.7%)
(8.3%)
7
6
3
2
3
(58.3%)
(50.0%)
(25.0%)
(16.7%)
(25.0%)
5
(41.7%)
11
(91.7%)
4
4
3
(33.3%)
(33.3%)
(25.0%)
7
5
(58.3%)
(41.7%)
2
(16.7%)
2
(16.7%)
2
(100.0%)
4
2
1
3
(33.3%)
(16.7%)
(8.3%)
(25.0%)
145
Table 47: Continuing care and ending choice point practices: Factors that support practice and
barriers to successful implementation (N = 12)
Frequency (Percentage)
What are the major factors that support practice # 1
11 (91.7%)
• Support from child welfare workers
11 (91.7%)
• Support from management
(91.7%)
11
• Support from families and children
(75.0%)
9
• Support from foster families
(66.7%)
8
• Support from contract agencies
(58.3%)
7
• Support from other out-of-home placement providers
(58.3%)
7
Support from the mental health system
(58.3%)
7
Support from state government officials
(58.3%)
7
Support from county government officials
(50.0%)
6
• Stable funding
(50.0%)
6
Support
from
the
community
•
(50.0%)
6
• Support from the health system
(50.0%)
6
Support
from
criminal
justice
system
•
(41.7%)
5
• Support from the substance abuse system
(25.0%)
3
• Evaluations show that practice is successful
(16.7%)
2
• Other
What have been the barriers to successful implementation of
this practice?
(16.7%)
2
• No barriers
(16.7%)
2
• Funding is not stable
(8.3%)
1
• Insufficient support from state government officials
Insufficient
support
from
county
government
•
(8.3%)
1
officials
(8.3%)
1
• Insufficient support from child welfare workers
(8.3%)
1
• Insufficient support from foster families
(8.3%)
1
Insufficient
support
from
the
community
•
(8.3%)
1
• Insufficient support from the criminal justice system
(50.0%)
6
• Other
•
•
•
146
Both Beginning and Continuing/Ending Choice Point Practices (N = 4)
Four respondents identified practices that were related to both beginning and
continuing/ending choice points as their most effective practices for children and families of
color in the CWS. The four beginning and continuing/ending choice point practices fell into two
categories: 1) community-based services (n = 3, 75.0%); and 2) strength based assessments (n =
I, 25.0%). Please see Table 48.
Table 48: Both beginning and continuing/ending choice point practices: Types of practices (N =
4
Table 49 provides qualitative information on community based services. The target
population for these services includes low-income populations, families at risk for entry into the
child welfare system, families already in the child welfare system, and the community. These
practices were described as being located in the communities in which clients reside and include
a wide range of services such as, health, substance abuse, public assistance, domestic violence
and home parenting services. Practices were also described as collaborative and aimed at
preventing an out-of-home placement or reunifying children with their families.
These practices were described as effective with children and families of color because
services are "user friendly;" clients have easy access to many services that are housed in one
location. Community based services are free and outreach for these services is performed
through advertisements into different communities. Moreover, because these services are
typically located within clients' own communities, services are "community-specific" and clients
may run into their own community members while accessing services, thus making the services
less intrusive for clients. Additionally, these practices were described as flexible and able to
address "changing community needs."
Table 50 provides information on family-centered strength based assessments. This
. practice was described as being targeted to Hispanic, African American, American Indian and
Asian American/Pacific Islander clients. It is focused on family strengths with an empowerment
perspective.
Quantitative results from survey responses are presented in Table 51. Practices that are
both beginning and continuing/ending practices typically take place at a contract agency site (n =
2, 50.0%), or the birth family's homes (n = 2, 50.0%). One hundred percent (n = 4) of these
practices are targeted to: children of all ages, families of any ethnicity, families with substance
abuse problems, families with a mental health problem, families with domestic violence, families
with income problems, or families with a disability. Seventy five percent (n = 3) ofpractices
have the following goals: preventing an out-of-home placement; reunifying children with their
birth families; assessing level of risk; improving psychosocial functioning of families; improving
parenting skills; addressing substance abuse problems; addressing mental health problems;
addressing socioeconomic problems; and addressing housing problems.
147
Information on the history of the practices is presented in Table 52. The practices were
first implemented in response to a particular problem (n = 1, 25.0%); funds becoming available
(n = 1, 25.0%); or new policies being implemented (n = 1, 25.0%). Child welfare workers (n = 2,
50.0%); the child welfare director (n = 2, 50.0%); and contract agencies (n = 2; 50.0%) were
identified as having first initiated the practices. The year the practice was first implemented
ranged from 1967 to 1997, and 66.7% (n = 2) of respondents indicated that the practice had
changed significantly since it was first implemented.
Factors related to the implementation of both beginning and continuing/ending choice
point practices are presented in Table 53. All practices (n = 4, 100.0%) are permanent and most
are primarily funded through federal funds (n = 4, 100.0%). Fifty percent of respondents (n = 2)
indicated that the practice is mandated and 33.3% contract with another agency to deliver the
practice. Seventy five percent of respondents indicated that the practice is widespread throughout
the child welfare system. All respondents (n = 4, 100.0%) reported that the practice has been
evaluated, and 50.0% (n = 2) indicated that evaluations demonstrated that the practice is effective
with the target population.
Factors that support the practice and barriers to successful implementation are presented
in Table 54. One hundred percent of respondents (n = 4) indicated that support from the
following entities are major factors that support the practices: community, county government
officials, mental health system, substance abuse system, health system, and families and
children. Seventy five percent of respondents (n = 3) indicated that unstable funding is a barrier
to successful implementation of the practices.
148
Table 49: Both beginning and continuini /ending practices: Community based services
Target population
Description of practice
Name of
practice
Community- Low-income
Federal and state funded programs that are free to
populations
low-income clients, including the health
based
department, substance abuse treatment, Aids
services
department, WIC, immunizations, planned
parenthood. Services include domestic violence
support, clothing donations, utility assistance,
HEAD Start program; sexual assault crisis, and in home parenting programs, and teenage parenting
programs.
Children and Families at There are many Family Resource centers. A
Family
Resource
risk for entry into the
collaborative of agencies applies for funds from the
CWS and children and
funding source (though RFP). The Family
Centers
families already in
Resource Centers have a number of agencies
CWS. Most clients are
working together within one site - so it provides
Hispanic, the second
just one-stop for families to receive services. The
Family Resources Centers are located in the
most are API
communities and the neighborhoods where the
clients live. Services are meant to be accessible and
less intimidating than county government services.
Family
Community, Families
Commmunity specific family based services located
in six geographical areas in the County with embed
and Children
Resource
access to public child welfare, mental health,
Centers
alcohol and other drug and physical health services.
Why effective with children and families of
color
Because services are free and they advertise
into all areas of the community, and there are
satellite offices in many locations; so clients
have access to the services, which is
extremely helpful. They also cross-refer so we
all depend on one another. We have a large
Hispanic popuiation so information is
available in Spanish. Transportation is
accessible.
Because wherever the families are they can
reach out to the resources that meet their
need-there are a multitidue of agencies in
one place-it's a one-stop shop. Clients are
more likely to run into their own community
members at the Family Resource Centers. It is
user friendly and less intrusive than other
services.
Practice is community specific with flexibility
to address changing community needs.
Subsequently,_ it offers increased accessibility
to services, intervention/prevention and
family support and advocacy.
149
Table 50: Both beginning and continuin1 /ending practices: Strength based assessments
Target population
Description of practice
Name of
practice
Family
Hispanic, AA, American Focused on family strengths with the empowerment
Indian and Asian
perspective.
centered
strength
based
assessments
Why effective with children and families of
color
Missing
150
Table 51: Both beginning and continuing care/ending choice point practices: Location, target
population and goals (N = 4)
Frequency (Percentage)
Where does practice #1 take place?
2
(50.0%)
• Contract agency site
2
(50.0%)
• Birth family's home
1
(25.0%)
County
child
welfare
office
•
(25.0%)
1
• Foster family's home
(25.0%)
1
Another
out-of-home
placement
site
•
2
(50.0%)
• Other
Who is the target population for practice # 1?
4 (100.0%)
• Children of all ages
4
(100.0%)
• Families of any ethnicity
4 (100.0%)
• Families with substance abuse problems
4 (100.0%)
• Families with a mental health problem
(100.0%)
4
Families
with
domestic
violence
•
4 (100.0%)
• Families with income problems
4 (100.0%)
• Families with a disability
(25.0%)
1
Other
What are the purposes and goals of practice # 1?
3 (75.0%)
• Prevent out-of-home placement
3 (75.0%)
• Reunify children with their birth families
3 (75.0%)
• Assess level of risk
3 (75.0%)
• Improve psychosocial functioning of family
(75.0%)
3
Improve
parenting
skills
•
3 (75.0%)
• Address substance abuse problems of family
(75.0%)
3
• Address mental health problems of family
3 (75.0%)
• Address socioeconomic problems of family
3 (75.0%)
• Address housing problems of family
2
(50.0%)
Ensure
a
permanent
placement
for
the
child
•
3 (75.0%)
• Other
•
151
Table 52: Both beginning and continuing care/ending choice point practices:
History of practice (N = 4)
Frequency (Percentage)
Why was practice #1 first implemented?
1 (25.0%)
• In response to a particular problem
1 (25.0%)
• Funds became available
1 (25.0%)
• New policies were implemented
1 (25.0%)
• Other
Who first initiated practice #1?
2 (50.0%)
• Child welfare workers
(50.0%)
2
• Director of the child welfare system
2 (50.0%)
• Contract agencies
1 (25.0%)
Clients of the CWS
1 (25.0%)
Foster families
1 (25.0%)
Out-of-home placement providers
(25.0%)
1
• Mental health system
1 (25.0%)
• Substance abuse system
1 (25.0%)
• Health system
(25.0%)
1
Criminal
justice
system
•
2 (50.0%)
• Other
When was practice #1 first implemented?
1 (33.3%)
• 1997
1 (33.3%)
• 1995
1 (33.3%)
• 1967
Has practice # 1 changed significantly between the time it
was first implemented and now?
2 (66.7%)
Yes
•
•
•
•
152
Table 53: Both beginning and continuing care/ending choice point practices: Factors related to
implementation (N = 4)
Frequency (Percentage)
Current implementation status of practice # 1
4 (100.0%)
• Permanent
What is the primary funding source for practice # 1?
4 (100.0%)
• Federal funds
(50.0%)
2
State
funds
•
(50.0%)
2
• County funds
(25.0%)
1
• Nongovernmental grant
Is practice # 1 mandated?
(50.0%)
2
• Yes
Are social workers provided with training in how to
implement practice #1?
(50.0%)
2
• Yes
Do you contract with another agency or system to deliver
practice #1?
(33.3%)
1
• Yes
(33.3%)
1
No
•
1 (33.3%)
• Combination of internal and external resources
Is practice #1?
(75.0%)
3
• Widespread throughout the child welfare system
(25.0%)
1
Targeted
to
a
relatively
small
group
of
clients
•
Is there a specific geographic area or community in which
this practice is targeted?
(50.0%)
2
• Yes
Has practice #1 been evaluated to determine its
effectiveness?
4 (100.0%)
• Yes
If yes, what were the results of the evaluation?
(50.0%)
2
Effective with target population
• Evaluation did not provide evidence that practice was
(25.0%)
1
effective
1 (25.0%)
• Other
•
153
Table 54: Both beginning and continuing care/ending choice point practices: Factors that support
practice and barriers to successful implementation (N = 4)
Frequency (Percentage)
What are the maior factors that support practice # I
4 (100.0%)
• Support from the community
4 (100.0%)
• Support from county government officials
4
(100.0%)
• Support from the mental health system
4 (100.0%)
• Support from substance abuse svstem
4 (100.0%)
• Support from health svstem
4 (100.0%)
Support
from
families
and
children
•
3 (75.0%)
• Support from child welfare workers
(75.0%)
3
• Support from management
3 (75.0%)
• Support from foster families
3 (75.0%)
• Support from contract a,zencies
3 (75.0%)
• Support from the criminal iustice system
(75.0%)
3
• Evaluations show that practice is successful
(50.0%)
2
• Support from state government officials
(25.0%)
1
• Stable funding
(25.0%)
1
• Support from other out-of-home placement providers
What have been the barriers to successful implementation of
this practice?
3 (75.0%)
• Funding is not stable
1 (25.0%)
• Insufficient suooort from state government officials
• Evaluations are inconclusive as to the success of the
1 (25.0%)
practice
154
Organizational or Structural Practices (N = 14)
Fourteen respondents identified organizational or structural practices as their most
effective practices for children and families of color in the CWS. The fourteen
organizational/structural practices fell into three categories: 1) practices related to cultural
competency of staff (n = 10, 71.3%), 2) collaboration with American Indian tribes (n = 3,
21.3%), and 3) vertical case management (n = 1, 7.4%).
Table 55: Or anization or structural
Collaboration with American Indian Tribes
Vertical case mana ement
3
1
Qualitative data on cultural competency practices are presented in Table 56. The target
population for these services varies, some are focused on the Latino population or the Spanish
speaking population, while others are focused on all children in the system and one is focused on
social workers themselves. Practices related to cultural competency involve the use of
bilingual/bicultural staff or competent translators in the delivery of services, as well as providing
information on civil rights, providing training to social workers and matching clients and
workers on ethnicity/culture.
Many respondents mentioned the importance of having bilingual/bicultural staff and the
need to be able to communicate with clients in their own language as reasons for why these
practices are effective with children and families of color. The importance of assessing and
delivering services in the clients' language is described as resulting in more accurate
assessments, more meaningful services and better connections between worker and client.
Cultural competency trainings for social workers are described as bringing awareness and
knowledge to staff.
Table 57 provides qualitative data on organizational or structural practices related to
collaboration with American Indian tribes. These practices are targeted to Native American
children and families and generally include multidisciplinary teams involving tribal
representatives, legal professionals and child welfare staff who come together to better serve
Native American children and ensure the proper implementation of the Indian Child Welfare
Act.
Practices related to collaboration with American Indian tribes were described as effective
for children and families of color because they respect cultural heritage; encourage cultural
awareness for Native American children; result in more complete assessments; better service
planning and fewer contested hearings.
· Table 58 provides qualitative information on vertical case management. This practice is
targeted toward all clients of the child welfare system and it involves having social workers carry
the same cases all the way through the system. This practice was described as effective with
children and families of color because there is consistency in the case and a better relationship
155
between worker and client. It was also noted that this practice is effective with the overall
·
population as well.
Quantitative data from survey responses is provided in Table 59. Most
organizational/structural practices take place throughout the child welfare system (n = 8, 57.1 %)
and are targeted to children of all ages (n = 12, 85.7%). Many are targeted to families of any
ethnicity (n = 7, 50.0%), while some are targeted to Latino families (n = 4, 28.6%), or American
Indian families (n = 3, 21.4%). Additionally, families with substance abuse problems are also a
focus of these practices (n = 7, 50.0%), as are families with mental health problems (n = 7,
50.0%). Goals of the practice include preventing an out-of-home placement (n = 7, 50,0%);
ensuring a permanent placement (n = 7, 50.0%); reunifying children and families (n = 6, 42.9%);
addressing substance abuse problems of families (n = 6,42.9%); and improving parenting skills,
(n = 6, 42.9%).
Information on the history of the practices is presented in Table 60. Most practices were
first implemented in response to a particular problem (n = 7, 50.0%); and the director of the child
welfare system was most frequently mentioned as the person who first initiated the practice (n =
7, 50.0%), followed by child welfare workers (n = 6, 42.9%). The year the practice was first
implemented range from 1972 to 2002 and 46.2% of counties indicated that the practice had
changed significantly since it had first been implemented.
Table 61 provides data on factors related to implementation. Most practices are
permanent (n = 12, 92.3%) and the majority are funded through federal dollars (n = 8, 57.1 %).
Most counties also indicated that these practices are mandated (n = 8, 61.5%), and 66.7% (n = 8)
of counties indicated that social workers are provided with training in how to implement the
practice. Most counties do not contract with another agency to deliver the practice (n = 9, 64.3%)
and most also indicated that the practice is widespread throughout the child welfare system (n =
12, 92.9%). Only one county (7.1 %) indicated that there is a specific geographic area in which
the practice is targeted. Four counties (28.6%) reported that the practice has been evaluated to
determine its effectiveness and of those 66.7% (n = 2) reported that the practice was found to be
effective with the target population. Of those counties in which no evaluation has been
conducted, 50.0% (n = 5) indicated that an evaluation was not needed.
Factors that support the practices and barriers to successful implementation are presented
in Table 62. Support from child welfare workers (n = 11, 78.6%) was most commonly noted as a
major factor that supports the practice, followed by support from management (n = 9, 64.3%);
county government officials (n = 8, 57.1 %); families and children (n = 7, 50.0%); and the
community (n = 7, 50.0%). No barriers to successful implementation were noted by 28.6% of
counties (n = 4), while 14.3% (n = 2) reported insufficient support from the community and
14.3% (n = 2) noted insufficient support from child welfare workers.
156
Table 56: Organizational or structural practices: Practices related to cultural competency (n = 10)
Why effective with children and families of color
Description of practice
Target
Name of practice
population
One of things that has made it particularly
SpanishWe have a very high rate of return for
The use of
Hispanic children--one of our better
effective in this county is the size of the county.
competent
speaking
practices is the use of competent translators And our translator even though they may not be
translators
clients
and always having this available when
CW social workers they are from the community
and they are familiar with the community and the
working with a family. These are used in
the juvenile court as well. And we also have dynamics of that particular culture. For the most
part this had enhanced the practice.
bilingual staff as well.
Availability of
bilingual/bicultural
staff to serve the
Latino population.
Latino
population
At least one
bicultural/bilingual
worker in each unit
Mostly
Hispanic
children
B ilingual/B icultural
social work staff.
Spanish
speaking
population
Cultural knowledge and ability to converse in same
Hire and place bilingual/bicultural social
workers at office serving majority of Latino language.
families. Provides culturally sensitive
services to families, better assuring accurate
asessments of risk and safety for children.
When you are doing social work and having to
The agency really values language and
trying to accommodate that the best we can. assess and discuss very difficult things with people
Most of our receptionists are bilingual and
having to do with emotions and feelings these
things are very difficult to do through translation
our social service aides and social worker
because a lot of what a social worker does is their
I's are bilingual. Whenever we start to
use of self and being able to read body language
implement a program we always look to
bringing in bilingual/bicultural staff. We are and intonation and understand the subtlety of
limited in our resources but we do our best. language and what someone is saying. You are less
effective through translation. The communication
We also really try to ask this of our
is much better when services are conducted in the
providers as well.
clients language. And bicultural folks understand
the local jargon.
Ability to communicate with the families in their
Bilingual/bicultural staff assigned to
native language & an ability to understand the
bilingual/monolingual cases to ensure_
appropriate communication of issues, needs, culture helps the social worker provide appropriate
services and agency function/process.
resources & guidance to the family.
157
Table 56 continued: Organizational or structural practices: Practices related to cultural competency (n = 10)
Target
Name of practice
Why effective with children and families of color
Description of practice
population
B ilingual/B icultural Families
Recruiting and hiring bilingual/cultural staff Recruitment is a priority within service delivery
system and best practice philosophy .. .it becomes
staff
whose
to :work with the diverse population we
children
serve. Staff are assigned throughout our
integrated in every day work.
have been
child welfare programs/services.
Assessment of language and cultural
abused.
staffing needs is determined on an ongoing
basis and is used as a criteria to fill vacant
positions.
Federal Division 21
Regulation-Civil
rights
All
Emphasis of
Cultural
Competence
All cultures
We inform all clients of their civil rights.
This includes that they cannot be
discriminated against on the basis of
national origin, disabilities, race, age sex,
all those areas. We are supposed to talk to
all families about their civil rights. We ask
the family if they want a special skills
worker assigned-that would specialize in a
particular language or culture. We have
Black culture workers, Hispanic cultural
workers, and Asian, Russian special skills
workers. We ask all clients if they want a
special skills worker-this is for
language-but also culturally matching the
clients.
Our agency has a cultural competence
committee. We mandate 8 hours of training
for staff in orientation and an additional 16
hrs through our leadership academy.
Matching families with a social worker of their
ethnic background is helpful in getting families
back together. If the social worker understands the
culture-like ifthere are different discipline
techniques that are seen more in one culture, the
social worker may be able to help the child and
family more. Clients are a bit more comfortable,
social workers can confront them better-work
with the better to help bring about change.
Awareness, Knowledge
158
Table 56 continued: Organizational or structural practices: Practices related to cultural competency (n = 10)
Target
Name of practice
Description of practice
Why effective with children and families of color
population
Policy is to align
Applies
Our practice is to assign staff of similar
Ability to establish trusting relationship
case assignments to across all
ethnicity and we are able to do so most
staff from like
cws
often in our African American and Mexican
ethnicity when
components American population. If unable to do so,
we consider assigning a spanish speaking,
possible
non-Mexican-American. staff.
Cultural
Social
We require all social workers to attend
Missing
competence
workers
training on cultural competency on a regular
training
basis.
Missing
Civil Rights
Missing
Training- bringing awareness to staff
Training
159
Table 57: Organizational or structural practices: Collaboration with American Indian tribes (n = 3)
Why effective with children and families of color
Name of practice
Target
Description of practice
population
It respects their cultural background and keeps
Indian Child
American
A practice in which tribal representatives,
Indian
agency staff, kids' attorneys, County
them connected with their tribal roots, it respects
Welfare Act
that people can do things differently and still have
(ICWA)
children and Counsel, and others come together on a
monthly
basis
to
talk
about
issues
related
to
good outcomes. The tribe is actively solicited to be
Roundtable
families
involved with children and families in the CWS.
the implementation of the ICWA, how we
We pay for culturally appropriate services, like
can work to together on certifying tribal
tribal dances, and other non-traditional services.
homes, how to get services for American
Indian Children and what are the most
appropriate services. The group created a
handbook ofICWA regulations. Very
effective in opening dialogue between the
parties, sensitizing staff to issues around
services that would be more culturally
appropriate to American Indian children,
identifying and providing services, and
ensuring that kids can get to tribal homes.
Native
Indian Child Welfare Act Multidisciplinary More complete assessments, better service
ICWA
planning and less contested hearings.
Team
American
Native American families-in general they are
Some of the tribes have strong cultural
Collaboration with
Native
Americans
centers with health care and counseling and still at war with both the US and with each other. It
tribes and tribal
is effective because in spite of these things we are
so we try to work hard with them. Native
courts
able to overcome the barriers and we are able to
American children do best when placed
make use of the real resources of cultural
with a Native American family. The Tribal
court makes placement decisions and so we awareness for Native American people.
work with them to make decisions that we
Collaborating with the tribes depends on having the
are all in agreement about. We work closely right social worker work with the tribes-in
with the tribes in Native American cases.
general the tribes have been cooperative with the
cws.
160
Table 58: Organizational or structural practices: Vertical case management {n = 1)
Name of practice
Target
Description of practice
Why effective with children and families of color
population
Social Workers carry the same cases all the There's consistency-a lot of it is when cases are
Social workers
All
carry a generic case
way through the system A lot is lost when a transferred from one person to another there may
load-vertical case
case is transferred from worker to worker.
be different interpretations of the situation.
management
Families always know who their social
Effective with overall population-it would be
workers are and there is ownership
unfair to say that what we are doing is more
advantageous with group than another.
throughout the life of the case. There is a
better work product. There's a relationship
between the worker and clients.
161
T abl e 59 0 rgamzattona or structural practices: Location, tan et population and goals (N = 14)
Frequency (Percentage)
Where does practice #1 take place?
(57.l %)
8
• Throughout the child welfare system
County
child
welfare
office
(50.0%)
7
•
(42.9%)
6
• Birth family's home
Foster family's home
(42.9%)
6
(42.9%)
Court
6
5
(35.7%)
• Another out-of-home placement site
(14.3%)
2
• Contract agency site
5 (35.7%)
• Other
Who is the target population for practice # 1?
(85.7%)
12
Children of all ages
(7.1%)
1
Very
young
children
(0
to
3
years)
•
(7.1%)
I
• Young children (4 to 10 years)
(50.0%)
7
Families
of
any
ethnicity
•
(28.6%)
4
Latino families
(21.4%)
3
American Indian families
(50.0%)
7
• Families with substance abuse problems
(50.0%)
7
• Families with a mental health problem
(42.9%)
6
• Families with domestic violence
(28.6%)
4
• Families with income problems
(21.4%)
3
• Families with a disability
(7.1%)
1
Other
What are the purposes and goals of practice #1?
(50.0%)
7
• Prevent out-of-home placement
(50.0%)
7
• Ensure a permanent placement for the child
(42.9%)
6
Reunify children with their birth families
(42.9%)
6
• Address substance abuse problems of family
(42.9%)
6
• Improve parenting skills
(35.7%)
5
Assess
level
of
risk
•
(35.7%)
5
• Improve psychosocial functioning of family
(35.7%)
5
Address
mental
health
problems
of
family
•
5 (35.7%)
• Address housing problems of family
(28.6%)
4
• Address socioeconomic problems of family
(57.1%)
8
Other
•
•
•
•
•
•
•
•
162
Table 60: Organizational or structural practices: History of practice (N = 14)
Frequency (Percentage)
Why was practice #1 first implemented?
7
(50.0%)
In response to a particular problem
(21.4%)
3
• New policies were implemented
1
(7.1%)
• Funds became available·
6
(42.9%)
• Other
Who first initiated practice # 1?
(50.0%)
7
• Director of the child welfare system
(42.9%)
6
• Child welfare workers
(7.1%)
1
Clients of the CWS
(7.1%)
1
• Foster families
(7.1%)
1
• Criminal justice system
1
(7.1%)
• Contract agencies
(28.6%)
4
• Other
When was practice #1 first implemented?
(7.1%)
1
• 2002
2 (14.3%)
2000
(7.1%)
1
• 1999
2
(14.3%)
• 1998
(7.1%)
1
• 1995
(7.1%)
1
• 1980
(7.1%)
1
1972
•
Has practice # 1 changed significantly between the time it
was first implemented and now?
(46.2%)
6
• Yes
•
•
•
163
Table 61: Organizational or structural practices: Factors related to implementation (N = 14)
Frequency (Percentage)
Current implementation status of practice # 1
12 (92.3%)
• Permanent
Other
1
(7.7%)
What is the primary funding source for practice # 1?
8
(57.1%)
• Federal funds
(42.9%)
6
State
funds
•
5 (35.7%)
• County funds
(28.6%)
4
• Other
Is practice #1 mandated?
8
(61.5%)
• Yes
Are social workers provided with training in how to
implement practice # 1?
(66.7%)
8
• Yes
Do you contract with another agency or system to deliver
practice # 1?
(64.3%)
9
• No
(21.4%)
3
Combination
of
internal
and
external
resources
•
(14.3%)
2
• Yes
Is practice # 1?
(92.9%)
13
• Widespread throughout the child welfare system
(7.1%)
1
Targeted
to
a
relatively
small
group
of
clients
•
Is there a specific geographic area or community in which
this practice is targeted?
(7.1%)
1
• Yes
Has practice # 1 been evaluated to determine its
effectiveness?
4
(28.6%)
• Yes
If yes, what were the results ofthe evaluation?
(66.7%)
2
• Effective with target population
• Evaluation did not provide evidence that practice was
1
(33.3%)
effective
If practice # 1 has not been evaluated, why not?
(35.7%)
5
• An evaluation was not considered to be needed
(21.4%)
3
• Insufficient resources
(7.1%)
1
Insufficient
time
•
(7.1%)
1
• Other
•
164
Table 62: Organizational or structural practices: Factors that support practice and barriers to
successful implementation (N = 14)
Frequency (Percentage)
What are the major factors that support practice # I
11 (78.6%)
• Support from child welfare workers
9 (64.3%)
Support
from
management
•
8 (57.1%)
• Support from county government officials
(50.0%)
7
• Support from families and children
7 (50.0%)
Support from the community
6 (42.9%)
• Support from foster families
6 (42.9%)
Support
from
state
government
officials
•
5 (35.7%)
• Support from the mental health system·
4
(28.6%)
• Support from the substance abuse system
4 (28.6%)
• Support from contract agencies
4 (28.6%)
• Support from the health system
(28.6%)
4
Stable
funding
•
3 (21.4%)
• Support from other out-of-home placement providers
(14.3%)
2
• Evaluations show that practice is successful
2 (14.3%)
• Support from the criminal justice system
2 (14.3%)
Other
What have been the barriers to successful implementation of
this practice?
-4 (28.6%)
• No barriers
2 (14.3%)
• Insufficient support from the community
2 (14.3%)
• Insufficient suooort from child welfare workers
(7.1%)
I
• Funding is not stable
(50.0%)
7
Other
•
•
•
165
Summary of Key Findings and Implications:
Descriptive Results
Respondents to the California statewide survey identified practices in four overall areas
as most effective for children and families of color: 1) beginning choice point practices, 2)
continuing or ending choice point practices, 3) both beginning and continuing/ending choice
point practices, and 4) organizational or structural practices.
·
Beginning Choice Point Practices
The beginning choice point practices identified by respondents fell into three categories:
1) family group conferencing or other team decision-making practices, 2) voluntary family
maintenance or other prevention/family preservation services and 3) risk assessment tools or
systems.
In general, results from the California statewide survey indicate that beginning choice
point practices are most often delivered at the county child welfare office, and they are targeted
to children ofall ages and ethnicities. Most of these practices are aimed at preventing an out-of-
home placement, assessing level of risk or improving the psychosocial functioning of the family.
Over half of these practices have been evaluated by individual counties to determine their
effectiveness, and of those that have been evaluated most are considered to be effective with the
target population.
Results ofthe California statewide survey indicate that beginning choice point practices
are considered to result in enhanced outcomes for children and families ofcolor because they
are inclusive, collaborative and objective. Yet very little empirical evidence exists to support the
effectiveness of these practices for children and families of color. No studies have evaluated the
impact offamily group conferencing or other team decision-making practices on children and
families ofcolor. However there is some evidence to suggest that inclusive practice, in which
parents are engaged in placement and other decisions, may result in more placement stability for
children, yet possible differential effects for diverse populations has not been researched
(Palmer, 1996). More research is needed to determine if family group conferencing/other team
decision-making practices and inclusive practice result in enhanced outcomes for children and
families of color.
Additionally, the effectiveness of voluntary family maintenance or other prevention/family
preservation services in preventing out-of-home placements has not been adequately
demonstrated, and the impact of these services for children and families of color has not been
sufficiently addressed in the research literature. It is interesting to note that results from the
California statewide survey indicate that only 15.4 percent of beginning choice point practices
take place in the birth family's home, yet the research literature suggests that home-based
prevention practices may be the most effective voluntary family maintenance practice, especially
for high-risk mothers. This finding may suggest that counties are not utilizing evidence-based
prevention efforts. Research on the reasons for this underutilization of home-based prevention
services is warranted.
166
Lastly, the effectiveness ofrisk assessment tools or systems in accurately assessing risk
for future maltreatment has not been consistently demonstrated in the research literature. Yet
research does indicate that the individual decision-making practices ofchild welfare workers are
inconsistent and subject to biases that may negatively impact children and families of color. As
such, the use of more objective assessments of risk in child welfare is promising, however more
research is needed to identify risk assessment tools and systems that are reliable and valid.
Continuing Care/Ending Choice Point Practices
The continuing care/ending choice point practices identified by respondents fell into
three categories: 1) placement-related practices, 2) the Family to Family Initiative, and 3)
intensive and collaborative interventions for high-risk children.
Results from the California statewide survey indicate that, in general, the majority of
continuing care/ending choice point practices take place throughout the child welfare system
and are targeted to children ofall ages and ethnicities. However, more continuing care/ending
choice point practices are targeted to specific age groups than the other practices identified by
respondents. The goal of most of these services is to reunify children with their birth families,
ensure a permanent placement or prevent and out-of-home placement. Only two respondents
reported that these practices have been evaluated to determine their effectiveness and both of
these evaluations indicated that the practices are effective with the target population.
These continuing care/ending choice point practices are described as effective with
children and families of color because they are culturally appropriate, collaborative, intensive
and they involve clients ' communities. Yet, research on the effectiveness of these services for
children and families of color is lacking. Placement-related practices that emphasize placement
of children in culturally appropriate homes have not been evaluated to determine their impact on
children of color. However, some studies have noted that social workers may be reluctant to
place children of color with White alternate caregivers (Brooks & James, 2003; Carter-Black
2002; Vidal de Haymes & Simon, 2003). Results from the California statewide survey regarding
the identification ofplacement-related practices that emphasize placement of children ofcolor in
culturally appropriate homes do suggest that culture and ethnicity may be important factors in
child welfare workers' placement decisions. This appears particularly true for Native American
children who were mentioned specifically by two counties.
Research on the effectiveness ofthe Family to Family Initiative for children andfami/ies
ofcolor has not been conducted. There is limited evidence to suggest that social network
interventions which seek to involve community members in services are potentially effective in
increasing social networks, improving parenting skills and increasing the likelihood of case
closure (Gaudin et al., 1990). However the Family to Family Initiative encompasses more than
social network interventions and research is needed to evaluate the effectiveness ofthis practice
for diverse populations.
Although children in out-of-home care are at an increased risk for a variety ofpoor
outcomes (Stevenson, 1999), the impact of intensive and collaborative interventions for high-risk
children have demonstrated only mixed results in the research literature. While some studies on
167
wraparound services report that children in these services are less likely to be removed from their
community and may experience a reduction in behavioral problems (Bruns, Burchard, & Yoe,
1995, as cited in Toffalo, 2000; Clark, Lee, Prange, & McDonald, 1996, as cited in Toffalo, ·
2000; VanDenBerg, 1993, as cited in Toffalo, 2000), other studies report no significant benefit
from wraparound services (Clarke, Schafer, Burchard, & Welkowitz, 1992, as cited in Toffalo,
2000). More research is needed to identify aspects ofthese interventions that are most effective,
especially for children ofcolor
Both Beginning and Continuing/Ending Choice Point Practices
Practices related to both beginning and continuing/ending choice points fell into two
categories: 1) community-based services, and 2) strength-based assessments.
Results from the California statewide survey suggest that many practices that are both
beginning and continuing/ending choice point practices take place at a contract agency site or
the birth family's home. These services are targeted to children of all ages and any ethnicity.
These practices have numerous goals including: preventing an out-of-home placement;
reunifying children with their birth families; assessing level of risk; improving psychosocial
functioning of families; improving pare~ting skills; addressing substance abuse problems;
addressing mental health problems; addressing socioeconomic problems; and addressing housing
problems. All four of the respondents indicated that these practices have been evaluated and 50
percent of the evaluations indicated that the practices are effective with the target population.
Respondents to the California statewide survey described these practices as effective with
children and families of color because they are easy for clients to access; they are communityspecific, and non-intrusive. These services have broad goals that focus on both prevention and
intervention for families in contact with the CWS and research on the impact of these services on
children and families of color is lacking. In general, research on prevention and intervention with
child welfare populations does not provide convincing evidence of their effectiveness. However,
because these practices are more of an approach to services, rather than services themselves,
more research is needed to determine if these approaches are effective with ethnically diverse
populations.
Organizational or Structural Practices
Organizational or structural practices/ell into three categories: 1) cultural competency
ofstaff, 2) collaboration with American Indian tribes, and 3) vertical case management.
According to responses from the California statewide survey, most
organizational/structural practices take place throughout the CWS; the majority are targeted to
children ofall ages and ethnicities, although some are specifically targeted to Latino or Native
American clients. The primary goals of these practices include preventing an out-of-home
placement and ensuring a permanent placement. Approximately one quarter of respondents
indicated that these practices have been evaluated to determine their effectiveness and of these,
approximately 66 percent reported that the practices are effective with the target population.
168
These practices are described by respondents to the California statewide survey as
effective with children and families ofcolor because they result in better communication and
relationships between worker and client; they respect cultural heritage; increase cultural
awareness and produce more accurate and complete assessments and services. Although
cultural competency appears intuitively beneficial for ethnically diverse children and families,
research on the impact of culturally competent staff and programs in producing successful
outcomes for children and families of color has not been conducted. There is evidence to suggest
that for Latino clients. in particular, language barriers can exclude them from necessary services
and produce significant delays in service delivery; and researchers have also discussed the
importance of language in accurate assessment and treatment of clients (Suleiman, 2003). As
such, culturally and linguistically appropriate services appear to increase service use and
accuracy, although their impact on client outcomes is unclear.
Research on the effectiveness ofcollaboration with American Indian tribes and vertical
case management practices on outcomes for children and families of color is limited Effective
collaboration between the CWS and American Indian tribes appears intuitively beneficial for
Native American children, however more research is needed to identify aspects ofeffective
collaboration and how this collaboration impacts outcomes for Native American children. A
vertical case management model of child welfare service delivery may result in enhanced
outcomes. For instance, in South County Offices, where a vertical case management model is
used, families are more likely to have their cases closed with family maintenance services than
they are in the Main County Office where a vertical case management model is not used (please
see Section IV for a full discussion). These results suggest that a vertical case management
model may be an effective child welfare practice.
Limitations and Conclusions
Descriptive results of the statewide survey must be considered within the context of
certain limitations. For instance, the survey instrument was designed specifically for this study
and so its reliability and validity is not known. The self-report nature of the data also presents the
possibility of personal biases in survey responses and since the professional titles of the
respondents varied from directors, to managers to supervisors, there is the possibility that
different perspectives and information may occur as a result of these differing job titles.
Additionally, although the response rate was high (77.6%), there is the possibility that the sample
may be biased due to some counties refusing to participate.
Overall, respondents described beginning choice point practices as effective with
children and families ofcolor because they are inclusive, collaborative and objective.
Continuing care/ending choice point practices were considered effective with children and
families of color because they are culturally appropriate, collaborative, intensive and involve
client's communities. Both beginning and continuing/ending choice point practices were
described as effective because they are easy for clients to access, they are community-specific
and non-intrusive. Lastly, organizational or structural practices were considered effective with
children and families ofcolor because they are thought to result in better communication and
relationships because worker and client, they respect cultural heritage, increase cultural
awareness and produce more accurate and complete assessments and services.
169
In general, results from the California statewide survey suggest that there is a significant
gap between research on child welfare practices and the best practices for children and families
of color being used in California county child welfare systems. Little systematic research has
evaluated the impact of many of the most effective practices for children and families of color
that were identified by respondents Although the research literature has not yet provided
convincing evidence of the effectiveness ofmany of the practices identified by respondents,
respondents to the statewide survey felt that these practices were their most effective with
children and families ofcolor. These findings suggest that some counties may be utilizing
promising practices for children and families ofcolor in the child welfare system, and more
rigorous evaluations of these practices is needed.
170
Multivariate Analysis:
Impact of Statewide Practices on Differential Representation
The statewide survey identified practices, which addressed disproportionate
representation of children of color and were considered effective by county child welfare
personnel. Although perceived as effective, few of the practices had been evaluated to determine
effectiveness from an independent point of view and in a controlled manner. Child welfare
personnel nominated them as being effective, but in most cases their impact had not been
established through systematic empirical research.
For present purposes, the impact of interest is how the practices identified through the
survey might affect differential representation ofchildren ofcolor. Some of the practices had
been implemented with children of color in mind and others had not, but all were considered by
child welfare personnel to be especially valuable in addressing the needs and circumstances of .
children of color. If child welfare personnel are correct in their belief that the practices serve
children of color particularly well, then it is reasonable to believe that the practices identified in
the survey will divert children of color from the child welfare system or will reduce the scope or
duration of their involvement.
By helping children of color to follow pathways around full-fledged involvement in the
child welfare system or reducing the intensity or length of child welfare participation, the
practices identified in the survey can be expected to reduce differential representation. We can
expect lower levels of overrepresentation if the practices are effective.
Statewide data are available from sources, which, in conjunction with data from our
survey, permit the degree of empirical association to be determined between practices and
disproportionate representation. The U.S. Census and CWS/CMS publish data, available for use
by researchers and others, describing California counties on the worldwide web. The web-based
data provide an opportunity to evaluate whether there is a link between disproportionate
representation and use of perceived-to-be-effective practices.
If a study is statewide in scope like the present one, then it furnishes a basis to go beyond
the experience of a single county. By comparing groups of counties and contrasting those that
implemented a practice with those that did not, our investigation encompasses child welfare
systems around the state. The results point to promising practices identified from a statewide
perspective.
In this section, we present results from several analyses linking characteristics of
practices presented previously with disproportionate representation of children of color in the
child welfare system. Our intention was to identify types of practices, practices, and practice
characteristics-distinctions to be explained later-- associated with levels of disproportionate
representation, and which therefore might be more effective in reducing disproportionate
representation of children of color.
171
Multivariate Methods
The unit of analysis for this study was the county. The method, broadly conceived, was to
conduct controlled tests of empirical associations between types of practices, practices, and
practice characteristics on the one hand, and levels of disproportionate representation on the
other hand.
For thinking about the study's approach in greater detail, it is useful to begin by
considering the meaning of disproportional representation. Disproportional representation is
said to occur when the proportion of children of color in a county's child welfare system is
higher or lower than expected. Typically, documenting the proportion of African Americans,
Latinos, or Asian American/Pacific Islander in the population at large usually establishes the
proportion that is expected. The difference between the representation of a group in the
population at large and its representation in the child welfare system defines its level of
disproportional representation.
In our study we calculated a measure of this difference using regression analysis. We
analyzed a group's representation in the child welfare system after statistically adjusting for it
representation in the population at large. Following a research design like ours, a statistic results
from regression which defines the difference between expected and actual levels of child welfare
representation.
Disproportional representation is affected by factors other than those under the direct
control ofchild welfare authorities. The stresses under which minority families live and the
resources available to them for example, can contribute to disproportionate representation.
Counties can differ in the levels ofstress, resources, and similar differentiating characteristics.
Child welfare systems too are diverse. Some are larger and more centralized than others, for
example; policies and proc~dures vary.
It is difficult to establish definitively whether practices are effective because ofthe
differences apart from the practices between counties and child welfare systems associated with
disproportionate representation. When a link is observed between a practice and
disproportionate representation, it remains possible that a characteristic other than the practice
itself is responsible for that link. The observed link between a practice and disproportionate
representation might be coincidental. Controlling/or key differences between counties and child
welfare systems reduces the chances of associations occurring through coincidence.
We examined many indicators describing counties and their child welfare systems in
order to identify differences that were especially important to take into account. Among
indicators we screened were population size and density, wages and costs of goods and services
in the county, the geographical location of the county, size of the child welfare system and
proportion of the child welfare population in kincare. We sought key indicators from a large
number we considered reflecting differences between the counties and their child welfare
systems and linked to disproportionate representation.
172
Through this process we found two indicators that were consistently and strongly
associated with disproportionate representation: urban-rural status and per-capita income.
Counties that were urban and had higher per-capita income were notably more likely to have
higher levels ofdisproportionate representation.
We used these data to conduct a controlled analysis of the association between practice
types, practices, and practice characteristics by constructing regression equations. In each
equation, the proportion of Latino, African American, or Asian American/Pacific Islander
children served as the dependent variable. The proportion of Latinos, African Americans, or
Asian American/Pacific Islander in the county was entered first as an independent variable, as
discussed above, in order to subtract an expected level of representation from the actual
representation of each group in each county's child welfare system. We then added to the
equation each county's urban-rural status and per-capita income. Finally, we added each practice
type, practice, or practice characteristic suitable for analysis.
This enabled us to determine how much each practice type, practice, or practice
characteristic was associated with greater or lower level of representation apart from the level of
representation expected from its group representation in the county, and after adjusting for the
county's urban-rural classification and its per-capita level of income. Thus, the approach
adjusted for the three key factors: group representation in the population, whether the county was
classified as urban or rural, and per-capita income of county residents.
In most cases the three control variables alone explained disproportionate representation
in a statistical sense extremely well. Success in accounting/or differences between counties in
statistical models implies that many important differences between counties were explained and
that the possibility that associations arose by coincidence was low.
We considered practice types, practices, and practice characteristics in a three-phase
analysis. Phase 1 addressed our organization ofpractices into beginning choice point practices,
continuing care/ending choice point practices, organizational or structural practices, and both
beginning and continuing/ending choice point practices. We considered disproportionate
representation according to whether practices were beginning, continuing/ending, organizational
or structural, or both beginning or ending.
In phase 2, we considered individual practices in an attempt to consider the impact of
practices on disproportionate presentation in finer-grained detail. We were hampered in our
attempt: very few counties reported instituting certain of the practices.
Ultimately, we chose to consider at the individual practice level only those where four or
more counties indicated implementing them. Four counties is a small number for this purpose
and limit our ability to detect genuine differences that might have been present at statistically
significant levels. It proved fruitful however, and provides an important basis for discussion of
individual practices. The five practices we considered were: family group conference or other
team decision making practices (n=7), voluntary family maintenance or other prevention/family
preservation services (n=4), placement related practices (n=6), family to family initiative (n=4),
and cultural competency of staff (n= 10).
173
We asked parallel questions in all phase I and phase 2 analyses. Adjusting for urbanrural status and per-capita income, for Latinos, African Americans, and Asian Americans how is
disproportionate representation affected by type of practice? What was the impact on
disproportionate representation ofwhether a practice was one offive specific practices: family
group conference or other team decision making practices, voluntary family maintenance or
other prevention/family preservation services, placement related practices, family to family
initiative, and cultural competency of staff?
In phase three of our analysis we asked questions without regard to type ofpractice. We
considered the impact on disproportionate representation-whatever the particular practice
type-- ofdifferences in practice location, target population, purposes, reasons for
implementation, initiating persons and agencies, implementation timing, implementation status,
location and scope of the practice, geographic and community targeting,. evaluation, support,
and barriers.
There are many practice characteristics describing practices. To reduce the number of
characteristics for purposes of analysis, we clustered characteristics into logically related,
empirically associated groupings using factor analysis, a statistical procedure for clustering
individual questions to which respondents have given similar answers. When a question was
discovered not to fit well within factor-analysis created clusters of questions, we eliminated it
from the cluster and considered it in subsequent analysis individually.
We then conducted regression analyses like those conducted in phases 1 and 2. The
question was: adjusting for urban-rural status ahd per-capita income, for Latinos, African
Americans, and Asian Americans, how is disproportionate representation affected by practice
characteristics? What was the impact on disproportionate representation of differences in
practice location, target population, purposes, reasons for implementation, initiating persons and
agencies, implementation timing, implementation status, location and scope of the practice,
geographic and community targeting, evaluation, support, and barriers?
There are two ways in which practice types, practices, and practice characteristics can
be linked to disproportionate representation. One is a negative association between the
variables, indicating lower than expected overrepresentation. Negative association means that
implementing counties have relatively low or nonexistent minority overrepresentation and
suggests that the practice type, practice, or practice characteristic is successful.
Another kind ofassociation between a practice type, practice, or practice characteristic
and overrepresentation is a positive association. Positive association between variables means
that there is greater than expected overrepresentation associated with the practice type, practice,
or practice characteristic-that overrepresentation is relatively high.
Practices linked to greater overrepresentation cannot be considered successful. How does
a practice come to be linked to greater overrepresentation? It is likely that positive association
between a practice and overrepresentation comes about because instead of reducing
overrepresentation, some practices are implemented reactively, that is, in response to
174
overrepresentation. These practices do not reduce overrepresentation, although they may do so in
the future if given sufficient time. At present, we note only that a significant positive association
with overrepresentation is probably reactive and does not indicate success.
Multivariate Results
Latino Overrepresentation
Latinos were overrpresented in about 60% of counties and the degree of their
overrepresentation was as high as 28 percentage points.
Practice types. For Latinos, there was a statistically significant difference according to
whether practices were beginning choice point practices, continuing care/ending choice point
practices, organizational or structural practices, or both beginning and continuing/ending choice
point practices.
Both significant differences involved beginning choice point practices. Beginning choice
point practices were associated with about 8 percentage points less overrepresentation than both
beginning and continuing/ending choice point practices and about 9 percentage points less
overrepresentation than organizational or structural practices (See Table 63).
Practices. One statistically significant difference emerged when we considered the five
individual practices that could be analyzed. Practices .involving family group conference or other
team decision making were associated with 7 percentage points less overrepresentation than were
practices involving culturally competency of staff (See Table 63).
Practice characteristics. For Latinos, there were statistically significant difference in
target population, purposes, initiating persons and agencies, and implementation timing (See
Table 64). Latino overrepresentation was greater when areas were targeted for intervention (area
where most referrals came from; with low socioeconomic status residents; with high crime; with
large ethnic representation), when clients, foster families, and out of home placement providers
were involved in the development of the practice, when service systems were involved in the
development of the practice (mental health, substance abuse, health, criminal justice, contract
agencies). Overrepresentation was less also when "other" problems were addressed as the
purpose of the practice and when it began more recently.
African American Overrepresentation
African Americans were overrepresented in about 64% of counties and their degree of
their overrepresentation was as high as 65 percentage points.
Practice types. For African Americans, there was no statistically significant difference in
overrepresentation according to whether practices were beginning choice point practices,
continuing care/ending choice point practices, organizational or structural practices, or both
beginning and continuing/ending choice point practices (See Table 63).
175
Practices. One statistically significant difference appeared when we considered the five
individual practices, which could be analyzed. Family to Family initiatives were associated with
8-percentage points greater overrepresentation than were the remaining practices considered (See
Table 63).
Practice characteristics. For African Americans, there were differences in target
population, purposes, and initiating persons and agencies, and geographic and community
targeting. African American overrepresentation was less when areas were targeted for
intervention (area where most referrals came from; with low socioeconomic status residents;
with high crime; with large ethnic representation), ethnic minority. populations (Latino, African
American, Asian American/Pacific Islander, American Indian, families of any ethnicity), and
when very young children were targeted. African American overrepresentation was less when
foster families and clients or contract agencies first initiated the practice. African American
overrepresentation was less also when the purpose the intervention was to improve the
psychosocial functioning of the birth family (See Table 64).
Asian American/Pacific Islander Underrepresentation
In no counties were Asian American/Pacific Islanders overrerpresented. Instead, they
were equally represented in a few counties and in most counties Asian American/Pacific Islander
were underrepresented.
Practice types. For Asian American/Pacific Islanders, there was no statistically
significant difference in overrepresentation according to whether practices were beginning
choice point practices, continuing care/ending choice point practices, organizational or structural
practices, and both beginning and continuing/ending choice point practices (See Table 63).
Practices. Nor was there a statistically significant difference when we considered the five
individual practices that could be analyzed (See Table 63).
Practice characteristics. For Asian American/Pacific Islanders, there were differences in
target population, implementation timing of the practice. Asian American/Pacific Islanders under
representation was greater when areas were targeted for intervention (area where most referrals
came from; with low socioeconomic status residents; with high crime; with large ethnic
representation), ethnic minority populations (Latino, African American, Asian American/Pacific
Islanders, American Indian, families of any ethnicity). Asian American/Pacific Islander under
representation was less when system management and child welfare workers expressed support
for the intervention, and when it had begun earlier (See Table 64).
176
Table 63: Practice Types, Practices, and Disproportionate Representation
Pracfice Types
Ethnic Group
Variable
B SE Beta
T
Latino
Beginning vs
.09 .04 .14
1.98*
Organizational/
Structural
Beginning vs
.08 .03 .20 2.61 **
Beginning/Continuing
Practices
Variable:
B
SE Beta T
Family Group Conference/ .07 .03 .15
2.13*
Team Decision Making vs
Cultural Competence
African American Family to Family vs. other -.08 .04 .13
2.23*
Ethnic Group:
Latino
Table 64: Practice Characteristics and Disproportionate Representation
Ethnic Group:
Latino
Variable:
Area Targeting
Target:Other Problems
Clients Involved in
Development/Evolution
Systems Involved in
Development/Evolution
Initated Earlier
African American Ethnic Targeting
Area Targeting
Targeting: Very young
Foster family-client
Initiated
Contract agency
Initiated
Goal: Improve psychoSocial Functioning Birth
Family
Asian American/ Ethnic Targeting
Pacific Islander
Area Targeting
Management-Worker
Support
B
.04
-.05
.03
SE Beta T
.01 .19
3.07**
.26 -.13 -2.06*
2.77**
.01 .17
.03
.01
.16
2.61 **
-.00
-.04
-.02
.06
-.08
.00
.01
.01
.03
.04
-.16
-.23
-.12
.11
-.11
-2.69**
-4.20**
-1.93*
1.92**
-1.94**
-.08 .04 -.13
-2.11 *
-.05
.02 -.13
-2.11 *
.00
.00 .22
1.99*
.00
-.01
.00 .19
.00 -.29
1.90*
-3.24**
Notes:
Regressions: Ordinary Least-Squares. All models R >.80 (p>.01).
Factors= Principal Components, 1 factor per analysis, Eignenmvalue> 1. Regression-based factor scores used in
subsequent regression. Variables with loadings >.50 were eliminated from analysis and analyzed separately in
subsequent regression.
177
Summary of Key Findings and Implications:
Multivariate Analysis
When considering the 43 counties responding to the survey of practices perceived
effective in working with children of color, significant associations were uncovered between
practice types, practices, and practice characteristics on the one hand, and disproportionate
representation on the other. These associations remained after adjusting for urban rural status and
per-capita income, differences between the counties strongly linked to minority child welfare
representation. Some point to effective practices-practices that bring about minority
underrepresentation. Different patterns ofassociation were found for Latinos, African
Americans, and Asian American/Paci.fie Islanders.
For Latinos, beginning choice point practices had a beneficial effect on
overrepresentation. In particular, family group conference or other team decision-making
interventions predicted lower levels ofoverrepresentation.
Bringing together interested parties early on in child welfare involvement may divert
cases from child welfare or otherwise minimize involvement, perhaps by facilitating
communication and mobilizing otherwise overlooked resources. Conceivably, the intervention is
especially valuable for Latinos in that its family focus is congruent with cultural values, and
because of the intervention's potential to overcome linguistic and cultural barriers to
understanding.
Notably, these interventions do not target at-risk neighborhoods as a special focus of
concern. Indeed, Latino overrepresentation was greater, not less, when interventions were
implemented targeting areas with high concentrations ofdistressed families.
Other findings indicated that practices targeting at-risk populations came about when
clients, foster families, and out of home placement providers were engaged in development of
the practice, along with mental health, substance abuse, health, criminal justice, contract
agencies. It appears that when they perceived a need and had an opportunity to do so,
stakeholders mobilized and focused on high need areas and populations as intervention targets.
However, they appear do this in response to Latino overrepresentation, such that
implementation is linked to greater overrepresentation. The positive association at-risk targeting
and disproportional representation indicates that these practices do not reduce disproportional
representation. In all likelihood, the positive association means instead that stakeholders
perceive Latino overrepresentation and initiate programs in response.
For African Americans, targeting ofat-risk areas and African American populations was
beneficial in reducing overrepresentation, as well as targeting of very young children. Less
overrepresentation was linked also to whether foster families and clients and contract agencies
initiated an intervention, and whether it focused on improving psychosocial outcomes of birth
families. The findings suggest a picture almost the reverse of that found for Latinos.
178
For African Americans, there appear to be lower levels of overrepresentation than
expected too when key community stakeholders help to initiate the interventions that focus on
high-risk areas and populations. When interventions have a goal ofstrengthening the well being
of birth families, then implementation is associated with lesser overrepresentation.
On the other hand, counties that implementedfamily-to-family initiatives displayed
greater overrepresentation than counties that did not. Fami/y to family programs appear to be
used reactively, in response to high perceived levels of overrepresentation.
It is important to note however that in counties where interventions were successful
levels of overrepresentation remained high. Successful interventions were those associated with
lower overrepresentation than expected. Improvement was relative: in absolute terms there
remained a high degree of African American overrepresentation.
For Asian American/Pacific Islander populations findings suggested that, as with
Latinos, targeting ofhigh need areas was less beneficial than not engaging in this kind of
intervention. Involving clients and service systems in initiation, again more common for
interventions targeting high need areas, was also relatively unsuccessful. On the other hand,
management and worker support for a practice was associated with greater success.
Asian American/Pacific Islanders are, overall, underrepresented in child welfare systems
and one must interpret our findings in light ofthis fact. There were fewer significant differences
between practice types, practices, and practice characteristics for Asian American/Pacific
Islanders than for Latinos or African Americans. This may reflect a lesser sense of urgency felt
by administrators and workers in the face of Asian American/Paci fie Islander disproportionate
representation that is proportionally low than for Latino and African American disproportionate
representation, which is proportionately high.
A need remains to conceive and mount culturally and linguistic programs for Asian
American/Pacific Islander populations. This aspect of the study, with its focus on
disproportionate representation rather than on other indicators of system performance, may not
have been as good a vehicle as other parts of the study to identify program needs for Asian
American/Pacific Islander populations as for overrepresented Latino and African American
populations.
Certain limitations must be kept in mind when considering the present findings. The
study encompasses forty-three California counties. The number of counties implementing any
practice or practice type was smaller than ideal to undertake truly powerful statistical tests
sensitive to all underlying differences. As a matter of practical necessity, the study's design was
such that some of the links of practice types, practices, and practice characteristics with
disproportionate representation might have arisen by coincidence. The possibility that some
unmeasured characteristic explained some of the associations of interest revealed in the study
was reduced, but cannot entirely be ruled out.
Despite these limitations, the study produced meaningful patterns of findings connecting
programs and disproportionate representation. The associations proved sensitive to the
179
differential experience of different ethnic groups. They furnish a useful context of fact to further
consider the problem of disproportionate representation in Santa Clara County.
180
VI. PHASE 3 SUMMARY, RECOMMENDATIONS AND
CONCLUSION
Phases 1 and 2 of this study focused on examining reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where it occurs. The
focus ofPhase 3 was on key practices that take place within the CWS that either promote or
reduce disporportiona/ity among children and families ofcolor.
Overall, research on the impaci of child welfare practices for children and families of
color is limited. In general, much of the research literature is inconclusive and fails to support the
effectiveness of child welfare practices in improving outcomes for children and families involved
in the CWS. Very few studies describe differences in outcomes by ethnic/racial groups and most
of the research fails to take into account the effect of culture and ethnicity on the delivery and
impact of child welfare practices. This gap in the research literature is striking and has important
implications for children and families of color in the CWS. More information about child welfare
practices including programs, services, strategies, policies, and/or tools that either increase or
reduce racial/ethnic disproportionality can contribute to the development of more effective
services and supports for children and families of color in the CWS.
The three overall aims addressed in Phase 3 were: 1) to assess identified key practices for
children and families of color in Santa Clara County's CWS, 2) to examine the influence of the
court system on service recommendations for children and families and, 3) to conduct a
statewide comparative analysis of identified key practices that may affect the disproportionate
representation of children ofcolor in the CWS.
In order to address Aim 1 which involved an assessment of key practices in Santa Clara
County's CWS, four related analyses were conducted. First, the CWRT conducted analyses to
understand the types of practices and case characteristics that are associated with successful
outcomes for children and families from various racial/ethnic groups. In order to assess
successful outcomes, we used data from our Phase 2 Case Record Review sample and defined
success according to US DHHS guidelines, i.e., safety, permanency and family and child wellbeing. Results indicated that there were very few ethnic differences in successful outcomes.
In regards to safety, there was a significant relationship between the number of times
removed from family in current episode and ethnicity where African American children had the
highest average number of times removed from family in current episode and Asian
American/Pacific Islander children had the lowest average. In terms of family and child wellbeing, there was a marginally significant relationship between families' capacity to provide for
their children's needs and ethnicity, with African American families appearing to have a more
enhanced capacity while Asian American/Pacific Islander families seem to have a more
diminished capacity.
It is important to note that major ethnic differences do exist in Santa Clara County's
CWS. Phase 1 and 2 confirmed that certain ethnic groups are over-represented (i.e., Latinos,
African Americans) while others are under-represented (i.e., Asian American/Pacific Islanders)
in the county's CWS. We also found that there are significant ethnic differences in many
181
demographic, system-related, and psychosocial characteristics. Many of these variations appear
at different choice points in the system (such as psychosocial status at the beginning of the case,
and assignment to voluntary family maintenance). Results ofour successful case outcomes
analysis indicate that ethnic differences at the conclusion ofthe case are not evident.
While results suggest there is a clear disproportion of ethnic groups in the system, once in
the system, the children and families are generally faring the same. Their relatively
homogeneous experience may be a result of the "one-size-fits all" limited array of services
offered. Although this may seem equitable and logistically efficient, it appears ultimately
ineffective in serving a diverse group of children and families. Also, to determine better how
"well" these children are actually doing while in the system, the availability of comparative data,
clear operational definitions for "success," and more reliable and valid data are needed.
Implications and Recommendations
•
In order to better assess whether safety, permanency, and family and child wellbeing are maintained and enhanced, records should include more complete and
measurable information. We acknowledge that child welfare case records were
not originally intended for research. Yet, if the agencies responsible for providing
effective services wish to evaluate the performance of their programs more
reliably and validly, improving information collection and tracking should be
considered.
•
The creation of a case summary check-out form used at case closure would be one
way to track child outcomes. The check-out form would indicate the child and
family's status in major areas of psychosocial well-being (i.e., health, education,
and finances), in addition to the existing case outcomes. Given the amount of
missing data assessing these domains describing case and child status, and the
relative scarcity of information recorded in the last two court hearings, this form
would enable measurement and evaluation of "successful cases and outcomes" in
a more consistent objective manner.
•
The amount of missing data and the manner in which information is assessed also
implies the need for more reliable and valid methods of recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their families or of
those adopted high or low? And compared to what? Implementing better record
keeping and data collection over time on key characteristics, based on an agreedupon operational definition would better address the question of effectiveness and
performance. Implementing these practices across counties and across states
would also facilitate evaluation.
Second, a qualitative analysis using both focus group and interview data was conducted
to obtain insights about current child welfare practices used within the DFCS, how these specific
182
practices might positively or negatively impact children and families of color, and what practices
might be created, enhanced or maintained to better serve children and families of color.
Results provided information on numerous practices and contextual factors that have
either a negative or positive impact on children andfamilies of color in Santa Clara County's
child welfare system. Study participants recommended the use of in-home services for children
and families of color. There is some evidence in the research literature that supports the
effectiveness of in-home services (see literature review), however these services appear to be
most effective when they are intensive, long-term and delivered by health or social service
professionals and when the workers delivering these services receive a high level of supervision
(Barth, 1991; McGuigan et al. 2003; Olds et al., 1997). Participants also identifiedfamily
conferencing as an effective practice. These study findings are consistent with limited research in
this area that suggests that inclusive practices, in which parents are engaged in placement and
other decisions, may result in more placement stability for children, although possible
differential effects for diverse populations has not been researched (Palmer, 1996).
Participants generally expressed optimism about the adoption ofteam decision making
(I'DM) as a vehicle for improving decision-making with diverse families, although some social
workers expressed concern about time required for implementation. Research suggests that TDM
can be labor intensive and can be difficult to implement effectively within the context of the
child welfare system (Sieppert et al., 2000). Finally, participants noted that both formal and
informal collaborative relationships with other services delivery systems, such as domestic
violence and substance abuse treatment systems, are critical for successful interventions with
children and families. Furthermore, participants recognized the importance of training and
cross-training to enhance the capacity of helping professionals in child welfare and other
systems to better address the needs of children and families of color.
Interestingly, most practices and contextual factors mentioned by social workers had both
positive and negative aspects. Many of the practices and contextual factors that negatively
impact children and families of color were actually barriers to successful implementation of the
practices that social workers felt have a positive impact on children and families of color.
Although practices such as orientation ofnew clients; family and team decision-making;
preventive services; substance abuse services; cultural competency; collaboration with other
agencies and systems; a strength-based approach; and the use ofrelative placements were
described as having a positive impact on children and families ofcolor, certain negative
practices and contexts impeded their effective implementation. For instance, having to screen out
large numbers of inappropriate referrals; inconsistency in decision~making practices; a shortage
of services, particularly substance abuse and preventive services; lack of client access to
services; difficult protocols for placing children with kin; gaps in cultural competency; time
limits; and agency-level factors such as heavy caseloads, staff shortages, substantial amounts of
paperwork, lack of access to information about resources; and confusion about the agency's
overall mission and key policies all impeded implementation of the best practices. These findings
suggest that in addition to building on current positive practices for children and families of
color, CWS stakeholders should also actively work both internally and in collaboration with
partners to reduce barriers to these best practices.
183
Indeed, in order to address these barriers to effective implementation of best practices,
study participants mentioned numerous recommendations that centered around certain key
themes. In general, there was an emphasis on improving decision-making through increased
accountability, training and cross-training to reduce bias, and increased use ofgroup decisionmaking or family involvement in decision-making. Study participants also stressed a need to
improve and expand prevention, diversion and concrete services, as well as an overall need to
expand culturally and linguistically competent services, and develop ways to improve
availability and access to services. Once a case is opened, social workers also felt it valuable to
provide an orientation to clients so that they are more aware of the CWS and court processes and
can better navigate the system. Recommendations related to organizational factors were also
noted, including reducing caseloads; increasing support and streamlining service delivery;
clarification of the mission ofDFCS, as well as key child welfare policies; fostering an
organizational culture that is strength-based and community-based, and more overall training
for social workers.
Implications and Recommendations
•
The findings from the qualitative component of this study affirm the value of
many efforts that are already in place, such as culturally specific ER response
units and family resource centers, and other that are in the process of
implementation including efforts to reduce caseloads, initiation of team decision
making, and participation in the Family-to-Family initiative.
•
Practices that are inclusive, collaborative, culturally appropriate and involve
client's communities appear to be best suited to meeting the needs of children and
families of color. Although the research literature does not provide convincing
evidence of the effectiveness of these practices, the fact that respondents reported
these practices as those that are most effective with children and families of color
suggests that they may serve as promising models for children and families of
color and that more rigorous evaluations of these practices are needed.
•
The findings also point to opportunities to further strengthen services for children
and families of color. Recommendations related to training, institutionalizing
group orientation for clients, and other suggestions from study participants could
be used to inform practice and planning.
•
It would be beneficial for a team of managers, line staff, family representatives,
and community members to review the recommendations suggested by focus
group and interview participants to identify and prioritize possible practices for
adoption based on their feasibility and utility for children and families of color.
Third, in an effort to better understand the influence of contextual factors on child
welfare practices and outcomes for children and families of color, a comparison of Santa Clara
County's Main Offices and South County office was conducted. This comparison included both
quantitative and qualitative analyses. Quantitative data from CWS/CMS were used to examine
differences in case characteristics between the Main County Offices and South County Offices.
184
A qualitative comparison of focus group data from the Offices and South County Offices was
also conducted to discern how workers perceive practices in these two geographical locations.
These analyses help shed light on the potential impact of agency and community context on case
characteristics, child welfare practices and outcomes for children and families of color in Santa
Clara County's CWS.
Unfortunately, the South County location subsample was too small to allow analyses by
ethnic group. However, some key case characteristics are significantly related to county
location. Specifically, in South County Offices, a higher proportion ofcases are in voluntary
services, cases are shorter in duration, fewer workers are assigned during the course of the case,
and a higher proportion ofcases are in family maintenance at case closure. These findings are
also statistically noteworthy given the small subsample size available from SC. There were also
other differences in case characteristics by county location that did not attain statistical
significance but are noteworthy for further investigation. Children served in the South County
Offices appear to have fewer number of unique placement homes in the current episode, have a
shorter average stay per placement, a shorter length oftime in out-ofhome placement, and are
younger at time ofcase closure.
These findings imply that there is a difference in the style and quality of services
provided in the South County Office location. Qualitative findings are congruent with the
quantitative data and provide some insight into the dynamics that may contribute to this
difference. The higher number ofcases in voluntary services, the larger proportion ofcases in
family maintenance at case close, and the shorter duration ofcases is consistent with what South
County study participants described as a culture ofcommitment to maintaining and reunifying
families. This commitment was described as integral to the philosophy of managers and social
workers and reflected in the .expectations communicated to social workers by most supervisors
and by peers. The practice of providing an orientation to the child welfare system, time limits,
the courts, and how to work with social workers may also contribute to the differences found in
the quantitative analysis. Study participants frequently described South County as similar to a
small county or rural area. This geographic difference, and the resulting sense of "connection to
the community, " was identified as a factor in the greater sense ofshared philosophy and
emphasis on prevention of out-of-home placements. Other studies suggest that region may
influence practice. For example, one study (Drake, 1996) also found that rural areas are more
likely to offer preventive services than are urban areas. The finding that fewer workers are
assigned during the course of the case is also consistent with descriptions of the vertical case
management model. This model calls for social workers to carry the same case from case
opening to case closure.
Implications and Recommendations
•
Future research should include a larger subsample from South County and include
methods to understand the reasons behind these location differences. Also, a
larger South County Office location sample may enable more detailed
comparisons by ethnicity.
185
•
Practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County
Offices and adapted to other regions of Santa Clara County, as recommended by
study participants. This practice may be particularly helpful to low-income
families and families of color that may be intimidated by or unfamiliar with child
welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for other
regions of Santa Clara County, though the success of this practice may be linked
to the "small county" dynamic described by South County study participants.
Some participants in focus groups that took place at the main office mentioned
this model as promising while others asserted that specialized expertise in
different areas of social work was an advantage in providing quality services that
are informed by "the most current information."
•
Further examination of ways in which the culture of commitment to maintaining
and reunifying families is created, communicated, and continued merits attention.
In addition, an exploration of how this culture might be adapted to other parts of
the county would be beneficial.
Fourth, in order to identify case characteristics and system-related factors predictive of
reunification of children with their families, two exploratory models were tested using
multivariate analyses. The first model was composed of primarily demographic and systemrelated variables to predict reunification (child's gender, ethnicity, age at time of case opening,
the number of workers assigned across the history of the case, the time length of the case, the
number of unique placement homes assigned, the number of times removed from the family, and
the county location). Four of these variables predicted reunification: child's ethnicity, number of
workers assigned throughout the case, length ofthe case, and number of unique placement
homes.
Asian American/Pacific Islanders were less likely than Whites, African Americans, and
Latinos to be reunified with their families. Given that Asian American/Pacific Islanders are one
of the minority groups least acculturated to mainstream society, their cases may be perceived as
more severe or problematic because of the different cultural attitudes toward child rearing and
discipline. Differences in parenting styles may reduce the likelihood of reunification compared to
the other major ethnic groups. However, as Phase 2 results indicated Asian American/Pacific
Islanders are also the most likely to be placed in Voluntary Family Maintenance (Hines et al.,
2002), the CWS appears to value the preservation of Asian American/Pacific Islander families
when first referred to the system (perhaps recognizing the challenges of culturally appropriate
services, language issues, and other circumstances related to the Asian American/Pacific Islander
community) but once in the system, Asian American/Pacific Islander family cases are not easily
resolved and circumstances may persist that place the child at risk for further abuse of neglect.
The fewer number ofworkers assigned was also related to reunification. This finding
indicates consistency in service and the relationship between the social worker and the family
186
may improve the chances ofreunification. Perhaps this is due to greater familiarity and thus more
effective advocacy for the child and family. This finding could also ·be a function of particularly
challenging family cases already less likely to be reunified needing a greater set of workers over
time (e.g., due to worker burnout in the case, or specialization needed across the case).
Shorter case length was also predictive ofreunification. Presumably, less severe cases
could be resolved in a timely fashion, usually concluding with reunification. Following a similar
logic above regarding the number of workers, a longer case may imply more problems, thus
involving more time for resolution, but also possibly a case already protracted into a situation
where reunification is unhealthy for the child.
The fewer number of unique placement homes assigned was related to reunification.
Change in placements may be due to systemic conditions (i.e. logistics and short term
availability of space) but multiple placements can also indicate a persistent problem with a child
adapting to a placement and accepting care offered. As this may again indicate a more severe
case, reunification could also become less likely.
To identify better the influence of case characteristics and indicators of success, we added
6 variables to the model collected through our in-depth case record reviews and utilized the
additional information coded for Phase 3: a family's enhanced capacity to provide for their
children's needs, children receiving appropriate educational services, children receiving adequate
services to meet physical and mental health needs, number of referrals, number of previous times
in the CWS, and assignment to Family Maintenance or Family Reunification services. The
location variable was not included in this model as only one case coded from the South County
Office area.
Similar to the first model, Asian American/Pacific Islanders were still less likely than
Whites, African Americans, and Latinos to be reunified. Also, shorter time length of a case
again predicted reunification. However, in this expanded model with success indicators, the
number of workers assigned and number of unique placement homes assigned were not
statistically related to reunification. The child's age at time of case opening and the number of
unique placement homes did approach significance. Interestingly, the variables related to
success cases and outcomes were not significant.
Implications and Recommendations
•
These findings indicate that particular factors are important to consider when
predicting reunification, specifically ethnicity and the length of time a case
remains open. More severe cases (those needing more time for resolution) may be
more problematic, thus reducing the chances of reunification.
•
The finding that Asian American/Pacific Islanders are less likely to be reunified
than Whites, African Americans and Latinos, highlights a main ethnic difference.
In Phase 2 we discovered that Asian American/Pacific Islanders were more likely
than the other ethnic groups to be enrolled in Voluntary Family Maintenance
services (Hines et al., 2002). However, our results here imply that when Asian
187
American/Pacific Islander children are removed from the home, their children are
less likely to be reunified. This may also indicate extreme outcomes where Asian
American/Pacific Islander children are either initially diverted from the system or
once in the mainstream of the system are less ~ikely to be reunited with their
families.
•
It was interesting that the successful case and outcome variables were not
predictive of reunification. This could be due to the significant impact of other
variables overshadowing the influence of success indicators. It could also be due
to our use of reunification as a dependent variable. As we discussed in the
Successful Cases Analysis section, reunification is not necessarily synonymous
with success, and thus these variables may indeed be unrelated.
•
Due to missing data across our set of variables, a number of were available for
this multivariate analysis. More complete information would have improved
confidence in our findings.
•
Better measurement and a more comprehensive evaluation of family history,
abuse and neglect circumstances, and other indicators could produce a stronger
predictive model.
•
Although only a preliminary set of models, these findings are useful for
identifying areas in the CWS and family circumstances that can be studied
further, especially in conjunction with the specific goal of reunifying children
with their families.
Our second aim focused on examining the influence of the court system on service
recommendations for children and families ofcolor. Fey,t studies have examined the potential
influence of the court system on child welfare practices and outcomes, however as noted in the
literature review (Section III), the court system is likely to have a significant impact on the
trajectory of some child welfare cases. In Phase 3, we examined the role of the court system in
child welfare practice by exploring the types of court ordered changes that are made to social
worker recommendations at the initial jurisdictional/dispositional hearing, and how changes
might vary by ethnicity of the case. In addition, we explored the relationship between the child
welfare and court systems.
Quantitative results indicated that the court made changes to the initial social worker
recommendations at the jurisdictional/dispositional hearing in approximately halfof the cases.
Ethnicity was not related to whether or not a change occurred Yet, these results may not
necessarily indicate that the system and courts treat children and families from different
ethnicities the same, but rather that when the judge decides to make a change from the social
worker recommendations, these changes appear to occur equally across ethnic groups. It was the
opinion of members of the Court Officer Unit focus group that ethnicity does not contribute to
decisions in court, but rather it is circumstances that drive the case.
188
Additionally, although quantitative results may imply that judges accept social worker
recommendations about half of the time, findings from the focus group with the Court Officer
Unit suggest that many factors would complicate an accurate measurement of why social worker
recommendations may or may not be followed. Factors that may impede accurate assessment of
the agreement between child welfare and court personnel on service plan recommendations
include the strong influence of individual stakeholders involved, and formal and informal case
plan negotiation tactics, all of which contribute to a much more complicated process.
Qualitative .findings from the agency-wide focus groups suggested that collaboration
between the child welfare and court systems is problematic. Agency-wide focus group
participants felt that the child welfare and court system have different perspectives on the needs
and circumstances of children and families in the CWS and that the court system may have
unrealistic expectations of families. Additionally, many agency-wide focus group participants
described the ways in which child welfare and court personnel interact as ineffective. Some
social workers felt that some court personnel would dismiss their assessments and
recommendations and may try to pressure social workers to change their recommendations. A
somewhat similar finding was noted by Knepper and Barton, ( 1997) who found that although
judges tended to accept social worker recommendations, the relationship between social workers
and the court plays a significant role. Their study found that when social workers adhered to the
group norms of the court, judges rewarded them by not ordering "unrealistic" practices and
allowing reasonable time frames for mental health evaluations to be completed. Similarly,
agency-wide focus group .findings seem to suggest that ifsocial workers go against the implicit
rules of the court system, thenjudges may override their recommendations, but if they adhere to
expectations of the court system than recommendations are accepted.
Implications and Recommendations
•
Although the quantitative and qualitative information described many of the
circumstances behind why social worker recommendations are not upheld by the
court, it is very difficult to ascertain the degree to which each circumstances
influences a court ordered change. More examination of this area is needed, as is
further investigation of ways in which relationships between court personnel and
child welfare workers have an impact on the experiences of children and families
of color involved in the CWS.
•
Better and more exhaustive record keeping may not be the solution, as many
informal and undocumented actions in the negotiation process occur regularly,
and are confidential, thus making much of the data unavailable for research and
evaluation purposes.
•
Assessing other factors such as courtroom environment, stakeholder
characteristics, and other judicial issues would be advantageous in later research.
Although obtaining this information is a formidable task, these factors should be
include in the future, as measuring primarily system-related factors may not be
sufficient in efforts to explain outcomes in the CWS.
189
Our third aim was to conduct a statewide comparative analysis of identified key practices
that may affect ethnic disproportionality in the CWS. A statewide survey was administered to
child welfare directors, managers or supervisors in California counties who were asked to
identify and describe effective practices for children and families of color in their child welfare
systems. In addition, a quantitative analysis that identified the link between these practices and
county levels of disproportionate representation of children of color in county child welfare
systems was conducted.
Overall findings from the statewide comparative analysis indicate that certain types of
practices f:Ire considered by county child welfare personnel as effective with diverse populations,
and that these practices are related to disproportionate representations ofchildren of color in
county child welfare systems. Descriptive results revealed four overall types of practices
considered to be most effective with children and families of color, as well as various specific
practices within these types of practices including: 1) beginning choice point practices, which
included family group conferencing or other team decision-making practices, voluntary family
maintenance or other prevention/family preservation services, and risk assessment tools or
systems, 2) continuing care/ending choice point practices, which included, placement related
practices, the Family to Family Initiative, and intensive and collaborative interventions for highrisk children, 3) both beginning and continuing/ending choice point practices, which included,
community-based services, and strength based assessments, and 4) organizational or structural
practices which included, cultural competency of staff, collaboration with American Indian tribes
and vertical case management.
In general, very little research on the effectiveness of the practices identified by
respondents has been conducted, and in cases where evaluations have been performed, results are
generally inconclusive. More research is needed to evaluate the effectiveness of these services
for diverse populations.
Multivariate analyses revealed several significant associations between types of
practices, specific practices and practice characteristics and levels ofdisproportionate
representation ofchildren of color in county child welfare systems, when other factors such as
urban-rural status and per capita income were statistically controlled
Beginning choice point practices were found to have a beneficial effect on Latino
overrepresentation. In particular, family group conference or other team decision-making
interventions predicted lower levels of Latino overrepresentation. However, Latino
overrepresentation was greater, not less, when interventions were implemented targeting areas
with high concentrations of distressed families. Latino overrepresentation was also greater when
clients, foster families, out-of-home placement providers, and other service systems were
involved in the development of the practice. As such, it appears that practices may be developed
because stakeholders perceive Latino overrepresentation and initiate programs in response.
In contrast to Latinos, targeting practices to at-risk areas for African American
populations was beneficial in reducing overrepresentation, as well as targeting of very young
children. Less overrepresentation was linked also to whether foster families and clients and
contract agencies initiated an intervention, and whether it focused on improving psychosocial
190
outcomes of birth families. The findings suggest a picture almost the reverse of that found for
Latinos.
Additionally, there appear to be lower levels ofAfrican American overrepresentation
than expected when key community stakeholders help to initiate the interventions that focus on
high-risk areas and populations. When interventions have a goal of strengthening the well being
of birth families, then implementation is associated with less African American
overrepresentation. Yet when counties implemented Family-to-Family initiatives greater African
American overrepresentation resulted. Again, the implementation of Family to Family programs
may be used reactively, in response to high perceived levels of overrepresentation.
As with Latinos, targeting ofhigh need areas was less beneficial for Asian
American/Pacific Islander children than not engaging in this kind of intervention. Involving
clients and service systems in initiation, again more common for interventions targeting high
need areas, was also relatively unsuccessful. On the other hand, management and worker
support for a practice was associated with greater success.
Implications and Recommendations
•
County directors and child welfare administrators identified practices with similar
characteristics as those identified within Santa Clara County, i.e., those that are
inclusive, collaborative, culturally appropriate and involve client's communities
appear to be best suited to meeting the needs of children and families of color.
However, further research and evaluation is necessary to determine which
practices most effectively meet the needs of individual racial/ethnic groups.
•
While multivariate analyses suggest that certain practices do predict levels of over
and underrepresentation of children of color in the child welfare system,
additional research to determine the mechanisms through which these practices
impact this disproportionate representation is needed.
•
Additional explorations into the relationship between practices, disproportionate
representation of children of color in the child welfare system and county
characteristics would also help shed light on the process through which certain
practices impact children and families of color.
Phase 3 Conclusion
Results from Phase 2 indicated that different racial/ethnic groups experience unique
pathways through the CWS (Hines et al., 2002). At every point in the system, various child
welfare practices and services impact children and families. Phase 2 results indicated that
children and families of color in Santa Clara County's CWS tend to be recommended services
that are traditional and formal in nature and that do not appear to meet the wide range of needs
experienced by these highly diverse racial/ethnic family groups. Indeed, children and families of
color in the CWS represent a high risk group who are often impacted by myriad psychosocial
191
challenges yet services recommended for these families tend to be limited to a one-size-fits all
approach.
Moreover, there exists relatively little empirical evidence that traditional child welfare
services are effective in maintaining or reunifying children and improving family functioning.
Even less research has been conducted on the impact of child welfare practices for children and
families of color. Yet because children and families of color are disproportionately represented
in the CWS, the effectiveness of child welfare practices for these groups is of particular interest.
Research suggests that children and families of color tend to ·have longer stays in the CWS, they
receive fewer and less comprehensive services, and have poorer case outcomes than White
children and families (Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996).
This bleak outlook for children and families of color in the CWS creates an urgent need to
understand child welfare practices and services that result in enhanced outcomes for these
families. It is especially important to identify the types of individual and contextual factors that
hinder or support the effectiveness of these practices.
The primary goal of Phase 3 was to assess key practices for children and families of color
in the CWS at both the county and state levels in order to provide information that might help
generate strategies aimed at reducing ethnic/racial disparities and develop more effective
services and supports for children and families of color. While little research exists that
empirically supports policies, programs and practices that attempt to reduce ethnic/racial
disproportionality, some programs and practices appear promising and merit further exploration.
In general, these practices aim to reduce the need for out-of-home placement and increase family
and community participation in decision-making. Please see Chart 1 for a summary of promising
child welfare practices from Phase 3.
In our analysis of Santa Clara County child welfare practices, results based on focus
group and interview data indicated that in-home services, family conferencing, and team decision
making {TDM) were viewed by participants as ways to improve decision-making with families
from diverse cultural and ethnic groups. Participants noted that both formal and informal
collaborative relationships with other services delivery systems, such as domestic violence and
substance abuse treatment systems, are critical for successful interventions with children and
families. Furthermore, participants recognized the importance of training and cross-training to
enhance the capacity of helping professionals in child welfare and other systems to better address
the needs of children and families of color.
Our analysis comparing South County and the Main Office enabled us to examine
contextual factors that may have an impact on child welfare practices, in this case, small vs. large
and urban vs. rural. Results indicate that children in South County appear to have better
outcomes, i.e., they have fewer number of unique placement homes in the current episode, have a
shorter average stay per placement, a shorter length of time in out-of-home placement, and are
younger at time of case closure. Study participants described South County as "a small county or
rural area." This geographic difference, and factors related to an organizational culture that
emphasize a commitment to maintaining and reunifying families, the practice of providing an
orientation to the child welfare system, as well as the finding that fewer workers are assigned
192
Chart 1: Promising Child Welfare Practices from Phase 3
Beginning Choice Points
Promis.ine Practices
Team decision making, case conferencing, family group conferencing
and/or family involvement in decision-making
Orientation for parents on child welfare and court processes
.: ; :.-~::_:,.' . ::,. ~~ :~=. ,,:--.;-:.:~ : _,~~.;;..:.::,,>!.~ :•..~·.: ~'."...,--..,_::11 ': ·:,._f~:;..-.-~.. 1" _;.:i:-:..:.~~ ', ~"',:.
4
Continuing Care/Ending
Choice Points
..... :,' 1::..;.1.~'',':": ~-j,,d•,~- ·,')·~.J-'.f;.i~·-."- : •. : · ,l ·--~~- ~~:. ,,."·'· \
,:...:;...:_.
Contextual Factors or
Organizational/
Structural Practices
-~
Voluntary Family Maintenance/Family Preservation Services
Child care and in-home services
Risk assessment tool or system
.i ~-~~~!:__··:~·. :,-~---..~:.-:,---;~· l- -:.:;.;;:_'~-·~. ~,.;.:~. ~~-: ·~~-;.• ~; .."~'-=J~-~~·•:~\"; ?:;""': ,"":"-....,._
._h• .. ,~ :):. ·:~ '·.~ 1.-::~.}-"bic;=::,-:..' :~~-~.'=-!~·..•
i~.-~~:-~~:·:-::::,:_.-··:..:'.\t'~.:. --~::
;_.,;._. :t.::_: ;.,.::.:..
.. ,••·.,.
Parenting education services involving both child and parent
Cultural matching for children in foster care
Recruitment, training and coaching for foster parents
Placement-related practices
Family-to-Family Initiative
Intensive and collaborative interventions for high risk children
Community based services (both beginning & continuing/ending)
Strength-based assessments (both beginning & continuing/ending)
,,f·J ·l 5.-.,l'""·"t:-~'\i' ':_"':. .. ,_·7~·i-a·t_:~.J""'t..:;::----~-z.:Jt:;:-,;;. ..1;.:,?;:;c-:.:• r:~-~:!~~-l. ... _" '.'~-~=.r~:t~--"":_1;.-e,~:;:v !" l"1i•,;-.,.,r...,,.~.;..; .... :··-;;-~~·~~~•-W""""· ,\. -~·---...~~-=-~;1i!;._~~~-~-'
-. -~:A
.,.,,t
of
Cultural competency staff
Collaboration with American Indian tribes
'.< ~--: -•:•. -~J:•;;.-:.-"·•:-:-,;:.:.:
~-: .:·:,,:-t:-:
.i.: .~·1•_:.:-;~;:.c...:.::::..~· ~·:.:·-::-~•~'!"r.--~:~"!'" :-:~~· ~~y;-~:~:~·:1•.,::_;-;:,-,:-:j.~.:::::
Main County Focus Groups
Main County Focus Groups
Main County Focus Groups
Statewide Survey
Statewide Survey
Statewide Survey
Statewide Survey
Statewide Survey
f_.,:,,.: ... ,J t: ~-r'-:-_:r:~_-.:.:-...~· ;.. ,-x•_.;$.!.it...,S.---!":~:?,_;,,:;_ --;=-·=•:~::::..i.,;Jtf;'°.fi-;.:r'£.,.::_-~;::·.::.-:i.1~ ~-.--e,;,".~:.1?;;: ::;_:;-:
Cultural matching between staff and clients, and cultural consultation
Supportive supervision for staff
Collaboration with other agencies or systems
Strength-based approach to services
Small county dynamic
Culture of commitment to maintaining and reunifying families
Vertical case management
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Study Component
Main County Focus Groups
South County Focus Groups
Statewide Survey
Main County Focus Groups
South County Focus Groups
Statewide Survey
Main County Focus Groups
Statewide Survey
•
•
•
•
•
•
•
•
•
•
Main County Focus Groups
Main County Focus Groups
South County Focus Groups
Main County Focus Groups
Main County Focus Groups
South County Focus Groups
South County Focus Groups
South County Focus Groups
Statewide Survey
Statewide Survey
Statewide Survey
193
during the course of the case may all be related to these positive outcomes for children in South
County. It is interesting to note, however that our multivariate analyses that examined case
characteristics and system-related factors related to reunification indicated that when taking other
factors into account, geographical location was not a significant predictor of reunification.
Rather, fewer workers, shorter case duration, and fewer placements were the strongest predictors.
More research on factors related to agency culture and geographical location as they impact child
welfare practice and outcomes for children and families of color is clearly warranted.
Research indicates that the court system is likely to have an impact on the practices that
are provided to children and families in the CWS. In an analysis conducted to determine the
extent to which court orders changed social worker recommendations at the
jurisdictional/dispositional hearing, results indicated that the court made changes to the initial
social worker in approximately half of the cases. Results further indicated that that ethnicity was
not related to whether or not a change occurred. Findings from qualitative analysis exploring the
relationship between the child welfare and court system painted suggested that the collaboration
between the child welfare system and courts is problematic. More research is this area is needed,
as is further investigation of ways in which relationships between court personnel and child
welfare workers have an impact on the experiences of children and families of color involved in
the CWS.
Results on statewide practices indicate that certain types of practices are considered by
county child welfare personnel to be effective with diverse populations, and that these practices
are related to disproportionate representations of children of color in county child welfare
systems. Similar to practices identified within Santa Clara County's DFCS, practices at the
statewide level were described as effective with children and families of color because they are
inclusive, collaborative, culturally appropriate, and involve client's communities. Although the
research literature does not provide convincing evidence of the effectiveness of these practices,
the fact that respondents reported these practices as their most effective with children and
families of color suggests that they may serve as promising models for children and families of
color and that more rigorous evaluations of these practices are needed. Phase 3 also attempted to
assess how the practices identified through the survey might affect differential representation of
children of color in the CWS. Significant associations were detected between practice types,
practices and practice characteristics and disproportionate representation and interestingly,
different patterns of association were found for Latinos, African Americans and Asian
American/Pacific Islanders.
While researchers have described the existence of ethnic/racial disparities in every aspect
of the CWS, little attention has been given to ways to address the existing disproportinality. It is
our hope that the information contained in Phase 3 of the Children of Color Study will help in the
development of more effective programs and practices for families and children of color at both
the county and state level. Overall, our findings from Phase 3 suggest that efforts to address
racial and ethnic disparities in the child welfare system should involve a more concentrated focus
on child and family well-being, involving the creation of family-centered and community-based
services that are inclusive and collaborative and specifically designed to meet the needs of the
diverse cultural and ethnic groups in the child welfare population.
194
VII. SUMMARY OF PHASES 1, 2 AND 3: OVERALL
CONCLUSIONS AND IMPLICATIONS FOR SANTA CLARA
COUNTY'S CHILD WELFARE SYSTEM
The central focus of inquiry for this project was to identify factors related to the
disproportionate representation of children of color in Santa Clara County's CWS. In order to
address the complexity of this issue, the CWRT elected to employ a multiphase/multimethod
approach that examined factors at the individual, family and system levels. The following
summarizes findings from Phases 1, 2 and 3 of the study and provides overall implications for
the County based on our results.
Phase 1, an exploratory phase, drew from three sources: research literature at the national,
state and county levels; Santa Clara County's management information system (CWS/CMS); and
focus group discussions with professionals in Santa Clara County who provide child welfare
services, as well as parents, caregivers and youth who are recipients of CWS services. Results
indicated that factors related to racial/ethnic disproportionality were multiple and complex and
necessitated investigation at the individual, family, system and community level. Based on
preliminary analyses and available literature, we argued that children from different racial/ethnic
groups were likely to be treated differently at specific decision-making or key choice points in
the system. The concern about possible differential treatment of children of color was
substantiated by the focus group results, as well as by CWS/CMS data that indicated that there
were significant differences in the number of months in placements by race/ethnicity. For
example, results indicated that African American children spent significantly more time in
placement than their White, Hispanic/Latino, Asian American/Pacific Islander, Vietnamese, and
Filipino peers. Also, according to preliminary findings from CWS/CMS, in Santa Clara County,
following placements in~ relative home, the second most frequent placement for African
American, Native American, White and Hispanic/Latino youth, was a Foster Family Agency
(FFA); for Asian American/Pacific Islander and Vietnamese youth it was a Foster Family Home
(FFH). Almost 14% of Native American and 18% of Asian American/Pacific Islander youth in
OHP were placed in the Children's Shelter in December, 2000, a percentage that was higher than
any of the other ethnic/racial groups. We also argued that little was known about specific
pathways through the system and ways in which the pathways differed for various racial and
ethnic groups. Our findings from Phase I indicated that much of the research on children in the
CWS had focused on factors related to movement in and out of the system, but that little
information existed on the actual experiences of children in care and the individual and family
characteristics that are associated with these experiences. While focus group members
frequently stated that once a child of color entered the system, it was very hard for that child to
exit, little information existed that could explain what happened once that child was in the
system.
Phase 2 was designed to focus on the actual experiences of children in care and the
individual and family-related characteristics that were associated with these experiences. Results
based on extensive, in-depth reviews of 403 closed child welfare case records, a parallel
descriptive analysis of I 720 closed cases within the CWS/CMS database,·and key informant
interviews with managers and supervisors in the County's DFCS indicated that families
195
belonging to each of four different racial/ethnic groups (Latino, African American,
Asian/Pacific Islander and White) were characterized by different constellations of risk factors,
and that once in the system, the children had different experiences that resulted in different
outcomes. Additionally, results indicated that services recommended for the families and
children were limited to a small array of traditional services - a one-size fits all approach - that
did not necessarily meet the needs of the culturally diverse families and children. (Please see
Attachment 8 for a summary chart of Phase 2 findings.) We concluded that the paucity of
recommended services indicated that children and families of color were not likely to be
provided with sufficient preventive and supportive services and that the traditional child welfare
services might not meet the particular needs of these unique and diverse groups. These findings
are supported by research literature indicating relatively little empirical evidence that traditional
child welfare services are effective in maintaining or reunifying children and improving family
functioning. Even less research has been conducted on the impact of child welfare practices for
children and families of color. Yet because children and families of color are disproportionately
represented in the CWS, we contend that the effectiveness of child welfare practices for these
groups is of particular interest.
Phase 3 focused on agency practice and ways in which current practice interacted with
child, family and cultural characteristics of different cultural/ethnic groups. The prhnary goal of
Phase 3 was to assess key practices for children and families of color in the CWS both at the
County and state levels in order to provide information that might help generate strategies aimed
at reducing ethnic/racial disparities and develop more effective services and supports for children
and families of color. Methods included semi-structured in-depth interviews and focus groups
with DFCS supervisors and managers, social workers and parents involved in the CWS; a
qualitative and quantitative comparison of child welfare practices and outcomes between the
DFCS South County and Main County offices; a reanalysis of cases from Phase 2 with
successful outcomes and an analysis of factors that contributed to those positive outcomes; a
supplemental data collection for our case record review sample so as to include information on
court ordered changes to the social worker recommendations at the jurisdictional/dispositional
hearing; and, a statewide survey of county child welfare directors, managers or supervisors in
California counties and quantitative analyses using a statewide database consisting of countylevel characteristics.
As mentioned earlier, Phases 1 and 2 confirmed that certain ethnic groups were overrepresented (i.e., Latinos, African Americans) while others were under-represented (i.e., Asian
American/Pacific Islanders) in the County's CWS. Phase 2 results also indicated that there were
significant ethnic differences in demographic, system-related, and psychosocial characteristics
and that many of these variations appeared at early choice points in the system. During Phase 3,
we were interested in examining ethnic/racial differences in outcomes at later stages of the
system. Using data from our Phase 2 Case Record Review sample and defining successful
outcomes according to US DHHS guidelines (i.e., safety, permanency, and family and child
well-being), we found that there were minimal ethnic differences in successful outcomes. Phase
3 results suggested that when success was defined according to DHHS guidelines, ethnic
differences at the conclusion of the case were not evident. One interpretation for this finding
might be that while there is a clear disproportion of ethnic groups in the system, once in the
system, children as a whole, regardless ofrace/ethnicity, are faring at about the same level.
196
Given that Phase 2 results found that children and families of color in Santa Clara County's
CWS were receiving a "one-size fits all" approach to services, an approach that may seem
equitable and logistically efficient, but perhaps, ultimately ineffective in serving a diverse group
of children and families.
In order to further examine child welfare practices at the County level and to gain a better
understanding than the one gleaned solely through the use of CWS/CMS data, we conducted a
series of semi-structured in-depth interviews with DFCS managers and supervisors and focus
groups with social workers and parents involved in the CWS. Our results indicated that in-home
services, family conferencing, and team decision-making (TOM) were viewed by participants as
ways to improve decision-making with families from diverse cultural and ethnic groups.
Participants noted that both formal and informal collaborative relationships with other service
delivery systems, such as domestic violence and substance abuse treatment systems, were critical
for successful interventions with children and families. Furthermore, participants recognized the
importance of training and cross-training to enhance the capacity of helping professionals in
child welfare and other systems to better address the needs of children and families of color.
We were also interested in gathering information from counties throughout California
regarding promising practices for families and children of color. This information was gathered
through a statewide survey administered to child welfare directors and program administrators.
Similar to findings on practices identified within Santa Clara County's DFCS, practices at the
statewide level were described as effective with children and families of color because they are
inclusive, collaborative, culturally appropriate and involve client's communities. Although the
research literature does not provide convincing evidence of the effectiveness of these practices,
the fact that respondents reported these practices as those that are most effective with children
and families of color suggests that they may serve as promising models for children and families
of color and that more rigorous evaluations of these practices are needed (Please see Chart I on
p. 192 of the Phase 3 Final Report for a summary of promising practices for children and
families of color in the CWS).
Although practices such as orientation of new clients; family and team decision-making;
preventive services; substance abuse services; cultural competency; collaboration with other
agencies and systems; a strength-based approach; and the use of relative placements were
described as having a positive impact on children and families of color, certain negative practices
and contexts were cited as possibly impeding their effective implementation. For instance,
having to screen out large numbers of inappropriate referrals; inconsistency in decision-making
practices; a shortage of services, particularly substance abuse and preventive services; lack of
client access to services; difficult protocols for placing children with kin; gaps in cultural
competency; time limits; and agency-level factors such as heavy caseloads, staff shortages,
substantial amounts of paperwork, lack of access to information about resources; and confusion
about the agency's overall mission and key policies all impeded implementation of the best
practices. These findings suggest that in addition to building on current positive practices for
children and families of color, CWS stakeholders should also actively work both internally and
in collaboration with partners to reduce barriers to these best practices.
197
Results from focus groups conducted during Phase 1 and key informant interviews held
during Phase 2, as well as current research literature suggest that agency organizational culture
and context may have an impact on child welfare practices and resulting family and child
outcomes. During Phase 3, we conducted both quantitative and qualitative analyses comparing
South County and the Main Office that enabled us to examine various contextual factors at the
organizational level, in this case, small vs. large and urban vs. rural, and their impact on
outcomes for children and families of color. Results indicated that children in South County
appeared to have better outcomes, i.e., they had fewer number of unique placement homes in the
current episode, had a shorter average stay per placement, a shorter length of time in out-ofhome placement, and were younger at time of case closure. Qualitative findings were congruent
with the quantitative data and provided some insight into the dynamics that might contribute to
this difference. South County study participants described a culture of commitment to
maintaining and reunifying families. This commitment was described as integral to the
philosophy of managers and social workers and reflected in the expectations communicated to
social workers by most supervisors and by peers. The practice of providing an orientation to the
child welfare system, time limits, the courts, and how to work with social workers might also
contribute to the differences found in the quantitative analysis. Study participants frequently
described South County as similar to a small county or rural area. This geographic difference,
and factors related to an organizational culture that emphasizes a commitment to maintaining and
reunifying families, the practice of providing an orientation to the child welfare system, as well
as the finding that fewer workers were assigned during the course of the case may all be related
to these positive outcomes for children in South County. Further, our multivariate analyses that
examined case characteristics and system-related factors related to reunification indicated that
when taking other factors into account, fewer workers, shorter case duration, and fewer
placements were the strongest predictors of family reunification.
The issue of disproportionate involvement of children of color in the CWS has long been
an issue of concern for CWS workers, clients, researchers and government and community
groups. More recently, it has been the focus of much national attention. Santa Clara County's
Children of Color Study is one of the few to examine this issue at the local level. We applaud the
County Board of Supervisors.and DFCS for their openness and willingness to undertake such a
project. We acknowledge that there are multiple stakeholders who may be interested in
identifying a "magic bullet" or a single, straightforward explanation for the system's
racial/ethnic disproportionality. Single, straightforward explanations are appealing as they might
lead to quickly implemented solutions. However, results of this study indicate that rather than
one primary causal factor, there appear to be numerous aitd interrelated factors associated with
the disproportionate involvement of children of color in Santa Clara County's CWS. It is our
contention, that factors that operate simultaneously and in complex ways at the individual,
family and system level with differing patterns across diverse racial and ethnic groups contribute
to disproportionality throughout the system. Overall, findings suggest that efforts to address
racial and ethnic disparities in the CWS should focus on the diverse needs of the different ethnic
and racial groups involved in the CWS and the design of more culturally specific and effective
prevention and intervention programs to meet those needs.
198
Implications for Santa Clara County's Child Welfare System
The following are suggestions based on results from all three phases of the Children of Color
Study. It is our hope that these recommendations may be helpful in guiding Santa Clara
County's continuing efforts to address the complex issue of racial/ethnic disparities in its CWS.
1. There is a need to expand available child welfare services to better serve families and
children ofcolor. Phase 2 results indicated that reliance on a small array of
traditional formal services does not appear to meet the needs of the highly diverse
ethnic/racial family groups involved in the County's CWS.
•
More preventive and early intervention services for vulnerable families of color
should be implemented and carefully evaluated. The findings from the qualitative
component of Phase 3 affirmed the value of many efforts that are already in place,
such as culturally specific ER response units and family resource centers, and
other that are in the process of implementation including efforts to reduce
caseloads, initiation of team decision making, and participation in the Family-toFamily initiative.
In addition, home visitation services may be exceptionally helpful in addressing
the needs of vulnerable, at-risk families from diverse racial/ethnic groups.
Research has indicated that programs of home visitation that promote positive
health-related behaviors in mothers of young children, competent care of their
children and linkage with needed health care and hu~an services, reduce rates of
criminality, problems related to substance abuse and child abuse and neglect
among young, unmarried, isolated, poor mothers.
• _The use of more non-traditional, culturally sensitive services is clearly -warranted.
Key informants in all three phases of the study underscored this point. A paucity
of social services, particularly multi-lingual services, was cited as a significant
barrier for many families of color. Interviewees discussed the shortage of
substance abuse treatment programs, particularly those geared for women with
children and people whose primary language is not English. Multi-lingual and
culturally appropriate domestic violence services, parenting classes, and other
social services were considered in need of development.
•
Piloting new, innovative services and evaluating their success with different
ethnic/racial groups could contribute to a more diverse array of culturally specific
programs for families and children.
•
Phase 3 results also indicated that practices that are inclusive, collaborative,
culturally appropriate and involve client's communities appear to be best suited to
meeting the needs of children and families of color. Although the research
literature does not provide convincing evidence of the effectiveness of these
199
practices, the fact that respondents reported these practices as those that are most
effective with children and families of color suggests that they may serve as
promising models for children and families of color.
•
More services targeted to fathers and programs that are formulated and delivered
within a family-based framework would be of benefit to families and children of
color involved in the CWS.
•
In addition to building on current positive practices, and initiating new programs
for children and families of color, CWS stakeholders should actively work
internally and in collaboration with partners to reduce barriers to implementing
such practices. Barriers such as large numbers of inappropriate referrals;
inconsistency in decision-making practices; a shortage of services, particularly
substance abuse and preventive services; lack of client access to services; difficult
protocols for placing children with kin; time limits; and agency-level factors such
as heavy caseloads, staff shortages, substantial amounts of paperwork, lack of
access to information about resources; and confusion about the agency's overall
mission and key policies were cited during Phase 3 focus groups as impediments
to the implementation of cited best practices.
2. There is a need to involve multiple social service systems in a comprehensive and
coordinated effort to meet the needs of children and families of color. Results from
Phases 2 and 3 indicated that the problems experienced by families across the
different racial/ethnic groups span multiple systems including: mental health, juvenile
justice, adult criminal justice, substance abuse, and welfare. Statistics presented in
the literature review section of the Phase 2 Final Report also indicate that families of
color are involved in systems other than child welfare in high numbers. Prevention
and intervention efforts should involve a deliberate and organized coordination of
these multiple systems. During Phase 3 interviews and focus groups, collaboration
with other agencies and systems concerned with the safety and welfare of children
and families of color was highlighted as a beneficial practice for children and families
of color.
•
Developing interagency formal agreements and connections with liaisons in
other agencies would be beneficial as it can facilitate collaboration and
minimize the potential for agencies to give conflicting messages and mandates
to clients. Phase 3 participants suggested that social workers were better able
to divert cases or get a more complete picture of the family when two agencies
are working together. Other agencies mentioned in this context include
probation, hospital, medical providers, family and mental health agencies,
CalWORKS, substance abuse treatment agencies, domestic violence agencies
and other entities that may be involved with clients.
•
The Greenbook project, an initiative to better coordinate domestic violence
and child welfare services, could serve as a model for interdisciplinary
problem solving and policy development. The Greenbook project successes
200
could help inform evolving collaboration with other key systems, such as the
substance abuse treatment system.
•
Other coll,aborative efforts such as outstationed social workers, co-location of
services and Family Drug Court could serve as models for best practices for
children and families of color. Social workers and supervisors, interviewed
during Phase 3, described the family drug court as a model for effectively
working with substance abusing families in the child welfare system.
Participants described having ER social workers at different sites such as the
Family Violence Center and police stations as a valued practice.
3. Ways in which Agency organizational context and culture impacts families and
children of color merits attention. It is possible that characteristics of a large
burea~cratic organization impede the development and facilitation of practices that
would be of benefit to families and children of color. For example, organizational
goals such as having a unified agency mission, a shared organizational culture and
personal contact with clients. may be difficult to achieve in the face of complex and
competing demands inherent in a large organization.
•
Practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County
Offices and adapted to other regions of Santa Clara County, as recommended by
study participants. This practice may be particularly helpful to low-income
families and families of color that may be intimidated by or unfamiliar with child
welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for other
regions of Santa Clara County, though the success of this practice may be linked
to the "small county" dynamic described by South County study participants.
Some participants in focus groups that took place at the main office mentioned
this model as promising while others asserted that specialized expertise in
different areas of social work was an advantage in providing quality services that
are informed by "the most current information."
•
Further examination of ways in which the culture of commitment to maintaining
and reunifying families is created, communicated, and continued merits attention.
In addition, an exploration of how this culture might be adapted to other parts of
the county would be beneficial.
4. More attention that focuses on ways in which system level changes at the federal,
state and local levels have an impact on families and children ofcolor is warranted.
With the passage of the Multi-Ethnic Placement Act-Interethnic Adoption Provision
(MEPA-IEP, 1996) and the Adoption and Safe Families Act (AFSA, 1997), safety,
permanency and expedited placements have taken precedence in the CWS. Results
based on focus group and interview data from all three phases of the study indicated
201
that the shift toward expedited placement and an emphasis on permanency has
certainly influenced the culture and ways in which social services are being delivered
in California in general, and in Santa Clara in particular and consequently raises
several major concerns for children of color in the County's CWS. In combination
with the new regulations, the characteristics of families, children and communities of
color including chronic poverty, substance abuse, lack of community social
organization or racial segregation that increase chances of entering the CWS may
create overwhelming barriers to successful reunification for children of color in the
CWS and keep them in the system longer.
•
Regular assessment of ways in which shortened timelines, early termination of
parental rights, bypass criteria, as well as changes in adoption regulations and
incentives have had an impact .on children and families of color would be
beneficial.
•
A close monitoring of the effect of new policies, implementation of new programs
or changes in the agency's organizational structure on outcomes for families and
children of color would contribute to a better understanding of the differential
impact these innovations might have on different racial and ethnic groups.
5. In order to better assess whether safety, permanency, and family and child well-being
are maintained and enhanced, records should include more complete and measurable
information. We acknowledge that child welfare case records were not originally
intended for research and that that improving the collection and condition of child
welfare data is no easy task given the issues of limitations of the data system,
compatibility among systems of data recording and storage, and training needs.
However, concerted efforts in this area are necessary if accurate and useful
information is to be obtained and used to provide feedback on the effectiveness of
programs and services in improving outcomes for families and children in the
County's CWS.
•
The amount of missing data and the manner in which information is assessed also
implies the need for more reliable and valid methods of recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their families or of
those adopted high or low? And compared to what? Implementing better record
keeping and data collection over time on key characteristics, based on an agreedupon operational definition would better address the question of effectiveness and
performance. Implementing these practices across counties and across states
would also facilitate evaluation.
6. Future research efforts should be directed toward examining neighborhood,
community, and other macro-level factors particular to Santa Clara County and ways
in which these factors interact with individual, family and system-related
characteristics to propel children into and maintain them in the CWS. Research
202
findings consistently point to a relationship between poverty and child maltreatment.
Characteristics associated with communities and neighborhoods of poverty including;
living in a high crime area, living in public housing, having larger numbers of
dependent children, and receiving welfare benefits might place children of color at an
increased risk of entering and staying in the CWS.
It was our intention in conceptualizing the original design and scope of work for the
Children of Color Study to examine community level factors and their interaction
with individual and family characteristics during Phase 3 of the study. However, due
to the interests of various stakeholders, the community component was not included
in the current study. It is our contention that in order to understand more fully the
disproportionate involvement of children and families of color in the CWS, the
processes by which individual, family and system level factors interact with
characteristics that are, perhaps unique to Santa Clara County and impact families and
children of color and their subsequent entry into the CWS need to be examined.
203
References
Akin, B. A., & Gregoire, T. K. (1997). Parents' views on child welfare's response to
addiction. Families in Society: The Journal of Contemporary Human Services, 393-404.
Alter, C. F. (1985). Decision-making factors in cases of child neglect. Child Welfare,
64(2), 99-111.
American Humane Association (2003). What is Family Group Decision Making? Online,
retrieved July 1, 2003, from:
http://www.americanhumane.org?site?PageServer?pagemane=pc_fgdm_what_is.
Ards, S. D., Myers, S. L., Malkis, A., Sugrue, E., & Zhou, L., (2003). Racial
disproportionality in reported and substantiated child abuse and neglect: An examination of
systematic bias. Children and Youth Services Review, 25(5/6) 375-392.
Atkinson, L., & Butler, S. (1996). Court-ordered assessment: Impact of maternal
noncompliance in child maltreatment cases. Children and Youth Services Review, 20(3), 185190.
Avery, R. J. (2000). Perceptions and practice: Agency e{forts for the hardest-to-place
children. Children and Youth Services Review, 22(6), 399-420.
Baird, C., & Wagner, D. (2000). The relative validity of actuarial- and consensus-based
risk assessment systems. Children and Youth Services Review, 22(11/12), 839-871.
Banach, M. ( 1998). The best interests of the child: Decision-making factors. Families in
Society: The Journal of Contemporary Human Services, 79(3), 331-341.
Barth, R. P. (1991). An experimental evaluation of in-home child abuse prevention
services. Child Abuse and Neglect, 15, 363-375.
Barth, R. P., Snowden, L. R., Broeck, E.T., Clancey, T., Jordan, C., & Barusch, A. S.
( 1987). Contributors to reunification or permanent out-of-home care for physically abused
children. Journal of Social Service Research, 9(2/3), 31-45.
Bath, H. I., Richey, C. A., Haapala, D. A. (1992). Child age and outcome correlates in
intensive family pr~servation services. Children and Youth Services Review, 14, 389-406.
Bay Area Social Services Consortium (2000). Concurrent Planning: The Influence of
Bypass and Poor Prognosis Indicators on Child Welfare Outcomes. Berkeley, CA: Author.
Beeman, S. K., Hagemeister, A. K., & Edleson, J. L. (1999). Child protection and
battered women's services: From conflict to collaboration. Child Maltreatment, 4(2), 116-126.
204
Berry, M. (1992). An evaluation of family preservation services: Fitting agency services
to family needs. Social Work. 37(4), 314-321.
Berry, M. (1991 ). The assessment of imminence of risk of placement: Lessons from a
family preservation program. Children and Youth Services Review. 13, 239-256.
Bowen, K. (2000). Child abuse and domestic violence in families of children seen for
suspected sexual abuse. Clinical Pediatrics. 39, 33-40.
Brown, T. G., Werk, A., Caplan, T., & Seraganian, P. (1999). Violent substance abusers
in domestic violence treatment. Violence and Victims. 14(2), 179-190.
Brooks, D., & James, S. (2003). Willingness to adopt Black foster children: Implications
for child welfare policy and recruitment of adoptive parents. Children and Youth Services
Review, 25(5/6), 463-489.
Brunk, M., Henggeler, S. W. & Whelan, J.P. (1987). Comparison ofmultisystemic
therapy and parent training in the brief treatment of child abuse and neglect. Journal of
Consulting and Clinical Psychology. 55(2), 171-178.
California Department of Social Services (2003). Structured Decision Making. Online,
retrieved July 14, 2003 from: http://www.childsworld.ca.gov/Structured 352.htm.
California Department of Social Services (2001). Wraparound Services Pilot. Fiscal
Questions and Answers. Online, retrieved, May 5, 2001 from:
http://childsworld.org/services/famcenwrap.htm
California Department of Social Services (1999). Panel on Race and Bias in Risk
Assessment Thirteenth National Roundtable on Child Protective Services July 7-9, San
Francisco, CA.
Camasso, M. J., & Jagannathan, R. (2000). Modeling the reliability and predictive
validity of risk assessment in child protective services. Children and Youth Services Review,
22(11/12), 873-896.
Carlo, P. (1993). Parent education vs. parent involvement: Which type of efforts work
best to reunify families? Journal of Social Service Research, 17(1/2), 135-150.
Carten, A. J. (1996). Mothers in recovery: rebuilding families in the aftermath of
addiction. Social Work 41(2) 214-223.
Carter-Black, J. (2002). Transracial adoption and foster care placement: Worker
perception and attitude. Child Welfare, 71(2), 337-370.
Casey Foundation (2003a). Team Decision Making. Online, retrieved July 13, 2003 from:
http://aecf.org/initiatives/familytofamily/tools/decisions/htm.
205
Casey Foundation (2003b). Family to Family Home Page. Online, retrieved July 13, 2003
from: http://www.aecf.org/initiatives/familytofamily/
Chaffin, M., Bonner, B. L., & Hill, R. F. (2001). Family preservation and family support
programs: Child maltreatment outcomes across client risk levels and program types. Child Abuse
and Neglect. (25), 1269-1289.
Chao, R., & Tseng, V. (2002). Parenting of Asians. In M. H. Bornstein (Ed.). Handbook
of Parenting Vol. 4 Social Conditions and Applied Parenting (pp. 59-93). Mahwah, NJ:
Lawrence Erlbaum Associates.
Cash, S. J., & Berry, M. (2002). Family characteristics and child welfare services: Does
the assessment drive service provision? Families in Society, 83(5/6), 499-507.
Children NOW. (2001). California Report Card 2001: County Information. Online,
retrieved, August 20 2003, from: http://www.childrennow.org/california/rc-2001/county-pdflist.htm#top2.
Children's Research Center (2003). Structured Decision Making. Online, retrieved
July15, 2003 from: http://www.nccd-crc.org/crc/decision.htm.
Clark, H. W. (2001). Residential substance abuse treatment for pregnant and postpartum
women and their children: Treatment and policy implications. Child Welfare, 80(2), 179-198.
Close, M. M. (1983). Child welfare and people of color: Denial of equal access. Social
Work Research and Abstracts, 19(4), 13-20.
Cohen, E. P. (2003). Framework for culturally competent decision-making in child
welfare. Child Welfare, 82(2), 143-155.
Courtney, M., Piliavin, I., Grogan-Kaylor, A., & Nesmith, A. (2001). Foster youth
transitions to adulthood: A longitudinal view of youth leaving care. Child Welfare, 80, 685-717.
Courtney, M. E. (1997). The politics and realities oftransracial adoption. Child Welfare,
76(6), 749-779.
Courtney, M. E., Barth, R. P., Berrick, J. D., Brooks, D., Needell, B., & Park, L. (1996).
Race and child welfare services: Past research and future directions. Child Welfare 75(2) 99-137.
Courtney, M. E. (1994). Factors associated with the reunification of foster children with
their families. Social Service Review, 68(1), 81-108.
Cuddeback, G. S., & Orme, J. G. (2002). Training and services for kinship and
nonkinship foster families. Child Welfare, 91(6), 879-909.
206
Dalgleish, L. I., & Drew, E. C. (1989). The relationship of child abuse ind~cators to the
assessment of perceived risk and to the court's decision to separate. Child Abuse and Neglect,
ll, 491-506.
Davidson-Arad, B. (2001). Predicted changes in children's quality of life in decisions
regarding the removal of children at risk from their homes. Children and Youth Services Review,
23(2), 127-143.
DePanfilis, D. (1996). Social isolation of neglectful families: A review of social support
assessment and intervention models. Child Maltreatment, 1(1 ), 37-52.
DePanfilis, D., & Zuravin, S. J. (2001). Assessing risk to determine the need for services.
Children and Youth Services Review, 23(1 ), 3-20.
DePanfilis, D., & Scannapieco, M. (1994). Assessing the safety of children at risk of
maltreatment: Decision-making models. Child Welfare, 73(3), 229-245.
Drake, B. (1996). Predictors of preventive services provision among unsubstantiated
cases. Child Maltreatment, 1(2), 168-175.
Dore, M. M. & Lee, J.M. (1999). The role of parent training with abusive and neglectful
parents. Family Relations, 48(3), 313-325.
dosReis, S., Zito, J.M., Safer, D. J., Soeken, K. L. (2001). Mental health services for
youths in foster care and disabled youths. American Journal of Public Health, 91(7). 10941100.
Doueck, H.J., English, D. J., DePanfilis, D., Moote, G. T. (1993). Decision-making in
child protective services: A comparison of selected risk-assessment systems. Child Welfare,
72(5), 441---452.
Downing, J. D., Wells, S. J., & Fluke, J. (1990). Gatekeeping in child protective services:
A survey of screening policies. Child Welfare, 69(4), 357-369.
Dunford, F. W. (2000). The San Diego navy experiment: An assessment of interventions
for men who assault their wives. Journal of Consulting and Clinical Psychology, 68(3), 468476.
Dutton, D. G. (1986). The outcome of court-mandated treatment for wife assault: A
quasi-experimental evaluation. Violence and Victims, 1(3), 163-175.
Eckenrode, J., Ganzel, B., Henderson, C.R., Smith, E., Olds, D. L., Powers, J., Cole, R.,
Kitzman, H., & Sidora, K. (2000). Preventing child abuse and neglect with a program of nurse
home visitation: The limiting effects of domestic violence. The Journal of the American Medical
Association, 284(11), 1385-1396.
207
Eckenrode, J., Powers, J., Doris, J., Munsch, J., & Bolger, N. (1988). Substantiation of
child abuse and neglect reports. Journal of Consulting and Clinical Psychology, 56(1 ), 9- 16.
English, D. J., Marshall, D. B., Coghlan, L., Brummel, S., & Orme, M. (2002). Causes
and consequences of the substantiation decision in Washington state child protective s~rvices.
Children and Youth Services Review, 24(11), 817-851.
Farmer, E. M., Wagner, H. R., Bums, B. J., & Richards, J. T. (2003). Treatment foster
care in a system of care: Sequences and correlates of residential placements. Journal of Child and
Family Studies, 12(1), 11-25.
Federal Interagency Forum on Child and Family Statistics. America's Children: Key
National Indicators of Well-being, 2001. Federal Interagency Forum on Child and Family
Statistics, Washington DC: U.S. Government Printing Office.
Fein, E., & Staff, I. (1993). Last best chance: Findings from a reunification services
program. Child Welfare, 72(1), 25-40.
Fetsch, R. J., Schultz, C. J., & Wahler, J. J. (1999). A preliminary evaluation of the
Colorado rethink parenting and anger management program. Child Abuse and Neglect. 23(4),
353-360.
Findlater, J.E., & Kelly, S. (1999). Reframing child safety in Michigan: Building
collaboration among domestic violence, family preservation, and child protection teams. Child
Maltreatment, 4(2), 167-174.
359.
Flanigan, J. ( 1954). The Critical Incident Technique. Psychological Bulletin. 4, 327-
Fluke, J. D., Yuan, Y. T., Hedderson, J., & Curtis, P.A. (2003). Disproportionate
representation of race and ethnicity in child maltreatment: Investigation and victimization.
Children and Youth Services Review, 25(5/6), 359-373.
Forehand, R., & Kotchick, B. A. (1996).·Cultural diversity: A wake-up call for parent
training. Behavior Therapy, 27, 187-206.
Fraser, J. A., Armstrong, K. L., Morris, J.P. & Dadds, M. R. (2000). Homes visiting
intervention for vulnerable families with newborns: Follow-up results of a randomized controlled
trial. Child Abuse and Neglect, 24(11), 1399-1429.
Garbarino, J., & Kostelny, K. (1992). Child maltreatment as a community problem. Child
Abuse and Neglect, 16, 455-464.
Garland, A. F., & Besinger, B. A. (1997). Racial/ethnic differences in court referred
pathways to mental health services for children in foster care. Children and Youth Services
Review. 19(8), 651-666.
208
Garland, A. F., Hough, R., Landsverk, J. A., McCabe, K. M., Yeh, M., Ganger, W. C., &
Reynolds, B. (2000). Racial/ethnic differences in mental health care utilization among children
in foster care. Children's Services: Social Policy, Research and Practice, 3, 133-146.
Gaudin, J.M., Wodarski, J. S., Arkinson, M. K., & Avery, L. S. (1990). Remedying child
neglect: Effectiveness of social network interventions. The Journal of Applied Sciences, 15( 1),
97-123.
Gillespie, J.M., Byrne, B., Workman, L. J. (1995). An intensive reunification program
for children in foster care. Child and Adolescent Social Work Journal, 12(3), 213-228.
Gregoire, K. A. & Schultz, D. J. (2001). Substance-abusing child welfare parents:
Treatment and child placement outcomes. Child Welfare 80(4), 433-452
Goerge, R. M. (1990). The reunification process in substitute care. Social Service
Review, 64(3), 422-457.
Gordon, A. L., McKinley, S. E., Satterfield, M. L., & Curtis, P. (2003). A first look at the
need for enhanced support services for kinship caregivers. Child Welfare, 81(1), 77-96.
Guterman, N. B. (1999). Enrollment strategies in early home visitation to prevent
physical child abuse and neglect and the "universal versus targeted" debate: A meta-analysis of
population-based and screening-based programs. Child Abuse and Neglect, 23(9), 863-890.
Haller, D. L., Knisely, J. S., Elswick, R. K., Dawson, K. S., Scholl, S. H. (1997).
Perinatal substance abusers: Factors influencing treatment retention. Journal of Substance Abuse
Treatment, 14(6), 513-519.
Harman, J. S. (2001). Mental health care utilization and expenditures by children in foster
care. The Journal of the American Medical Association, 285(5), 524.
Harris, M. S., & Courtney, M. E. (2003). The interaction of race, ethnicity, and family
structure with respect to the timing of family reunification. Children and Youth Services Review,
25(5/6), 409-429.
Hines, A. M., Lee, P.A., Drabble, L., Snowden, L. R., & Lemon, K. (2002). An
Evaluation of Factors Related to the Disproportionate Representation of Children of Color in
Santa Clara County's Child Welfare System: Phase 2 Final Report. Report submitted to Santa
Clara County, CA Department of Social Services, Family and Children's Services, August 31,
2002.
Hochstadt, N. J., Harwicke, N. J. (1985). How effective is the multidisciplinary
approach? A follow-up study. Child Abuse and Neglect, 9, 365-372.
209
Hosley, C. A., Gensheimer, L., & Yang, M. (2003). Building effective working
relationships across culturally and ethnically diverse communities. Child Welfare. 82(2), 157168.
Jellinek, M. S., Murphy, M., Poitrast, F., Quinn, D., Bishop, S. J., & Goshko, M. (1992).
Serious child mistreatment in Massachusetts: The course of 206 children through the courts.
Child Abuse and Neglect, 16, 179-185.
Johnson, P ., & Cahn, K. ( 1992). Improving child welfare practice through improvements
in attorney-social worker relationships. Child Welfare, 74(2), 383-394.
Jones, L. (1998). The social and family correlates of successful reunification of children
in foster care. Children and Youth Services Review, 20(4), 305-323.
Jones, L. (1993). Decision making in child welfare: A critical review of the literature.
Child and Adolescent Social Work Journal, 10(3), 241-262.
Kapp, S. A., McDonald, T. P., & Diamond, K. L. (2001). The path to adoption for
children of color. Child Abuse and Neglect, 25, 215-229.
Karoll, B. R. & Poertner, J. (2002). Judges' caseworkers', and substance abuse
counselors' indicators of family reunification with substance-affected parents. Child Welfare,
71(2), 249-269.
Karski, R. L. (1999). Key decisions in child protective services: Report investigation and
court referral. Children and Youth Services Review, 21(8), 643-656.
Kemp, S. P., & Bodonyi, J.M. (2002). Beyond termination: Length of stay and predictors
of permanency for legally free children. Child Welfare, 81(1), 59-86.
Kinard, E. M. (2002). Services for maltreated children: variations by maltreatment
characteristics. Child Welfare, 81(4), 617-645.
Kitzman, H., Olds, D, L., Henderson, C. R., Hanks, C., Cole, R., Tatelbaum, R.,
McConnochie, K. M., Sidora, K., Luckey, D. W., Shaver, D., Engelhardt, K., James, D., &
Barnard, K. (1997). Effect of prenatal and infancy home visitation by nurses on pregnancy
outcomes, childhood injuries, and repeated childbearing: A randomized controlled trial. The
American Medical Association, 278(8), 644-652.
Knepper, P. E., & Barton, S. M. (1997). The effect of courtroom dynamics on child
maltreatment proceedings. Social Service Review, 71(2), 288-309.
Knight, D. K., Logan, S. M., & Simpson, D. D. (2001). Predictors of program completion
for women in residential substance abuse treatment. American Journal of Drug and Alcohol
Abuse, 27(1), 1-15.
210
Kolbo, J. R., & Strong, E. (1997). Multidisciplinary team approaches to the investigation
and resolution of child abuse and neglect: A national study. Child Maltreatment, 2(1), 61-72.
Landsman, M. J. (2002). Rural child welfare practice from an organization-inenvironment perspective. Child Welfare, 81 (5), 790-819.
Landy, S., & Munro, S. (1998). Shared parenting: Assessing the success of a foster parent
program aimed at family reunification. Child Abuse and Neglect. 22(4), 305-318.
Lee, B. J., & Goerge, R. M. (1999). Poverty, early childbearing, and child maltreatment:
A multinomial analysis. Children and Youth Services Review, 21(9/10), 755-780.
Lewandowski, C. A., & GlenMaye, L. F. (2002). Teams in child welfare settings:
Interprofessional and collaborative processes. Families in Society, 83(3), 245-256.
Lewandowski, C. A., & Pierce, L. (2002). Assessing the effect of family-centered out-ofhome care on reunification outcomes. Research on Social Work Practice, 12(2), 205-221.
Lightbourne, W. (2002). Greenbook Project Update. Report submitted to the Santa Clara
County Board of Supervisors.
Lindsey, D. (1991). Factors affecting the foster care placement decision: An analysis of
national survey data. American Journal of Orthopsychiatry, 61 (2), 272-281.
Littell, J. H. (1997). Effects of the duration, intensity, and breadth of family preservation
services: A new analysis of data from the Illinois Family First Experiment. Children and Youth
Services Review, (19)2, 17-39.
Littell, J. H., & Schuennan, J. R. (1995). A synthesis of research on family preservation
and family reunification programs. Online, retrieved January 27, 2003 from:
http://aspe.hhs.gov/hsp/cyp/fp litrev .htm
Little Hoover Commission (2003). Still in Our Hands: A Review of Efforts to Refonn
Foster Care in California. Online, retrieved July 24, 2003 from:
http://www.lhc.ca.gov/lhcdir/report168.html
Maluccio, A. N. & Ainsworth, F. (2003). Drug use by parents: A challenge for family
reunification practice. Children and Youth Services Review, 25(7), 511-533.
Martin, L. M., Peters, C. L., & Glisson, C. (1998). Factors affecting case management
recommendations for children entering state custody. Social Service Review,--, 521-544.
McGuigan, W., Katzev, A. R., & Pratt, C. C. (2003). Multi-level determinants of
retention in a home-visiting child abuse prevention program. Child Abuse & Neglect, 27, 363380.
211
McGuigan, W., M., & Pratt, C. C. (2001). The predictive impact of domestic violence on
three types of child maltreatment. Child Abuse & Neglect. 25, 869-883.
McKay, M. M. (1994). The link between domestic violence and child abuse: Assessment
and treatment considerations. Child Welfare, 73(1), 29-39.
McMurty, S., & Lie, G. (1992). Differential exit rate of minority children in foster care.
Social Work Research a~d Abstracts, 28(1), 42--48.
McPhatter, A. R., & Ganaway, T. L. (2003). Beyond the rhetoric: Strategies for
implementing culturally effective practices with children, families and communities. Child
Welfare, 82(2), 103-124.
Mederos, F., & Woldeguiorguis, I. (2003). Beyond cultural competence: What child
protection managers need to know and do. Child Welfare, 82(2), 125-142.
Menanteau-Horta, D., & Yigzaw, M. (2002). Indicators of social well-being and
elements of child welfare in Minnesota rural counties. Child Welfare, 81(5), 709-735.
Metsch, L. R., Wolfe, H.P., Fewell, R., McCoy, C. B., Elwood, W. N., Wohler-Torres,
B. et al. (2001). Treating substance-using women and their children in public housing:
Preliminary evaluation findings. Child Welfare, 80(2), 199-220.
Miller, 0. A., & Gatson, J. (2003). A model of culture-centered child welfare practice.
Child Welfare, 82(2), 235-249.
Merrel, T. M., Elliot, J. D., Murphy, C. M., & Taft, C. T. (2003). Cognitive behavioral
and supportive group treatments for partner-violent men. Behavior Therapy, 34, 77-95.
Murphey, D. A., & Braner, M. (2000). Linking child maltreatment retrospectively to birth
and home visit records: An initial examination. Child Welfare, 79(6), 711-728.
Needell, B., Brookhart, M.A., Lee, S. (2003). Black children and foster care placement
in California. Children and Youth Services Review, 25(5/6) 393--408.
Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A., Lery,
B., Shaw, T., Dawson, W., Piccus, W., Magruder, J., & Kim, H. (2003). Child Welfare Services
Reports for California. Online, retrieved June 28, 2003, from: University of California at
Berkeley Center for Social Services Research website. URL:
<http://cssr .berkeley .edu/CWSCMSreports/>
Nicholson, B. C., Brenner, V., & Fox, R. A. (1999). A community-based parenting
program with low-income mothers of young children. Families in Society, 80(3), 247-253.
Olds, D. L., Eckenrode, J., Henderson, C.R., Kitzman, H., Powers, J., Cole, R., Sidora,
K., Morris, P., Pettitt, L. M., & Luckey, D. (1997). Long-term effects of home visitation on
212
material life course and child abuse and neglect: Fifteen-year follow-up of a randomized trial.
The Journal of the American Medical Association, 278(8) 637-644.
Ondersma, S. J., Malcoe, L. H., & Simpson, S. M., (2001). Child protective services'
response to prenatal drug exposure: Results from a nationwide survey. Child Abuse and Neglect.
~657-668.
Palmer, S. E. (1996). Placement stability and inclusive practice in foster care: An
empirical study. Children and Youth Services Review, 18(7), 589-601.
Palmer, S. E., Brown, R. A., & Barrera, M. E. (1992). Group treatment for aqusive
husbands: Long-term evaluation. American Journal ofOrthopsychiatry, 62(2), 276-283.
Patterson, G. R., Chamberlain, P., & Reid, J.B. (1982). A comparative evaluation of a
parent-training program. Behavior Therapy. 13, 638-650.
Pecora, P. J., Fraser, M. W., Bennett, R. B., & Haapala, D. A. (1991). Placement rates of
children and families served by the Intensive Family Preservation Services Programs. In M. W.
Fraser, P. J. Pecora, & D. A. Haapala (Eds.), Families in Crisis: The Impact oflntensive Family
Preservation Services (pp.149-180). New York: Aldine De Gruyter.
Pennell, J., & Burford, G. (2000). Family group·decision making: Protecting children and
women. Child Welfare. 79(2), 131-158.
Rossi, P.H., Schuerman, J., & Budde, S. (1999). Understanding decisions about child
maltreatment. Evaluation Review, 23(6), 579-598.
Reiff, M. I. (2001 ). Mental health care utilization and expenditures by children in foster
care. Journal of Developmental and Behavioral Pediatrics, 22(2), 151.
Reyes, B. I. (Ed.). (2001). A Portrait of Race and Ethnicity in California. San Francisco:
Public Policy Institute of California.
Russel, R. (1988). Role perceptions of attorneys and caseworkers in child abuse cases in
juvenile court. Child Welfare, 67(3), 205-216.
Sanchirico, A., & Jablonka, K. (2000). Keeping foster children connected to their
biological parents: The impact of foster parent training and support. Child and Adolescent Social
Work Journal, 17(3), 185-203.
Scannapieco, M. (1994). Home-based services program: Effectiveness with at risk
families. Children and Youth Services Review, 16(5/6), 363-377.
Schriver, J.M. (1998). Human Behavior and the Social Environment. Allyn and Bacon:
Boston.
213
Schuerman, J., Rossi, P.H., & Budde, S. (1999). Decisions on placement and family
preservation: Agreement and targeting. Evaluation Review, 23(6), 599-618.
Sheppard, D. I., & Zangrillo, P.A. (1996). Coordinating investigations of child abuse .
.Public Welfare, Winter, 21-31.
Sieppert, J. D., Hudson, J., & Unrau, Y. (2000). Family group conferencing in child
welfare: Lessons from a demonstration project. Families in Society, 81(4), 382-391.
Smith, B. D. (2003). How parental drug use and drug treatment compliance relate to
family reunification. Child Welfare, 82(3), 335-365.
Spaid, W. M., Fraser, M. W., & Lewis, R. E. (1991 ). Changes in family functioning: Is
participation in Intensive Family Preservation Services correlated with changes in attitudes or
behaviors? In M. W. Fraser, P. J. Pecora, & D. A. Haapala (Eds.), Families in Crisis: The Impact
oflntensive Family Preservation Services (pp.131-148). New York: Aldine De Gruyter.
Stevenson, J. ( 1999). The treatment and long-term sequelae of child abuse. Journal of
Child Psychology and Psychiatry, 40(1), 89-111.
Suleiman, L. P. (2003). Beyond cultural competence: Language access and Latino civil
rights. Child Welfare, 82(2). 185-200.
Taft, C. T., Murphy, C. M., Elliot, J. D., & Keaser, M. C. (2001). Race and demographic
factors in treatment attendance for domestically abusive men. Journal of Family Violence, 16(4),
385-400.
Templeman, S. B., & Mitchell, L. (2002). Challenging the one-size-fits-all myth:
Findings and solutions from a statewide focus group of rural social workers. Child Welfare,
fil.(5), 757-772.
Thieman, A. A., & Dail, P. W. (1997). Predictors of out-of-home placement in a family
preservation program: Are welfare recipients particularly vulnerable? Policy Studies Journal,
25(1) 124-139.
Toffalo, D. A. (2000). An investigation of treatment integrity and outcomes in
wraparound services. Journal of Child and Family Studies, 9(1), 351-361.
Tumlin, K. C., & Geen, R. (2000). The Decision to Investigate: Understanding State
Child Welfare Screening Policies and Practices. Online, retrieved, February 7, 2003 from:
http://www.urban.org/url.cfm?ID=309534
U.S. Census Bureau (2001). Overview of Race and Hispanic Origin, Census 2000 Brief.
Online, retrieved November 20, 2002 from: http://www.census.gov/prod/2001pubs/c2kbr01l.pdf.
214
U.S. Census Bureau (2000). State and County Quick Facts: California. Online, retrieved
March 3, 2003 from: http://quickfacts.census.gov/qfd/states/06000.html
U.S. Department of Health and Human Services, Administration for Children, Youth and
Families, Children's Bureau (2003a). Decision-making in Unsubstantiated Child Protective
Services Cases. Washington DC: U.S. Government Printing Office.
U.S. Department of Health and Human Services. (2003b). Protecting the Well-being of
Children. Online, retrieved April 16, 2003 from:
http://www.scf.hhs.gov/news/facts/chilwelfpr.htm
U.S. Department of Health and Human Services. (2002). The Adoption and Foster Care
Analysis Reporting System Report. Online, retrieved July 16, 2003 from:
·
http://www.scf.hhs.gov/programs/cb/publications/afcars/report7 .htm.
U.S. Department of Health and Human Services, Administration for Children, Youth and
Families (2001). Child Maltreatment 2001. Washington DC: U.S. Government Printing Office.
U.S. General Accounting Office (1998). Agencies Face Challenges Securing Stable
Homes for Children of Substance Abusers. GAO/HEHS 98-182. Washington DC: U.S.
Government Printing Office.
U.S. General Accounting Office (1997). Child Welfare: States' Progress in Implementing
Family Preservation and Support Services. GAO/HEHS 97-34. Washington DC: U.S.
Government Printing Office.
Vidal de Haymes, M., & Simon, S. (2003). Transracial adoption: Families identify issues
and needed support services. Child Welfare, 82(2), 251-272.
Wald, M. S., Woolverton, M. (1990). Risk assessment: The emperor's new clothes?
Child Welfare, 69(6), 483-511.
Walton, E., Fraser, M. W., Lewis, R. E., Pecora, P. J., & Walton, W. K. (1993). In-home
family-focused reunification: An experimental study. Child Welfare, 72(5), 473-487.
Wells, K., & Guo, S. (2003). Mothers' welfare and work income and reunification with
children in foster care. Children and Youth Services Review, 25(3), 203-224.
Wells, K., & Guo, S. (1999). Reunification and reenty of foster children. Children and
Youth Services Review, 21(4), 273-294.
Wells, K., & Whittington, D. (1993). Child and family functioning after intensive family
preservation services. Social Service Review, March, 55-83.
Wells, S. J., Fluke, J. D., & Brown, C.H. (1995). The decision to investigate: Child
protection practice in 12 local agencies. Children and Youth Services Review, 17(4), 523-546.
215
Westat, Inc. (2001a). Caseworker decision-making. In Assessing the context of
permanency and reunification in the foster care system. Online, retrieved January 27, 2003 from:
http://aspe.hhs.gov/hsp/fostercare-reunifO 1/chapter7 .htm
Westat, Inc. (2001b). Evaluation of family preservation and reunification programs.
Online, retrieved January 27, 2003 from: http://aspe.hhs.gov/hsp/fampres94/index.htm
Westat, Inc. (2001c). The role of race in parental reunification. In Assessing the context
of permanency and reunification in the foster care system. Online, retrieved January 27, 2003
from: http://aspe.hhs.gov/hsp/fostercare-reunito l /chapter6.htm
Whitney, P., & Davis, L. (1999). Child abuse and domestic violence in Massachusetts:
Can practice be integrated in a Public Child Welfare Setting? Child Maltreatment, 4(2), 158166.
Wolfe, D. A., Edwards, B., Manion, I., & Koverola, C. (1988). Early intervention for
parents at risk of child abuse and neglect: A preliminary investigation. Journal of Consulting and
Clinical Psychology, 56(1), 40-47.
Wolock, I. (1982). Community characteristics and staff judgments in child abuse and
neglect cases. Social Work Research and Abstracts, 18(2), 9-15.
1
I.
INTRODUCTION
An analysis of national data on the ethnic composition of children in out-of-home-placement
reveals that, compared to their presence in the general population, there is a disproportionate
involvement of children of color in the public child welfare system (CWS). Statistics indicate that
African American and Native American children are overrepresented in the CWS. African
American children represent 15 percent of the general child population, yet comprise approximately
42 per~ent of children in the CWS and Native American children constitute approximately 1 percent
of the child population and represent approximately 2 percent of the CWS. Though not all states
provide data on Hispanic/Latino children in the CWS, aggregate statistics from those that do reveal
that while Hispanic/Latino children make up approximately 16 percent of the national child
population, they comprise 15 percent of the CWS population. Asian American/Pacific Islanders and
Whites also tend to be underrepresented at the national level. Asian American/Pacific Islander
children comprise approximately 4 percent of the general child population and 1 ·percent of the
CWS, while White children who are approximately 64 percent of the general e:hild population,
constitute 36% of children in the CWS (Federal Interagency Forum on Child and Family Statistics,
2001; U.S. Census Bureau, 2001; U.S. Department of Health and Human Services, 1999; U.S.
Department of Health.and Human Services, 2001).
Further analysis at the California level provides important information on differences and
trends that may go undetected at the national level. In California, a large and ethnically diverse
state, African Americans constitute 6 percent of the general population (32 percent of whom are
below the age of 18), but represent approximately 32.9 percent of children in supervised foster care.
Hispanics/Latinos comprise approximately 32 percent of the general population in California with
· 43 percent being below the age of 18, and represent 36.2 percent of children in the state CWS.
Native Americans represent 1.0 percent of California's population and constitute approximately 1.2
percent of the children in the CWS. Asian American/Pacific Islanders are under-represented in
California's CWS, as they constitute approximately 11 percent of California's general population,
with approximately 30 percent being younger than 18, but represent just 2.1 percent of the children
in the CWS. While Whites comprise approximately 47 percent of the population in the state, with
only 20 percent being below the age of 18, they constitute 27.5 percent of the children in
California's CWS (Needell, Webster, Cuccaro-Alamin, Armijo, Lee, Brookhart et al., 2003.; Reyes,
2001; U.S. Census Bureau, 2000).
In Santa Clara County, there also exists a disproportionate representation of children of
color in the County's CWS. When compared to the general population, African American,
Hispanic/Latino and Native American children are overrepresented in Santa Clara County's CWS,
while Asian American/Pacific Islander and White children are underrepresented. African Americans
represent 4 percent of the general child population in the county, and are 13.6 percent of children in
supervised foster care. Hispanic/Latino children represent 30 percent of the general child population
in Santa Clara County and constitute 55.3 percent of the child welfare cases. Native Americans are
approximately 0.5 percent of Santa Clara County's population and represent 0.6 percent of children
in the CWS. Asian American/Pacific Islander children represent 21 percent of the general county
child population and 5.4 percent of children in the CWS; Whites constitute 45 percent of the general
child population and 25.0 percent of the child welfare population (Needell et al., 2003; U.S. Bureau
of the Census, 2000).
2
Project Description
The disproportionate involvement of children of color in the CWS has long been an issue of
concern for CWS workers, clients, researchers and government and community groups; yet few
studies have systematically investigated the factors associated with this disproportionality. In an
effort to better understand the factors related to the disproportionate number of children of color in
the CWS in Santa Clara County, the Department of Family and Children Services (DFCS)
contracted with the Child Welfare Research Team (CWRT) in the College of Social Work at San
Jose State University to conduct a three-year study on this topic.
The overall question posed by DFCS was: What are the primary reasons why children of
color are disproportionately represented in Santa Clara County's Child Welfare System? In order to
address the complexity of this question, the CWRT elected to employ a multiphase/multimethod
approach, beginning with an initial exploratory phase that was completed in May 2001. Data for
Phase I were gathered from three sources: research literature at the national, state and county levels;
Santa Clara County's management information system (CWS/CMS); and focus group discussions
with professionals in Santa Clara County who provide child welfare services, as well as parents,
caregivers and youth who are recipients ofCWS services. Four central themes emerged from Phase
I findings {please see Phase I Final Report for a complete description of the findings from which
these themes were generated). The four emerging themes from Phase I included:
I). Little is known about specific pathways through the CWS and ways in which these
pathways differ for various racial/ethnic groups. Much of the research on children in the
CWS focuses on factors relates to movement in and out of the system. The specific
pathways through the system for children of various ethnicities is largely unknown, and may
provide important insights into the factors related to the disproportionate involvement of
children of color. Focus group participants in Phase I frequently stated that once a child of
color enters the system, it is very hard for that child to exit.
2). Various racial/ethnic groups may receive different treatment at key decision-making
points in the system. Differential treatment of children of color in both the initial reporting of
child maltreatment and throughout the various stages within the CWS may exist. CWS/CMS
data for Santa Clara County corroborate the possible differential treatment of children of
color at various points in the system. Phase I analysis revealed that African American
children spent significantly more time in placement than their White, Hispanic/Latino, and
Asian American/Pacific Islander peers. Differences in placement type were also foundalthough kincare (placement with a relative) was the most frequent out of home placement
(OHP) type across ethnicities, the second most frequent OHP type varied among
racial/ethnic groups. For African American, Native American, White, and Hispanic/Latino
children, the second most common OHP type was a Foster Family Agency. For Asian
American/Pacific Islander children, the second most frequent placement was a Foster
Family Home. Asian American/Pacific Islander (18%) children were also placed at the
Children's Shelter at higher percentages than children of other racial/ethnic groups. In
addition, focus group participants in Phase I expressed concern for possible racial/ethnic
differences in ways in which services and resources were allocated and ways in which the
judicial system handled cases.
3
3). System level changes recently initiated at the federal, state and local level will
undoubtedly have an impact on children ofcolor in the CWS. Recent federal policies,
including the Multi-Ethnic Placement Act-Interethnic Adoption Provision (MEPA-IEP,
1996); the Adoption and Safe Families Act (AFSA, 1997); the Personal Responsibility and
Work Opportunity Reconciliation Act (PRWORA, 1996); as well as state and local county
policies all impact the ways in which families and children of color enter and stay in the
CWS. Recent policy shifts toward expedited permanent placements for children in out-ofhome care and shortened timelines for reunification, as well as new restrictions for
Temporary Assistance to Needy Families (TANF) recipients (more than half of children in
foster care are from welfare eligible families) may increase the likelihood of children of
color entering and staying in the CWS-although much more research is still needed to
determine this association.
4). Factors related to the disproportionate representation ofchildren of color in the CWS
are multiple and complex. No clear consensus from the research literature exists on how
families and children of color become and stay involved in the CWS. Available evidence
suggests that, rather than one primary cause, there appear to be numerous and interrelated
factors associated with disproportionate rates of children of color in the CWS. Factors found
to be associated with CWS involvement for children of all ethnicities, include but may not
be limited to: parental mental illness substance abuse, and incarceration; spousal abuse;
living in poverty; living in neighborhoods with concentrated poverty, and living in
communities with low levels of social organization. For children of color specifically,
research also indicates that race and class biases in initial reporting and subsequent CWS
service delivery do exist.
Conclusions from Phase 1 indicated that more research was needed that focused on the
actual experiences of children in care and the individual and family-related characteristics that are
associated with these experiences. Phase 1 results also indicated that CWS/CMS data alone would
not provide the information needed to address these research aims. Thus, Phase 2, which concluded
in August 2002 focused on characteristics of children and families of color in the CWS and
experiences they had as they wended their way through the system. The methodology for Phase 2
included extensive, in-depth case record reviews with a sample of 403 closed child welfare case
records, enabling us to collect information on the individual child and family, as well as systemrelated characteristics not included in the CWS/CMS database. The case record review data
provided extensive information on decisions at key points in the system and children's progress
through the system, as well as services ordered for children at the time of case disposition and at
subsequent hearings. In addition, a parallel descriptive analysis of 1720 closed cases within the
CWS/CMS database allowed us to examine the entire histories of children as they progressed
through the CWS and to conduct analyses that were not possible with the CWS/CMS crosssectional dataset that we used in Phase 1. Key informant interviews with managers and supervisors
in the county's DFCS at key decision points in the system provided additional information
pertaining to cultural and environmental aspects. of the agency. (Please see Phase 2 Final Report for
a complete description of the study methods and findings). Based on analyses conducted during
Phase 2, five overall conclusions were drawn:
1). Families belonging to each ofthe four racial/ethnic groups studied, including: African
Americans, Hispanic/Latinos, Asian American/Paci.fie Islanders and Whites are
characterized by unique constellations ofriskfactors.
4
2). The characteristics that emerged in our analyses as key characteristics of each ofthe
four racial/ethnic groups have been associated in the research literature as risk/actors/or
child abuse and neglect. These characteristics have also been associated with increased
serious behavioral and adjustment problems in children and adolescents.
3). Once in the child welfare system, children in each ofthe four racial/ethnic groups follow
different pathways and experience different outcomes.
4). The services ordered/or families ofcolor are generally limited to aone-size-fits all
approach and to a small array of available services. The reliance on traditional formal
services does not appear to meet the needs of these culturally diverse families and children.
5). There is a need to involve multiple social service systems in a comprehensive and
coordinated effort to meet the needs ofchildren and families ofcolor.
The current report provides findings from Phase 3 that was originally to be conducted from
September 2002 to August 31, 2003. However, due to contractual changes involving the Scope of
Work, the CWRTwas not able to begin work until the end of January 2003. Thus, the research
involved in Phase 3 was conducted over a shorter seven-month tfme period. This third phase of the
research project focused on agency practice and ways in which current practice interacts with child,
family and cultural characteristics of different racial/ethnic groups. Phase 3 of the study addressed
three overall aims: 1) to assess identified key practices for children and families of color in Santa
Clara County's CWS, 2) to examine the influence of the court system on service recommendations
for children and families and, 3) to conduct a statewide comparative analysis of identified key
practices that may affect the disproportionate representation of children of color in the CWS.
The primary methodology for Phase 3 included semi-structured in-depth interviews and
focus groups with DFCS supervisors and managers, social workers and parents involved in the
CWS; a qualitative and quantitative comparison of child welfare practices and outcomes between
the DFCS South County and Main County offices; a reanalysis of cases from Phase 2 with
successful outcomes and an analysis of factors that contributed to those positive outcomes; a
supplemental data collection for our case record review sample to gather information on the role of
the court system in child welfare practice, including court ordered changes to the social worker
recommendations at the jurisdictional/dispositional hearing; and, a statewide survey of county child
welfare directors, managers or supervisors in California counties and quantitative analyses using a
statewide database consisting of county-level characteristics.
Overview of the Report
The next section of this report provides an overview of the study objectives and approach for
Phase 3, as well as the specific research questions that guide the analyses presented in each section.
The third section reviews the recent research on evidence-based child welfare practices and their
application to children and families of color. Section four focuses on the impact of child welfare
practices on children and families of color in the Santa Clara County's Department of Family and
Children Services (DFCS). Several related analyses were conducted to understand the types of
practices and case characteristics that are associated with more successful outcomes for children and
families from various racial/ethnic groups; the types of child welfare practices used in Santa Clara
County; how practices in the Main County Offices differ from South County Office practices; how
5
practices interact with child, family and cultural characteristics of differing racial/ethnic groups to
produce various outcomes; and the role of the court system on child welfare practices. Section five
describes results from our statewide comparative analysis of effective practices for children and
families of color in the child welfare system. A summary of the report, as well as recommendations
and conclusions are contained in section six.
Definition of Terms used in this Report
Racial/ethnic group refers to cultural heritage and country of origin. Racial/ethnic group is used
interchangeably with the terms, children of color and ethnic group.
Disproportionate representation ofchildren ofa particular racial/ethnic group refers to the
difference between the proportion of children of a particular racial or ethnic group in the CWS and
the proportion of children of a particular racial or ethnic group in the general population.
Overrepresentation of children ofa particular racial/ethnic group exists when the proportion of
children of a certain racial or ethnic group in the CWS exceeds its proportion in the general
population.
Underrepresentation ofchildren of a particular racial/ethnic group exists when the proportion of
children of a certain racial or ethnic group in the CWS is less than its proportion in the general
population.
African American refers to children of African American heritage. The term Black is included in
this category.
Asian American/Pacific Islander refers to children of Asian American and Pacific Islander heritage.
The following population groups are included: Asian Indian, Cambodian, Chinese, Filipino,
Guamanian, Hawaiian, Hmong, Japanese, Korean, Laotian, Polynesian, Samoan and Vietnamese
and other Asian American/Pacific Islanders.
Hispanic/Latino/a refers to children of Mexican American and Latin American heritage. The terms
Hispanic and Latino/a are included in this category.
Native American refers to children of Native American heritage and includes those designated as
Alaska Natives/Aleuts. The term American Indian is included in this category.
White refers to children of European heritage. The term Caucasian is included in this category.
Other category - Those in the Other ethnic subgroup included Native Americans, Ethiopians,
White-Armenians, and White-Middle Easterners
6
II.
OVERVIEW OF STUDY OBJECTIVES AND APPROACH
FORPHASE3
While Phases 1 and 2 of this study focused on investigating reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where they occur, the
overarching goal of Phase 3 was to focus on practices that take place within the CWS that either
promote or reduce disproportionality. More information about child welfare practices including
programs, services, strategies, policies, and/or tools can help address racial/ethnic disproportionality
by guiding the development of more effective services and supports for children and families of
color in the CWS.
In an effort to examine ways to address disproportionality within the CWS, the CWRT explored
three overall aims during Phase 3. The first aim included an assessment of identified key practices
for children and families ofcolor in the Santa Clara County's CWS. Specific research questions
related to this aim include:
•
•
•
•
•
•
What are the key practices in Santa Clara County's CWS that may have an impact on
children and families of color?
What empirical evidence exists for the effectiveness of these practices?
Have these practices been evaluated regarding their efficacy with varying racial/ethnic
groups?
How do these practices interact with child, family and cultural characteristics of differing
racial/ethnic groups involved in Santa Clara County's CWS?
How do practices in South County vs. the Main Office differ? (i.e., are there differences in
worker socialization, training, values/attitudes, stated and unstated expectations norms that
contribute to differences in practice and client outcomes?)
What practice characteristics are necessary in order to promote more positive outcomes for
children and families from different racial/ethnic groups involved in Santa Clara county's
CWS?
The methodology used to explore these research questions included 1) a review of the current
research literature on evidence-based practice for children in the CWS and the application of these
practices to children and families of color, 2) semi-structured in-depth interviews with DFCS
supervisors and managers, 3) semi-structured focus groups with DFCS social workers and parents
in the CWS, 4) a qualitative and quantitative comparison of child welfare practices and outcomes
between the DFCS South County and Main County offices, and 5) a review of cases from Phase 2
of the research with successful outcomes and an analysis of factors that contributed to positive
outcomes.
The second aim ofPhase 3 included an exploration ofthe influence of the court system on
child welfare practice, including court ordered changes to the social worker service
recommendations at the jurisdictional/dispositional hearing. We were interested in whether the
court system imposed any changes to the social workers' recommendations offered at the
jurisdictional/dispositional hearing in a child's case, and if these changes differed by the child's
7
ethnicity, as well as the nature of the relationship between the child welfare and court systems.
Specific research questions related to this aim include:
•
•
•
•
To what degree does the court system impose changes to the social workers' case plan
recommendations?
What types of changes to the social workers recommendations does the court system
order?
Do court ordered changes to the social workers' recommended case plan differ by child's
ethnicity?
How does the relationship between the child welfare and court system positively or
negatively impact child welfare practices and outcomes for children and families of color?
The methodology used to explore these research questions included case record reviews of the 403
closed child welfare cases used in Phase 2 of the research. Cases were reviewed again and
information on court ordered changes to the social worker recommendations at the
jurisdictional/dispositional hearing was gathered. Additionally, agency-wide focus groups provided
information on the how the relationship between the child welfare and the court system has an
impact on practices and outcomes for children and families of color.
The third aim ofthe research involved a statewide comparative analysis of identified key
practices that may affect the disproportionate representation of children ofcolor in the CWS.
Specific research questions associated with this aim include:
•
•
•
What are key practices in California counties that may have an impact on children and
families of color?
What practices are considered most promising by individual counties for children and
families of color?
Comparing counties with large racial/ethnic representation, what factors predict
over/underrepresentation of various racial/ethnic groups?
The methodology used to explore these research questions included 1) a statewide survey of county
child welfare directors, managers or supervisors in California counties and 2) quantitative analyses
using a statewide database consisting of county-level characteristics.
Phase 3 focused on exploring ways to address the disproportionality of children and families
of color within the CWS by examining practices that may either contribute to or reduce ethnic/racial
disparities. In order to accomplish the aims outlined above, we combined qualitative and
quantitative research methods and examined both practices within the system, as well as contextual
factors that can potentially have an impact on what transpires within the system. The report begins
with a review of the literature on effective practices for children and families in the CWS, followed
by a close up view of practices within Santa Clara County and then broadens to include a wider
perspective on statewide child welfare practices aimed at children and families of color. The
following section will describe the background and literature related to effective practices for
children and families of color in the CWS.
8
III. BACKGROUND AND LITERATURE REVIEW
Different racial/ethnic groups experience unique pathways through the child welfare system
(CWS) (Hines et al., 2002). At every point in the system, various child welfare practices and
services impact children and families. Phase 2 results indicated that children and families of color in
Santa Clara County's CWS tend to be recommended services that are traditional and formal in
nature and that do not appear to meet the wide range of needs experienced by these highly diverse
racial/ethnic family groups. Indeed, children and families of color in the CWS represent a high risk
group who are often impacted by myriad psychosocial challenges (see Phase 2 report for a full
description), yet services recommended for these families tend to be limited to a one-size-fits all
approach.
Moreover, there exists relatively little empirical evidence that traditional child welfare
services are effective in maintaining or reunifying children and improving family functioning. Even
less research has been conducted on the impact of child welfare practices for children and families
of color. Yet because children and families of color are disproportionately represented in the CWS,
the effectiveness of child welfare practices for these groups is of particular interest. Research
suggests that children and families of color tend to have longer stays in the CWS, they receive
fewer and less comprehensive services, and have poorer case outcomes than White children and
families (Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996). This bleak
outlook for children and families of color in the CWS creates an urgent need to understand child
welfare practices and services that result in enhanced outcomes for these families. It is also crucial
to identify the types of individual and contextual factors that hinder or support the effectiveness of
these practices.
This section will review the current research on evidence-based child welfare practices and
their application to children and families of color. Practices will be discussed in relation to key
choice points in the CWS including, beginning, continuing care, and ending choice points. Please
see Figure 1 for a flow chart of possible pathways and key choice points in Santa Clara County's
CWS. Additionally, contextual factors including cultural competency of workers and services; the
unique needs of rural child welfare populations; and the role ofthe court system will be discussed.
Across all stages of the CWS from beginning to end, highlights of practices that appear to be most
effective will be presented and the ways in which these best practices may impact children and
families of color will be discussed.
9
Figure 1: Possible Pathways and Key Choice Points In the Child Welfare System
Referral
(Report of Abuse or Neglect)
Emergency Response
(Screening)
Case Closed
Case Opened:
Further
Investigation
Jurisdictional Hearing:
Decision Regarding Substantiation
of Abuse/Neglect
Case Dismissed:
Unsubstantiated
Allegations
Family
Maintenance
Case Opened:
Voluntary Family
Maintenance
Case Closed:
Short-Term Services
Provided and Family
Stabilized Within 6
Months
Dispositional Hearing:
Substantiated Allegations,
Further Action and Services
Family Reunification:
Placement in FFH, FFA, Group
Home, Kincare, or Other Facility
Bypass
Hearings: Including 6-Month Review
Case Closed:
Family Stabilized
Through Family
Maintenance
Case Closed:
Family Stabilized
Through Family
Reunification
Case Closed:
Permanent Placement into
Adoption, Guardianship,
or Long Term Care
10
Impact of Child Welfare Practices on Children and Families of Color
Beginning Choice Points
When a child is reported to the CWS, there are various pathways that she or he may take
through the system (see figure 1). At the screening stage, the case may be closed altogether with no
further action; it may be opened for six months of voluntary family maintenance services; or the
case may be opened for further investigation. If the case is opened for further investigation, a
jurisdictional/dispositional hearing typically occurs in which there is a judicial finding on whether
the allegation of maltreatment is substantiated. If allegations are found to be unsubstantiated, the
case may be dismissed, however, if substantiation occurs, children may be placed in out-of-home
care, with family reunification services being ordered, or children may be maintained in the home
with orders for family maintenance services. In some cases, families may be bypassed for services
and have their parental rights terminated at this early stage in the system.
Following the pathway that children may take through the CWS, the screening process is the
first entry point into the CWS. On a national level, the U.S. Department of Health and Human
services ([U.S. DHHS], 2001) reports that approximately 2.7 million referrals were received by the
CWS in 2001; 32.7 percent of these referrals were screened out and closed with no further action.
Other national studies have found similar rates of screening out referrals, ranging from 36 percent to
42 percent (Tumlin & Geen, 2000; Wells, Fluke, & Brown, 1995).
Research suggests that certain case characteristics are associated with the decision to open a
case for further investigation or to refer to voluntary family maintenance services. For instance,
reports involving more than one child, children under age two, severe injury, a minor or other type
of injury or physical condition, allegation of sexual abuse, previous reports to the CWS, (Wells et
al., 1995); as well as child welfare cases involving AFDC receipt; alleged parental drug use; a
female primary victim; reports made by a family member; and reports with direct evidence (Karski,
1999) all have a higher likelihood of investigation than cases not involving these characteristics.
Additionally, cases involving children of color are opened for investigation at a higher rate
than cases involving White children. In Wells et al.'s (1995) analysis of 12 child welfare sites,
African American children· were significantly more likely to have reports investigated than children
of other ethnicities, however when the data was dis-aggregated and examined by sites, it was found
that almost the entire sample of African American children reported to the CWS came from two
sites that exhibited particularly high investigation rates. Yet, in an analysis of data from five states,
Fluke, Yuan, Hedderson, and Curtis (2003), found that African American children were
significantly overrepresented at the stage of investigation in each of the five states, as well as in all
counties within the five states, while White children tended to be underrepresented at the stage of
investigation. Rates of disproportionality at the investigation stage for other racial/ethnic groups,
including Asian American/Pacific Islander, Hispanic and Native American children showed much
wider variation. Depending on the state, these groups were both underrespresented and
overrepresented at the investigation stage.
Factors associated with the provision of voluntary services are somewhat less well
understood. In a statewide survey in Missouri, Drake (1996) analyzed factors associated with the
provision of preventive voluntary services to 3,442 children and families in the CWS, in
11
comparison to cases that were substantiated by the court. Cases were divided into three types of
maltreatment: sexual abuse, physical abuse and neglect. Across all three types of maltreatment,
cases from rural areas were significantly more likely to receive preventive services than cases from
urban or metro areas. Additionally, predictors of preventive service provision in sexual abuse cases
and neglect cases included those that were referred by a non-anonymous source. Physical abuse
cases were more likely to receive preventive services when a mandated reporter reported them.
Neglect cases were more likely to receive preventive services when they involved White children
and families, parents who were unemployed and children and families from non-poor
neighborhoods (Drake, 1996).
For children who are not offered voluntary family maintenance services, investigations
result in a decision regarding whether to substantiate the maltreatment and take further action or to
not substantiate and dismiss the case. Nationally, 27.5 percent of investigated referrals are
substantiated (U.S. DHSS, 2001). State data suggest similar substantiation rates. One study
conducted in New York state revealed a 30 percent substantiation rate across types of maltreatment,
with a 48 percent substantiation rate for physical abuse, 39 percent for sexual abuse and 28 percent
for neglect (Eckenrode, Powers, Doris, Munsch, & Bolger, 1988). Interestingly, there appears to be
little relationship between screening rates and substantiation rates. States with higher screening in
rates do not appear to have higher rates of substantiation of child maltreatment (Tumlin & Geen,
2000).
In California, the overall rate of substantiation of all referrals is 21.2 percent (Needell et al.,
2003). Statewide rates of substantiation between various racial/ethnic groups appear similar; Black
children have a 22 percent substantiation rate, White children 21.4 percent, Hispanic children 23 .4
percent, Asian children 24.5 percent, and Native American children 25.3 percent. Across all ethnic
groups, younger children have higher rates of substantiation than do older children (Needell et al.,
2003).
Santa Clara County's overall rate of substantiation in 2002 was 17.1 percent--a rate
somewhat lower than the state average. Santa Clara County also appears to have greater disparities
between racial/ethnic groups in substantiation rates. For instance, Black children in Santa Clara
County have an overall substantiation rate of 21 percent, White children 15 .4 percent, Hispanic
children 18 percent, Asian children 15 .1 percent, and Native American children 15 .I percent. Rates
of substantiation between racial/ethnic groups in Santa Clara County also depend on the age of the
child. As is the case within the state, younger children have a higher likelihood of substantiation,
however, in Santa Clara County it appears that Black children between O and 2 years of age have a
relatively high rate of substantiation (76.8%), whereas, White children 0-2 have a substantiation rate
of 62.6 percent, Hispanic children 52.6 percent; Asian children 46 percent, and Native American
children 66.6 percent (Needell et al., 2003).
Certain factors appear to be associated with the decision to substantiate an allegation. For
instance, reports from professional reporters tend to be substantiated at a higher rate than other
reports (Eckenrode et al, 1988; English, Marshall, Coghlan, Brummel, & Orme, 2002), as well as
cases referred by law enforcement (English et al., 2002). Poverty and early child bearing have been
identified as strong predictors of substantiation (Lee & George 1999). A history of prior referrals or
involvement in the CWS are also associated with higher rates of substantiation (Eckenrode et al.,
1988; English et al., 2002), and one statewide survey in Washington state found that referrals from
rural areas were more likely to be substantiated, although this effect was small (English et al.,
12
2002). Additionally, cooperation with child welfare agencies has been found to be associated with a
lower likelihood of substantiation (English et al., 2002). Factors associated with substantiation also
depend to some degree on the type of maltreatment alleged. For instance, sexual abuse reports that
involve older, female children in large households tend to have a higher substantiation rate than
cases not involving these characteristics, and the same is true for neglect reports that involve
younger children with previous child welfare referrals (Eckenrode et al., 1988).
·
Cases involving children of color appear to have an elevated likelihood of substantiation.
One large study conducted in New York state found that for allegations of physical abuse or
neglect, African American and Hispanic cases were substantiated at a higher rate than White cases
(Eckenrode, et al., 1988). In a more recent study conducted in Minnesota all types of maltreatment
were substantiated at a higher rate for non-White groups than for White groups; this effect remained
even after factors such as type of reporter, type of maltreatment, geographic location and
characteristics of the child and perpetrator were statistically controlled (Ards, Myers, Malkis,
Sugrue, & Zhou, 2003). Similarly, in Drake's (1996) statewide analysis of child welfare cases in
Missouri, non-White racial status was a significant predictor of substantiation even when other case
characteristics were statistically controlled. In contrast, Fluke et al.'s (2003) analysis of
disproportionate representation of race/ethnicity in investigation and victimization rates in five
states found little disproportionaility in rates of substantiation of maltreatment.
In addition to substantiation of an allegation, decisions at the beginning stages of the CWS
are also focused on whether to place a child in out-of-home care. Studies have consistently found
certain factors to be associated with the decision to remove a child from home. Children of color, in
particular African American children have an elevated likelihood of out-of-home placement
(Needell, Brookhart, & Lee, 2003; U.S. DHHS, 2001). Other factors associated with the decision to
remove a child include inadequate or low income (Berry, 1991; Lindsey, 1991 ); cases involving
prior maltreatment reports (Dalgleish & Drew, 1989; U.S. DHHS, 2001); cases involving families
with many children (Berry, 1991); lack of cooperation from the family (Dalgleish & Drew, 1989;
Davidson-Arad, 2001; Karski, 1999); cases involving young children, mothers as perpetrators, and
reports made by educational personnel are all associated with court intervention and placement is
substitute care (U.S. DHHS, 2001).
Few studies have examined factors associated with the decision to bypass family reunification
services and terminate parental rights at the early stages of the CWS. In California, any one of 13
bypass criteria can be used to deny reunification services and pursue an alterative permanent
placement. These criteria include:
1. Parents' whereabouts are unknown for six months.
2. Parental mental disability that prevents the utilization of services.
3. A sibling has been removed from the home, returned and then removed from parental custody
again.
4. Parent has caused the death of a child.
5. Parent caused severe emotional damage.
6. Parent committed severe physical or sexual abuse to the child, sibling or half-sibling.
7. Reunification has been terminated on a sibling or half-sibling due to indicators# 3, 5 or 6.
8. Parent has been found guilty of rape that conceived the child.
9. Willful abandonment constituting serious danger to the child.
I 0. Parent failed to reunify with a sibling or half-sibling.
11. Parent convicted of a violent felony.
13
12. Parent has extensive history of substance abuse and resisted treatment for three
years prior, or failed to benefit twice.
13. Parent does not want family maintenance or family reunification services, or does not want
the child removed.
In the Bay Area Social Services Consortium's ([BASSC], 2000) report on the use of bypass
criteria within San Mateo and Santa Clara Counties, approximately half of the cases reviewed had
bypass criteria present. Twenty six percent of these bypass cases were due to parental substance ·
abuse, 18 percent were due to a sibling with parental rights terminated or with a permanent plan,
and 13 percent had a parental mental disability that prevented the utilization of services (BASSC,
2000). However, not all cases with a bypass criterion were actually bypassed. Of the cases with a
bypass criterion present, 73 percent of cases involving parental mental disability; 38 percent of
cases with a sibling who had parental rights terminated or who has a permanent plan; and 13 percent
of cases involving parental substance abuse were actually bypassed. Overall 25 percent of the entire
sample were bypassed and denied reunification services. Additionally, BASSC (2000) reported that
the use of bypass criteria is not empirically or clinically valid and that bypass criteria are not
predictive of a lowered likelihood for reunification. Moreover, the report suggested that that since
many bypass indicators are common within the child welfare population, their use in denying family
reunification services may not be warranted.
Beginning Choice Points: Im pact on Children and Families of Color
Research clearly indicates that certain/actors are associated with various decisions at
beginning choice points and these factors disproportionately impact children and families ofcolor.
In general, the most consistent factors found to increase the likelihood of investigation,
substantiation or out-of-home placement include cases involving more than one child, young
children, previous reports to the child welfare system, and families with low incomes (Eckenrode et
al., 1988; English et al., 2002; Karski, 1999; Lee & George, 1999; Wells et al., 1995). Children and
families of color are at an increased risk for all of these factors, but in addition, studies have
indicated that even when these factors are statistically controlled, children of color, and in particular
African American children are more likely to have an allegation investigated and substantiated and
are also more likely to be placed out of the home (Ards et al., 2003; Drake, 1996; Wells et al.,
1995). Additionally, there is evidence to suggest that in cases of neglect, children and families of
color may receive voluntary family maintenance services less frequently than White children and
families (Drake, 1996). Taken together these findings suggest that beginning choice point decisions
may be biased against children and families ofcolor. Indeed it appears from the research that once
a child ofcolor is reported, they are more likely than White children to be set on a pathway through
the child welfare system that involves court intervention and out ofhome removal.
Decision-Making Practices at Beginning Choice Points
The types of decision-making practices that child welfare workers use at these early stages
of the CWS are crucial in determining the trajectory of the pathway that children and families take
through the system. However, decisions on whether to investigate a referral, provide voluntary
services, substantiate an allegation, remove a child or bypass a case are often complex, ambiguous
and multifaceted. Child welfare workers must simultaneously consider numerous factors, including
child safety, the best interests of the child, parental rights and preserving family integrity. These
14
factors must be considered in a complicated context; children and families in the CWS often face
multiple and complex challenges and child welfare policies and mandates fluctuate frequently.
Given this difficult context, it is not surprising that studies have found a significant lack of
consistency in child welfare decision-making practices across various stages of the system (Rossi,
Schuerman, & Budde, 1999; Tumlin & Geen, 2000; Westat, 2001a). For instance, disparities in
screening out rates between states have been found to range from a low of 5 percent to a high of 78
percent (Tumlin & Geen, 2000), while disparities between counties have been found to range from a
low of 1 percent to a high of 71 percent (Wells, et al., 1995). There is some evidence to suggest that
child welfare workers may use more restrictive screening practices when resources or staff are
reduced or when the number of referrals is high (Curtis, Boyd, Liepold, & Petit, 1995 as cited in
Tumlin & Geen, 2000; Hutchinson, 1989, as cited in Jones, 1993). Moreover, screening decisionmaking practices tend to be influenced by the average severity level of cases within the agency as a
whole and characteristics of the surrounding community. In one survey, workers in 11 child welfare
agencies in the state of New Jersey rated vignettes of child maltreatment reports as less severe in
agencies that handled the more severe cases of child maltreatment. These agencies were also located
in areas that were more socially and economically disadvantaged-suggesting that characteristics
specific to agency sites or communities influe·nce screening and substantiation decision-making
practices (Wolock, 1982).
Although virtually all states and counties screen child maltreatment reports to some degree,
screening practices and policies vary greatly between states and counties. National surveys suggest
that a majority of states do have some form of written agency policies to guide screening and
investigation decisions (Downing, Wells & Fluke, 1990); yet few states have formal or explicit
procedures or instruments delineating reports that should be screened out (Tumlin & Geen, 2000).
Approximately 65 percent of states have a single review screening process in which the decision to
investigate is made by one worker (Tumlin & Geen, 2000). Moreover, the training of child welfare
workers who screen referrals has also been found to vary widely from state to state; with some
states having educational or work experience requirements for screening workers, while others do
not (Tumlin & Geen, 2000).
Studies investigating consistency between child welfare workers in their decision-making
practices indicate wide variability in these practices. For instance, Rossi et al. (1999) explored
decision-making practices among 130 child welfare workers and experts in three states. Results
indicated that although certain factors were associated with the decision to maintain a child in the
home, including families with no prior referrals and cases involving employed caretakers who
showed signs of engagement in services, the thresholds used to trigger an out-of-home placement
decision were not consistent. For instance, some participants decided on out-of-home care in cases
involving just one prior referral, while others recommended out-of~home placements in cases
involving two or more prior referrals. Using the same data set, Schuerman, Rossi and Budde (1999)
calculated the level of agreement between workers on decisions about services, as ranging from a
high of 64 percent to a low of 48 percent, indicating a considerable lack of consensus among
workers regarding out-of-home placement decisions.
Studies also suggest that child welfare workers consider numerous factors in their decisionmaking practices and that these factors are generally not considered in isolation from one another
(English et al., 2002). For instance, a recent study by the U.S. Department of Health and Human
15
Services (2003a) identified four overall factors that influence decisions regarding whether to
substantiate an allegation, including:
1) Case factors, such as child characteristics, severity of the maltreatment, chronicity of
maltreatment, caretaker characteristics, caretaker/child relationship, social and economic
factors, and perpetrator access to the child.
·
2) Decision-maker factors, such as caseworkers' individual tendencies, experience level,
relationships with co-workers, self-assessments of skills and degree of importance
placed on state policy.
3) Organizational factors, such as workload stress, and degree of supervisor support.
4) External factors, such as state policies.
All of these factors were found to impact the decision-making process of child welfare workers in
complex ways that may differ between individual workers, counties and states (U.S. DHHS, 2003).
Yet there is some evidence to suggest that decision-making practices within a particular
agency may be more consistent than between agencies. Alter ( 1985) has suggested that child
welfare workers tend to use heuristics or assumptions to guide decision-making and that workers
within a particular agency often develop a mutual, although unspoken understanding of the kinds of
cases that should be investigated. Indeed there does appear to be greater consistency in initial
decision-making practices within agencies than between agencies. For instance, Wells et al. 's
( 1995) study of 12 child welfare sites, found wide variability between the sites in their investigation
decision-making practices, yet when all other factors were held constant, the site of the report was
more significant than any other factor in predicting an investigation. These results suggest that the
type of decision-making practices common to a particular child welfare agency play an important
role in decision-making practices related at the beginning stages of the CWS.
In addition to the importance of agency norms in decision-making practices, the labels used
to describe families may also trigger certain decisions. For instance, Martin, Peters, and Glisson
(1998) surveyed 100 child welfare workers in five states and found that workers did not make
service and placement recommendations for children based on a uniform psychosocial assessment.
Instead, decisions were influenced more by the labels attached to children upon entry into care and
judicial decisions. Children whose files contained language such as "substance abuser, child of
substance abuser, sex offender, victim of sexual abuse and victim of physical abuse" (Martin et al.,
1998, p. 3), were more likely to be recommended mental health services and more restrictive
placements-regardless of their actual need as indicated in the psychosocial assessment. Similarly,
children who were adjudicated delinquent or who were placed in out-of-home care due to behavior
problems rather than maltreatment were also recommended mental health services and more
restrictive placements, regardless of their actual need.
The influence of agency norms, the use of implicit assumptions and the complexity inherent
in child welfare decisions has prompted many practioners and researchers to call for more objective
or collaborative decision-making practices. Structured risk assessment tools, multidisciplinary
teams, and family group conferencing are considered by many to offer a more objective, balanced
and inclusive method of decision-making at beginning choice points.
16
Decision-Making Practices: Risk Assessment Tools
Structured risk assessment tools are designed to assist child welfare workers in making
accurate assessments at various stages in the CWS. Risk assessment tools generally evaluate factors
that are considered to be associated with various outcomes including such items as prior referrals
and prior substantiated cases; the presence of substance abuse, mental health or economic problems,
family constellation, the type, severity and chronicity of the maltreatment and circumstances
surrounding the current incident (Camasso & Jagannathan, 2000) Risk assessment tools fall into
two general categories: 1) consensus based systems in which risk factors in the instrument are based
on the consensus ofjudgments made by experts in the field, and 2) actuarial systems in which risk
factors are identified based on empirical evidence of factors associated with future maltreatment
(Baird & Wagner 2000). There is some evidence to suggest that actuarial systems more accurately
capture levels of risk in families referred to the CWS (Baird & Wagner, 2000).
The use of risk assessment tools varies greatly from state to state and county to county, and
these tools can be used at various decision points in the CWS from intake to case closing (Wald &
Woolverton, 1990). Although risk assessment tools are widely used in CWS agencies, their ability
to accurately assess levels of risk is an area of considerable debate in the child welfare field (Wald
& Woolverton, 1990). Several researchers have noted that risk assessment instruments have only
minimal reliability and validity in predicting case outcomes (Camasso & Jagannathan 2000;
DePanfilis & Scannapieco, 1994; Doueck, English, DePanfilis, & Moote, 1993; Wald &
Wooverton, 1990).
In 1998, certain California counties, including Santa Clara County began using a
comprehensive risk assessment system, known as Structured Decision-Making (SOM). SDM has
several components including response priority, safety assessment, risk assessment, family needs
and strengths assessment, case planning and management, case reassessment, workload based
resource allocation, and the use of management information systems (Children's Research Center,
2003). These components are designed to assist workers in assessing the severity of allegations,
making structured decisions, managing workloads, minimizing the trauma associated with
maltreatment and preventing the recurrence of maltreatment (California Department of Social
Services, 2003). The risk assessment tool used in this system is an actuarial instrument that is
intended to estimate the likelihood that maltreatment will recur. In addition to the risk assessment,
SOM also calls for a family needs assessment in which family's strengths and resources are to be
considered in conjunction with the risk assessment. Although peer reviewed evaluations of the
SOM project are not available, the Children's Research Center (as cited by the California
Department of Social Services, 2003) reports that a 12-month evaluation on the SOM project was
conducted in Michigan. Results indicated that cases receiving SOM had 27 percent fewer new
referrals, 54 percent fewer new substantiated allegations, 40 percent fewer children removed to
foster care, and 42 percent fewer child injuries than children not receiving SOM. Although these
results are promising, more evaluations are needed to discern the effectiveness of SOM, specifically
for children and families of color.
Decision-Making Practices: Multidisciplinary Teams/Family Group Conferencing
The use of multidisciplinary teams to assist in child welfare decision-making has gained
popularity in recent years. Multidisciplinary teams are designed to better coordinate efforts between
child welfare, law enforcement, legal professionals and medical professionals in making decisions
17
about child welfare cases. These efforts are intended to improve the efficiency and effectiveness of
decision-making activities and also in some cases-to reduce the potential for traumatizing children
through repeated investigatory interviews (Sheppard & Zangrillo, 1996). Approximately two thirds
of states have been found to use multidisciplinary teams in making child welfare decisions (Kolbo
& Strong, 1997).
Research on the effectiveness of multidisciplinary teams in improving decision-making in
child welfare is lacking. In one of the few studies to address this issue, Hochstadt and Harwicke
(1985) evaluated the use of multidisciplinary teams for 180 children referred to the CWS for
suspected child maltreatment. The multidisciplinary team included child welfare workers, medical
professionals and psychological personnel who completed a multidisciplinary evaluation and
participated in a case conference in which specific service recommendations and treatment plans
were made. Although no comparison group was used in the study, findings suggested that the use of
multidisciplinary teams was associated with a high percentage of children and families obtaining
recommended services on their treatment plans. One hundred percent of children with a
recommendation to return home did so, and 92 percent with a recommendation to foster care
followed that recommendation. Other service recommendations that were received included visiting
nurse (76%), additional medical assessment (64%), infant program (64%) and individual
psychotherapy (59%). The authors suggest that the use of multidisciplinary teams can positively
influence the receipt of recommended services.
In an effort to discern how multidisciplinary teams in child welfare work together,
Lewandowski and Glen Maye (2002), conducted a survey of 165 child welfare and community
professionals who were a part of a multidisciplinary child welfare team in an urban Midwest
county. The sample included social workers, counselors, community members, education
professionals, law enforcement, legal professionals and medical professionals. Findings revealed
that 72 percent of respondents had team meetings at regularly scheduled times with support staff
assigned to the team, and 53 percent had a designated team chairperson. Forty eight percent of the
sample indicated that decision-making was most likely to occur around case planning, 32 percent
indicated that decision-making was least likely to occur at intake, and 28 percent reported that
decision-making occurs at all stages. Inadequate communication was identified as the greatest
barrier to effective team decision-making, followed by inadequate resources and inadequate
professional respect. Lastly, although most respondents (63%) agreed that family members should
be active and equal participants in the team decision-making process, only 35.5 percent reported
that family members attended team meetings.
Recently, more concerted attempts to involve family members and extended support systems
in decision-making have occurred. Many child welfare systems are moving toward the use of family
group conferences to improve decision-making. The use of family group conferences began in New
Zealand in 1989, when legislation mandating their use was implemented (Sieppert, Hudson, &
Unrau, 2000). Family group conferences, also referred to as family group decision-making, are
designed to bring together family members, relatives, and other support systems in order to make
decisions about a case. The family group conference is intended to identify the family's strengths
and resources; to develop a plan to ensure child safety and improve family functioning; and to foster
cooperation, collaboration and communication between families and professionals (American
Humane Association, 2003: Pennell, & Buford, 2000). These methods are based on the principle
that families themselves possess the most information about what decisions should be made; the
approach is intended to be family centered, strengths based~ and takes into consideration issues of
18
culture and community (American Humane Association, 2003).
In addition to family group conferencing, recent efforts have also been aimed at ''team
decision-making," a closely related concept that appears to widen the scope of participants in the
decision-making process. Team decision-making is one component of the Casey Foundation's
Family to Family Initiative, a broad child welfare effort aimed at improving outcomes for children
in the CWS (see page 31 for a discussion of the Family to Family Initiative). The team decisionmaking component of the initiative involves convening a broad range of participants to share
information about the family and develop a plan for· child safety, family maintenance or
reunification. The team decision-making meeting can involve any participants who can contribute to
the decision-making process, including family members, neighbors, foster parents, service
providers, child welfare workers and supervisors, other resource staff or other community
representations (Casey Foundation, 2003a).
Research on the effectiveness of family group conferencing and team decision-making on
outcomes for children and families is lacking, although some studies have investigated the process
of implementing family group conferences. For instance Sieppert, et al. (2000) investigated the
process of implementing a family group conferencing pilot project in Canada. The sample included
23 child welfare families with court involvement and one family group conference was conducted
for each family. Mothers were the most frequent family members to attend the conferences (91 %),
while 52 percent involved a father, 70 percent involved grandmothers, 57 percent involved aunts,
and ~5 percent involved grandfathers. Other members of the conferences included child welfare
workers (83%), and foster parents (30%). Findings revealed that family group conferences tended to
be resource-intensive. Workers performed a wide range of activities in preparing for and
implementing the family group conferences. An average of 17 telephone contacts, 2.4 in-person
contacts, and 3 letter contacts were made prior to each family conference. Workers also spent time
traveling to meet with prospective conference attendees. On average, workers spent 7 hours
preparing for each conference. Family conferences lasted an average of2.75 hours. At the end of
the conference, families were asked to rate their experience of the group. Seventy percent reported
being highly satisfied with conference preparation and members in attendance; 85 percent felt a
high degree of freedom to speak out in the group; 76 percent reported feeling involved in the
decision-making process and 65 percent felt highly satisfied with the decisions that resulted from
the conference. Unfortunately the ethnicity the families involved in the family group conferences
was not provided.
Decision-Making Practices at Beginning Choice Points:
Impact on Children and Families of Color
In the beginning stages of the child welfare system, the decision-making practices ofchild
welfare workers can have a substantial impact on children and families ofcolor. However in
general there appears to be considerable inconsistency in the types of decision-making practices
used in the child welfare field (Rossi et al., 1999; Schuerman et al., 1999; Tumlin & Geen, 2000;
Westat, 2001 a). Although many child welfare systems are moving toward more objective or
multidisciplinary approaches to decision-making, a majority of states still use a single review
process in which decisions to investigate are made by one worker (Tumlin & Geen, 2000); and few
states have formal policies and procedures for decision-making at early stages of the CWS (Tumlin
& Geen, 2000).
19
However, making child welfare decisions in isolation can have deleterious effects on
children and families of color. Research suggests that child welfare workers often use implicit
assumptions to guide decision-making. Ways in which each worker thinks about diverse families,
and the overall spoken and unspoken expectations and norms of a particular child welfare agency
can have an enormous impact on decision-making practices (Cohen, 2003). As Cohen (2003) notes,
The cultural and racial background of families influences the specific factors that workers
consider in assessing the severity of risk and level of intervention. Decisions are more likely
to be made on the basis of deficits in available resources, accepted agency practice, personal
values and biases, and notions of an ideal family, than by application of consistent case rules
(Chohen, 2003, p. 145).
It is speculated that in some cases, race/ethnicity may impact decision-making by resulting in a
higher level of intervention, while in other cases race/ethnicity may result in less intervention effort.
Much ofthis variation appears to depend on the worker and the context in which the worker makes
decisions (Cohen, 2003).
The use of risk assessment tools and multidisciplinary teams/family group conferencing may
have the potential to improve decision-making practices affecting children and families of color.
Although research on the effectiveness of risk assessment tools in assessing risk for future
maltreatment has not been consistently demonstrated in the research literature, the move toward
more objective ways of assessing risk at beginning stages of the child welfare system may be
promising for children of color. For instance, in one study conducted by the California Department
of Social Services (1999), 2,500 substantiated cases from counties that used the SOM system, were
evaluated to assess whether the risk assessment instrument used in SOM had any differential impact
on African American, Hispanic, or White children and families. Findings indicated that Whites had
slightly higher scores on the risk assessment instrument than did African Americans or Hispanics
and the authors concluded that the risk assessment instrument did not disparately affect African
Americans or Hispanics. While this finding is encouraging, other studies have suggested that risk
assessment instruments in general are not always able to accurately predict case outcomes Camasso
& Jagannathan 2000; DePanfilis & Scannapieco, 1994; Doueck et al., 1993; Wald & Wooverton,
1990). More research is needed on the impact of risk assessment tools in general, and their impact
on children and families of color in particular.
Additionally, more research on the effectiveness ofmultidisciplinary teams and family group
conferencing is needed The family centered, strength based and inclusive philosophy offamily
group conferences may be a particularly useful strategy for children and families ofcolor, many of
who!11 are experiencing numerous risk factors in a highly stressed environment. The active
engagement offamilies and support systems in decision-making at early stages of the child welfare
system may potentially result in enhanced outcomes for children and families ofcolor.
Yet there is some evidence to suggest that these approaches can be labor, time and resource
intensive (Sieppert et al., 2000). As such, the effective implementation of these practices may
require the use of increased support staff to assist in scheduling and arranging meetings, as well as
identifying persons to be included in the meeting. Additionally, because beginning choice point
decision-making practices are often made quickly and in the context ofcrisis situations, the effective
use ofmultidisciplinary teams and family group conferences requires coordinated efforts and
appropriate support staff. Clearly, more research is needed to determine how group decision-
20
making can be effectively integrated into child welfare agencies and how these practices impact
outcomes for children and families ofcolor.
Voluntary Family Maintenance Practices/Family Preservation Services
In cases where maltreatment is not substantiated, families may be referred to voluntary
family maintenance services. Because the overall goal of voluntary family maintenance services is
to prevent an out-of-home placement and further involvement with the CWS, these services are
considered beginning choice point practices. Much of the research literature on services and
practices aimed at preventing out-of-home placement refer to these services as "family preservation
services," (FPS) or "intensive family preservation services" (IFPS). Many researchers and
practioners have noted that services falling under the rubric of "family preservation services" can be
quite diverse, although they do tend to share the common goal of preventing child maltreatment and
out-of-home placements (United States General Accounting Office, 1997). The model of FPS that is
most often described in the research literature includes short-term, intensive and home-based
services that are aimed at family skill-building, while also including concrete services, such as
public assistance, child care, employment training or transportation assistance (Littell, 1997).
In general, there are numerous shortcomings in the research on the effectiveness of FPS;
many studies use small sample sizes, no comparison groups and often fail to adequately describe the
nature of the intervention and the characteristics of the clients. As a result, evaluations of FPS often
yield conflicting results. While some studies report that FPS do not result in lower placement rates
or improved family functioning (Chaffin, Bonner, & Hill, 2001; Littell & Schuerman, 1995; Littell,
1997; Westat, Inc. 2001b); other studies suggest that FPS do improve outcomes for children and
families (Pecora, Fraser, Bennett, & Haapala, 1991; Scannapieco, 1994; Spaid, Fraser & Lewis,
1991; Wells & Whittington, 1993).
There is also evidence to suggest that FPS outcomes vary depending on case characteristics.
A number of different studies have noted that FPS clients who experience an out-of-home
placement typically have more risk factors and less resources than families who avoid placement.
The particular risk factors affecting FPS clients who experience out-of-home placements have been
found to vary widely. For instance, FPS cases involving substance abuse, mental illness, criminal
behavior, disability, low income, or previous CWS involvement have all been linked to an increased
likelihood of an out-of-home placement (Fraser, et al., 1991; Thieman & Dail, 1997; Thieman et al.,
1990, as cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman,
1995).
Voluntary Family Maintenance Practices/Family Preservation Services:
Impact on Children and Families of Color
Research on voluntary family maintenance practices and FPS has not systemically
evaluated the efficacy of these programs for diverse children and families ofcolor. Many studies
fail to perform analyses examining the possible differential outcomes of these services for children
and families of color. In general, research on FPS has not consistently demonstrated that these
services are effective in preventing out-of-home placement or improving family functioning
(Chaffin et al., 2001; Littell & Schuerman, 1995; Littell, 1997; Westat, Inc. 2001b), and some
researchers have concluded that there is no reliable research that supports the effectiveness of these
services (Littell & Schuerman, 1995). However, some studies do suggest that FPS may be most
21
effective with lower-risk families (Fraser et al., 1991; Thieman & Dail, 1997; Thieman et al., 1990,
as cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman, 1995).
Yet because children and families of color referred to the child welfare system often possess
numerous risk factors (Hines et al., 2002), the use of traditional FPS for these diverse clients may
not be effective. Research does suggest that FPS that are most effective with high-risk clients are
those that are more intensive, and those that involve high client participation and the delivery of
concrete services in the home (Berry, 1992; Cash & Berry, 2002; Nelson & Landsman, 1992, as
cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman, 1995).
Home-Based Prevention Efforts
Research suggests that the use of home-based services may be effective in preventing
maltreatment and further involvement in the CWS (Olds, Eckenrode, Henderson, Kitzman, Powers,
Cole et al., 1997). Most research on the use of in-home services has focused on their efficacy for
pregnant women or mothers of young children. While the nature ofin-home services may vary, in
general these services are aimed at improving health behaviors of pregnant and parenting women,
facilitating parenting skills, increasing support systems and encouraging healthy parent child
attachments (Olds et al., 1997). Many practioners and researchers have suggested that the perinatal
period is an important window of opportunity and that services offered during this time have
increased potential to offset a number of negative outcomes, including child maltreatment
(Guterman, 1999).
For instance, in one study involving pregnant women in rural New York state, 324 women
participated in a home nurse visitation program that sought to improve health related behaviors
during pregnancy, parenting behaviors and the personal life course development of the mother
(Olds et al., 1997). Specific services included linking families with health and human services;
involving family members and friends in the care of the child; educational services to increase
parenting abilities; and assisting mothers in completing education, obtaining employment and
family planning. Participants received services for two years; outcomes assessed at a 15-year
follow-up indicated that compared to pregnant women who received no home visitation services,
mothers in the program had significantly fewer verified reports of child maltreatment in which they
were the identified perpetrator. The program appeared to have even more of a protective effect on
unmarried women and women experiencing income problems. These groups were the least likely to
have a verified child maltreatment report. Women in the program were also less likely to have
substance abuse problems, had fewer contacts with the criminal justice system, had fewer
subsequent pregnancies and were less likely to use public assistance (Olds et al., 1997).
Interestingly, a separate analysis of the same data set indicated that the protective effects of
home visitation in reducing the likelihood of child maltreatment did not hold true for mothers
experiencing domestic violence. Mothers in the home visitation program who reported more than 28
incidents of domestic violence during the 15-year follow-up period did not experience a reduced
likelihood of verified child maltreatment (Eckenrode, Ganzel, Henderson, Sinith, Olds, Powers et
al., 2000). These results suggest that different risk factors may have varying degrees of influence on
outcomes for home visitation program participants.
In addition to child maltreatment prevention, in-home services have also been identified as a
useful public health strategy to improve health related behaviors and outcomes for mothers and
young children. Kitzman, Olds, Henderson, Hanks, Cole, Tatelbaum et al. (1997) evaluated a home-
22
visiting program that involved in-home services delivered by nurses to 1,139 low-income, primarily
African American women during pregnancy and the first two years of the child's life. Results
indicated that, compared to women who had no home-visiting intervention, women who received
in:--home services experienced fewer injuries and fewer hospitalizations for their children, and also
experienced fewer pregnancies and a lower rate of pregnancy induced hypertension.
Not surprisingly, .research also suggests that the nature of the interventions provided through
in-home programs plays a role in their effectiveness. For instance, some studies on in-home services
report no increased benefit from their use (Barth, 1991; Fraser, Armstrong, Morris, & Dadds, 2000).
However it appears from the literature that in-home programs are most effective when health or
social services professionals, as opposed to paraprofessionals, deliver the services. In-home services
that are long term, generally lasting longer than 6 months are also considered to be more effective
(Barth, 1991; Olds et al., 1997).
Home-Based Prevention Efforts:
Impact on Children and Families of Color
Although differential outcomes for various racial/ethnic groups receiving home-based
prevention services are not often cited in the research, studies do generally support the effectiveness
of these services for vulnerable parenting women. Participation in home-based services has been
linked to a reduced likelihood ofchild welfare system involvement, substance abuse problems,
contact with the criminal justice system, public assistance usage, subsequent pregnancies (Olds et
al., 1997), as well as fewer child injuries and hospitalizations, and better health during pregnancy
(Kitzman et al., 1997).
Additionally, there is evidence to suggest that home-based services are effective with highrisk populations and families ofcolor. For instance Olds et al. (1997) found that home-based
interventions were associated with even better outcomes among unmarried and low SES mothers,
and Fraser et al. (2000) reported that mothers with poor attachment to their child and who had a low
sense of competence were more likely to remain in a home-based program. Kitzman et al. ( 1997)
also noted that in-home services were associated with better outcomes for a sample of primarily
African American mothers. Furthermore, McGuigan, Katzev, and Pratt (2003) reported that older
Hispanic mothers had the greatest likelihood of remaining in a home-based program, suggesting
that in-home services may be particularly useful with families of color.
However research also suggests that the effectiveness of home-based interventions depends
to some degree on the types of risk factors present and the components of the interventions. For
instance, the presence of domestic violence has been found to eliminate the protective effects of
home-based interventions (Ekenrode et al., 2000); and unmarried mothers with less than a high
school education; mothers who received no prenatal care; and mothers who have had multiple births
may be less likely to benefit from home-based programs than mothers without these challenges
(Murphey & Braner, 2000). Furthermore, mothers living in communities with a high rate of
violence have been found to have an increased likelihood of dropping out of home-based programs
(McGuigan et al., 2003); and teenage mothers and those with insecure housing also appear to have
an increased risk of dropping out of home-based programs (Fraser et al., 2000). It also appears
from the literature that home-based programs may be most effective when they last more than 6
months; when they are delivered by health or social service professionals (Barth, 1991; Olds et al.,
23
1997); and when there is a high level ofsupervision for the workers delivering the services
(McGuigan et al., 2003).
In general, home-based prevention efforts appear to be a promising best practice for
vulnerable child welfare populations. More research is needed to discern what the necessary
components of intervention are and to what degree these interventions differentially impact diverse
groups. Additionally, it is important to note that home-based interventions may be time, labor and
resource intensive and as such, child welfare agencies may have difficulty implementing these
services. However, clients of these services clearly appear to benefit and home-based programs do
tend to result in enhanced outcomes for parents and children.
Continuing Care and Ending Choice Points
If child maltreatment is substantiated, children typically enter the CWS under some form of
court supervision. Generally families who are not bypassed for services are assigned to either family
maintenance or family reunification services (see figure 1). Under the Adoption and Safe Families
Act of 1997, families generally receive up to 18 months of family reunification services, at which
point they are either reunified or parental rights are terminated and the child enters post permanency
planning services. As such, continuing care and ending choice point practices generally include
family reunification services, including specific treatments for parents; as well as services for
children in out of home care,· services for alternate caregivers, and permanency planning services
for children who are not reunified with their birth parents.
Factors Associated with Family Reunification
Studies have clearly identified certain factors that are predictive of reunification or nonreunification. Research consistently demonstrates that White children are reunified at a higher rate
than are Black children (Courtney, 1994; Jones, 1998; McMurty & Lie, 1992; Wells & Guo, 1999;
Westat, Inc., 2001c); and young children tend to be reunified more often than older children
(Courtney, 1994; Goerge, 1990; Westat, Inc., 2001c). Harris and Courtney (2003) report that single
parent African American families are particularly unlikely to reunify, while two-parent Hispanic
families have a significantly higher likelihood of reunification than two-parent Caucasian or African
American families. Socioeconomic status also plays a role in reunification; receipt of AFDC and
lower socioeconomic status is associated with a deceased likelihood of reunification (Barth,
Snowden, Broeck, Clancey, Jordan, & Barusch, 1987; Courtney, 1994; Jones, 1998), and Westat,
Inc. (2001c) found that parents with a high school education who were currently employed were
more likely to be reunified than parents who do not have these attributes. Interestingly, Wells and
Guo (2003) report that mothers who move from welfare to work actually reunify with their children
at a slower rate than other groups. Some studies have also found that children with disabilities or
health problems are less likely to be reunified than children without these problems (Courtney,
1994; McMurty & Lie, 1992; Wells & Guo, 1999). Additionally, parents who are non-compliant
with court orders and treatment are significantly more likely to lose custody of their children than
are court compliant parents (Atkinson & Butler, 1996; Jellinek, Murphy, Poitrast, Quinn, Bishop &
Goshko, 1992; Smith, 2003).
Demographic information indicates that most children in family reunification services come
from single parent homes, often headed by the mother (Fein & Staff, 1993; Gillespie, Byrne, &
24
Workman, 1995; Lewandowski & Pierce, 2002; Walton, Fraser, Lewis, Pecora, & Walton, 1993).
Financial problems are common among family reunification clients, (Fein & Staff 1993; Gillespie et
al., 1995; Walton et al., 1993); as are previous out-of-home placements (Lewandowski & Pierce,
2002; Walton et al., 1993). In California, 16. l percent of Black children are in family reunification
services; 25. l percent of White children; 24.8 percent of Hispanic children, 20.3 percent of Asian
children, and 22 percent of Native American children are in family reunification services (Needell
et al., 2003). Children in family reunification services in California also tend to be young; 43.8
percent of children under the age of one are in family reunification services; 32.9 percent between
the ages of one and two; and 27.6 percent between the ages of three and five are in family
reunification services. In Santa Clara County, 15.7 percent ofBlack children are in family
reunification services, 24.3 percent of White children, 25 percent ofHispanic children, 20 percent
of Asian children and 13.3 percent ofNative American children are in family reunification services.
In Santa Clara County, 33.3 percent of children under the age of one are in family reunification
services, 32.8 percent between the ages of one and two; and 30.7 percent between the ages of three
and five are in family reunification services (Needell et al., 2003).
Family Reunification Services
Family reunification services are generally focused on addressing the broad range of issues
that have contributed to the removal of the child, or that prevent the child from returning home
safely. These services are grounded in the notion that the parent child attachment is of primary
importance and the continuity of these attachments is critical to the well-being and development of
the child (Maluccio & Ainsworth, 2003). Family reunification services are generally offered for a
period lasting 6-18 months and they may include a wide range of practices that vary greatly from
county to county and state to state. As is the case with family preservation practices, a large number
of different interventions fall under the definition of "family reunification services;" some family
reunification interventions are modeled after FPS and are intensive, short term and family-centered;
while others focus on the delivery of specific servi_ces designed to ameliorate particular problems.
As with research on the effectiveness of family preservation services, research on the
effectiveness of FRS is lacking; most family reunification studies have small sample sizes, no
comparison groups, no clear definition of the intervention and often fail to measure family and child
well-being in addition to reunification rates (Littell & Schuerman, 1995). Studies evaluating the
impact ofFRS demonstrate mixed results, with many researchers concluding that there is no solid
evidence that FRS are effective in returning children to their birth families (Littell & Schuerman,
1995). Indeed rates of reunification for participants in FRS have been found to range all the way
from 25 percent to 100 percent (Littell & Schuerman, 1995).
For instance, Fein and Staff (1993) evaluated the effectiveness of a fairly intensive family
reunification program in which services were delivered by both a social worker and a family
support worker; workers carried small caseloads and provided a broad range of both concrete and
clinical services, such as parenting training, budgeting assistance, transportation, help with job
training, as well as counseling, or support for substance abuse treatment. Although no comparison
group was used in the study, results indicated that 38 percent of children were reunified with their
families after one year of participation in services-and of these 10 percent were returned to OHC
at the end of the second year.
25
Some researchers have concluded that family reunification services that are intensive in
nature, sometimes following a family preservation model, may be most effective. Walton et al.
( 1993 ), evaluated the effectiveness of family reunification services that were Iimited to 90 days and
were delivered in the home of the birth parents by workers carrying small caseloads. Interventions
averaged 3.1 hours a week and were focused on concrete services, as well as psychosocial training.
Findings indicated that 93 percent of families receiving FRS reunified after the 90-day treatment
period, compared to 28 percent of a co~parison group receiving regular child welfare services. At a
12-month follow-up, the FRS children group (75%) was still more likely to be living in the home,
than the comparison group (49%).
Similarly, Gillespie et al. (1995) also reported promising results from an intensive
reunification program that was modeled after family preservation services. Services were provided
at the birth parents home, for approximately 8 to 10 hours a week and included a combination of
concrete and clinical services. In addition, this program also provided support and training to the
foster parents caring for the children, increased contact between the foster parents and social
workers, as well as the facilitation ofjoint meetings between foster parents and birth parents to
improve consistency of parenting and to share· information. Birth parents were also allowed
increasing numbers of visits based on their participation in services. Although no comparison group
was used, findings indicated that 79 percent of the children were reunified, and at 12-month followup 71 percent of the sample was sill living at home.
Less encouraging results were noted by Lewandowski and Pierce (2002) in their evaluation
of the effects of Missouri's Family Centered Out Of Home Care (FCOHC) Program, a program that
incorporates family-centered principles aimed at involving families in reunification case plans as
soon as the child enters out of home care. Services were intensive, time-limited and strengths-based.
Progress was monitored through team meetings involving the families, social workers, and other
representatives from social service systems. Differences between families receiving the FCOHC
and comparison groups receiving standard child welfare services were significant although not in
the expected direction. The FCOHC group actually experienced fewer reunifications and more
returns to out of home care, than the comparison group. Yet results did indicate that, on average,
children in the FCOHC group who spent more than seven days in out of home care, also spent fewer
days overall in out of home care than the comparison group.
In addition to the impact of family reunification services themselves, the decision-making
practices of workers also play a role in whether or not families reunify. In Westat's (2001a) analysis
of reunification decision-making practices among nine workers in three public child welfare
agencies in Washington DC, workers reported they could generally tell within the first few months
of a reunification case whether reunification would actually occur. Factors identified by workers as
important in their decision to return a child home included, the parents' compliance with services
and visitations, the strong desire of parents to have their children returned, the safety of the child
and the child's wishes and well-being. Workers also indicated that parents who were noncompliant
with services or non-responsive to efforts at reunification and cases involving young children were
most likely to lead to a decision to terminate parental rights. In general workers in this study felt
that time limits for reunification were a positive factor in the decision-making process because time
restrictions allow them to terminate parental rights more quickly in cases where reunification
appears unlikely; however workers also noted that some families with more serious problems could
be reunified if they received services beyond federal time limits. However since the sample size in
this study was small, these results should be interpreted with some caution.
26
Family Reunification Services:
Impact on Children and Families of Color
The impact offamily reunification services for children and families ofcolor is not
adequately addressed in the research literature. Although we know that children and families of
color tend to reunify less frequently than White children, the types ofservices that increase the
likelihood of reunification for children andfamilies of color are not well understood. In general
FRS research does not clearly support the effectiveness of these interventions, either on family
reunification or on family functioning (Littell & Schuerman, 1995; Lewandowski & Pierce, 2002).
Furthermore, the actual interventions that are associated with "family reunification services" are
often not clearly defined.
Research does suggest that families who fail to reunify often have serious social and
economic risk factors that do not appear to be adequately addressed by current family reunification
services. For instance, families with a decreased likelihood of reunification are those that have
income problems (Barth et al., 1987; Courtney, 1994; Jones, 1998); parents with less than a high
school education; unemployed parents (Westat, Inc., 2001c); and children with health problems or
disabilities (Courtney, 1994; McMurty & Lie, 1992; Wells & Guo, 1999). Yet these characteristics
also describe the general child welfare population. Clearly more research is needed to discern the
types of FRS interventions that are effective with these high-risk groups, as well as the possible
differential outcomes for children and families ofcolor.
Specific Family Reunification Services: Parenting Education, Substance Abuse, Domestic
Violence, and Family to Family/Social Network Interventions
In addition to research on the effectiveness of family reunification services as a whole, some
studies have focused on the effects of specific services aimed at improving particular problems such
as parenting deficits, substance abuse, domestic violence and poor social networks. As with family
reunification services, studies on specific treatments to reunify parents with their children are
limited and often have methodological problems such as small sample sizes, lack of comparison
groups, lack of treatment consistency and lack of adequate long-term follow-up.
Parenting Education Services
Programs aimed at teaching parents new parenting skills are frequently included in family
reunification case plans. For instance, Phase 2 case record review data indicated that in Santa Clara
County, 70.2 percent of mothers were recommended parenting education services at the
jurisdictional/dispositional hearing (Hines et al., 2002). In general parenting education services are
didactic in nature and focus on increasing participants' knowledge of effective parenting skills.
Research on the effectiveness of parenting education programs for families in the child welfare
system is lacking. Most parenting education evaluations have examined the impact of these
programs on parents whose children have specific behavioral problems-not for families who have
had children removed as a result of maltreatment. Research on parenting education for nonmaltreating parents has yielded some promising results (Dore & Lee, 1999). However, only a
handful of studies have evaluated parenting education services for child welfare parents. Results
from these studies suggest that child welfare parents may experience increased benefits from
parenting education programs that combine a didactic approach with another parenting education
27
service (Brunk, Henggeler, & Whelan, 1987; Carlo, 1993; Dore & Lee, 1999; Wolfe, Edwards,
Manion, & Koverola, 1988).
More rigorous studies of traditional didactic parenting education services have been
performed with samples involving parents whose children have behavioral problems. Most of these
studies suggest that low risk parents may experience significant short term benefits from parenting
education services (Petsch, Schultz, & Wahler, 1999; Nicholson, Brenner, & Fox, 1999; Patterson,
Chamberlain, & Reid, 1982). However, the research also suggests that high-risk parents typically
experience little benefit from parenting education services (Dore & Lee, 1999).
Parenting Education Services:
Impact on Children and Families of Color
Research on the impact ofparenting education services for children and families ofcolor is
lacking. Studies fail to mention the possible impact ofcultural differences in parenting behaviors on
the success or failure ofparenting education services (Forehand & Kotchick, 1996). Parenting
education appears to be most effective for lower-risk families, however, numerous stressors such as
poverty, substance abuse, domestic violence, low education levels, and community violence often
impact parents involved in the child welfare system. Dore and Lee (1999) note these risk factors
may make cognitive behavioral and didactic approaches to parenting education particularly unsuited
to these populations:
For those who are highly stressed by poverty and its accompanying pressures, even
sustaining a commitment to the training process can be overwhelming. Those parents who
lack the cognitive foundation needed for more didactic and learning based approaches seem
unable to benefit fully from behaviorally based training models (Dore & Lee, 1999, p. 322).
More promising approaches to parenting education for maltreating parents appear to involve a
multifaceted approach that includes not only instructional training in parenting behaviors, but also
experiential training and interventions that address the emotional and psychosocial stressors
affecting these high-risk parents (Carlo, 1993; Brunk et al., 1987; Dore & Lee, 1999; Wolfe et al.,
1988).
Indeed, there is some limited evidence to suggest that parents in the child welfare system
may experience increased benefits from a parenting education program than combines didactic
parenting education with another parenting related service such as behavior training, experiential
training or multistystemic therapy (Brunk et al., 1987; Carlo, 1993; Wolfe et al., 1988). However,
more research is needed in this area. For instance, it is unclear which parenting related services are
most effective when paired with didactic training and it is also unclear how well parents retain
information learned. Long-term follow-up studies that track the retention of information in
parenting education services for child welfare parents, as well as child and family outcomes, are
needed.
Additionally, few studies have addressed the role of culture and ethnicity in the effectiveness
ofparenting education services/or parents in the child welfare system. Forehand and Kotchick
(1996) note, " ... different cultures have different histories that shape cultural theories about
parenting. Without an awareness of the values of a particular culture, attempts to implement
parenting programs with its members may fail" (Forehand & Kotchick, 1996, p.197). As such,
28
research needs to focus on how parenting education outcomes may differ for different cultural and
ethnic groups and what types ofparenting education approaches might be most successful for
diverse families.
Substance Abuse Treatment
Parental substance use is common among the child welfare population. In 1998, the United
States General Accounting Office [U.S. GAO] conducted a study of two state CWS systems in
California and Illinois, and reported that 65 percent of open foster care cases in California and 74
percent in Illinois, involved a family where one or both parents had a substance abuse problem for
which treatment was required. Research indicates that cases involving parental substance abuse
reunify at a significantly lower rate than do cases not involving substance abuse (Smith, 2003), and
in general, children from substance abusing households experience poorer outcomes than children
from non-substance abusing households (U.S. GAO, 1997).
As a result of the widespread substance use among child welfare parents and the poor
outcomes for cases involving parental substance abuse, many family reunification case plans
include requirements for some form of substance abuse treatment. These interventions can include
various components such as drug testing, 12-step groups, outpatient treatment or inpatient
treatment. For instance, Phase 2 case record review results indicated that 40.9 percent of mothers in
Santa Clara County are ordered random drug or alcohol testing, 40.0 percent are ordered substance
abuse or alcohol treatment and 26.8 percent are ordered 12-step groups (Hines et al., 2002).
Unfortunately few studies have evaluated the impact of substance abuse services for child welfare
parents (Gregoire & Schultz, 2001 ). More research efforts have focused on the effectiveness of
these services for pregnant and parenting women who may not necessarily be involved in the CWS.
In one of the few studies that has evaluated substance abuse treatment among child welfare
parents, Gregoire and Schultz (2001) analyzed the treatment and placement outcomes for 167 child
welfare parents in Lancaster County Pennsylvania. Findings revealed that those participants who
had support from significant others were more likely to complete substance abuse treatment and
also experienced better sobriety outcomes. Moreover, parents who were sober at nine months
follow-up were more likely than parents who were not sober to have child custody or to have
maintained parental rights.
Ways in which child welfare workers make decisions about cases involving substanceaffected parents also impacts the likelihood of reunification. Karoll and Poertner (2002) identified
certain factors considered to be associated with readiness for reunification in their qualitative study
of judge's, caseworker's and substance abuse counselor's indicators of reunification. Factors
associated with professional's indication ofreadiness for reunification included: 1) parental
motivation, 2) evidence of substance abuse recovery, 3) evidence of competence and reliability, 4)
the presence of social support, 5) evidence of improved parenting skills and 6) legal aspects of child
loss and prospects for reunification. However there was considerable variability among respondents
in how many clean drug tests are necessary for reunification. Although the median response was 8
months of clean tests, approximately 30 percent felt 6 months was sufficient, while 32 percent felt
that 12 months of clean tests were needed before reunification should occur.
Although research on the effectiveness of substance abuse treatment for child welfare
parents is scant, more research efforts have focused on the effectiveness of these services for
29
pregnant and parenting women who may not necessarily be involved in the CWS. However, since
most child welfare agencies respond to referrals for prenatal drug exposure efforts to offset drug use
among pregnant and parenting women is an important concern for the CWS (Ondersma, Malcoe, &
Simpson, 2001).
Overall, research on the effectiveness of substance abuse services for pregnant and parenting
women suggests these services may be beneficial, however the degree of benefit appears to depend
on the type and intensity of the intervention, as well as the case characteristics. For instance, there is
some evidence that suggests that residential treatment may be more beneficial than outpatient
treatment for women with substance abuse problems. (Haller et al., 1997). Moreover, parenting
women experience better outcomes in substance abuse treatment when they are able to reside in
residential treatment with their children (Clark, 2001; Metsch et al., 2001). Studies also indicate that
women are more likely to complete treatment and experience better outcomes when they have the
support of significant others (Gregoire & Schultz, 2001 ), and when family and significant others are
involved in treatment (Carten, 1996). Parenting women have also identified support and
encouragement from non-judgmental staff as particularly important in their ability to successfully
complete treatment (Akin & Gregoire, 1997; Carten, 1996).
Substance Abuse Treatment:
Impact on Children and Families of Color
Research on the effectiveness ofsubstance abuse treatment for parents color in the child
welfare system is lacking. In general, studies do not describe differential outcomes for various
racial/ethnic groups in these treatments and few studies have actually examined the impact of
substance abuse treatment on reunification outcomes for children and families of color in the child
welfare system. However, research does suggest that parenting women involved in substance abuse
treatment are a high-risk group. Most participants of these services are young single women of
color, with low educational levels and income problems (Carten 1996; Clark, 200 I; Knight et al.,
2001; Metsch et al., 2001 ). In general, research on the effectiveness of substance abuse treatment
services for these women has yielded promising results; however the degree of treatment impact
appears to depend on the type and intensity of the intervention, as well as case characteristics.
Studies suggest that participants in substance abuse treatment services are more likely to maintain
child custody (Gregoire & Schultz, 2001); and maintain sobriety (Clark, 2001; Metsch et al., 2001).
However, women with less than a high school education, criminal justice system involvement and
poor social supports are less likely to complete treatment (Knight et al., 2001 ).
Overall, research suggests that high-risk parenting women benefit from substance abuse
treatment that is comprehensive; allows them to involve their children and other family members or
supports; and is delivered by supportive, nonjudgmental staff. Although research has not
specifically evaluated the role ofethnic diversity in substance abuse treatment services for
parenting women in the child welfare system, the fact that many of the participants of these services
are women ofcolor does suggest that these services may be beneficial for a diverse population.
However, more research is needed to determine the role of culture in treatment, as well as possible
differential outcomes for various ethnic/racial groups.
Domestic Violence Treatment
Child welfare professionals are becoming increasingly aware of the importance of
30
addressing issues of domestic violence for families attempting to reunify with their children. Rates
of domestic violence among families in the CWS vary from 38 percent to 54 percent (Bowen, 2000;
McGuigan & Pratt 2001 ). Phase 2 results from case record review data indicated that in Santa Clara
County, 45.4 percent of cases involve domestic violence (Hines et al., 2002). Research from the
child welfare field on the effectiveness of domestic violence treatments in reunifying families and
improving outcomes for families is extremely lacking. Instead, child welfare research efforts related
to domestic violence have primarily focused on describing barriers to effective collaboration
between domestic violence and child welfare professionals and attempts to overcome these barriers.
Indeed, practioners and researchers have noted that child maltreatment and spousal abuse
service systems have historically operated in isolation from one another and typically have different
service delivery approaches (Beeman, Hagemeister, & Edleson, 1999; McKay, 1994). Recent
efforts to increase collaboration between the child welfare and domestic violence fields has focused
on cross-training between the two professions, as well as simplified processes for referring families
to child welfare or domestic violence services (Findlater & Kelly, 1999; Whitney & Davis, 1999).
Additionally, the creation of specific child welfare policies for families experiencing domestic
violence and the use of multidisciplinary task forces to address ongoing issues related to
collaboration between fields has also been described as helpful in increasing coordination and
collaboration (Findlater & Kelly, 1999; Whitney & Davis, 1999).
·
Santa Clara County has implemented efforts to merge child welfare and domestic violence
fields in order to better serve families affected by both child maltreatment and domestic violence.
Santa Clara County is implementing the Greenbook Project, a multidisciplinary effort involving
representatives from child welfare, the domestic violence community, law enforcement and the
juvenile court. The overall goal of the project is to change the way these groups assist families who
experience both domestic violence and child maltreatment. During the first year of implementation,
the Greenbook Project placed domestic violence advocates from Next Door Solutions to Domestic
Violence at the County Department of Social Services. These advocates provide consultation to
child welfare workers on issues of domestic violence at every level of the child welfare system.
Additionally, the Greenbook Project also visited the Miami-Dade County Dependency Court
Intervention Program for Family Violence to learn key aspects of a domestic violence advocacy
program. Other activities of the Greenbook Project include cross-training between child welfare,
domestic violence, law enforcement and legal personnel, improving cultural competency,
developing guiding principles, changing agency policy and practice, and integrated parenting
programs (Lightbourne, 2002).
The collaboration of child welfare and domestic violence fields to better meet the needs of
families affected by both issues is promising. However, in general these efforts are not focused on
specific treatments for the batterer. Indeed, child welfare research has not addressed the
effectiveness of domestic violence services for child welfare parents. Yet studies from related fields
have evaluated effectiveness of domestic violence treatment for batterers who may not necessarily
be involved in the CWS. Domestic violence treatment for batters typically involves a group
treatment that includes psycho-educational and cognitive behavioral techniques designed to change
faulty beliefs, reduce anger, improve communication and increase non-violent coping skills
(Morrel, Elliot, Murphy, & Taft, 2003). Less common are psychodynamic or support groups.
Findings from studies on the effectiveness of these treatments are mixed and much of the research is
limited due to small sample sizes, lack of comparison groups and findings that rely on self-reports.
Some studies have reported that targeted domestic violence services are effective in reducing
31
violent behaviors (Dutton, 1986; Palmer, Brown, & Barrera, 1992), yet other studies report no
benefits from these treatments (Dunford, 2000). As with many other treatments, domestic violence
treatment tends to be most effective for participants with relatively few other risk factors (Morrel et
al., 2003; Taft, Murphy, Elliot, & Keaser, 2001).
Domestic Violence Treatment: Impact on Children and Families of Color
Research on the impact ofcross-system collaboration and specific domestic violence
treatment for children and families of color is lacking. Although child welfare research describing
cross-system collaboration between domestic violence professionals and child welfare professionals
is promising, these-studies do not address the role ofcultural issues on domestic violence, or the
treatment of batterers and battered women. Similarly, research from related fields describing
outcomes of didactic domestic violence treatment also rarely consider the role of culture or
ethnicity.
Research suggests that the domestic violence field and the child welfare field have
historically had different approaches to treatment that have hindered collaborative efforts (Beeman
et al., 1999). Recently, more concerted efforts to link the two fields have occurred. Typically these
efforts involve the use of multidisciplinary task forces to address issues related to collaboration
(Findlater & Kelly, 1999; Lighbourne, 2002); cross-training (Findlater & Kelly, 1999; Lighbourne,
2002; Whitney & Davis, 1999); simplified processes to refer battered women to child maltreatment
prevention programs and to refer battered child welfare mothers to domestic violence services
(Findlater & Kelly, 1999; Whitney & Davis, 1999); and, the creation ofspecific child welfare
policies stating procedures for families experiencing both child maltreatment and domestic violence
(Findlater & Kelly, 1999; Whitney & Davis, 1999). Although these efforts have great potential to
impact children and families of color in the child welfare system, research has not yet addressed this
topic.
The collaborative efforts between child welfare and domestic violence tend to focus on
services and assistance to women and children, with less ofan emphasis on treatments for the
perpetrators of domestic violence. As with other didactic interventions, such as parenting education,
there is evidence to suggest that domestic violence treatment for batterers is more effective with
low-risk groups than with high-risk groups (Morrel et al., 2003). However, research also suggests
that most men who are participating in domestic violence treatment are a high risk group who often
experience multiple stressors (Brown, Werk, Caplan, & Seraganian, 1999). Moreover, the presence
of certain risk factors appears to reduce the likelihood of treatment completion (Taft et al., 2001).
Additionally, one study found that African American men were most likely to drop out of domestic
violence treatment, even after other factors were statistically controlled, suggesting that these
treatments may not be as effective for African Americans (Taft et al., 2001). However, in general,
more research is needed to discern the impact ofdomestic violence treatments/or various
racial/ethnic groups in the child welfare system.
Social Network Interventions/Family to Family Initiative
Some child welfare services have focused on the role of social network interventions on
improving outcomes for children and families. Social network interventions are designed to enhance
both formal and informal social networks, including increased networking among family members,
32
friends, neighbors, as well as more structured networking involving support groups (Gaudin,
Wodarski, Arkinson, & A very, 1990). These interventions are based on research indicating that
people living in neighborhoods in which there are high child maltreatment rates tend to describe
these communities as lacking in formal and informal support systems (Garbarino & Kostelny,
1992). Additionally, DePanfilis (1996) has noted that in addition to a lack of social support systems,
neglectful families may be particularly distrustful of any support systems that are available.
In an effort to address these issues, social network interventions, such as the Family to
Family Initiative have been implemented. Although the goals of Family to Family are extensive and
go beyond just social network interventions, one of the initiative's main emphases is on
"strengthening the network of families available to care for abused and neglected children in their
own communities" and "building partnerships with at-risk neighborhoods toward that end" (Annie
E. Casey Foundation, 2003b, p. 1). Indeed, many interventions provided through Family to Family
are designed to strengthen neighborhood networks for birth and foster families in the child welfare
system. These strengthened networks are considered to have the potential to assist families in
learning new skills and becoming more self-sufficient. The tactics used to increase social networks
for families, and also increase collaboration between the child welfare system and local
communities are varied. In general there is an emphasis on identifying families' naturally occurring
support networks and involving these networks in services. Interventions can be focused on
identifying and utilizing "natural helpers" in the communities and neighborhoods where birth and
foster families reside. Natural helpers are considered to understand the culture of a particular
community better than professional helpers. Efforts are also aimed at increasing overall
collaboration between child welfare agencies and the communities in which child welfare clients
reside.
Outcome studies on the effectiveness of Family to Family interventions are not available.
However, Gaudin et al. (1990) evaluated the effectiveness of the Social Network Intervention
Project (SNIP) for 52 neglectful families in the child welfare system. The intervention, which lasted
approximately 10 months, included first an assessment of existing formal and informal support
networks in the lives of the families, and then a series of interventions designed to enhance these
networks. Direct interventions were made with informal supports, including family members,
friends or neighbors. Efforts were also made to identify and utilize informal helpers existing in the
families' neighborhoods. Form~l support networks were also encouraged through the use of support
groups, parent aides and social skills training. Post-test results from the evaluation indicated that
compared to a comparison group who received normal child welfare interventions, the families
receiving the social network services fared better on a number of outcomes assessed by their
caseworkers. Intervention parents were rated as having improved parenting skills and attitudes, and
they experienced significant increases in the size of their formal and informal support networks. At
the end of the intervention, 59 percent of the families had their child welfare cases closed, while
23 .5 percent of the comparison group has their cases closed.
Although the use of social network interventions may be a promising child welfare practice,
Gaudin et al. (1990) notes that there are organizational and professional barriers to the
implementation of such interventions. For instance, results from the SNIP evaluation indicated that
it was difficult for social workers implementing the project to maintain a focus on social network
interventions within the context of a public child welfare agency. The predominate model of
services in the agency tended to focus on case management services, and barriers such as high
33
caseloads, heavy paperwork, frequent staff turnover, and lack of collaboration between human
service agencies all contributed to difficulties in implementing the social network intervention.
Family to Family/Social Network Interventions:
Impact on Children and Families of Color
Researchers have not evaluated the impact ofsocial network interventions, such as the
Family to Family Initiative on children and families ofcolor. Given that children and families of
color involved in the child welfare system are likely to reside in impoverished communities, the use
ofsocial network interventions with this population is ofparticular interest. Research exploring the
assets and informal supports imbedded in impoverished communities is limited. It certainly appears
likely that strengthening support networks for vulnerable families would be beneficial, however
research on the mechanisms through which to increase support networks for families is lacking.
Indeed, the exact interventions associated with utilizing informal helpers, strengthening
neighborhood networks and increasing overall collaboration between child welfare agencies and
the communities in which child welfare clients reside are not clearly defined. As Gaudin et al.
(1990) has suggested, there may be significant barriers to effective implementation of these
interventions within the crisis-driven, resource strapped child welfare field. More research is
needed to first discern the types of interventions associated with social network practices, as well as
the ways in which outcomes from these services may differ for various racial/ethnic groups.
Mental Health Services for Children in Out-of-Home Care
Children in the child welfare system are at an increased risk for a variety of psychological
and behavioral problems (Stevenson, 1999). Studies have documented the long-term negative
consequences of both child maltreatment and lengthy stays in out-of-home care (Courtney, Piliavin,
Grogan-Kaylor, & Nesmith, 2001; Stevenson, 1999). In order to ameliorate these negative
consequences, many children in out-of-home care are referred to mental health services. Research
suggests that mental health services are the most commonly recommended treatment for children in
the child welfare system (Kinard, 2002). Studies have also found that children in foster care are
more likely to have a mental health diagnosis and to use mental health services than comparison
groups. For instance, children in foster care are more likely than children receiving public assistance
or children receiving SSI to have a mental health diagnosis (Harman, 2001; dosReis, 2001);
children in foster care tend to have more mental health hospitalizations; and use more mental health
services overall than children receiving public assistance (Reiff, 2001 ).
Yet despite the relatively high rate of mental health service usage among children in the
CWS, the nature and impact of these services are not well understood. Indeed mental health services
can include a broad range of interventions, including support groups, individual counseling, family
counseling, residential services or inpatient services. One model of services, Treatment Foster Care
(TFC), has been the topic of child welfare research. TFC, also referred to as therapeutic foster care
is an intensive, family centered intervention that seeks to place youth with behavioral and emotional
disorders in community and family settings. Children are typically placed with foster parents and
the placement is intended to be both therapeutic and family-centered. Intensive services are
delivered to parents, alternate caregivers and the child. Some studies have suggested that TFC can
be an effective transitional placement for children moving from restrictive settings into less
restrictive placements (Chamberlain, 1994, as cited in Farmer, Wagner, Bums, & Richards 2003;
Clarke, Hawkins, Murphy, & Sheeber, 1993, as cited in Farmer et al., 2003). However, not all
34
studies have found TFC to be effective. For instance, Farmer et al. (2003) used administrative data
to analyze the impact ofTFC on foster youth with emotional or behavioral disorders. Findings
indicated that use of TFC did not result in reduced use of group homes.
In addition to TFC, some child welfare interventions for children focus on a "wraparound"
process that is targeted to high risk children who may have emotional or behavioral problems and
who are currently placed in a level 12 or higher group home, or are at risk for entry into a more
restrictive placement (California Department of Social Services [COSS], 2001). Wraparound
services are described as a "family-centered, strength-based, needs-driven planning process for
creating individualized services and supports for children and their families" (COSS, 2001, p. 1).
Wraparound interventions are collaborative in nature, often involving multiple social service
systems in the planning and delivery of a wide range of services. Few studies have systematically
investigated the efficacy of wraparound services for children and families in the child welfare
system. However, some research suggests that children in wraparound are less likely to be removed
from their community and may experience a reduction in behavioral problems (Bruns, Burchard, &
Yoe, 199S, as cited in Toffalo, 2000; Clark, Lee, Prange, & McDonald, 1996, as cited in Toffalo,
2000; VanDenBerg, 1993, as cited in Toffalo, 2000). Yet other research has reported no significant
benefit from wraparound services (Clarke, Schafer, Burchard, & Welkowitz, 1992, as cited in
Toffalo, 2000).
Mental Health Services:
Impact on Children of Color in Out-of-Home Care
Although research is lacking in identifying effective mental health practices for children in
foster care, studies have demonstrated that children ofcolor are less likely to be recommended for
these services and to use these services. In Santa Clara County, Phase 2 case record review results
indicated that 36.2 percent of children were recommended individual counseling and 19 .6 percent
were recommended other counseling services. Differences between racial/ethnic groups in
recommended counseling services were also found. For instance, 18 percent of Black children were
recommended individual counseling, 46.3 percent of White children, 31.9 percent of Latino
children; and 35.5 percent of Asian American/Pacific Islander children were recommended
individual counseling, suggesting that children of color, and in particular African American children
have a decreased likelihood of being recommended individual counseling (Hines et al., 2002).
Differences in service recommendation and provision were also noted by Garland and
Besinger ( 1997) in their analysis of racial/ethnic differences in court referred pathways to mental
health services for children in foster care. Results suggested a trend for White children to be more
likely than African American or Hispanic children to be court ordered counseling/psychotherapy.
White children were also significantly more likely than African American or Hispanic children to
receive counseling during their first eight months in out-of-home care. And although not significant,
White children had higher rates of service use in general than did African American or Hispanic
children. Racial/ethnic differences in mental health service recommendations and service usage
have been found to remain even after the effects of other demographic factors are controlled
(Garland, Hough, Landsyerk, McCabe, Yeh, Ganger et al., 2000).
Research on the effectiveness of mental health practices for children in out-of-home care is
lacking, and even less is known about the impact of these services for children of color.
Comprehensive treatments for high-risk children, such as treatment foster care and wraparound
35
services have not been adequately researched and have in-general not consistently demonstrated
positive effects. Studies on other mental health interventions for children ofcolor in the child
welfare system are lacking. However, research does suggest that children of color are less likely
than White children to be referred to mental health services (Garland & Besinger, 1997; Hines et al.,
2002). The reduced likelihood for children ofcolor to be referred to mental health services is of
particular concern because children ofcolor experience long stays in the child welfare system and
are at an increased risk for a variety ofpoor outcomes (Courtney et al., 1996). As such, more
efforts should be aimed at ensuring that children ofcolor are referred to and receive appropriate
mental health support, and more research is needed to evaluate the impact ofthese services on
outcomes for children ofcolor.
Foster Parent and Kinship Support Services
Although the majority of research on child welfare practices has focused on interventions
with birth families and children, some research has emphasized the importance of services and
training for alternative caregivers, including foster parents and kin caregivers (Cuddeback & Orme,
2002; Gordon, McKinley, Satterfield, & Curtis, 2003). Some studies have suggested that kin
caregivers may receive fewer services than non- kin caregivers (Gordon et al., 2003; Scannapieco,
Hegar, & McAlpine, 1997, as cited in Cuddeback, 2002). However, other research has reported no
differences between services delivered to kin and non-kin caregivers (Cuddeback & Orme, 2002).
Research suggests that the use of kin placements is common in California and in Santa Clara
County. In 2002, 36.4 percent of children in supervised foster care in California were in a kin
placement and in Santa Clara, 40.1 percent were placed with kin (Needell et al., 2003).
Additionally, children and families of color are especially likely to be placed with kin. For instance,
point in time reports indicate that in California, 40.5 percent of Black children in supervised foster
care are placed with kin; 37.0 percent ofNative American; 35.8 percent of Hispanic, 29.5 percent of
White and 29.4 percent of Asian children are placed with kin. In Santa Clara County, 66.7 percent
ofNative American children are placed with kin; 41.9 percent of Hispanic; 41.4 percent of Black;
35.6 percent of Asian; and 32.5 percent of White children (Needell et al., 2003).
The prevalence of kin placements has prompted some researchers to explore the service and
support needs of kin caregivers. Gordon et al. (2003) noted the unique needs of kin caregivers in
their analysis of focus groups with 3 7 kin caregivers. Findings revealed that unlike non-kin
caregivers, many kin caregivers did not expect to be caring for the children in their care; they had to
adjust their family roles and many described this adjustment as sudden and significant. Kin
caregivers also described having to adjust to the involvement of the child welfare agency in their
lives; many reported feeling unappreciated by the child welfare agency and excluded from the
decision-making process for their kin children. They also described a great need for additional
support services as they took on the unexpected responsibility of caring for their kin children.
Other studies have noted that non-kin caregivers also experience unique challenges. For
instance, although non-kin caregivers are expected to support reunification efforts, many may have
conflictual feelings about supporting contact between the children in their care and the birth
families (Sanchirico & Jalonka, 2000). As a result, some child welfare practices have focused on
providing specialized training and support to non-kin caregivers to assist in the process of
reunification. For instance, Sanchirico and Jalonka (2000) surveyed 650 non-kin caregivers in New
York State and noted that only 19 percent received both specialized training to be a foster parent
36
and support and guidance from the agency. Those who received both training and support also
performed the most activities related to encouraging contact between children and their birth
families. Foster parents who received either training or support but not both also exhibited some
tendency to encourage parent child contact, however those who received both training and support
were significantly more likely to support reunification efforts. These results suggest that specialized
services delivered to foster parents can have a beneficial impact on reunification.
Additionally, other research has focused on evaluating more concerted efforts to involve
foster families in the reunification process. For instance, Landy and Munro ( 1998) evaluated the
impact of the Shared Parenting program. The Shared Parenting program encourages foster families
to work as a team with birth parents in order to improve the parenting skills of the parents. Foster
families were encouraged to provide support and guidance to birth parents through on-going contact
and involvement. An improved relationship between foster and birth families was expected to
increase birth parents motivation and interest in learning new skills, thereby increasing
reunification. Thirteen families participated in the evaluation. Results revealed difficulties in
retaining parents in the program. Over half of the participants did not complete the program, either
because of lack of compliance or because workers discharged families due to changes in status that
decreased the possibility of reunification. Outcomes for parents who remained in the program
revealed that those with higher socioeconomic status, families who moved less, families with no
health problems and families with less risk factors were more likely to reunify.
Although research suggests that kin and non-kin caregivers have different service needs,
studies have demonstrated that overall, both kin and non-kin caregivers have numerous unmet
service needs. For instance, Cuddeback and Orme's (2002) conducted a national evaluation of the
service experiences and needs of kin and non-kin caregivers, and found that both groups reported
numerous unmet service needs. Most alternate caregivers expressed a need for day care, respite
care, liability insurance, heath care, and transportation. Additionally, the majority of caregivers in
both groups reported feeling unprepared to care for their foster children. The authors concluded "the
levels of training and services received were inadequate for both groups of caregivers" (Cuddeback
& Orme, 2002, p. 900).
Foster and Kinship Support Services:
Impact on Children and Families of Color
The impact offoster and kinship support services ofchildren and families ofcolor has not been
adequately researched. However, research clearly indicates that kin and non-kin caregivers are in
need of increased training and supportive services. Kin caregivers may need increased support as
they take on the unexpected responsibility and role of a caregiver (Gordon et al., 2003), while nonkin caregivers may need additional support and training to facilitate the reunification process.
(Sanchirico & Jalonka, 2000). In general, both kin and non-kin caregivers describe the need for
increased services, such as day care, respite care, liability insurance, heath care, transportation,
specialized training and increased participation in the child welfare decision making process
(Cuddeback & Orme, 2002; Gordon et al., 2003; Sanchirico & Jalonka, 2000).
Research on the impact offoster and kinship support services on child welfare outcomes is
lacking. However, there is some evidence to suggest that interventions with foster parents are
associated with better reunification outcomes (Sanchirico & Jalonka, 2000). Yet more research is
clearly needed to discern the impact ofservices to both kin and non-kin caregiver, especially for
37
different racial/ethnic groups.
Permanency Planning Practices
The federal Adoption and Safe Families Act of 1997 shortened reunification time frames
and mandated the use of concurrent planning-a process in which both a reunification plan and an
alternative permanent placement plan are created. For children who are not reunified, the permanent
placement plan with its goal of finding a suitable alternative permanent placement for the child
becomes a major focus of services. As such, permanency planning and placement practices often
·
impact children at the ending stages of the CWS.
Very few studies have investigated the types of practices that are used at the ending stages in
the system and how these practices impact children of color. However, the impact of child welfare
practices on children who are not reunified is particularly important for children and families of
color. The U.S. Department of Health and Human Services (2002) reports that approximately 23
percent of children in foster care have had parental rights terminated and children of color are
disproportionately represented in· this group. Nationally, 43 percent of legally free children are
African American, 34 percent are White, 13 percent are Hispanic, 2 percent are Native American
and 1 percent are Asian Pacific Islander (U.S. DHHS, 2002). In California, 68.2 percent of Black
children in foster care are in permanent placement services, 48.4 percent of White children; 41.5
percent of Hispanic children; 32.5 percent of Asian children, and 53.9 percent of Native American
children are in permanent placement services (Needell et al., 2003). Santa Clara County has similar
racial/ethnic rates of permanent placement clients; 64.4 percent of Black children are in permanent
placement services; 44.7 percent of White children; 41.3 percent ofHispanic children; 30.6 percent
of Asian children and 43.3 percent of Native American children are in permanent placement
services (Needell et al., 2003).
Federal policy mandates that these children have safe and permanent alternative homes,
however, in reality many wait months and years for a permanent placement and some will reside in
long term foster care until they emancipate from the system at age 18. In Kemp and Bodonyi's
(2002) analysis of legally free children in the state of Washington, the median length of stay in outof-home care for these children was 50. 7 months and demographic factors were significantly related
to permanency outcomes. For instance, older children, boys, African Americans, and Asian Pacific
Islanders were all less likely to achieve a permanent placement than others groups. Interestingly,
Latino children were significantly more like than White children to achieve a permanent placement.
The interaction of demographic characteristics was also significant in predicting permanency. Older
African American boys had an even higher chance of not achieving a permanent placement. Other
research corroborates these findings. For instance, African American children have been found to be
less likely to be adopted than White or Latino children and African Americans who are freed for
adoption wait significantly longer periods of time for a permanent placement than do children from
other racial/ethnic groups (Barth, Courtney & Berry, 1994, as cited in Kemp & Bodonyi, 2002;
Kapp, McDonald, & Diamond, 2001).
The federal Multi Ethnic Placement Act (MEPA) of 1994 and the Interethnic Adoption
Provisions (IEP) of 1996 were enacted to increase the number of children who were adopted by
preventing discrimination in the placement of children on the basis of ethnicity and allowing for
increased use oftransracial adoption. The impetus behind MEPA-IEP was a belief that children of
color, especially African Americans were languishing in care, while workers searched for same race
38
foster or adoptive homes. MEPA-IEP also called on child welfare agencies to increase efforts to
recruit foster and adoptive parents who reflect the ethnic and cultural backgrounds of their children
in care. Although MEPA-IEP may have the potential to reduce the time legally free children spend
in substitute care, rates of adoption for children of color in the CWS do not appear to have improved
since its enactment. Indeed, Courtney (1997) suggests that transracial adoption policies and
practices may not actually have a large impact on children of color. Because many legally free
children of color in out-of-home care are older and may have disabilities or health problems, many
potential adoptive parents may choose not to adopt these children. Thus, even though child welfare
agencies may seek transracial adoptions, the actual pool of available adoptive parents may be too
limited to have an impact on children of color waiting for a permanent placement (Courtney, 1997).
In an effort to understand the factors that contribute to adoptive parents willingness to adopt
Black children, Brooks and James (2003) surveyed 541 participants of the California Long-Range
Adoption Study, a longitudinal study of adopters in California. Results indicated significant
differences between participants who were willing to adopt Black children and those who were not.
Of the sample 36 percent were willing to adopt a Black child. In general, parents who were willing
to adopt Black children were also more willing to adopt children of varying ages and special needs
children; they were more likely to be motivated to adopt due to religious or humanitarian reasons
and many eventually did adopt Black children from foster care. Interestingly, approximately 50
percent of parents who indicated they were unwilling to adopt a Black child also reported that
adoption workers or agencies discouraged them from adopting Black children, while only 11
percent of those parents who were willing to adopt Black children were encouraged by their worker
or agency to do so.
Additionally, parents who adopt transracially may experience unique challenges and have
special service needs. In an effort to identify the special needs of transracial adoptive parents Vidal
de Haymes and Simon (2003) conducted interviews with 20 youth and their adoptive parents. Youth
in the sample were either African American (70%), Latino (15%) or biracial (15%), while 95
percent of adoptive parents were White and 5 percent were Latino. Findings revealed that parents
often felt that others misjudged their intentions for adopting a child of color and that their families
"deserved pity" (Vidal de Haynes & Simon, 2003, p. 257). Many families reported that child
welfare workers, especially workers of color did not adequately support their decision to adopt a
child of color. Parents felt that workers viewed their desire to adopt transracially as curious, and
many felt that they received only minimal assistance in their efforts to adopt a child of color.
Parents identified numerous needed pre and post adoptive supports and resources. The most
frequently noted needed services were formal linkages to cultural activities, products, or services
that could educate parents on cultural issues and also provide a link for the child to their own
cultural group. Other needed support services included establishing a network of parents who
adopted transracially, educating workers to support transracial adoption, providing parenting classes
to address issues of cross cultural parenting, and developing a mechanism to help White parents
address their own possible racism, as well as racism from their family and friends.
In an effort to understand how worker perceptions may impact the likelihood of transracial
adoption for children of color, Carter-Black (2002) conducted an ethnographic study often Black
child welfare workers. Although the sample size makes generalizing the results problematic,
findings revealed that workers felt Black children need to develop an ethnic identity and an
understanding of their cultural heritage and that the CWS in general does not meet the needs of
Black clients and poor clients. The recruitment, support and maintenance of Black adoptive and
39
foster parents by the agency were seen as lacking. Many also felt that MEPA-IEP could actually be
harmful to Black children because it may result in agencies being less culturally aware or sensitive
to diverse populations.
Avery (2000) also explored the impact of worker perceptions on permanent placement
practices in her study on the "hardest to place children" in New York State. Eighty case files were
examined and children's workers completed a questionnaire on their perceptions and practices
related to these children. Demographics of the sample are consistent with other studies on children
who have lengthy stays in out-of-home care. Seventy three percent of the sample were male, 60
percent were Black, 30 percent White and 10 percent of another ethnicity, many had substantial
disabilities; approximately 80 percent had siblings in the CWS and they spent approximately 11. 79
years in substitute care. Responses to survey questions revealed that 41.4 percent of workers
believed that children in the sample were ultimately not adoptable. These beliefs appeared to result
in reduced efforts on behalf of the workers to recruit potential adoptive parents. Worker attitudes
toward appropriate placements revealed that 44 percent felt placement with gay or lesbian parents
was not appropriate, and 42 percent felt transracial adoption was not appropriate. When asked to
identify resources that would be helpful in placing the children, workers most often identified
respite services, intensive adoptive parent training, intensive post-placement services, more
effective recruitment techniques and higher adoption subsidies. These findings suggest that workers
negative perceptions of hard to place children appear to result in reduced efforts to recruit adoptive
parents.
Permanency Planning Practices:
Impact on Children and Families of Color
In general, child welfare research has focused relatively little attention on practices at the
later stages of the child welfare system. Although child welfare services have the goal of ensuring a
permanent placement for children, research indicates that many children who have had parental
rights terminated may wait years in out-of-home care before a permanent placement is found (Kemp
& Bodonyi, 2002), and many children remain in out-of-home care until they emancipate at age 18
(Courtney et al., 2001). Yet the impact of permanency planning practices on outcomes for children
of color is of great importance. Research indicates that African American children are
overrepresented among children who have had parental rights terminated and are waiting for a
permanent placement. African American and Asian Pacific Islander children both have a reduced
likelihood ofachieving a permanent placement, and older African American boys have an even
lower likelihood ofa permanent placement (Kemp & Bodonyi 2002).
Some researchers have noted that MEPA-IEP and the issue oftransracial adoption may
actually have little impact on children of color in permanent placement services. Because the pool
of adoptive parents willing to adopt legally free children of color (who are often older, and may
have disabilities or health problems) may be relatively small, policies aimed at transracial adoption
may not have their intended impact (Courtney, 1997). Yet other research suggests that the
perceptions and practices of workers may have an important influence on the likelihood of
achieving a permanent placement. Indeed, negative perceptions of hard to place children may result
in reduced efforts to recruit adoptive parents (A very, 2000). Additionally, parents who have adopted
transracially report that workers often did not adequately support their decision to adopt
transracially (Vidal de Haymes & Simon, 2003), and Black child welfare workers have expressed
some hesitancy regarding MEPA-IEP, believing that it could be harmful to Black children because
40
it may result in agencies being less culturally aware or sensitive to diverse populations (CarterBlack, 2002). However, in general more research is needed to discern the types of permanent
placement practices that impact children and families of color.
Impact of Contextual Factors on Child Welfare Practices
for Children and Families of Color
Contextual factors can impact both the delivery of child welfare practices and the outcomes
of these.practices. Although numerous contextual factors exist, this section will focus on three
areas: 1) cultural competency of child welfare workers and practices, 2) child welfare practice in
rural areas, and 3) the role of the court system. The potential impact of these contextual factors on
practices and outcomes for children and families of color will be discussed.
The Role of Cultural Competency in Child Welfare Practice
Child welfare workers often work with children and families from a wide range of cultures
other than their own. Many practioners and researchers have noted that effective child welfare
practices are those that acknowledge and incorporate the importance of culture in the delivery of
services (Miller & Gaston, 2003). Indeed, Miller and Gaston (2003) note that inherent assumptions
within the child welfare system are grounded in Anglo-Saxon values and cultural norms about child
rearing and family. Child welfare legislation and policies often follow European standards of
culture and White, middle class family values are the standard through which ethnically diverse
parents and children are compared. As such, children and families exhibiting alternative values may
be seen as deviant by the system. These conflicts in attitudes regarding acceptable parenting
behavior may contribute to ineffective or harmful child welfare practices (Miller & Gatson, 2003).
In an effort to combat ethnocentrism in the child welfare system, many agencies have placed
increasing importance on ensuring that workers, programs, policies and practices are "culturally
competent." In general, the term cultural competence refers to an ability to recognize and respect
similarities and differences in beliefs, interpersonal styles, values, norms, and behaviors of various
ethnic and cultural groups (Roberts, 1990, as cited in Schriver, 1998).
Although cultural competency is widely viewed as an important aspect of effective child
welfare practice, research on the cultural competency of child welfare systems and the impact of
this competency on children and families is lacking. Some studies describe the problems diverse
clients experience in trying to access culturally and linguistically competent services, other studies
have described ways to improve cultural competency in child welfare agencies, while other research
has addressed mechanisms for child welfare agencies to work more collaboratively with ethnically
diverse communities.
The issue of access to culturally and linguistically appropriate services for children and
families of color is of tremendous importance. ·Suleiman (2003) notes that for Latino clients in
particular, language barriers can exclude them from necessary services and produce significant
delays in service delivery. Indeed language barriers are implicated as a factor in the underutilization
of services among Hispanic clients (Gomez, Ruiz, & Rumbaut, 1985, as cited in Suleiman,.2003),
and many researchers have discussed the importance of language in accurate assessment and
treatment of clients (Suleiman, 2003). Suleiman (2003) further notes that denial of access to
41
services and programs for people with limited English proficiency is in violation of the Civil Rights
Act of 1964; yet federally funded agencies have wide discretion in the degree to which language
needs of clients are addressed.
As such, the responsibility to provide and monitor culturally competent and linguistically
appropriate services has largely fallen on child welfare systems themselves. Some studies have
proposed various models of achieving cultural competence in child welfare agencies. For instance,
Mederos and Woldeguiorguis (2003) discussed the role of child welfare managers in producing
culturally competent child welfare agencies. Recommendations for increasing cultural competence
included first and foremost, the managers' recognition that disproportionality in services and
outcomes for children of color does exist. Indeed, other researchers have noted that a prerequisite to
true cultural competence within an agency is a commitment to cultural competence principles and
practices among top level staff (McPhatter & Ganaway, 2003). Additionally, to produce agency
wide cultural competence, managers must prioritize staff diversity through recruitment and
maintenance of bilingual and bicultural staff. Child welfare managers also need to reach out to
diverse communities to form collaborations with community representatives. This collaboration can
help the child welfare system better understand cultural differences in parenting behaviors, as well
as assist in the creation of policies and practices that are relevant to those communities.
Indeed, other studies have also documented a need for improved collaboration between child
welfare agencies and culturally diverse communities. For instance, Hosley, Gensheimer, and Yang
(2003), described the implementation process of two programs aimed at improving collaborations
between the child welfare system and two Southeast Asian communities in Minnesota. Both
programs sought to improve practices and outcomes for children and parents from these
communities. The collaborations required committed involvement from numeroll:s agencies, as well
as parents and youth from the communities. Language barriers made the use of bilingual and
bicultural staff at collaborative meetings extremely important. Effective collaboration depended on
a number of different factors, including the need for a trusting relationship between all parties, the
incorporation of different cultural styles in the structure of meetings, decision-making and
communication styles, the need to have bilingual and bicultural staff involved in the process and
that active involvement from the youth and parents is crucial to successful collaboration.
Cultural Competency:
Im pact on Children and Families of Color
Research has not yet investigated the impact of culturally competent child welfare practices
on children and families ofcolor, however there is evidence to suggest that practices that are not
culturally or linguistically appropriate result in underutilization of services and ineffective
treatments and are in violation of clients civil rights (Gomez et al., 1985, as cited in Suleiman,
2003; Suleiman, 2003). Some researchers have described processes for achieving cultural
competence within child welfare agencies. A wide range of activities aimed at increasing cultural
competence have been identified, including support from top level staff, the prioritization of staff
diversity, and forming collaborations between child welfare and ethnically and culturally diverse
communities. These activities are intended to combat the child welfare system's legacy of using
European standards of culture and White, middle class family values in assessing diverse children
and families (Miller & Gatson, 2003).
42
However achieving cultural competency in child welfare can be a difficult task. For instance
it may be challenging for child welfare agencies to recruit and retain bilingual/bicultural staff and
the process of building collaborations with diverse communities may be time consuming and labor
intensive. Yet the impact of cultural competency on children and families of color is of tremendous
importance. More research is needed to discern how child welfare agencies recruit and retain
bilingual/bicultural staff, as well as how collaborations are built between child welfare and
ethnically diverse communities. Most importantly, research is needed to systematically evaluate the
impact ofcultural competence on practices and outcomes for children and families ofcolor.
Rural Child Welfare Practices
Much of the research literature on the effectiveness of child welfare practices is conducted
in urban settings. Researchers have suggested that urban child welfare research is often not relevant
to rural communities who may face uniquely different challenges than their urban counterparts
(Templeman & Mitchell, 2002). Indeed several studies have suggested that rural communities are
particularly vulnerable to a number of different risk factors. Rural communities in comparison to
urban communities have an elevated rate of poverty, there are fewer employment opportunities in
rural communities, and rural communities have been found to have lower levels of social well-being
than urban communities (Menanteau-Horta, & Yigzaw, 2002).
In addition to community differences, the characteristics of urban and rural child welfare
workers also differ. For instance, Landsman (2002) surveyed child welfare workers in 115 rural and
urban communities in Missouri and found several differences between the two groups. Findings
indicated that agency size played an important role in practice, many of the differences between
rural and urban child welfare practice were also noted between smaller and larger agencies. Rural
and small agencies possessed workers with a greater number of years spent in their current position
compared to urban and large agencies, they also experienced greater autonomy, decision-making,
growth opportunities, and agency supportiveness, as well as more job satisfaction and
organizational commitment. In contrast, urban and large agencies had more workers with a social
work degree, more work overload, degree of specialization and adequacy of resources, they also had
higher community stress and availability ofjobs. These findings suggest that child welfare workers
in rural communities are generalist practioners who may be more satisfied with their job than their
urban counterparts, while urban workers tend to have more resources and a higher degree of
specialization.
Other studies have suggested that rural child welfare agencies provide more services and
intervene more often than urban child welfare systems. For instance, research suggests that rural
child welfare agencies have a higher likelihood of offering preventive services to children and
families that do urban agencies (Drake, 1996); they tend to have higher levels of substantiation of
maltreatment (English et al., 2002); and rural child welfare agencies are more likely to intervene in
cases involving drug exposed infants (Ondersma et al., 2001 ).
The differences between urban and rural communities and child welfare agencies have
important implications for child welfare practice. Templeman and Mitchell (2002) have noted that
the application of urban child welfare practices to rural communities is inappropriate. In their
qualitative study of child welfare workers, they report the need for rural specific child welfare
practices. For instance, participants noted that families in rural communities are often
geographically isolated, they often have incomplete knowledge of available resources, lack of
43
access to specialized services, and restricted mobility. Thus, traditional child welfare services that
call on families to travel to and from a variety of specialized interventions are not generally feasible
with this population. However respondents also noted that rural communities possess certain assets
that should be utilized for more effective rural child welfare practice. Most importantly, rural
communities were seen by social workers as having a strong network of informal and at times,
formal supports, including linkages between neighborhoods, and the use of faith based
organizations. More effective child welfare practice for rural communities was described as using
these natural supports to build better collaboration between child welfare agencies and rural
communities. This collaboration was seen as a mechanism through which child welfare could
perform outreach to the community.
Rural Child Welfare Practices:
Impact on Children and Families of Color
The impact ofrural child welfare practices on children and families ofcolor is not
sufficiently addressed in the research literature. In general, rural communities have an increased
rate of poverty, fewer employment opportunities and lower levels of social well-being than do urban
communities (Menanteau-Horta, & Yigzaw, 2002). Differences between child welfare workers in
rural and urban areas have also been found (Landsman, 2002), and there is some evidence to
suggest that rural child welfare agencies intervene at a higher rate than do urban child welfare
agencies (Drake, 1996; English et al., 2002; Ondersma et al., 2001). These differences between
urban and rural communities and child welfare agencies suggest that applying an urban model of
child welfare practice to rural communities is not justified (Templeman & Mitchell, 2002). The
unique needs of these communities and their child welfare populations call for increased research
efforts to identify effective practices for children and families of color in rural areas.
The Role of the Court System in Cbild·Welfare Practice
For children who become dependents of the court, the judicial system may significantly
impact practices and outcomes. In child maltreatment proceedings, judges typically have authority
over decisions about services and certain case outcomes. Indeed, studies have noted that
noncompliance with court orders is a strong predictor of termination of parental rights (Atkinson &
Butler, 1996; Jellinek et al., 1992; Smith, 2003). Other studies report that the courts can have an
important influence over the services to which ch-ildren are referred (Garland & Besinger, 1997).
As such, the ways in which court systems and child welfare systems interact is an important
contextual issue affecting practices and outcomes. In general there is little research on the ways in
which child welfare and court personnel collaborate and how this collaboration, or lack of
collaboration impacts children and families of color. However the few studies that have addressed
this issue report that the two professions experience numerous barriers to effective collaboration.
·For instance, in Russell's (1988) survey of 98 caseworkers and 69 attorneys, significant differences
in role perception were found between the two groups. Child welfare workers and attorneys
disagreed about who should perform a number of different tasks. A majority of both groups felt that
they should have primary responsibility for the same tasks, including requesting authorization for a
dependency hearing, deciding on the allegations, explaining and interpreting court proceedings to
parents, deciding if children should testify, entering into agreements with parents regarding case
disposition, and recommending dispositions to the court. Both groups expressed a need to have
attorney and child welfare worker roles clarified and a need for cross training between the two
44
groups. Additionally, in open-ended responses to survey questions, both attorney and child welfare
respondents expressed some irritation over their interactions with one another. Attorneys felt that
child welfare workers should "stop trying to be attorneys" (Russell, 1988, p. 212), and child welfare
workers expressed that "we know about what lawyers do, but they don't know about what we do.
Therefore training of lawyers to casework responsibilities would be helpful" (Russell, 1988, p. 212).
In an attempt to address barriers to collaboration between child welfare and the court
system, Johnson and Cahan (1992) reported on the "Children Can't Wait Project," a training
program aimed at cross training for attorneys and social workers, and overall system modifications.
The program convened a series of seminars for child welfare workers and attorneys. Seminars
included cross training and open explorations of one another's expectations. Social workers in the
seminars expressed a need for attorneys and court personnel to increase respect and understanding
of their profession, while court personnel described a need for social workers to provide more
timely and accurate case preparation and to be knowledgeable of legal definitions. Other seminar
activities included creating action plans to expedite court proceedings by improving communication
between attorneys and workers, and modifying court proceedings themselves. Although both
attorneys and social workers had good attendance at the initial seminars of the project, attorneys had
much less involvement in multidisciplinary groups working to implement the project. Outcomes of
these collaborative efforts were not described.
Additionally, Knepper and Barton (1997) explored courtroom dynamics of court personnel
and child welfare professionals during court proceedings. District judges, circuit judges, circuit
clerks, Foster Care Review Board chairs, Court Appointed Special Advocates, and attorneys from
five sites in one state completed surveys. Interviews with court personnel and caseworkers were
also conducted, as well as focus groups with foster parents and caseworkers, court observations, and
case record reviews. Findings revealed that in general, caseworkers felt that judges respected their
opinions and judgments about cases. Yet judges also demonstrated great influence on certain
decisions and actions. When caseworkers adhered to group norms, judges often rewarded them by
not ordering "unrealistic" practices (Knepper & Barton, 1997, p. 298), and providing more
reasonable time frames for mental health evaluations to be completed. Many respondents also noted
that court proceedings were often used to finalize agreements, rather than negotiate; they lasted an
average of 10 minutes with prosecutors and other attorneys doing most of the talking. Survey results
indicated that court personnel generally do not receive training on child maltreatment issues prior to
beginning their job. Approximately 77 percent of district judges and 79 percent of circuit judges
reported that there were no training requirements for attorneys working in child welfare.
The Role of the Court System in Child Welfare Practice:
Impact on Children and Families of Color
Few studies have evaluated the potential impact of the court system on children andfami/ies
ofcolor in the child welfare system. However the court system clearly does play a significant role in
case processes and outcomes (Garland & Besinger 1997; Knepper & Barton, 1997). For instance,
Garland and Besinger ( 1997) noted that White children in the child welfare system tended to be
more likely than African American or Hispanic children to be court ordered
counseling/psychotherapy, however no racial/ethnic differences were found in social worker
recommendations in counseling/psychotherapy.
45
Research also indicates that there are numerous barriers to effective collaboration between
court personnel and child welfare professionals, including lack ofclarity in their different roles
(Russell, 1988). Studies suggest that collaboration between these two groups might be improved
through cross training and clarification of roles (Cahan, 1992; Russell, 1988). More research is
needed to discern how the court system impacts child welfare practices as well as outcomes for
children and families of color.
Summary of Key Findings and Implications
This section reviewed research on evidence-based child welfare practices and their
application to children and families of color, as well as the impact of contextual factors on child
welfare practices for children and families of color. Overall, research on the effectiveness ofchild
welfare practices across the various choice points in the CWS is somewhat limited and
unfortunately many studies fail to assess the possible differential impact ofchild welfare practices
on diverse ethnic groups. Nevertheless, certain practices evaluated in the literature are emerging as
promising practices for children and families of color and there is a need to continue to promote
evidence-based practice in child welfare.
Following the pathway that children may take through the child welfare system, research on
child welfare practices at the beginning choice points. including screening, investigation and
provision ofvoluntary services reveals mixed results on the effectiveness of these practices.
Research does indicate that cases involving children of color are investigated and substantiated at a
higher rate than cases involving White children (Ards, et al., 2003; Drake, 1996; Eckenrode et al.,
1988; Wells, et al., 1995). Children of color are also more likely than White children to be placed in
out-of-home care (Needell et al., 2003; U.S. DHHS, 2001).
Studies also demonstrate that there is little overall consistency in the decision-making
practices of individual child welfare workers (Rossi et al., 1999; Schuerman et al., 1999). Practices
designed to improve decision-making including the use of structured risk assessment tools, or
multidisciplinary teams/family group conferencing may have the potential to improve decisionmaking practices, yet research has not yet evaluated the effectiveness of these interventions with
ethnically diverse populations.
Perhaps the most promising beginning choice point practice to emerge from the literature is
the use of home-based prevention services. Research does demonstrate that home-based prevention
efforts are effective in preventing maltreatment and further involvement in the CWS, as well as
improving health related behaviors and outcomes for mothers and young children (Kitzman et al.,
1997; Olds et al., 1997).
Research on continuing care and ending choice point practices provide limited evidence
that these practices are successful in reunifyingfamilies or ensuring a permanent placement/or
children. However, research does consistently demonstrate that Black children are reunified at a
lower rate than are White children (Courtney, 1994; Jones, 1998; McMurty & Lie, 1992; Wells &
Guo, 1999; Westat, Inc., 2001c). Evaluations oftraditionalfamily reunification services, and
specific treatments, such as parenting education, substance abuse. domestic violence or Family to
Family/social network interventions appear to suggest that these services have limited overall
success, and tend to be most successful with lower-risk parents (Dore & Lee, 1999; Morre! et al.,
46
2003). However, research does suggest that substance abuse services for parenting women in
particular, are most effective when they are residential, fairly intensive and allow mothers to live
with their children while in treatment (Clark, 2001; Haller et al., 1997; Metsch et al., 2001 ).
Research on mental health services for children in out-of-home care suggests that children
ofcolor are less likely to be recommended for and to use child welfare services while in out-ofhome care (Garland & Besinger, 1997; Garland et al., 2000). Studies on intensive interventions for
children in out-of-home care, including wraparound services and treatment foster care, have yielded
mixed results (Clarke et al., 1992: Farmer et al., 2003). Additionally, although research does
indicate that foster and kinship alternate caregivers are in need of a variety of services, research on
the effectiveness of these services is not available (Cuddeback & Orme, 2002). Similarly, research
on permanency planning practices is limited. There is evidence that children of color are more
likely than White children to have lengthier stays in the CWS once parental rights have been
terminated (Barth et al., 1994; Kapp et al., 2001). Research also suggests that some child welfare
workers may fail to encourage transracial adoption or placement (Avery, 2000; Brooks & James,
2003).
Additionally, contextual factors such as cultural competency of child welfare workers and
practices; child welfare practice in rural areas; and the role of the court system all impact the
effectiveness of child welfare practices. Cultural competence in child welfare settings, although
widely acknowledged as important, has not been adequately researched. Studies do suggest that the
lack of linguistically appropriate services may result in the under-utilization of services, or
inaccurate or incomplete assessments and interventions (Gomez et al., 1985; Suleiman, 2003).
Moreover, research also suggests a need to differentiate between rural and urban areas when
researching and discussing child welfare practices (Templeman & Mitchell, 2002). Rural areas in
comparison to urban areas have been found to have unique risk factors and child welfare agencies
in rural and urban areas have been found to differ in a number ofways (Drake, 1996; English et
al., 2002; Menanteau-Horta & Yigzaw, 2002; Ondersma et al., 2001). As such, there is a need to
define and evaluate rural-specific child welfare practices (Templeman & Mitchell, 2002).
Although more research is needed to discern how the court system affects child welfare
practice, research does suggest thaf there are numerous barriers to effective collaboration between
court and child welfare systems. Cross-training and clarification of roles have been identified as
important factors in improving the relationship between the child welfare and court systems (Ca~an,
1992; Russell, 1988).
In general, much of the research literature on the impact of child welfare practices is
inconclusive. Yet very few studies describe differences in outcomes by ethnic/racial groups and
most of the research fails to even mention the effect of culture and ethnicity on the delivery and
impact of child welfare practices. This gap in the research literature is striking and has important
implications for children and families of color in the CWS. The lack of adequate knowledge of the
types of practices, services, programs, policies, tools, or strategies that are effective for children and
families of color hinders our ability to remedy the disproportionate representation of children of
color in the CWS and may leave many children and families of color without adequate or
appropriate services. The next section focuses on child welfare practices in Santa Clara County's
child welfare system.
47
IV. SANTA CLARA COUNTY CHILD WELFARE PRACTICES
In Phase 3, one area of investigation focused on the impact of child welfare practices on
children and families of color in the Santa Clara County's Department of Family and Children
Services (DFCS). Several related analyses were conducted to understand the types of practices and
case characteristics that are associated with more successful outcomes for children and families
from various racial/ethnic groups; the types of child welfare practices used in Santa Clara County;
how practices in the Main County Offices differ from South County Office practices; how practices
interact with child, family and cultural characteristics of differing racial/ethnic groups to produce
various outcomes; and the role of the court system on child welfare practices. As such, this section
contains five overall analyses:
I) A successful case analysis. This analysis examined case record review data collected in
Phase 2 to determine the types of practices and case characteristics associated with
successful case outcomes.
2) A qualitative analysis of Santa Clara County child welfare practices. Focus group and
interview data from Santa Clara County workers were collected to obtain information
about current practices, how these specific practices might positively or negatively
impact children and families of color, and what practices might be created, enhanced or
maintained to better serve children and families of color.
3) A comparison of Main County Offices versus South County Offices. This analysis used
both quantitative data from CWS/CMS, as well as qualitative data from focus groups in
South County to compare case characteristics and practices between the two areas of the
county.
4) A multivariate analysis of factors associated with reunification. This analysis
examined system-related factors and case characteristics predictive of reunification of
children with their families.
5) An analysis of the role of the court system in child welfare practice. This analysis
used case record review data collected in Phase 3 to determine whether and what type of
court-ordered changes are made to the social worker case plan recommendations at the
jurisdictional/dispositional hearing. Results from agency-wide focus groups related to
the relationship between the child welfare and court systems are also provided.
Taken together, these five sections provide extensive and important information on Santa Clara
County child welfare practices and factors that are associated with successful outcomes for children
and families of color.
48
Successful Cases Analysis
In Phase 3, one aim was to define and assess successful cases and outcomes in Santa Clara
County's Child Welfare System. An analysis of factors associated with successful case outcomes
and how these factors and outcomes differ by racial/ethnic group helps shed light on the types of
practices and conditi~ns necessary to promote enhanced outcomes for children and families of
color.
Methods
First, our task was to operationally define "successful cases and outcomes" for this
quantitative portion, which we based on the United States Department of Health and Human
Services (US DHHS) Children's Bureau standardized outcomes. These standardized outcomes are
part of the U.S. DHHS' child and family services reviews and the Title IV-E foster care eligibility
reviews (U.S. DHHS, 2003b). These levels of success and outcomes are:
Safety
• Children are first and foremost protected from abuse and neglect
• Children are safely maintained in their homes whenever possible
Permanency
• Children have permanency and stability in their living situations
• The continuity of family relationships and connections is preserved for families
Family and Child Well-Being
• Families have enhanced capacity to provide for their children's needs
• Children receive appropriate services to meet their educational needs
• Children receive adequate services to meet their physical and mental health needs
Second, from our in-depth reviews of the 403 cases sampled for Phase 2, we matched basic
information and additional data collected from the court reports during Phase 3 with the US DHHS
guidelines.· Although information from every court hearing available within our sample was
recorded, we relied primarily on data from the jurisdictional/dispositional hearing and the last two
hearings of each case to evaluate "successful cases and outcomes." We then expanded our analyses
started in Phase 2 to evaluate "successful cases and outcomes" and differences by ethnicity for this
current phase.
Results
Regarding safety, referring to the US DHHS criteria mentioned above, we used three
variables to indicate "children are first and foremost protected from abuse and neglect." Number of
prior referrals and previous number of times in the CWS indicated a risk of harm affecting these
children. The number of times removed from the family indicated a family environment that was
assessed as harmful and thus warranted additional removal.
49
We also used the initial case assignment from the jurisdictional/dispositional hearing to
denote whether "children are safely maintained in their homes whenever possible." A similar
variable, case closure type, which also refers to maintenance of the child in the home, will be
discussed in the permanency section.
Number of Prior Referrals
Of the 403 cases, 48 had missing data. Of the remaining 355 cases, 118 (33.2%) had no
prior referrals, 111 (31.2%) had 1 or 2 prior referrals, and the remaining cases (35.6%) had more
than 2 prior referrals. This indicated that more than half of the cases had prior referrals, which are
used by social workers in establishing potential risk to child safety. One-third of the cases had more
than two prior referrals. Please see Table la.
. Re erralsa
Table 1a: Frequency of Pnor
Number of Prior
Frequency
Referrals
0
118
I
75
2
36
3
35
4
24
5
18
14
6
7
6
8
7
9
4
3
10
11
5
12
3
5
13
14
1
15
0
16
0
17
0
18
1
a Based on 355 cases with valid information
Percentage
33.2%
21.1%
10.1%
9.9%
6.8%
5.1%
3.9%
1.7%
2.0%
1.1%
.8%
1.4%
.8%
1.4%
0.3%
0.0%
0.0%
0.0%
0.3%
There was not a statistically significant relationship between number of prior referrals and
ethnicity. An average of2.46 prior referrals (sd = 3.14) were made before the case was opened for
the current episode. Please see Table 1b.
50
T able lb N umberof Pnor
. Re£erra ls B~Y Ethme
. Groupa
Total
African
White
American
Prior Referrals
Avg.
2.46
2.14
2.64
(sd)
(3.14)
(2.71)
(3.27)
Latino
Asian
American/PI
Other
2.46
(3.25)
1.62
(2.14)
2.81
(3.37)
a Based on 355 cases with valid information
Previous Times in the CWS
Of the 403 cases, 93 had missing data. Of the remaining 310 cases, 204 (65.8%) had no
prior experience with the CWS while 106 (34.2%) did have prior experience with the CWS. This
indicates that one-third of cases had an elevated risk for abuse or neglect as evidenced by prior
substantiated abuse or neglect. Please see Table 2a.
Table 2a: Previous Times in the CWS 8
Previous Times
Frequency
In the CWS
0
204
85
1
2
10
3
6
4
4
5
0
1
6
a
Based on 310 cases with valid information
Percentage
65.8%
27.4%
3.2%
1.9%
1.3%
0.0%
0.3%
There was not a significant relationship between the number of previous times in the CWS
and ethnicity. On average, children across ethnic groups were previously in the CWS 0.47 times (sd
= 0.83). Please see Table 2b.
Table 2b: Previous Number of Times in CWS by Ethnicity a
White
Latino
Total
African
American
Average
(sd)
a
0.47
(0.83)
0.25
(0.49)
0.50
(0.92)
0.46
(0.78)
Asian
American/Pl
Other
0.63
(1.06)
0.56
(0.82)
Based on 310 cases with valid information
Number of Times Removed from Family in Current Episode
Ofthe 403 cases, 15 had missing data. Of the remaining 388 cases, 257 (66.2%) were
removed once and 131 (33.8%) were removed more than once, indicating that in one-third of the
cases, the safety level in the home regarding the protection from abuse and neglect was not
consistent. Please see Table 3a.
51
·1 .m Current ,p1so
Table 3a: T"1mes Removed fr om Fam11y
Number of
Frequency
Percentage
Times Removed
1
257
66.2%
2
98
25.3%
3
24
6.2%
4
7
1.8%
5
2
0.5%
a Based on 388 cases with valid infonnation
There was a significant relationship between the number of times removed from family in
current episode and ethnicity. Across all ethnic groups, children were removed an average of 1.45
times (sd = 0.74). African American children had the highest average number of times removed
from family in current episode with 1.67 times (sd = 0.90), and Asian American/Pacific Islander
children had the lowest average with 1.26 times (sd = 0.51). Please see Table 3b.
Table 3b: Number of Times Removed from Family in Current Episode bv Ethnicitva
White
Total
African
Latino
Asian
American
American/PI
Other
Average
(sd)
1.64
(0.92)
1.45
(0.74)
1.67
(0.90)
1.36
(0.68)
1.44
(0.68)
1.26
(0.51)
a Based on 388 cases with valid infonnation
Case Assignment After Jurisdictional/Dispositional Hearing
Of the 403 cases, 12 had missing data. Of the remaining 391 valid cases, 88 (22.5%) were
assigned to family maintenance. This indicates that less than one-quarter of cases are safely
maintained in the home whenever possible. There was not a significant relationship between case
assignment after the jurisdictional/dispositional hearing and ethnicity. Overall, most cases were
assigned to family reunification (71.1 %, 278 of 391 cases). Please see Table 4.
• a
Table 4: Case Assijl;nment aft er Jurisd1ctiona 1spos1t1ona Hearmg
White
Latino
Total
African
American
Family
34
27
88
7
(22.5%)
Maintenance
(14.9%)
(25.8%)
(19.4%)
Asian
American/PI
9
(29.0%)
Other
11
(26.2%)
Family
Reunification
278
(71.1 %)
38
(80.9%)
91
(68.9%)
101
(72.7%)
19
(61.3%)
29
(69.0%)
Other
25
(6.4%)
2
(4.3%)
7
(5.3%)
11
3
(9.7%)
2
(4.8%)
a Based on 391 cases with valid infonnation
(7.9%)
52
Permanency
Regarding permanency, referring to the US DHHS criteria mentioned above, we used three
·variables to indicate whether "children have permanency and stability in their living situations."
The number of unique placement homes and the number of placements in the current episode
indicated the consistency of the children's living situation while in the CWS. Since permanency and
stability also imply a solution of consistency within particular settings, case closure type was also
examined. To evaluate the outcome of whether "the continuity of family relationships and
connections is preserved for families" we also examined the proportion of reunified families within
the case closure type.
Number of Unique Placement Homes in Current Episode
Of the 403 cases, 15 had missing data. Of the remaining 388 valid cases, 162 (41.8%) had
only one or two unique placement homes. 224 (58.2%) had more than two unique homes, which
indicated that less than half of the children in the CWS are in a relatively permanent living situation.
Please see Table 5a.
T a bl e 5a: N um berofU ntQUe
.
Placement H omesa
Percentage
Number of
Frequency
Placement Homes
55
14.2%
1
2
27.6%
107
16.0%
62
3
11.9%
46
4
7.0%
5
27
6.2%
24
6
2.1%
8
7
2.8%
8
11
1.8%
9
7
2.3%
10
9
0.5%
11
2
1.0%
12
4
2.1%
13
8
0.8%
14
3
0.5%
15
2
0.0%
16
0
0.8%
17
3
0.3%
18
I
0.3%
1
19
0.3%
20
1
0.0%
21
0
22
0.3%
1
0.3%
23
1
0.0%
24
0
0.3%
25
1
53
26
27
28
29
30
31
32
33
34
35
0
2
0
0
0
0
0
1
0
1
a Based on 388 cases with valid information
0.0%
0.5%
0.0%
0.0%
0.0%
0.0%
0.0%
0.3%
0.0%
0.3%
There was not a significant relationship between number of unique placement homes in
current episode and ethnicity. Across all ethnic groups, there was an average of 4.54 (sd = 4.67)
unique placement homes in the child's current episode. Please see Table 5b.
. a
Table5 b: Number of Umque Placement Homes biy Ethn'ICity'
White
Total
African
Latino
American
Average
(sd)
4.54
(4.67)
5.35
(6.48)
4.70
(4.60)
4.44
(4.26)
Asian
American/Pl
Other
3.29
(3.22)
4.36
(4.59)
a Based on 388 cases with valid information
Number of Placements in Current Episode
Of the 403 cases, 16 had missing data. Of the remaining 387 valid cases, 222 (57.4%) had
only one or two placements. However, 165 cases had more than two placements indicating that
42.6% of children in CWS are in a relatively inconsistent living situation. Please see Table 6a.
.
Placement Homesa
Table 6a: Num bero f Umque
Percentage
Frequency
Number of
Placements
20.4%
1
79
37.0%
143
2
14.5%
56
3
10.9%
42
4
5.2%
20
5
2.8%
11
6
2.1%
7
8
0.8%
8
3
2.1%
8
9
1.0%
4
10
0.3%
11
1
0.0%
12
0
54
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
a
4
2
1
0
0
1
0
1
1
0
1
0
0
0
0
0
0
0
0
0
1
Based on 387 cases with valid information
1.0%
0.5%
0.3%
0.0%
0.0%
0.3%
0.0%
0.3%
0.3%
0.0%
0.3%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.3%
There was not a significant relationship between the number of placements in current
episode and ethnicity. Across ethnic groups the average number of placements in current episode
was 3.31 (sd = 3.34). Please see Table 6b.
Table 6b: Number of Placements in Current Episod e By Eth nicitya
White
Latino
African
Total
American
Average
(sd)
a Based
3.31
(3.34)
3.27
(3.23)
3.77
(4.06)
2.99
(2.78)
Asian
American/PI
Other
3.00
(3.16)
3.20
(2.83)
on 387 cases with valid information
Case Closure Type
Of the 403 cases, 386 had valid information for the case closure type. Including those
adopted, reunified with family, or placed with a relative or guardian, 183 (47.4%) of the cases have
permanency and stability in their living situation. There was not a significant relationship between
case closure type and ethnicity. Please see Table 7.
55
.. a
T a ble 7 Case Closure Type By Ethmc1ty'
Total
African
American
Family
Stabilized
(FM)
White
Latino
Asian
American/PI
Other
98
(25.4%)
13
(27.1%)
40
(30.5%)
31
(23.5%)
5
(16.1%)
9
(20.5%)
44
(11.4%)
5
(10.4%)
(9.9%)
13
21
(15.9%)
2
(6.5%)
3
(6.8%)
Reunified
with parent or
guardian,
court or noncourt specified
25
(6.5%)
3
(6.3%)
11
(8.4%)
6
(4.5%)
0
(0.0%)
5
(11.4%)
Adoption
16
(4.1%)
4
(8.3%)
6
(4.6%)
2
(1.5%)
2
(6.5%)
2
(4.5%)
Emancipation
57
(14.8%)
6
(12.5%)
11
(8.4%)
25
(18.9%)
7
(22.6%)
8
(18.2%)
27
(7.0%)
6
(12.5%)
7
(5.3%)
10
(7.6%)
(3.2%)
I
3
(6.8%)
119
(30.8%)
11
(22.9%)
43
(32.8%)
37
(28.0%)
14
(45.2%)
14
(31.8%)
Guardianship
established or
placement
with relative
Incarceration,
runaway or
medical
services
Court ordered
termination
and other
a Based
on 386 cases with valid information
Reunification
Of the 403 cases, 386 had valid information to determine whether children were reunified
with their family at case closure. Of these 386 cases, 263 (68.1 %) were not reunified with family
56
and 123 (31.9%) were. This indicates the continuity of family relationships and connections are
preserved for less than one-third of these families. Please see Table 8a.
. h F amllY
·1 a
T able 8a: Propo rf10nof Ch"ld
1 ren R eum"fl1ed wit
Frequency
Percentage
Not Reunified
with Family
263
68.1%
Reunified with
Family
123
31.9%
a Based on 386 cases with valid information
There was not a significant relationship between reunification with family and ethnicity.
Please see Table 8b. However, it is important to note that in the multivariate analyses that appears
later in this report (please see page 103), significant variations in reunification by ethnicity were
found even after controlling for other demographic and system-related factors.
• •
a
T a ble 8b Ch"ld
1 ren Reum"fi1ed By Ethmc1ty·
Total
African
American
White
Latino
Asian
American/Pl
Other
Not Reunified
with Family
263
(68.1%)
32
(66.7%)
80
(61.1 %)
95
(72.0%)
26
(83.9%)
30
(68.2%)
Reunified
with Family
123
(31.9%)
16
(33.3%)
51
(38.9%)
37
(28.0%)
5
(16.1%)
14
(31.8%)
a Based on
386 cases with valid information
Family and Child Well-Being
Regardingfamily and child well-being, referring to the US DHHS criteria mentioned above,
we created a summary score based on 16 variables categorized in 4 domains ( overall well-being,
economic status, health status, and substance abuse related to the mother and the father) evaluated
from the last two hearings to indicate if "families have enhanced capacity to provide for their
children's needs." Although we originally wanted to examine each domain separately, the
extremely sparse information available from the last two hearings of each case prohibited these
comparisons at such a detailed level.
We then evaluated whether the child's education status was positive from the last two
hearings to indicate if "children receive appropriate services to meet their educational needs.
57
Finally, we created a summary score to evaluate the status of the child. This was based on 8
variables within similar domains above (child's overall well-being, economic status, health status,
and substance abuse) to indicate whether "children receive adequate services to meet their physical
and mental health needs."
Families Enhanced Capacity to Provide For Their Children's Needs
Of the 403 cases, 156 had valid information to determine if families had an enhanced
capability to provide for their children's needs. A positive score on this quantified information
meant that this capacity was generally enhanced, and a negative score meant the capacity was
generally diminished. The possible range for this summary score was -2.0 to +1.0. The average
score was 0.01 (sd = 0.71), which indicated that on the whole, case outcomes seemed neutral in
terms of family's capacities to provide for their children's needs. There was a marginally significant
relationship with ethnicity. African American families appear to have a more enhanced capacity
while Asian American/Pacific Islander families seem to have a more diminished capacity. Please
see Table 9.
Table 9: Family's Capacity to Provide for Their Children's Needs By Ethnicitya
Total
African
White
Latino
Asian
American
American/Pl
Other·
Average
(sd)
0.07
(0.62)
0.01
(0.71)
0.35
(0.59)
0.00
(0.77)
-0.04
(0.69)
-0.50
(0.53)
a Based on 156 cases with valid information
Children Receiving Appropriate Educational Services
Of the 403 cases, 76 had valid information to determine if children received appropriate
services to meet their educational needs. A positive score meant that their educational needs were
generally met, and a negative score meant that their educational needs were generally unmet. The
possible range for this summary score was -1.0 to +1.0. The average score was 0.30 (sd = 0.71)
which indicated that educational needs were being somewhat met. There was not a significant
relationship with ethnicity. Please see Table 10.
Table 10: Children Receivin~ Appropriate Services to Meet Educational Needs By Ethnicitya
White
Asian
Other
Total
African
Latino
American/PI
American
Average
(sd)
a Based
0.30
(0.92)
0.75
(0.71)
on 76 cases with valid information
0.32
(0.89)
0.03
(0.98)
0.50
(0.93)
1.00
(0.00)
58
Children Receiving Adequate Services to Meet Physical and Mental Health Needs
Of the 403 cases, 156 had valid information to determine if children received adequate
services to meet their physical and mental health needs. A positive score in the summary variable
meant that their needs were generally met and a negative score that their needs were generally
unmet. The possible range for this score was -1.0 to +1.0. The average score was 0.11 (sd = 0.71)
indicating that on the whole, the case outcomes seemed neutral in regards to meeting children's
needs. There was not a significant relationship with ethnicity. Please see Table 11.
s
T a ble 11 Ch"ld
1 ren R ece1vmii Ad equate erv1ces t o Meet Needsa
White
Total
African
Latino
American
Average
(sd)
0.11
(0.71)
-0.05
(0.76)
0.21
(0.69)
0.07
(0.71)
Asian
American/PI
Other
0.13
(0.64)
0.07
(0.83)
a Based on 156 cases with valid information
Summary of Key Findings and Implications
In Phase 2, we analyzed major case characteristics on the main sample of 1720 cases
(primarily from CWS/CMS data) and additional psychosocial factors measuring child and family
well-being for the 403 cases relying on data gathered directly from the case files. We conducted
ethnic comparisons for both samples. In Phase 3, our aim was to evaluate successful cases and
outcomes using the US DHHS guidelines and assess ethnic differences. Results indicate that there
were minimal ethnic differences
Variations of Safety, Permanency, and
Family and Child Well-Being by Ethnicity
In regards to safety, there was a significant relationship between the number oftimes
removed from family in current episode and ethnicity where African American children had the
highest average number of times removed from family in current episode and Asian
American/Pacific Islander children had the lowest average.
In terms offamily and child well-being, there was a marginally significant relationship
between a families capacity to provide for their children's needs and ethnicity, with African
American families appearing to have a more enhanced capacity while Asian American/Pacific
Islander families seem to have a more diminished capacity. However, from this portion of our
analysis, the general absence of significant differences by ethnic group assessed from the end of the
case implies that certain aspects leading to a successful case and outcome are not related to
ethnicity.
It is important to note that major ethnic differences do exist in Santa Clara County's CWS,
but many of these variations appear at different choice points in the system (such as psychosocial
status at the beginning of the case, and assignment to voluntary family maintenance). In addition, a
combination of factors experienced through the course of the case may be related to success or
59
positive outcomes. However, our results suggest that ethnic differences at the conclusion of the
case are not evident, if we define success according to DHHS guidelines.
Implications Related to the Disproportionate Representation of
Children of Color in the CWS
Given the relatively few ethnic differences in evaluating successful cases and outcomes,
what does this imply about ethnicity and the CWS? The absence of ethnic differences in this
segment does not negate the previous findings in Phase 1 and 2 confirming that certain ethnic
groups are over-represented (i.e., Latinos, African Americans) while others are under-represented
(i.e., Asian American/Pacific Islanders) in the county's CWS. Phase 2 results also indicate that
there are significant ethnic differences in many demographic, system-related, and psychosocial
characteristics. Yet, what does the absence of ethnic differences in successful cases and outcomes
mean in context? While results suggest there is a clear disproportion of ethnic groups in the system,
once in the system, the children and families are generally faring the same. Their relatively
homogeneous experience may be a result ofthe "one-size-fits all" limited array of services offered
Although this may seem equitable and logistically efficient, it appears ultimately ineffective in
serving a diverse group of children and families. Also, to determine better how "well" these
children are actually doing while in the system, the availability of comparative data, clear
operational definitions for "success," and more reliable and valid data are needed.
Limitations and Recommendations
The major limitations for this successful cases and outcomes component include the lack of
information available (especially at case closure), to establish an operational definition of a
"successful case and outcome," and the need for baseline and comparative data.
Data Issues
To assess better whether safety, permanency, and family and child well-being are
maintained and enhanced, child welfare records should include more complete and measurable
information. We acknowledge that child welfare case records were not originally intended for
research. Yet, if the agencies responsible for providing effective services wish to evaluate the
performance of their programs more reliably and validly, improving information collection and
tracking should be considered.
One step would be to create a case summary check-out form used at case closure or in an
exit-interview type situation. The check-out form would indicate the child and family's status in
major areas of psychosocial well-being (i.e., health, education, and finances), in addition to the
existing case outcomes. Given the amount of missing data assessing these domains describing case
and child status, and the relative scarcity of information recorded in the last two hearings
(presumably the place where outcomes of a case would be recorded) this form would enable
measurement and evaluation of "successful cases and outcomes" in a more consistent objective
manner. The amount of missing data and the manner in which information is assessed also implies
the need for more reliable and valid methods of recording of information.
Operational Definition of Successful Case and Outcome
60
Having a clear operational definition of a "successful case and outcome" is vital to program
evaluation and research. When we used the US DHHS criteria, the initial impression maybe that a
relatively low proportion of"successful cases and outcomes" seems to exist in regards to safety,
permanency, and well-being as for example only 32% of the children are reunified with their
families. However, again availability of comparative data along with the consistent use of a welldefined operational definition of "success" is needed to understand the actual success rates for
children and youth in the CWS; Two questions arise from our results and interpretation. First, what
actually is "success?" For example, adoption for one case may be a failure if a family has improved
its ability to support and nurture a child, but was denied reunification because the system still
deemed those parents unfit. Adoption for another case may be a success if the system permanently
removes a child from a persistently dangerous or unhealthy family environment. The same
difficulty in evaluation occurs when a family is reunified (presumably a success) but the child later
falls victim to reabuse (failure). Finding an agreed-upon operational definition may be first a
theoretical or philosophical question rather than a research one, but is important for the County to
consider if their programs are to determine their effectiveness
The second question concerns a judgment of Santa Clara County's CWS performance
according to the proportion of "successful cases and outcomes." Given the above limitations and
issues, Phase 3 cannot give a final answer to how well the county's CWS is serving children and
families. This brings us to our third limitation and set of recommendations, the need for baseline
and comparative data.
Baseline and Comparative Data
It was difficult to assess the successful cases and outcomes in Santa Clara County's CWS.
Are the proportions of children reunified with their families or of those ado'pted high or low? And
compared to what? Implementing better record keeping and data collection over time on key
characteristics, based on an agreed-upon operational definition would better answer the question of
effectiveness and performance. Implementing these practices across counties and across states
would also facilitate evaluation.
The recent Little Hoover Commission report (2003) urges reform in the CWS, especially for
the county and state to improve the leadership and management structure guiding the system (Little
Hoover Commission, 2003). As part of this reform, an accurate evaluation of success for children
and families involved in the CWS is needed.
61
Qualitative Analysis of Santa Clara County Child Welfare Practices
The qualitative component of this study consisted of both focus groups and interviews and
was designed to obtain insights about current child welfare practices used within the DFCS, how
these specific practices might positively or negatively impact children and families of color, and
what practices might be created, enhanced or maintained to better serve children and families of
color.
Methods
For the purposes of both the focus groups and the interviews, "practices" were defined
broadly to include programs, services, strategies, policies, or tools. Qualitative methods, including
focus groups and interviews, are ideally suited for capturing and documenting common experiences
about what "works" and what might be improved from the perspective of individuals involved in
different parts of the child welfare system, including Social Workers, supervisors and clients. These
methods may also help to illuminate quantitative findings, offer information about formal and
informal local practices, and provide a context for understanding a variety of factors that might
influence practices. The methods employed for gathering qualitative data through focus groups and
through interviews are described below.
A total of 16 structured focus group interviews were conducted during the course of this
study. Thirteen of the fucus groups were composed of Social Workers, specifically workers with
titles of Social Worker II or III. The focus groups were structured to explore practices at different
choice points in the child welfare system. Specific topics for focus groups included the following:
• Emergency Response/Dependent Intake (early choice points related to entry into the system and
initial assessment and investigation).
• Diversion (provision of voluntary family maintenance, informal supervision, early intervention,
and other efforts to divert families to services without court intervention).
• Case Assignment (early choice points related to cases that are substantiated and opened).
• Mandated Services (including court ordered parenting classes, substance abuse treatment,
domestic violence services, mental health counseling or other services).
• Time Limits (including local, state and federal time limits and the impact on practices with
children and families of color).
In order to ensure that concerns of clients were adequately represented in the project, three
focus groups were conducted with parents in the child welfare system. Individual interviews were
also conducted with supervisors with experience in each of the choice points described above.
Individual interviews were used to supplement and affirm findings from the focus groups.
Social Worker Focus Group Sampling and Recruitment
Prospective participants for the Social Worker focus groups were randomly selected from a
list of workers with title of Social Worker II or Social Worker III in different units/division. Lists of
randomly selected Social Workers were generated using SPSS (Statistical Package for Social
Sciences). This random selection was conducted "without replacement," so that if Social Workers
were randomly selected for a specific focus group topic, they would be excluded from random
samples for other focus groups. Table 12 below provides a list of the 13 Social Worker focus
62
groups, the child welfare units from which focus group participants were selected, and the number
of focus group participants.
Approximately 15 to 20 Social Workers were randomly selected to receive an invitation to
attend each of the 13 Social Worker focus groups. In recognition of possible scheduling conflicts,
emergencies, or vacations of invitees, we invited more Social Workers than were expected to attend
each focus group. Prospective focus group participants were sent an initial invitation by email two
weeks to one month in advance of the focus group. Non-responding Social Workers were sent a
second invitation to participate in the focus group. Finally, prospective focus group participants who
agreed to attend a focus group were sent a confirmation email and/or phone call two to three days
prior to the scheduled group. Each focus group was comprised of approximately 5 - 8 members,
though some were smaller. In particular, the two case assignment focus groups were small, with two
and three participants in each of the groups respectively. A total of 67 Social Workers participated
in the focus group interviews.
Table 12 Emp oyee Focus Group Compos1tion
Focus Group
Units Included in Selection Process
Diversion
ER/DI
Case Assignment
Mandated
Services
Time Limits
Court Services
• Early Intervention/Intensive Family Services/
Home Supervision
• Informal Supervision
Family Resource Centers
• Asian Pacific Family Resource Center
• Nuestra Casa Family Resource Center
• Uiirani Family Resource Center
Emergency Response
• Screening
• Emergency Response
Court Services
• Dependent Intake
Emergency Response
• Emergency Response
Court Services
• Dependent Intake
Continuing and Child Welfare Services Bureau
• Court Supervision
• Domestic Violence Unit - Court Supervision
• Continuing
• Spanish Speaking Services
• Continuing Multilingual
Adoptions Family Assessment and Development
• Adoptions and Family Drug Court
Continuing and Child Welfare Services Bureau
• Court Supervision
• Domestic Violence Unit - Court Supervision
• Continuing
• Spanish Speaking Services
• Continuing Multilingual
Adoptions Family Assessment and Development
• Pre-Adoption and Drug Court
• Family Assessment and Development
Number of
Groups
Number of
Particioants
2
14
2
5
2
10
2
11
2
14
63
South County
TOTALS
•
Familv Linkage
South County Region
• Gilroy Family Center
• Rosanna Child Welfare Unit
• Monterey Road Child Welfare Unit
• Highland Child Welfare
3
13
13
67
Social Worker Focus Group Procedures
The Social Worker focus groups were de.signed to examine practices that impact children
and families of color at different choice points or units within the child welfare system. To this end,
the focus group questions explored the context in which services are provided, specific practices
that might positively or negatively impact children and families of color, and recommendations
about practices that might be maintained, enhanced or initiated to better serve children and families
of color (please see Attachment 1 for the Social Worker Focus Group Interview Guide). Each of the
focus groups took place in a room located at a DFCS office. Ten focus groups were conducted at the
main administrative office in downtown San Jose and three were conducted at Gilroy Family
Resource Center in South County. The focus groups were approximately 1½ hours in length.
All focus group participants received and signed a consent form that explained the study and
confirmed their voluntary participation in the focus. group. All focus group participants also
completed a brief demographic questionnaire. The summary of focus group participant
demographics is provided in Attachment 2.
Family Focus Group Sampling and Recruitment
Three focus groups were conducted with families involved in the child welfare system. Two
focus groups were conducted with parents of children from populations that are over-represented in
the child welfare system in Santa Clara County, specifically Latino and African American children
and families. Each of these focus groups was organized in partnership with one of the family
resource centers and non-random, purposive sampling methods were employed. Participants for the
Latino parents' focus group were recruited among recent graduates of a monolingual Spanish
speaking parenting class at the Nuestra Casa Resource Center. Four parents agreed to participate in
the focus group and two were able to attend. Staff ofUjirani Family Resource Center recruited
participants for the African American parents focus group. Five parents and four staff attended the
focus group. Comments from parents and staff were documented separately. Finally, a focus group
of foster parents was convened to explore the needs of children of color in out of home placement.
The six focus group participants in the foster family focus group were all women; 2 participants
identified as Latina, 1 identified as Latina/Native American, 2 identified as Caucasian/European
American, and 1 identified as African American.
Family Focus Group Procedures
The three parent focus groups were approximately I hour in length. Questions explored in
these group interviews were focused on what has worked well for clients in getting services from
Child Welfare; what has not worked well and what services should be kept, improved or created to
better meet the needs of families. Please see Attachment 3 for the Family Focus Group Interview
64
Guide. Participating parents signed a consent form and each parent received a $10 gift certificate
from Target as a token of appreciation for contributing to the study. The family focus groups are
described in Table 13 below.
Ta ble 13 Famity
·1 FOCUS Group Compos1t10n
Focus Group
Sponsoring Organization
Latino/ Mono-Lingual
Spanish Speaking Parents
African American Parents
Foster Families
TOTALS
Nuestra Casa
Ujirani Family Resource
Foster Home Licensing,
Recruitment & Intake - Resource
Families Support Team
Number
of
Groups
1
Number of
Participants
1
1
5
6
3
13
2
Individual Interviews Sampling and Recruitment
Fourteen in-depth interviews, of approximately one hour in length, were conducted with
supervisors in different positions in the Department of Family and Children's Services. Prospective
supervisor interviewees were randomly selected using the same sampling procedures described
earlier. In several cases, because the total number of supervisors was small, all supervisors in some
practice areas (such as Emergency Response or Dependent Intake) were included on the list of
prospective invitees. Based on the assumption that some supervisors would decline or be unable to
complete interviews, more participants were invited than target number of interviews ( 15). A total
of23 supervisors were invited to participate in interviews and 14 completed interviews.
Individual Interview Procedures
The interviews used a critical incident approach that is designed to elicit descriptions of
specific behaviors that help formulate the critical requirements of a specific activity (Flanigan,
1954), such as effective child welfare practice with children and families of color. The critical
incident technique has been used to research decision-making in child welfare (Banach, 1998) and
is a valuable tool for obtaining information about specific behaviors and practices from
practitioners, who are in the best position to evaluate which of these practices "work" or "don't
work." Rather than gathering general opinions and hunches, this method allows interviewees to
recall specific experiences related to the provision of services and to delineate factors related to
success or failure in outcomes, such as outcomes related to child welfare intervention for children
and families of color.
Interviewees responded to three (3) open ended questions and several "probe" questions that
solicited detail about specific critical incidents in practice with children and families of color. In
general, the questions explored: 1) factors and practices associated with a case perceived by the
interviewee to represent a successful outcome with an African American, Latino or Asian
American/Pacific Islander family, 2) factors and practices associated with a case considered
unsuccessful, and 3) specific practices that result in or could potentially result in enhanced
65
outcomes for children and families of color (please see Attachment 4 for the Supervisor Individual
Interview Guide). Demographic information was collected at the close of the interview.
The supervisors interviewed for this study represented different practice areas ofDFCS
including Emergency Response (5), Court Services (3), Continuing and Child Welfare Services
Bureau (2), Adoptions Family Assessment and Development (1 ), and South County (3).
Interviewees all had extensive experience working in the Department of Social Services and
possessed an average of 17.5 years working with County DFCS (range of 5.5 years to 33 years). All
fourteen interviewees possessed an MSW degree and some also were LCSW's. Half of the
interviewees were men and half were women. Four of the interviewees were Latino/a, five were
Caucasian/Euro American, three were African American, and two were Asian American/Pacific
Islander. Ten of the key informants are fluent in speaking another language, and seven use these
language skills in their work. A summary of demographic information for the 14 fnterviewees is
provided in Table 14.
T a ble 14 Superv1sor Interv1ewee D emograp. 1cs
Bureau/Unit
Gender
n(¾)
Emergency Response= 5 (35.7%)
Court Services
=3 (21.4%)
Continuing
= 2 (14.3%)
Adoptions
= I (7.1%)
South County
= 3 (21.4%)
n(¾)
Males= 7 (50.%)
Females= 7 (50.%)
Ethnicity
n(¾)
= 3 (21.4%)
African American
Asian/Pacific Islander= 2 (14.3%)
Latino/Hispanic
=4 (28.6%)
White
= 5 (35.7%)
YearsCurrent
Position
mean
(std. dev.)
3.7
(2.4)
YearsCounty
mean
(std. dev.)
17.5
(8.9)
Qualitative Analysis Procedures
Two members of the Child Welfare Research Team, a trained focus group leader and a
research assistant, were present during each of the focus groups with the exception of three of the
smaller focus groups that were facilitated and documented by one focus group leader. Notes were
taken during both focus group and individual supervisor interviews. Audio tape recordings were
made during social worker focus groups and individual supervisor interviews to ensure accuracy of
the transcribed notes. Data for the qualitative portion of this study consist of the transcribed notes
taken during each of the focus group and individual interviews. Qualitative methods of data analysis
were employed to analyze the focus group and individual supervisor interview data. First, major
themes were summarized for each of the choice points in the social worker focus group and
supervisor interview data. Second, themes from the family focus groups were analyzed separately.
Third, a separate analysis of themes in the qualitative data from South County Offices was
conducted to provide insights and supplement quantitative analysis of possible differences between
South County Offices and Main County Offices within DFCS. Finally, common themes across all
focus groups and interviews were identified and documented. The findings of the analysis of themes
from focus groups and interviews are presented below and are followed by specific
recommendations that emerged across focus groups and interviews.
.66
Qualitative Results: Themes from Focus Groups and Interviews
Participants in focus groups and interviews were asked to describe practices that may
negatively or positively impact children and families of color in the Santa Clara County's CWS, as
well as contextual factors that may affect the delivery of these services or client outcomes. This
section describes the main themes that emerged from these focus groups and interviews. Following
the pathway that children may take through the CWS, both negative and positive practices are
described at beginning choice points as well as continuing care/ending choice points. Lastly,
important contextual factors will also be discussed in relation to the delivery and impact of these
practices. 1
Beginning Choice Points:
Practices that Negatively Impact Children and Families of Color
The main themes related to practices that negatively impact children and families of color at
beginning choice points included, I) inappropriate referrals to the CWS, 2) inconsistency in
decision-making practices, and 3) inadequate prevention of child maltreatment or entry into the
cws.
Large numbers of inappropriate referrals were noted as having a negative impact on children
and families of color. Study participants reported that, "Many mandated reporters have no clear
definition or evidence of abuse," and that low-income children and families are frequently referred
into the system because of issues more closely associated with poverty than abuse or neglect. Focus
group participants stressed the high number of referrals from schools "based on their observations
of children wearing old or dirty clothes," "because children had lice," or in one case "because a
child brought an expired can of food to a food drive, the child was thought to therefore be
neglected." Some participants commented that low-income children from wealthier school districts
were disproportionately reported to child protective services. "A child who is at a marginal
economic level will be seen differently at a school in Los Gatos as opposed to a school anywhere
else."
Although study participants stressed that inappropriate referrals are generally screened out
by the Screening Unit, repeated referrals can be damaging because they create a perception of
increased risk that may or may not be accurate. As the number of inappropriate referrals increases
for a particular family the likelihood ofDFCS opening a case may also increase. This situation was
described as having a potentially negative impact on children and families of color in particular
because they are often more likely to have socioeconomic problems. Some respondents also
suggested that children of color are more likely to be scrutinized by law enforcement. "Police are
more harsh on African American, API and Latino Families, especially if they don't speak English,"
commented another participant.
These inappropriate referrals were described as negatively impacting screening practices
because social workers often have to spend a great deal of time screening out inappropriate
1
Please note that focus group and interview participants discussed contextual factors related to the court system. These
findings are presented in the section entitled "The Role of the Court System in Child Welfare Practice" which begins on
page 109.
67
referrals. This situation then often reduces the amount of time social workers have to identify and
serve families who are in need of voluntary services.
Additionally, study participants felt that, in some cases, inconsistency in decision-making
practices during beginning choice points negatively impact children and families of color. Although
decision-making practices occur at all stages of the CWS, many participants discussed these
practices in relation to clients' entry into the CWS. Although staff generally noted the utility of
structured risk assessment tools used at the entry point into the CWS, they also stressed that these
tools are limited and that social workers need to develop recommendations using both assessment
tools and their own judgment. In relation to the latter, participants expressed concern that decisionmaking is often determined by the biases and perceptions of individual social workers. Social
workers named specific risk assessment tools that are used at beginning stages of the CWS as
useful, such as the decision tree use4 during screening and a written protocol used in Early
Intervention. At the same time, social workers suggested that application of assessment tools might
vary among social workers based on a number of factors including individual worker bias, size of
the unit and unit level, shared expectations about assessment, and level of supervisor support and
oversight.
Moreover, study participants noted that, "Decision-making all depends on the worker and
his or her attitudes and values," and "We can so easily sway the picture so it is positive or negative
by how we write up our narratives for the court reports; the report has a lot of influence and
sometimes peoples' personal issues get in the way of making informed decisions." Furthermore,
interviewees noted that social workers are not immune to stereotypes of different groups in the
dominant culture and in the media and that social workers often judge families based on their own
pers_onal experiences and perceptions about what is "good enough" for children. Some· social
workers pointed out that, for example, workers from other cultures might misinterpret
communication norms in African American families as abusive. "Sometimes African American
families have norms about speaking loudly, yelling, and threatening (like I am going to pop
you) ... but kids are rarely hurt ... an African American social worker might be less likely to
[negatively] interpret these kind of verbalizations than a social worker from a different
background."
According to study participants, decision-making practices are also influenced by
socioeconomic factors. Families experiencing unemployment, lack of access to resources such as
daycare, or problems obtaining health care for children may be initially assessed as having a higher
risk for maltreatment. For example, "older siblings are taking care of younger siblings because
parents can't afford childcare," noted one participant. "If there were more good daycare there
wouldn't be as many neglect cases, even once in the system we [workers] can't get them [clients]
child care," added another. Social workers also underscored the impact of local high housing costs
on low-income families. This problem was encapsulated by the comments of one worker: "Many
families partner up with other individuals to maintain housing and just because there are numerous
people living in the home, relatives sharing beds, doesn't mean that this is abuse or neglect. It's
poverty." The case example below typified the illustrations provided by study participants:
A Social Worker 1 was covering a case and she was concerned about the family because
there were four children living in a basement with their family. This was a Chicano family.
She reported that the children might need to be removed. I went out to this social worker
and saw completely different things. The basement was an apartment. The place was small,
68
but had couches and a crib for the baby; it was cluttered but not dirty ... Maybe another
worker might have justified what this worker saw and removed the kids for perhaps no good
reason at all.
The issue of social worker attitudes toward clients was particularly salient to family focus
group participants. Several parents commented on the notable variation in how they are treated by
different social workers. A remark that typified this theme was made by one parent who said, "some
social workers do their job because they really want to help other and some just do it for the money
and don't seem to care about others." Parents also described feeling undermined by social workers
who appeared to have a negative attitude about the client. One client summarized, "some of the
social workers are disrespectful, unfriendly, they look down on the clients and make you feel
humiliated." They elaborated, "It gets to the point where it makes the client feel like 'I don't want to
go in there,' but I have to go in there for my child and when you come out you're in tears."
Inadequate prevention of child maltreatment or entry into the CWS was also identified as
having a negative impact on children and families of color at beginning choice points. Participants
commented on the need for services in the community that would help to prevent entry into the
child welfare system. Focus group members stated that there is a need for education about child
welfare laws and definitions of child abuse, particularly for cultures in which norms related to
discipline of children differ from expectations in the dominant culture. Participants suggested that
such programs could be offered through Social Service agencies but also in community based social
institutions, such as churches, and that some of the communities that have benefited from this type
of preventive education in the past include African American, Asian American/Pacific Islander
(e.g., Vietnamese, Chinese, and Filipino), Latino, and immigrant communities.
Study participants also emphasized the problems in obtaining services without opening a
case. For example, one worker stated that "to get into a wrap-around service from an outside
agency, such as schools or probation, without the child being a dependent takes months, by this
point the problem will have escalated." Many social workers stated that opening a case had both
benefits and drawbacks: families had greater access to services and support through resources such
as Family Preservation Funds but were at greater risk for having a child removed because of time
limits or establishing a history with DFCS.
Several social workers also noted that it is actually more work for them to maintain a child
in the home with Voluntary Family Maintenance (VFM) services than it is to remove a child from
the home and refer to court intervention. Additionally, for families who do make it into VFM or
Informal Supervision (IS) services, focus group participants pointed out that resources were not
available to provide follow-up support to families to prevent re-entry into the system once their
cases are closed. "The family may still have needs but we have to close it. We can only keep it in IS
for 6 months (or 3 months in VFM); it relates to funding because you lose federal and state money
when a case is closed and you are on County money-and there is none," exemplified comments
from interviews. Consequently, "clients may end up back in services." Ideally "workers would be
able to provide some additional help to them [families] to help themselves."
Time lags between case assignment to Voluntary Family Maintenance and social workers
actually receiving the case were also noted as an issue. "By the time the social workers gets the
case, three weeks have gone by and the social worker has only 2 ½ months left [out of three months
of services under Voluntary Family Maintenance]," elaborated one social worker.
69
Beginning Choice Points:
Practices that Positively Impact Children and Families of Color
The main themes related to practices that positively impact children and families of color at
beginning choice points included, I) orientation of clients to the CWS, 2) team decision-making, 3)
family involvement in decision-making, and 4) the use of child care and in-home services.
Providing an orientation for parents to help them understand the components of the child
welfare systems, the court process, current laws and policies related to time limits, and how to work
with social workers was named as a best practice for families first entering the CWS. Social
workers in South County portrayed this as a best practice that was formalized into an orientation
group while social workers in other areas of the county described conducting orientations with
clients individually. Social workers and supervisors advocated for strengthening and
institutionalizing the practice of providing an orientation for clients throughout the county. Some
participants suggested that an orientation workshop could be provided or co-lead by someone
"outside the system" that might be perceived by clients as more neutral.
It would be helpful for parents to go to a workshop to educate them about time lines.
Parents don't get it immediately. Sometimes by the time the client "gets it, " it may too late.
Also sometimes social workers don 't explain it well. Sometimes clients are just traumatized
about what all this means and you have to walk them through the process for the first 2
months and you have 6 months time limit - it's very difficult. The workshop should also
teach the clients how to read a court report and what the recommendations mean.
Team decision-making <TDM), which is currently being initiated in Santa Clara County,
was viewed by participants as a promising tool for mediating individual bias and cultural
misunderstanding by social workers. "Even with the decision tree in place we know it hasn't been
working because there are still so many children of color and there is still a lot of the worker's
subjective perspective put into the report; hopefully TDM can reduce this." TDM was also
perceived as a vehicle to "help get everyone on the same page" with professionals in other systems,
such as schools and service providers. Staffing to discuss cases within units and with
interdisciplinary teams was identified as a positive practice for children and families of color that
could be enhanced by TDM. Some social workers expressed concern about the amount of time that
might be required by TDM's and that "it may be hard to get everyone together in a limited time."
Study participants identified family involvement in decision-making. including the use of
family conferences as a positive practice, particularly for children and families of color. Participants
suggested this was particularly important for cultural groups that place high value on extended
family. "In the Latino culture it's appropriate to extend an invitation to grandparents, godparents,
and all parental figures," illustrated one social worker. Study participants stressed the importance of
extending family involvement to include both biological and non-biological family members. For
example, some participants noted that African American families often have non-biological aunts,
uncles, or other extended family members that should be included in family conferencing. Many
social workers and supervisors viewed the developing family-to-family initiative as a promising
strategy for expanding and institutionalizing family participation in decision-making.
70
Last, the use of childcare and in-home services were also noted as a positive practice. For
instance, in one social worker's opinion, families with adequate childcare have about an 80 percent
participation rate in diversion services, compared to 50 percent participation rate for parents without
adequate childcare. Social workers, supervisors, and families discussed the importance of childcare,
in relation to the ability of families to succeed in services. "It's really helpful to have child care
through Calworks; We need to make sure that is not cut because people can't work or train if there
is no one to watch their kids -This is a huge thing."
Social workers also noted that the provision of in-home services allows for "working with
clients in their own environment" and providing opportunities for one-on-one mentoring and role
modeling new behavior. Social workers identified agencies and programs that provide in-home
parenting classes as more effective than traditional didactic intervention practices alone. Social
workers also stated that providing in-home visits and follow up services, with the support of Social
Worker l's, was an effective practice that is underemployed because of heavy caseloads and scarcity
of resources.
Beginning Choice Points:
Recommendations to Improve Practices
Recommendations to improve practices at beginning choice points in the CWS included 1)
reducing bias and increasing accountability in decision-making, 2) training professionals in other
systems, 3) providing client orientations to families first entering the CWS, and 4) investing in
prevention and early intervention. Specifically,
Reduce Bias and Increase Accountability in Decision Making
•
•
•
•
Implement team decision-making in a manner that includes consideration of client culture.
Examine ways to facilitate accountability for team decisions.
Explore mechanisms to ensure accountability in decision-making and minimize the possible
influence of individual or supervisor bias in decision-making.
Continue the practice of facilitating cultural consultations between staff, with the support of
supervisors, including consultations across units.
Continue and expand practices related to family conferencing.
Training Professionals in Other Systems
•
•
•
Provide or collaborate in providing training for mandated reporters, including schoolteachers
and health care professionals.
Provide or collaborate in cross training with other systems and agencies "so we can understand
one another's needs and views-what we need from each other." Cross training may also serve
"to increase communication" or even "help with preventing some families from coming into the
system, including families that are already being seen by the agencies."
Provide or facilitate more training for police regarding removal, working with social workers to
assess risk before removal, coordinating with social workers regarding child and family
interviews, and ensuring sufficient training for law enforcement officers assigned to conducting
sexual abuse interviews.
71
Provide Client Orientation
•
•
•
Educate parents referred into the system about child abuse and neglect laws. Provide workshops
in their own language on topics such as court processes, how to work with social workers,
timelines, and how to navigate the system.
Consider having such workshops lead or co-lead by someone outside of the system that may be
perceived by clients as more neutral.
Hire and Train Social Worker 1's to assist with family education and orientation.
Invest in Prevention and Early Intervention
•
•
•
Explore mechanisms to continue providing prevention services, including education of diverse
communities about child abuse and elder abuse.
Explore mechanisms to fund counseling and intervention services for families at risk that are
less restrictive than federally funded programs such as Family Preservation, which frequently
require formally opening a case to allow families to access resources.
Continue support and explore opportunities to expand voluntary and diversion services,
including early intervention and voluntary services at the family resource centers should
continue. These efforts are critical to "determine if families should be in the system" and ensure
opportunities for less restrictive placements for children.
Continuing Care/Ending Choice Points:
Practices that Negatively Impact Children and Families of Color
Study participants noted three overall practices that negatively impact children and families
of color at continuing care/ending choice points, including 1) problems with service plan
recommendations, 2) inconsistency in assessment of substance abusing families and need for
stronger linkages with substance abuse services, 3) barriers and service gaps related to relative and
foster placements.
Problems with service plan recommendations, including a "cookie cutter approach" to
services, and unrealistic expectations of families were noted as practices that have a negative impact
on children and families of color. For instance, several study participants suggested that using a
"cookie cutter approach" or "template for recommendations" is a practice among both social
workers and district attorneys that negatively impacts children and families of color. Participants
offered a number of explanations for this dynamic including "social workers being jaded, believing
that since they had worked with a similar family before, that the family is the same and the
interventions should be the same." Similar observations from social workers in other focus groups
suggested that "sometimes workers will get challenged if they don't take the cookie-cutter
approach," and sometimes "DA's think all parents need counseling, parenting class, etc." Several
social workers called for greater flexibility in working with families and "more creativity regarding
case plans needs. For example, some parents may just need more time to spend with their children."
Others cited examples where clients were mandated to services that were not appropriate including
a mother mandated to services before an assessment was conducted and a father "who is in a 52week batters' program-who had one incident of losing his temper and hitting his child-and
72
though hitting her was not okay he would be better served in family therapy and an advanced
parenting class."
Additionally, some service plan recommendations were also described as placing unrealistic
expectations on families that can set them up for eventual failure. These practices tend to impact
low-income families and were perceived to disproportionately impact children and families of color.
First, service plans frequently require "a laundry list of mandated services" that are unrealistic for
families. "I have a lot of clients who when I sit down with them and go over the service plan with
them for the first time they say 'I'm overwhelmed, I can't do this." "Families are in a panic to get
everything done on the service plan from ER to continuation," explained one social worker focus
group participant. Other participants noted that a family might be mandated to substance abuse
services, parenting classes,.and mental health services while seeking employment and new housing.
Some families "just give up," observed participants. Moreover, clients are often required to attend
services that conflict with their job schedules, particularly clients whose schedules change from
week to week. "The expectation for clients to reschedule their work schedule is difficult, especially
those that receive hourly wages, and causes even more stress for the family."
Many participants expressed concerns that high expectations imposed on families are
reflective of and "ideal" rather than "good enough" standard for families. One study participant
summarized this dynamic: "We are not required to have 'perfect parents,' we need minimal
standards of parenting .. .Just because you or I would not parent that way does not mean there is a
risk to the child." "From a client's side, they are being told by a court system and a social system
that the way that they were brought up and the way that things were done in their [clients'] family is
wrong and not acceptable," noted a social worker focus group participant.
Finally, study participants cited insufficient resources to support the success of families
attending mandated services as a problem. For example, participants in several focus groups stated
that the shortage of bus passes and increased paperwork required to obtain the limited passes that
were available negatively impact families, particularly children and families of color. "Most of
these clients don't have a vehicle so for the low income families it is challenging for them to get
their children back ... how are they supposed to fulfill their required programs without
transportation?"
Inconsistency in assessment of substance abusing families and need for stronger linkages
with substance abuse services was also noted as negatively impacting children and families of color
at the continuing care/ending choice point. Participants stressed the importance of effectively
addressing the needs of families with substance abuse problems who they perceived to comprise a
majority of caseloads. Study participants described a number of success stories in which families
were reunified when parents completed substance abuse treatment and also highlighted the fact that
when appropriate substance abuse services are provided they can be an effective best practice.
At the same time, study participants suggested that protocols for decision-making with
substance abusing families and formal agreements with substance abuse treatment agencies were
inadequate. Participants also espoused different standards about when it might be appropriate to
reunify a child with a parent or place a child with a relative in recovery. Furthermore, participants
suggested that there was a need for developing protocols for addressing relapse with substance
abusers in conjunction with both treatment programs and the courts. "We [social workers] pay lip
service, we say relapse is a part of recovery, but the system does not reflect that." Finally, several
73
study participants noted a lack of clarity and consistency in standards for assessing risk in the case
of drug exposed infants, suggesting that race and class bias may influence both initial referral and
assessment of parents of drug exposed infants in child welfare.
Study participants also commented on the scarcity of substance abuse services, particularly
residential programs for women and children. Although a variety of substance abuse treatment
services are offered throughout Santa Clara County, participants pointed out that the need fo·r
services was greater than availability, particularly for clients whose primary language is not
English. Several participants, particularly in South County, noted that a centralized intake process
for entree into treatment was a barrier for many families because of time and transportation issues.
Additionally, social workers felt that barriers and service gaps related to relative and foster
placements also negatively impacted children and families of color at continuing care/ending choice
point practices. While the use of relative placements was described as an effective practice for
children and families of color, study respondents proposed that poverty issu~s and timelines
converged with other factors to create substantial obstacles to relative placement. First, pressure to
move children out of the children's shelter, in the context of little staff support to locate prospective
relative placements, often results in lost opportunities to place a child with a family member. Some
social workers described feeling "in a double bind" and, in some cases, deeply frustrated because of
having to place a child in a foster care situation that was perceived as less positive for the child.
Second, even if a relative placement is identified, it is often "impossible to complete the relative
placement paperwork within 24 hours, with the required background checks and fingerprinting."
Third, although social workers can seek a Director's Exemption to place children with families
where members may have had a criminal background, many social workers expressed concern
about the fact that even a distant criminal history can create barriers to appropriate relative
placements. Finally, new standards for relative placement ''are the same as a licensed foster home
and are unrealistic for many families; this causes children to not be placed with relatives."
In general, although study participants expressed commitment to ensuring appropriate
relative placements for children, the study participants suggested that the requirements represented
by the relative packet "doesn't take into account economics or cultural practices" and
"disproportionately impacts families of color." These concerns were repeated across focus groups
and interviews and were epitomized by the following observations and examples:
The only people that canfit into it [relative packet] are those with money ... the packet
doesn't take into account the families we deal with. Many families are able to put four
children in a bedroom and they 're able to do fine, each child has their own space. It
[relative packet] does not allow for any arrangement like that at all. "
There was a case with 10 and 12 year old Hispanic/Native American girls who were
abandoned and went to live with a relative. The relative contacted CPS and stated that the
mother had not returned and wanted to know what to do with the children. The worker was
unable to assess the relative 's home immediately; therefore, the children were then placed
in a foster home. The licensed foster home was horrible, they ffoster parent] placed the
older girl on the floor on a crib mattress. It was terrible and all because they couldn't
continue to stay where they hadbeen, because I [worker] couldn't get the clearances fast
enough.
74
Three young Mexican American girls needed a place to live. The aunt was willing to take
them in and the children were comfortable with her. The uncle had a criminal history and
was a recovering heroin addict, who had been clean for approximately 20 years. On paper
the uncle looked really bad, the worker was unable to clear his past. The children were
placed in foster care and may be adopted if the mother doesn't comply with the service plan.
In addition to barriers to relative placements, social workers pointed out a need to develop
strategies for working with foster care and adoptive families. Respondents noted that these alternate
caregivers are often deeply disappointed when the children in their care are reunified. Some social
workers suggested working more closely with fost-adopt families about the probabilities for
reunification. Foster family and social worker focus groups also expressed a need to provide
training to foster families in practical strategies for caring for children across cultural groups and a
need to develop mechanisms for experienced foster parents to mentor new foster parents.
Continuing Care/Ending Choice Points:
Practices that Positively Impact Children and Families of Color
Social workers described three overall practices that positively impact children and families
of color at the continuing care/ending choice points including 1) parenting education services that
involve both parents and children in services, 2) cultural matching for children in foster care, and 3)
recruitment, training and coaching for foster families.
Parenting education services that involve both parents and children in services were
described as practices that positively impact children and families of color. Social workers and
supervisors noted that parenting classes that involved children and provided opportunities for
parent-child interaction were also a promising practice in comparison to traditional parenting
classes. Some social workers suggested that it might be helpful to "wait until children have been
returned to order parenting classes" or to arrange for children's participation in classes.
Additionally, social workers described cultural matching for children in foster care in which
workers attempt to place children in a home with foster parents of a similar background as a
preferred practice. This was perceived to be critical in relation to matching the primary language of
the child and particularly important in the context of federal and state time lines and concurrent
planning. At the same time, both social workers and foster parents stressed the importance of a safe
and caring placement for children, whether a cultural match was possible or not. Social workers
expressed concern that the particular issues and needs of children, including but not exclusive to
culture, are sacrificed to the expedience of "finding any placement." Foster parents echoed a
concern about considering both the needs of the child and the experience of the foster parents when
arranging a "match." One foster parent caring for a special needs child of a similar background
noted that she would have been overwhelmed and tempted to abandon the idea of foster parenting if
her first foster child had been particularly challenging. Foster parents stressed that they and other
foster parents often have successful experiences with children of different backgrounds and that
they would "never want to see a child wait for a permanent home or left in a shelter because of a
cultural difference."
Social workers and foster parents suggested that recruitment, training and coaching for
foster parents is critical to working effectively with children and families of color. Specifically,
foster families need to know about how to care for children placed in their home. For example,
75
foster parents caring for African American children for the first time required basic information
about hair care. Foster parents suggested that best practices for working with children of color
included obtaining coaching and training from other foster parents, learning about the child's
culture and the individual child, and collaborating with social workers to interface with birth
parents. Study participants noted that tensions between foster families and birth families were
common but that positive practices for children required minimizing these potential problems.
Foster families described "success stories" in which foster parents maintained supportive
relationships with children who were returned to their family of origin.
Social workers and foster parents both noted the need to recruit diverse foster families,
particularly Latino and African American families in Santa Clara County. One foster mother who
identified as Latina suggested that some families in her community might be reluctant to become
foster parents because they are more likely to already have children than Anglo families. Study
participants identified outreach to prospective foster families in underrepresented communities as an
important continuing practice.
Continuing Care/Ending Choice Points:
Recommendations to Improve Practices
Recommendations to improve practices at continuing care/ending choice points included, 1)
providing or facilitating access to concrete support for families, 2) facilitating increased availability
and access to services, and 3) supporting and developing foster families. Specifically,
Provide or Facilitate Access to Concrete Sugport for Families
•
•
•
Examine options for restoring access to bus passes for clients, both in terms of increasing
number of available passes and decreasing paperwork required from social workers to obtain
passes for clients.
Explore ways to increase access to childcare for parents participating in voluntary or mandated
services.
Facilitate access to services related to transitional housing, jobs, financial assistance (e.g.,
CalWORKS), and for families as well as educational interventions for children in the child
welfare system.
Facilitate Increased Availability/Access to Services
•
•
•
Work with policy makers, courts, communities, and other service delivery systems to identify
and address issues related to insufficient services for families. Areas where services are lacking
or limited in relation to the perceived need include substance abuse treatment (particularly
residential treatment for women and children), transitional and drug free living, mental health
services for children, services for young children exposed to trauma or substance abuse, after
school programs and other resources for children.
Collaborate with other systems to protect funding of key services in the context of state budget
problems, such as services for Oto 5 programs.
Work with contract agencies to facilitate inclusion of children in parenting classes to allow for
both didactic and experiential learning opportunities.
76
•
Sustain and increase in-home services, such as intensive wrap around services and one-on-one
mentoring du.ring home visits.
Support and Develop Foster Families
•
•
•
Continue and expand efforts to recruit foster families that are representative of diverse
populations in the child welfare system and that match the languages of different children and
families (e.g., additional Spanish speaking and Vietnamese speaking foster families).
Provide training for foster families in practical strategies for caring for children across cultural
groups.
Continue to develop mechanisms for experienced foster parents to mentor new foster parents.
Contextual Factors that Negatively Impact Practices for Children and Families of Color
Study participants noted three overall contextual factors that negatively impact child welfare
practices for children and families of color 1) agency level factors that detract from social work
practice, 2) a need for more bilingual and bicultural social workers and services, and 3) time limit
constraints.
Certain agency level factors that detract from social work practice were mentioned by many
study participants as negatively impacting practices for children and families of color. For instance,
heavy caseloads and staff shortages were identified as critical factors impeding effective practices
for children and families of color. Children and families of color, particularly famili~s who may be
less trusting or more intimidated by the system, may require language specific services, or who may
have multiple vulnerabilities were deemed very likely to be impacted by the heavy caseloads of
social workers. One social worker who works primarily with Spanish speaking families observed
that many of her clients require "a lot of hand holding" in navigating through the system. Given
heavy caseloads, even dedicated social workers that invest substantial time with some families will
have other families "slip through the cracks." One participant explained:
The weight ofthe caseload is so heavy-there 's no time to do social work. I always feel I
have to rush. Every minute ofyour day is booked. There is no time to explain court
processes to clients.
Many participants commented positively on agency plans to reduce caseloads. This was
recognized as a critical move toward improving overall effectiveness in day-to-day social work
practice. Yet some social workers commented on some of the weaknesses in the current method for
determining caseloads. "There's no equity in the way the agency does the point system for
determining workers' caseloads .... it's based on a 100 point system and you could have 70 points
that could be 35 children in permanent placement, which is a lot of children to manage, whereas
someone else could also have 70 points in Family Reunification (FR) cases, which is not the same
workload." Others suggested that there appeared to be "no rhyme or reason to how cases get
assigned-workers with a lot of points will still get assigned cases-even when a different worker
has a lower number of points." Social workers noted that the point system accounts for the number
of children in a caseload and also assigns a higher number of points to FR cases than to Permanent
Placement (PP) and Family Maintenance (FM) cases. However, they point out that the point system
77
does not necessarily account for the severity of the case and for individual differences between
cases. "Even a case with low points can be very complicated and take up a lot of your time."
Social workers also highlighted the lack of adequate support staff to work effectively with
families, including help with transportation, investigating relative placements, and follow up with
clients at home as negatively impacting children and families of color. Social workers also
commented on the emotional strain on the job and risk for burnout that is exacerbated by too little
support.
We don't have enough support staff-not enough Social Worker J's. It's a lot harder to
manage a good FR case without a SW I to document things, help with transport, etc ... it
becomes more challenging.
This leads you on the road to burnout-I always want to do my best-I want to work harder
to get the parents to be successful. There are some workers who are tremendously burned
out.
Additionally, social workers noted that substantial amounts of required paperwork can
detract from building relationships with clients. Several social workers suggested that, as an
example, writing interim court reports in addition to required reports was duplicative and used time
that would be better spent with clients. Social workers also pointed out that not having easy access
to centralized and updated information about resources and services for families impedes effective
practice. The need for centralized and current resource information was noted as important to
avoiding "reinventing the wheel" in relation to researching resources that other workers already
knew about.
Moreover, confusion about the agency's overall mission, as well as confusion about key
agency and child welfare policies was also described as negatively impacting child welfare practices
for children and families of color. A need for greater clarity about the mission ofDFCS was
expressed. "Who is the client, the child or the parent?" asked one participant. Participants varied in
their perspectives about the balance between child protection and family preservation. Social
workers also expressed confusion over certain key policies including concurrent planning, home
assessment protocols and protocols for voluntary versus court placements. Social workers suggested
that clearer guidelines related to implementation of concurrent planning were needed. Additionally,
participants reported that a more detailed written protocol on home assessments would be
beneficial, ''the policy is not in writing, there is no written protocol and the most recent memo has
DRAFT written on it." Consequently, it is "difficult completing assessments and explaining the
policy to relatives, when the workers aren't clear about it themselves." Last, uncertainty about
protocols for voluntary versus court placements were noted. "There are misconceptions about
voluntarily placing a child and when that child is placed into protective custody... about whether
this protocol is required for all placements, it seems to go back and forth," commented one social
worker. Another noted "I don't have the understanding about what is a voluntary placement versus
a child taken into protective custody; it seems that ifwe get involved in any kind of removal that it
generates a protective custody."
A need for more bilingual and bicultural social workers and services was also noted as a
contextual factor that negatively impacts children and families of color. Although study participants
generally commented on the diversity ofDFCS staff as a strength, there was a perceived need for
78
continued growth in this area. "There are not enough workers who speak the different languages of
the families," noted focus group participants. Some social workers suggested that "the clients are
left in limbo when a worker who speaks their language is not found" and "when a worker is found
that does speak the language of the client then two months have passed by; due to time line
constraints the worker is forced to make decisions without knowing much about the family."
"English speaking workers assigned to Spanish-speaking clients is a problem-even if the parents
and child speak English, a lot of the extended family may not," noted one worker. Reliance on
interpreters was identified as problematic because it can be difficult to find interpreters within a
limited amount of time which can lead to an inability of clients to obtain a timely answer to a
question from a worker because it may take days to find an interpreters. Moreover, there can be
problems with interpreters who are culturally inappropriate, who do not translate appropriately, or
who fail to read all of the court report to the clients.
In spite of these limitations, additional qualified interpreters and translators are needed to
facilitate communication with clients and to provide materials in client's primary language. Study
participants stressed the need for increasing language capacity both through social work staff and
through access to qualified interpreters. Access to qualified interpreters remains critical to working
in a region with representation of a growing number of cultural and language groups. "Lack of
interpreters causes continuance of cases," noted one social worker. Study participants offered a
number of illustrations of the impact that language barriers had on clients. One example is offered
below.
There are only two Korean interpreters in the South Bay. In one case with a Koreanfamily,
Korean interpreter could not be found for a court hearing. This father had to take offfrom
work five days in a row to go to court, because each day an interpreter wasn't there. This
impacted the father because ofthe time he had to take offofwork. Each day he became
more and more frustrated By the time the judge saw the case, the judge saw him at his
most frustrated point.
In addition to the need for bilingual and bicultural social workers, study participants also
expressed a need for more culturally competent services in general. "There is a need for more
culturally diverse counselors and therapists," noted study participants. Barriers to successful
compliance with mandated services include lack of counselors who have knowledge of different
cultures and long waiting lists for the few culturally diverse counselors in the area, according to
study participants. Furthermore, social worker participants observed that White families appeared to
attach less stigma and experience less resistance to accepting mental health or counseling services
than families of color. "Lack of services designed for the needs of a specific ethnicity or family can
cause Dependent Intake to hold onto cases longer due to being unable to find the necessary
resources." Some social workers suggested that existing services for English speaking clients varied
in quality and that contract agencies also differed in the degree to which they prioritized access to
services for DFCS clients.
A shortage of services for monolingual non-English speaking clients was identified as a
substantial problem for many families. "Families stay in the system longer because of lack of
Spanish speaking and other language services ... this can be frustrating for families, they want to
give up." Study participants described struggling tenaciously to find appropriate services for
families. Examples included searching for domestic violence and substance abuse services in
Tagolog, finding a Vietnamese-speaking counselor, and locating a group home for a deaf child.
79
Several participants expressed concerns that services provided in English for clients whose primary
language is non-English were ineffectual and, on occasion, harmful to families such as when
"English-speaking counselors who work with monolingual families will use the children as
translators." Spanish speaking services were perceived to be an issue throughout the county and
were particularly described as pressing by South County study participants. During one focus group
discussion of this issue a participant provided an example:
A Hispanic monolingual mother living in Sunnyvale does not have transportation [to
mandated services]. There are no Spanish resources in her area. They are all in the San
Jose/Santa Clara area, which causes the family to commute farther. It is difficult finding
resources in her area and with no bus passes transportation is a big issue.
Study participants also described time limit constraints as having a negative impact on
children and families of color. Focus group participants noted that it was often difficult to work
toward family reunification in the context of federal and state time limits. Participants uniformly
affirmed the importance of expediting placement of a child into a permanent, safe and stable home
and to seeking the least restrictive placement options for children. At the same time, they suggested
that timelines made it difficult to have enough time to work with families that require services in
order to safely reunify. "Fifteen months is not enough time to work with the client's
issues/problems. The reality of what people are dealing with is not there in the law," was a typical
observation by social workers. Several social workers and interviewees offered examples of
children who were placed into adoption who would otherwise have been able to remain with their
families if time limits were less rigid: "Court time limits played a large role in the adoption of the
child ... the time limits affected the outcome of the family and did not take into account how difficult
it is to change the way one had lead their life in a matter of a few short months."
The issue of access to timely services in the context of time limits was also perceived as an
important issue. An interviewee described a successful case in which a Latina client with a
substance abuse problem was able to reunify largely because the social worker was able to get the
mother into substance abuse treatment despite waiting lists: "There are waiting lists for services and ·
not enough time to successfully complete treatment with the waiting lists and the time limits ...This
family was lucky; the worker was able to get the mother into treatment within 2-3 weeks, so being
at the right place at the right time accounts for this [success]." Conflicts between child welfare time
lines were perceived as particularly common in relation to addressing the needs of substance
abusing parents that represent a majority of cases. "The Federal government only gives families in
permanent planning 12 months to complete services and parents only have 6 months for children
under 3 years of age; this is especially difficult for parents who abuse substances."
Study participants suggested that time limits disproportionately impacted families of color:
"Children of color are brought into the system because parents are on drugs, they aren't being
watched and are in an unsafe environment at that point. The children stay in the system because the
expectation is for the parents to all of a sudden to get better because their children are in the system,
'so if you [parents] don't get better we [social workers] are going to keep your children. This is
unrealistic, especially with the court timelines," exemplified comments made by many study
participants. Families whose primary language is other than English were also perceived as being at
greater risk for failing to reunify: "It is difficult to meet timelines with certain services that are not
offered in different languages; clients are placed on waiting lists or those services don't exist."
80
Contextual Factors that Positively Impact Practices for Children and Families of Color
Respondents noted four overall contextual factors that positively impact child welfare
practices for children and families of color including 1) cultural matching and cultural consultation,
2) supportive supervision 3) collaboration with other agencies and systems, and 4) a strength-based
approach to services.
Cultural matching and cultural consultation was noted as a best practice for children and
families of color. Although study participants did describe a need for more bilingual and bicultural
staff, they also commended the agency's efforts to provide a cultural match between clients and
social workers. Cultural matching was viewed as an effective strategy for facilitating development
of rapport with clients and minimizing the potential for cultural bias in assessment. Specifically,
social workers and supervisors valued the culturally specific Emergency Response (ER) units and
acknowledged efforts to match clients and social workers as much possible in Dependent Intake
(DI). At the same time, respondents asserted that it is not always possible to assign to clients
workers of a similar cultural background. In this context, social workers described obtaining
cultural consultation from other child welfare professionals as a best practice.
At the same time, study participants acknowledged that clients might prefer to interface with
social workers of a different culture than their own. For example, several social workers that
specialize in working with Asian American/Pacific Islander groups commented that some families
prefer to be assigned to a social worker of a different background out of a sense of privacy, pride, or
concern about judgment. Both social worker and client focus group participants noted that cultural
matching was not a guarantee against bias. Social workers are not immune to being judgmental
about clients of their own culture or biased in relation to other issues such as socioeconomic status.
In some cases, clients described their experiences with social workers from a different background
as positive and that the quality of respect was more important that sharing a common background.
To this end, continuing training on cultural competence and working effectively with specific
population groups was also considered a key practice.
High quality, supportive supervision that integrates coaching in working effectively with
diverse families was viewed as critical to practice with children and families of color. "Some
supervisors want to hear about cases and are supportive-others just want to know if you got the
reports done on time," summarized one participant. In addition, supervisors were identified as key
players in the process of brokering cultural consultations with other social workers, particularly
between social workers from different units.
Collaboration with other agencies and systems concerned with the safety and welfare of
children and families of color was also highlighted as a beneficial practice for children and families
of color. For instance, developing interagency formal agreements and connections with liaisons in
other agencies was described as beneficial because it facilitates collaboration and minimizes the
potential for agencies to give conflicting messages and mandates to clients. Social workers are "able
to divert cases or get a better picture of the family when two agencies are working together to work
with the families, which helps not only us but also the client." Other agencies mentioned in this
context include probation, hospital, medical providers, family and mental health agencies,
CalWORKS, substance abuse treatment agencies, domestic violence agencies and other entities that
may be involved with clients.
81
Additionally, collaboration with other agencies to address systemic issues was described as a
best practice. Social workers noted that community meetings are held on a monthly basis for all
community agencies and are attended by program managers, but workers are encouraged to attend.
Participation in policy level discussion between systems was considered to be an important part of
collaboration. Several participants mentioned the Greenbook project, an initiative to better
coordinate domestic violence and child welfare services, as a model for interdisciplinary problem
solving and policy development. The Greenbook project successes could help inform evolving
collaboration with other key systems, such as the substance abuse treatment system.
Other collaborative efforts such as outstationed social workers, co-location of services and
Family Drug Court were noted as best practices for children and families of color. Participants
described having ER social workers at different sites such as the Family Violence Center and police
stations as a valued practice. Moreover, South County social workers noted "we are located at the
resource centers where there are community providers here that are not child welfare ... we've got
AA meetings, parenting classes-everything is centralized here and that makes it comfortable for
the families." Although social workers mentioned that co-location of services have been reduced
because of budget changes and other factors, it remains a positive practice for children and families,
particularly low-income families and families of color that may have less time and fewer resources
to access services. Lastly, social workers and supervisors described the family drug court as a model
for effectively working with substance abusing families in the child welfare system. Several
examples of "successful" cases involved families assigned to family drug court.
A strength-based approach to services was also mentioned as an important best practice that can
be used at all points in the CWS. Indeed, strength-based practice was lauded as a fundamental tenet
of social work that is crucial for working with children and families of color. Specific elements of
this practice named by study participants include, looking at the unique characteristics of each
family; building solutions with families and showing that they [families] do have some control;
being willing to try new things, especially when clients make the suggestions; involving the parent
in the removal of the children, a strategy that is empowering to parents and less traumatic for the
children; demonstrating a commitment to invest time with families on the "front-end" to build
rapport, facilitate client understanding of the child welfare system, and optimize chances for
maintaining or reunifying the family; working with families to identify relative placements when
out of home placements are required and coaching relatives through the application process; and
lastly, demonstrating a willingness to advocate for the child and family with the court and other
service delivery systems.
Contextual Factors:
Recommendations to Improve Practices for Children and Families of Color
Study participants described five overall recommendations to improve practices for children
and families of color as they relate to contextual factors, including 1) reducing caseloads, increasing
support and streamlining service delivery, 2) maintaining and expanding multifaceted approaches to
providing culturally competent practice, 3) increasing linguistic capacity of systems and services for
families, 4) fostering an organizational culture that is strength-based and community based, 5)
enhancing collaboration, and 6) training social workers. Specifically,
Reduce Caseloads, Increase Support and Streamline Service Delivery
82
•
•
•
•
•
•
•
•
•
•
Continue efforts to reduce social worker caseloads. Reduction in caseloads was uniformly
identified as key to providing quality services to clients, particularly children and families of
color that may have complex needs and that may initially have less trust and/or less'familiarity
with child welfare and court systems. "Social workers need time to do their job and to help
clients access culturally appropriate services and programs."
Provide additional Social Worker I support for in-home assessment/support, follow up with
families, and overall cas~ support.
Provide dedicated staff, and explore other options, to expedite requirements to qualify relative
placements. For example, policies requiring relative placement standards similar to foster
families might be reevaluated (e.g., illustrations of this as a problem included an example of two
young girls were denied relative placement because they shared a queen size bed rather than
having separate beds). In addition, develop strategies for overcoming barriers to relative
placement, such as an expedited process for Director's Exemption in the case of a distant
criminal history of a family member in the prospective home.
Examine opportunities to facilitate quicker access for social workers to criminal background
checks.
Develop procedure or protocols to facilitate communication between units, particularly in
relation to transferring ER cases to voluntary family maintenance informal supervision.
Examine options for maximizing the skills of social workers with specific cultural or language
expertise. For example, social workers in ER and DI stated that support across units that
involved overtime was recently prohibited. However, social workers suggested that this practice
failed to reduce the total amount of overtime expended by social workers while limiting
flexibility in obtaining culturally and linguistically skilled consultation and assistance.
Develop centralized, and updated resources lists for a wide range of services, including service
providers with expertise in serving diverse cultural and language groups. This practice was
identified as key to avoiding duplication of effort and ensuring access to quality services for
clients. "Sometimes we make a lot of phone calls just to find out what someone two aisles from
you already knew." In addition, update Court Services Handbook, which was perceived by
several social workers to be out of date.
Examine opportunities to provide flexibility in office hours, alternate social workers to cover
cases when social workers are on vacation, or other mechanisms to reduce potential burnout.
Provide social workers access to adequate transportation (e.g., "better cars") for trav~l related to
working with families.
Develop protocols that reduce paperwork so "there is more time to do social work." For
example, interim court reports could be eliminated or reduced, templates for psychosocial
assessments or other guidelines for court reports could be provided, and social workers could
"get assistance especially with filing."
Maintain and Expand Multifaceted Approach to Providing Culturally Competent Practice
•
•
Continue practice of attempting to provide a cultural match for clients as a strategy for
facilitating rapport and, in some cases, minimizing cultural or class bias.
At the same time, maintain recognition that 1) some clients may prefer to see a Social worker
who is not of their background, 2) many clients value "being treated with respect" above being
assigned a social worker of the same background, and 3) the diversity of the population in Santa
Clara County may not always allow for "matching" clients and social workers in relation to
race/ethnicity or other demographics. Ensure training and support for all social workers in
83
•
•
cultural competence and addressing the needs of different population groups among all social
workers
Continue working to ensure diverse staff in units throughout the agency.
Continue using the family resource centers, including provision of services for mandated clients.
Increase Linguistic Capacity of Systems and Services for Families
•
•
•
•
•
Facilitate development of services specifically designed for families whose primary language is
not English in key areas including mental health, domestic violence, substance. abuse,
counseling services, parenting without violence classes, and other services.
Continue to hire bilingual/bicultural staff. Need for greater linguistic capacity was reported for
a number of different language groups including for Spanish speaking, Vietnamese, and Filipino
clients.
Minimize need for use of translators. When needed, provide high quality translators who are
able to communicate concepts related to the law, the social work system, timelines, and court
processes.
Provide forms and written materials in multiple languages.
Examine opportunities for establishing guidelines for contracting or collaborating agencies in
relation to linguistic services, such as procedures disallowing use of children as translators for
monolingual speaking parents.
Foster an Organizational Cult~re that is Strength-Based and Community-Based
•
•
•
•
Continue to identify mechanisms, such as the Family-to-Family Initiative, to build community
involvement in decision-making.
Communicate a commitment to assuring the least restrictive placement for children from
management, through supervision, and to line staff. Provide training and support for social
workers to advocate for clients in court and other settings.
Examine strategies to minimize what study participants described as internal disincentives to
ensuring the least restrictive placement for children including "County Council whose role it is
to protect the agency sometimes telling social workers to change recommendations" and the
greater paperwork and advocacy required for less restrictive options.
Ensure client access to information about complaint procedures and access to the Ombudsman.
Enhance Collaboration
•
•
Continue to develop collaborative case planning, linkages, information sharing and policy
development with other service delivery systems, including schools, service providers,
probation, CalWORKS, housing, and other systems. Collaborative planning and problem
solving was perceived as particularly crucial in an environment of sever fiscal restraints.
Develop protocols to facilitate access to mandated services for clients in the context of federal
and state timelines, such as formal agreements and protocols to expedite access to substance
abuse treatment.
Train Social Workers
84
•
•
•
•
•
•
•
•
Require cultural competence training for all new staff and mandate continuing education for all
social workers. Include, as part of the training, opportunities for self-reflection and examination
of personal bias related to culture of other issues frequently confronted in families, such as
substance abuse. In these areas, provide training that is high quality, dynamic, and delivered by
skilled trainers who are knowledgeable about both their subject area and child welfare.
Provide education within the agency about different cultural norms in parenting and family
strengths.
Provide training on specific issues that impact all families and that may differentially impact
children and families of color including policies and practices for addressing drug exposed
infants, substance abuse, domestic violence, sexual abuse, and other topics. In addition, provide
training on procedures and practical considerations for working with diverse families, such as
training on guardianship.
Provide training on immigration law and practical strategies for working with families, such as
identification of immigration status of the child at an early stage and working with international
relative placement.
Training on these topic areas should integrate consideration of cultural differences and
evidence-based practices for different populations.
Provide training for all staff including clerical staff and eligibility workers.
Facilitate forums for social workers to share resources and effective practice strategies.
Create opportunities for social worker access to training. "We can't go to trainings because we
have too much else to do-so we're denied professional development opportunities."
Summary of Key Findings and Implications
Study participants described numerous practices and contextual factors that have either a
negative or positive impact on children and families ofcolor in Santa Clara County's child welfare
system. There is some evidence in the research literature that supports certain best practices
mentioned by study participants. For instance, general research supports the effectiveness of inhome services (see literature review), however these services appear to be most effective when they
are intensive, long-term and delivered by health or social service professionals and when the
workers delivering these services receive a high level of supervision (Barth, 1991; McGuigan et al.
2003; Olds et al., 1997). In addition, some research has suggested that parenting classes for child
welfare parents are more effective when they pair a didactic approach with a more experiential
approach, such as parent and child interaction (Carlo, 1993).
Participants also identifiedfamily conferencing as an effective practice. These study
findings are consistent with limited research in this area that suggests that inclusive practices, in
which parents are engaged in placement and other decisions, may result in more placement stability
for children, although possible differential effects for diverse populations has not been researched
(Palmer, 1996). Participants generally expressed optimism about the adoption of team decision
making (J'DM) as a vehicle for improving decision-making with diverse families, although some
social workers expressed concern about time required for implementation. Research does suggest
that TOM can be labor intensive and can be difficult to implement effectively within the context of
the child welfare system (Sieppert et al., 2000). Finally, participants noted that both formal and
informal collaborative relationships with other services delivery systems, such as domestic violence
and substance abuse treatment systems, are critical for successful interventions with children and
families. Furthermore, participants recognized the importance oftraining and cross-training to
85
enhance the capacity ofhelping professionals in child welfare and other systems to better address
the needs ofchildren and families ofcolor.
Interestingly, most practices and contextual factors mentioned by social workers had both
positive and negative aspects. Many of the practices and contextual that negatively impact children
and families of color were actually barriers to successful implementation of the practices that social
workers felt have a positive impact on children and families of color. Although practices such as
orientation ofnew clients; family and team decision-making; preventive services; substance abuse
services; cultural competency; collaboration with other agencies and systems; a strength-based
approach; and the use ofrelative placements were described as having a positive impact on
children and families ofcolor, certain negative practices and contexts impeded their effective
implementation. For instance, having to screen out large numbers of inappropriate referrals;
inconsistency in decision-making practices; a shortage ofservices, particularly substance abuse
and preventive services; lack ofclient access to services; difficult protocols for placing children
with kin; gaps in cultural competency; time limits; and agency-level factors such as heavy ·
caseloads, staffshortages, substantial amounts ofpaperwork, lack of access to information about
resources; and confusion about the .agency's overall mission and key policies all impeded
implementation ofthe best practices. These findings suggest that in addition to building on current
positive practices for children and families of color, CWS stakeholders should also actively work
both internally and in collaboration with partners to reduce barriers to these best practices.
Indeed, in order to address these barriers to effective implementation of best practices, study
participants mentioned numerous recommendations that centered around certain key themes. In
general, there was an emphasis on improving decision-making through increased accountability,
training and cross-training to reduce bias, and increased use of group decision-making or family
involvement in decision-making. Study participants also stressed a need to improve and expand
prevention, diversion and concrete services, as well as an overall need to expand culturally and
linguistically competent services, and develop ways to improve availability and access to services.
Once a case is opened, social workers also felt it valuable to provide an orientation to clients so that
they are more aware of the CWS and court processes and can better navigate the system.
Recommendations related to organizational factors were also noted, including reducing caseloads;
increasing support and streamlining service delivery; clarification .ofthe mission ofDFCS, as well
as key child welfare policies; fostering an organizational culture that is strength-based and
community-based, and more overall training for social workers.
Limitations and Conclusions
Qualitative data, such as focus groups and interviews, provide rich insights into practices
that may positively or negatively impact children and families of color. In any such endeavor, there
are possible biases that may impact the results. For example, given the likelihood of time conflicts
and emergencies, we invited more social workers to participate in the focus groups than we
expected to attend. It is not_possible to determine if other factors influenced participation. Workers
with a particular interest in this topic may have made a special effort to attend. In addition, the
perspectives and opinions of the social workers, supervisors and families who participated in the
study may not be entirely reflective of the populations they represented. At the same time,
participants were representative of different units and facets of the child welfare system, different
race/ethnic groups, and varied levels of experience. The findings ofthis study affirm the value of
many efforts that are already in place, such as culturally specific ER response units and family
86
resource centers, and other that are in the process of implementation including efforts to reduce
caseloads, initiation ofteam decision making, and participation in the Family-to-Family initiative.
The findings ofthe study also point to opportunities to further strengthen services for children and
families ofcolor. Recommendations related to training, institutionalizing group orientation for
clients, and other suggestions from study participants could be used to inform practice and planning.
It would be beneficial for a team ofmanagers, line staff, family representatives, and community
members to review these findings to identify and prioritize possible practices for adoption based on
their feasibility and utility for children and families ofcolor.
87
Comparison of the Main County Offices and South County Offices
In an effort to better understand the influence of contextual factors on child welfare practices
and outcomes for children and families of color, a comparison of Santa Clara County's Main
Offices and South County Offices was conducted. This comparison included both quantitative and
qualitative analyses. Quantitative data from CWS/CMS were used to examine differences in case
characteristics between the Main County Offices and South County Offices. A qualitative
comparison of focus group data from the Offices and South County Offices was also conducted to
discern how workers perceive practices in these two geographical locations. These analyses help
shed light on the potential impact of agency and community context on case characteristics, child
welfare practices and outcomes for children and families of color in Santa Clara County's CWS.
Quantitative Methods
In Phase 2, the CWRT continued its investigation of the disproportionate representation of
children of color in the child welfare system by analyzing a sample of cases closed during an 18month period (January 2000 through June 2001 ). Key case characteristics were analyzed, including
ethnic comparisons. Please see the Phase 2 report for methodology details and case characteristics
by ethnicity (Hines et al., 2002).
One of the recommendations that emerged from Phase 2 of the study was to explore possible
differences in case characteristics and practices between the Main County Offices and the South
County Offices. In Phase 3, we used the basic characteristics from the 1720 cases analyzed in Phase
2 while adding the Santa Clara County location information provided via the Child Welfare System
/ Computer Management System (CWS/CMS).
Analyses
A series of exploratory, bivariate analyses were conducted with county location as the
distinguishing variable, Main County (MC) versus South County (SC). The purpose of this initial
set of analyses was to identify differences in the CWS related to county location in order to help
describe variations not only by ethnicity, but by other case and service characteristics. The
following sections will describe (1) the proportion of cases with each county location, (2) general
characteristics by county location, and (3) significant differences related to county location.
Definition of Terms
County Location - Two overall geographical areas within Santa Clara County designated as either
Main County Offices (centered primarily in the Northern part of the County, including the main
office on Julian Street in San Jose) or South County Offices (based in Gilroy).
Quantitative Results
County Location
Of the 1720 cases, 1633 (94.9%) had valid county location information with 87 missing
(5.1%). Of the 1633 valid cases, 1487 (91.1%) were served in MC, 50 in SC (3.1%) and 96 (5.9%)
88
in both county locations. Please see Table 15.
•
a
T able 15 : P roport1on o fC ases biy County Location
Total
Main (MC)
8
1633
1487
(91.1%)
South (SC)
Both
96
(5.9%)
50
(3.1%)
a Based on 1633 valid cases, excluding 87 cases with missing location
For the primary ethnic comparison, we retained three county location designations, MC, SC,
and Both (cases served in MC and SC). However, given the range in the length of time children
designated as Both were served in MC and SC, and the focus on comparing the two stated
geographical areas, subsequent analyses excluded this third category.
County Locatio~ and Child. Characteristics
Child's Ethnicity
There was a significant relationship between the child's ethnicity and county location. As
noted in Phase 1 and 2, Latinos compose the largest ethnic group served in the county's CWS.
When examined by location, those served in MC were more likely to be Latinos (44.7%, 661 of
1480 MC cases). This proportion of Latinos was higher in SC (68.0%, 34 of 50 SC cases) and Both
(62.1 %, 59 of 95 Both cases). African American, Asian American/Pacific Islander, and Other
children's cases were least likely located in SC and Both locations. Please see Table 16.
Table 16: Ethnic Group By County Location
Total
Child's Ethnicitya
•
•
N size
•
White
•
476
(29.3%)
Latino
754
(46.4%)
•
•
African American
1625
177
(10.9%)
Asian American
/Pacific Islander
156
(9.6%)
Other ethnicity
62
(3.8%)
Main
South
Both
1480
50
1
(2.0%)
95
6
(6.3%)
13
(26.0%)
29
(30.5%)
34
(68.0%)
59
(62.1%)
1
(2.0%)
0
(0.0%)
1
(2.0%)
1
(1.1%)
170
(11.5%)
434
(29.3%)
661
(44.7%)
155
(10.5%)
60
(4.1%)
a Based on 1625 cases with valid MC and SC county locations and county location infonnation
Unfortunately, the South County Office subsample was too small to allow more detailed analyses by
ethnicity. However, we continued to pursue our comparison of the two primary county locations by
other key variables.
89
Child Demographics
There was not a significant relationship between gender and county location. Within MC,
52.2% was female (776 of 1486 cases) compared to SC where 50.0% was female (25 of50 cases).
Within SC, 47.8% was male (710 of 1486 cases) compared to SC where 50.0% was male (25 of 50
cases).
There was not a significant relationship between age and county location. The average age
of the MC sample was 5.95 years (sd = 5.16), while the average age of the SC sample was 6.57
years (sd = 4.19).
There was not a significant relationship between child's language and county location. The
majority of the MC and SC samples spoke English. However, in SC, Spanish was the only other
language other than English spoken. Please see Table 17.
T able 17 Ch.Id
1 Demograph.1cs
Gender
Total
Main
South
1536
1486
801
(52.1%)
776
(52.2%)
50
25
(50.0%)
735
(47.9%)
710
(47.8%)
25
(50.0%)
5.97
(5.13)
5.95
(5.16)
6.57
(4.19)
8
•
•
N Size
Female
•
Male
Average age in years (sd/
Child's Languagec
•
•
N Size
1524
1474
50
English
•
•
1308
(85.8%)
1262
(85.6%)
46
(92.0%)
Asian American/
Pacific Islander
62
(4.1%)
62
(4.2%)
0
(0.0%)
Spanish
•
144
(9.4%)
4
(8.0%)
Other
140
(9.5%)
10
(0.7%)
•
IO
(0.7%)
a Based on 1536 cases with valid MC and SC county location and gender information
b Based on 1537 cases with valid MC and SC county location and age information
c Based on 1524 cases with valid MC and SC county location and language information.
0
(0.0%)
90
County Location and System-Related Factors
Voluntary Status and Reason for Removal
There was a significant relationship between voluntary status and county location. The
proportion of cases in voluntary services was higher in SC at 88.0% (44 of50 cases) compared to
41% in MC (600 of 1462cases).
There was not a significant relationship between the reason for removal and county location.
Please see Table 18.
T a ble 18 Vo luntary status an dReason or Remova
Total
8
Voluntary Status
N Size
1512
Main
South
1462
50
•
•
Voluntary
•
644
(42.6%)
600
(41.0%)
44
(88.0%)
Not Voluntary
868
(57.4%)
862
(59.0%)
6
(12.0%)
1206
1196
10
Reason for Removal b
•
N Size
General Neglect
•
161
(13.3%)
158
(13.2%)
3
(30.0%)
Severe Neglect
•
172
(14.3%)
172
(14.4%)
0
(0.0%)
Physical Abuse
•
300
(24.9%)
297
(24.8%)
3
(30.0%)
Sexual Abuse
•
73
(6.1%)
73
(6.1%)
0
(0.0%)
Emotional Abuse
•
40
(3.3%)
40
(3.3%)
0
(0.0%)
Other Maltreatment
460
(38.1%)
456
(38.1%)
4
(40.0%)
•
a Based on 1512 cases with valid MC and SC county location and voluntary service information
b Based
on 1206 cases with valid MC and SC county location and removal information
91
Initial Out-of-Home Placement Facility Type
There was not a significant relationship between initial out-of-home placement facility type
and county location. Please see Table 19.
T a bl e 19 : Imt1a
.. 1Out-o f.-Home Placement Facility Type
Total
Main
South
1206
1196
10
Initial Out of Home Placement
Facility Type8
N Size
•
•
Foster Family Home
•
258
(21.6%)
258
(21.6%)
0
(0.0%)
Group Home
•
66
(5.5%)
66
(5.5%)
0
(0.0%)
Children's Shelter/
359
(29.8%)
353
(29.5%)
6
(60.0%)
Relative Home
•
343
(28.4%)
341
(28.5%)
2
(20.0%)
Foster Family Agency
123
(10.3%)
I
•
124
(10.3%)
Guardian Home/Court
56
(4.6%)
55
(4.6%)
•
a
Receiving Home/ Non-EA
/AFDC
Specified Home
(10.0%)
1
(10.0%)
Based on 1206 cases with valid MC and SC county location, and placement information
Number of Times removed from Family and
Number of Unique Placement Homes in Current Episode
There was not a significant relationship between the number of times a child was removed
from their family in current episode and county location. On average, children in the sample were
removed 1.35 times (sd = 0.70).
There was not a significant relationship between the number of unique placement homes in
current episode and county location. The sample had an average of 3.56 (sd = 3.86) unique
placement homes in current episode. Although MC cases had a higher average of placements (3.57,
sd = 3.88) compared to SC (1.90, sd = 1.10), this difference did not reach statistical significance,
most likely due to the small number of SC cases available for comparison. Please see Table 20.
92
Table 20: Number of Times removed from Family and Number of Unique Placement Homes in
C urrentE.
,p1sode
Average Number of Times
Removed from Family in
Current Episode (sd)8
Average Number of Unique
Placement Homes in Current
Episode (sd)b
Total
1.35
(0.70)
Main
1.35
(0.70)
South
1.10
(0.32)
3.56
(3.86)
3.57
(3.88)
1.90
(1.10)
a Based 1228 cases with valid MC and SC county location and removal information
b
Based on 1228 cases with valid MC and SC county location and placement information
Average Stay (in days) per Placement Facility, Number of Placements, and Total Length of
Time in Out of Home placement in Current Episode
There was not a significant relationship between the average stay (in days) per placement
facility in current episode and county location. The average number of days a child stayed in one
placement in their current episode was 244.06 (sd = 480.43). Although MC cases had a higher
average of245.73 days in placement (sd = 482.03) compared to 41.54 days (sd = 44.71) in SC, this
difference did not reach statistical significance, again most likely due to the small number of SC
cases available for comparison.
There was not a significant relationship between the number of placements in episode and
county location. On average, children in the sample had 2.84 placements (sd = 2.95). Although MC
cases had a higher average of2.85 placements (sd = 2.95) compared to 1.80 (sd = 1.32) for SC, this
difference did not reach statistical significance, again most likely due to the small number of SC
cases available for comparison.
There was not a significant relationship between the total length of time (in months) in outof-home placement in current episode and county location. On average, children in the sample spent
13.16 months (sd = 20.40) in out-of-home placement. Although MC cases had a higher average of
13.24 months in placement (sd = 20.47) compared to 3.28 months (sd = 4.24) for SC, this difference
did not reach statistical significance, again most likely due to the small number of SC cases
available for comparison. Please see Table 21.
93
Table 21: Average Stay (in days) per Placement Facility, Number of Placements, and Total Length
. 0 ut o f H ome p1acement m
. Current E,ptso
. de
0 fT1mem
Average stay (in days) per
Placement Facility Type in
Current Episode (sd)8
Number of placements in
Total
Main
South
244.06
(480.43)
245.73
(482.03)
41.54
(44.71)
2.84
1.80
(1.32)
3.28
(4.24)
Current Episode (sdl
(2.95)
2.85
(2.95)
Total length of time (in months)
in Out of Home Placement in
Current Episode (sdt
13.16
(20.40)
13.24
(20.47)
a Based on 1228 cases with valid MC and SC county locations and placement information
Based on 1206 cases with valid MC and SC county locations and placement information
c Based on 1200 cases with valid MC and SC county locations and time information
b
Number of Episodes, Total Length of Case (in years), and Number of Workers Assigned to
Case over Time in the Current Episode
There was not a significant relationship between the number of episodes and county
location. Children in the sample had an average of 1.19 (sd = 0.47) episodes.
There was a significant relationship between the total length of the case (in years) and
county location. On average, children in the sample had a case length of 1.87 years (sd = 2. 78). MC
cases lasted longer with an average of 1.93 years (sd
SC.
= 2.80) compared to 0.18 years (sd = 0.39) for
There was a significant relationship between the total number ofworkers assigned to case
over time in current episode and county location. Children in the sample had an average of 4.91
workers (sd = 2.75). MC cases were assigned a higher average of 5.03 different workers (sd = 2.72)
compared to 1.48 workers (sd = 0.84) for SC. Please see Table 22.
Table 22: Number of Episodes, Total Length of Case (in years), and Number of Workers Assigned
tC
heeurrent E.d
,p1so e
mt
0 ase over T 1me ·
South
Main
Total
1.16
1.19
1.19
Average Number of Episodes (sdt
(0.37)
(0.48)
(0.47)
0.18
Average Total Length of Case (in
1.93
1.87
(0.39)
(2.78)
(2.80)
years) (sdl
Average Number of Workers
Assigned to Case over Time in the
Current Episode (sdt
4.91
(2.75)
5.03
(2.72)
Based on 1512 cases with valid MC and SC county locations and episode information
Based on 1512 cases with valid MC and SC county locations and case time information
c Based on 1532 cases with valid MC and SC county locations and worker information
a
b
1.48
(0.84)
94
Last Out-of-Home Placement Type
There was not a significant relationship between last placement type and county location.
Please see Table 23.
T a ble 23 Last O ut-o f.H
- ome Placement Type, andA.2e at T1me of Case Closure
Total
Main
Last Out of Home
Placement Type8
1218
1228
• N Size
South
10
•
Foster Family Home
•
260
(21.2%)
260
(21.3%)
0
(0.0%)
Group Home
•
77
(6.3%)
77
(6.3%)
0
(0.0%)
Children's Shelter
•
354
(28.8%)
6
(60.0%)
Relative Home
•
346
(28.2%)
348
(28.6%)
344
(28.2%)
2
(20.0%)
Foster Family Agency
•
133
(10.8%)
132
(10.8%)
1
(10.0%)
Guardian Home/
Court Specified Home
58
(4.7%)
57
(4.7%)
I
a
(10.0%)
Based on 1228 cases with valid MC and SC county locations and placement information
Age at Time of Case Closure and Service Type at Case Closure
There was not a significant relationship between age at time of case closure for current
episode an.d county location. On average, children in the sample were 8.19 years-of-age (sd = 5. 72)
at the time of case closure. Although MC cases closed with an older child's average age of 8.23
years (sd = 5.76) compared to 7.15 years (sd = 4.22) for SC children, this difference did not reach
statistical significance.
There was a significant relationship between service type at closure of case and county
location. Overall, the majority of cases ended in family maintenance (59.9%, 920 of 1537 cases).
However, SC had a higher proportion of cases in family maintenance (92.0%, 46 of 50 cases)
compared to MC (58.8%, 874 of 1487 cases). Please see Table 24.
95
. T ype at Case Closure
T a ble 24 A.ge at T1meofC ase Closure andS erv1ce
Main
Total
Age at Time of Case
8.19
8.23
South
Closure for the Current
(5.72)
(5.76)
7.15
(4.22)
Episode (sdt
Service Type at Case
Closureb
1537
1487
50
•
•
N Size
Emergency Response
•
39
(2.5%)
29
(2.6%)
0
(0.0%)
Family Maintenance
•
920
(59.9%)
874
(58.8%)
46
(92.0%)
Family Reunification
•
67
(4.4%)
64
(4.3%)
3
(6.0%)
Permanent Placement
511
(33.2%)
510
(34.3%)
1
(2.0%)
a Based on 1482 cases with valid MC and SC county locations and child's age at case closure information
b Based on 1537 cases with valid MC and SC county locations and service type information
Case Closure Type
There was a significant relationship between case closure type and county location. The
most common case closure type was court ordered termination and other (32.1 %, 482 of 1502
cases), followed by family maintenance (31.6%, 475 of 1502 cases). However, 62% (31 of 50
cases) ~fSC closed with family maintenance compared to 30.6% (444 of 1452 cases) in MC.
Please see Table 25.
96
T able 25 Case Closure T ype
Case Closure Type
• N Size
8
Total
1502
Main
1452
South
50
•
Adoption
•
246
(16.4%)
246
(16.9%)
0
(0.0%)
Emancipation
93
(6.2%)
0
(0.0%)
31
(62.0%)
•
•
•
•
•
a
Family Stabilized
(FM)
475
(31.6%)
93
(6.4%)
444
(30.6%)
Guardianship
99
(6.6%)
97
(6.7%)
2
(4.0%)
61
(4.1%)
59
(4.1%)
2
(4.0%)
Incarceration,
Runaway or Medical
services
46
(3.1%)
45
(3.1%)
1
(2.0%)
Court Ordered
Termination and Other
482
(32.1%)
468
(32.2%)
14
(28.0%)
Established or
Placement with
Relative
Reunified.with Parent
or Guardian, Court or
non-Court Specified
Based on 1502 cases with valid MC and SC county locations and case closure information
97
Qualitative Comparison of the Main County Offices and South County Offices
In addition to the quantitative comparison of the Main County Offices and South County
Offices, qualitative focus group data were collected to obtain information on how workers perceive
practices in the two geographical areas, and how these practices may differ.
Qualitative Methods
Prospective participants for the Social Worker focus groups were randomly selected from a
list of workers with title of Social Worker II or Social Worker III in different units/divisions. The
random selection and recruitment process is described in detail in the section describing qualitative
methods (please see page 61 for a full description of focus group qualitative methods). As described
above, a total of 67 Social Workers participated in the 13 focus group interviews. Of this number,
13 Social Workers from various units in South County Offices participated in three South County
focus groups. All the focus group participants in the South County sample were female. Of these, 2
(15.4%) identified as African American, 1 identified as Asian American/Pacific Islander (7.7%), 4
identified as Latina (26.9%), and 6 identified as white/Caucasian (32.8%). A majority of the
participants (n=l2, 92.3%) had masters level education, primarily MSW degrees (Please see
Attachment 2 for additional information about focus group demographics)
Procedures
Procedures for all focus groups are described in detail in the Qualitative Methods section of
this report. Procedures for all focus groups, including South County Office focus groups were
identical. As described above, ten focus groups were conducted at the main administrative office in
downtown San Jose and three were conducted at Gilroy Family Resource Center in South County.
Analysis
The process for analysis of the South County Offices data compared to other focus groups
was parallel to that described above in the Qualitative Methods section. The data for the qualitative
portion of this study consist of the transcribed notes taken during each of the focus groups and
individual interviews. Qualitative methods of data analysis were employed to analyze the focus
group and interview data. First, major themes were summarized for each of the choice points or
units in the Social Worker focus group and supervisor interview data. A separate analysis of the
qualitative data from South County Offices was conducted to identify themes that may have been
different or more.prominent in these series of 3 focus groups compared to the other 10 social worker
focus groups. The purpose of this additional analysis of the qualitative data was to obtain insights
and supplement quantitative analysis of possible differences between South County Offices and
Main County Offices within DFCS. In addition to gathering information about practices that may
differ between the Main County Offices and the South County Offices, findings from the qualitative
data also provide insights that may help explain quantitative differences in the relationship between
location and other variables, such as differences in the proportion of voluntary family maintenance
cases.
98
Qualitative Results
Many of the contextual and practice issues described by social workers and supervisors in
South County Offices paralleled comments by study participants in other regions of Santa Clara
County, such as concerns related to client poverty, housing shortages, and insufficient substance
abuse and other treatment resources, particularly for non-English speaking language groups. At the
same time, a few themes emerged that were either stronger or unique to South County Offices.
These are described below.
Context of Service Provision in South County Offices
Demographics of South County
According to study participants, South County has a large Latino community that is strongly
represented among child welfare clients. In relation to these demographics, participants emphasized
the need for bilingual and bicultural child welfare professionals. They also reported a paucity of
culturally and language specific counseling services for monolingual Spanish speaking participants.
In addition, access to counseling and services for undocumented families in the child welfare
system were perceived as a problem. Participants noted that one strength of the Family Resource
Center is the bilingual and bicultural social work staff. At the same time, participants saw a need to
expand capacity to work with other communities. "Our resource center is reflective of the majority
community of Latinos, but some clients may not feel represented," explained one South County
focus group member. Specifically, several social workers suggested that the current and evolving
demographic changes in the region called for greater representation of African American and Asian
American/Pacific Islander social workers.
Small County Dynamic
South County Offices were perceived as different from "the main" office in part because
dynamics related to the smaller size of the region. Participants described feeling more connected to
the community, other service providers and clients because of the "a small-town atmosphere."
Participants noted that, "The community is smaller and there is more interaction within the
community and with individuals from other agencies; this tends to make people work with clients
differently." For example, one social worker commented that she often sees clients and colleagues
at the grocery store and others mentioned that clients freely come into the child welfare offices and
Family Resource Centers asking for assistance.
Furthermore, study participants in South County consistently reported a high level of both
formal and informal consultation with one another. In addition to relying on other workers and
supervisors to assess families and develop case plans, social workers working in close physical
proximity are aware of one another's cases and routinely exchange informal support and advice
about working with clients and accessing resources. Some participants noted that they might be
"less likely to connect" in larger facilities or offices "built like a maze." The absence of security for
entering the agency building "changes the atmosphere because clients can just walk in, there is an
open door policy; this changes agency attitude in working with clients and removes barriers for
clients."
99
Culture of Commitment to Maintaining and Reunifying Families
South County study participants described a shared philosophy that the vast majority of
children referred to the child welfare system can and should be with their birth parents, whether it
be through family maintenance or through family reunification. Specific values and norms
embedded in the philosophy and practice of the region include emphasis on using family strengths,
including families in the process of making decisions, and aggressively seeking alternatives to
removal. Study participants underscored the importance of this approach in working with children
and families of color. Although these principles are valued throughout the child welfare system,
participants in South County stated that these principles were infused throughout the system and
were used consistently to guide practice and social workers' accountability in assessment and
decision-making. For example, social workers described substantial involvement of supervisors and
coworkers when removal of children into protective custody is required. Study participant
comments suggested that the smaller size of the agency helps to solidify a sense of shared
philosophy and reinforce practice norms.
When asked to articulate how these values and norms are sustained and passed on, workers
cited several factors. First, study participants emphasized the importance of consistency in
messages, policies, and direction from the Program Manager and supervisors. Second, social
workers suggested that these expectations related to seeking the least restrictive placement for
children are communicated beginning with the initial hiring and orientation process through
everyday practice. "People who get hired have the same philosophy as South County ... and
supervisors ingrain values into workers," noted participants. Third, social workers stated that it was
normative for workers to obtain both formal and informal support when they are making decisions.
"Workers do a lot of staffing and hash things out; we don't make decisions alone," noted one
participant. "By using co-workers and supervisors in decision-making, you gain another
perspective," observed another. Integrating the perspectives of other workers in decision-making
was identified as crucial to mediating potential bias or misunderstanding related to cultural or
socioeconomic class differences. Finally, social workers suggested that they received substantial
support and consultation from supervisors in decision-making and that supervisor values and
practices were perceived as critical to the practices of line staff. For example, supervisor
involvement in decision-making was described as a positive practice, particularly in relation to
taking children into protective custody and out of home placement. "My supervisor was on the
phone the whole time I had to remove a child," was offered as an example from one social worker.
Such oversight was perceived as a sign of support rather than solely a mechanism to assure social
worker accountability.
Practices in Service Provision in South County Offices
Vertical Case Management
Focus group participants and supervisor interviewees named the vertical case management
model, in which social workers assigned to a case work with the family throughout their case rather
than transferring cases to a different unit, as an effective practice with clients, particularly with
children and families of color. Some study participants suggested that the vertical case management
model works well in a small community. "It works because it puts social workers in the community
with the family, making social workers accountable ... it forces social workers to have to deal with
100
that family not just in a [case] process sense but in a real social work sense of developing
relationships," summarized one supervisor interviewee. Social Workers suggested that maintaining
continuity through a specific case enhanced trust with clients and created a greater sense of
accountability. The opportunity to develop and sustain rapport with specific clients was deemed to
be particularly important for marginalized groups, including children and families of color. Families
that return to the system are also generally reassigned to the same social worker. "In this case a new
worker may not have rapport or know family strengths ... A worker who has worked with a family in
the past knows them well enough to call them on their stuff." At the same time, respondents also
described vertical case management as work intensive and requiring additional support to ensure its
success, such as the help of Social Worker I staff.
Study participants in the focus groups that took place at the main administrative office in
San Jose location appeared to be familiar with the vertical case management model. Some
mentioned that it might be a possibility as a positive practice for children and families of color in
other parts of the county while others stressed the importance of maintaining specialized units with
staff who have expertise and an opportunity to stay current with changing laws, policies and
practices in their area.
Orientation of Clients
Social Workers in South County frequently refer parents to a client orientation that helps
parents understand the child welfare system and policies related to timelines for reunification. The
orientation is offered in English and in Spanish. It is offered at different times and provides 6 hours
of information over three sessions. Three areas are addressed in the orientation: timelines and facets
of the child welfare system (ER, DI, FR, FM, adoption), how to work with the social worker, and
resources for children and families. The orientation is not mandatory and is not offered at all sites.
One focus group participant examined outcomes (identified as either reunification or closure of
case) and found that the orientation appeared to be particularly helpful for families in the Spanish
speaking orientation. In sum, the orientation for clients was recognized as an important resource for
families, particularly families of color, who might feel confused or intimidated by the child welfare
system. At the same time, workers appeared to feel that there were opportunities to further
strengthen, institutionalize, and evaluate this practice.
Case Conferencing
Study participants stressed as a best practice the formal and informal process of meeting
with co-workers and supervisors to review cases and participate in the decision-making process.
Accountability and support in decision-making was viewed as an important factor in minimizing
potential cultural or other bias about clients and in maximizing creative thinking about how best to
work with children and families. Inclusion of families in decision-making was also noted as
positive practice that was considered particularly important for children and families from cultures
that place a high value on extended family, including both biological and fictive kin. "If you have a
voluntary family maintenance case, you have a family conference," observed one social worker.
a
101
Summary of Key Findings and Implications:
Quantitative and Qualitative Comparison
of Main County Offices and South County Offices
Unfortunately, the South County location subsample was too small to allow analyses by
ethnic group. However, some key case characteristics are significantly related to county location.
Specifically, in South County Offices, a higher proportion ofcases are in voluntary services, cases
are shorter in duration, fewer workers are assigned during the course of the case, and a higher
proportion ofcases are in family maintenance at case closure. These findings are also statistically
noteworthy given the small subsample size available from SC. There were also other differences in
case characteristics by county location that did not attain statistical significance but are noteworthy
for further investigation. Children served in the South County Offices appear to have a fewer
number of unique placement homes in the current episode, have a shorter average stay per
placement, a shorter length oftime in out-of-home placement, and are younger at time ofcase
closure.
These findings imply that there is a difference in the style and quality of services provided in
the South County Office location. Qualitative findings are congruent with the quantitative data and
provide some insight into the dynamics that may contribute to this difference. The higher number of
cases in voluntary services, the larger proportion ofcases in family maintenance at case closure,
and the shorter duration ofcases is consistent with what South County study participants described
as a culture ofcommitment to maintaining and reunifying families. This commitment was described
as integral to the philosophy of managers and social workers and reflected in the expectations
communicated to social workers by most supervisors and by peers. The practice ofproviding an
orientation to the child welfare system, time limits, the courts, and how to work with social workers
may also contribute to the differences found in the quantitative analysis. Study participants
frequently described South County as similar to a small county or rural area. This geographic
difference, and the resulting sense of "connection to the community,,, was identified as a/actor in
the greater sense ofshared philosophy and emphasis on prevention ofout-of-home placements.
Other studies suggest that region may influence practice. For example, one study (Drake, 1996) also
found that rural areas are more likely to offer preventive services than are urban areas. The finding
that fewer workers are assigned during the course of the case is also consistent with descriptions of
the vertical case management model. This model calls for social workers to carry the same case
from case opening to case closure. assigned to a case works with that family from the initiation to
closure of the case.
Limitations and Recommendations
Only about 3 % of cases were designated from SC, and the third location designation of
"both" was dropped. A key informant from South County explained that many of the cases
categorized as "both" may have been opened by ER workers from the Main County Office but
carried in South County prior to assignment of South County staff to a specific region of the county.
Similarities in demographics between the South County Office cases and the "both" categories
would appear to be congruent with this hypothesis. However, since this explanation could not be
verified in the data that was available, the cases designated as both were not combined with other
South County data. Future research should include a larger subsample from SC and add methods to
102
understand the reasons behind these location differences. Also, a larger South County Office
location sample may enable more detailed comparisons by ethnicity.
Some.practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County Offices and adapted to
other regions ofSanta Clara County, as recommended by study participants. This practice may be
particularly helpful to low-income families and families of color that may be intimidated by or
unfamiliar with child welfare and related systems. Other practices, such as vertical case
management, may hold promise for other regions ofSanta Clara County, though the success ofthis
practice may be linked to the "small county'' dynamic described by South County study
participants. Some participants in focus groups that took place at the main office mentioned this
model as promising while others asserted that specialized expertise in different areas of social work
was an advantage in providing quality services that are informed by "the most current information."
Further examination of how the culture ofcommitment to maintaining and reunifying families is
created, communicated, and continued and how this might be adapted to other parts of the county
warrants attention.
103
Multivariate Analysis of Fae.tors Predicting Reunification
Preliminary Models
In order to identify system-related factors and case characteristics predictive of reunification
of children with their families, two exploratory models were tested using multivariate analyses. The
first was with a larger sample of 1720 closed cases, and the second with 403 cases from a case
record review sample. Both samples were collected during Phase 2 and additional data was
collected and analyzed during Phase 3.
Results: Closed Case Sample
We chose 8 demographic and system-related variables available through CWS/CMS to
predict reunification: child's gender, ethnicity, age at time of case opening, the number of workers
assigned across the history of the case, the length of the case, the number of unique placement
homes assigned, the number of times removed from the family, and the county location (Main
County Offices versus South County Offices).
For our analysis, due to missing data across our set of variables, 1190 cases out of I 720
were available for logistic regression. Our model was significant in predicting the reunification of a
child with the family (X2 = 176.27, df = 11, p = .001, Cox & Snell R = 0.14). Specifically, Asian
American/Pacific Islanders were less likely than Whites, African Americans, and Latinos to be
reunified, and the fewer number of workers assigned, shorter length of a case, and fewer number of
unique placement homes assigned were related to reunification (while controlling for each variable
within the model). Please see Table 26.
104
. . Reum.fi1cation in Larger Sample Model
T a ble 26 Log1st1c
. . R egress1on Pred1ctmg
Model
Cox & Snell R:
0.14
176.27
11
.001
xl.
df
p
Predictors
B and Significance
Gender (female baseline)
Ethnicity
African American vs White
Latino vs White
Asian vs White
Other vs White
Latino vs African American
Asian vs African American
Other vs African American
Asian vs Latino
Other vs Latino
Other vs Asian
Age at case opening
Worker Number
Time LenITTh of Case
Number of Unique Placement Homes
Number of Removals
County Location (Main vs South)
Using two-tailed tests * p < .05
** p < .01
Model Coding: 1 = reunified, 0 = not reunified
Odds Ratio
0.15
1.16
0.19
0.22
-0.59*
-0.33
0.02
-0.78**
-0.53
-0.81 ***
-0.55
0.26
-0.01
-0.19***
-0.02***
-0.10**
-0.02
0.40
*** p < .001
1.21
1.24
0.55
0.72
1.02
0.46
0.59
0.45
0.58
1.30
0.99
0.82
0.98
0.91
0.98
1.49
Summary of Key Findings and Implications:
Closed Case Sample
Our model composed of primarily demographic and system-related variables to predict
reunification (child's gender, ethnicity, age at time of case opening, the number of workers assigned
across the history of the case, the length of the case, the number of unique placement homes
assigned, the number of times removed from the family, and the county location) was significant.
Specifically, four of these variables predicted reunification: child's ethnicity, number ofworkers
assigned throughout the case, length of the case, and number of unique placement homes.
Asian American/Pacific Islanders were less likely than Whites, African Americans, and
Latinos to be reunified with their families. Given that Asian American/Pacific Islanders are one of
the minority groups least acculturated to mainstream society, many oftheir cases might be
perceived as more severe because ofthe differences between traditional Asian and mainstream
American cultural attitudes toward child rearing and discipline. Culturally, many traditional Asian
105
families have a high level ofparental control and tend to use corporal punishment as part oftheir
parenting (Chao & Tseng, 2002). Many Asian parents may also be unaware about the laws defining
abuse in this society. In these cases, cultural sensitivity by investigators and caseworkers would
help determine if these Asianfamily cases are indeed severe or if they are over-interpreted as such
because ofa philosophical difference between the professional and the family in parenting and
discipline. Future empirical studies could help determine the extent to which abuse severity or
cultural differences explain this relatively low reunification rate among Asian American families.
However, as Phase 2 results indicated Asian American/Pacific Islanders are also the most likely to
be placed in Voluntary Family Maintenance (Hines et al., 2002), the CWS appears to value the
preservation of Asian American/Pacific Islander families when first referred to the system (perhaps
recognizing the challenges of culturally appropriate services, language issues, and other
circumstances related to the Asian American/Pacific Islander community) but once in the system,
Asian American/Pacific Islander family cases are not easily resolved and circumstances may persist
that place the child at risk for further abuse ofneglect.
The fewer number ofworkers assigned was also related to reunification. This finding
indicates consistency in service and the relationship between the social worker and the family may
improve the chances ofreunification. Perhaps this is due to greater familiarity and thus more
effective advocacy for the child and family. This finding could also be a function of particularly
challenging family cases already less likely to be reunified needing a greater set of workers over
time (e.g., due to worker burnout in the case, or specialization needed across the case).
Shorter case length was also predictive of reunification. Presumably, less severe cases
could be resolved in a timely fashion, usually concluding with reunification. Following a similar
logic above regarding the number of workers, a longer case may imply more problems, thus
involving more time for resolution, but also possibly a case already protracted into a situation where
reunification is unhealthy for the child.
The fewer number of unique placement homes assigned was related to reunification.
Change in placements may be due to systemic conditions (i.e. logistics and short term availability of
space) but multiple placements can also indicate a persistent problem with a child adapting to a
placement and accepting care offered. As this may again indicate a more severe case, reunification
could also become less likely.
Due to missing data across our set of variables, 1190 cases out of 1720 were available for
this multivariate analysis. More complete information would have improved confidence in our
findings. Also, as we have made references to the possibility of severity ofa case influencing
reunification, better measurement and a more comprehensive evaluation offamily history, abuse
and neglect circumstances, and other indicators could produce a stronger predictive model. In the
future, we may be able to include more psychosocial and case variables in a larger sample to build
upon our current findings.
Results: Case Record Review Sample
To identify better the influence of case characteristics and indicators of success (discussed in
the Successful Cases Analysis, beginning on page 48) we added 6 variables to the model with the
larger sample collected through our in-depth case record reviews and utilized the additional
106
information coded for Phase 3: a family's enhanced capacity to provide for their children's needs,
children receiving appropriate educational services, children receiving adequate services to meet
physical and mental health needs, number of referrals, number of previous times in the CWS, and
assignment to Family Maintenance or Family Reunification services. The location variable was not
included in this model as only one case coded from the South County Office area.
For our analysis, again due to missing data across our set of variables, 246 out of 403 cases
were available for logistic regression. Our expanded model was also significant in predicting the
reunification of a child with the family (X2 = 63.87, df= 16, p = .001, Cox & Snell R2 = 0.23).
Similar to the larger sample's model, Asian American/Pacific Islanders were still less likely than
Whites, African Americans, and Latinos to be reunified. Also, shorter time length of a case again
predicted reunification. However, in this expanded model with success indicators, the number of
workers assigned and number of unique placement homes assigned were not statistically related to
reunification. The child's age at time of case opening and the riumber of unique placement homes
did approach significance. Interestingly, the variables related to success cases and outcomes
( discussed in the Successful Cases Analysis, beginning on page 48) were not significant. Please see
Table 27.
107
. Reum.fi1caf10n .m C ase R ecord Rev1ew
. Sampe
Table 27 Log1st1c
. . R egress1on Pre d.1ctmg
1 Model
Model
Cox & Snell R'-
0.23
63.87
16
.001
xi.
df
p
Predictors
B and Significance
Gender (female baseline)
Ethnicity
African American vs White
Latino vs White
Asian vs White
Other vs White
Latino vs African American
Asian vs African American
Other vs African American
Asian vs Latino
Other vs Latino
Other vs Asian
Age at case opening
Worker Number
Time Length of Case
Number of Unique Placement Homes
Number of Removals
Number of Referrals
Previous Times in the CWS
Case Assignment (FM vs FR)
Child Receiving Adequate Educational
Services
Child Receiving Adequate Services to
Meet Physical and Mental Health Needs
Family's Enhanced Capacity to Provide
for Child's Needs
Using two-tailed tests* p < .05
** p < .01
+ approached significance at p < . l 0
Model Coding: I = reunified, 0 = not reunified
Odds Ratio
0.29
1.34
0.43
-0.19
-1.63*
-0.36
-0.61
-2.06**
-0.79
-1.44*
-0.17
1.27
-0.06+
-0.06
-0.03***
-0.ll..
-0.05
-0.04
0.12
-0.50
0.10
1.21
1.53
0.83
0.20
0.54
0.13
0.46
0.24
0.84
3.56
0.95
0.95
0.98
0.88
1.05
0.96
1.13
0.61
1.10
-0.18
0.84
-0.13
0.88
*** p < .001
108
Summary of Key Findings and Implications:
Case Record Review Sa~ple
These findings indicate that particular factors are important to consider when predicting
reunification, specifically ethnicity and the length oftime a case remains open. Similar to the model
in the larger closed case sample, this implies that more severe cases (those needing more time for
resolution) may be more problematic, thus reducing the chances ofreunification. We saw in the
larger sample model that number of workers and unique placement homes were related to
reunification, but the number of workers was no longer significant and number of placement homes
approached significance. In addition, the younger the child at case opening the more likely
reunification would occur. This could be that an effort is made to reunify younger children with
their families given that they are in an especially crucial stage of psychological and emotional
development related to the attachment with parents. These variables should be explored in more
detail.
The finding again that Asian American/Pacific Islanders are less likely to be reunified than
Whites, African Americans and Latinos, highlights a main ethnic difference. In Phase 2 we
discovered that Asian American/Pacific Islanders were more likely than the other ethnic groups to
be enrolled in Voluntary Family Maintenance services (Hines et al., 2002). However, our results
here imply that when Asian American/Pacific Islander children are removed from the home, their
children are less likely to be reunified. This may also indicate extreme outcomes where Asian
American/Pacific Islander children are either initially diverted from the system or once in the
mainstream ofthe system are less likely to be reunited with their families.
It was interesting that the successful case and outcome variables were not predictive of
reunification. This could be due to the significant impact of other variables overshadowing the
influence of success indicators. It could also be due to our use of reunification as a dependent
variable. As we discussed in the Successful Cases Analysis (see page 48) reunification is not
necessarily synonymous with success, and thus these variables may indeed be unrelated.
Again, due to missing data across our set of variables, 246 cases out of 403 were available
for this multivariate analysis. Although only a preliminary set of models, these findings are useful
for identifying areas in the CWS and family circumstances that can be studied further, especially in
conjunction with the specific goal of reunifying children with their families. For future research,
reasons as to why major system-related factors are not predictive would also be valuable.
109
The Role of the Court System in Child Welfare Practice
Few studies have examined the potential influence of the court system on child welfare
practices and outcomes, however as noted in the literature review (Section III), the court system is
likely to have a significant impact on the trajectory of child welfare cases. In order to examine this
issue further, this analysis examines the role of the court system in child welfare practice.
Specifically, two overall issues are addressed. First, an analysis of quantitative case record review
data, and qualitative data from one focus group conducted with the Court Officer's Unit will
describe court ordered changes to social worker recommendations at the initial
jurisdictional/dispositional hearing; the types of changes the court orders; if these changes differ by
child's ethnicity; and possible explanations for these changes. Secondly, a qualitative analysis from
the agency-wide focus groups (described on page 61) will focus on themes related to the
relationship between the child welfare and court systems.
Quantitative and Qualitative Methods:
Analysis of Court-Ordered Changes
Quantitative Analysis
- The quantitative analysis of the role of the court system in child welfare practice included a
new case record review of the same sample of 403 closed child welfare cases that were included in
Phase 2 of the research.
Sampling
Case record reviews include the same sample of 403 closed child welfare cases that was
analyzed in Phase 2 of the research. This sample was created from 6761 total case closures over an
18-month period, and a data file containing 1753 unique cases representing one child per family and
one case opening was constructed. This data file was used to obtain our target sample of 403 cases
selected randomly, guided by stratification according to ethnicity, age and service type. Of the 403
cases reviewed for this court analysis, 12 cases were closed, but not reviewed because they were
with a worker and not at the retention center; and 13 were not reviewed because they had been reopened.
Data Collection
Data collection took place at the Santa Clara County's Social Services Record Retention
Center. A Senior Research Assistant worked with the Retention Center Manager to ensure that the
sample of case records from Phase 2 of the research were re-pulled for review. Cases were reviewed
by three research assistants, two of these research assistants participated in the Phase 2 case record
review data collection and so had already been trained in case record review methods and one
research assistant was newly trained to complete the case record reviews.
Instrumentation
A data extraction form was created and pilot tested (please see Attachment 5 for a copy of
the Court Cas~ Record Data Extraction Form). The form contained questions related to court
110
ordered service or visitation changes for the child, mother, father, alternate caregiver, or the whole
case.
Analysis
We addressed this aim through a 3-part series of exploratory descriptive and bivariate
analyses. First, we identified whether there were any court ordered changes and if court ordered
changes in general were related to ethnicity. A single change in recommended services for a case,
regardless of magnitude, would classify the case as one with a "court ordered change."
Second, we examined major changes and if these were related to ethnicity. We considered
any court ordered change in the three main service areas (those concerning bypass, family
maintenance, and family reunification) as a major change.
Third, we developed 4 summary scores consisting of the number of changes imposed
separately concerning the child, mother, and father, and then for the entire case. Across each case
there were 85 variables over 9 potential areas where a change could be ordered in addition to the
three main service areas. These areas were psychological evaluation, counseling services, health
services, drug/alcohol testing, substance abuse/alcohol treatment, parenting education, support
groups, domestic violence treatment, and visitation. Changes in these areas were assessed and
summed into the 4 scores for court ordered changes aforementioned: child changes, mother
changes, father changes, and total changes. These scores were also examined for ethnic differences.
Definition of Terms
Court Ordered Change - Major changes were defined as court orders to initiate bypass,
family maintenance, or family reunification services for the mother, father, and/or the family. The
remaining changes included those across 9 areas represented by 85 variables: psychological
evaluation, counseling services, health services, drug/alcohol testing, substance abuse/alcohol
treatment, parenting education, support groups, domestic violence treatment, and visitation. A court
ordered change could also include placing more restrictions or conditions on service
recommendations.
Qualitative Analysis
A separate focus group with representatives from the Court Office Unit was conducted
toward the end of the study in order to obtain feedback on the implications and meanings of the
quantitative findings. The focus group with the Court Office Unit contained 9 participants,
including two Social Workers (22.2%) Ill's, five social work supervisors (55.6%), one Program
Manager (I 1. 1%) and one Legal Clerk (1 1. 1%). The mean number of years working in child
welfare among the Court Officer's focus group was 11.6 years and ranged from 4 to 22. Focus
group participants had a mean of 3.2 years in their current positions, with a range of .5 to 7 years.
Seven participants had MS W's (77.8%), one had a BSW (11.1 %), and one reported unspecified
postgraduate education (I 1.1 %). Two of the participants identified as Latino (22.2%), two identified
as Asian American/Pacific Islander (22.2%), four identified as White (44.4%), and one identified as
mixed race (11. I%). Eight of the participants were female (88.9%) and one was male (11.1 %).
111
Participants in the Court Officer Unit focus group were asked a series of questions designed
to provide insights into the "story behind the statistics." Specifically, Court Officers were invited to
share their experiences in relation to overall factors that contribute to court ordered changes and
comment on specific findings from quantitative data. Please see Attachment 6 for the specific
questions used to guide the-Court Officer Unit focus group discussion.
Notes from the Court Officer Unit focus group were transcribed and analyzed for key
themes related to court changes. In addition, focus group respondent interpretations and insights
about results of the quantitative data were examined and summarized.
Quantitative Results:
Analysis of Court Ordered Changes
Frequency of Court Ordered Changes
Of the 403 cases, 178 (44.2%) recorded at least one court ordered change from the original
social worker recommendations from the j urisdictionaVd_ispositional hearing, 200 (49 .6%) abided
by the original recommendations, and 25 (6.2%) were not available for evaluating the additional
court ordered changes.
Court Ordered Changes and Child's Ethnicity
We eliminated the cases with missing information to do our comparison by ethnicity,
leaving us with 378 valid cases. There was no significant relationship between the existence of a
court ordered change and ethnicity. Please see Table 28.
t me Groupa
Table 28 Proport1ons ofC ourt 0 rderedChanges b1y Eh.
White
Latino
Total
African
American
Court Change
8
Asian
American/PI
Other
No
200
(52.9%)
25
(53.2%)
65
(51.6%)
68
(51.9%)
18
(62.1%)
24
(53.3%)
Yes
178
(47.1%)
22
(46.8%)
61
(48.4%)
63
(48.1%)
11
(37.9%)
21
(46.7%)
Based on 378 cases with valid infonnation
Types of Service Changes
We examined the types of court changes and distinguished three to be relatively major given
their impact on the case: bypass, family maintenance, and family reunification. Of the 178 cases,
the most common change with these three types was the implementation of family reunification
services: 27 cases (152%) involved family reunification, 2 (1.1 %) cases regarded bypass, and 1
case (0.6%) regarded family maintenance. Ethnic comparisons were conducted, but no relation~hips
were found with these major services. Please see Table 29.
112
Table 29 Bypass, Family Maintenance, and Family Reunification Court Changes by Ethnic Group8
Total
African
White
Latino
Asian
Other
American
American/PI
Bypass
No
Yes
11
20
(95.2%)
0
1
(4.8%)
176
(98.9%)
22
(100.0%)
61
(100.0%)
62
(98.4%)
(100.0%)
2
0
(0.0%)
0
(0.0%)
1
(1.6%)
(0.0%)
(1.1%)
Family
Maintenance
(99.4%)
(100.0%)
22
60
(98.4%)
63
(100.0%)
(100.0%)
11
21
(100.0%)
1
(0.6%)
0
(0.0%)
1
(1.6%)
0
(0.0%)
0
(0.0%)
0
(0.0%)
No
151
(84.8%)
16
(72.7%)
55
(90.2%)
52
(82.5%)
10
(90.9%)
18
(85.7%)
Yes
27
(15.2%)
6
(27.3%)
6
(9.8%)
11
(17.5%)
1
(9.1%)
3
(14.3%)
No
Yes
177
Family
Reunification
a Based on 178 cases with court change information
Court Changes Regarding Children, Mothers, and Father
Four scores were calculated to describe the number of changes imposed by the court, each
for the child, mother, and father, and finally a total score representing the entire case. The number
of changes related to the child's circumstances ranged from Oto 4 with an average of 0.31 (sd =
0.65). Court ordered changes related to the mother were most common, ranging from Oto 8
changes with an average of 0. 77 (sd = 1.28). The number of changes related to the father's
circumstances ranged from Oto 7 with an average of0.67 (sd = 1.26). The number of total court
ordered changes related to a case ranged from 1 to 16 with an average of2.53 (sd = 2.22). No
significant ethnic differences were found regarding the number of court-imposed changes. Please
see Table 30.
113
Table 30: Number of Court Chan9:es by Ethnicity'a
Total
African
White
American
Child Related
Changes
Avg.
0.31
0.23
0.38
(sd)
(0.65)
(0.53)
(0.78)
Mother
Related
Changes
Avg.
(sd)
Latino
Asian
American/PI
Other
0.27
(0.54)
0.36
(0.50)
0.33
(0.73)
0.77
(1.28)
1.32
(l.94)
0.69
(0.99)
0.83
(1.30)
0.18
(0.40)
0.57
(1.29)
Father Related
Changes
Avg.
(sd)
0.67
(1.26)
0.82
(1.92)
0.61
(1.14)
0.62
(1.10)
0.55
(1.21)
0.90
(1.30)
Total Case
Changes
Avg.
(sd)
2.53
(2.22)
3.36
(3.36)
2.39
(1.63)
2.54
(2.39)
1.82
(1.33)
2.43
(2.01)
a Based
on 178 cases with court change information
Qualitative Results:
Analysis of Court Ordered Changes
Quantitative findings were presented to a focus group made up of9 representatives of the
Agency's Court Officers Unit for discussion and feedback. Quantitative results suggested that
approximately half of the cases had indicated a court ordered change from the recommendations of
the social worker describedfrom the jurisdictional/dispositional hearing. From our focus group
with members from the Court Officer Unit, the consensus was that the interpretation of the
measured proportion of court ordered changes found in the quantitative analysis, depended on what
we were actually measuring. The proportion is ·dependent on a variety of circumstances as to why
there was a change from the social worker recommendations including clerical errors needing
correction, changes due to modifications in the law, simple (albeit common) adjustments to services
such as visitation terms in addition to actual substantial changes due to conflicting opinions
concerning the direction of the case could all be related to a change between social worker
recommendations and court ordered.
Findings from the focus group with the Court Officer Unit provide additional information
regarding changes made to social worker recommendations at the jurisdictional/dispositional
hearing. First, capturing the details needed to conclude why social worker recommendations may
or may not be followed is extremely difficult. This is because variations by judge, courtroom
114
environment, social worker credibility and reputation, assigned lawyer (i.e., court appointed vs.
privately retained), as well as formal and informal case plan negotiation tactics frequently occurring
would complicate research. Second, the personal dynamics among the stakeholders (socials
workers, lawyers, judges, and families), including the strength of their positions and philosophical
standpoint can be such an influence that systemic conditions may be poor predictors of case
outcomes and success as they are related to the judicial process. Third, that despite the complex
circumstances involved, judges, guided by social workers, determine the best possible service plan,
while being aware of interpersonal dynamics and individual characteristics that may obscure
reasonable decisions. According to the focus group, social workers have a significant influence in
the court process and usually receive a majority of services they request for their clients.
Quantitative data indicated that few changes involved bypass, family maintenance, or family
reunification. The proportion of changes in these main service areas seemed reasonable to the focus
group. The low percentage of change in bypass, FM, and FR assignment should be low given the
legal criteria to qualify for such a change, as well as the intricate legal procedures.
When presented with quantitative findings that indicated no relationship between ethnicity
and court ordered changes in major service areas, or the overall number of court-ordered changes,
it was the opinion ofmembers of the focus group that ethnicity does not contribute to decisions in
court, but rather it is circumstances that drive the case.
Qualitative Methods:
Relationship Between the Child Welfare and Court Systems
Qualitative results describing the relationship between the child welfare and court systems
are gleaned from the agency-wide focus groups. These focus groups were structured to explore
practices at different choice points in the child welfare system. Please see page 61 for a full
description of agency-wide focus group methods and results.
Qualitative Results:
Relationship Between the Child Welfare and Court Systems
Results from agency-wide focus groups (please see page 61 for a full description) suggest
that the relationship between DFCS workers and district attorneys andjudges is problematic at
times. Many participants described difficulties in effective collaboration between the court system
and the child welfare system. Social workers suggested that problems between the court system and
the child welfare system may reflect fundamentally different perspectives on the needs and
circumstances of children and families in the CWS, and perhaps unrealistic expectations regarding
service plans. For instance, one social worker noted, "From a client's side, they are being told by a
court system and a social system that the way that they were brought up and the way that things
were done in their [clients'] family is wrong and not acceptable," noted a social worker focus group
participant. Several participants suggested that courts directly and indirectly set an unrealistic
standard for families, exemplified by one participant's comments:
"The court system just makes these orders and that's it. They don't consider anything else
except for this narrow focus on child and parent. The social worker has to look at resources
for the child, for the family, and language barriers. The court doesn't consider the impact it
115
has on the family, who has to pay for the services, shrinking or expanding resources. The
court is in an isolated bubble anything outside that bubble is not in their framework "
Regarding the services ordered, another social worker commented, the "Court wants it all
done at once ...the court doesn't understand that for whatever reason the client may be only capable
of doing so much within the first 6 months." Additionally once the court does order a service or
case plan, social workers report that it is difficult to adjust mandated services "because the court
requires justification for removing the order; also it is difficult convincing the DA of [the need for]
new recommendations."
In general, focus group and interview participants expressed a need to bridge these gaps
between the court system and the child welfare system. Although some social workers described
having strong collaborative relationships with the court system, many described tensions with
attorneys and other court representatives in relation to recommendations for mandated services and
disposition of cases. Social workers also conveyed that it was common for court representatives to
dismiss their psychosocial family assessments and push for recommendations that the social
workers deemed disconnected from the reality and needs of families. "Legal aids rarely visit the
children or the parents, they need to be more reality-based," commented on worker. "Social workers
and the court system are both looking at the case from different views," noted another. Some
workers depicted the district attorneys as having more of a focus on "winning" a case than on
ensuring the best outcomes for the children and family.
Working with the DA is a challenge. Their role is to advocate for the child-but they often
do only one interview and they don't know the particulars ofthe case. There can be
antagonism between the social worker and the DA-there should be less antagonism
between the DA and the social worker and more cooperation.
Social workers also expressed concern about the fairness ofthe system. For example, "In court
cases, private attorneys are more likely to have their client's case dismissed compared to court
appointed attorneys." Many participants conveyed that many court representatives and social
workers exhibit a "gap between standards that common people have and that educated professional
people have," which may disadvantage low-income families and families of color. "Attorneys.look
at the information provided by the social worker and make sweeping generalizations and judgments
about families," summarized one social worker. One supervisor captured concerns about failing to
recognize the context of client's lives:
When there is child abuse or neglect we raise our eyebrows to it, we go in there and want
everything to change but we 're not providing a basicfoundationfor change. We 're not
developing the community, we 're not providing better housing. We tell people to go out and
get suitable housing, but we 're not providing it.
Language issues also impact the experience offamilies in the court. "Families are unable to
communicate with their lawyers because of language barriers; English only attorneys can't
communicate with a monolingual Spanish speaking client," summarized one worker. Other workers
observed that "Families may not like to leave a message for an attorney because of the language
factor" and "the court process can be confusing for English speaking families, but more so to
Spanish speaking families." Furthermore, "the court reports are all in English and the clients don't
have a translated version that they can look at."
116
Additionally, social workers expressed concern that client outcomes are often dependent on
working relationships or attitudes ofDA 's. "There are some DA's that I have a great rapport
with-but a few are very judgmental and I feel like they are not supportive of the family." The
DA's are perceived as having substantial power. "The DA's word is very strong in court because
they are supposed to be speaking on behalf of the child," stated one social worker. Other social
workers suggested that some DA's pressured social workers to change recommendations or would
attempt to discredit them in the courtroom, particularly when the DA's case was weak.
Agency-wide focus group results also provided recommendations regarding ways to enhance
collaborative relationships between the court system and the child welfare system. These
recommendations included,
•
•
•
•
•
Explore opportunities to develop standards for practice between the courts and the child welfare
system including development of shared policies, practices and procedures. Such efforts could
help address common concerns of social workers that generally called for a "switch from courtcentered practice to family/community-centered practice."
Develop a systems level mechanism for problem solving to address conflicts between social
workers and DA's and "set a minimum standard or tone for what is appropriate behavior in the
court."
Facilitate development of role clarity between child welfare and the courts to address
widespread concern that decision-making is often not based on the social worker assessments
and, instead, is often based on perceptions ofDA's who may not have spent sufficient time with
children and who may not have evaluated family strengths.
Facilitate a dialog with both court and child welfare workers about standards for "good enough"
parenting and the dynamics of poverty in decision-making.
Explore opportunities for support of social workers in the court process through training,
enhanced supervision, or coaching from experienced colleagues.
Summary of Key Findings and Implications:
The Role of the Court System in Child Welfare Practice
In Phase 3, we examined the role of the court system in child welfare practice by exploring
the types of court ordered changes that are made to social worker recommendations at the initial
jurisdictional/dispositional hearing, as well as the relationship between the child welfare and court
systems. Quantitative results indicated that the court made changes to the initial social worker
recommendations at the jurisdictional/dispositional hearing in approximately halfof the cases.
Ethnicity was not related to whether or not a change occurred. These results do not necessarily
indicate that the system and courts treat children and families from different ethnicities the same,
but rather that when the judge decides to make a change from the social worker recommendations,
these changes appear to occur equally across ethnic groups. It was the opinion of members of the
Court Officer Unit focus group that ethnicity does not contribute to decisions in court, but rather it
is circumstances that drive the case.
Findings from the focus group with the Court Officer Unit suggest that many factors
complicate an accurate measurement ofwhy social worker recommendations may or may not be
117
followed. Factors that may impede accurate assessment of the agreement between child welfare and
court personnel on service plan recommendations include the strong influence of individual
stakeholders involved, and formal and informal case plan negotiation tactics all of which paint a
much more complicated picture.
Qualitative findings from the agency-wide focus groups suggested that collaboration
between the child welfare and court systems is problematic. Agency-wide focus group participants
felt that the child welfare and court system have different perspectives on the needs and
circumstances of children and families in the CWS and that the court system may have unrealistic
expectations of families.
Additionally, many agency-wide focus group participants described the ways in which child
welfare and district attorneys and judges interact as ineffective. Some social workers felt that some
district attorneys and judges would dismiss their assessments and recommendations and may try to
pressure social workers to change their recommendations. A somewhat similar finding was noted by
Knepper and Barton, (1997) who found that although judges tended to accept social worker
recommendations, the relationship between social workers and the court system plays a significant
role in their decisions. Their study found that when social workers adhered to the group norms of
the court, judges rewarded them by not ordering "unrealistic" practices and allowing reasonable
time frames for mental health evaluations to be completed. Similarly, agency-wide focus group
findings seem to suggest that if social workers go against the implicit rules ofthe court system, .then
judges may override their recommendations, but ifthey adhere to expectations ofthe court system
than recommendations are accepted.
Limitations and Recommendations
From these exploratory findings, issues arose regarding limitations and recommendations.
First, although quantitative and qualitative information described many of the circumstances behind
why social worker recommendations are not upheld by the court, it is very difficult to ascertain the
degree to which each circumstance influences a court ordered change. Second, better and more
exhaustive record keeping may not be the solution, as many informal and undocumented actions in
the negotiation process occur regularly, and are confidential, thus making much of the data
unavailable for research and evaluation purposes. Third, assessing other factors such as courtroom
environment, stakeholder characteristics, and other judicial issues would be advantageous in later
research. Although obtaining this information is a formidable task, these factors should be included
in the future, as measuring primary systemic factors no longer seem adequate to explain outcomes
in the CWS.
118
V.
STATEWIDE COMPARATIVE ANALYSIS
The statewide comparative analysis of effective practices for children and families of color
in the child welfare system included a statewide survey of child welfare directors, managers or
supervisors in California counties who were asked to describe effective practices for children and
families of color in their child welfare systems. In addition, this statewide comparative analysis,
included a quantitative analysis that identified the link between these practices and levels of
disproportionate representation of children of color in county child welfare systems. As such, two
overall analyses were included in the statewide comparative analysis of effective practices for
children and families of color in the child welfare system, 1) descriptive information from the
statewide survey on the types of practices, practices and practice characteristics identified by survey
respondents, and 2) multivariate results that link types of practices, practices and practice
characteristics to over or underrepresentation of children of color within counties. Additionally,
contextual information on the estimated proportions of children of color in county populations and
county child welfare systems within selected regions of California is provided, as well as
information on the estimated proportion of children within the overall county population who are in
the CWS.
Methods
The statewide survey on effective practices for children and families of color in the child
welfare system was designed to gather information on 1) key practices in California counties that
may have an impact on children and families of color, and 2) practices considered most promising
for children and families of color. For the purposes of the survey, "practices" were defined broadly
to include programs, services, strategies, policies, or tools.
Data Collection
Child welfare directors, managers or supervisors were mailed an invitation to participate in
the survey, and a copy of the survey. Two weeks later, a follow-up email that contained the survey
as an attachment and telephone calls were made to encourage participation. Counties that failed to
respond to the survey received approximately 3 additional follow-up emails or telephone calls
requesting their participation or a referral to someone else in the agency who might be able to
participate.
Respondents were offered four options for returning the survey to the CWRT, either by 1)
completing the survey on their computer and emailing it as an attachment, 2) completing the survey
over the telephone with a member of the CWRT, 3) faxing the completed survey, or 4) mailing the
completed survey. Surveys that were completed over the telephone were tape recorded in order to
ensure the accuracy of information and once information was verified, the tapes were erased. Fortyfive counties responded to the survey, resulting in a 77.6% response rate. Of the 45 counties who
completed the survey, 23 returned it via email, 10 by telephone, 7 by fax, and 5 returned the survey ·
by mail.
119
Sample
The sample included child welfare directors, managers or supervisors from California
counties. Participants for the statewide survey were identified through the 2003 County Welfare and
Social Service Directory. An invitation to participate and a copy of the survey were mailed to each
child welfare director, manager or supervisor in each of the 58 counties in California. Of the 13
counties that declined to participate, 4 indicated that their child welfare population is not diverse
and they did not have any special programs for children and families of color; 3 replied that the
survey was too time consuming, 3 did not reply to the survey, 2 indicated that they would complete
the survey, but did not return it and one county indicated that they were goirig through a reorganization and could not complete the survey because of these changes.
Instrumentation
A survey entitled "Statewide Survey on Effective Practices for Children and Families of
Color in the Child Welfare System" was developed and tested in a pilot study involving five
counties (please see Attachment 7 for a copy of the Statewide Survey). Feedback provided from the
pilot counties resulted in a reduction in the number of questions and minor changes in wording. The
survey contains questions related to caseload characteristics of counties and the most effective
practices for children and families of color in each county's child welfare system. The survey asks
respondents to provide demographic information on children in various components of the county's
child welfare system. Respondents are also asked to identify up to five of the most effective
practices for children and families of color in their county's child welfare system, ranking them
from #1 (most effective), #2 (second most effective) and so on. Multiple choice questions on the
most effective practice (#1) are also included in the survey.
Analysis
Quantitative analyses were conducted using the Statistical Package for the Social Sciences
(SPSS) software. Descriptive results were generated by running frequencies and percentages for
variables of interest. Multivariate results were generated by conducting controlled tests of empirical
associations between types of practices, practices, and practice characteristics and levels of
disproportionate representation of children of color in county CWS's.
Qualitative data describing the most effective practices were analyzed through content
analysis methods. The types of practices within each of the four categories were identified and
descriptive information about their target population, description and reasons why_these practices
are effective with children and families of color were delineated.
Contextual Tables:
County Child Populations and CWS Involvement·
The following tables provide contextual information on selected county child populations
and CWS involvement by ethnic group. Please note that the counties presented in these tables are
not necessarily the counties who responded to the statewide survey. Also please note that because
the most recent information available on county child populations is for the year 2001, information
on CWS populations for the year 2001 is also provided. As such, current 2003 proportions of
120
children of color in California county CWS may vary slightly from the 2001 proportions provided
in the following tables.
This descriptive information is presented in order to provide a context through which to
view the results from the statewide survey. Because compiling information on all 58 California
counties would be cumbersome, information on selected counties is provided. Table 31 a provides
information on the estimated proportions of children of color in county populations and the CWS
within 13 Bay Area counties. Table 31 b described estimated proportions of children of color in
county populations and the CWS within selected counties by California regions. Last, Table 31 c
provides information on the estimated proportion of children within the population in the CWS
·
among selected counties by region.
121
I . Popu laf10n and Ch"ld
We lfiare S1ysem
t
W"th·
T able 31 a: E sf1mat ed Propo rf10ns o fCh"ld
I
I m Bay A rea C ounf1es
I reno fC oorm
Ethnic Group
County
White
African American
Latino
Asian
Native American
%in
%in
%in
%in
%in
%in
%in
%in
%in
%in
Population
cws Population
cws
Population
cws
Population
cws
Population
cws
Alameda
33.7%
18.0%
19.0%
66.1%
24.6%
11.9%
22.3%
2.9%
0.4%
0.7%
54.1%
38.5%
11.4%
Contra Costa
45.2%
20.2%
12.7%
13.9%
2.0%
0.4%
0.5%
Marin
71.3%
35.7%
3.2%
19.4%
18.7%
9.4%
6.6%
3.3%
0.2%
0.3%
Mendocino
67.3%
62.1%
0.7%
4.9%
24.7%
18.5%
1.4%
0.2%
6.0%
12.9%
Monterey
30.1%
4.3%
14.2%
52.3%
5.8%
32.0%
57.6%
1.7%
0.4%
0.2%
Napa
67.4%
1.5%
11.8%
61.2%
32.3%
17.4%
4.3%
2.8%
0.6%
0.0%
San Benito
43.9%
32.1%
0.5%
1.2%
53.1%
60.7%
2.1%
1.2%
0.4%
1.2%
San Francisco
27.6%
11.6%
12.9%
70.4%
22.2%
12.3%
37.0%
4.8%
0.2%
0.8%
San Mateo
38.5%
26.7%
4.3%
29.8%
33.0%
32.0%
23.9%
9.4%
0.3%
0.1%
Santa Clara
35.4%
27.4%
3.5%
12.2%
32.8%
52.0%
28.1%
7.4%
0.2%
0.8%
Santa Cruz
53.8%
46.1%
1.1%
3.4%
41.0%
49.5%
3.8%
0.2%
0.3%
0.2%
41.2%
42.4%
49.5%
15.0%
18.9%
9.1%
16.0%
Solano
3.7%
0.6%
0.6%
Sonoma
70.6%
57.4%
2.0%
9.0%
21.8%
28.8%
4.5%
1.0%
1.0%
3.2%
State of
California
37.7%
30.3%
7.1%
31.3%
42.9%
33.8%
11.7%
2.8%
0.5%
1.1%
Notes:
I. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in January 2001 according to Needell, B.• Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A.,
Lery, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. & Kim, H. (2003). Child Welfare Services Reports for California. Retrieved (8/29/2003), from University of California at Berkeley Center for Social
Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>.
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org.
3. The Bay Area Counties are the 13 identified by the CWDA, and are also categorized as "big" counties in California except San Benito which is "small."
.
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California County Data Book summarized information on children aged 0-17
whereas the CWS data may include children as old as 19. Some totals may not equal l 00% due to missing data or rounding.
122
• and Ch.Id
.
. Popu Iat1on
.
Ta ble 31b Est1mate
dP roport1ons o fCh.ld
1 We lfiare S,ystem w·Ith.m Se lectedC ount1es B,y Reg1on
1 reno fC o lorm
Ethnic Group
County
White
African American
Latino
Asian
Native American
%in
%in
%in
%in
%in
% in
%in
%in
% inCWS
% in
Population
cws Population cws Population cws Population cws Population
Bay Area
33.7%
18.0%
19.0%
24.6%
11.9%
22.3%
2.9%
0.4%
Alameda
66.1%
0.7%
27.6%
12.9%
22.2%
12.3%
37.0%
4.8%
0.2%
0.8%
San Francisco
11.6%
70.4%
52.0%
28.1%
Santa Clara
35.4%
27.4%
3.5%
12.2%
32.8%
7.4%
0.2%
0.8%
Mountain
0.8%
10.1%
4.7%
0.8%
0.0%
86.3%
91.4%
0.4%
2.3%
Calaveras5
0.8%
3.6%
12.0%
12.0%
3.2%
1.8%
Placer
83.3%
79.0%
0.7%
0.8%
2.8%
17.9%
17.1%
14.9%
3.1%
Sacramento
53.7%
43.4%
12.5%
35.2%
1.0%
1.0%
Northern
8.9%
13.8%
7.9%
7.5%
1.3%
1.7%
6.8%
Butte
75.1%
73.9%
1.8%
78.4%
1.0%
2.3%
8.0%
5.1%
4.3%
0.3%
8.2%
Humboldt
65.6%
22.4%
79.3%
1.1%
1.2%
15.0%
26.8%
1.1%
0.0%
3.5%
61.0%
8.5%
Lassen5
Southern
10.8%
2.9%
0.2%
60.1%
38.9%
0.4%
Los Angeles
19.8%
15.2%
9.0%
42.3%
Oran,ie
40.9%
39.1%
1.8%
43.3%
48.0%
13.7%
0.2%
6.8%
5.3%
0.2%
35.0%
San Die,io
46.0%
35.1%
7.1%
24.5%
36.6%
9.9%
3.0%
0.5%
2.2%
42.2%
4.8%
42.2%
41.5%
2.4%
15.1%
50.2%
1.0%
0.4%
0.2%
Santa Barbara
Valley
31.1%
28.0%
5.7%
22.3%
48.5%
45.9%
13.8%
2.1%
0.8%
1.7%
Fresno
44.9%
44.0%
15.0%
35.9%
35.0%
3.9%
13.9%
6.4%
0.4%
0.1%
Merced
0.8%
0.5%
53.4%
50.3%
2.3%
13.0%
35.1%
32.4%
8.4%
2.6%
Stanislaus
State of California
37.7%
30.3%
7.1%
31.3%
42.9%
33.8%
11.7%
2.8%
0.5%
1.1%
Notes:
I. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in January 2001 according to Needell, B., Webster, D., Cuccaro-Alamin, S., Annijo, M., Lee, S., Brookhart, A.,
Lery, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. & Kim, H. (2003). Child Welfare Services Reports for California. Retrieved [8/29/2003), from University of California at Berkeley Center for Social
Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>.
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org.
3. Superscript "s" denotes a CWDA categorized "small" county as opposed to "big" county.
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California County Data Book summarized infonnation on children aged 0-17
whereas the CWS data might include children as old as 19. Some totals may not equal I00% due to missing data or rounding.
123
Table 31c:
Estimated Proportion of Children Within the Population in the Child Welfare System Among
.
Selecte d Count.1es B R eg1on
Region and County
Number of Children
Percentage of Child
inCWS
Pooulation in the CWS
Bay Area
San Mateo
875
0.47%
Monterey
641
0.52%
San Benitos
84
0.54%
Napa
178
0.60%
Marin
361
0.70%
Solano
821
0.71%
Sonoma
941
0.83%
Santa Cruz
588
0.88%
Santa Clara
4,401
0.94%
Contra Costa
3,061
1.29%
Alameda
6,033
1.54%
San Francisco
3,050
1.97%
Mendocino
628
2.77%
Mountain
Placer
843
1.28%
Calaverass
128
1.33%
Sacramento
8,436
2.45%
Northern
Lassens
82
1.11%
Humboldt
1.27%
389
Butte
827
1.62%
Southern
Santa Barbara
516
0.48%
Orange
7,316
0.89%
9,935
1.19%
San Die20
Los Ane;eles
52,630
1.80%
Valley
Stanislaus
1280
0.89%
Merced
822
1.11%
Fresno
4757
1.80%
State of California
139,823
1.41%
Notes:
1. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in January 2001 according to
Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A., Lery, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. &
Kim, H. (2003). Child Welfare Services Reports for California. Retrieved [8/29/2003], from University of California at Berkeley Center for
Social Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>
2. Children in the population refer to those aged 0 to 17 according to the California County Data Book 2001, Children Now, www.childrennow.org
3. Superscript "s" denotes a CWDA categorized "small" county as opposed to "big" county
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California
County Data Book summarized information on children aged 0-17 whereas the CWS data might include children as old as 19.
124
Descriptive Results:
Effective Practices for Children and Families of Color
Two counties did not provide information on effective practices; as such, a total of 43
effective practices for children and families of color in the child welfare system were identified.
Practices identified by the respondents as the most effective (# 1) for children and families of color
were divided into one of four categories: 1) beginning choice point practices (N = 13), 2) continuing
or ending choice point practices (N = 12), 3) organizational or structural practices (N = 14) and 4)
both beginning and continuing choice point practices (N = 4). Descriptive results for each of these
four categories are presented separately. In addition, descriptive statistics on the location, target
population, goals, history, factors related to implementation, factors that support the practice and
barriers to successful implementation of the practice are presented.
It should also be noted that many counties reported that they do not target services to any
specific racial/ethnic group and that the practices they identified as being most effective with
children and families of color, were also in general, their most effective practices for their overall
child welfare population.
Beginning Choice Point Practices (N = 13)
Thirteen counties identified practices that take place at early choice points as their most
effective practices for children and families of color in the CWS. The 13 beginning choice point
practices identified by the counties fell into three categories of practices: 1) family group
conferencing, or other team decision-making services (n = 7, 53.8%) 2) voluntary family
maintenance or other prevention/family preservation services (n = 4, 30.8%), and 3) risk assessment
tools or systems to guide decision-making (n = 2, 15.4%). Please see Table 32.
Family group conference or other team decision making
ractice
Voluntary family maintenance or other prevention/family
reservation service
Risk assessment tool ors stem
7
53.8%
4
2
30.8%
15.4%
Table 33 provides qualitative data on the beginning choice point practices that are related to
family group conferencing or other team decision-making practices. Most respondents described the
target population as children and families who have been detained or are at risk for court-ordered
services, although some respondents described using family group conferencing for all families
involved with the CWS, not just those at risk for court ordered out-of-home placement.
Most respondents described family group conferencing or other team decision-making
practices as involving the family, extended family or the support system in the group conference.
Some respondents described staff meetings or multidisciplinary team meetings that did not involve
the family. Survey respondents described the practice as a collaborative effort with the goal of
producing a plan for the child and family. Many respondents described the process as strengths
based and focused on how best to help children and families.
125
In general, respondents felt that the practice was effective with children and families of color
because it involves clients in the decision-making process. This inclusive model of practice was
described as particularly helpful for children and families of color because they may be "more
distrustful of the system," and also because inclusive practice may be more culturally appropriate
with some families. Additionally, team meetings not involving family members were described as
being beneficial to children and families of color because more accurate decisions are expected
from bringing numerous perspectives together.
Table 34 provides qualitative data on voluntary maintenance or other prevention/family
preservation practices. The target population for these services included families at risk for courtordered out-of-home placements, as well as one program aimed at expectant mothers and parenting
women. These practices were described as intensive in nature and focused on preventing child
maltreatment and out-of-home placements.
Voluntary family maintenance or other prevention/family preservation services were
described as effective with children and families of color because their goal is to keep children with
their birth families. Additionally, these services were viewed as effective for children and families
of color because they are collaborative in nature, involving other systems or organizations such as,
Native American tribes, mental health and probation systems.
Table 35 provides qualitative data on risk assessment tools or systems. These practices are
targeted to all.children reported to the child welfare system for suspected maltreatment. Risk
assessment practices were described as tools to help make decisions at critical points by assessing a
number of different risk factors.
Risk assessment tools or systems were described as effective with children and families of
color because they help to assess risk in an objective way, thereby minimizing potential bias based
on race/ethnicity.
Table 36 provides quantitative results on all of the beginning choice point practices
identified from the statewide survey. In general, beginning choice point practices are most often
delivered at the county child welfare office (n = 7, 53.8%); they tend to be targeted to children of all
ages (n = 12, 92.3%); as well as families of any ethnicity (n = 13, 100.0%). Families with substance
abuse problems (n = 9, 69.2%) are also targeted for these beginning choice point practices, as are
families with domestic violence, (n = 9, 69.2%), families with a mental health problem (n = 9,
69.2%), families with income problems (n = 8, 61.5%), and families with a disability (n = 8,
61.5%).
The most frequently identified goal of the practices was to prevent out-of-home placement
(n = 11, 84.6%); other goals included to assess level ofrisk (n = 10, 76.9%); improve psychosocial
functioning of family (n = 10, 76.9%); address substance abuse problems of family (n = 8, 61.5%);
address socioeconomic problems of family (n = 8, 61.5%); improve parenting skills, (n = 7, 53.8%);
address mental health problems of family (n = 7, 53.8%); ensure a permanent placement for the
child (n = 7, 53.8%), and reunify children with their birth families (n = 6, 46.2%)
Table 37 provides data on the history of the beginning choice point practices. Most were
first implemented in response to a particular problem (n = 7, 53.8%) and most were first initiated by
126
the director of the child welfare system (n = 8, 61.5%). The year the practices were first
implemented ranged from 1992 to 2002, and 15.4% (n = 2) of respondents reported that the practice
has changed significantly between now and the time it was first implemented.
Table 38 provides information on factors related to implementation of beginning choice
point practices. The majority of these practices are permanent (n = 10, 76.9%); most are funded
through state funds (n = 7, 53.8%) and 38.5% (n = 5) are mandated. Most social workers are
provided with training in how to implement the practice (n = 12, 92.3%); 15.4% (n = 2) of counties
contract with another agency or system to deliver the practice and 69.2% (n = 9) indicated that the
practice is widespread throughout the child welfare system. No respondents reported that the
practice is delivered in a specific geographic area or community. The majority of respondents (n =
8, 61.5%) reported the practice has been evaluated to determine its effectiveness and of those, 87 .5
% (n = 7) indicated that the practice is effective with the target population. Of those counties where
the practice had not been evaluated, 80.0% (n = 4) indicated that an evaluation is not needed.
Factors that support the practices and barrier to successful implementation are presented in
Table 39. The majority ofrespondents felt that families and children help support the successful
implementation of the practice (n = 10, 76.9%). Support from the mental heath system was also
frequently noted (n = 9, 69.2%); as was support from the substance abuse system (n = 8, 61.5%),
contract agencies (n = 7, 53.8%), county government officials (n = 7, 53.8%), and evaluations that
show practice is successful (n = 7, 53.8%). Few barriers to successful implementation were noted,
38.5% (n = 5) of respondents indicated that there were no barriers to successful implementation, and
38.5% (n = 5) also reported unstable funding is a barrier.
127
Table 33: Beginning choice point practices: Family group conference or other team decision-making practice (N = 7)
Target
Description of practice
Name of
Why effective with children and families of color
practice
population
Family
Available to all
The Family Group Conference is a tool for solving
I believe that this practice works well for families
Group
consumers who
family problems that is based on a simple, traditional
of color because it is based on principles shared
Conferencing have family,
belief: The combination of family strengths and
by those families. It involves much extended
extended family, community support can keep children safe and well
family, and is centered around a meal (family
friends, support cared for. A Family Conference is a gathering of family alone time) .. It is inclusive of all family members,
members, friends, and community specialists who join
including elders. Meetings can be held at the
together to improve the care and protection of a specific location that best fits the needs of the family.
child or children.
Children of all
Family
Family Conferencing is a program that uses strengthThis program allows families to assist in the
Conferencing ages
based methodology by allowing families to participate
decision process as to what is best for the family.
in the decision-making process as to how best help their
children. Family Conferencing is a gathering of
extended family members, other intimate support
people, and service providers in order to share
information and give the family an opportunity to
create a plan for their children.
Involving the extended family, non-relateq extended
Family
All families
These families are often even more distrustful of
Group
involved with
family and all service providers in a conference in
the "system" than most families; this practice
Conferencing CWS at any
which the family takes responsibility for making a plan allows the family to exert more control over the
stage of the
for the child/
direction of services and the planning for the
children.
process can use
this program
Children at risk This program targets all children (including children of Team decision-making/goal setting/evaluation
Youth Pilot
of placement, or color) at risk of placement, or at risk of being moved to process which includes the family and all support
Project
(YPP)
at risk of more
a more restrictive placement. The program uses monies persons they wish to include, as well as the
restrictive
that would be spent on foster care to provide services to various service providers working with the
family.
placement
maintain the child in the home or in a less restrictive
placement. The program uses a team decision-making
and goal-setting model, which includes the family and
supports.
128
Table 33 continued: BeJ?;inning
Target
Name of
practice
population
Placement
All detained
Meetings
children and
their families
choice point practices: Family group conference or other team decision-making practice (N = 7)
Description of practice
Why effective with children and families of color
After a child is detained, a placement meeting is held as
soon as possible to determin the "next best placement"
for that child. In addition to Social Workers and other
staff, parents and family members are included
whenever possible in this process, so that their input
can be provided in this decision. The safety of the child
and culturally relevant placements are goals ..
All children,
Detention
No child is detained by CPS or remains detained by law
Case Staffing CPS may
enforcement without being discussed as soon as
consider
possible in a staffing. The participants include the line
detaining or who ER worker in the matter, ER Supervisor, On-going
have been
Case Supervisor, and CPS Program Manager. The ER
detained by law worker presents the issues of protection, risk, current
enforcement in
situation, the court worker/supervisor, lends legal
an emergency
documentation evidence perspective, all others ask
questions, add comments with the ultimate goal of
consensus on whether to proceed with intervention and
what the intervention looks like.
All ER referrals Consensus decision making with workers, key program
Team
considered for
supervisors, mental health clinician, PI-IN, family
Assessment
voluntary or
Planning
decision meeting coordinator, and often other service
court-ordered
providers. Type & level of CWS intervention
services
determined after exploring safety/risk, family
strengths/needs, and available
resources/services/strategies. This is a collaborative
process which may be utilized for other key case
decisions.
Families and relatives are able to participate in
the decision making and are able to provide input
in to the "next best placement" for a child, which
can include placement with relatives or in a
culturally relevant home if the child cannot be
safely returned to his or her parents.
Many perspectives and opinions are brought to
bear at inception, promoting more objectivity and
balance regarding FACTS and EVIDENCE of
risk and best interests of the child. Less
possibility of stereotyping, subjectivity, ignorance
of cultural factors.
This decision-making forum is used for ALL
cases moving to on-going services. The core
group attends pretty consistently and represents
multiple perspectives, resulting in more consistent
and well thought out decision making regarding
case handling and service needs.
129
Table 34: Bee;inning choice point practices Voluntary family maintenance or other prevention/family preservation service (N = 4)
Target
Description of practice
Why effective with children and families of color
Name of
population
practice
Families at high Social worker and social work assistant are dedicated to American Indian Tribe has facilitated family
Voluntary
conferencing for tribal families.
risk of court
program. Services are based on strength-based and
Family
Maintenance intervention due family-focused approaches, including family
conferencing and family de_cision-making. Promoting
Program
to abuse and
Safe and Stable Families funding used for contracting
neglect issues.
services to include psych assessments, intensive case
management services, and a half-time family advocate.
High Risk
Expectant
Program in which pregnant mothers with a CPS history Missing
Mothers and
or involvement coupled with drug or alcohol use are
Infant
children under
targeted for intensive services and treatment to avoid
Program
placing their children into the foster care system.
age 1
Mothers of newborns are also a target population for
this program.
Family Preservation practice is an integrated,
Built a community consensus around the effective
Family
Abused and
comprehensive approach to strengthening and
service delivery model for Family Preservation in
Preservation neglected
the County; Supported public/private
families at
preserving families who are at risk of or already
collaborative planning and service delivery
imminent risk of experiencing problems in family functioning with the
goal of assuring the physical, emotional, social,
placement
efforts; Preventive-oriented; Collaboration with
other County Departments such as DPSS, Mental
educational, cultural and spiritual development of
Health and Probation; Strong and continued
children in a safe and nurturing environment.
leaderhip within the Department of Children and
Families Services.
Provide intervention and make referrals to other
We do not have foster homes in our county. We
Missing
Keep
agencies so that family can remain intact, if possible.
have seen children go through foster care system
families
and the majority do not do well.
together
130
Table 35: Beginning choice point practices Risk assessment tool or system (N = 2)
Description of practice
Name of
Target
population
practice
Research-based & validated tools for decision-making
All new intake
Structured
at critical points in child welf~re and Foster Care
to CPS
Decision
systems developed by the Children's Research Center
(currently) will
Making
in collaboration with COSS & California Counties.
System
soon extend to
(SDM)
all active cases
The risk assessment process looks at the child factors,
Best practice All children
including age, disability, illnesses as well as the
reported
risk
parent's ability to care-take, the history of abuse, also
assessment
speaking with the child, the parent and collateral
during the
contacts and then making a decision. We do not have a
ER phase
universal risk assessment tool, the workers are trained
in this process. We have an ER team that does this
practice.
Why effective with children and families of color
Objective nature of factors assessed.
Children are assessed in an objective wayconsidering their factors that place them at risk,
rather than being targeted for their ethnicity.
131
Table 36: Beginning choice point practices: Location, target population and goals (N = 13)
Frequency (Percentage)
Where does practice #1 take place?
7 (53.8%)
• County child welfare office
(30.8%)
4
• Throughout the child welfare system
2 (15.4%)
• Birth family's home
2 (15.4%)
• Contract agency site
1 (7.7%)
Foster
family's
home
•
I (7.7%)
• Another out-of-home placement site
2 (15.4%)
• Other
Who is the target population for practice # 1?
12 (92.3%)
• Children of all ages
(7.7%)
I
• Very young children
(100%)
13
• Families of any ethnicity
9 (69.2%)
• Families with substance abuse problems
9 (69.2%)
Families with domestic violence
(69.2%)
9
• Families with a mental health problem
8 (61.5%)
• Families with income problems
8 (61.5%)
• Families with a disabilitv
6 (46.2%)
• Other
What are the purposes and goals of practice #1?
11 (84.6%)
• Prevent out-of-home placement
10 (76.9%)
• Assess level of risk
10 (76.9%)
• Improve psychosocial functioning of family
(61.5%)
8
Address substance abuse problems of family
(61.5%)
8
Address
socioeconomic
problems
of
familv
•
(61.5%)
8
• Address housing problems of family
(53.8%)
7
• Improve parenting skills
(53.8%)
7
• Address mental health problems of family
(53.8%)
7
• Ensure a permanent placement for the child
(46.2%)
6
Reunify
children
with
their
birth
families
•
(7.7%)
1
• Other
•
•
132
Table 37: Beginning choice point practices: History of practice (N = 13)
Frequency (Percentage)
Why was practice# 1 first implemented?
7 (53.8%)
• In response to a particular problem
4
(30.8%)
• Funds became available
4 (30.8%)
• New policies were implemented
Who first initiated practice # 1?
8 (61.5%)
• Director of the child welfare system
4 (30.8%)
• Child welfare workers
3 (23.l %)
Other
•
When was practice #1 first implemented?
1
(7.7%)
• 2002
3 (23.1%)
• 2001
(7.7%)
1
2000
•
2 (15.4%)
• 1999
1
(7.7%)
• 1998
(7.7%)
1
• 1995
1
(7.7%)
• 1992
Has practice # 1 changed significantly between the time it
was first implemented and now?
2 (15.4%)
• Yes
133
Table 38: Beginning choice point practices: Factors related to implementation (N = 13)
Frequency (Percentage)
Current implementation status of practice # 1
IO (76.9%)
• Permanent
(23.1%)
3
Dependent
on
re-funding
•
What is the primary funding source for practice #1?
(53.8%)
7
• State funds
(38.5%)
5
• Federal funds
(23.1%)
3
County
funds
•
(15.4%)
2
• Other
Is practice # 1 mandated?
(38.5%)
5
• Yes
Are social workers provided with training in how to
implement practice # 1?
12 (92.3%)
• Yes
Do you contract with another agency or system to deliver
practice # 1?
2 (15.4%)
• Yes
Is practice # 1?
9 (69.2%)
• Widespread throughout the child welfare system
4
(30.8%)
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
(0.0%)
0
• Yes
Has practice # 1 been evaluated to determine its
effectiveness? .
(61.5%)
8
• Yes
If yes, what were the results of the evaluation?
(87.5%)
7
• Effective with target population
(12.5%)
1
Other
•
If practice #1 has not been evaluated, why not?
(80.0%)
4
• An evaluation was not considered to be needed
(20.0%)
1
• Other
134
Table 39: Beginning choice point practices: Factors that support practice and barriers to successful
implementation (N = 13)
Frequency (Percentage)
What are the major factors that support practice # 1
10 (76.9%)
• Support from families and children
(69.2%)
9
• Support from the mental health system
(61.5%)
8
• Support from the substance abuse system
(53.8%)
7
• Support from contract agencies
(53.8%)
7
• Support from county government officials
(53.8%)
7
Evaluations
show
that
practice
is
successful
•
(46.2%)
6
• Support from the community
(46.2%)
6
Support from state government officials
(46.2%)
6
Stable funding
(46.2%)
6
Support from the health system
(30.8%)
4
Support
from
the
criminal
justice
system
•
(30.8%)
4
• Support from foster families
3
(23.1%)
• Support from other out-of-home placement providers
What have been the barriers to successful implementation of
this practice?
(38.5%)
5
• No barriers
(38.5%)
5
Funding
is
not
stable
•
(15.4%)
2
• Insufficient support from child welfare workers
(7.7%)
1
• Insufficient support from management
1
(7.7%)
• Insufficient support from families and children
• Evaluations are inconclusive as to the success of the
(7.7%)
1
practice
•
•
•
135
Continuing Care/Ending Choice Point Practices (N = 12)
Twelve counties identified continuing care/ending choice point practices as their most
effective with children and families of color in the CWS. The 12 continuing care/ending choice
point practices identified by the counties fell into three categories of practices: 1) placement related
practices (n = 6, 50.0%); 2) the Family to Family Initiative (n = 4; 33.3%); and 3) intensive and
collaborative interventions for high-risk children (n = 2; 16.7%). Please see Table 40.
Table 40: Continuin care/endin
Intensive and collaborative interventions for high risk
children
2
16.7%
Table 41 contains qualitative data on placement related continuing care/ending choice point
practices. The target populations of these services varies, some are targeted to all children in out-ofhome care, others are targeted specifically to Native American children, one is targeted to children
in group homes (The Village Project) or at risk for entry into group homes and one is targeted to
parents with young children likely to enter out-of-home care (Shared Family Care).
Descriptions of the practices also varied. For instance, many participants noted that the first
priority is to place children with relatives; some discussed the role of culture and ethnicity in
making placement decisions. One practice {The Village Project) is described as a multidisciplinary
effort aimed at avoiding higher levels of care for children or working toward a lower level of care.
Lastly, one practice (Shared Family Care) is aimed at providing a placement for the whole family in
a mentor, foster home.
Respondents provided a variety of opinions regarding why the practice is effective with
children and families of color. Specifically, many mentioned the importance of placing children in
culturally appropriate homes-whether with relatives or with foster parents. The Shared Family
Care program is thought to be effective with children and families of color because it allows
families with few resources to be placed together and The Village Project was considered to be
effective with children and families of color because of a multicultural staff and flexibility of
funding sources.
Table 42 provides descriptive information on the Family to Family Initiative. The target
population for this practice varies somewhat; some respondents identified children and families in
foster care as the target population, others identified care providers and resource homes or all clients
and care providers. The Family to Family Initiative is described as engaging and involving
communities in preventing out-of-home placements, reunification, and providing resource homes
within children's neighborhoods. Many also described the practice as encompassing several aspects
of the child welfare system with the goal of improving the overall child welfare system.
136
Family to Family was described as effective with children and families of color because it
involves community and family support and is aimed at reducing disparities associated with gender,
and ethnicity and increasing the number of children placed in their own communities.
Table 43 provides qualitative data on continuing/ending choice point practices that are
aimed at intensive and collaborative interventions for high-risk children. The target population for
these services is children who are at risk for a more restrictive placement. These practices were
described as collaborative in nature-involving systems such as public health, community based
organizations, probation, behavioral health and faith based organizations. Interventions are
comprehensive, intensive and culturally competent.
Intensive and collaborative interventions for high-risk children were described as being
effective with children and families of color because the services are individualized to each family's
needs and are intensive in nature These practices were also described as "engaging families in a
non-threatening and respectful manner," and assessing the types of strengths and resources already
being used by the family. Children and families of color were described as benefiting from these
services because they are respectful of culture and involve a family-driven process
Quantitative results from survey responses are provided in Table 44. Most respondents
indicated that these continuing care/ending choice point practices take place throughout the child
welfare system (n = 7, 58.3%) and are targeted to children of all ages (n = 9, 75.0%), as well as
older children (11 to 18 years) (n = 5, 41.7%), young children (4 to 10 years), (n = 3, 35.0%), and
very young children (0 to 3 years) (n= 3, 25.0%). Most are targeted to families of any ethnicity (n =
10, 83.3%), with far fewer being targeted to African American families (n = 2, 16.7%), or American
Indian families (n = 2, 16.7%). Families with substance abuse problems were also identified as a
target population (n = 9, 75.0%) as were families with a mental health problem (n = 9, 75.0%),
families with domestic violence (n = 9, 75.0%), families with income problems (n = 7, 58.3%), and
families with a disability (n = 7, 58.3%).
The most frequently noted goal of the practice is to reunify children with their birth families
(n = 10, 83.3%), followed by ensuring a permanent placement for children (n = 8, 66.7%), and
preventing an out-of-home placement (n = 8, 66.7%). Other commonly occurring goals included
improving parenting skills (n =7, 58.3%); improving psychosocial functioning of family (n = 6,
50.0%); addressing socioeconomic problems of family (n = 6, 50.0%); and addressing housing
problems of family (n = 6, 50.0%).
Table 45 provides information on the history of the continuing care/ending choice point
practices. The majority of practices were first implemented due to new policies (n = 8, 66.7%); the
director of the child welfare system was most often identified as the person who first initiated the
practice (n = 11, 91.7%), followed by child welfare workers (n = 4, 33.3%). Many practices were
first implemented in 2001 (n = 4, 33.3%), and 33.3% (n = 4) had changed significantly since first
being implemented.
Factors related to implementation are presented in Table 46. The majority of practices are
permanent (n = 11, 91.7%), and are funded by state funds (n = 7; 58.3%). Of the practices, 41.7% (n
= 5) are mandated and 91. 7% (n = 11) of counties provide social workers with training on how to
implement the practices. Of the counties, 33.3% (n = 4) contract with another agency to implement
the practices, and 58.3% (n = 7) of respondents indicated that the practice is widespread throughout
137
the child welfare system. Very few respondents indicated that there is a specific geographic
community in which the practice is targeted (n = 2, 16.7%), and only 16.7% (n = 2) indicated that
the practice has been evaluated to determine its effectiveness. Of those practices that had been
evaluated, 100.0% (n = 2) reported that the practice is effective with the target population. For
practices that had not been evaluated, 33.3% (n = 4) have not been evaluated because of insufficient
resources.
Table 47 provides information on factors that support the practice and barriers to successful
implementation of the practice. The most commonly mentioned supportive factors included support
from child welfare workers (n = 11, 91.7%), support from management (n = 11, 91.7%), and
support from families and children (n = 11, 91.7%). Other factors that support the practice include
support from foster families (n = 9, 75.0%); contract agencies (n = 8, 66.7%); other out-of-home
placement providers (n = 7, 58.3%), the mental health system (n = 7, 58.3%), state government
officials (n = 7, 58.3%); and county government officials (n = 7, 58.3%). Many respondents did not
note barriers to successful implementation of the practice, no barriers were reported by 16.7% (n =
2) of respondents, and 16.7% (n = 2) reported that unstable funding is a barrier to successful
implementation.
138
Table 41: Continuing care/ending choice point practices: Placement related practices (n = 6)
Why effective with children and families
Description of practice
Name of practice Target population
of color
Our county is predominately Caucasian,
All children who
In all cases, we must first consider any willing,
Culturally
so children of color are already truly a
able and appropriate relatives for placement. If
require out-ofsensitive
this is not possible, we then try to place in a home minority. Placing them in a home that is
placements
home placement
sensitive to that is very important.
of the same ethnicity, religion, etc. If this is not
possible, then we let the foster parents know that
their culture is an important issue, and we are
expecting them to continue to allow the children
to live in a culturally sensitive environment.
For families with few resources it allows
Mothers/Fathers
This program provides placement for the family
Shared Family
Care
with young
in a mentor home. Usually one parent (mother in for placement of child and parent
together. Most of our families with few
children likely to
most cases) is subject to intervention by child
resources are those living in communities
welfare. Rather than place the child in foster
enter out of home
impacted by poverty, drugs, and crimes.
care, it is assessed that the mother is willing to
care
Families of color primarily live in these
participate in Shared Family Care and will abide
with the conditions of working and living with a
areas.
mentor family.
Children are kept in their communities
All Children
Identify as quickly as possible family members
Place children
with relatives of same culture and
for
placement
with relatives
practices promote stability for children
They are treated the same as all other
Native American Native Am/lCWA Attempt to place all children with families.
cases-dignity-compassion-provided
Counseling/assessment/service orientated.
children placed
services- acknowledgement of their
with families
concerns and problems.
through ICWA
139
Table 41 continued: Continuing care/ending cho.ice point practices: Placement related practices (n = 6)
Name of practice Target population Description of practice
Why effective with children and families
of color
Multidisciplinary effort to work with families and
S8163
Kids in group
A multicultural staff in the county and
program-"The
kids to avoid higher levels of care or get them
homes level 13 or
in the community-based providers and
village project"
14 or at risk for
flexibility of funding sources-money
into a lower level of care-like a foster home or
going into a group back into their home. The staff are bilingual and
can be used for activities outside the
home
bi-cultural and they are able to make connections group home, so it can be used for things
with multi-cultural families. They also work hand to prevent a group home placement and
and hand with a parent advocacy group-and they not just for kids already in a group home.
advocate for children and families especially kids
with educational difficulties.
ICWA-·
Native Americans When a Native American child is removed from
We are able to place many of our Native
placement of
the home of parents, we contact any relatives in
American children in homes that comply
Native American
with ICWA.
order to place the child in compliance with
children
ICWA.
140
Table 42: Continuing care/ending choice point practices: Family to Family Initiative (n = 4)
Name of practice Target population
Description of practice
Family to Family
All children in out
of home
placements
Family to Family
Initiative
All clients and
care providers
Family to Family
Initiative
Children &
families in foster
care
Family to family
Resource homes
and the
community
Regional initiatives underway to work with local
communities to prevent removal and/or maintain
children in their community/school if they must
be removed. Community collaboration on child
safety and well being.
The use of the following Family to Family
strategies are applied to all aspects of child
welfare placement practices in this County: Team
decision making, Recruitment, Retention, and
Support of Care Providers, Self-Evaluation and
Community Partnership.
Community-based, family focused and designed
to meet 9 key outcomes toward improving our
foster care system including reducing disparities
associated with ethnicity, gender & age.
This is an effort to engage the community in
providing resource homes for foster children who
are being placed in homes from their
neighborhoods in which they are being removed.
The community is being engaged and as well as
community based organizations.
Why effective with children and families
of color
Community and family
involvement/support.
Missing
Outcomes include: reduction in
disparities associated with sex, gender &
ethnicity, increasing kids placed in their
own neighborhoods & communities,
Missing
141
Table 43: Continuing care/ending choice point practices: Intensive and collaborative interventions for high-risk children (n = 2)
Name of practice Target population Description of practice
Why effective with children and families
of color
Children's
Children who are
To develop a child 'who is emotionally
As with all families, intensive services,
System of Care
in danger of being disturbed/behaviorally disturbed' through a
individualized plans
placed in a higher collaborative interagency system of care with
services that are "youth-centered;" family
level of care i.e.
Group homes
involved & focused; culturally competent;
community based and comprehensive.
Wraparound
Children who are
This practice really values families individually
The practice engages with families in a
services
placed or at risk of and values their culture. It provides intensive and respectful and non-threatening manner
placement into
supportive in-home services to keep the child at
and asking them what supports, what
home or return them home. It's also been a
level 14 group
people, in your life and in your
home care
collaborative effort between probation, behavioral community that are important to you and
health, public health, CBO's, and faith-based
how can we incorporate those aspects of
organizations.
your life into this child and family plan to
help your child stay safely at home. And
by asking those questions and evaluating
what the family uses to thrive and asking
them to make that a part of the service
planning for their child is kind of a
foreign way of working with the families.
It is a family-driven, needs based process,
rather than a professionally-driven
process. There is respect for the family's
culture.
142
Table 44: Continuing care and ending choice point practices:
Location, target population and goals (N = 12)
Where does practice # 1 take place?
Throughout the child welfare system
• County child welfare office
• Birth family's home
• Foster family's home
• Court
• Another out-of-home placement site
• Other
Who is the target population for practice # 1?
• Children of all ages
• Very young children (0 to 3 years)
• Young children (4 to 10 years)
• Older children (11 to 19 years)
Families of any ethnicity
African American families
• American Indian families
• Asian American/Pacific Islander families
• Latino families
• White families
Families of another ethnicity
• Families with substance abuse problems
• Families with a mental health problem
• Families with domestic violence
• Families with income problems
Families with a disability
• Other
What are the purposes and goals of practice #I?
• Reunify children with their birth families
• Ensure a permanent placement for the child
• Prevent out-of-home placement·
• Improve parenting skills
• Improve psychosocial functioning of family
• Address socioeconomic problems of family
• Address housing problems of family
• Address substance abuse problems of family
• Address mental health problems of family
• Assess level of risk
• Other
•
•
•
•
•
Frequency (Percentage)
7
4
4
4
4
4
(58.3%)
(33.3%)
(33.3%)
(33.3%)
(33.3%)
(33.3%)
(4L7%)
9
3
3
(75.0%)
(25.0%)
(35.0%)
(41.7%)
(83.3%)
(16.7%)
(16.7%)
(8.3%)
(8.3%)
(8.3%)
(16.7%)
(75.0%)
(75.0%)
(75.0%)
(58.3%)
(58.3%)
(50.0%)
5
5
IO
2
2
1
I
I
2
9
9
9
7
7
6
IO
8
8
7
6
6
6
5
5
4
4
(83.3%)
(66.7%)
(66.7%)
(58.3%)
(50.0%)
(50.0%)
(50.0%)
(41.7%)
(41.7%)
(33.3%)
(33.3%)
143
Table 45: Continuing care and endin2 choice point practices: History of practice (N = 12)
Frequency (Percentage)
Why was practice # 1 first implemented?
8
(66.7%)
• New policies were implemented
(25.0%)
3
• Funds became available
2
(16.7%)
• In response to a particular problem
(50.0%)
6
• Other
Who first initiated practice #1?
11 (91.7%)
• Director of the child welfare system
4 (33.3%)
• Child welfare workers
2 (16.7%)
• Mental health system
1
(8.3%)
• Foster families
1
(8.3%)
Substance
abuse
system
•
(8.3%)
1
• Contract a2encies
(8.3%)
1
• Criminal justice system
4 (33.3%)
Other
When was practice #1 first implemented?
1
(8.3%)
• 2003
2 (16.7%)
• 2002
4 (33.3%)
2001
•
2 (16.7%)
• 2000
Has practice # 1 changed significantly between the time it
was first implemented and now?
4 (33.3%)
• Yes
•
144
Table 46: Continuing care and ending choice point practices:
Factors related to implementation (N = 12)
Current implementation status of practice # 1
• Permanent
• Demonstration
What is the primary funding source for practice # 1?
• State funds
• County funds
• Federal funds
• Nongovernmental grant
• Other
Is practice #1 mandated?
• Yes
Are social workers provided with training in how to
implement practice # 1?
• Yes
Do you contract with another agency or system to deliver
practice # 1?
• Yes
• No
Combination of internal and external resources
-Is practice # 1?
• Widespread throughout the child welfare system
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
• Yes
Has practice # 1 been evaluated to determine its
effectiveness?
• Yes
If yes, what were the results of the evaluation?
• Effective with target population
If practice # 1 has not been evaluated, why not?
• Insufficient resources
• Insufficient time
• An evaluation was not considered to be needed
• Other
.-
Frequency (Percentage)
11
1
(91.7%)
(8.3%)
7
6
3
2
3
(58.3%)
(50.0%)
(25.0%)
(16.7%)
(25.0%)
5
(41.7%)
11
(91.7%)
4
4
3
(33.3%)
(33.3%)
(25.0%)
7
5
(58.3%)
(41.7%)
2
(16.7%)
2
(16.7%)
2
(100.0%)
4
2
1
3
(33.3%)
(16.7%)
(8.3%)
(25.0%)
145
Table 47: Continuing care and ending choice point practices: Factors that support practice and
· barriers to successful implementation (N = 12)
Frequency (Percentage)
What are the maior factors that support practice # 1
11 (91.7%)
• Support from child welfare workers
11 (91.7%)
• Support from management
11 (91.7%)
• Support from families and children
(75.0%)
9
• Support from foster families
8 (66.7%)
• Support from contract agencies
(58.3%)
7
Support
from
other
out-of-home
placement
providers
•
(58.3%)
7
• Support from the mental health system
(58.3%)
7
• Support from state government officials
(58.3%)
7
• Support from county govemmenfofficials
(50.0%)
6
• ·Stable funding
(50.0%)
6
Support
from
the
community
•
(50.0%)
6
• Support from the health system
6.
(50.0%)
• Support from criminal justice system
5 (41.7%)
• Support from the substance abuse system
(25.0%)
3
• Evaluations show that practice is successful
(16.7%)
2
• Other
What have been the barriers to successful implementation of
this practice?
(16.7%)
2
• No barriers
(16.7%)
2
• Funding is not stable
1
(8.3%)
• Insufficient support from state government officials
Insufficient
support
from
county
government
•
(8.3%)
I
officials
(8.3%)
1
• Insufficient support from child welfare workers
(8.3%)
I
• Insufficient support from foster families
(8.3%)
I
Insufficient
support
from
the
community
•
(8.3%)
I
• Insufficient support from the criminal justice system
(50.0%)
6
• Other
146
Both Beginning and Continuing/Ending Choice Point Practices (N = 4)
Four respondents identified practices that were related to both beginning and
continuing/ending choice points as their most effective practices for children and families of
color in the CWS. The four beginning and continuing/ending choice point practices fell into two
categories: 1) community-based services (n = 3, 75.0%); and 2) strength based assessments (n =
1, 25.0%). Please see Table 48.
Table 48: Both beginning and continuing/ending choice point practices: Types of practices (N =
4
Table 49 provides qualitative information on community based services. The target
population for these services includes low-income populations, families at risk for entry into the
child welfare system, families already in the child welfare system, and the community. These
practices were described as being located in the communities in which clients reside and include
a wide range of services such as, health, substance abuse, public assistance, domestic violence
and home parenting services. Practices were also described as collaborative and aimed at
preventing an out-of-home placement or reunifying children with their families.
These practices were described as effective with children and families of color because
services are "user friendly;" clients have easy access to many services that are housed in one
location. Community based services are free and outreach for these services is performed
through advertisements into different communities. Moreover, because these services are
typically located within clients' own communities, services are "community-specific" and clients
may run into their own community members while accessing services, thus making the services
less intrusive for clients. Additionally, these practices were described as flexible and able to
address "changing community needs."
Table 50 provides information on family-centered strength based assessments. This
practice was described as being targeted to Hispanic, African American, American Indian and
Asian American/Pacific Islander clients. It is focused on family strengths with an empowerment
perspective.
Quantitative results from survey responses are presented in Table 51. Practices that are
both beginning and continuing/ending practices typically take place at a contract agency site (n =
2, 50.0%), or the birth family's homes (n = 2, 50.0%). One hundred percent (n = 4) of these
practices are targeted to: children of all ages, families of any ethnicity, families with substance
abuse problems, families with a mental health problem, families with domestic violence, families
with income problems, or families with a disability. Seventy five percent (n = 3) of practices
have the following goals: preventing an out-of-home placement; reunifying children with their
birth families; assessing level of risk; improving psychosocial functioning of families; improving
parenting skills; addressing substance abuse problems; addressing mental health problems;
addressing socioeconomic problems; and addressing housing problems.
147
Information on the history of the practices is presented in Table 52. The practices were
first implemented in response to a particular problem (n = 1, 25.0%); funds becoming available
(n = 1, 25.0%); or new policies being implemented (n = 1, 25.0%). Child welfare workers (n = 2,
50.0%); the child welfare director (n = 2, 50.0%); and contract agencies (n = 2; 50.0%) were
identified as having first initiated the practices. The year the practice was first implemented
ranged from 1967 to 1997, and 66. 7% (n = 2) of respondents indicated that the practice had
changed significantly since it was first implemented.
Factors related to the implementation of both beginning and continuing/ending choice
point practices are presented in Table 53. All practices (n = 4, 100.0%) are permanent and most
are primarily funded through federal funds (n = 4, 100.0%). Fifty percent of respondents (n = 2)
indicated that the practice is mandated and 33.3% contract with another agency to deliver the
practice. Seventy five percent of respondents indicated that the practice is widespread throughout
the child welfare system. All respondents (n = 4, 100.0%) reported that the practice has been
evaluated, and 50.0% (n = 2) indicated that evaluations demonstrated that the practice is effective
with the target population.
Factors that support the practice and barriers to successful implementation are presented
in Table 54. One hundred percent of respondents (n = 4) indicated that support from the
following entities are major factors that support the practices: community, county government
officials, mental health system, substance abuse system, health system, and families and
children. Seventy five percent of respondents (n = 3) indicated that unstable funding is a barrier
to successful implementation of the practices.
148
Table 49: Both be_ginnin_g and continuini /ending practices: Community based services
Name of
Target population
Description of practice
practice
Community- Low-income
Federal and state funded programs that are free to
based
populations
low-income clients, including the health
services
department, substance abuse treatment, Aids
department, WIC, immunizations, planned
parenthood. Services include domestic violence
support, clothing donations, utility assistance,
HEAD Start program; sexual assault crisis, and in home parenting programs, and teenage parenting
programs.
Family
Children and Families at There are many Family Resource centers. A
Resource
risk for entry into the
collaborative of agencies applies for funds from the
Centers
CWS and children and
funding source (though RFP). The Family
families already in
Resource Centers have a nun;iber of agencies
CWS. Most clients are
working together within one site - so it provides
just one-stop for families to receive services. The
Hispanic, the second
most are API
Family Resources Centers are located in the
communities and the neighborhoods where the
clients live. Services are meant to be accessible and
less intimidating than county _government services.
Community, Families
Commmunity specific family based services located
Family
and Children
in six geographical areas in the County with embed
Resource
access to public child welfare, mental health,
Centers
alcohol and other drug and physical health services.
Why effective with children and families of
color
Because services are free and they advertise
into all areas of the community, and there are
satellite offices in many locations; so clients
have access to the services, which is
extremely helpful. They also cross-refer so we
all depend on one.another. We have a large
Hispanic population so information is
available in Spanish. Transportation is
accessible.
Because wherever the families are they can
reach out to the resources that meet their
need-there are a multitidue of agencies in
one place-it's a one-stop shop. Clients are
more likely to run into their own community
members at the Family Resource Centers. It is
user friendly and less intrusive than other
services.
Practice is community specific with flexibility
to address changing community needs.
Subsequently, it offers increased accessibility
to services, intervention/prevention and
family support and advocacy.
149
Table 50: Both beginning and continuin, lendine practices: Streneth based assessments
Description of practice
Target population
Name of
practice
Family
Hispanic, AA, American Focused on family strengths with the empowerment
perspective.
Indian and Asian
centered
strength
based
assessments
Why effective with children and families of
color
Missing
150
Table 51: Both beginning and continuing care/ending choice point practices: Location, target
population and goals (N = 4)
Frequency (Percentage)
Where does practice # 1 take place?
2
(50.0%)
• Contract agency site
2
(50.0%)
• Birth family's home
1
(25.0%)
County
child
welfare
office
•
1 · (25.0%)
• Foster family's home
(25.0%)
1
• Another out-of-home placement site
2
(50.0%)
• Other
Who is the target population for practice #1?
4 (100.0%)
• Children of all ages
4 (100.0%)
• Families of any ethnicity
(100.0%)
4
Families
with
substance
abuse
problems
•
4 (100.0%)
Families with a mental health problem
4 (100.0%)
Families with domestic violence
4 (100.0%)
• Families with income problems
4 (100.0%)
• Families with a disability
I
(25.0%)
Other
•
What are the purposes and goals of practice #1?
3 (75.0%)
• Prevent out-of-home placement
3 (75.0%)
• Reunify children with their birth families
3 (75.0%)
Assess
level
of
risk
•
3 (75.0%)
• Improve psychosocial functioning of family
(75.0%)
3
• Improve parenting skills
3 (75.0%)
• Address substance abuse problems of family
3 (75.0%)
• Address mental health problems of family
3 (75.0%)
• Address socioeconomic problems of family
3 (75.0%)
• Address housing problems of family
(50.0%)
2
• Ensure a permanent placement for the child
3 (75.0%)
• Other
•
•
151
Table 52: Both beginning and continuing care/ending choice point practices:
History of practice (N = 4)
Frequency (Percentage)
Why was practice #1 first implemented?
1 (25.0%)
• In response to a particular problem
1 (25.0%)
• Funds became available
1
(25.0%)
New
policies
were
implemented
•
1 (25.0%)
• Other
Who first initiated practice # 1?
2 (50.0%)
• Child welfare workers
2 (50.0%)
Director
of
the
child
welfare
system
•
2 (50.0%)
• Contract agencies
1 (25.0%)
• Clients of the CWS
1 (25.0%)
• Foster families
1 (25.0%)
• Out-of-home placement providers
1
(25.0%)
Mental
health
system
•
1 (25.0%)
• Substance abuse system
1 (25.0%)
• Health system
1 (25.0%)
• Criminal justice system
2 (50.0%)
• Other
When was practice # 1 first implemented?
1 (33.3%)
• 1997
(33.3%)
1
1995
•
1 (33.3%)
• 1967
Has practice # 1 changed significantly between the time it
was first implemented and now?
2 (66.7%)
• Yes
152
Table 53: Both beginning and continuing care/ending choice point practices: Factors related to
implementation (N = 4)
Frequency (Percentage)
Current implementation status of practice # 1
4 (100.0%)
Permanent
What is the primary funding source for practice # 1?
4 (100.0%)
• Federal funds
2
(50.0%)
• State funds
(50.0%)
2
County funds
(25.0%)
1
Nongovernmental grant
Is practice # 1 mandated?
2
(50.0%)
• Yes
Are social workers provided with training in how to
implement practice #1?
(50.0%)
2
• Yes
Do you contract with another agency or system to deliver
practice #I?
(33.3%)
I
• Yes
(33.3%)
1
• No
(33.3%)
1
Combination of internal and external resources
Is practice # 1?
(75.0%)
3
• Widespread throughout the child welfare system
(25.0%)
1
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
2
(50.0%)
• Yes
Has practice # 1 been evaluated to determine its
effectiveness?
4 (100.0%)
Yes
If yes, what were the results of the evaluation?
2
(50.0%)
• Effective with target population
• Evaluation did not provide evidence that practice was
1
(25.0%)
effective
(25.0%)
1
Other
•
•
•
•
•
•
153
Table 54: Both beginning and continuing care/ending choice point practices: Factors that support
practice and barriers to successful implementation (N = 4)
Frequency (Percentage)
What are the major factors that support practice # I
4 (100.0%)
• Support from the community
4 (100.0%)
• Support from county government officials
4
(100.0%)
• Support from the mental health system
4 (100.0%)
• Support from substance abuse system
4 (100.0%)
• Support from health system
4 (100.0%)
Support
from
families
and
children
•
(75.0%)
3
• Support from child welfare workers
(75.0%)
3
• Support from management
(75.0%)
3
• Support from foster families
(75.0%)
3
• Support from contract agencies
(75.0%)
3
Support
from
the
criminal
justice
system
•
(75.0%)
3
• Evaluations show that practice is successful
(50.0%)
2
• Support from state government officials
(25.0%)
1
• Stable funding
(25.0%)
1
• Support from other out-of-home placement providers
What have been the barriers to successful implementation of
this practice?
(75~0%)
3
• Funding is not stable
1
(25.0%)
• Insufficient support from state government officials
• Evaluations are inconclusive as to the success of the
(25.0%)
1
practice
154
Organizational or Structural Practices (N = 14)
Fourteen respondents identified organizational or structural practices as their most
effective practices for children and families of color in the CWS. The fourteen
organizational/structural practices fell into three categories: I) practices related to cultural
competency of staff (n = 10, 71.3%), 2) collaboration with American Indian tribes (n = 3,
21.3%), and 3) vertical case management (n = I, 7.4%).
Table 55: Or anization or structural
Collaboration with American Indian Tribes
Vertical case mana ement
3
I
Qualitative data on cultural competency practices are presented in Table 56. The target
population for these services varies, some are focused on the Latino population or the Spanish
speaking population, while others are focused on all children in the system and one is focused on
social workers themselves. Practices related to cultural competency involve the use of
bilingual/bicultural staff or competent translators in the delivery of services, as well as providing
information on civil rights, providing training to social workers and matching clients and
workers on ethnicity/culture.
Many respondents mentioned the importance of having bilingual/bicultural staff and the
need to be able to communicate with clients in their own language as reasons for why these
practices are effective with children and families of color. The importance of assessing and
delivering services in the clients' language is described as resulting in more accurate
assessments, more meaningful services and better connections between worker and client.
Cultural competency trainings for social workers are described as bringing awareness and
knowledge to staff.
Table 57 provides qualitative data on organizational or structural practices related to
collaboration with American Indian tribes. These practices are targeted to Native American
children and families and generally include multidisciplinary teams involving tribal
representatives, legal professionals ~nd child welfare staff who come together to better serve
Native American children and ensure the proper implementation of the Indian Child Welfare
Act.
Practices related to collaboration with American Indian tribes were described as effective
for children and families of color because they respect cultural heritage; encourage cultural
awareness for Native American children; result in more complete assessments; better service
planning and fewer contested hearings.
Table 58 provides qualitative information on vertical case management. This practice is
targeted toward all clients of the child welfare system and it involves having social workers carry
the same cases all the way through the system. This practice was described as effective with
children and families of color because there is consistency in the case and a better relationship
155
between worker and client. It was also noted that this practice is effective with the overall
population as well.
Quantitative data from survey responses is provided in Table 59. Most
organizational/structural practices take place throughout the child welfare system (n = 8, 57.1 %)
and are targeted to children of all ages (n = 12, 85.7%). Many are targeted to families of any
ethnicity (n = 7, 50.0%), while some are targeted to Latino families (n = 4, 28.6%), or American
Indian families (n = 3, 21.4%). Additionally, families with substance abuse problems are also a
focus of these practices (n = 7, 50.0%), as are families with mental health problems (n = 7,
50.0%). Goals of the practice include preventing an out-of-home placement (n = 7, 50,0%);
ensuring a permanent placement (n = 7, 50.0%); reunifying children and families (n = 6, 42.9%);
addressing substance abuse problems of families (n = 6,42.9%); and improving parenting skills,
(n = 6, 42.9%).
Information on the history of the practices is presented in Table 60. Most practices were
first implemented in response to a particular problem (n = 7, 50.0%); and the director of the child
welfare system was most frequently mentioned as the person who first initiated the practice (n =
7, 50.0%), followed by child welfare workers (n = 6, 42.9%). The year the practice was first
implemented range from 1972 to 2002 and 46.2% of counties indicated that the practice had
changed significantly since it had first been implemented.
Table 61 provides data on factors related to implementation. Most practices are
permanent (n = 12, 92.3%) and the majority are funded through federal dollars (n = 8, 57.1 %).
Most counties also indicated that these practices are mandated (n = 8, 61.5%), and 66.7% (n = 8)
of counties indicated that social workers are provided with training in how to implement the
practice. Most counties do not contract with another agency to deliver the practice (n = 9, 64.3%)
and most also indicated that the practice is widespread throughout the child welfare system (n =
12, 92.9%). Only one county (7.1 %) indicated that there is a specific geographic area in which
the practice is targeted. Four counties (28.6%) reported that the practice has been evaluated to
determine its effectiveness and of those 66.7% (n = 2) reported that the practice was found to be
effective with the target population. Of those counties in which no evaluation has been
conducted, 50.0% (n = 5) indicated that an evaluation was not needed.
Factors that sugport the gractices and barriers to successful imglementation are presented
in Table 62. Support from child welfare workers (n = 11, 78.6%) was most commonly noted as a
major factor that supports the practice, followed by support from management (n = 9, 64.3%);
county government officials (n = 8, 57.1 %); families and children (n = 7, 50.0%); and the
community (n = 7, 50.0%). No barriers to successful implementation were noted by 28.6% of
counties (n = 4), while 14.3% (n = 2) reported insufficient support from the community and
14.3% (n = 2) noted insufficient support from child welfare workers.
156
Table 56: Organizational or structural practices: Practices related to cultural competency (n = 10)
Why effective with children and families of color
Target
Description of practice
Name of practice
population
We have a very high rate of return for
One of things that has made it particularly
SpanishThe use of
effective in this county is the size of the county.
speaking
Hispanic children--one of our better
competent
clients
practices is the use of competent translators And our translator even though they may not be
translators
CW social workers they are from the community
and always having this available when
and they are familiar with the community and the
working with a family. These are used in
the juvenile court as well. And we also have dynamics of that particular culture. For the most
part this had enhanced the practice.
bilingual staff as well.
Availability of
bilingual/bicultural
staff to serve the
Latino population.
Latino
population
At least one
bicultural/bilingual
worker in each unit
Mostly
Hispanic
children
B ilingual/B icultural
social work staff.
Spanish
speaking
population
Hire and place bilingual/bicultural social
Cultural knowledge and ability to converse in same
workers at office serving majc;>rity of Latino language.
families. Provides culturally sensitive
services to families, better assuring accurate
asessments of risk and safety for children.
The agency really values language and
When you are doing social work and having to
trying to accommodate that the best we can. assess and discuss very difficult things with people
having to do with emotions and feelings these
Most of our receptionists are bilingual and
our social service aides and social worker
things are very difficult to do through translation
I's are bilingual. Whenever we start to
because a lot of what a social worker does is their
use of self and being able to read body language
implement a program we always look to
bringing in bilingual/bicultural staff. We are and intonation and understand the subtlety of
limited in our resources but we do our best. language and what someone is saying. You are less
effective through translation. The communication
We also really try to ask this of our
is much better when services are conducted in the
providers as well.
clients language. And bicultural folks understand
the local jargon.
Bilingual/bicultural staff assigned to
Ability to communicate with the families in their
bilingual/monolingual cases to ensure
native language & an ability to understand the
appropriate communication of issues, needs, culture helps the social worker provide appropriate
resources & guidance to the family.
services and agency function/process.
157
Table 56 continued: Organizational or structural practices: Practices related to cultural competency (n = 10)
Why effective with children and families of color
Target
Description of practice
Name of practice
population
Recruiting and hiring bilingual/cultural staff Recruitment is a priority within service delivery
B ilingual/B icultural Families
to work with the diverse population we
system and best practice philosophy .. .it becomes
whose
staff
integrated in every day work.
children
serve. Staff are assigned throughout our
have been
child welfare programs/services.
Assessment of language and cultural
abused.
staffing needs is determined on an ongoing
basis and is used as a criteria to fill vacant
positions.
Federal Division 21
Regulation-Civil
rights
All
Emphasis of
Cultural
Competence
All cultures
We inform all clients of their civil rights.
This includes that they cannot be .
discriminated against on the basis of
national origin, disabilities, race, age sex,
all those areas. We are supposed to talk to
all families about their civil rights. We ask
the family if they want a special skills
worker assigned-that would specialize in a
particular language or culture. We have
Black culture workers, Hispanic cultural
workers, and Asian, Russian special skills
workers. We ask all clients if they want a
special skills worker-this is for
language-but also culturally matching the
clients.
Our agency has a cultural competence
committee. We mandate 8 hours of training
for staff in orientation and an additional 16
hrs through our leadership academy.
Matching families with a social worker of their
ethnic background is helpful in getting families
back together. If the social worker understands the
culture-like if there are different discipline
techniques that are seen more in one culture, the
social worker may be able to help the child and
family more. Clients are a bit more comfortable,
social workers can confront them better-work
with the better to help bring about change.
Awareness, Knowledge
158
Table 56 continued: Organizational or structural practices: Practices related to cultural competency (n = 10)
Name of practice
Target
Description of practice
Why effective with children and families of color
population
Policy is to align
Applies
Our practice is to assign staff of similar
Ability to establish trusting relationship
case assignments to across all
ethnicity and we are able to do so most
cws
often in our African American and Mexican
staff from like
ethnicity when
components American population. If unable to do so,
possible
we consider assigning a spanish speaking,
non-Mexican-American. staff.
Cultural
Social
We require all social workers to attend
Missing
competence
workers
training on cultural competency on a regular
training
basis.
Civil Rights
Missing
Missing
Training- bringing awareness to staff
Training
159
Table 57: Organizational or structural practices: Collaboration with American Indian tribes (n = 3)
Target
Description of practice
Name of practice
Why effective with children and families of color
population
Indian Child
American
A practice in which tribal representatives,
It respects their cultural background and keeps
Indian
agency staff, kids' attorneys, County
Welfare Act
them connected with their tribal roots, it respects
(ICWA)
children and Counsel, and others come together on a
that people can do things differently and still have
families
Roundtable
monthly basis to talk about issues related to good outcomes. The tribe is actively solicited to be
the implementation of the ICWA, how we
involved with children and families in the CWS.
can work to together on certifying tribal
We pay for culturally appropriate services, like
tribal dances, and other non-traditional services.
homes, how to get services for American
Indian Children and what are the most
appropriate services. The group created a
handbook of ICWA regulations. Very
effective in opening dialogue between the
parties, sensitizing staff to issues around
services that would be more culturally
appropriate to American Indian children,
identifying and providing services, and
ensuring that kids can get to tribal homes.
ICWA
Native
Indian Child Welfare Act Multidisciplinary More complete assessments, better service
Team
planning and less contested hearings.
American
Some of the tribes have strong cultural
Collaboration with
Native
Native American families-in general they are
centers
with
health
care
and
counseling
and
tribes and tribal
Americans
still at war with both the US and with each other. It
so we try to work hard with them. Native
is effective because in spite of these things we are
courts
American children do best when placed
able to overcome the barriers and we are able to
with a Native American family. The Tribal
make use of the real resources of cultural
court makes placement decisions and so we awareness for Native American people.
work with them to make decisions that we
Collaborating with the tribes depends on having the
are all in agreement about. We work closely right social worker work with the tribes-in
with the tribes in Native American cases.
general the tribes have been cooperative with the
cws.
160
Table 58: Organizational or structural practices: Vertical case management (n = 1)
Name of practice
Target
Description of practice
Why effective with children and families of color
population
Social workers
All
Social Workers carry the same cases all the There's consistency-a lot of it is when cases are
carry a generic case
way through the system A lot is lost when a transferred from one person to another there may
load-vertical case
case is transferred from worker to worker.
be different interpretations of the situation.
management
Families always know who their social
Effective with overall population-it would be
workers are and there is ownership
unfair to say that what we are doing is more
throughout the life of the case. There is a
advantageous with group than another.
better work product. There's a relationship
between the worker and clients.
161
T a bl e 59 0 rgamzat1ona or structural practices: Location, tan et population and goals (N = 14)
Frequency (Percentage)
Where does practice #1 take place?
Throughout the child welfare system
(57.1%)
8
County child welfare office
7
(50.0%)
(42.9%)
6
• Birth family's home
Foster
family's
home
(42.9%)
6
•
(42.9%)
6
• Court
(35.7%)
5
• Another out-of-home placement site
(14.3%)
2
• Contract agency site
5 (35.7%)
• Other
Who is the target population for practice #I?
12
(85.7%)
• Children of all ages
(7.1%)
I
• Very young children (0 to 3 vears)
(7.1%)
I
• Young children (4 to IO years)
(50.0%)
7
Families
of
any
ethnicity
•
(28.6%)
4
• Latino families
(21.4%)
3
• American Indian families
(50.0%)
7
• Families with substance abuse problems
(50.0%)
7
• Families with a mental health problem
(42.9%)
6
Families
with
domestic
violence
•
4
(28.6%)
• Families with income problems
(21.4%)
3
• Families with a disability
(7.1%)
I
• Other
What are the purposes and goals of practice #I?
(50.0%)
7
• Prevent out-of-home placement
(50.0%)
7
• Ensure a permanent placement for the child
(42.9%)
6
Reunify
children
with
their
birth
families
•
(42.9%)
6
• Address substance abuse problems of family
(42.9%)
6
• Improve parenting skills
(35.7%)
5
• Assess level of risk
5 (35.7%)
• Improve psychosocial functioning of family
(35.7%)
5
Address
mental
health
problems
of
family
•
(35.7%)
5
• Address housing problems of familv
(28.6%)
4
Address
socioeconomic
problems
of
family
•
(57.1%)
8
• Other
•
•
162
Table 60: Organizational or structural practices: History of practice (N = 14)
Frequency (Percentage)
Why was practice #1 first implemented?
(50.0%)
7
• In response to a particular problem
(21.4%)
3
• New policies were implemented
1
(7.1%)
• Funds became available
(42.9%)
6
• Other
Who first initiated practice # 1?
(50.0%)
7
• Director of the child welfare system
6
(42.9%)
Child
welfare
workers
•
(7.1%)
1
• Clients of the CWS
(7.1%)
1
• Foster families
1
(7.1%)
• Criminal justice system
(7.1%)
1
• Contract agencies
(28.6%)
4
Other
When was practice # 1 first implemented?
(7.1%)
1
• 2002
2 (14.3%)
• 2000
(7.1%)
1
1999
•
2 (14.3%)
• 1998
(7.1%)
1
• 1995
1
(7.1%)
• 1980
1 (7.1%)
1972
Has practice #1 changed significantly between the time it
was first implemented and now?
(46.2%)
6
• Yes
•
•
163
T able 6 1: 0 rgamzational or structural practices: Factors related to implementation (N = 14)
Frequency (Percentage)
Current implementation status of practice # 1
12 (92.3%)
• Permanent
(7.7%)
1
• Other
What is the primary funding source for practice # 1?
(57.1%)
8
• Federal funds
(42.9%)
State
funds
6
•
(35.7%)
County funds
5
(28.6%)
4
Other
•
Is practice # 1 mandated?
(61.5%)
8
Yes
Are social workers provided with training in how to
implement practice # 1?
(66.7%)
8
• Yes
Do you contract with another agency or system to deliver
practice #1?
(64.3%)
9
• No
(21.4%)
3
Combination
of
internal
and
external
resources
•
(14.3%)
2
• Yes
Is practice #1?
(92.9%)
13
• Widespread throughout the child welfare system
1
(7.1%)
Targeted
to
a
relatively
small
group
of
clients
•
Is there a specific geographic area or community in which
this practice is targeted?
(7.1%)
1
• Yes
Has practice # 1 been evaluated to determine its
effectiveness?
(28.6%)
4
• Yes
If yes, what were the results of the evaluation?
(66.7%)
2
• Effective with target population
Evaluation did not provide evidence that practice was
(33.3%)
1
effective
If practice # 1 has not been evaluated, why not?
(35.7%)
5
• An evaluation was not considered to be needed
(21.4%)
3
• Insufficient resources
(7.1%)
I
Insufficient
time
•
(7.1%)
I
• Other
•
•
•
164
Table 62: Organizational or structural practices: Factors that support practice and barriers to
successful implementation (N = 14)
Frequency (Percentage)
What are the major factors that support practice # 1
11 (78.6%)
• Support from child welfare workers
9 (64.3%)
• Support from manag;ement
8 (57.1%)
• Support from county government officials
(50.0%)
Support
from
families
and
children
7
•
7 (50.0%)
• Support from the community
6 (42.9%)
• Support from foster families
6 (42.9%)
• Support from state R;ovemment officials
5 (35.7%)
• Support from the mental health system
4
(28.6%)
Support
from
the
substance
abuse
system
•
4 (28.6%)
• Support from contract agencies
(28.6%)
4
Support
from
the
health
system
•
4 (28.6%)
• Stable funding
3 (21.4%)
• Support from other out-of-home placement providers
2 (14.3%)
• Evaluations show that practice is successful
2 (14.3%)
• Support from the criminal justice system
2 (14.3%)
• Other
What have been the barriers to successful implementation of
this practice?
4 (28.6%)
• No barriers
2 (14.3%)
• Insufficient support from the community
2 (14.3%)
• Insufficient support from child welfare workers
(7.1%)
1
• Funding is not stable
7 (50.0%)
• Other
165
Summary of Key Findings and Implications:
Descriptive Results
Respondents to the California statewide survey identified practices in four overall areas
as most effective for children and families of color: 1) beginning choice point practices, 2)
continuing or ending choice point practices, 3) both beginning and continuing/ending choice
point practices, and 4) organizational or structural practices.
Beginning Choice Point Practices
The beginning choice point practices identified by respondents fell into three categories:
1) family group conferencing or other team decision-making practices, 2) voluntary family
maintenance or other prevention/family preservation services and 3) risk assessment tools or
systems.
In general, results from the California statewide survey indicate that beginning choice
point practices are most often delivered at the county child welfare office, and they are targeted
to children of all ages and ethnicities. Most of these practices are aimed at preventing an out-of-
home placement, assessing level of risk or improving the psychosocial functioning of the family.
Over half of these practices have been evaluated by individual counties to determine their
effectiveness, and of those that have been evaluated most are considered to be effective with the
target population.
Results of the California statewide survey indicate that beginning choice point practices
are considered to result in enhanced outcomes for children and families of color because they
are inclusive, collaborative and objective. Yet very little empirical evidence exists to support the
effectiveness of these practices for children and families of color. No studies have evaluated the
impact offamily group conferencing or other team decision-making practices on children and
families ofcolor. However there is some evidence to suggest that inclusive practice, in which
parents are engaged in placement and other decisions, may result in more placement stability for
children, yet possible differential effects for diverse populations has not been researched
(Palmer, 1996). More res~arch is needed to determine if family group conferencing/other team
decision-making practices and inclusive practice result in enhanced outcomes for children and
families of color.
Additionally, the effectiveness of voluntary family maintenance or other prevention/family
preservation services in preventing out-of-home placements has not been adequately
demonstrated, and the impact ofthese services for children and families of color has not been
sufficiently addressed in the research literature. It is interesting to note that results from the
California statewide survey indicate that only 15.4 percent of beginning choice point practices
take place in the birth family's home, yet the research literature suggests that home-based
prevention practices may be the most effective voluntary family maintenance practice, especially
for high-risk mothers. This finding may suggest that counties are not utilizing evidence-based
prevention efforts. Research on the reasons for this underutilization of home-based prevention
services is warranted.
166
Lastly, the effectiveness ofrisk assessment tools or systems in accurately assessing risk
for future maltreatment has not been consistently demonstrated in the research literature. Yet
research does indicate that the individual decision-making practices ofchild welfare workers are
inconsistent and subject to biases that may negatively impact children and families ofcolor. As
such, the use of more objective assessments of risk in child welfare is promising, however more
research is needed to identify risk assessment tools and systems that are reliable and valid.
Continuing Care/Ending Choice Point Practices
The continuing care/ending choice point practices identified by respondents fell into
three categories: 1) placement-related practices, 2) the Family to Family Initiative, and 3)
intensive and collaborative interventions for high-risk children.
Results from the California statewide survey indicate that, in general, the majority of
continuing care/ending choice point practices take place throughout the child welfare system
and are targeted to children of all ages and ethnicities. However, more continuing care/ending
choice point practices are targeted to specific age groups than the other practices identified by
respondents. The goal of most of these services is to reunify children with their birth families,
ensure a permanent placement or prevent and out-of-home placement. Only two respondents
reported that these practices have been evaluated to determine their effectiveness and both of
. these evaluations indicated that the practices are effective with the target population.
These continuing care/ending choice point practices are described as effective with
children and families of color because they are culturally appropriate, collaborative, intensive
and they involve clients' communities. Yet, research on the effectiveness of these services for
children and families of color is lacking. Placement-related practices that emphasize placement
of children in culturally appropriate homes have not been evaluated to determine their impact on
children of color. However, some studies have noted that social workers may be reluctant to
place children of color with White alternate caregivers (Brooks & James, 2003; Carter-Black
2002; Vidal de Haymes & Simon, 2003). Results from the California statewide survey regarding
the identification ofplacement-related practices that emphasize placement ofchildren of color in
culturally appropriate homes do suggest that culture and ethnicity may be important factors in
child welfare workers 'placement decisions. This appears particularly true for Native American
children who were mentioned specifically by two counties.
Research on the effectiveness ofthe Family to Family Initiative for children andfamilies
ofcolor has not been conducted. There is limited evidence to suggest that social network
interventions which seek to involve community members in services are potentially effective in
increasing social networks, improving parenting skills and increasing the likelihood of case
closure (Gaudin et al., 1990). However the Family to Family Initiative encompasses more than
social network interventions and research is needed to evaluate the effectiveness ofthis practice
for diverse populations.
Although children in out-of-home care are at an increased risk for a variety ofpoor
outcomes (Stevenson, 1999), the impact of intensive and collaborative interventions for high-risk
children have demonstrated only mixed results in the research literature. While some studies on
167
wraparound services report that children in these services are less likely to be removed from their
community and may experience a reduction in behavioral problems (Bruns, Burchard, & Yoe,
1995, as cited in Toffalo, 2000; Clark, Lee, Prange, & McDonald, 1996, as cited in Toffalo,
2000; VanDenBerg, 1993, as cited in Toffalo, 2000), other studies report no significant benefit
from wraparound services (Clarke, Schafer, Burchard, & Welkowitz, 1992, as cited in Toffalo,
2000). More research is needed to identify aspects ofthese interventions that are most effective,
especially for children ofcolor
Both Beginning and Continuing/Ending Choice Point Practices
Practices related to both beginning and continuing/ending choice points fell into two
categories: 1) community-based services, and 2) strength-based assessments.
Results from the California statewide survey suggest that many practices that are both
beginning and continuing/ending choice point practices take place at a contract agency site or
the birth family's home. These services are targeted to children of all ages and any ethnicity.
These practices have numerous goals including: preventing an out-of-home placement;
reunifying children with their birth families; assessing level of risk; improving psychosocial
functioning of families; improving parenting skills; addressing substance abuse problems;
addressing mental health problems; addressing socioeconomic problems; and addressing housing
problems. All four of the respondents indicated that these practices have been evaluated and 50
percent of the evaluations indicated that the practices are effective with the target population.
Respondents to the California statewide survey described these practices as effective with
children and families ofcolor because they are easy for clients to access; they are communityspecific, and non-intrusive. These services have broad goals that focus on both prevention and
intervention for families in contact with the CWS and research on the impact of these services on
children and families of color is lacking. In general, research on prevention and intervention with
child welfare populations does not provide convincing evidence of their effectiveness. However,
because these practices are more of an approach to services, rather than services themselves,
more research is needed to determine if these approaches are effective with ethnically diverse
populations.
Organizational or Structural Practices
Organizational or structural practices fell into three categories: 1) cultural competency
of staff, 2) collaboration with American Indian tribes, and 3) vertical case management.
According to responses from the California statewide survey, most
organizational/structural practices take place throughout the CWS; the majority are targeted to
children ofall ages and ethnicities, although some are specifically targeted to Latino or Native
American clients. The primary goals of these practices include preventing an out-of-home
placement and ensuring a permanent placement. Approximately one quarter of respondents
indicated that these practices have been evaluated to determine their effectiveness and of these,
approximately 66 percent reported that the practices are effective with the target population.
168
These practices are described by respondents to the California statewide survey as
effective with children and families ofcolor because they result in better communication and
relationships between worker and client; they respect cultural heritage; increase cultural
awareness and produce more accurate and complete assessments and services. Although
cultural competency appears intuitively beneficial for ethnically diverse children and families,
research on the impact of culturally competent staff and programs in producing successful
outcomes for children and families of color has not been conducted. There is evidence to suggest
that for Latino clients in particular, language barriers can exclude them from necessary services
and produce significant delays in service delivery; and researchers have also discussed the
importance of language in accurate assessment and treatment of clients (Suleiman, 2003). As
such, culturally and linguistically appropriate services appear to increase service use and
accuracy, although their impact on client outcomes is unclear.
Research on the effectiveness ofcollaboration with American Indian tribes and vertical
case management practices on outcomes for children and families of color is limited. Effective
collaboration between the CWS and American Indian tribes appears intuitively beneficial for
Native American children, however more research is needed to identify aspects ofeffective
collaboration and how this collaboration impacts outcomes for Native American children. A
vertical case management model of child welfare service delivery may result in enhanced
outcomes. For instance, in South County Offices, where a vertical case ·management model is
used, families are more likely to have their cases closed with family maintenance services than
they are in the Main County Office where a vertical case management model is not used (please
see Section IV for a full discussion). These results suggest that a vertfoal case management
model may be an effective child welfare practice.
Limitations and Conclusions
Descriptive results of the statewide survey must be considered within the context of
certain limitations. For instance, the survey instrument was designed specifically for this study
and so its reliability and validity is not known. The self-report nature of the data also presents the
possibility of personal biases in survey responses and since the professional titles of the
respondents varied from directors, to managers to supervisors, there is the possibility that
different perspectives and information may occur as a result of these differing job titles.
Additionally, although the response rate was high (77.6%), there is the possibility that the sample
may be biased due to some counties refusing to participate.
Overall, respondents described beginning choice point practices as effective with
children and families ofcolor because they are inclusive, collaborative and objective.
Continuing care/ending choice point practices were considered effective with children and
families of color because they are culturally appropriate, collaborative, intensive and involve
client's communities. Both beginning and continuing/ending choice point practices were
described as effective because they are easy for clients to access, they are community-specific
and non-intrusive. Lastly, organizational or structural practices were considered effective with
children and families of color because they are thought to result in better communication and
relationships because worker and client, they respect cultural heritage, increase cultural
awareness and produce more accurate and complete assessments and services.
169
In general, results from the California statewide survey suggest that there is a significant
gap between research on child welfare practices and the best practices for children and families
of color being used in California county child welfare systems. Little systematic research has
evaluated the impact of many of the most effective practices for children and families of color
that were identified by respondents Although the research literature has not yet provided
convincing evidence of the effectiveness of many ofthe practices identified by respondents,
respondents to the statewide survey felt that these practices were their most effective with
children and families ofcolor. These findings suggest that some counties may be utilizing
promising practices for children and families ofcolor in the child welfare system, and more
rigorous evaluations ofthese practices is needed.
·
170
Multivariate Analysis:
Impact of Statewide Practices on Differential Representation
The statewide survey identified practices, which addressed disproportionate
representation of children of color and were considered effective by county child welfare
personnel. Although perceived as effective, few of the practices had been evaluated to determine
effectiveness from an independent point of view and in a controlled manner. Child welfare
personnel nominated them as being effective, but in most cases their impact had not been
established through systematic empirical research.
For present purposes, the impact of interest is how the practices identified through the
survey might affect differential representation of children ofcolor. Some of the practices had
been implemented with children of color in mind and others had not, but all were considered by
child welfare personnel to be especially valuable in addressing the needs and circumstances of
children of color. If child welfare personnel are correct in their belief that the practices serve
children of color particularly well, then it is reasonable to believe that the practices identified in
the survey will divert children of color from the child welfare system or will reduce the scope or
duration of their involvement.
By helping children of color to follow pathways around full-fledged involvement in the
child welfare system or reducing the intensity or length of child welfare participation, the
practices identified in the survey can be expected to reduce differential representation. We can
expect lower levels of overrepresentation if the practices are effective.
Statewide data are available from sources, which, in conjunction with data from our
survey, permit the degree of empirical association to be determined between practices and
disproportionate representation. The U.S. Census and CWS/CMS publish data, available for use
by researchers and others, describing California counties on the worldwide web. The web-based
data provide an opportunity to evaluate whether there is a link between disproportionate
representation and use of perceived-to-be-effective practices.
If a study is statewide in scope like the present one, then it furnishes a basis to go beyond
the experience of a single county. By comparing groups of counties and contrasting those that
implemented a practice with those that did not, our investigation encompasses child welfare
systems around the state. The results point to promising practices identified from a statewide
perspective.
In this section, we present results from several analyses linking characteristics of
practices presented previously with disproportionate representation of children of color in the
child welfare system. Our intention was to identify types of practices, practices, and practice
characteristics-distinctions to be explained later-- associated with levels of disproportionate
representation, and which therefore.might be more effective in reducing disproportionate
representation of children of color.
171
Multivariate Methods
The unit of analysis for this study was the county. The method, broadly conceived, was to
conduct controlled tests of empirical associations between types of practices, practices, and
practice characteristics on the one hand, and levels of disproportionate representation on the
other hand.
For thinking about the study's approach in greater detail, it is useful to begin by
considering the meaning of disproportional representation. Disproportional representation is
said to occur when the proportion of children of color in a county's child welfare system is
higher or lower than expected. Typically, documenting the proportion of African Americans,
Latinos, or Asian A~erican/Pacific Islander in the population at large usually establishes the
proportion that is expected. The difference between the representation of a group in the
population at large and its representation in the child welfare system defines its level of
disproportional representation.
In our study we calculated a measure of this difference using regression analysis. We
analyzed a group's representation in the child welfare system after statistically adjusting for it
representation in the population at large. Following a research design like ours, a statistic results
from regression which defines the difference between expected and actual levels of child welfare
representation.
Disproportional representation is affected by factors other than those under the direct
control ofchild welfare authorities. The stresses under which minority families live and the
resources available to them for example, can contribute to disproportionate representation.
Counties can differ in the levels ofstress, resources, and similar differentiating characteristics.
Child welfare systems too are diverse. Some are larger and more centralized than others, for
example; policies and procedures vary.
It is difficult to establish definitively whether practices are effective because of the
differences apart from the practices between counties and child welfare systems associated with
disproportionate representation. When a link is observed between a practice and
disproportionate representation, it remains possible that a characteristic other than the practice
itself is responsible for that link. The observed link between a practice and disproportionate
representation might be coincidental. Controlling/or key differences between counties and child
welfare systems reduces the chances ofassociations occurring through coincidence.
We examined many indicators describing counties and their child welfare systems in
order to identify differences that were especially important to take into account. Among
indicators we screened were population size and density, wages and costs of goods and services
in the county, the geographical location of the county, size of the child welfare system and
proportion of the child welfare population in kincare. We sought key indicators from a large
number we considered reflecting differences between the counties and their child welfare
systems and linked to disproportionate representation.
172
Through this process we found two indicators that were consistently and strongly
associated with disproportionate representation: urban-rural status and per-capita income.
Counties that were urban and had higher per-capita income were notably more likely to have
higher levels ofdisproportionate representation.
We used these data to conduct a controlled analysis of the association between practice
types, practices, and practice characteristics by constructing regression equations. In each
equation, the proportion of Latino, African American, or Asian American/Pacific Islander
children served as the dependent variable. The proportion of Latinos, African Americans, or
Asian American/Pacific Islander in the county was entered first as an independent variable, as
discussed above, in order to subtract an expected level of representation from the actual
representation of each group in each county's child welfare system. We then added to the
equation e~ch county's urban-rural status and per-capita income. Finally, we added each practice
type, practice, or practice characteristic suitable for analysis.
This enabled us to determine how much each practice type, practice, or practice
characteristic was associated with greater or lower level of representation apart from the level of
representation expected from its group representation in the county, and after adjusting for the
county's urban-rural classification and its per-capita level of income. Thus, the approach
adjusted for the three key factors: group representation in the population, whether the county was
classified as urban or rural, and per-capita income of county residents.
In most cases the three control variables alone explained disproportionate representation
in a statistical sense extremely well. Success in accounting for differences between counties in
statistical models implies that many important differences between counties were explained and
that the possibility that associations arose by coincidence was low.
We considered practice types, practices, and practice characteristics in a three-phase
analysis. Phase 1 addressed our organization ofpractices into beginning choice point practices,
continuing care/ending choice point practices, organizational or structural practices, and both
beginning and continuing/ending choice point practices. We considered disproportionate
representation according to whether practices were beginning, continuing/ending, organizational
or structural, or both beginning or ending.
In phase 2, we considered individual practices in an attempt to consider the impact of
practices on disproportionate presentation in finer-grained detail. We were hampered in our
attempt: very few counties reported instituting certain of the practices.
Ultimately, we chose to consider at the individual practice level only those where four or
more counties indicated implementing them. Four counties is a small number for this purpose
and limit our ability to detect genuine differences that might have been present at statistically
significant levels. It proved fruitful however, and provides an important basis for di~cussion of
individual practices. The five practices we considered were: family group conference or other
team decision making practices (n=7), voluntary family maintenance or other prevention/family
preservation services (n=4), placement related practices (n=6), family to family initiative (n=4),
and cultural competency of staff (n=l 0).
173
We asked parallel questions in all phase 1 and phase 2 analyses. Adjusting for urbanrural status and per-capita income, for Latinos, African Americans, and Asian Americans how is
disproportionate representation affected by type of practice? What was the impact on
disproportionate representation ofwhether a practice was one of.five specific practices:family
group conference or other team decision making practices, voluntary family maintenance or
other prevention/family preservation services, placement related practices, family to family
initiative, and cultural competency ofstaff?
In phase three ofour analysis we asked questions without regard to type ofpractice. We
considered the impact on disproportionate representation-whatever the particular practice
type-- of differences in practice location, target population, purposes, reasons for
implementation, initiating persons and agencies, implementation timing, implementation status,
location and scope of the practice, geographic and community targeting, evaluation, support,
and barriers.
There are many practice characteristics describing practices. To reduce the number of
characteristics for purposes of analysis, we clustered characteristics into logically related,
empirically associated groupings using factor analysis, a statistical procedure for clustering
individual questions to which respondents have given similar answers. When a question was
discovered not to fit well within factor-analysis created clusters of questions, we eliminated it
from the cluster and considered it in subsequent analysis individually.
We then conducted regression analyses like those conducted in phases I and 2. The
question was: adjusting for urban-rural status and per-capita income, for Latinos, African
Americans, and Asian Americans, how is disproportionate representation affected by practice
characteristics? What was the impact on disproportionate representation of differences in
practice location, target population, purposes, reasons for implementation, initiating persons and
agencies, implementation timing, implementation status, location and scope of the practice,
geographic and community targeting, evaluation, support, and barriers?
There are two ways in which practice types, practices, and practice characteristics can
be linked to disproportionate representation. One is a negative association between the
variables, indicating lower than expected overrepresentation. Negative association means that
implementing counties have relatively low or nonexistent minority overrepresentation and
suggests that the practice type, practice; or practice characteristic is successful.
Another kind of association between a practice type, practice, or practice characteristic
and overrepresentation is a positive association. Positive association between variables means
that there is greater than expected overrepresentation associated with the practice type, practice,
or practice characteristic-that overrepresentation is relatively high.
Practices linked to greater overrepresentation cannot be considered successful. How does
a practice come to be linked to greater overrepresentation? It is likely that positive association
between a practice and overrepresentation comes about because instead of reducing
overrepresentation, some practices are implemented reactively, that is, in response to
174
overrepresentation. These practices do not reduce overrepresentation, although they may do so in
the future if given sufficient time. At present; we note only that a significant positive association
with overrepresentation is probably reactive and does not indicate success.
Multivariate Results
Latino Overrepresentation
Latinos were overrpresented in about 60% of counties and the degree of their
overrepresentation was as high as 28 percentage points.
Practice types. For Latinos, there was a statistically significant difference according to
whether practices were beginning choice point practices, continuing care/ending choice point
practices, organizational or structural practices, or both beginning and continuing/ending choice
point practices.
Both significant differences involved beginning choice point practices. Beginning choice
point practices were associated with about 8 percentage points less overrepresentation than both
beginning and continuing/ending choice point practices and about 9 percentage points less
overrepresentation than organizational or structural practices (See Table 63).
Practices. One statistically significant difference emerged when we considered the five
individual practices that could be analyzed. Practices involving family group conference or other
team decision making were associated with 7 percentage points less overrepresentation than were
practices involving culturally competency of staff (See Table 63).
Practice characteristics. For Latinos, there were statistically significant difference in
target population, purposes, initiating persons and agencies, and implementation timing (See
Table 64). Latino overrepresentation was greater when areas were targeted for intervention (area
where most referrals came from; with low socioeconomic status residents; with high crime; with
large ethnic representation), when clients, foster families, and out of home placement providers
were involved in the development of the practice, when service systems were involved in the
development of the practice (mental health, substance abuse, health, criminal justice, contract
agencies). Overrepresentation was less also when "other" problems were addressed as the
purpose of the practice and when it began more recently.
African American Overrepresentation
African Americans were overrepresented in about 64% of counties and their degree of
their overrepresentation was as high as 65 percentage points.
Practice types. For African Americans, there was no statistically significant difference in
overrepresentation according to whether practices were beginning choice point practices,
continuing care/ending choice point practices, organizational or structural practices, or both
beginning and continuing/ending choice point practices (See Table 63).
175
Practices. One statistically significant difference appeared when we considered the five
individual practices, which could be analyzed. Family to Family initiatives were associated with
8-percentage points greater overrepresentation than were the remaining practices considered (See
Table 63).
Practice characteristics. For African Americans, there were differences in target
population, purposes, and initiating persons and agencies, and geographic and community
targeting. African American overrepresentation was less when areas were targeted for
intervention (area where most referrals came from; with low socioeconomic status residents;
with high crime; with large ethnic representation), ethnic minority populations (Latino, African
American, Asian American/Pacific Islander, American Indian, families of any ethnicity), and
when very young children were targeted. African American overrepresentation was less when
foster families and clients or contract agencies first initiated the practice. African American
overrepresentation was less also when the purpose the intervention was to improve the
psychosocial functioning of the birth family (See Table 64).
Asian American/Pacific Islander Underrepresentation
In no counties were Asian American/Pacific Islanders overrerpresented. Instead, they
were equally represented in a few counties and in most counties Asian American/Pacific Islander
were underrepresented.
Practice types. For Asian American/Pacific Islanders, there was no statistically
significant difference in overrepresentation according to whether practices were beginning
choice point practices, continuing care/ending choice point practices, organizational or structural
practices, and both beginning and continuing/ending choice point practices (See Table 63).
Practices. Nor was there a statistically significant difference when we considered the five
individual practices that could be analyzed (See Table 63).
Practice characteristics. For Asian American/Pacific Islanders, there were differences in
target population, implementation timing of the practice. Asian American/Pacific Islanders under
representation was greater when areas were targeted for intervention (area where most referrals
came from; with low socioeconomic status residents; with high crime; with large ethnic
representation), ethnic minority populations (Latino, African American, Asian American/Pacific
Islanders, American Indian, families of any ethnicity). Asian American/Pacific Islander under
representation was less when system management and child welfare workers expressed support
for the intervention, and when it had begun earlier (See Table 64).
176
Table 63: Practice Types, Practices, and Disproportionate Representation
pract1ce
. Types
Ethnic Group
Variable
B SE Beta
T
Latino
Beginning vs
.09 .04 .14
1.98*
Organizational/
Structural
Beginning vs
.08 .03 .20 2.61 **
Beginning/Continuing
Practices
Variable:
B
Family Group Conference/ .07
Team Decision Making vs
Cultural Competence
African American Family to Family vs. other -.08
Ethnic Group:
Latino
SE
.03
Beta T
.15
2.13*
.04
.13
2.23*
Table 64: Practice Characteristics and Disproportionate Representation
Ethnic Group:
Latino
Variable:
Area Targeting
Target:Other Problems
Clients Involved in
Development/Evolution
Systems Involved in
Development/Evolution
Initated Earlier
African American Ethnic Targeting
Area Targeting
Targeting: Very young
Foster family-client
Initiated
Contract agency
Initiated
Goal: Improve psychoSocial Functioning Birth
Family
Asian American/ Ethnic Targeting
Pacific Islander
Area Targeting
Management-Worker
Support
B
.04
-.05
.03
SE Beta T
.01 .19
3.07**
.26 -.13 -2.06*
2.77**
.01 .17
.03
.01
.16
2.61 **
-.00 .00 -.16
-.04 .01 -.23
-.02 .01 -.12
.06 .03 .11
-.08 .04 -.11
-2.69**
-4.20**
-1.93*
1.92**
-1.94**
-.08 .04 -.13
-2.11 *
-.05
.02 -.13
-2.11 *
.00
.00
.00
-.01
.00 .19
.00 -.29
.22
1.99*
1.90*
-3.24**
Notes:
Regressions: Ordinary Least-Squares. All models R >.80 (p>.01).
Factors = Principal Components, 1 factor per analysis, Eignenmvalue> 1. Regression-based factor scores used in
subsequent regression. Variables with loadings >.50 were eliminated from analysis and analyzed separately in
subsequent regression.
177
Summary of Key Findings and Implications:
Multivariate Analysis
When considering the 43 counties responding to the survey of practices perceived
effective in working with children of color, significant associations were uncovered between
practice types, practices, and practice characteristics on the one hand, and disproportionate
representation on the other. These associations remained after adjusting for urban rural status and
per-capita income, differences between the counties strongly linked to minority child welfare
representation. Some point to effective practices-practices that bring about minority
underrepresentation. Different patterns ofassociation were found for Latinos, African
Americans, and Asian American/Pacific Islanders.
For Latinos, beginning choice point practices had a beneficial effect on
overrepresentation. In particular, family group conference or other team decision-making
interventions predicted lower levels ofoverrepresentation.
Bringing together interested parties early on in child welfare involvement may divert
cases from child welfare or otherwise minimize involvement, perhaps by facilitating
communication and mobilizing otherwise overlooked resources. Conceivably, the intervention is
especially valuable for Latinos in that its family focus is congruent with cultural values, and
because of the intervention's potential to overcome linguistic and cultural barriers to
understanding.
Notably, these interventions do not target at-risk neighborhoods as a special focus of
concern. Indeed, Latino overrepresentation was greater, not less, when interventions were
implemented targeting areas with high concentrations ofdistressed families.
Other findings indicated that practices targeting at-risk populations came about when
clients, foster families, and out of home placement providers were engaged in development of
the practice, along with mental health, substance abuse, health, criminal justice, contract
agencies. It appears that when they perceived a need and had an opportunity to do so,
stakeholders mobilized and focused on high need areas and populations as intervention targets.
However, they appear do this in response to Latino ove"epresentation, such that
implementation is linked to greater overrepresentation. The positive association at-risk targeting
and disproportional representation indicates that these practices do not reduce disproportional
representation. In all likelihood, the positive association means instead that stakeholders
perceive Latino overrepresentation and initiate programs in response.
For African Americans, targeting ofat-risk areas and African American populations was
beneficial in reducing overrepresentation, as well as targeting ofvery young children. Less
overrepresentation was linked also to whether foster families and clients and contract agencies
initiated an intervention, and whether it focused on improving psychosocial outcomes of birth
families. The findings suggest a picture almost the·reverse of that found for Latinos.
178
For African Americans, there appear to be lower levels ofoverrepresentation than
expected too when key community stakeholders help to initiate the interventions that focus on
high-risk areas and populations. When interventions have a goal of strengthening the well being
of birth families, then implementation is associated with lesser overrepresentation.
On the other hand, counties that implemented family-to-family initiatives displayed
greater overrepresentation than counties that did not. Family to family programs appear to be
used reactively, in response to high perceived levels of overrepresentation.
It is important to note however that in counties where interventions were successful
levels of overrepresentation remained high. Successful interventions were those associated with
lower overrepresentation than expected. Improvement was relative: in absolute terms there
remained a high degree of African American overrepresentation.
For Asian American/Pacific Islander populations findings suggested that, as with
Latinos, targeting of high need areas was less beneficial than not engaging in this kind of
intervention. Involving clients and service systems in initiation, again more common for
interventions targeting high need areas, was also relatively unsuccessful. On the other hand,
management and worker support for a practice was associated with greater success.
Asian American/Pacific Islanders are, overall, underrepresented in child welfare systems
and one must interpret our findings in light ofthis fact. There were fewer significant differences
between practice types, practices, and practice characteristics for Asian American/Pacific
Islanders than for Latinos or African Americans. This may reflect a lesser sense of urgency felt
by administrators and workers in the face of Asian American/Pacific Islander disproportionate
representation that is proportionally low than for Latino and African American disproportionate
representation, which is proportionately high.
A need remains to conceive and mount culturally and linguistic programs for Asian
American/Pacific Islander populations. This aspect of the study, with its focus on
disproportionate representation rather than on other indicators of system performance, may not
have been as good a vehicle as other parts of the study to identify program needs for Asian
American/Pacific Islander populations as for overrepresented Latino and African American
populations.
Certain limitations must be kept in mind when considering the present findings. The
study encompasses forty-three California counties. The number of counties implementing any
practice or practice type was smaller than ideal to undertake truly powerful statistical tests
sensitive to all underlying differences. As a matter of practical necessity, the study's design was
such that some of the links of practice types, practices, and practice characteristics with
disproportionate representation might have arisen by coincidence. The possibility that some
unmeasured characteristic explained some of the associations of interest revealed in the study
was reduced, but cannot entirely be ruled out.
Despite these limitations, the study produced meaningful patterns of findings connecting
programs and disproportionate representation. The associations proved sensitive to the
179
differential experience of different ethnic groups. They furnish a useful context of fact to further
consider the problem of disproportionate representation in Santa Clara County.
180
VI. PHASE 3 SUMMARY, RECOMMENDATIONS AND
CONCLUSION
Phases 1 and 2 of this study focused on examining reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where it occurs. The
focus ofPhase 3 was on key practices that take place within the CWS that either promote or
reduce disporportiona/ity among children and families ofcolor.
Overall, research on the impact of child welfare practices for children and families of
color is limited. In general, much of the research literature is inconclusive and fails to support the
effectiveness of child welfare practices i'n improving outcomes for children and families involved
in the CWS. Very few studies describe differences in outcomes by ethnic/racial groups and most
of the research fails to take into account the effect of culture and ethnicity on the delivery and
impact of child welfare practices. This gap in the research literature is striking and has important
implications for children and families of color in the CWS. More information about child welfare
practices _including programs, services, strategies, policies, and/or tools that either increase or
reduce racial/ethnic disproportionality can contribute to the development of more effective
services and supports for children and families of color in the CWS.
The three overall aims addressed in Phase 3 were: 1) to assess identified key practices for
children and families ofcolor in Santa Clara County's CWS, 2) to examine the influence ofthe
court system on service recommendations for children and families and, 3) to conduct a
statewide comparative analysis of identified key practices that may affect the disproportionate
representation of children ofcolor in the CWS.
In order to address Aim 1 which involved an assessment of key practices in Santa Clara
County's CWS, four related analyses were conducted. First, the CWRT conducted analyses to
understand the types of practices and case characteristics that are associated with successful
outcomes for children and families from various racial/ethnic groups. In order to assess
successful outcomes, we used data from our Phase 2 Case Record Review sample and defined
success according to US DHHS guidelines, i.e., safety, permanency and family and child wellbeing. Results indicated that there were very few ethnic differences in successful outcomes.
In regards to safety, there was a significant relationship between the number of times
removed from family in current episode and ethnicity where African American children had the
highest average number of times removed from family in current episode and Asian
American/Pacific Islander children had the lowest average. In terms of family and child wellbeing, there was a marginally significant relationship between families' capacity to provide for
their children's needs and ethnicity, with African American families appearing to have a more
enhanced capacity while Asian American/Pacific Islander families seem to have a more
diminished capacity.
It is important to note that major ethnic differences do exist in Santa Clara County's
CWS. Phase 1 and 2 confirmed that certain ethnic groups are over-represented (i.e., Latinos,
African Americans) while others are under-represented (i.e., Asian American/Pacific Islanders)
in the county's CWS. We also found that there are significant ethnic differences in many
181
demographic, system-related, and psychosocial characteristics. Many of these variations appear
at different choice points in the system (such as psychosocial status at the beginning of the case,
and assignment to voluntary family maintenance). Results ofour successful case outcomes
analysis indicate that ethnic differences at the conclusion ofthe case are not evident.
While results suggest there is a clear disproportion of ethnic groups in the system, once in
the system, the children and families are generally faring the same. Their relatively
homogeneous experience may be a result of the "one-size-fits all" limited array of services
offered. Although this may seem equitable and logistically efficient, it appears ultimately
ineffective in serving a diverse group of children and families. Also, to determine better how
"well" these children are actually doing while in the system, the availability ofcomparative data,
clear operational definitions for "success," and more reiiable and valid data are needed.
Implications and Recommendations
•
In order to better assess whether safety, permanency, and family and child wellbeing are maintained and enhanced, records should include more complete and
measurable information. We acknowledge that child welfare case records were
not originally intended for research. Yet, if the agencies responsible for providing
effective services wish to evaluate the performance of their programs more
reliably and validly, improving information collection and tracking should be
considered.
•
The creation of a case summary check-out form used at case closure would be one
way to track child outcomes. The check-out form would indicate the child and
family's status in major areas of psychosocial well-being (i.e., health, education,
and finances), in addition to the existing case outcomes. Given the amount of
missing data assessing these domains describing case and child status, and the
relative scarcity of information recorded in the last two court hearings, this form
would enable measurement and evaluation of "successful cases and outcomes" in
a more consistent objective manner.
•
The amount of missing data and the manner in which information is assessed also
implies the need for more reliable and valid methods of recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their families or of
those adopted high or low? And compared to what? Implementing better record
keeping and data collection over time on key characteristics, based on an agreedupon operational definition would better address the question of effectiveness and
performance. Implementing these practices across counties and across states
would also facilitate evaluation.
Second, a qualitative analysis using both focus group and interview data was conducted
to obtain insights about current child welfare practices used within the DFCS, how these specific
182
practices might positively or negatively impact children and families of color, and what practices
might be created, enhanced or maintained to better serve children and families of color.
Results provided information on numerous practices and contextual factors that have
either a negative or positive impact on children and families ofcolor in Santa Clara County's
child welfare system. Study participants recommended the use of in-home services for children
and families of color. There is some evidence in the research literature that supports the
effectiveness of in-home services (see literature review), however these services appear to be
most effective when they are intensive, long-term and delivered by health or social service
professionals and when the workers delivering these services receive a high level of supervision
(Barth, 1991; McGuigan et al. 2003; Olds et al., 1997). Participants also identifiedfamily
conferencing as an effective practice. These study findings are consistent with limited research in
this area that suggests that inclusive practices, in which parents are engaged in placement and
other decisions, may result in more placement stability for children, although possible
differential effects for diverse populations has not been researched (Palmer, 1996).
Participants generally expressed optimism about the adoption ofteam decision making
(/'DM) as a vehicle for improving decision-making with diverse families, although some social
workers expressed concern about time required for implementation. Research suggests that TOM
can be labor intensive and can be difficult to implement effectively within the context of the
child welfare system (Sieppert et al., 2000). Finally, participants noted that both formal and
informal collaborative relationships with other services delivery systems, such as domestic
violence and substance abuse treatment systems, are critical for successful interventions with
children and families. Furthermore, participants recognized the importance of training and
cross-training to enhance the capacity of helping professionals in child welfare and other
systems to better address the needs ofchildren and families ofcolor.
Interestingly, most practices and contextual factors mentioned by social workers had both
positive and negative aspects. Many of the practices and contextual factors that negatively
impact children and families of color were actually barriers to successful implementation of the
practices that social workers felt have a positive impact on children and families of color.
Although practices such as orientation ofnew clients; family and team decision-making;
preventive services; substance abuse services; cultural competency; collaboration with other
agencies and systems; a strength-based approach; and the use ofrelative placements were
described as having a positive impact on children and families ofcolor, certain negative
practices and contexts impeded their effective implementation. For instance, having to screen out
large numbers of inappropriate referrals; inconsistency in decision-making practices; a shortage
of services, particularly substance abuse and preventive services; lack of client access to
services; difficult protocols for placing children with kin; gaps in cultural competency; time
limits; and agency-level factors such as heavy caseloads, staff shortages, substantial amounts of
paperwork, lack of access to information about resources; and confusion about the agency's
overall mission and key policies all impeded implementation of the best practices. These findings
suggest that in addition to building on current positive practices for children and families of
color, CWS stakeholders should also actively work both internally and in collaboration with
partners to reduce barriers to these best practices.
183
Indeed, in order to address these barriers to effective implementation of best practices,
study participants mentioned numerous recommendations that centered around certain key
themes. In general, there was an emphasis on improving decision-making through increased
accountability, training and cross-training to reduce bias, and increased use ofgroup decisionmaking or family involvement in decision-making. Study participants also stressed a need to
improve and expand prevention, diversion and concrete services, as well as an overall need to
expand culturally and linguistically competent services, and develop ways to improve
availability and access to services. Once a case is opened, social workers also felt it valuable to
provide an orientation to clients so that they are more aware of the CWS and court processes and
can better navigate the system. Recommendations related to organizational factors were also
noted, including reducing caseloads; increasing support and streamlining service delivery;
clarification of the mission ofDFCS, as well as key child welfare policies; fostering an
organizational culture that is strength-based and community-based, and more overall training
for social workers.
Implications and Recommendations
•
The findings from the qualitative component of this study affirm the value of
many efforts that are already in place, such as culturally specific ER response
units and family resource centers, and other that are in the process of
implementation including efforts to reduce caseloads, initiation of team decision
making, and participation in the Family-to-Family initiative.
•
Pract~ces that are inclusive, collaborative, culturally appropriate and involve
client's communities appear to be best suited to meeting the needs of children and
families of color. Although the research literature does not provide convincing
evidence of the effectiveness of these practices, the fact that respondents reported
these practices as those that are most effective with children and families of color
suggests that they may serve as promising models for children and families of
color and that more rigorous evaluations of these practices are needed.
•
The findings also point to opportunities to further strengthen services for children
and families of color. Recommendations related to training, institutionalizing
group orientation for clients, and other suggestions from study participants could
be used to inform practice and planning.
•
It would be beneficial for a team of managers, line staff, family representatives,
and community members to review the recommendations suggested by focus
group and interview participants to identify and prioritize possible practices for
adoption based on their feasibility and utility for children and families of color.
Third, in an effort to better understand the influence of contextual factors on child
welfare practices and outcomes for children and families of color, a comparison of Santa Clara
County's Main Offices and South County office was conducted. This comparison included both
quantitative and qualitative analyses. Quantitative data from CWS/CMS were used to examine
differences in case characteristics between the Main County Offices and South County Offices.
184
A qualitative comparison of focus group data from the Offices and South County Offices was
also conducted to discern how workers perceive practices in these two geographical locations.
These analyses help shed light on the potential impact of agency and community context on case
characteristics, child welfare practices and outcomes for children and families of color in Santa
Clara County's CWS.
Unfortunately, the South County location subsample was too small to allow analyses by
ethnic group. However, some key case characteristics are significantly related to county
location. Specifically, in South County Offices, a higher proportion ofcases are in voluntary
services, cases are shorter in duration, fewer workers are assigned during the course ofthe case,
and a higher proportion ofcases are in family maintenance at case closure. These findings are
also statistically noteworthy given the small subsample size available from SC. There were also
other differences in case characteristics by county location that did not attain statistical
significance but are noteworthy for further investigation. Children served in the South County
Offices appear to have fewer number of unique placement homes in the current episode, have a
shorter average stay per placement, a shorter length oftime in out-ofhome placement, and are
younger at time of case closure.
These findings imply that there is a difference in the style and quality of services
provided in the South County Office location. Qualitative findings are congruent with the
quantitative data and provide some insight into the dynamics that may contribute to this
difference. The higher number ofcases in voluntary services, the larger proportion ofcases in
family maintenance at case close, and the shorter duration ofcases is consistent with what South
County study participants described as a culture ofcommitment to maintaining and reunifying
families. This commitment was described as integral to the philosophy of managers and social
workers and reflected in the expectations communicated to social workers by most supervisors
and by peers. The practice of providing an orientation to the child welfare system, time limits,
the courts, and how to work with social workers may also contribute to the differences found in
the quantitative analysis. Study participants frequently described South County as similar to a
small county or rural area. This geographic difference, and the resulting sense of "connection to
the community, " was identified as a factor in the greater sense ofshared philosophy and
emphasis on prevention ofout-ofhome placements. Other studies suggest that region may
influence practice. For example, one study (Drake, 1996) also found that rural areas are more
likely to offer preventive services than are urban areas. The finding that fewer workers are
assigned during the course ofthe case is also consistent with descriptions of the vertical case
management model. This model calls for social workers to carry the same case from case
opening to case closure.
Implications and Recommendations
•
Future research should include a larger subsample from South County and include
methods to understand the reasons behind these location differences. Also, a
larger South County Office location sample may enable more detailed
comparisons by ethnicity.
185
•
Practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County
Offices and adapted to other regions of Santa Clara County, as recommended by
study participants. This practice may be particularly helpful to low-income
families and families of color that may be intimidated by or unfamiliar with child
welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for other
regions of Santa Clara County, though the success of this practice may be linked
to the "small county" dynamic described by South County study participants.
Some participants in focus groups that took place at the main office mentioned
this model as promising while others asserted that specialized expertise in
different areas of social work was an advantage in providing quality services that
are informed by "the most current information."
•
Further examination of ways in which the culture of commitment to maintaining
and reunifying families is created, communicated, and continued merits attention.
In addition, an exploration of how this culture might be adapted to other parts of
the county would be beneficial.
Fourth, in order to identify case characteristics and system-related factors predictive of
reunification of children with their families, two exploratory models were tested using
multivariate analyses. The first model was composed of primarily demographic and systemrelated variables to predict reunification (child's gender, ethnicity, age at time of case opening,
the number of workers assigned across the history of the case, the time length of the case, the
number of unique placement homes assigned, the number of times removed from the family, and
the county location). Four ofthese variables predicted reunification: child's ethnicity, number of
workers assigned throughout the case, length ofthe case, and number of unique placement
homes.
Asian American/Pacific Islanders were less likely than Whites, African Americans, and
Latinos to be reunified with their families. Given that Asian American/Pacific Islanders are one
of the minority groups least acculturated to mainstream society, their cases may be perceived as
more severe or problematic because of the different cultural attitudes toward child rearing and
discipline. Differences in parenting styles may reduce the likelihood of reunification compared to
the other major ethnic groups. However, as Phase 2 results indicated Asian American/Pacific
Islanders are also the most likely to be placed in Voluntary Family Maintenance (Hines et al.,
2002), the CWS appears to value the preservation of Asian American/Pacific Islander families
when first referred to the system (perhaps recognizing the challenges of culturally appropriate
services, language issues, and other circumstances related to the Asian American/Pacific Islander
community) but once in the system, Asian American/Pacific Islander family cases are not easily
resolved and circumstances may persist that place the child at risk for further abuse of neglect.
The fewer number ofworkers assigned was also related to reunification. This finding
indicates consistency in service and the relationship between the social worker and the family
186
may improve the chances of reunification. Perhaps this is due to greater familiarity and thus more
effective advocacy for the child and family. This finding could also be a function of particularly
challenging family cases already less likely to be reunified needing a greater set of workers over
time (e.g., due to worker burnout in the case, or specialization needed across the case).
Shorter case length was also predictive ofreunification. Presumably, less severe cases
could be resolved in a timely fashion, usually concluding with reunification. Following a similar
logic above regarding the number of workers, a longer case may imply more problems, thus
involving more time for resolution, but also possibly a case already protracted into a situation
where reunification is unhealthy for the child.
The fewer number of unique placement homes assigned was related to reunification.
Change in placements may be due to systemic conditions (i.e. logistics and short term
availability ofspace) but multiple placements can also indicate a persistent problem with a child
adapting to a placement and accepting care offered. As this may again indicate a more severe
case, reunification could also become less likely.
To identify better the influence of case characteristics and indicators of success, we added
6 variables to the model collected through our in-depth case record reviews and utilized the
additional information coded for Phase 3: a family's enhanced capacity to provide for their
children's needs, children receiving appropriate educational services, children receiving adequate
services to meet physical and mental health needs, number of referrals, number of previous times
in the CWS, and assignment to Family Maintenance or Family Reunification services. The
location variable was not included in this model as only one case coded from the South County
Office area.
Similar to the first model, Asian American/Pacific Islanders were still less likely than
Whites, African Americans, and Latinos to be reunified. Also, shorter time length ofa case
again predicted reunification. However, in this expanded model with success indicators, the
number of workers assigned and number of unique placement homes assigned were not
statistically related to reunification. The child's age at time of case opening and the number of
unique placement homes did approach significance. Interestingly, the variables related to
success cases and outcomes were not significant.
Implications and Recommendations
•
These findings indicate that particular factors are important to consider when
predicting reunification, specifically ethnicity and the length of time a case
remains open. More severe cases (those needing more time for resolution) may be
more problematic, thus reducing the chances of reunification.
•
The finding that Asian American/Pacific Islanders are less likely to be reunified
than Whites, African Americans and Latinos, highlights a main ethnic difference.
In Phase 2 we discovered that Asian American/Pacific Islanders were more likely
than the other ethnic groups to be enrolled in Voluntary Family Maintenance
services (Hines et al., 2002). However, our results here imply that when Asian
187
American/Pacific Islander children are removed from the home, their children are
less likely to be reunified. This may also indicate extreme outcomes where Asian
American/Pacific Islander children are either initially diverted from the system or
once in the mainstream of the system are less likely to be reunited with their
families.
•
It was interesting that the successful case and outcome variables were not
predictive of reunification. This could be due to the significant impact of other
variables overshadowing the influence of success indicators. It could also be due
to our use of reunification as a dependent variable. As we discussed in the
Successful Cases Analysis section, reunification is not necessarily synonymous
with success, and thus these variables may indeed be unrelated.
•
Due to missing data across our set of variables, a number of were available for
this multivariate analysis. More complete information would have improved
confidence in our findings.
•
Better measurement and a more comprehensive evaluation of family history,
abuse and neglect circumstances, and other indicators could produce a stronger
predictive model.
•
Although only a preliminary set of models, these findings are useful for
identifying areas in the CWS and family circumstances that can be studied
further, especially in conjunction with the specific goal of reunifying children
with their families.
Our second aim focused on examining the influence of the court system on service
recommendations for children and families ofcolor. Few studies have examined the potential
influence of the court system on child welfare ·practices and outcomes, however as noted in the
literature review (Section III), the court system is likely to have a significant impact on the
trajectory of some child welfare cases. In Phase 3, we examined the role of the court system in
child welfare practice by exploring the types of court ordered changes that are made to social
worker recommendations at the initial jurisdictional/dispositional hearing, and how changes
might vary by ethnicity of the case. In addition, we explored the relationship between the child
welfare and court systems.
Quantitative results indicated that the court made changes to the initial social worker
recommendations at the jurisdictional/dispositional hearing in approximately halfof the cases.
Ethnicity was not related to whether or not a change occurred Yet, these results may not
necessarily indicate that the system and courts treat children and families from different
ethnicities the same, but rather that when the judge decides to make a change from the social
worker recommendations, these changes appear to occur equally across ethnic groups. It was the
opinion of members of the Court Officer Unit focus group that ethnicity does not contribute to
decisions in court, but rather it is circumstances that drive the case.
188
Additionally, although quantitative results may imply that judges accept social worker
recommendations about half of the time, findings from the focus group with the Court Officer
Unit suggest that many factors would complicate an accurate measurement of why social worker
recommendations may or may not be followed. Factors that may impede accurate assessment of
the agreement between child welfare and court personnel on service plan recommendations
include the strong influence of individual stakeholders involved, and formal and informal case
plan negotiation tactics, all of which contribute to a much more complicated process.
Qualitative findings from the agency-wide focus groups suggested that collaboration
between the child welfare and court systems is problematic. Agency-wide focus group
participants felt that the child welfare and court system have different perspectives on the needs
and circumstances of children and families in the CWS and that the court system may have
unrealistic expectations of families. Additionally, many agency-wide focus group participants
described the ways in which child welfare and court personnel interact as ineffective. Some
social workers felt that some court personnel would dismiss their assessments and
recommendations and may try to pressure social workers to change their recommendations. A
somewhat similar finding was noted by Knepper and Barton, ( 1997) who found that although
judges tended to accept social worker recommendations, the relationship between social workers
and the court plays a significant role. Their study found that when social workers adhered to the
group norms of the court, judges rewarded them by not ordering "unrealistic" practices and
allowing reasonable time frames for mental health evaluations to be completed. Similarly,
agency-wide focus group findings seem to suggest that if social workers go against the implicit
rules ofthe court system, then judges may override their recommendations, but if they adhere to
expectations ofthe court system than recommendations are accepted
Implications and Recommendations
•
Although the quantitative and qualitative information described many of the
circumstances behind why social worker recommendations are not upheld by the
court, it is very difficult to ascertain the degree to which each circumstances
influences a court ordered change. More examination of this area is needed, as is
further investigation of ways in which relationships between court personnel and
child welfare workers have an impact on the experiences of children and families
of color involved in the CWS.
•
Better and more exhaustive record keeping may not be the solution, as many
informal and undocumented actions in the negotiation process occur regularly,
and are confidential, thus making much of the data unavailable for research and
evaluation purposes.
•
Assessing other factors such as courtroom environment, stakeholder
characteristics, and other judicial issues would be advantageous in later research.
Although obtaining this information is a formidable task, these factors should be
include in the future, as measuring primarily system-related factors may not be
sufficient in efforts to explain outcomes in the CWS.
189
Our third aim was to conduct a statewide comparative analysis of identified key practices
that may affect ethnic disproportiona/ity in the CWS. A statewide survey was administered to
child welfare directors, managers or supervisors in California counties who were asked to
identify and describe effective practices for children and families of color in their child welfare
systems. In addition, a quantitative analysis that identified the link between these practices and
county levels of disproportionate representation of children of color in county child welfare
systems was conducted.
Overall findings from the statewide comparative analysis indicate that certain types of
practices are considered by county child welfare personnel as effective with diverse populations,
and that these practices are related to disproportionate representations ofchildren of color in
county child welfare systems. Descriptive results revealed four overall types of practices
considered to be most effective with children and families of color, as well as various specific
practices within these types of practices including: 1) beginning choice point practices, which
included family group conferencing or other team decision-making practices, voluntary family
maintenance or other prevention/family preservation services, and risk assessment tools or
systems, 2) continuing care/ending choice point practices, which included, placement related
practices, the Family to Family Initiative, and intensive and collaborative interventions for highrisk children, 3) both beginning and continuing/ending choice point practices, which included,
community-based services, and strength based assessments, and 4) organizational or structural
practices which included, cultural competency of staff, collaboration with American India~ tribes
and vertical case management.
In general, very little research on the effectiveness of the practices identified by
respondents has been conducted, and in cases where evaluations have been performed, results are
generally inconclusive. More research is needed to evaluate the effectiveness of these services
for diverse populations.
Multivariate analyses revealed several significant associations between types of
practices, specific practices and practice characteristics and levels of disproportionate
representation ofchildren ofcolor in county child welfare systems, when other factors such as
urban-rural status and per capita income were statistically controlled
Beginning choice point practices were found to have a beneficial effect on Latino
overrepresentation. In particular, family group conference or other team decision-making
interventions predicted lower levels of Latino overrepresentation. However, Latino
overrepresentation was greater, not less, when interventions were implemented targeting areas
with high concentrations of distressed families. Latino overrepresentation was also greater when
clients, foster families, out-of-home placement providers, and other service systems were
involved in the development of the practice. As such, it appears that practices may be developed
because stakeholders perceive Latino overrepresentation and initiate programs in response.
In contrast to Latinos, targeting practices to at-risk areas for African American
populations was beneficial in reducing overrepresentation, as well as targeting ofvery young
children. Less overrepresentation was linked also to whether foster families and clients and
contract agencies initiated an intervention, and whether it focused on improving psychosocial
190
outcomes of birth families. The findings suggest a picture almost the reverse of that found for
Latinos.
Additionally, there appear to be lower levels ofAfrican American overrepresentation
than expected when key community stakeholders help to initiate the interventions that focus on
high-risk areas and populations. When interventions have a goal of strengthening the well being
of birth families, then implementation is associated with less African American
overrepresentation. Yet when counties implemented Family-to-Family initiatives greater African
American overrepresentation resulted. Again, the implementation of Family to Family programs
may be used reactively, in response to high perceived levels of overrepresentation.
As with Latinos, targeting ofhigh need areas was less beneficial for Asian
American/Pacific Islander children than not engaging in this kind of intervention. Involving
clients and service systems in initiation, again more common for interventions targeting high
need areas, was also relatively unsuccessful. On the other hand, management and worker
support for a practice was associated with greater success.
Implications and Recommendations
•
County directors and child welfare administrators identified practices with similar
characteristics as those identified within Santa Clara County, i.e., those that are
inclusive, collaborative, culturally appropriate and involve client's communities
appear to be best suited to meeting the needs of children and families of color.
However, further research and evaluation is necessary to determine which
practices most effectively meet the needs of individual racial/ethnic groups.
•
While multivariate analyses suggest that certain practices do predict levels of over
and underrepresentation of children of color in the child welfare system,
additional research to determine the mechanisms through which these practices
impact this disproportionate representation is needed.
•
Additional explorations into the relationship between practices, disproportionate
representation of children of color in the child welfare system and county
characteristics would also help shed light on the process through which certain
practices impact children and families of color.
Phase 3 Conclusion
Results from Phase 2 indicated that different racial/ethnic groups experience unique
pathways through the CWS (Hines et al., 2"002). At every point in the system, various child
welfare practices and services i'mpact children and families. Phase 2 results indicated that
children and families of color in Santa Clara County's CWS tend to be recommended services
that are traditional and formal in nature and that do not appear to meet the wide range of needs
experienced by these highly diverse racial/ethnic family groups. Indeed, children and families of
color in the CWS represent a high risk group who are often impacted by myriad psychosocial
191
challenges yet services recommended for these families tend to be limited to a one-size-fits all
approach.
Moreover, there exists relatively little empirical evidence that traditional child welfare
services are effective in maintaining or reunifying children and improving family functioning.
Even less research has been conducted on the impact of child welfare practices for children and
families of color. Yet because children and families of color are disproportionately represented
in the CWS, the effectiveness of child welfare practices for these groups is of particular interest.
Research suggests that children and families of color tend to have longer stays in the CWS, they
receive fewer and less comprehensive services, and have poorer case outcomes than White
children and families (Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996).
This bleak outlook for children and families of color in the CWS creates an urgent need to
understand child welfare practices and services that result in enhanced outcomes for these
families. It is especially important to identify the types of individual and contextual factors that
hinder or support the effectiveness of these practices.
The primary goal of Phase 3 was to assess key practices for children and families of color
in the CWS at both the county and state levels in order to provide information that might help
generate strategies aimed at reducing ethnic/racial disparities and develop more effective
services and supports for children and families of color. While little research exists that
empirically supports policies, programs and practices that attempt to reduce ethnic/racial
disproportionality, some programs and practices appear promising and merit further exploration.
In general, these practices aim to reduce the need for out-of-home placement and increase family
and community participation in decision-making. Please see Chart 1 for a summary of promising
child welfare practices from Phase 3.
In our analysis of Santa Clara County child welfare practices, results based on focus
group and interview data indicated that in-home services, family conferencing, and team decision
making (TDM) were viewed by participants as ways to improve decision-making with families
from diverse cultural and ethnic groups. Participants noted that both formal and informal
collaborative relationships with other services delivery systems, such as domestic violence and
substance abuse treatment systems, are critical for successful interventions with children and
families. Furthermore, participants recognized the importance of training and cross-training to
enhance the capacity of helping professionals in child welfare and other systems to better address
the needs of children and families of color.
Our analysis comparing South County and the Main Office enabled us to examine
contextual factors that may have an impact on child welfare practices, in this case, small vs. large
and urban vs. rural. Results indicate that children in South County appear to have better
outcomes, i.e., they have fewer number of unique placement homes in the current episode, have a
shorter average stay per placement, a shorter length of time in out-of-home placement, and are
younger at time of case closure. Study participants described South County as "a small county or
rural area." This geographic difference, and factors related to an organizational culture that
emphasize a commitment to maintaining and reunifying families, the practice of providing an
orientation to the child welfare system, as well as the finding that fewer workers are assigned
192
Chart 1: Promising Child Welfare Practices from Phase 3
Beginning Choice Points
Promisine Practices
Team decision making, case conferencing, family group conferencing
and/or family involvement in decision-making
Orientation for parents on child welfare and court processes
.;_:; ~-. ,. ~t.-t~~- -~~;:~ ~· .. "-\;.-::.:
:'.~-:_-:t,.~-~ ':. <' /,:._:- !-<. ,.t;>..:.-. :~. <-:: ·.: _~- i:-.::':
Continuing Care/Ending
Choice Points
;:.: . .,,.-,1;;":•.;_, :.t_;~.,~.:.,.;:. ..:;.\:;~ ·-- 1~~-~·::..:.-.... ·tt'"~~.::-.=i~-~ 'f.-tr./'~_1 :.·::::;...~ 7~:-~ ...
Contextual Factors or
Organizational/
Structural Practices
-~:c:i!:_;-:,:~
Voluntary Family Maintenance/Family Preservation Services
Child care and in-home services
Risk assessment tool or system
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
cr.··~~~~~:;,::~::;:-,.:.~:s.-~- ·· ~.i:..;;:,:t-:·. •·-::-:~.:;;.:-:;, t: .:~::::·'.:\,:·:.:o1,h'!:~t1!~-.:;:.~.=::.:.r:·-~.:_::~t· .! ~,-·= /'- ~:!,i.,-;-:_~"::f,.;~;-:i·,:::-:::;.:,S'.:..~-~-.;.1~:-~""E~1..:;:::-:-;.:..J-k_ 1:~-.;:':j:':~.:..-.:": ·J,:~ --,.·/:; -~··":.~..-fC?::._~
Parenting education services involving both child and parent
Cultural matching for children in foster care
Recruitment, training and coaching for foster parents
Placement-related practices
Family-to-Family Initiative
Intensive and collaborative interventions for high risk children
Community based services (both beginning & continuing/ending)
Strength-based assessments (both beginning & continuing/ending)
•
•
Study Component
Main County Focus Groups
South County Focus Groups
Statewide Survey
Main County Focus Groups
South County Focus Groups
Statewide Survey
Main County Focus Groups
Statewide Survey
"!'" -.. ...
~r:;-::::tJ;.~.--,-:-7~---,.~~2'.--...... -• _
...... .:il ,.._.: .. c'.
Main County Focus Groups
Main County Focus Groups
Main County Focus Groups
Statewide Survey
Statewide Survey
Statewide Survey
Statewide Survey
Statewide Survey
r.t: . - ·-":"!=i.~"·";;;t:".:'!..._...~,..~~~~~!.'::.~~;,;,-.,t;:.:i:!i\.~-:~....-:~:-::.~-.1;,~-::~~1,~h=-~~L!.:$:"."~~~1::;.;;~.:..~~r.,,;~~}.~?'i~~·~t-~Y5~·LGt:~'T~;;.h--I~~~~;;:-..-~ :i:~-,~:tr.:.3.'~=-n..~-z:~~~~..11....-~..,17;~;~::tr~
Cultural matching between staff and clients, and cultural consultation
Supportive supervision for staff
Collaboration with other agencies or systems
Strength-based approach to services
Small county dynamic
Culture of commitment to maintaining and reunifying families
Vertical case management
Cultural competency of staff
Collaboration with American Indian tribes
•
•
•
•
•
•
•
•
•
•
'"'.-;";i,.-~·-~·7••.. •• ..........-.,...,,r•• '",
~~~~;.-:·~~M..i:~"~;:-,.1.-:-z.:.;s;-~'-!':1
Main County Focus Groups
Main County Focus Groups
South County Focus Groups
Main County Focus Groups
Main County Focus Groups
South County Focus Groups
South County Focus Groups
South County Focus Groups
Statewide Survey
Statewide Survey
Statewide Survey
193
during the course of the case may all be related to these positive outcomes for children in South
County. It is interesting to note, however that our multivariate analyses that examined case
characteristics and system-related factors related to reunification indicated that when taking other
factors into account, geographical location was not a significant predictor of reunification.
Rather, fewer workers, shorter case duration, and fewer placements were the strongest predictors.
More research on factors related to agency culture and geographical location as they impact child
welfare practice and outcomes for children and families of color is clearly warranted.
Research indicates that the court system is likely to have an impact on the practices that
are provided to children and families in the CWS. In an analysis conducted to determine the
extent to which court orders changed social worker recommendations at the
jurisdictional/dispositional hearing, results indicated that the court made changes to the initial
social worker in approximately half of the cases. Results further indicated that that ethnicity was
not related to whether or not a change occurred. Findings from qualitative analysis exploring the
relationship between the child welfare and court system painted suggested that the collaboration
between the child welfare system and courts is problematic. More research is this area is needed,
as is further investigation of ways in which relationships between court personnel and child
welfare workers have an impact on the experiences of children and families of color involved in
theCWS.
Results on statewide practices indicate that certain types of practices are considered by
county child welfare personnel to be effective with diverse populations, and that these practices
are related to disproportionate representations of children of color in county child welfare
systems. Similar to practices identified within Santa Clara County's DFCS, practices at the
statewide level were described as effective with children and families of color because they are
inclusive, collaborative, culturally appropriate, and involve client's communities. Although the
research literature does not provide convincing evidence of the effectiveness of these practices,
the fact that respondents reported these practices as their _most effective with children and
families of color suggests that they may serve as promising models for children and families of
color and that more rigorous evaluations of these practices are needed. Phase 3 also attempted to
assess how the practices identified through the survey might affect differential representation of
children of color in the CWS. Significant associations were detected between practice types,
practices and practice characteristics and disproportionate representation and interestingly,
different patterns of association were found for Latinos, African Americans and Asian
American/Pacific Islanders.
While researchers have described the existence of ethnic/racial disparities in every aspect
of the CWS, little attention has been given to ways to address the existing disproportinality. It is
our hope that the information contained in Phase 3 of the Children of Color Study will help in the
development of more effective programs and practices for families and children of color at both .
the county and state level. Overall, our findings from Phase 3 suggest that efforts to address
racial and ethnic disparities in the child welfare system should involve a more concentrated focus
on child and family well-being, involving the creation of family-centered and community-based
services that are inclusive and collaborative and specifically designed to meet the needs of the
·
diverse cultural and ethnic groups in the child welfare population.
194
VII. SUMMARY OF PHASES 1, 2 AND 3: OVERALL
CONCLUSIONS AND IMPLICATIONS FOR SANTA CLARA
COUNTY'S CHILD WELFARE SYSTEM
The central focus of inquiry for this project was to identify factors related to the
disproportionate representation of children of color in Santa Clara County's CWS. In order to
address the complexity of this issue, the CWRT elected to employ a multiphase/multimethod
approach that examined factors at the individual, family and system levels. The following
summarizes findings from Phases 1, 2 and 3 of the study and provides overall implications for
the County based on our results.
Phase 1, an exploratory phase, drew from three sources: research literature at the national,
state and county levels; Santa Clara County's management information system (CWS/CMS); and
focus group discussions with professionals in Santa Clara County who provide child welfare
services, as well as parents, caregivers and youth who are recipients of CWS services. Results
indicated that factors related to racial/ethnic disproportionality were multiple and complex and
necessitated investigation at the individual, family, system and community level. Based on
preliminary analyses and available literature, we argued that children from different racial/ethnic
groups were likely to be treated differently at specific decision-making or key choice points in
the system. The concern about possible differential treatment of children of color was
substantiated by the focus group results, as well as by CWS/CMS data that indicated that there
were significant differences in the number of months in placements by race/ethnicity. For
example, results indicated that African American children spent significantly more time in
placement than their White, Hispanic/Latino, Asian American/Pacific Islander, Vietnamese, and
Filipino peers. Also, according to preliminary findings from CWS/CMS, in Santa Clara County,
following placements in a relative home, the second most frequent placement for African
American, Native American, White and Hispanic/Latino youth, was a Foster Family Agency
(FFA); for Asian American/Pacific Islander and Vietnamese youth it was a Foster Family Home
(FFH). Almost 14% of Native American and 18% of Asian American/Pacific Islander youth in
OHP were placed in the Children's Shelter in December, 2000, a percentage that was higher than
any of the other ethnic/racial groups. We also argued that little was known about specific
pathways through the system and ways in which the pathways differed for various racial and
ethnic groups. Our findings from Phase 1 indicated that much of the research on children in the
CWS had focused on factors related to movement in and out of the system, but that little
information existed on the actual experiences of children in care and the individual and family
characteristics that are associated with these experiences. While focus group members
frequently stated that once a child of color entered the system, it was very hard for that child to
exit, little information existed that could explain what happened once that child was in the
system.
Phase 2 was designed to focus on the actual experiences of children in care and the
individual and family-related characteristics that were associated with these experiences. Results
based on extensive, in-depth reviews of 403 closed child welfare case records, a parallel
descriptive analysis of 1720 closed cases within the CWS/CMS database, and key informant
interviews with managers and supervisors in the County's DFCS indicated that families
195
belonging to each of four different racial/ethnic groups (Latino, African American,
Asian/Pacific Islander and White) were characterized by different constellations of risk factors,
and that once in the system, the children had different experiences that resulted in different
outcomes. Additionally, results indicated that services recommended for the families and
children were limited to a small array of traditional services - a one-size fits all approach - that
did not necessarily meet the needs of the culturally diverse families and children. (Please see
Attachment 8 for a summary chart of Phase 2 findings.) We concluded that the paucity of
recommended services indicated that children and families of color were not likely to be
provided with sufficient preventive and supportive services and that the traditional child welfare
services might not meet the particular needs of these unique and diverse groups. These findings
are supported by research literature indicating relatively little empirical evidence that traditional
child welfare services are effective in maintaining or reunifying children and improving famUy
functioning. Even less research has been conducted on the impact of child welfare practices for
children and families of color. Yet because children and families of color are disproportionately
represented in the CWS, we contend that the effectiveness ofchild welfare practices for these
groups is of particular interest.
Phase 3 focused on agency practice and ways in which current practice interacted with
child, family and cultural characteristics of different cultural/ethnic groups. The primary goal of
Phase 3 was to assess key practices for children and families of color in the CWS both at the
County and state levels in order to provide information that might help generate strategies aimed
at reducing ethnic/racial disparities and develop more effective services and supports for children
and families of color. Methods included semi-structured in-depth interviews and focus groups
with DFCS supervisors and managers, social workers and parents involved in the CWS; a
qualitative and quantitative comparison of child welfare practices and outcomes between the
DFCS South County and Main County offices; a reanalysis of cases from Phase 2 with
successful outcomes and an analysis of factors that contributed to those positive outcomes; a
supplemental data collection for our case record review sample so as to include information on
court ordered changes to the social worker recommendations at the jurisdictional/dispositional
hearing; and, a statewide survey of county child welfare directors, managers or supervisors in
California counties and quantitative analyses using a statewide database consisting of countylevel characteristics.
As mentioned earlier, Phases 1 and 2 confirmed that certain ethnic groups were overrepresented· (i.e., Latinos, African Americans) while others were under-represented (i.e., Asian
American/Pacific Islanders) in the County's CWS. Phase 2 results also indicated that there were
significant ethnic differences in demographic, system-related, and psychosocial characteristics
and that many of these variations appeared at early choice points in the system. During Phase 3,
we were interested in examining ethnic/racial differences in outcomes at later stages of the
system. Using data from our Phase 2 Case Record Review sample and defining successful
outcomes according to US DHHS guidelines (i.e., safety, permanency, and family and child
well-being), we found that there were minimal ethnic differences in successful outcomes. Phase
3 results suggested that when success was defined according to DHHS guidelines, ethnic
differences at the conclusion of the case were not evident. One interpretation for this finding
might be that while there is a clear disproportion of ethnic groups in the system, once in the
system, children as a whole, regardless of race/ethnicity, are faring at about the same level.
196
Given that Phase 2 results found that children and families of color in Santa Clara County's
CWS were receiving a "one-size fits all" approach to services, an approach that may seem
equitable and logistically efficient, but perhaps, ultimately ineffective in serving a diverse group
of children and families.
In order to further examine child welfare practices at the County level and to gain a better
understanding than the one gleaned solely through the use ofCWS/CMS data, we conducted a
series of semi-structured in-depth interviews with DFCS managers and supervisors and focus
groups with social workers and parents involved in the CWS. Our results indicated that in-home
services, family conferencing, and team decision-making (TOM) were viewed by participants as
ways to improve decision-making with families from diverse cultural and ethnic groups.
Participants noted that both formal and informal collaborative relationships with other service
delivery systems, such as domestic violence and substance abuse treatment systems, were critical
for successful interventions with children and families. Furthermore, participants recognized the
importance of training and cross-training to enhance the capacity of helping professionals in
child welfare and other systems to better address the needs of children and families of color.
We were also interested in gathering information from counties throughout California
regarding promising practices for families and children of color. This information was gathered
through a statewide survey administered to child welfare directors and program administrators.
Similar to findings on practices identified within Santa Clara County's DFCS, practices at the
statewide level were described as effective with children and families of color because they are
inclusive, collaborative, culturally appropriate and involve client's communities. Although the
research literature does not provide convincing evidence of the effectiveness of these practices,
the fact that respondents reported these practices as those that are most effective with children
and families of color suggests that they may serve as promising models for children and families
of color and that more rigorous evaluations of these practices are needed (Please see Chart 1 on
p. 192 of the Phase 3 Final Report for a summary of promising practices for children and
families of color in the CWS).
Although practices such as orientation of new clients; family and team decision-making;
preventive services; substance abuse services; cultural competency; collaboration with other
agencies and systems; a strength-based approach; and the use of relative placements were
described as having a positive impact on children and families of color, certain negative practices
and contexts were cited as possibly impeding their effective implementation. For instance,
having to screen out large numbers of inappropriate referrals; inconsistency in decision-making
practices; a shortage of services, particularly substance abuse and preventive services; lack of
client access to services; difficult protocols for placing children with kin; gaps in cultural
competency; time limits; and agency-level factors such as heavy caseloads, staff shortages,
substantial amounts of paperwork, lack of access to information about resources; and confusion
about the agency's overall mission and key policies all impeded implementation of the best
practices. These findings suggest that in addition to building on current positive practices for
children and families of color, CWS stakeholders should also actively work both internally and
in collaboration with partners to reduce barriers to these best practices.
197
Results from focus groups conducted during Phase 1 and key informant interviews held
during Phase 2, as well as current research literature suggest that agency organizational culture
and context may have an impact on child welfare practices and resulting family and child
outcomes. During Phase 3, we conducted both quantitative and qualitative analyses comparing
South County and the Main Office that enabled us to examine various contextual factors at the
organizational level, in this case, small vs. large and urban vs. rural, and their impact on
outcomes for children and families of color. Results indicated that children in South County
appeared to have better outcomes, i.e., they had fewer number of unique placement homes in the
current episode, had a shorter average stay per placement, a shorter length of time in out-ofhome placement, and were younger at time of case closure. Qualitative findings were congruent
with the quantitative data and provided some insight into the dynamics that might contribute to
this difference. South County study participants described a culture of commitment to
maintaining and reunifying families. This commitment was described as integral to the
philosophy of managers and social workers and reflected in the expectations communicated to
social workers by most supervisors and by peers. The practice of providing an orientation to the
child welfare system, time limits, the courts, and how to work with social workers might also
contribute to the differences found in the quantitative analysis. Study participants frequently
described South County as similar to a small county or rural area. This geographic difference,
and factors related to an organizational culture that emphasizes a commitment to maintaining and
reunifying families, the practice of providing an orientation to the child welfare system, as well
as the finding that fewer workers were assigned during the course of the case may all be related
to these positive outcomes for children in South County. Further, our multivariate analyses that
examined case characteristics and system-related factors related to reunification indicated that
when taking other factors into account, fewer workers, shorter case duration, and fewer
placements were the strongest predictors of family reunification.
The issue of disproportionate involvement of children of color in the CWS has long been
an issue of concern for CWS workers, clients, researchers and government and community
groups. More recently, it has been the focus of much national attention. Santa Clara County's
Children of Color Study is one of the few to examine this issue at the local level. We applaud the
County Board of Supervisors and DFCS for their openness and willingness to undertake such a
project. We acknowledge that there are multiple stakeholders who may be interested in
identifying a "magic bullet" or a single, straightforward explanation for the system's
racial/ethnic disproportionality. Single, straightforward explanations are appealing as they might
lead to quickly implemented solutions. However, results of this study indicate that rather than
one primary causal factor, there appear to be numerous and interrelated factors associated with
the disproportionate involvement of children of color in Santa Clara County's CWS. It is our
contention, that factors that operate simultaneously and in complex ways at the individual,
family and system level with differing patterns across diverse racial and ethnic groups contribute
to disproportionality throughout the system. Overall, findings suggest that efforts to address
racial and ethnic disparities in the CWS should focus on the diverse needs of the different ethnic
and racial groups involved in the CWS and the design of more culturally specific and effective
prevention and intervention programs to meet those needs.
198
Implications for Santa Clara County's Child Welfare System
The following are suggestions based on results from all three phases of the Children of Color
Study. It is our hope that these recommendations may be helpful in guiding Santa Clara
County's continuing efforts to address the complex issue of racial/ethnic disparities in its CWS.
l. There is a need to expand available child welfare services to better serve families and
children ofcolor. Phase 2 results indicated that reliance on a small array of
traditional formal services does not appear to meet the needs of the highly diverse
ethnic/racial family groups involved in the County's CWS.
•
More preventive and early intervention services for vulnerable families of color
should be implemented and carefully evaluated. The findings from the qualitative
component of Phase 3 affirmed the value of many efforts that are already in place,
such as culturally specific ER response units and family resource centers, and
other that are in the process of implementation including efforts to reduce
caseloads, initiation of team decision making, and participation in the Family-toFamily initiative.
In addition, home visitation services may be exceptionally helpful in addressing
the needs of vulnerable, at-risk families from diverse racial/ethnic groups.
Research has indicated that programs of home visitation that promote positive
health-related behaviors in mothers of young children, competent care of their
children and linkage with needed health care and human services, reduce rates of
criminality, problems related to substance abuse and child abuse and neglect
among young, unmarried, isolated, poor mothers.
•
The use of more non-traditional, culturally sensitive services is clearly -warranted.
Key informants in all three phases of the study underscored this point. A paucity
of social services, particularly multi-lingual services, was cited as a significant
barrier for many families of color. Interviewees discussed the shortage of
substance abuse treatment programs, particularly those geared for women with
children and people whose primary language is not English. Multi-lingual and
culturally appropriate domestic violence services, parenting classes, and other
social services were considered in need of development.
•
Piloting new, innovative services and evaluating their success with different
ethnic/racial groups could contribute to a more diverse array of culturally specific
programs for families and children.
•
Phase 3 results also indicated that practices that are inclusive, collaborative,
culturally appropriate and involve client's communities appear to be best suited to
meeting the needs of children and families of color. Although the research
literature does not provide convincing evidence of the effectiveness of these
199
practices, the fact that respondents reported these practices as those that are most
effective with children and families of color suggests that they may serve as
promising models for children and families of color.
•
More services targeted to fathers and programs that are formulated and delivered
within a family-based framework would be of benefit to families and children of
color involved in the CWS.
•
In addition to building on current positive practices, and initiating new programs
for children and families of color, CWS stakeholders should actively work
internally and in collaboration with partners to reduce barriers to implementing
such practices. Barriers such as large numbers of inappropriate referrals;
inconsistency in decision-making practices; a shortage of services, particularly
substance abuse and preventive services; lack of client access to services; difficult
protocols for placing children with kin; time limits; and agency-level factors such
as heavy caseloads, staff shortages, substantial amounts of paperwork, lack of
access to information about resources; and confusion about the agency's overall
mission and key policies were cited during Phase 3 focus groups as impediments
to the implementation of cited best practices.
2. There is a need to involve multiple social service systems in a comprehensive and
coordinated effort to meet the needs ofchildren and families ofcolor. Results from
Phases 2 and 3 indicated that the problems experienced by families across the
different racial/ethnic groups span multiple systems including: mental health, juvenile
justice, adult criminal justice, substance abuse, and welfare. Statistics presented in
the literature review section of the Phase 2 Final Report also indicate that families of
color are· involved in systems other than child welfare in high numbers. Prevention
and intervention efforts should involve a deliberate and organized coordination of
these multiple systems. During Phase 3 interviews and focus groups, collaboration
with other agencies and systems concerned with the safety and welfare of children
and families of color was highlighted as a beneficial practice for children and families
of color.
•
Developing interagency formal agreements and connections with liaisons in
other agencies would be beneficial as it can facilitate collaboration and
minimize the potential for agencies to give conflicting messages and mandates
to clients. Phase 3 participants suggested that social workers were better able
to divert cases or get a more complete picture of the family when two agencies
are working together. Other agencies mentioned in this context include
probation, hospital, medical providers, family and mental health agencies,
CalWORKS, substance abuse treatment agencies, domestic violence agencies
and other entities thatrriay be involved with clients.
•
The Greenbook project, an initiative to better coordinate domestic violence
and child welfare services, could serve as a model for interdisciplinary
problem solving and policy development. The Greenbook project successes
200
could help inform evolving collaboration with other key systems, such as the
substance abuse treatment system.
•
Other collaborative efforts such as outstationed social workers, co-location of
services and Family Drug Court could serve as models for best practices for
children and families of color. Social workers and supervisors, interviewed
during Phase 3, described the family drug court as a model for effectively
working with substance abusing families in the child welfare system.
Participants described having ER social workers at different sites such as the
Family Violence Center and police stations as a valued practice.
3. Ways in which Agency organizational context and culture impacts families and
children ofcolor merits attention. It is possible that characteristics of a large
bureaucratic organization impede the development and facilitation of practices that
would be of benefit to families and children of color. For example, organizational
goals such as having a unified agency mission, a shared organizational culture and
personal contact with clients may be difficult to achieve in the face of complex and
competing demands inherent in a large organization.
•
Practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County
Offices and adapted to other regions of Santa Clara County, as recommended by
study participants. This practice may be particularly helpful to low-income
families and families of color that ·may be intimidated by or unfamiliar with child
welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for other
regions of Santa Clara County, though the success of this practice may be linked
to the "small county" dynamic described by South County study participants.
Some participants in focus groups that took place at the main·office mentioned
this model as promising while others asserted that specialized expertise in
different areas of social work was an advantage in providing quality services that
are informed by "the most current information."
•
Further examination of ways in which the culture of commitment to maintaining
and reunifying families is created, communicated, and continued merits attention.
In addition, an exploration of how this culture might be adapted to other parts of
the county would be beneficial.
4. More attention thatfocuses on ways in which system level changes at the federal,
state and local levels have an impact on families and children ofcolor is warranted
With the passage of the Multi-Ethnic Placement Act-Interethnic Adoption Provision
(MEPA-IEP, 1996) and the Adoption and Safe Families Act (AFSA, 1997), safety,
permanency and expedited placements have taken precedence in the CWS. Results
based on focus group and interview data from all three phases of the study indicated
201
that the shift toward expedited placement and an emphasis on permanency has
certainly influenced the culture and ways in which social services are being delivered
in California in general, and in Santa Clara in particular and consequently raises
several major concerns for children of color in the County's CWS. In combination
with the new regulations, the characteristics of families, children and communities of
color including chronic poverty, substance abuse, lack of community social
organization or racial segregation that increase chances of entering the CWS may
create overwhelming barriers to successful reunification for children of color in the
CWS and keep them in the system longer.
•
Regular assessment of ways in which shortened timelines, early termination of
parental rights, bypass criteria, as well as changes in adoption regulations and
incentives have had an impact on children and families of color would be
beneficial.
•
A close monitoring of the effect of new policies, implementation of new programs
or changes in the agency's organizational structure on outcomes for families and
children of color would contribute to a better understanding of the differential
impact these innovations might have on different racial and ethnic groups.
5. In order to better assess whether safety, permanency, and family and child well-being
are maintained and enhanced, records should include more complete and measurable
information. We acknowledge that child welfare case records were not originally
intended for research and that that improving the collection and condition of child
welfare data is no easy task given the issues of limitations of the data system,
compatibility among systems of data recording and storage, and training needs.
However, concerted efforts in this area are necessary if accurate and useful
information is to be obtained and used to provide feedback on the effectiveness of
programs and services in improving outcomes for families and children in the
County's CWS.
•
The amount of missing data and the manner in which information is assessed also
implies the need for more reliable and valid methods of recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their families or of
those adopted high or low? And compared to what? Implementing better record
keeping and data collection over time on key characteristics, based on an agreedupon operational definition would better address the question of effectiveness and
performance. Implementing these practices across counties and across states
would also facilitate evaluation.
6. Future research efforts should be directed toward examining neighborhood,
community, and other macro-level factors particular to Santa Clara County and ways
in which these factors interact with individual, family and system-related
characteristics to propel children into and maintain them in the CWS. Research
202
findings consistently point to a relationship between poverty and child maltreatment.
. Characteristics associated with communities and neighborhoods of poverty including;
living in a high crime area, living in public housing, having larger numbers of
dependent children, and receiving welfare benefits might place children of color at an
increased risk of entering and staying in the CWS.
It was our intention in conceptualizing the original design and scope of work for the
Children of Color Study to examine community level factors and their interaction
with individual and family characteristics during Phase 3 of the study. However, due
to the interests of various stakeholders, the community component was not included
in the current study. It is our contention that in order to understand more fully the
disproportionate involvement of children and families of color in the CWS, the
processes by which individual, family and system level factors interact with
characteristics that are, perhaps unique to Santa Clara County and impact families and
children of color and their subsequent entry into the CWS need to be examined.
203
References
Akin, B. A., & Gregoire, T. K. (1997). Parents' views on child welfare's response to
addiction. Families in Society: The Journal of Contemporary Human Services, 393-404.
Alter, C. F. ( 1985). Decision-making factors in cases of child neglect. Child Welfare,
64(2), 99-111.
American Humane Association (2003). What is Family Group Decision Making? Online,
retrieved July 1, 2003, from:
http://www.americanhumane.org?site?PageServer?pagemane=pc_ fgdm _ what_is.
Ards, S. D., Myers, S. L., Malkis, A., Sugrue, E., & Zhou, L., (2003). Racial
disproportionality in reported and substantiated child abuse and neglect: An examination of
systematic bias. Children and Youth Services Review, 25(5/6) 375-392.
Atkinson, L., & Butler, S. (1996). Court-ordered assessment: Impact of maternal
noncompliance in child maltreatment cases. Children and Youth Services Review, 20(3), 185190.
Avery, R. J. (2000). Perceptions and practice: Agency efforts for the hardest-to-place
children. Children and Youth Services Review, 22(6), 399--420.
Baird, C., & Wagner, D. (2000). The relative validity of actuarial- and consensus-based
risk assessment systems. Children and Youth Services Review, 22(11/12), 839-871.
Banach, M. (1998). The best interests of the child: Decision-making factors. Families in
Society: The Journal of Contemporary Human Services, 79(3), 331-341.
Barth, R. P. (1991). An experimental evaluation of in-home child abuse prevention
services. Child Abuse and Neglect 15, 363-375.
Barth, R. P., Snowden, L. R., Broeck, E.T., Clancey, T., Jordan, C., & Barusch, A. S.
(1987). Contributors to reunification or permanent out-of-home care for physically abused
children. Journal of Social Service Research, 9(2/3), 31--45.
Bath, H. I., Richey, C. A., Haapala, D. A. (1992). Child age and outcome correlates in
intensive family preservation services. Children and Youth Services Review, 14, 389--406.
Bay Area Social Services Consortium (2000). Concurrent Planning: The Influence of
Bypass and Poor Prognosis Indicators on Child Welfare Outcomes. Berkeley, CA: Author.
Beeman, S. K., Hagemeister, A. K., & Edleson, J. L. (1999). Child protection and
battered women's services: From conflict to collaboration. Child Maltreatment, 4(2), 116-126.
204
Berry, M. (1992). An evaluation of family preservation services: Fitting agency services
to family needs. Social Work, 37(4), 314-321.
Berry, M. (1991). The assessment of imminence of risk of placement: Lessons from a
family preservation program. Children and Youth Services Review, 13, 239-256.
Bowen, K. (2000). Child abuse and domestic violence in families of children seen for
suspected sexual abuse. Clinical Pediatrics, 39, 33-40.
Brown, T. G., Werk, A., Caplan, T., & Seraganian, P. (1999). Violent substance abusers
in domestic violence treatment. Violence and Victims, 14(2), 179-190.
Brooks, D., & James, S. (2003). Willingness to adopt Black foster children: Implications
for child welfare policy and recruitment of adoptive parents. Children and Youth Services
Review, 25(5/6), 463-489.
Brunk, M., Henggeler, S. W. & Whelan, J.P. (1987). Comparison ofmultisystemic
therapy and parent training in the brief treatment of child abuse and neglect. Journal of
Consulting and Clinical Psychology. 55(2), 171-178.
California Department of Social Services (2003). Structured Decision Making. Online,
retrieved July 14, 2003 from: http://www.childsworld.ca.gov/Structured 352.htm.
California Department of Social Services (2001). Wraparound Services Pilot, Fiscal
Questions and Answers. Online, retrieved, May 5, 2001 from:
http://childsworld.org/services/famcenwrap.htm
California Department of Social Services (1999). Panel on Race and Bias in Risk
Assessment Thirteenth National Roundtable on Child Protective Services July 7-9, San
Francisco, CA.
Camasso, M. J., & Jagannathan, R. (2000). Modeling the reliability and predictive
validity of risk assessment in child protective services. Children and Youth Services Review,
22(11/12), 873-896.
Carlo, P. ( 1993). Parent education vs. parent involvement: Which type of efforts work
best to reunify families? Journal of Social Service Research, 17(1/2), 135-1°50.
Carten, A. J. (1996). Mothers in recovery: rebuilding families in the aftermath of
addiction. Social Work 41(2) 214-223.
Carter-Black, J. (2002). Transracial adoption and foster care placement: Worker
perception and attitude. Child Welfare, 71(2), 337-370.
Casey Foundation (2003a). Team Decision Making. Online, retrieved July 13, 2003 from:
http://aecf.org/initiatives/familytofamily/tools/decisions/htm.
205
Casey Foundation (2003b). Family to Family Home Page. Online, retrieved July 13, 2003
from: http://www.aecf.org/initiatives/familytofamily/
Chaffin, M., Bonner, B. L., & Hill, R. F. (2001). Family preservation and family support
programs: Child maltreatment outcomes across client risk levels and program types. Child Abuse
and Neglect, (25), 1269-1289.
Chao, R., & Tseng, V. (2002). Parenting of Asians. In M. H. Bornstein (Ed.). Handbook
of Parenting Vol. 4 Social Conditions and Applied Parenting (pp. 59-93). Mahwah, NJ:
Lawrence Erlbaum Associates.
Cash, S. J., & Berry, M. (2002). Family characteristics and child welfare services: Does
the assessment drive service provision? Families in Society, 83(5/6), 499-507.
Children NOW. (2001). California Report Card 2001: County Information. Online,
retrieved, August 20 2003, from: http://www.childrennow.org/califomia/rc-2001/county-pdflist.htm#top2.
Children's Research Center (2003). Structured Decision Making. Online, retrieved
July 15, 2003 from: http://www.nccd-crc.org/crc/decision.htm.
Clark, H. W. (2001). Residential substance abuse treatment for pregnant and postpartum
women and their children: Treatment and policy implications. Child Welfare, 80(2), 179-198.
Close, M. M. ( 1983). Child welfare and people of color: Denial of equal access. Social
Work Research and Abstracts, 19(4), 13-20.
Cohen, E. P. (2003). Framework for culturally competent decision-making in child
welfare. Child Welfare, 82(2), 143-155.
Courtney, M., Piliavin, I., Grogan-Kaylor, A., & Nesmith, A. (2001). Foster youth
transitions to adulthood: A longitudinal view of youth leaving care. Child Welfare, ,80, 685-717.
Courtney, M. E. (1997). The politics and realities oftransracial adoption. Child Welfare,
76(6), 749-779.
Courtney, M. E., Barth, R. P., Berrick, J. D., Brooks, D., Needell, B., & Park, L. (1996).
Race and child welfare services: Past research and future directions. Child Welfare 75(2) 99-137.
Courtney, M. E. (1994). Factors associated with the reunification of foster children with
their families. Social Service Review, 68(1), 81-108.
Cuddeback, G. S., & Orme, J. G. (2002). Training and services for kinship and
nonkinship foster families. Child Welfare, 91(6), 879-909.
206
Dalgleish, L. I., & Drew, E. C. (1989). The relationship of child abuse indicators to the
assessment of perceived risk and to the court's decision to separate. Child Abuse and Neglect,
.U,491-506.
Davidson-Arad, B. (2001). Predicted changes in children's quality of life in decisions
regarding the removal of children at risk from their homes. Children and Youth Services Review,
23(2), 127-143.
DePanfilis, D. (1996). Social isolation of neglectful families: A review of social support
assessment and intervention models. Child Maltreatment, 1(1), 37-52.
DePanfilis, D., & Zuravin, S. J. (2001). Assessing risk to determine the need for services.
Children and Youth Services Review, 23(1 ), 3-20.
DePanfilis, D., & Scannapieco, M. (1994). Assessing the safety of children at risk of
maltreatment: Decision-making models. Child Welfare, 73(3), 229-245.
Drake, B. (1996). Predictors of preventive services provision among unsubstantiated
cases. Child Maltreatment, 1(2), 168-175.
Dore, M. M. & Lee, J.M. (1999). The role of parent training with abusive and neglectful
parents. Family Relations, 48(3), 313-325.
dosReis, S., Zito, J.M., Safer, D. J., Soeken, K. L. (2001). Mental health services for
youths in foster care and disabled youths. American Journal of Public Health, 91 (7). 10941100.
Doueck, H.J., English, D. J., DePanfilis, D., Moote, G. T. (1993). Decision-making in
child protective services: A comparison of selected risk-assessment systems. Child Welfare,
72(5), 441-452.
Downing, J. D., Wells, S. J., & Fluke, J. (1990). Gatekeeping in child protective services:
A survey of screening policies. Child Welfare, 69(4), 357-369.
Dunford, F. W. (2000). The San Diego navy experiment: An assessment of interventions
for men who assault their wives. Journal of Consulting and Clinical Psychology, 68(3), 468476.
Dutton, D. G. (1986). The outcome of court-mandated treatment for wife assault: A
quasi-experimental evaluation. Violence and Victims, 1(3), 163-175.
Eckenrode, J., Ganzel, B., Henderson, C.R., Smith, E., Olds, D. L., Powers, J., Cole, R.,
Kitzman, H., & Sidora, K. (2000). Preventing child abuse and neglect with a program of nurse
home visitation: The limiting effects of domestic violence. The Journal of the American Medical
Association, 284(11), 1385-1396.
207
Eckenrode, J., Powers, J., Doris, J., Munsch, J., & Bolger, N. (1988). Substantiation of
child abuse and neglect reports. Journal of Consulting and Clinical Psychology, 56(1), 9-16.
English, D. J., Marshall, D. B., Coghlan, L., Brummel, S., & Orme, M. (2002). Causes
and consequences of the substantiation decision in Washington state-child protective services.
Children and Youth Services Review, 24(11), 817-851.
Farmer, E. M., Wagner, H. R., Bums, B. J., & Richards, J. T. (2003). Treatment foster
care in a system of care: Sequences and correlates of residential placements. Journal of Child and
Family Studies, 12(1), 11-25.
Federal Interagency Forum on Child and Family Statistics. America's Children: Key
National Indicators of Well-being, 2001. Federal Interagency Forum on Child and Family
Statistics, Washington DC: U.S. Government Printing Office.
Fein, E., & Staff, I. (1993). Last best chance: Findings from a reunification services
program. Child Welfare, 72(1), 25-40.
Fetsch, R. J., Schultz, C. J., & Wahler, J. J. (1999). A preliminary evaluation of the
Colorado rethink parenting and anger management program. Child Abuse and Neglect, 23(4),
353-360.
.
Findlater, J.E., & Kelly, S. (1999). Reframing child safety in Michigan: Building
collaboration among domestic violence, family preservation, and child protection teams. Child
Maltreatment, 4(2), 167-174.
359.
Flanigan, J. (1954). The Critical Incident Technique. Psychological Bulletin. 4, 327-
Fluke, J. D., Yuan, Y. T., Hedderson, J., & Curtis, P.A. (2003). Disproportionate
representation of race and ethnicity in child maltreatment: Investigation and victimization.
Children and Youth Services Review, 25(5/6), 359-373.
Forehand, R., & Kotchick, B. A. (1996). Cultural diversity: A wake-up call for parent
training. Behavior Therapy, 27, 187-206.
Fraser, J. A., Armstrong, K. L., Morris, J.P. & Dadds, M. R. (2000). Homes visiting
intervention for vulnerable families with newborns: Follow-up results of a randomized controlled
trial. Child Abuse and Neglect, 24(11), 1399-1429.
Garbarino, J., & Kostelny, K. (1992). Child maltreatment as a community problem. Child
Abuse and Neglect 16, 45 5-464.
Garland. A. F., & Besinger, B. A. (1997). Racial/ethnic differences in court referred
pathways to mental health services for children in foster care. Children and Youth Services
Review, 19(8), 651-666.
208
Garland, A. F., Hough, R., Landsverk, J. A., McCabe, K. M., Yeh, M., Ganger, W. C., &
Reynolds, B. (2000). Racial/ethnic differences in mental health care utilization among children
in foster care. Children's Services: Social Policy, Rese·arch and Practice, 3, 133-146.
Gaudin, J.M., Wodarski, J. S., Arkinson, M. K., & Avery, L. S. (1990). Remedying child
neglect: Effectiveness of social network interventions. The Journal of Applied Sciences, 15(1 ),
97-123.
Gillespie, J.M., Byrne, B., Workman, L. J. (1995). An intensive reunification program
for children in foster care. Child and Adolescent Social Work Journal, 12(3), 213-228.
Gregoire, K. A. & Schultz, D. J. (2001). Substance-abusing child welfare parents:
Treatment and child placement outcomes. Child Welfare 80(4), 433-452
Goerge, R. M. (1990). The reunification process in substitute care. Social Service
Review, 64(3), 422--457.
Gordon, A. L., McKinley, S. E., Satterfield, M. L., & Curtis, P. (2003). A first look at the
need for enhanced support services for kinship caregivers. Child Welfare, 81 ( 1), 77-96.
Guterman, N. B. (1999). Enrollment strategies in early home visitation to prevent
physical child abuse and neglect and the "universal versus targeted" debate: A meta-analysis of
population-based and screening-based programs. Child Abuse and Neglect, 23(9), 863-890.
Haller, D. L., Knisely, J. S., Elswick, R. K., Dawson, K. S., Scholl, S. H. (1997).
Perinatal substance abusers: Factors influencing treatment retention. Journal of Substance Abuse
Treatment, 14(6), 513-519.
Harman, J. S. (2001). Mental health care utilization and expenditures by children in foster
care. The Journal of the American Medical Association, 285(5), 524.
Harris, M. S., & Courtney, M. E. (2003). The interaction of race, ethnicity, and family
structure with respect to the timing of family reunification. Children and Youth Services Review,
25(5/6), 409--429.
Hines, A. M., Lee, P.A., Drabble, L., Snowden, L. R., & Lemon, K. (2002). An
Evaluation of Factors Related to the Disproportionate Representation of Children of Color in
Santa Clara County's Child Welfare System: Phase 2 Final Report. Report submitted to Santa
Clara County, CA Department of Social Services, Family and Children's Services, August 31,
2002.
Hochstadt, N. J ., Harwicke, N. J. ( 1985). How effective is the multidisciplinary
approach? A follow-up study. Child Abuse and Neglect, 9, 365-372.
209
Hosley, C. A., Gensheimer, L., & Yang, M. (2003). Building effective working
relationships across culturally and ethnically diverse communities. Child Welfare, 82(2), 157168.
Jellinek, M. S., Murphy, M., Poitrast, F., Quinn, D., Bishop, S. J., & Goshko, M. (1992).
Serious child mistreatment in Massachusetts: The course of 206 children through the courts.
Child Abuse and Neglect, 16, 179-185.
Johnson, P., & Cahn, K. (1992). Improving child welfare practice through improvements
in attorney-social worker relationships. Child Welfare, 74(2), 383-394.
Jones, L. (1998). The social and family correlates of successful reunification of children
in foster care. Children and Youth Services Review, 20(4), 305-323.
Jones, L. (1993). Decision making in child welfare: A critical review of the literature.
Child and Adolescent Social Work Journal, 10(3), 241-262.
Kapp, S. A., McDonald, T. P., & Diamond, K. L. (2001). The path to adoption for
children of color. Child Abuse and Neglect, 25, 215-229.
Karoll, B. R. & Poertner, J. (2002). Judges' caseworkers', and substance abuse
counselors' indicators of family reunification with substance-affected parents. Child Welfare,
71(2), 249-269.
· Karski, R. L. (1999). Key decisions in child protective services: Report investigation and
court referral. Children and Youth Services Review, 21(8), 643-656.
Kemp, S. P., & Bodonyi, J.M. (2002). Beyond termination: Length of stay and predictors
of permanency for legally free children. Child Welfare, 81(1), 59-86.
Kinard, E. M. (2002). Services for maltreated children: variations by maltreatment
characteristics. Child Welfare, 81(4), 617-645.
Kitzman, H., Olds, D, L., Henderson, C. R., Hanks, C., Cole, R., Tatelbaum, R.,
McConnochie, K. M., Sidora, K., Luckey, D. W., Shaver, D., Engelhardt, K., James, D., &
Barnard, K. (1997). Effect of prenatal and infancy home visitation by nurses on pregnancy
outcomes, childhood injuries, and repeated childbearing: A randomized controlled trial. The
American Medical Association, 278(8), 644-652.
Knepper, P. E., & Barton, S. M. (1997). The effect of courtroom dynamics on child
maltreatment proceedings. Social Service Review, 71(2), 288-309.
Knight, D. K., Logan, S. M., & Simpson, D. D. (2001). Predictors of program completion
for women in residential substance abuse treatment. American Journal ofD.rug and Alcohol
Abuse, 27(1), 1-15.
210
Kolbo, J. R., & Strong, E. (1997). Multidisciplinary team approaches to the investigation
and resolution of child abuse and neglect: A national study. Child Maltreatment, 2(1 ), 61-72.
Landsman, M. J. (2002). Rural child welfare practice from an organization-inenvironment perspective. Child Welfare, 81 (5), 790-819.
Landy, S., & Munro, S. (1998). Shared parenting: Assessing the success of a foster parent
program aimed at family reunification. Child Abuse and Neglect, 22(4), 305-318.
Lee, B. J., & Goerge, R. M. (1999). Poverty, early childbearing, and child maltreatment:
A multinomial analysis. Children and Youth Services Review, 21(9/10), 755-780.
Lewandowski, C. A., & GlenMaye, L. F. (2002). Teams in child welfare settings:
lnterprofessional and collaborative processes. Families in Society, 83(3), 245-256.
Lewandowski, C. A., & Pierce, L. (2002). Assessing the effect of family-centered out-ofhome care on reunification outcomes. Research on Social Work Practice, 12(2), 205-221.
Lightbourne, W. (2002). Greenbook Project Update. Report submitted to the Santa Clara
County Board of Supervisors.
Lindsey, D. (1991). Factors affecting the foster care placement decision: An analysis of
national survey data. American Journal of Orthopsychiatry, 61 (2), 272-281.
Littell, J. H. (1997). Effects of the duration, intensity, and breadth of family preservation
services: A new analysis of data from the Illinois Family First Experiment. Children and Youth
Services Review, (19)2, 17-39.
Littell, J. H., & Schuerman, J. R. (1995). A synthesis of research on family preservation
and family reunification programs. Online, retrieved January 27, 2003 from:
http://aspe.hhs.gov/hsp/cyp/fp litrev .htm
Little Hoover Commission (2003). Still in Our Hands: A Review of Efforts to Reform
Foster Care in California. Online, retrieved July 24, 2003 from:
http://www.lhc.ca.gov/lhcdir/report 168 .html
Maluccio, A. N. & Ainsworth, F. (2003). Drug use by parents: A challenge for family
reunification practice. Children and Youth Services Review, 25(7), 511-533.
Martin, L. M., Peters, C. L., & Glisson, C. (1998). Factors affecting case management
recommendations for children entering state custody. Social Service Review,--, 521-544.
McGuigan, W., Katzev, A. R., & Pratt, C. C. (2003). Multi-level determinants of
retention in a home-visiting child abuse prevention program. Child Abuse & Neglect. 27, 363380.
211
McGuigan, W., M., & Pratt, C. C. (2001 ). The predictive impact of domestic violence on
three types of child maltreatment. Child Abuse & Neglect, 25, 869-883.
McKay, M. M. (1994). The link between domestic violence and child abuse: Assessment
and treatment considerations. Child Welfare, 73(1), 29-39.
McMurty, S., & Lie, G. (1992). Differential exit rate of minority children in foster care.
Social Work Research and Abstracts, 28(1 ), 42-48.
McPhatter, A. R., & Ganaway, T. L. (2003). Beyond the rhetoric: Strategies for
implementing culturally effective practices with children, families and communities. Child
Welfare, 82(2), 103-124.
Mederos, F., & Woldeguiorguis, I. (2003). Beyond cultural competence: What child
protection managers need to know and do. Child Welfare, 82(2), 125-142.
Menanteau-Horta, D., & Yigzaw, M. (2002). Indicators of social well-being and
elements of child welfare in Minnesota rural counties. Child Welfare, 81 (5), 709-73 5.
Metsch, L. R., Wolfe, H.P., Fewell, R., McCoy, C. B., Elwood, W. N., Wohler-Torres,
B. et al. (2001 ). Treating substance-using women and their children in public housing:
Preliminary evaluation findings. Child Welfare, 80(2), 199-220.
Miller, 0. A., & Gatson, J. (2003). A model of culture-centered child welfare practice.
Child Welfare, 82(2), 235-.249.
Morrel, T. M., Elliot, J. D., Murphy, C. M., & Taft, C. T. (2003). Cognitive behavioral
and supportive group treatments for partner-violent men. Behavior Therapy, 34, 77-95.
Murphey, D. A., & Braner, M. (2000). Linking child maltreatment retrospectively to birth
and home visit records: An initial examination. Child Welfare, 79(6), 711-728.
Needell, B., Brookhart, M.A., Lee, S. (2003). Black children and foster care placement
in California. Children and Youth Services Review, 25(5/6) 393-408.
Needell, B., Webster, D., Cuccaro~Alamin, S., Armijo, M., Lee, S., Brookhart, A., Lery,
B., Shaw, T., Dawson, W., Piccus, W., Magruder, J., & Kim, H. (2003). Child Welfare Services
Reports for California. Online, retrieved June 28, 2003, from: University of California at
Berkeley Center for Social Services Research website. URL:
<http://cssr.berkeley.edu/CWSCMSreports/>
Nicholson, B. C., Brenner, V., & Fox, R. A. (1999). A community-based parenting
program with low-income mothers of young children. Families in Society, 80(3), 247-253.
Olds, D. L., Eckenrode, J., Henderson, C.R., Kitzman, H., Powers, J., Cole, R., Sidora,
K., Morris, P., Pettitt, L. M., & Luckey, D. (1997). Long-term effects of home visitation on
212
material life course and child abuse and neglect: Fifteen-year follow-up of a randomized trial.
The Journal of the American Medical Association, 278(8) 637-644.
Ondersma, S. J., Malcoe, L. H., & Simpson, S. M., (2001). Child protective services'
response to prenatal drug exposure: Results from a nationwide survey. Child Abuse and Neglect,
25, 657-668.
Palmer, S. E. (1996). Placement stability and inclusive practice in foster care: An
empirical study. Children and Youth Services Review, 18(7), 589-601.
Palmer, S. E., Brown, R. A., & Barrera, M. E. (1992). Group treatment for abusive
husbands: Long-term evaluation. American Journal of Orthopsychiatry, 62(2), 276-283.
Patterson, G. R., Chamberlain, P., & Reid, J.B. (1982). A comparative evaluation of a
parent-training program. Behavior Therapy, 13, 638-650.
Pecora, P. J., Fraser, M. W., Bennett, R. B., & Haapala, D. A. (1991). Placement rates of
children and families served by the Intensive Family Preservation Services Programs. In M. W.
Fraser, P. J. Pecora, & D. A. Haapala (Eds.), Families in Crisis: The Impact oflntensive Family
Preservation Services (pp.149-180). New York: Aldine De Gruyter.
Pennell, J., & Burford, G. (2000). Family group decision making: Protecting children and
women. Child Welfare, 79(2), 131-158.
Rossi, P. H., Schuerman, J., & Budde, S. (1999). Understanding decisions about child
maltreatment. Evaluation Review, 23(6), 579-598.
Reiff, M. I. (2001). Mental health care utilization and expenditures by children in foster
care. Journal of Developmental and Behavioral Pediatrics. 22(2), 151.
Reyes, B. I. (Ed.). (2001). A Portrait of Race and Ethnicity in California. San Francisco:
Public Policy Institute of California.
Russel, R. (1988). Role perceptions of attorneys and caseworkers in child abuse cases in
juvenile court. Child Welfare, 67(3), 205-216.
Sanchirico, A., & Jablonka, K. (2000). Keeping foster children connected to their
biological parents: The impact of foster parent training and support. Child and Adolescent Social
Work Journal, 17(3), 185-203.
Scannapieco, M. (1994). Home-based services program: Effectiveness with at risk
families. Children and Youth Services Review, 16(5/6), 363-377.
Schriver, J. M. ( 1998). Human Behavior and the Social Environment. Allyn and Bacon:
Boston.
213
Schuerman, J., Rossi, P.H., & Budde, S. (1999). Decisions on placement ·and family
preservation: Agreement and targeting. Evaluation Review, 23(6), 599-618.
Sheppard, D. I., & Zangrillo, P.A. (1996). Coordinating investigations of child abuse.
Public Welfare, Winter, 21-31.
Sieppert, J. D., Hudson, J., & Unrau, Y. (2000). Family group conferencing in child
welfare: Lessons from a demonstration project. Families in Society. 81(4), 382-391.
Smith, B. D. (2003). How parental drug use and drug treatment compliance relate to
family reunification. Child Welfare, 82(3), 335-365.
Spaid, W. M., Fraser, M. W., & Lewis, R. E. (1991). Changes in family functioning: Is
participation in Intensive Family Preservation Services correlated with changes in attitudes or
behaviors? In M. W. Fraser, P. J. Pecora, & D. A. Haapala (Eds.), Families in Crisis: The Impact
oflntensive Family Preservation Services (pp.131-148). New York: Airline De Gruyter.
Stevenson, J. (1999). The treatment and long-term sequelae of child abuse. Journal of
Child Psychology and Psychiatry, 40(1 ), 89-11 I.
Suleiman, L. P. (2003). Beyond cultural competence: Language access and Latino civil
rights. Child Welfare, 82(2). 185-200.
Taft, C. T., Murphy, C. M., Elliot, J. D., & Keaser, M. C. (2001). Race and demographic
factors in treatment attendance for domestically abusive men. Journal of Family Violence, 16(4),
385-400.
Templeman, S~ B., & Mitchell, L. (2002). Challenging the one-size-fits-all myth:
Findings and solutions from a statewide focus group of rural social workers. Child Welfare,
fil.(5}, 757-772.
Thieman, A. A., & Dail, P. W. (1997). Predictors of out-of-home placement in a family
preservation program: Are welfare recipients particularly vulnerable? Policy Studies Journal,
25(1) 124-139.
Toffalo, D. A. (2000). An investigation of treatment integrity and outcomes in
wraparound services. Journal of Child and Family Studies, 9(1), 351-361.
Tumlin, K. C., & Geen, R. (2000). The Decision to Investigate: Understanding State
Child Welfare Screening Policies and Practices. Online, retrieved, February 7, 2003 from:
http://www.urban.org/url.cfm?ID=309534
U.S. Census Bureau (2001). Overview of Race and Hispanic Origin, Census 2000 Brief.
Online, retrieved November 20, 2002 from: http://www.census.gov/prod/2001pubs/c2kbr011.pdf.
214
U.S. Census Bureau (2000). State and County Quick Facts: California. Online, retrieved
March 3, 2003 from: http://quickfacts.census.gov/qfd/states/06000.html
U.S. Department of Health and Human Services, Administration for Children, Youth and
Families, Children's Bureau (2003a). Decision-making in Unsubstantiated Child Protective
Services Cases. Washington DC: U.S. Government Printing Office.
U.S. Department of Health and Human Services. (2003b). Protecting the Well-being of
Children. Online, retrieved April 16, 2003 from:
http://www.scf.hhs.gov/news/facts/chilwel fpr .htm
U.S. Department of Health and Human Services. (2002). The Adoption and Foster Care
Analysis Reporting System Report. Online, retrieved July 16, 2003 from:
http://www.scf.hhs.gov/programs/cb/publications/afcars/report7 .htm.
U.S. Department of Health and Human Services, Administration for Children, Youth and
Families (2001). Child Maltreatment 2001. Washington DC: U.S. Government Printing Office.
U. S. General Accounting Office (1998). Agencies Face Challenges Securing Stable
Homes for Children of Substance Abusers. GAO/HEHS 98..;182. Washington DC: U.S.
Government Printing Office.
U.S. General Accounting Office (1997). Child Welfare: States' Progress in Implementing
Family Preservation and Support Services. GAO/HEHS 97-34. Washington DC: U.S.
Government Printing Office.
Vidal de Haymes, M., & Simon, S. (2003). Transracial adoption: Families identify issues·
and needed support services. Child Welfare, 82(2), 251-272.
Wald, M. S., Woolverton, M. (1990). Risk assessment: The emperor's new clothes?
Child Welfare, 69(6), 483-511.
Walton, E., Fraser, M. W., Lewis, R. E., Pecora, P. J., & Walton, W. K. (1993). In-home
family-focused reunification: An experimental study. Child Welfare, 72(5), 473--487.
Wells, K., & Guo, S. (2003). Mothers' welfare and work income and reunification with
children in foster care. Children and Youth Servic.es Review, 25(3), 203-224.
Wells, K., & Guo, S. (1999). Reunification and reenty of foster children. Children and
Youth Services Review, 21(4), 273-294.
Wells, K., & Whittington, D. (1993). Child and family functioning after intensive family
preservation services. Social Service Review, March, 55-83.
Wells, S. J., Fluke, J. D., & Brown, C.H. (1995). The decision to investigate: Child
protection practice in 12 local agencies. Children and Youth Services Review, 17(4), 523-546.
215
Westat, Inc. (2001a). Caseworker decision-making. In Assessing the context of
permanency and reunification in the foster care system. Online, retrieved January 27, 2003 from:
http://aspe.hhs.gov/hsp/fostercare-reunifO 1/chapter? .htm
Westat, Inc. (2001b). Evaluation of family preservation and reunification programs.
Online, retrieved January 27, 2003 from: http://aspe.hhs.gov/hsp/fampres94/index.htm
Westat, Inc. (2001c). The role of race in parental reunification. In Assessing the context
of permanency and reunification in the foster care system. Online, retrieved January 27, 2003
from: http://aspe.hhs.gov/hsp/fostercare-reunifO 1/chapter6.htm
Whitney, P., & Davis, L. (1999). Child abuse and domestic violence in Massachusetts:
Can practice be integrated in a Public Child Welfare Setting? Child Maltreatment, 4(2), 158166.
Wolfe, D. A., Edwards, B., Manion, I., & Koverola, C. (1988). Early intervention for
parents at risk of child abuse and neglect: A preliminary investigation. Journal of Consulting and
Clinical Psychology, 56(1), 40--47.
Wolock, I. (1982). Community characteristics and _staff judgments in child abuse and
neglect cases. Social Work Research and Abstracts, 18(2), 9-15.
Representation of Children of Color in
Santa Clara County's Child Welfare System:
Child Welfare Practices and
Ethnic/Racial Disproportionality in the Child Welfare System
Phase 3
Final Report
Submitted to the County of Santa Clara
Social Services Agency
Department of Family and Children's Services
By
The Child Welfare Research Team
College of Social Work
San Jose State University
September 2, 2003
(revised November 17, 2003)
Principal Contributors
Child Welfare Research Team
Alice M Hines, Ph.D.
Principal Investigator
Peter Allen Lee, Ph.D.
Co-Investigator and Primary
Data Analyst
Laurie Drabble, Ph.D.
Co-Investigator and Primary
Researcher on Qualitative
Component
Kathy Lemon, MS. W.
Senior Research Assistant
Julian Chow, Ph.D.
School of Social Welfare,
University of California at
Berkeley, Consultant
Alfred Perez, MS. W.
The Pew Commission on
Children in Foster Care
Georgetown University,
Consultant
Lonnie R. Snowden, Ph.D.
School of Social Welfare,
University of California at
Berkeley, Consultant
TABLE OF CONTENTS
PAGE
EXECUTIVE SUMMARY
ACKNOWLEDGEMENTS
I.
Introduction
i
xi
II.
Overview ofStudy Objectives and Approach for Phase 3
6
III.
Background and Literature Review
8
Impact of Child Welfare Practices on Children and Families of Color
10
Beginning Choice Points
10
Beginning Choice Points: Impact on Children and Families of Color
13
Decision-Making Practices at Beginning Choice Points
13
Decision-Making Practices: Risk Assessment Tools
Decision-Making Practices: Multidisciplinary Teams/Family Group
Conferencing
16
16
Decision-Making Practices at Beginning Choice Points: Impact on
Children and Families of Color
18
Voluntary Family Maintenance Practices/Family Preservation Services
Voluntary Family Maintenance Practices/Family Preservation Services:
20
20
Impact on Children and Families of Color
Home-Based Prevention Efforts
21
Home-Based Prevention Efforts: Impact on Children and Families of Color
Continuing Care and Ending Choice Points
22
23
Factors Associated with Family Reunification
23
Family Reunification Services
Family Reunification Services: Impact on Children and Families of Color
Parenting Education Services
Parenting Education Services: Impact on Children and Families of Color
Substance Abuse Treatment
Substance Abuse Treatment: Impact on Children and Families of Color
Domestic Violence Treatment
Domestic Violence Treatment: Impact on Children and Families of Color
24
26
26
27
28
29
29
31
TABLE OF CONTENTS
PAGE
Social Network Interventions/Family to Family Initiative
31
Family to Family/Social Network Interventions: Impact on Children and
Families of Color
33
Mental Health Services for Children in Out-of-Home Care
33
Mental Health Services: Impact on Children of Color in Out-of-Home Care
Foster Parent and Kinship Support Services
34
35
Foster and Kinship Support Services: Impact on Children and Families of
Color
36
Permanency Planning Practices
37
Permanency Planning Practices: Impact on Children and Families of Color
39
Impact of Contextual Factors on Child Welfare Practices for Children
40
and Families of Color
The Role of Cultural Competency in Child Welfare Practice
40
Cultural Competency: Impact on Children and Families of Color
Rural Child Welfare Practices
Rural Child Welfare Practices: Impact on Children and Families of Color
41
42
43
The Role of the Court System in Child Welfare Practice
The Role of the Court System in Child Welfare Practice: Impact on
43
44
Children and Families of Color
Summary of Key Findings and Implications
IV.
Santa Clara County Child Welfare Practices
45
47
Successful Cases Analysis
48
Methods
Safety Outcomes
Permanency Outcomes
Family and Child Well-Being Outcomes
Summary of Key Findings and Implications
Qualitative Analysis of Santa Clara County Child Welfare Practices
48
48
52
56
58
61
Methods
Beginning Choice Points: Practices that Negatively Impact Children and
61
66
Families of Color
TABLE OF CONTENTS
Beginning Choice Points: Practices that Positively Impact Children and
PAGE
69
Families of Color
Beginning Choice Points: Recommendations to Improve Practices
70
Continuing Care/Ending Choice Points:
71
Practices that Negatively Impact Children and Families of Color
Continuing Care/Ending Choice Points: Practices that Positively Impact
74
Children and Families of Color
Continuing Care/Ending Choice Points: Recommendations to Improve
75
Practices
Contextual Factors that Negatively Impact Practices for Children and
76
Families of Color
Contextual Factors that Positively Impact Practices for Children and
80
Families of Color
Contextual Factors: Recommendations to Improve Practices for Children
81
and Families of Color
Summary of Key Findings and Implications
84
Comparison of the Main County Offices and South County Offices
87
Quantitative Comparison
87
Quantitative Methods
87
County Location and Child Characteristics
88
County Location and System-Related Factors
90
Qualitative Comparison
97
Qualitative Methods
97
Context of Service Provision in South County Offices
98
Practices in Service Provision in South County Offices
99
Summary of Key Findings and Implications: Quantitative and Qualitative
101
Comparison of Main County Offices and South County Offices
Multivariate Analysis of Factors and Predicting Reunification:
103
Preliminary Models
Results: Closed Case Sample
I 03
Summary of Key Findings and Implications: Closed Case Sample
104
Results: Case Record Review Sample
I 05
Summary of Key Findings and Implications: Case Record Review Sample
108
TABLE OF CONTENTS
PAGE
The Role of the Court System in Child Welfare Practice
109
Quantitative and Qualitative Methods: Analysis of Court-Ordered Changes
109
Quantitative Results: Analysis of Court Ordered Changes
111
Qualitative Results: Analysis of Court Ordered Changes
113
Qualitative Methods: Relationship Between the Child Welfare and Court
114
Systems
Qualitative Results: Relationship Between the Child Welfare and Court
114
Systems
Summary of Key Findings and Implications: The Role of the Court System
116
in Child Welfare Practice
V.
Statewide Comparative Analysis
118
Methods
118
Contextual Tables: County CJiild Populations and CWS Involvement
119
Descriptive Results: Effective Practices for Children and Families of
124
Color
Beginning Choice Point Practices
124
Continuing Care/Ending Choice Point Practices
135
Both Beginning and Continuing Care/Ending Choice Point Practices
146
Organizational or Structural Practices
154
Summary of Key Findings and Implications: Descriptive Results
165
Multivariate Analysis: Impact of Statewide Practices on Differential
170
Representation
VI.
Multivariate Methods
171
Latino Overrepresentation
174
African American Overrepresentation
174
Asian American/Pacific Islander Underrepresentation
175
Summary of Key Findings and Implications: Multivariate Analysis
177
Phase 3 Summary and Conclusions
VII. Summary of Phases 1,2 and 3: Overall Conclusions and Implications for Santa
180
194
Clara County's Child Welfare System
References
203
TABLE OF CONTENTS
Attachments
1. Social Worker Focus Group Interview Guide
2. Summary Table of Focus Group Participant Demographics
3. Family Focus Group Interview Guide
4. Supervisor Individual Interview Guide
5. Court Case Record Review Data Extraction Form
6. Court Officer Unit Focus Group Interview Guide
7. Statewide Survey on Effective Practices for Children and Families of Color
8. Summary Chart of Phase 2 Findings
Executive Summary
Introduction
While child abuse and neglect appears to affect children of all racial and ethnic
origins (US Department of Health and Human Services, 1998; Sedlack & Broadhurst,
1996), an analysis of national, California and Santa Clara County data on the ethnicities
of children in out of home placement reveals that, compared to their presence in the
general population, there is a disproportionate involvement of children of color in the
public child welfare system (CWS). In Santa Clara County, in particular, when compared
to the general population, African American, Hispanic/Latino and Native American
children are overrepresented in Santa Clara County's CWS, while Asian
American/Pacific Islander and White children are underrepresented. African Americans
represent 4 percent of the general child population in the county, and are 13.6 percent of
children in supervised foster care. Hispanic/Latino children represent 30 percent of the
general child population in Santa Clara County and constitute 55.3 percent of the child
welfare cases. Native Americans are approximately 0.5 percent of Santa Clara County's
population and represent 0.6 percent of children in the CWS. Asian American/Pacific
Islander children represent 21 percent of the general county child population and 5.4
percent of children in the CWS; Whites constitute 45 percent of the general child
population and 25.0 percent of the child welfare population (Needell et al., 2003; U.S.
Bureau of the Census, 2000).
The disproportionate involvement of children of color in the CWS has long been
an issue of concern for CWS workers, clients, researchers and government and
community groups; yet few studies have systematically investigated the factors
associated with this disproportionality. In an effort to better understand the factors related
to the disproportionate number of children of color in the CWS in Santa Clara County,
the Department of Family and Children Services (DFCS) contracted with the Child
Welfare Research Team (CWRT) in the College of Social Work at San Jose State
University to conduct a three-year study on this topic.
The overall question posed by DFCS was: What are the primary reasons why
children of color are disproportionately represented in Santa Clara County's Child
Welfare System? In order to address the complexity of this question, the CWRT elected
to employ a multiphase/multimethod approach, beginning with an initial exploratory
phase that was completed in May 2001. Data for Phase I were gathered from three
sources: research literature at the national, state and county levels; Santa Clara County's
management information system (CWS/CMS); and focus group discussions with
professionals in Santa Clara County who provide child welfare services, as well as
parents, caregivers and youth who are recipients of CWS services. Four central themes
emerged from Phase I findings (please see Phase 1 Final Report for a complete
description of the findings from which these themes were generated).
Conclusions from Phase 1 indicated that more research was needed that focused
on the actual experiences of children in care and the individual and family-related
characteristics that are associated with these experiences. Phase 1 results also indicated
that CWS/CMS data alone would not provide the information needed to address these
research aims. Thus, Phase 2, which concluded in August 2002 focused on characteristics
of children and families of color in the CWS and experiences they had as they wended
their way through the system. The methodology for Phase 2 included extensive, in-depth
case record reviews with a sample of 403 closed child welfare case records, enabling us
to collect information on the individual child and family, as well as system-related
characteristics not included in the CWS/CMS database. The case record review data
provided extensive information on decisions at key points in the system and children's
progress through the system, as well as services ordered for children at the time of case
disposition and at subsequent hearings. In addition, a parallel descriptive analysis of 1720
closed cases within the CWS/CMS database allowed us to examine the entire histories of
children as they progressed through the CWS and to conduct analyses that were not
possible with the CWS/CMS cross-sectional dataset that we used in Phase 1. Key
informant interviews with managers and supervisors in the county's DFCS at key
decision points in the system provided additional information pertaining to cultural and
environmental aspects of the agency. (Please see Phase 2 Final Report for a complete
description of the study methods and findings).
The current report provides findings from Phase 3 that was originally to be
conducted from September 2002 to August 31, 2003. However, due to contractual
changes involving the Scope of Work, the CWRTwas not able to begin work until the
end of January 2003. Thus, the research involved in Phase 3 was conducted over a
shorter seven-month time period. This third phase of the research project focused on
agency practice and ways in which current practice interacts with child, family and
cultural characteristics of different racial/ethnic groups. Phase 3 of the study addressed
three overall aims: 1) to assess identified key practices for children and families of color
in Santa Clara County's CWS, 2) to examine the influence of the court system on service
recommendations for children and families and, 3) to conduct a statewide comparative
analysis of identified key practices that may affect the disproportionate representation of
children of color in the CWS.
The primary methodology for Phase 3 included semi-structured in-depth
interviews and focus groups with DFCS supervisors and managers, social workers and
parents involved in the CWS; a qualitative and quantitative comparison of child welfare
practices and outcomes between the DFCS South County and Main County offices; a
reanalysis of cases from Phase 2 with successful outcomes and an analysis of factors that
contributed to those positive outcomes; a supplemental data collection for our case record
review sample so as to include information on court ordered changes to the social worker
recommendations at the jurisdictional/dispositional hearing; and, a statewide survey of
county child welfare directors, managers or supervisors in California counties and
quantitative analyses using a statewide database consisting of county-level
characteristics.
ii
Overall Conclusions
Results from Phase 2 indicated that different racial/ethnic groups experience
unique pathways through the CWS (Hines et al., 2002). At every point in the system,
various child welfare practices and services impact children and families. Phase 2 results
indicated that children and families of color in Santa Clara County's CWS tend to be
recommended services that are traditional and formal in nature and that do not appear to
meet the wide range of needs experienced by these highly diverse racial/ethnic family
groups. Indeed, children and families of color in the CWS represent a high risk group
who are often impacted by myriad psychosocial challenges yet services recommended for
these families tend to be limited to a one-size-fits all approach.
Moreover, there exists relatively little empirical evidence that traditional child
welfare services are effective in maintaining or reunifying children and improving family
functioning. Even less research has been conducted on the impact of child welfare
practices for children and families of color. Yet because children and families of color
are disproportionately represented in the CWS, the effectiveness of child welfare
practices for these groups is of particular interest. Research suggests that children and
families of color tend to have longer stays in the CWS, that they receive fewer and less
comprehensive services, and have poorer case outcomes than White children and families
(Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996). This bleak
outlook for children and families of color in the CWS creates an urgent need to
understand child welfare practices and services that result in enhanced outcomes for these
families. It is especially important to identify the types of individual and contextual
factors that hinder or support the effectiveness of these practices.
Phases I and 2 of this study focused on examining reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where it
occurs. The primary goal ofPhase 3 was to assess key practices for children and
families ofcolor in the CWS at both the county and state levels in order to provide
information that might help generate strategies aimed at reducing ethnic/racial
disparities and develop more effective services and supports for children and families of
color.
In order to develop a context for examining practices in Santa Clara County's
CWS, we first conducted analyses to assess types ofpractices and case characteristics
associated with successful outcomes for children and families across different
racial/ethnic groups. Using data from our Phase 2 Case Record Review sample and
defining successful outcomes according to US DHHS guidelines (i.e., safety,
permanency, and family and child well-being), results indicated that there were minimal
ethnic differences in successful outcomes. It is important to note that major ethnic
differences do exist in Santa Clara County's CWS. Phase 1 and 2 confirmed that certain
ethnic groups are over-represented (i.e., Latinos, African Americans) while others are
under-represented (i.e., Asian American/Pacific Islanders) in the county's CWS. Phase 2
results also indicated that there are significant ethnic differences in demographic, systemrelated, and psychosocial characteristics and that many of these variations appear at early
iii
choice points in the system. Phase 3 results suggest that when success is defined
according to DHHS guidelines, ethnic differences at the conclusion of the case are not
evident. One interpretation for this finding might be that while there is a clear
disproportion of ethnic groups in the system, once in the system, children as a whole,
regardless of race/ethnicity, are generally faring the same. Phase 2 results found that
children and families of color in Santa Clara County's CWS are receiving a "one-size fits
all" approach to services (Hines et al., 2002). Although this may seem equitable and
logistically efficient, it appears ultimately ineffective in serving a diverse group of
children and families.
While little research exists that empirically supports policies, programs and
practices that attempt to reduce ethnic/racial disproportionality, some programs and
practices appear promising and merit further exploration. In general, these practices aim
to reduce the need for out-of-home placement and increase family and community
participation in decision making.
In our analysis of Santa Clara County child welfare practices, results based on
focus group and interview data indicated that in-home services, family conferencing, and
team decision making (TDM) were viewed by participants as ways to improve decisionmaking withfamiliesfrom diverse cultural and ethnic groups. Participants noted that both
formal and informal collaborative relationships with other services delivery systems, such
as domestic violence and substance abuse treatment systems, are critical for successful
interventions with children and families. Furthermore, participants recognized the
importance of training and cross-training to enhance the capacity of helping professionals
in child welfare and other systems to better address the needs of children and families of
color.
It is interesting that most practices and contextual factors mentioned by social
workers had both positive and negative aspects. Many of the practices and contextual
factors that negatively impact children and _families of color were actually barriers to
successful implementation of the practices that social workers felt have a positive impact
on children and families of color. Although practices such as orientation of new clients;
family and team decision-making; preventive services; substance abuse services; cultural
competency; collaboration with other agencies and systems; a strength-based approach;
and the use of relative placements were described as having a positive impact on children
and families of color, certain negative practices and contexts impeded their effective
implementation. For instance, having to screen out large numbers of inappropriate
referrals; inconsistency in decision-making practices; a shortage of services, particularly·
substance abuse and preventive services; lack of client access to services; difficult
protocols for placing children with kin; gaps in cultural competency; time limits; and
agency-level factors such as heavy caseloads, staff shortages, substantial amounts of
paperwork, lack of access to information about resources; and confusion about the
agency's overall mission and key policies all impeded implementation of the best
practices. These findings suggest that in addition to building on current positive practices
for children and families of color, CWS stakeholders should also actively work both
internally and in collaboration with partners to reduce barriers to these best practices.
iv
Our analysis comparing South County and the Main Office enabled us to examine
contextual factors that may have an impact on child welfare practices, in this case, small
vs. large and urban vs. rural. Results indicate that children in South County appear to
have better outcomes, i.e., they have fewer number of unique placement homes in the
current episode, have a shorter average stay per placement, a shorter length of time in
out-of-home placement, and are younger at time of case closure. Study participants
described South County as "a small county or rural area." This geographic difference,
and factors related to an organizational culture that emphasize a commitment to
maintaining and reunifying families, the practice of providing an orientation to the child
welfare system, as well as the finding that fewer workers are assigned during the course
of the case may all be related to these positive outcomes for children in South County. It
is interesting to note, however that our multivariate analyses that examined case
characteristics and system-related factors related to reunification indicated that when
taking other factors into account, geographical location was not a significant predictor of
reunification. Rather, fewer workers, shorter case duration, and fewer placements were
the strongest predictors. More research on factors related to agency culture and
geographical location as they impact child welfare practice and outcomes for children and
families of color is clearly warranted.
Research indicates that the court system is likely to have an impact on the
practices that are provided to children and families in the CWS. In our analysis
conducted to determine the extent to which court orders changed social worker
recommendations at the jurisdictional/dispositional hearing, results indicated that the
court made changes to the initial social worker in approximately halfofthe cases.
Results further indicated that that ethnicity was not related to whether or not a change
occurred These results do not necessarily indicate that the system and courts treat
children and families from different ethnicities the same, but rather that when the judge
decides to make a change from the social worker recommendations, these changes appear
to occur equally acros~ ethnic groups. It was the opinion of members of the Court
Officer Unit focus group that ethnicity does not contribute to decisions in court, but
rather it is circumstances that drive the case.
Findings from qualitative analysis exploring the relationship between the child
welfare and court system painted suggested that the collaboration between the child
welfare system and courts is problematic. Agency-wide focus group participants felt that
the child welfare and court system have different perspectives on the needs and
circumstances of children and families in the CWS and that the court system may have
unrealistic expectations of families. More research is this area is needed, as is further
investigation of ways in which relationships between court personnel and child welfare
workers have an impact on the experiences of children and families of color involved in
the CWS.
Results from our statewide survey indicate that certain types ofpractices are
considered by county child welfare personnel to be effective with divers_e populations,
and that these practices are related to disproportionate representations ofchildren of
V
color in county child welfare systems. Descriptive results revealed four overall types of
practices considered to be most effective with children and families of color, as well as
various specific practices within these types of practices including: 1) beginning choice
point practices, which included family group conferencing or other team decision-making
practices, voluntary family maintenance or other prevention/family preservation services,
and risk assessment tools or systems, 2) continuing care/ending choice point practices,
which included, placement related practices, the Family to Family Initiative, and
intensive and collaborative interventions for high-risk children, 3) both beginning and
continuing/ending choice point practices, which included, community-based services, and
strength based assessments, and 4) organizational or structural practices which included,
cultural competency of staff, collaboration with American Indian tribes and vertical case
management.
Similar to findings on practices identified within Santa Clara County's DFCS,
practices at the statewide level were described as effective with children and families of
color because they are inclus_ive, collaborative, culturally appropriate and involve client's
communities. Although the research literature does not provide convincing evidence of
the effectiveness of these practices, the fact that respondents reported these practices as
those that are most effective with children and families of color suggests that they may
serve as promising models for children and families of color and that more rigorous
evaluations of these practices are needed.
Phase 3 also attempted to assess how the practices identified through the
statewide survey might affect differential representation ofchildren of color in the CWS.
Our analyses uncovered significant associations between practice types, practices and
practice characteristics and disproportionate representation. It is interesting that different
patterns of association were found for Latinos, African Americans and Asian
American/Pacific Islanders.
Specifically, beginning choice point practices were found to have a beneficial
effect on Latino overrepresentation. In particular, family group conferencing or other
team decision-making interventions predicted lower levels of Latino overrepresentation.
However, Latino overrepresentation was greater, not less, when interventions were
implemented targeting areas with high concentrations of distressed families. Latino
overrepresentation was also greater when clients, foster families, out-of-home placement
providers, and other service systems were involved in the development of the practice. As
such, it appears that practices may be developed because stakeholders perceive Latino
overrepresentation and initiate programs in response.
In contrast to Latinos, targeting practices to at-risk areas for African American
populations was beneficial in reducing overrepresentation, as well as targeting of very
young children. Less overrepresentation was linked also to whether foster families and
clients and contract agencies initiated an intervention, and whether it focused on
improving psychosocial outcomes of birth families. The findings suggest a picture almost
the reverse of that found for Latinos. Additionally, there appear to be lower levels of
African American overrepresentation than expected when key community stakeholders
vi
help to initiate the interventions that focus on high-risk areas and populations. When
interventions have a goal of strengthening the well being of birth families, then
implementation is associated with less African American overrepresentation. Yet when
counties implemented Family-to-Family initiatives greater African American
overrepresentation resulted. Again, the implementation of Family to Family programs
may be used reactively, in response to high perceived levels of overrepresentation.
As with Latinos, targeting of high need areas was less beneficial for Asian
American/Pacific Islander children than not engaging in this kind of intervention.
Involving clients and service systems in initiation, again more common for interventions
targeting high need areas, was also relatively unsuccessful. On the other hand,
management and worker support for a practice was associated with greater success.
While researchers have described the existence of ethnic/racial disparities in every
aspect of the CWS, little attention has been given to ways to address the existing
disproportinality. It is our hope that the information contained in Phase 3 of the Children
of Color Study will help in the development of more effective programs and practices for
families and children of color at both the county and state level. Overall, our findings
from Phase 3 suggest that efforts to address racial and ethnic disparities in the child
welfare system should involve a more concentrated focus on child and family well-being,
involving the creation of family-centered and community-based services that are
inclusive and collaborative and specifically designed to meet the needs of the diverse
cultural and ethnic groups in the child welfare population.
Recommendations for the CWS
Based on Phase 3 findings, we make recommendations/suggestions in the following
areas:
1) Tracking successful case outcomes for families and children of color:
•
In order to better assess whether safety, permanency, and family and child
well-being are maintained and enhanced, records should include more
complete and measurable information. We acknowledge that child
welfare case records were not originally intended for research. Yet, if the
agencies responsible for providing effective services wish to evaluate the
performance of their programs more reliably and validly, improving
information collection and tracking should be considered.
•
The creation of a case summary check-out form used at case closure
would be one way to track child outcomes. The check-out form would
indicate the child and family's status in major areas of psychosocial wellbeing (i.e., health, education, and finances), ·in addition to the existing case
outcomes. Given the amount of missing data assessing these domains
describing case and child status, and the relative scarcity of information
vii
recorded in the last two court hearings, this form would enable
measurement and evaluation of "successful cases and outcomes" in a more
consistent objective manner.
•
The amount of missing data and the manner in which information is
assessed also implies the need for more reliable and valid methods of
recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their
families or of those adopted high or low? And compared to what?
Implementing better record keeping and data collection over time on key
-characteristics, based on an agreed-upon operational definition would
better address the question of effectiveness and performance.
Implementing these practices across counties and across states would also
facilitate evaluation.
·
2) Identifying promising practices within Santa Clara's CWS for families and children of
color:
•
The findings from the qualitative component of this study affirm the value
of many efforts that are already in place, such as culturally specific ER
response units and family resource centers, and other that are in the
process of implementation including efforts to reduce caseloads, initiation
of team decision making, and participation in the Family-to-Family
initiative.
•
Practices that are inclusive, collaborative, culturally appropriate and
involve client's communities appear to be best suited to meeting the needs
of children and families of color. Although the research literature does not
provide convincing evidence of the effectiveness of these practices, the
fact that respondents reported these practices as those that are most
effective with children and families of color suggests that they may serve
as promising models for children and families of color and that more
rigorous evaluations of these practices are needed.
•
The findings also point to opportunities to further strengthen services for
children and families of color. Recommendations related to training,
institutionalizing group orientation for clients, and other suggestions from
study participants could be used to inform practice and planning.
•
It would be beneficial for a team of managers, line staff, family
representatives, and community members to review the recommendations
suggested by focus group and interview participants to identify and
prioritize possible practices for adoption based on their feasibility and
utility for children and families of color.
viii
3) Identifying promising practices for families and children of color based on a
comparison between South County and the Main Office:
•
Future research should include a larger subsample from South County and
include methods to understand the reasons behind these location
differences. Also, a larger South County Office location sample may
enable more detailed comparisons by ethnicity.
•
Practices, such as providing a group orientation to families that may assist
them in navigating through the system may be easily strengthened in
South County Offices and adapted to other regions of Santa Clara County,
as recommended by study participants. This practice may be particularly
helpful to low-income families and families of color that may be
intimidated by or unfamiliar with child welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for
other regions of Santa Clara County, though the success of this practice
may be linked to the "small county" dynamic described by South County
study participants. Some participants in focus groups that took place at the
main office mentioned this model as promising while others asserted that
specialized expertise in different areas of social work was an advantage in
providing quality services that are informed by "the most current
information."
•
Further examination of ways in which the culture of commitment to
maintaining and reunifying families is created, communicated, and
continued merits attention. In addition, an exploration of how this culture
might be adapted to other parts of the county would be beneficial.
4) Identifying factors related to the Court System that may have an impact on practices
for families and children of color in the CWS:
•
Although the quantitative and qualitative information described many of
the circumstances related to whether or not recommendations may be
upheld by the court, it is very difficult to ascertain the degree to which
each circumstances influences a court ordered change. More examination
of this area is needed, as is further investigation of ways in which
relationships between court personnel and child welfare workers have an
impact on the experiences of children and families of color involved in the
cws.
ix
•
Better and more exhaustive record keeping may not be the solution, as
many informal and undocumented actions in the negotiation process occur
regularly, and are confidential, thus making much of the data unavailable
for research and evaluation purposes.
•
Assessing other factors such as courtroom environment, stakeholder
characteristics, and other judicial issues would be advantageous in future
research.
5) Identifying promising practices for families and children of color based on statewide
data:
•
County directors and child welfare administrators identified practices with
similar characteristics as those identified within Santa Clara County, i.e.,
those that are inclusive, collaborative, culturally appropriate and involve
client's communities appear to be best suited to meeting the needs of
children and families of color. However, further research and evaluation is
necessary to determine which practices most effectively meet the needs of
individual racial/ethnic groups.
•
While multivariate analyses suggest that certain practices do predict levels
.of over and underrepresentation of children of color in the child welfare
system, additional research to determine the mechanisms through which
these practices impact this disproportionate representation is needed.
•
Additional explorations into the relationship between practices,
disproportionate representation of children of color in the child welfare
system and county characteristics would also help shed light on the
process through which certain practices impact children and families of
color.
X
Acknowledgements
This report on Phase 3 of the Children of Color study could not have been
completed without the important contributions of so many people.
We wish to thank the former Director of Santa Clara County Social Services
Agency, Yolanda Rinaldo, the initiator of an important discussion with Dr. Sylvia R.
Andrew, the Dean of the College of Social Work at San Jose State University, about the
possibility of a joint research project with the Agency and the College.
Thank you to Leroy Martin, former Director of the Department of Family and
Children's Services, who provided valuable oversight for the research project until his
retirement and to Norma Doctor Sparks, current Director of the Department of Family
and Children's Services. Many thanks also to Will Lightbourne, Director, County of
Santa Clara, Social Services Agency.
We would also like to thank Zonia Sandoval Waldon, Deputy Director of the
Department of Family and Children's Services for her assistance throughout the third
year of the project. Thank you also to Ken Borelli, Sue Farr and Connie Vega.
Judi Boring and her staff, David Heinen and Yvonne Zing, were especially
helpful with providing access to the CWS/CMS system and arranging for the data set that
was used in this report. Thanks also to Diana Weir and her staff who were responsible
for gathering the additional CWS/CMS data for Phase 3. Judy Bushey was extremely
helpful in guiding us through the court report sections of the child welfare case records.
Thank you to Stan Lee and his staff who provided valuable training on the court report
section of the child welfare case records and to Maria Gonzalez who made the case
records available for coding.
Our special thanks to Dean Sylvia R. Andrew who provided members of the
research team with the precious resource of matched release time. Thank you also to
Simon Dominguez for his valuable feedback and encouragement.
The completion of Phase 3 was also made possible by our talented and dedicated
student research assistants: Lisa Navarrette, Priscilla Ribeiro, and Graciela Escobar. Their
work on the qualitative focus groups and supervisor interviews was truly invaluable.
Bernard Voon, who worked as a student research assistant for Phase 2. of the research
project, generously returned to help with supplemental case record reviews in Phase 3 as
a volunteer. Each of our student assistants brought noteworthy enthusiasm, dedication
and insight to the project. Their substantial contributions are deeply appreciated.
In addition, this research could not have been carried out without the cooperation
of the staff of the agency, and we would like to acknowledge the importance of the
xi
contributions of everyone who took the time to attend focus groups, participate in
individual interviews, or assist in organizing staff and family focus groups.
A special thank you to all the child welfare directors and administrators who took
the time to participate in our statewide survey on effective practices for families and
children of color. Their contributions were extremely valuable.
Thanks to Shaaron Gilson for her assistance as Project Manager and for her
thoughtful comments on an initial draft of the report and to Carol Pearce for her
assistance with the statewide survey and the case record review data collection.
Thank you to Jerri Carmo, Juliana Brooks, Bill Yabumoto and the staff at the San
Jose State Foundation for their help in administering this important project.
Finally, we would like to thank the Santa Clara County Board of Supervisors,
especially Blanca Alvarado and James Beall for their ongoing support of the Children of
Color study and their commitment to improving child welfare practices for children and
families in the system.
xii
I.
INTRODUCTION
An analysis of national data on the ethnic composition of children in out-of-home-placement
reveals that, compared to their presence in the general population, there is a disproportionate
involvement of children of color in the public child welfare system (CWS). Statistics indicate that
African American and Native American children are overrepresented in the CWS. African
American children represent 15 percent of the general child population, yet comprise approximately
42 percent of children in the CWS and Native American children constitute approximately 1 percent
of the child population and represent approximately 2 percent of the CWS. Though not all states
provide data on Hispanic/Latino children in the CWS, aggregate statistics from those that do reveal
that while Hispanic/Latino children make up approximately 16 percent of the national child
population, they comprise 15 percent of the CWS population. Asian American/Pacific Islanders and
Whites also tend to be underrepresented at the national level. Asian American/Pacific Islander
children comprise approximately 4 percent of the general child population and 1 percent of the
CWS, while White children who are approximately 64 percent of the general child population,
constitute 36% of children in the CWS (Federal Interagency Forum on Child and Family Statistics,
2001; U.S. Census Bureau, 2001; U.S. Department of Health and Human Services, 1999; U.S.
Department ofHealth and Human Services, 2001).
Further analysis at the California level provides important information on differences and
trends that may go undetected at the national level. In California, a large and ethnically diverse
state, African Americans constitute 6 percent of the general population (32 percent of whom are
below the age of 18), but represent approximately 32.9 percent of children in supervised foster care.
Hispanics/Latinos comprise approximately 32 percent of the general population in California with
43 percent being below the age of 18, and represent 36.2 percent of children in the state CWS.
Native Americans represent 1.0 percent of California's population and constitute approximately 1.2
percent of the children in the CWS. Asian American/Pacific Islanders are under-represented in
California's CWS, as they constitute approximately 11 percent of California's general population,
with approximately 30 percent being younger than 18, but represent just 2.1 percent of the children
in the CWS. While Whites comprise approximately 47 percent of the population in the state, with
only 20 percent being below the age of 18, they constitute 27.5 percent of the children in
California's CWS (Needell, Webster, Cuccaro-Alamin, Armijo, Lee, Brookhart et al., 2003; Reyes,
2001; U.S. Census Bureau, 2000).
In Santa Clara County, there also exists a disproportionate representation of children of
color in the County's CWS. When compared to the general population, African American,
Hispanic/Latino and Native American children are overrepresented in Santa Clara County's CWS,
while Asian American/Pacific Islander and White children are underrepresented. African Americans
represent 4 percent of the general child population in the county, and are 13.6 percent of children in
supervised foster care. Hispanic/Latino children represent 30 percent of the general child population
in Santa Clara County and constitute 55.3 percent of the child welfare cases. Native Americans are
approximately 0.5 percent of Santa Clara County's population and represent 0.6 percent of children
in the CWS. Asian American/Pacific Islander children represent 21 percent of the general county
child population and 5.4 percent of children in the CWS; Whites constitute 45 percent of the general
child population and 25.0 percent of the child welfare population (Needell et al., 2003; U.S. Bureau
of the Census, 2000).
2
Project Description
The disproportionate involvement of children of color in the CWS has long been an issue of
concern for CWS workers, clients, researchers and government and community groups; yet few
studies have systematically investigated the factors associated with this disproportionality. In an
effort to better understand the factors related to the disproportionate .number of children of color in
the CWS in Santa Clara County, the Department of Family and Children Services (DFCS)
contracted with the Child Welfare Research Team (CWRT) in the College of Social Work at San
Jose State University to conduct a three-year study on this topic.
The overall question posed by DFCS was: What are the primary reasons why children of
color are disproportionately represented in Santa Clara County's Child Welfare System? In order to
address the complexity of this question, the CWRT elected to employ a multiphase/multimethod
approach, beginning with an initial exploratory phase that was completed in May 2001. Data for
Phase 1 were gathered from three sources: research literature at the national, state and county levels;
Santa Clara County's management information system (CWS/CMS); and focus group discussions
with professionals in Santa Clara County who provide child welfare services, as well as parents,
caregivers and youth who are recipients of CWS services. Four central themes emerged from Phase
1 findings (please see Phase 1 Final Report for a complete description of the findings from which
these themes were generated). The four emerging themes from Phase 1 included:
1). Little is known about specific pathways through the CWS and ways in which these
pathways differ for various racial/ethnic groups. Much of the research on children in the
CWS focuses on factors relates to movement in and out of the system. The specific
pathways through the system for children of various ethnicities is largely unknown, and may
provide important insights into the factors related to the disproportionate involvement of
children of color. Focus group participants in Phase I frequently stated that once a child of
color enters the system, it is very hard for that child to exit.
2). Various racial/ethnic groups may receive different treatment at key decision-making
points in the system. Differential treatment of children of color in both the initial reporting of
child maltreatment and throughout the various stages within the CWS may exist. CWS/CMS
data for Santa Clara County corroborate the possible differential treatment of children of
color at various points in the system. Phase 1 analysis revealed that African American
children spent significantly more time in placement than their White, Hispanic/Latino, and
Asian American/Pacific Islander peers. Differences in placement type were also foundalthough kincare (placement with a relative) was the most frequent out of home placement
(OHP) type across ethnicities, the second most frequent OHP type varied among
racial/ethnic groups. For African American, Native American, White, and Hispanic/Latino
children, the second most common OHP type was a Foster Family Agency. For Asian
American/Pacific Islander children, the second most frequent placement was a Foster
Family Home. Asian American/Pacific Islander (18%) children were also placed at the
Children's Shelter at higher percentages than children of other racial/ethnic groups. In
addition, focus group participants in Phase 1 expressed concern for possible racial/ethnic
differences in ways in which services and resources were allocated and ways in which the
judicial system handled cases.
3
3). System level changes recently initiated at the federal, state and local level will
undoubtedly have an impact on children of color in the CWS. Recent federal policies,
including the Multi-Ethnic Placement Act-Interethnic Adoption Provision (MEPA-IEP,
1996); the Adoption and Safe Families Act (AFSA, 1997); the Personal Responsibility and
Work Opportunity Reconciliation Act (PRWORA, 1996); as well as state and local county
policies all impact the ways in which families and children of color enter and stay in the
CWS. Recent policy shifts toward expedited permanent placements for children in out-ofhome care and shortened timelines for reunification, as well as new restrictions for
Temporary Assistance to Needy Families (TANF) recipients (more than half of children in
foster care are from welfare eligible families) may increase the likelihood of children of
color entering and staying in the CWS-although much more research is still needed to
determine this association.
4). Factors related to the disproportionate representation of children ofcolor in the CWS
are multiple and complex. No clear consensus from the research literature exists on how
families and children of color become and stay involved in the CWS. Available evidence
suggests that, rather than one primary cause, there appear to be numerous and interrelated
factors associated with disproportionate rates of children of color in the CWS. Factors found
to be associated with CWS involvement for children of all ethnicities, include but may not
be limited to: parental mental illness substance abuse, and incarceration; spousal abuse;
living in poverty; living in neighborhoods with concentrated poverty, and living in
communities with low levels of social organization. For children of color specifically,
research also indicates that race and class biases in initial reporting and subsequent CWS
service delivery do exist.
Conclusions from Phase I indicated that more research was needed that focused on the
actual experiences of children in care and the individual and family-related characteristics that are
associated with these experiences. Phase 1 results also indicated that CWS/CMS data alone would
not provide the information needed to address these research aims. Thus, Phase 2, which concluded
in August 2002 focused on characteristics of children and families of color in the CWS and
experiences they had as they wended their way through the system. The methodology for Phase 2
included extensive, in-depth case record reviews with a sample of 403 closed child welfare case
records, enabling us to collect information on the individual child and family, as well as systemrelated characteristics not included in the CWS/CMS database. The case record review data
provided extensive information on decisions at key points in the system and children's progress
through the system, as well as services ordered for children at the time of case disposition and at
subsequent hearings. In addition, a parallel descriptive analysis of I 720 closed cases within the
CWS/CMS database allowed us to examine the entire histories of children as they progressed
through the CWS and to conduct analyses that were not possible with the CWS/CMS crosssectional dataset that we used in Phase 1. Key informant interviews with managers and supervisors
in the county's DFCS at key decision points in the system provided additional information
pertaining to cultural and environmental aspects of the agency. (Please see Phase 2 Final Report for
a complete description of the study methods and findings). Based on analyses conducted during
Phase 2, five overall conclusions were drawn:
1). Families belonging to each of the four racial/ethnic groups studied, including: African
Americans, Hispanic/Latinos, Asian American/Pacific Islanders and Whites are
characterized by unique constellations of risk factors.
4
2). The characteristics that emerged in our analyses as key characteristics ofeach of the
four racial/ethnic groups have been associated in the research literature as risk/actors/or
child abuse and neglect. These characteristics have also been associated with increased
serious behavioral and adjustment problems in children and adolescents.
3). Once in the child welfare system, children in each ofthe four racial/ethnic groups follow
different pathways and experience different outcomes.
4). The services orderedfor families ofcolor are generally limited to a one-size-fits all
approach and to a small array ofavailable services. The reliance on traditional formal
services does not appear to meet the needs of these culturally diverse families and children.
5). There is a need to involve multiple social service systems in a comprehensive and
coordinated effort to meet the needs ofchildren and families ofcolor.
The current report provides findings from Phase 3 that was originally to be conducted from
September 2002 to August 31, 2003. However, due to contractual changes involving the Scope of
Work, the CWRTwas not able to begin work until the end of January 2003. Thus, the research
involved in Phase 3 was conducted over a shorter seven-month time period. This third phase of the
research project focused on agency practice and ways in which current practice interacts with child,
family and cultural characteristics of different racial/ethnic groups. Phase 3 of the study addressed
three overall aims: 1) to assess identified key practices for children and families of color in Santa
Clara County's CWS, 2) to examine the influence of the court system on service recommendations
for children and families and, 3) to conduct a statewide comparative analysis of identified key
practices that may affect the disproportionate representation of children of color in the CWS.
The primary methodology for Phase 3 included semi-structured in-depth interviews and
focus groups with DFCS supervisors and managers, social workers and parents involved in the
CWS; a qualitative and quantitative comparison of child welfare practices and outcomes between
the DFCS South County and Main County offices; a reanalysis of cases from Phase 2 with
successful outcomes and an analysis of factors that contributed to those positive outcomes; a
supplemental data collection for our case record review sample to gather information on the role of
the court system in child welfare practice, including court ordered changes to the social worker
recommendations at the jurisdictional/dispositional hearing; and, a statewide survey of county child
welfare directors, managers or supervisors in California counties and quantitative analyses using a
statewide database consisting of county-level characteristics.
Overview of the Report
The next section of this report provides an overview of the study objectives and approach for
Phase 3, as well as the specific research questions that guide the analyses presented in each section.
The third section reviews the recent research on evidence-based child welfare practices and their
application to children and families of color. Section four focuses on the impact of child welfare
practices on children and families of color in the Santa Clara County's Department of Family and
Children Services (DFCS). Several related analyses were conducted to understand the types of
practices and case characteristics that are associated with more successful outcomes for children and
families from various racial/ethnic groups; the types of child welfare practices used in Santa Clara
County; how practices in the Main County Offices differ from South County Office practices; how
5
practices interact with child, family and cultural characteristics of differing racial/ethnic groups to
produce various outcomes; and the role of the court system on child welfare practices. Section five
describes results from our statewide comparative analysis of effective practices for children and
families of color in the child welfare system. A summary of the report, as well as recommendations
and conclusions are contained in section six.
Definition of Terms used in this Report
Racial/ethnic group refers to cultural heritage and country of origin. Racial/ethnic group is used
interchangeably with the terms, children of color and ethnic group.
Disproportionate representation ofchildren ofa particular racial/ethnic group refers to the
difference between the proportion of children of a particular racial or ethnic group in the CWS and
the proportion of children of a particular racial or ethnic group in the general population.
Overrepresentation ofchildren ofa particular _racial/ethnic group exists when the proportion of
children of a certain racial or ethnic group in the CWS exceeds its proportion in the general
population.
Underrepresentation ofchildren of a particular racial/ethnic group exists when the proportion of
children of a certain racial or ethnic group in the CWS is less than its proportion in the general
population.
African American refers to children of African American heritage. The term Black is included in
this category.
Asian American/Pacific Islander refers to children of Asian American and Pacific Islander heritage.
The following population groups are included: Asian Indian, Cambodian, Chinese, Filipino,
Guamanian, Hawaiian, Hmong, Japanese, Korean, Laotian, Polynesian, Samoan and Vietnamese
and other Asian American/Pacific Islanders.
Hispanic/Latino/a refers to children of Mexican American and Latin American heritage. The terms
Hispanic and Latino/a are included in this category.
Native American refers to children of Native American heritage and includes those designated as
Alaska Natives/Aleuts. The term American Indian is included in this category.
White refers to children of European heritage. The term Caucasian is included in this category.
Other category - Those in the Other ethnic subgroup included Native Americans, Ethiopians,
White-Armenians, and White-Middle Easterners
6
II.
OVERVIEW OF STUDY OBJECTIVES AND APPROACH
FORPHASE3
While Phases I and 2 of this study focused on investigating reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where they occur, the
overarching goal ofPhase 3 was to focus on practices that take place within the CWS that either
promote or reduce disproportionality. More information about child welfare practices including
programs, services, strategies, policies, and/or tools can help address racial/ethnic disproportionality
by guiding the development of more effective services and supports for children and families of
color in the CWS.
In an effort to examine ways to address disproportionality within the CWS, the CWRT explored
three overall aims during Phase 3. The first aim included an assessment of identified key practices
for children andfami/ies ofcolor in the Santa Clara County's CWS. Specific research questions
related to this aim include:
•
•
•
•
•
•
What are the key practices in Santa Clara County's CWS that may have an impact on
children and families of color?
What empirical evidence exists for the effectiveness of these practices?
Have these practices been evaluated regarding their efficacy with varying racial/ethnic
groups?
How do these practices interact with child, family and cultural characteristics of differing
racial/ethnic groups involved in Santa Clara County's CWS?
How do practices in South County vs. the Main Office differ? (i.e., are there differences in
worker socialization, training, values/attitudes, stated and unstated expectations norms that
contribute to differences in practice and client outcomes?)
What practice characteristics are necessary in order to promote more positive outcomes for
children and families from different racial/ethnic groups involved in Santa Clara county's
CWS?
The methodology used to explore these research questions included 1) a review of the current
research literature on evidence-based practice for children in the CWS and the application of these
practices to children and families of color, 2) semi-structured in-depth interviews with DFCS
supervisors and managers, 3) semi-structured focus groups with DFCS social workers and parents
in the CWS, 4) a qualitative and quantitative comparison of child welfare practices and outcomes
between the DFCS South County and Main County offices, and 5) a review of cases from Phase 2
of the research with successful outcomes and an analysis of factors that contributed to positive
outcomes.
The second aim of Phase 3 included an exploration ofthe influence of the court system on
child welfare practice, including court ordered changes to the social worker service
recommendations at the jurisdictional/dispositional hearing. We were interested in whether the
court system imposed any changes to the social workers' recommendations offered at the
jurisdictional/dispositional hearing in a child's case, and if these changes differed by the child's
7
ethnicity, as well as the nature of the relationship between the child welfare and court systems.
Specific research questions related to this aim include:
•
•
•
•
To what degree does the court system impose changes to the social workers' case plan
recommendations?
What types of changes to the social workers recommendations does the court system
order?
Do court ordered changes to the social workers' recommended case plan differ by child's
ethnicity?
How does the relationship between the child welfare and court system positively or
negatively impact child welfare practices and outcomes for children and families of color?
The methodology used to explore these research questions included case record reviews of the 403
closed child welfare cases used in Phase 2 of the research. Cases were reviewed again and
information on court ordered changes to the social worker recommendations at the
jurisdictional/dispositional hearing was gathered. Additionally, agency-wide focus groups provided
information on the how the relationship between the child welfare and the court system has an
impact on practices and outcomes for children and families of color.
The third aim of the research involved a statewide comparative analysis of identified key
practices that may affect the disproportionate representation ofchildren ofcolor in the CWS.
Specific research questions associated with this aim include:
•
•
•
What are key practices in California counties that may have an impact on children and
families of color?
What practices are considered most promising by individual counties for children and
families of color?
Comparing counties with large racial/ethnic representation, what factors predict
over/underrepresentation of various racial/ethnic groups?
The methodology used to explore these research questions included 1) a statewide survey of county
child welfare directors, managers or supervisors in California counties and 2) quantitative analyses
using a statewide database consisting of county-level characteristics.
Phase 3 focused on exploring ways to address the disproportionality of children and families
of color within the CWS by examining practices that may either contribute to or reduce ethnic/racial
disparities. In order to accomplish the aims outlined above, we combined qualitative and
quantitative research methods and examined both practices within the system, as well as contextual
factors that can potentially have an impact on what transpires within the system. The report begins
with a review of the literature on effective practices for children and families in the CWS, followed
by a close up view of practices within Santa Clara County and then broadens to include a wider
perspective on statewide child welfare practices aimed at children and families of color. The
following section will describe the background and literature related to effective practices for
children and families of color in the CWS.
8
III. BACKGROUND AND LITERATURE REVIEW
Different racial/ethnic groups experience unique pathways through the child welfare system
(CWS) (Hines et al., 2002). At every point in the system, various child welfare practices and
services impact children and families. Phase 2 results indicated that children and families of color in
Santa Clara County's CWS tend to be recommended services that are traditional and formal in
nature and that do not appear to meet the wide range of needs experienced by these highly diverse
racial/ethnic family groups. Indeed, children and families of color in the CWS represent a high risk
group who are often impacted by myriad psychosocial challenges (see Phase 2 report for a full
description), yet services recommended for these families tend to be limited to a one-size-fits all
approach.
Moreover, there exists relatively little empirical evidence that traditional child welfare
services are effective in maintaining or reunifying children and improving family functioning. Even
less research has been conducted on the impact of child welfare practices for children and families
of color. Yet because children and families of color are disproportionately represented in the CWS,
the effectiveness of child welfare practices for these groups is of particular interest. Research
suggests that children and families of color tend to have longer stays in the CWS, they receive
fewer and less comprehensive services, and have poorer case outcomes than White chHdren and
families (Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996). This bleak
outlook for children and families of color in the CWS creates an urgent need to understand child
welfare practices and services that result in enhanced outcomes for these families. It is also crucial
to identify the types of individual and contextual factors that hinder or support the effectiveness of
these practices.
This section will review the current research on evidence-based child welfare practices and
their application to children and families of color. Practices will be discussed in relation to key
choice points in the CWS including, beginning, continuing care, and ending choice points. Please
see Figure 1 for a flow chart of possible pathways and key choice points in Santa Clara County's
CWS. Additionally, contextual factors including cultural competency of workers and services; the
unique needs of rural child welfare populations; and the role of the court system will be discussed.
Across all stages of the CWS from beginning to end, highlights of practices that appear to be most
effective will be presented and the ways in which these best practices may impact children and
families of color will be discussed.
9
Figure 1: Possible Pathways and Key Choice Points In the Child Welfare System
Referral
(Report of Abuse or Neglect)
Emergency Response
(Screening)
Case Closed
Case Opened:
Further
Investigation
Jurisdictional Hearing:
Decision Regarding Substantiation
of Abuse/Neglect
Case Dismissed:
Unsubstantiated
Allegations
Family
Maintenance
Case Opened:
Voluntary Family
Maintenance
Case Closed:
Short-Term Services
Provided and Family
Stabilized Within 6
Months
Dispositional Hearing:
Substantiated Allegations,
Further Action and Services
Family Reunification:
Placement in FFH, FF A, Group
Home, Kincare, or Other Facility
Bypass
Hearings: Including 6-Month Review
Case Closed:
Family Stabilized
Through Family
Maintenance
Case Closed:
Family Stabilized
Through Family
Reunification
Case Closed:
Permanent Placement into
Adoption, Guardianship,
or Long Term Care
Impact of Child Welfare Practices on Children and Families of Color
Beginning Choice Points
When a child is reported to the CWS, there are various pathways that she or he may take
through the system (see figure I). At the screening stage, the case may be closed altogether with no
further action; it may be opened for six months of voluntary family maintenance services; or the
case may be opened for further investigation. If the case is opened for further investigation, a
jurisdictional/dispositional hearing typically occurs in which there is a judicial finding on whether
the allegation of maltreatment is substantiated. If allegations are found to be unsubstantiated, the
case may be dismissed, however, if substantiation occurs, children may be placed in out-of-home
care, with family reunification services being ordered, or children may be maintained in the home
with orders for family maintenance services. In some cases, families may be bypassed for services
and have their parental rights terminated at this early stage in the system.
Following the pathway that children may take through the CWS, the screening process is the
first entry point into the CWS. On a national level, the U.S. Department of Health and Human
services ([U.S. DHHS], 2001) reports that approximately 2.7 million referrals were received by the
CWS in 2001; 32.7 percent of these referrals were screened out and closed with no further action.
Other national studies have found similar rates of screening out referrals, ranging from 36 percent to
42 percent (Tumlin & Geen, 2000; Wells, Fluke, & Brown, 1995).
Research suggests that certain case characteristics are associated with the decision to open a
case for further investigation or to refer to voluntary family maintenance services. For instance,
reports involving more than one child, children under age two, severe injury, a minor or other type
of injury or physical condition, allegation of sexual abuse, previous reports to the CWS, (Wells et
al., 1995); as well as child welfare cases involving AFDC receipt; alleged parental drug use; a
female primary victim; reports made by a family member; and reports with direct evidence (Karski,
1999) all have a higher likelihood of investigation than cases not involving these characteristics.
Additionally, cases involving children of color are opened for investigation at a higher rate
than cases involving White children. In Wells et al. 's (1995) analysis of 12 child welfare sites,
African American children were significantly more likely to have reports investigated than children
of other ethnicities, however when the data was dis-aggregated and examined by sites, it was found
that almost the entire sample of African American children reported to the CWS came from two
sites that exhibited particularly high investigation rates. Yet, in an analysis of data from five states,
Fluke, Yuan, Hedderson, and Curtis (2003), found that African American children were
significantly overrepresented at the stage of investigation in each of the five states, as well as in all
counties within the five states, while White children tended to be underrepresented at the stage of
investigation. Rates of disproportionality at the investigation stage for other racial/ethnic groups,
including Asian American/Pacific Islander, Hispanic and Native American children showed much
wider variation. Depending on the state, these groups were both underrespresented and
overrepresented at the investigation stage.
Factors associated with the provision of voluntary services are somewhat less well
understood. In a statewide survey in Missouri, Drake ( 1996) analyzed factors associated with the
provision of preventive voluntary services to 3,442 children and families in the CWS, in
11
comparison to cases that were substantiated by the court. Cases were divided into three types of
maltreatment: sexual abuse, physical abuse and neglect. Across all three types of maltreatment,
cases from rural areas were significantly more likely to receive preventive services than cases from
urban or metro areas. Additionally, predictors of preventive service provision in sexual abuse cases
and neglect cases included those that were referred by a non-anonymous source. Physical abuse
cases were more likely to receive preventive services when a mandated reporter reported them.
Neglect cases were more likely to receive preventive services when they involved White children
and families, parents who were unemployed and children and families from non-poor
neighborhoods (Drake, 1996).
For children who are not offered voluntary family maintenance services, investigations
result in a decision regarding whether to substantiate the maltreatment and take further action or to
not substantiate and dismiss the case. Nationally, 27.5 percent of investigated referrals are
substantiated (U.S. DHSS, 2001). State data suggest similar substantiation rates. One study
conducted in New York state revealed a 30 percent substantiation rate across types of maltreatment,
with a 48 percent substantiation rate for physical abuse, 39 percent for sexual abuse and 28 percent
for neglect (Eckenrode, Powers, Doris, Munsch, & Bolger, 1988). Interestingly, there appears to be
little relationship between screening rates and substantiation rates. States with higher screening in
rates do not appear to have higher rates of substantiation of child maltreatment (Tumlin & Geen,
2000).
In California, the overall rate of substantiation of all referrals is 21.2 percent (Needell et al.,
2003). Statewide rates of substantiation between various racial/ethnic groups appear similar; Black
children have a 22 percent substantiation rate, White children 21.4 percent, Hispanic children 23.4
percent, Asian children 24.5 percent, and Native American children 25.3 percent. Across all ethnic
groups, younger children have higher rates of substantiation than do older children (Needell et al.,
2003).
Santa Clara County's overall rate of substantiation in 2002 was 17 .1 percent--a rate
somewhat lower than the state average. Santa Clara County also appears to have greater disparities
between racial/ethnic groups in substantiation rates. For instance, Black children in Santa Clara
County have an overall substantiation rate of21 percent, White children 15.4 percent, Hispanic
children 18 percent, Asian children 15.l percent, and Native American children 15.l percent. Rates
of substantiation between racial/ethnic groups in Santa Clara County also depend on the age of the
child. As is the case within the state, younger children have a higher likelihood of substantiation,
however, in Santa Clara County it appears that Black children between Oand 2 years of age have a
relatively high rate of substantiation (76.8%), whereas, White children 0-2 have a substantiation rate
of 62.6 percent, Hispanic children 52.6 percent; Asian children 46 percent, and Native American
children 66.6 percent (Needell et al., 2003).
Certain factors appear to be associated with the decision to substantiate an allegation. For
instance, reports from professional reporters tend to be substantiated at a higher rate than other
reports (Eckenrode et al, 1988; English, Marshall, Coghlan, Brummel, & Orme, 2002), as well as
cases referred by law enforcement (English et al., 2002). Poverty and early child bearing have been
identified as strong predictors of substantiation (Lee & George 1999). A history of prior referrals or
involvement in the CWS are also associated with higher rates of substantiation (Eckenrode et al.,
1988; English et al., 2002), and one statewide survey in Washington state found that referrals from
rural areas were more likely to be substantiated, although this effect was small (English et al.,
12
2002). Additionally, cooperation with child welfare agencies has been found to be associated with a
lower likelihood of substantiation (English et al., 2002). Factors associated with substantiation also
depend to some degree on the type of maltreatment alleged. For instance, sexual abuse reports that
involve older, female children in large households tend to have a higher substantiation rate than
cases not involving these characteristics,·and the same is true for neglect reports that involve
younger children with previous child welfare referrals (Eckenrode et al., 1988).
Cases involving children of color appear to have an elevated likelihood of substantiation.
One large study conducted in New York state found that for allegations of physical abuse or
neglect, African American and Hispanic cases were substantiated at a higher rate than White cases
(Eckenrode, et al., 1988). In a more recent study conducted in Minnesota all types of maltreatment
were substantiated at a higher rate for non-White groups than for White groups; this effect remained
even after factors such as type of reporter, type of maltreatment, geographic location and
characteristics of the child and perpetrator were statistically controlled (Ards, Myers, Malkis,
Sugrue, & Zhou, 2003). Similarly, in Drake's (1996) statewide analysis of child welfare cases in
Missouri, non-White racial status was a significant predictor of substantiation even when other case
characteristics were statistically controlled. In contrast, Fluke et al.'s (2003) analysis of
disproportionate representation of race/ethnicity in investigation and victimization rates in five
states found little disproportionaility in rates of substantiation of maltreatment.
In addition to substantiation of an allegation, decisions at the beginning stages of the CWS
are also focused on whether to place a child in out-of-home care. Studies have consistently found
certain factors to be associated with the decision to remove a child from home. Children of color, in
particular African American children have an elevated likelihood of out-of-home placement
(Needell, Brookhart, & Lee, 2003; U.S. DHHS, 2001). Other factors associated with the decision to
remove a child include inadequate or low income (Berry, 1991; Lindsey, 1991); cases involving
prior maltreatment reports (Dalgleish & Drew, 1989; U.S. DHHS, 2001); cases involving families
with many children (Berry, 1991); lack of cooperation from the family (Dalgleish & Drew, 1989;
Davidson-Arad, 2001; Karski, 1999); cases involving young children, mothers as perpetrators, and
reports made by educational personnel are all associated with court intervention and placement is
substitute care (U.S. DHHS, 2001).
Few studies have examined factors associated with the decision to bypass family reunification
services and terminate parental rights at the early stages of the CWS. In California, any one of 13
bypass criteria can be used to deny reunification services and pursue an alterative permanent
placement. These criteria include:
1. Parents' whereabouts are unknown for six months.
2. Parental mental disability that prevents the utilization of services.
3. A sibling has been removed from the home, returned and then removed from parental custody
again.
4. Parent has caused the death of a child.
5. Parent caused severe emotional damage.
6. Parent committed severe physical or sexual abuse to the child, sibling or half-sibling.
7. Reunification has been terminated on a sibling or half-sibling due to indicators# 3, 5 or 6.
8. Parent has been found guilty of rape that conceived the child.
9. Willful abandonment constituting serious danger to the child.
10. Parent failed to reunify with a sibling or half-sibling.
11. Parent convicted of a violent felony.
13
12. Parent has extensive history of substance abuse and resisted treatment for three
years prior, or failed to benefit twice.
13. Parent does not want family maintenance or family reunification services, or does not want
the child removed.
In the Bay Area Social Services Consortium's ([BASSC], 2000) report on the use of bypass
criteria within San Mateo and Santa Clara Counties, approximately half of the cases reviewed had
bypass criteria present. Twenty ·six percent of these bypass cases were due to parental substance
abuse, 18 percent were due to a sibling with parental rights terminated or with a permanent plan,
and 13 percent had a parental mental disability that prevented the utilization of services (BASSC,
2000). However, not all cases with a bypass criterion were actually bypassed. Of the cases with a
bypass criterion present, 73 percent of cases involving parental mental disability; 38 percent of
cases with a sibling who had parental rights terminated or who has a permanent plan; and 13 percent
of cases involving parental substance abuse were actually bypassed. Overall 25 percent of the entire
sample were bypassed and denied reunification services. Additionally, BASSC (2000) reported that
the use of bypass criteria is not empirically or clinically valid and that bypass criteria are not
predictive of a lowered likelihood for reunification. Moreover, the report suggested that that since
many bypass indicators are common within the child welfare population, their use in denying family
reunification services may not be warranted.
Beginning Choice Points: Impact on Children and Families of Color
Research clearly indicates that certain/actors are associated with various decisions at
beginning choice points and these factors disproportionately impact children and families of color.
In general, the most consistent factors found to increase the likelihood of investigation,
substantiation or out-of-home placement include cases involving more than one child, young
children, previous reports to the child welfare system, and families with low incomes (Eckenrode et
al., 1988; English et al., 2002; Karski, 1999; Lee & George, 1999; Wells et al., 1995). Children and
families of color are at an increased risk for all of these factors, but in addition, studies have
indicated that even when these factors are statistically controlled, children of color, and in particular
African American children are more likely to have an allegation investigated and substantiated and
are also more likely to be placed out of the home (Ards et al., 2003; Drake, 1996; Wells et al.,
1995). Additionally, there is evidence to suggest that in cases of neglect, children and families of
color may receive voluntary family maintenance services less frequently than White children and
families (Drake, 1996). Taken together these findings suggest that beginning choice point decisions
may be biased against children and families ofcolor. Indeed it appears from the research that once
a child ofcolor is reported, they are more likely than White children to be set on a pathway through
the child welfare system that involves court intervention and out of home removal.
Decision-Making Practices at Beginning Choice Points
The types of decision-making practices that child welfare workers use at these early stages
of the CWS are crucial in determining the trajectory of the pathway that children and families take
through the system. However, decisions on whether to investigate a referral, provide voluntary
services, substantiate an allegation, remove a child or bypass a case are often complex, ambiguous
and multifaceted. Child welfare workers must simultaneously consider numerous factors, including
child safety, the best interests of the child, parental rights and preserving family integrity. These
14
factors must be considered in a complicated context; children and families in the CWS often face
multiple and complex challenges and child welfare policies and mandates fluctuate frequently.
Given this difficult context, it is not surprising that studies have found a significant lack of
consistency in child welfare decision-making practices across various stages of the system (Rossi,
Schuerman, & Budde, 1999; Tumlin & Geen, 2000; Westat, 2001a). For instance, disparities in
screening out rates between states have been found to range from a low of 5 percent to a high of 78
percent (Tumlin & Geen, 2000), while disparities between counties have been found to range from a
low of 1 percent to a high of 71 percent (Wells, et al., 1995). There is some evidence to suggest that
child welfare workers may use more restrictive screening practices when resources or staff are
reduced or when the number of referrals is high (Curtis, Boyd, Liepold, & Petit, 1995 as cited in
Tumlin & Geen, 2000; Hutchinson, 1989, as cited in Jones, 1993). Moreover, screening decisionmaking practices tend to be influenced by the average severity level of cases within the agency as a
whole and characteristics of the surrounding community. In one survey, workers in 11 child welfare
agencies in the state of New Jersey rated vignettes of child maltreatment reports as less severe in
agencies that handled the more severe cases of child maltreatment. These agencies were also located
in areas that were more socially and economically disadvantaged-suggesting that characteristics
specific to agency sites or communities influence screening and substantiation decision-making
practices (Wolock, 1982).
Although virtually all states and counties screen child maltreatment reports to some degree,
screening practices and policies vary greatly between states and counties. National surveys suggest
that a majority of states do have some form of written agency policies to guide screening and
investigation decisions (Downing, Wells & Fluke, 1990); yet few states have formal or explicit
procedures or instruments delineating reports that should be screened out {Tumlin & Geen, 2000).
Approximately 65 percent of states have a single review screening process in which the decision to
investigate is made by one worker (Tumlin & Geen, 2000). Moreover, the training of child welfare
workers who screen referrals has also been found to vary widely from state to state; with some
states having educational or work experience requirements for screening workers, while others do
not {Tumlin & Geen, 2000).
Studies investigating consistency between child welfare workers in their decision-making
practices indicate wide variability in these practices. For instance, Rossi et al. (1999) explored
decision-making practices among 130 child welfare workers and experts in three states. Results
indicated that although certain factors were associated with the decision to maintain a child in the
home, including families with no prior referrals and cases involving employed caretakers who
showed signs of engagement in services, the thresholds used to trigger an out-of-home placement
decision were not consistent. For instance, some participants decided on out-of-home care in cases
involving just one prior referral, while others recommended out-of-home placements in cases
involving two or more prior referrals. Using the same data set, Schuerman, Rossi and Budde (1999)
calculated the level of agreement between workers on decisions about services, as ranging from a
high of 64 percent to a low of 48 percent, indicating a considerable lack of consensus among
workers regarding out-of-home placement decisions.
Studies also suggest that child welfare workers consider numerous factors in their decisionmaking practices and that these factors are generally not considered in isolation from one another
(English et al., 2002). For instance, a recent study by the U.S. Department of Health and Human
15
Services (2003a) identified four overall factors that influence decisions regarding whether to
substantiate an allegation, including:
1) Case factors, such as child characteristics, severity of the maltreatment, chronicity of
maltreatment, caretaker characteristics, caretaker/child relationship, social and economic
factors, and perpetrator access to the child.
2) Decision-maker factors, such as caseworkers' individual tendencies, experience level,
relationships with co-workers, self-assessments of skills and degree of importance
placed on state policy.
3) Organizational factors, such as workload stress, and degree of supervisor support.
4) External factors, such as state policies.
All of these factors were found to impact the decision-making process of child welfare workers in
complex ways that may differ between individual workers, counties and states (U.S. DHHS, 2003).
Yet there is some evidence to suggest that decision-making practices within a particular
agency may be more consistent than between agencies. Alter ( 1985) has suggested that child
welfare workers tend to use heuristics or assumptions to guide decision-making and that workers
within a particular agency often develop a mutual, although unspoken understanding of the kinds of
cases that should be investigated. Indeed there does appear to be greater consistency in initial
decision-making practices within agencies than between agencies. For instance, Wells et al.'s
(1995) study of 12 child welfare sites, found wide variability between the sites in their investigation
decision-making practices, yet when all other factors were held constant, the site of the report was
more significant than any other factor in predicting an investigation. These results suggest that the
type of decision-making practices common to a particular child welfare agency play an important
role in decision-making practices related at the beginning stages of the CWS.
In addition to the importance of agency norms in decision-making practices, the labels used
to describe families may also trigger certain decisions. For instance, Martin, Peters, and Glisson
(1998) surveyed 100 child welfare workers in five states and found that workers did not make
service and placement recommendations for children based on a uniform psychosocial assessment.
Instead, decisions were influenced more by the labels attached to children upon entry into care and
judicial decisions. Children whose files contained language such as "substance abuser, child of
substance abuser, sex offender, victim of sexual abuse and victim of physical abuse" (Martin et al.,
1998, p. 3), were more likely to be recommended mental health services and more restrictive
placements-regardless of their actual need as indicated in the psychosocial assessment. Similarly,
children who were adjudicated delinquent or who were placed in out-of-home care due to behavior
problems rather than maltreatment were also recommended mental health services and more
restrictive placements, regardless of their actual need.
The influence of agency norms, the use of implicit assumptions and the complexity inherent
in child welfare decisions has prompted many practioners and researchers to call for more objective
or collaborative decision-making practices. Structured risk assessment tools, multidisciplinary
teams, and family group conferencing are considered by many to offer a more objective, balanced
and inclusive method of decision-making at beginning choice points.
16
Decision-Making Practices: Risk Assessment Tools
Structured risk assessment tools are designed to assist child welfare workers in making
accurate assessments at various stages in the CWS. Risk assessment tools generally evaluate factors
that are considered to be associated with various outcomes including such items as prior referrals
and prior substantiated cases; the presence of substance abuse, mental health or economic problems,
family constellation, the type, severity and chronicity of the maltreatment and circumstances
surrounding the current incident (Camasso & Jagannathan, 2000) Risk assessment tools fall into
two general categories: 1) consensus based systems in which risk factors in the instrument are based
on the consensus ofjudgments made by experts in the field, and 2) actuarial systems in which risk
factors are identified based on empirical evidence of factors associated with future maltreatment
(Baird & Wagner 2000). There is some evidence to suggest that actuarial systems more accurately
capture levels of risk in families referred to the CWS (Baird & Wagner, 2000).
The use of risk assessment tools varies greatly from state to state and county to county, and
these tools can be used at various decision points in the CWS from intake to case closing (Wald &
Woolverton, 1990). Although risk assessment tools are widely used in CWS agencies, their ability
to accurately assess levels of risk is an area of considerable debate in the child welfare field (Wald
& Woolverton, 1990). Several researchers have noted that risk assessment instruments have only
minimal reliability and validity in predicting case outcomes (Camasso & Jagannathan 2000;
DePanfilis & Scannapieco, 1994; Doueck, English, DePanfilis, & Moote, 1993; Wald &
Wooverton, 1990).
In 1998, certain California counties, including Santa Clara County began using a
comprehensive risk assessment system, known as Structured Decision-Making (SOM). SDM has
several components including response priority, safety assessment, risk assessment, family needs
and strengths assessment, case planning and management, case reassessment, workload based
resource allocation, and the use of management information systems (Children's Research Center,
2003). These components are designed to assist workers in assessing the severity of allegations,
making structured decisions, managing workloads, minimizing the trauma associated with
maltreatment and preventing the recurrence of maltreatment (California Department of Social
Services, 2003). The risk assessment tool used in this system is an actuarial instrument that is
intended to estimate the likelihood that maltreatment will recur. In addition to the risk assessment,
SOM also calls for a family needs assessment in which family's strengths and resources are to be
considered in conjunction with the risk assessment. Although peer reviewed evaluations of the
SDM project are not available, the Children's Research Center (as cited by the California
Department of Social Services, 2003) reports that a 12-month evaluation on the SOM project was
conducted in Michigan. Results indicated that cases receiving SDM had 27 percent fewer new
referrals, 54 percent fewer new substantiated allegations, 40 percent fewer children removed to
foster care, and 42 percent fewer child injuries than children not receiving SDM. Although these
results are promising, more evaluations are needed to discern the effectiveness of SOM, specifically
for children and families of color.
Decision-Making Practices: Multidisciplinary Teams/Family Group Conferencing
The use of multidisciplinary teams to assist in child welfare decision-making has gained
popularity in recent years. Multidisciplinary teams are designed to better coordinate efforts between
child welfare, law enforcement, legal professionals and medical professionals in making decisions
17
about child welfare cases. These efforts are intended to improve the efficiency and effectiveness of
decision-making activities and also in some cases-to reduce the potential for traumatizing children
through repeated investigatory interviews (Sheppard & Zangrillo, 1996). Approximately two thirds
of states have been found to use multidisciplinary teams in making child welfare decisions (Kolbo
& Strong, 1997).
Research on the effectiveness of multidisciplinary teams in improving decision-making in
child welfare is lacking. In one of the few studies to address this issue, Hochstadt and Harwicke
(1985) evaluated the use of multidisciplinary teams for 180 children referred to the CWS for
suspected child maltreatment. The multidisciplinary team included child welfare workers, medical
professionals and psychological personnel who completed a multidisciplinary evaluation and
participated in a case conference in which specific service recommendations and treatment plans
were made. Although no comparison group was used in the study, findings suggested that the use of
multidisciplinary teams was associated with a high percentage of children and families obtaining
recommended services on their treatment plans. One hundred percent of children with a
recommendation to return home did so, and 92 percent with a recommendation to foster care
followed that recommendation. Other service recommendations that were received included visiting
nurse (76%), additional medical assessment (64%), infant program (64%) and individual
psychotherapy (59%). The authors suggest that the use of multidisciplinary teams can positively
influence the receipt of recommended services.
In an effort to discern how multidisciplinary teams in child welfare work together,
Lewandowski and Glen Maye (2002), conducted a survey of 165 child welfare and community
professionals who were a part of a multidisciplinary child welfare team in an urban Midwest
county. The sample included social workers, counselors, community members, education
professionals, law enforcement, legal professionals and medical professionals. Findings revealed
that 72 percent of respondents had team meetings at regularly scheduled times with support staff
assigned to the team, and 53 percent had a designated team chairperson. Forty eight percent of the
sample indicated that decision-making was most likely to occur around case planning, 32 percent
indicated that decision-making was least likely to occur at intake, and 28 percent reported that
decision-making occurs at all stages. Inadequate communication was identified as the greatest
barrier to effective team decision-making, followed by inadequate resources and inadequate
professional respect. Lastly, although most respondents (63%) agreed that family members should
be active and equal participants in the team decision-making process, only 35.5 percent reported
that family members attended team meetings.
Recently, more concerted attempts to involve family members and extended support systems
in decision-making have occurred. Many child welfare systems are moving toward the use of family
group conferences to improve decision-making. The use of family group conferences began in New
Zealand in 1989, when legislation mandating their use was implemented (Sieppert, Hudson, &
Unrau, 2000). Family group conferences, also referred to as family group decision-making, are
designed to bring together family members, relatives, and other support systems in order to make
decisions about a case. The family group conference is intended to identify the family's strengths
and resources; to develop a plan to ensure child safety and improve family functioning; and to foster
cooperation, collaboration and communication between families and professionals (American
Humane Association, 2003: Pennell, & Buford, 2000). These methods are based on the principle
that families themselves possess the most information about what decisions should be made; the
approach is intended to be family centered, strengths based, and takes into consideration issues of
18
culture and community (American Humane Association, 2003).
In addition to family group conferencing, recent efforts have also been aimed at "team
decision-making," a closely related concept that appears to widen the scope of participants in the
decision-making process. Team decision-making is one component of the Casey Foundation's
Family to Family Initiative, a broad child welfare effort aimed at improving outcomes for children
in the CWS (see page 31 for a discussion of the Family to Family Initiative). The team decisionmaking component of the initiative involves convening a broad range of participants to share
information about the family and develop a plan for child safety, family maintenance or
reunification. The team decision-making meeting can involve any participants who can contribute to
the decision-making process, including family members, neighbors, foster parents, service
providers, child welfare workers and supervisors, other resource staff or other community
representations (Casey Foundation, 2003a).
Research on the effectiveness of family group conferencing and team decision-making on
outcomes for children and families is lacking, although some studies have investigated the process
of implementing family group conferences. For instance Sieppert, et al. (2000) investigated the
process of implementing a family group conferencing pilot project in Canada. The sample included
23 child welfare families with court involvement and one family group conference was conducted
for each family. Mothers were the most frequent family members to attend the conferences (91 %),
while 52 percent involved a father, 70 percent involved grandmothers, 57 percent involved aunts,
and 35 percent involved grandfathers. Other members of the conferences included child welfare
workers (83%), and foster parents (30%). Findings revealed that family group conferences tended to
be resource-intensive. Workers performed a wide range of activities in preparing for and
implementing the family group conferences. An average of 17 telephone contacts, 2.4 in-person
contacts, and 3 letter contacts were made prior to each family conference. Workers also spent time
traveling to meet with prospective conference attendees. On average, workers spent 7 hours
preparing for each conference. Family conferences lasted an average of2.75 hours. At the end of
the conference, families were asked to rate their experience of the group. Seventy percent reported
being highly satisfied with conference preparation and members in attendance; 85 percent felt a
high degree of freedom to speak out in the group; 76 percent reported feeling involved in the
decision-making process and 65 percent felt highly satisfied with the decisions that resulted from
the conference. Unfortunately the ethnicity the families involved in the family group conferences
was not provided.
Decision-Making Practices at Beginning Choice Points:
Impact on Children and Families of Color
In the beginning stages of the child welfare system, the decision-making practices ofchild
welfare workers can have a substantial impact on children and families ofcolor. However in
general there appears to be considerable inconsistency in the types of decision-making practices
used in the child welfare field (Rossi et al., 1999; Schuerman et al., 1999; Tumlin & Geen, 2000;
Westat, 2001a). Although many child welfare systems are moving toward more objective or
multidisciplinary approaches to decision-making, a majority of states still use a single review
process in which decisions to investigate are made by one worker (Tumlin & Geen, 2000); and few
states have formal policies and procedures for decision-making at early stages of the CWS (Tumlin
& Geen, 2000).
19
However, making child welfare decisions in isolation can have deleterious effects on
children and families of color. Research suggests that child welfare workers often use implicit
assumptions to guide decision-making. Ways in which each worker thinks about diverse families,
and the overall spoken and unspoken expectations and norms of a particular child welfare agency
can have an enormous impact on decision-making practices (Cohen, 2003). As Cohen (2003) notes,
The cultural and racial background of families influences the specific factors that workers
consider in assessing the severity of risk and level of intervention. Decisions are more likely
to be made on the basis of deficits in available resources, accepted· agency practice, personal
values and biases, and notions of an ideal family, than by application of consistent case rules
(Chohen, 2003, p. 145).
It is speculated that in some cases, race/ethnicity may impact decision-making by resulting in a
higher level of intervention, while in other cases race/ethnicity may result in less intervention effort.
Much of this variation appears to·depend on the worker and the context in which the worker makes
decisions (Cohen, 2003).
The use of risk assessment tools and multidisciplinary teams/family group conferencing may
have the potential to improve decision-making practices affecting children and families of color.
Although research on the effectiveness of risk assessment tools in assessing risk for future
maltreatment has not been consistently demonstrated in the research literature, the move toward
more objective ways of assessing risk at beginning stages of the child welfare system may be
promising for children of color. For instance, in one study conducted by the California Department
of Social Services (1999), 2,500 substantiated cases from counties that used the SDM system, were
evaluated to assess whether the risk assessment instrument used in SDM had any differential impact
on African American, Hispanic, or White children and families. Findings indicated that Whites had
slightly higher scores on the risk assessment instrument than did African Americans or Hispanics
and the authors concluded that the risk assessment instrument did not disparately affect African
Americans or Hispanics. While this finding is encouraging, other studies have suggested that risk
assessment instruments in general are not always able to accurately predict case outcomes Camasso
& Jagannathan 2000; DePanfilis & Scannapieco, 1994; Doueck et al., 1993; Wald & Wooverton,
1990). More research is needed on the impact of risk assessment tools in general, and their impact
on children and families of color in particular.
Additionally, more research on the effectiveness ofmultidisciplinary teams and family group
conferencing is needed The family centered, strength based and inclusive philosophy offamily
group conferences may be a particularly useful strategy for children and families of color, many of
whom are experiencing numerous risk/actors in a highly stressed environment. The active
engagement offamilies and support systems in decision-making at early stages of the child welfare
system may potentially result in enhanced outcomes for children and families of color.
Yet there is some evidence to suggest that these: approaches can be labor, time and resource
intensive (Sieppert et al., 2000). As such, the effective implementation of these practices may
require the use of increased support staff to assist in scheduling and arranging meetings, as well as
identifying persons to be included in the meeting. Additionally, because beginning choice point
decision-making practices are often made quickly and in the context of crisis situations, the effective
use of multidisciplinary teams and family group conferences requires coordinated efforts and
appropriate support staff. Clearly, more research is needed to determine how group decision-
20
making can be effectively integrated into child welfare agencies and how these practices impact
outcomes for children and families ofcolor.
Voluntary Family Maintenance Practices/Family Preservation Services
In cases where maltreatment is not substantiated, families may be referred to voluntary
family maintenance services. Because the overall goal of voluntary family maintenance services is
to prevent an out-of-home placement and further involvement with the CWS, these services are
considered beginning choice point practices. Much of the research literature on services and
practices aimed at preventing out-of-home placement refer to these services as "family preservation
services," (FPS) or "intensive family preservation services" (IFPS). Many researchers and
practioners have noted that services falling under the rubric of "family preservation services" can be
quite diverse, although they do tend to share the common goal of preventing child maltreatment and
out-of-home placements (United States General Accounting Office, 1997). The model of FPS that is
most often described in the research literature includes short-term, intensive and home-based
services that are aimed at family skill-building, while also including concrete services, such as
public assistance, child care, employment training or transportation assistance (Littell, 1997).
In general, there are numerous shortcomings in the research on the effectiveness of FPS;
many studies use small sample sizes, no comparison groups and often fail to adequately describe the
nature of the intervention and the characteristics of the clients. As a result, evaluations of FPS often
yield conflicting results. While some studies report that FPS do not result in lower placement rates
or improved family functioning (Chaffin, Bonner, & Hill, 2001; Littell & Schuerman, 1995; Littell,
1997; Westat, Inc. 2001b); other studies suggest that FPS do improve outcomes for children and
families (Pecora, Fraser, Bennett, & Haapala, 1991; Scannapieco, 1994; Spaid, Fraser & Lewis,
1991; Wells & Whittington, 1993).
There is also evidence to suggest that FPS outcomes vary depending on case characteristics.
A number of different studies have noted that FPS clients who experience an out-of-home
placement typically have more risk factors and less resources than families who avoid placement.
The particular risk factors affecting FPS clients who experience out-of-home placements have been
found to vary wideJy. For instance, FPS cases involving substance abuse, mental illness, criminal
behavior, disability, low income, or previous CWS involvement have all been linked to an increased
likelihood of an out-of-home placement (Fraser, et al., 1991; Thieman & Dail, 1997; Thieman et al.,
1990, as cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman,
1995).
Voluntary Family Maintenance Practices/Family Preservation Services:
Impact on Children and Families of Color
Research on voluntary family maintenance practices and FPS has not systemically
evaluated the efficacy of these programs for diverse children and families ofcolor. Many studies
fail to perform analyses examining the possible differential outcomes of these services for children
and families of color. In general, research on FPS has not consistently demonstrated that these
services are effective in preventing out-of-home placement or improving family functioning
(Chaffin et al., 2001; Littell & Schuerman, 1995; Littell, 1997; Westat, Inc. 2001b), and some
researchers have concluded that there is no reliable research that supports the effectiveness of these
services (Littell & Schuerman, 1995). However, some studies do suggest that FPS may be most
21
effective with lower-risk families (Fraser et al., 1991; Thieman & Dail, 1997; Thieman et al., 1990,
as cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman, 1995).
Yet because children and families of color referred to the child welfare system often possess
numerous risk factors (Hines et al., 2002), the use of traditional FPS for these diverse clients may
not be effective. Research does suggest that FPS that are most effective with high-risk clients are
those that are more intensive, and those that involve high client participation and the delivery of
concrete services in the home (Berry, 1992; Cash & Berry, 2002; Nelson & Landsman, 1992, as
cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman, 1995).
Home-Based Prevention Efforts
Research suggests that the use of home-based services may be effective in preventing
maltreatment and further involvement in the CWS (Olds, Eckenrode, Henderson, Kitzman, Powers,
Cole et al., 1997). Most research on the use of in-home services has focused on their efficacy for
pregnant women or mothers of young children. While the nature of in-home services may vary, in
general these services are aimed at improving health behaviors of pregnant and parenting women,
facilitating parenting skills, increasing support systems and encouraging healthy parent child
attachments (Olds et al., 1997). Many practioners and researchers have suggested that the perinatal
period is an important window of opportunity and that services offered during this time have
increased potential to offset a number of negative outcomes, including child maltreatment
(Guterman, 1999).
For instance, in one study involving pregnant women in rural New York state, 324 women
participated in a home nurse visitation program that sought to improve health related behaviors
during pregnancy, parenting behaviors and the personal life course development of the mother
(Olds et al., 1997). Specific services included linking families with health and human services;
involving family members and friends in the care of the child; educational services to increase
parenting abilities; and assisting mothers in completing education, obtaining employment and
family planning. Participants received services for two years; outcomes assessed at a 15-year
follow-up indicated that compared to pregnant women who received no home visitation services,
mothers in the program had significantly fewer verified reports of child maltreatment in which they
were the identified perpetrator. The program appeared to have even more of a protective effect on
unmarried women and women experiencing income problems. These groups were the least likely to
have a verified child maltreatment report. Women in the program were also less likely to have
substance abuse problems, had fewer contacts with the criminal justice system, had fewer
subsequent pregnancies and were less likely to use public assistance (Olds et al., 1997).
Interestingly, a separate analysis of the same data set indicated that the protective effects of
home visitation in reducing the likelihood of child maltreatment did not hold true for mothers
experiencing domestic violence. Mothers in the home visitation program who reported more than 28
incidents of domestic violence during the 15-year follow-up period did not experience a reduced
likelihood of verified child maltreatment (Eckenrode, Ganzel, Henderson, Smith, Olds, Powers et
al., 2000). These results suggest that different risk factors may have varying degrees of influence on
outcomes for home visitation program participants.
In addition to child maltreatment prevention, in-home services have also been identified as a
useful public health strategy to improve health related behaviors and outcomes for mothers and
young children. Kitzman, Olds, Henderson, Hanks, Cole, Tatelbaum et al. (1997) evaluated a home-
22
visiting program that involved in-home services delivered by nurses to 1,139 low-income, primarily
African American women during pregnancy and the first two years of the child's life. Results
indicated that, compared to women who had no home-visiting intervention, women who received
in-home services experienced fewer injuries and fewer hospitalizations for their children, and also
experienced fewer pregnancies and a lower rate of pregnancy induced hypertension.
Not surprisingly, research also suggests that the nature of the interventions provided through
in-home programs plays a role in their effectiveness. For instance, some studies on in-home services
report no increased benefit from their use (Barth, 1991; Fraser, Armstrong, Morris, & Dadds, 2000).
However it appears from the literature that in-home programs are most effective when health or
social services professionals, as opposed to paraprofessionals, deliver the services. In-home services
that are long term, generally lasting longer than 6 months are also considered to be more effective
(Barth, 1991; Olds et al., 1997).
Home-Based Prevention Efforts:
Impact on Children and Families of Color
Although differential outcomes for various racial/ethnic groups receiving home-based
prevention services are not often cited in the research, studies do generally support the effectiveness
of these services for vulnerable parenting women. Participation in home-based services has been
linked to a reduced likelihood ofchild welfare system involvement, substance abuse problems,
contact with the criminal justice system, public assistance usage, subsequent pregnancies (Olds et
al., 1997), as well as fewer child injuries and hospitalizations, and better health during pregnancy
(Kitzman et al., 1997).
Additionally, there is evidence to suggest that home-based services are effective with highrisk populations and families of color. For instance Olds et al. (1997) found that home-based
interventions were associated with even better outcomes among unmarried and low SES mothers,
and Fraser et al. (2000) reported that mothers with poor attachment to their child and who had a low
sense of competence were more likely to remain in a home-based program. Kitzman et al. ( 1997)
also noted that in-home services were associated with better outcomes for a sample of primarily
African American mothers. Furthermore, McGuigan, Katzev, and Pratt (2003) reported that older
Hispanic mothers had the greatest likelihood of remaining in a home-based program, suggesting
that in-home services may be particularly useful with families of color.
However research a,so suggests that the effectiveness of home-based interventions depends
to some degree on the types of risk factors present and the components of the interventions. For
instance, the presence of domestic violence has been found to eliminate the protective effects of
home-based interventions (Ekenrode et al., 2000); and unmarried mothers with less than a high
school education; mothers Who received no prenatal care; and mothers who have had multiple births
may be less likely to benefit from home-based programs than mothers without these challenges
(Murphey & Braner, 2000). Furthermore, mothers living in communities with a high rate of
violence have been found to have an increased likelihood of dropping out of home-based programs
(McGuigan et al., 2003); and teenage mothers and those with insecure housing also appear to have
an increased risk of dropping out of home-based programs (Fraser et al., 2000). It also appears
from the literature that home-based programs may be most effective when they last more than 6
months; when they are delivered by health or social service professionals (Barth, 1991; Olds et al.,
23
1997); and when there is a high level of supervision for the workers delivering the services
(McGuigan et al., 2003).
In general, home-based prevention efforts appear to be a promising best practice for
vulnerable child welfare populations. More research is needed to discern what the necessary
components of intervention are and to what degree these interventions differentially impact diverse
groups. Additionally, it is important to note that home-based interventions may be time, labor and
resource intensive and as such, child welfare agencies may have difficulty implementing these
services. However, clients of these services clearly appear to benefit and home-based programs do
tend to result in enhanced outcomes for parents and children.
Continuing Care and Ending Choice Points
If child maltreatment is substantiated, children typically enter the CWS under some form of
court supervision. Generally families who are not bypassed for services are assigned to either family
maintenance or family reunification services (see figure 1). Under the Adoption and Safe Families
Act of 1997, families generally receive up to 18 months of family reunification services, at which
point they are either reunified or parental rights are terminated and the child enters post permanency
planning services. As such, continuing care and ending choice point practices generally include
family reunification services, including specific treatments for parents; as well as services for
children in out of home care, services for alternate caregivers, and permanency planning services
for children who are not reunified with their birth parents.
Factors Associated with Family Reunification
Studies have clearly identified certain factors that are predictive of reunification or nonreunification. Research consistently demonstrates that White children are reunified at a higher rate
than are Black children (Courtney, 1994; Jones, 1998; McMurty & Lie, 1992; Wells & Guo, 1999;
Westat, Inc., 2001c); and young children tend to be reunified more often than older children
(Courtney, 1994; Goerge, 1990; Westat, Inc., 2001c). Harris and Courtney (2003) report that single
parent African American families are particularly unlikely to reunify, while two-parent Hispanic
families have a significantly higher likelihood of reunification than two-parent Caucasian or African
American families. Socioeconomic status also plays a role in reunification; receipt of AFDC and
lower socioeconomic status is associated with a deceased likelihood of reunification (Barth,
Snowden, Broeck, Clancey, Jordan, & Barusch, 1987; Courtney, 1994; Jones, 1998), and Westat,
Inc. (2001c) found that parents with a high school education who were currently employed were
more likely to be reunified than parents who do not have these attributes. Interestingly, Wells and
Guo (2003) report that mothers who move from welfare to work actually reunify _with their children
at a slower rate than other groups. Some studies have also found that children with disabilities or
health problems are less likely to be reunified than children without these problems (Courtney,
1994; McMurty & Lie, 1992; Wells & Guo, 1999). Additionally, parents who are non-compliant
with court orders and treatment are significantly more likely to lose custody of their children than
are court compliant parents (Atkinson & Butler, 1996; Jellinek, Murphy, Poitrast, Quinn, Bishop &
Goshko, 1992; Smith, 2003).
Demographic information indicates that most children in family reunification services come
from single parent homes, often headed by the mother (Fein & Staff, 1993; Gillespie, Byrne, &
24
Workman, 1995; Lewandowski & Pierce, 2002; Walton, Fraser, Lewis, Pecora, & Walton, 1993).
Financial problems are common among family reunification clients, (Fein & Staff 1993; Gillespie et
al., 1995; Walton et al., 1993); as are previous out-of-home placements (Lewandowski & Pierce,
2002; Walton et al., 1993). In California, 16.l percent ofBlack children are in family reunification
services; 25.l percent of White children; 24.8 percent of Hispanic children, 20.3 percent of Asian
children, and 22 percent ofNative American children are in family reunification services (Needell
et al., 2003). Children in family reunification services in California also tend to be young; 43.8
percent of children under the age of one are in family reunification services; 32.9 percent between
the ages of one and two; and 27.6 percent between the ages of three and five are in family
reunification services. In Santa Clara County, 15.7 percent of Black children are in family
reunification services, 24.3 percent of White children, 25 percent of Hispanic children, 20 percent
of Asian children and 13.3 percent of Native American children are in family reunification services.
In Santa Clara County, 33.3 percent of children under the age of one are in family reunification
services, 32.8 percent between the ages of one and two; and 30.7 percent between the ages of three
and five are in family reunification services (Needell et al., 2003).
Family Reunification Services
Family reunification services are generally focused on addressing the broad range of issues
that have contributed to the removal of the child, or that prevent the child from returning home
safely. These services are grounded in the notion that the parent child attachment is of primary
importance and the continuity of these attachments is critical to the well-being and development of
the child (Maluccio & Ainsworth, 2003). Family reunification services are generally offered for a
period lasting 6-18 months and they may include a wide range of practices that vary greatly from
county to county and state to state. As is the case with family preservation practices, a large number
of different interventions fall under the definition of "family reunification services;'1 some family
reunification interventions are modeled after FPS and are intensive, short term and family-centered;
while others focus on the delivery of specific services designed to ameliorate particular problems.
As with research on the effectiveness of family preservation services, research on the
effectiveness ofFRS is lacking; most family reunification studies have small sample sizes, no
comparison groups, no clear definition of the intervention and often fail to measure family and child
well-being in addition to reunification rates (Littell & Schuerman, 1995). Studies evaluating the
impact ofFRS demonstrate mixed results, with many researchers concluding that there is no solid
evidence that FRS are effective in returning children to their birth families (Littell & Schuerman,
1995). Indeed rates of reunification for participants in FRS have been found to range all the way
from 25 percent to 100 percent (Littell & Schuerman, 1995).
For instance, Fein and Staff (1993) evaluated the effectiveness of a fairly intensive family
reunification program in which services were delivered by both a social worker and a family
support worker; workers carried small caseloads and provided a broad range of both concrete and
clinical services, such as parenting training, budgeting assistance, transportation, help with job
training, as well as counseling, or support for substance abuse treatment. Although no comparison
group was used in the study, results indicated that 38 percent of children were reunified with their
families after one year of participation in services-and of these 10 percent were returned to OHC
at the end of the second year.
25
Some researchers have concluded that family reunification services that are intensive in
nature, sometimes following a family preservation model, may be most effective. Walton et al.
(1993), evaluated the effectiveness of family reunification services that were limited to 90 days and
were delivered in the home of the birth parents by workers carrying small caseloads. Interventions
averaged 3.1 hours a week and were focused on concrete services, as well as psychosocial training.
Findings indicated that 93 percent of families receiving FRS reunified after the 90-day treatment
period, compared to 28 percent of a comparison group receiving regular child welfare services. At a
12-month follow-up, the FRS children group (75%) was still more likely to be living in the home,
than the comparison group (49%).
Similarly, Gillespie et al. (1995) also reported promising results from an intensive
reunification program that was modeled after family preservation services. Services were provided
at the birth parents home, for approximately 8 to 10 hours a week and included a combination of
concrete and clinical services. In addition, this program also provided support and training to the
foster parents caring for the children, increased contact between the foster parents and social
workers, as well as the facilitation of joint meetings between foster parents and birth parents to
improve consistency of parenting and to share information. Birth parents were also allowed
increasing numbers of visits based on their participation in services. Although no comparison group
was used, findings indicated that 79 percent of the children were reunified, and at 12-month followup 71 percent of the sample was sill living at home.
Less encouraging results were noted by Lewandowski and Pierce (2002) in their evaluation
of the effects of Missouri's Family Centered Out Of Home Care (FCOHC) Program, a program that
incorporates family-centered principles aimed at involving families in reunification case plans as
soon as the child enters out of home care. Services were intensive, time-limited and strengths-based.
Progress was monitored through team meetings involving the families, social workers, and other
representatives from social service systems. Differences between families receiving the FCOHC
and comparison groups receiving standard child welfare services were significant although not in
the expected direction. The FCOHC group actually experienced fewer reunifications and more
returns to out of home care, than the comparison group. Yet results did indicate that, on average,
children in the FCOHC group who spent more than seven days in out of home care, also spent fewer
days overall in out of home care than the comparison group.
In addition to the impact of family reunification services themselves, the decision-making
practices of workers also play a role in whether or not families reunify. In Westat's (2001a) analysis
of reunification decision-making practices among nine workers in three public child welfare
agencies in Washington DC, workers reported they could generally tell within the first few months
of a reunification case whether reunification would actually occur. Factors identified by workers as
important in their decision to return a child home included, the parents' compliance with services
and visitations, the strong desire of parents to have their children returned, the safety of the child
and the child's wishes and well-being. Workers also indicated that parents who were noncompliant
with services or non-responsive to efforts at reunification and cases involving young children were
most likely to lead to a decision to terminate parental rights. In general workers in this study felt
that time limits for reunification were a positive factor in the decision-making process because time
restrictions allow them to terminate parental rights more quickly in cases where reunification
appears unlikely; however workers also noted that some families with more serious problems could
be reunified if they received services beyond federal time limits. However since the sample size in
this study was small, these results should be interpreted with some caution.
26
Family Reunification Services:
Impact on Children and Families of Color
The impact offamily reunification services for children and families ofcolor is not
adequately addressed in the research literature. Although we know that children and families of
color tend to reunify less frequently than White children, the types ofservices that increase the
likelihood of reunification for children and families ofcolor are not well understood. In general
FRS research does not clearly support the effectiveness of these interventions, either on family
reunification or on family functioning (Littell & Schuerman, 1995; Lewandowski & Pierce, 2002).
Furthermore, the actual interventions that are associated with "family reunification services" are
often not clearly defined.
Research does suggest that families who fail to reunify often have serious social and
economic risk factors that do not appear to be adequately addressed by current family reunification
services. For instance, families with a decreased likelihood of reunification are those that have
income problems (Barth et al., 1987; Courtney, 1994; Jones, 1998); parents with less than a high
school education; unemployed parents (Westat, Inc., 2001c); and children with health problems or
disabilities (Courtney, 1994; McMurty & Lie, 1992; Wells & Guo, 1999). Yet these characteristics
also describe the general child welfare population. Clearly more research is needed to discern the
types of FRS interventions that are effective with these high-risk groups, as well as the possible
differential outcomes for children and families ofcolor.
Specific Family Reunification Services: Parenting Education, Substance Abuse, Domestic
Violence, and Family to Family/Social Network Interventions
In addition to research on the effectiveness of family reunification services as a whole, some
studies have focused on the effects of specific services aimed at improving particular problems such
as parenting deficits, substance abuse, domestic violence and poor social networks. As with family
reunification services, studies on specific treatments to reunify parents with their children are
limited and often have methodological problems such as small sample sizes, lack of comparison
groups, lack of treatment consistency and lack of adequate long-term follow-up.
Parenting Education Services
Programs aimed at teaching parents new parenting skills are frequently included in family
reunification case plans. For instance, Phase 2 case record review data indicated that in Santa Clara
County, 70.2 percent of mothers were recommended parenting education services at the
jurisdictional/dispositional hearing (Hines et al., 2002). In general parenting education services are
didactic in nature and focus on increasing participants' knowledge of effective parenting skills.
Research on the effectiveness of parenting education programs for families in the child welfare
system is lacking. Most parenting education evaluations have examined the impact of these
programs on parents whose children have specific behavioral problems-not for families who have
had children removed as a result of maltreatment. Research on parenting education for nonmaltreating parents has yielded some promising results (Dore & Lee, 1999). However, only a
handful of studies have evaluated parenting education services for child welfare parents. Results
from these studies suggest that child welfare parents may experience increased benefits from
parenting education programs that combine a didactic approach with another parenting education
27
service (Brunk, Henggeler, & Whelan, 1987; Carlo, 1993; Dore & Lee, 1999; Wolfe, Edwards,
Manion, & Koverola, 1988).
More rigorous studies of traditional didactic parenting education services have been
performed with samples involving parents whose children have behavioral problems. Most of these
studies suggest that low risk parents may experience significant short term benefits from parenting
education services (Petsch, Schultz, & Wahler, 1999; Nicholson, Brenner, & Fox, 1999; Patterson,
Chamberlain, & Reid, 1982). However, the research also suggests that high-risk parents typically
experience little benefit from parenting education services (Dore & Lee, 1999).
Parenting Education Services:
Impact on Children and Families of Color
Research on the impact ofparenting education services for children and families of color is
lacking. Studies fail to mention the possible impact ofcultural differences in parenting behaviors on
the success or failure ofparenting education services (Forehand & Kotchick, 1996). Parenting
education appears to be most effective for lower-risk families, however, numerous stressors such as
poverty, substance abuse, domestic violence, low education levels, and community violence often
impact parents involved in the child welfare system. Dore and Lee (1999) note these risk factors
may make cognitive behavioral and didactic approaches to parenting education particularly unsuited
to these populations:
For those who are highly stressed by poverty and its accompanying pressures, even
sustaining a commitment to the training process can be overwhelming. Those parents who
lack the cognitive foundation needed for more didactic and learning based approaches seem
unable to benefit fully from behaviorally based training models (Dore & Lee, 1999, p. 322).
More promising approaches to parenting education for maltreating parents appear to involve a
multifaceted approach that includes not only instructional training in parenting behaviors, but also
experiential training and interventions that address the emotional and psychosocial stressors
affecting these high-risk parents (Carlo, 1993; Brunk et al., 1987; Dore & Lee, 1999; Wolfe et al.,
1988).
Indeed, there is some limited evidence to suggest that parents in the child welfare system
may experience increased benefits from a parenting education program than combines didactic
parenting education with another parenting related service such as behavior training, experiential
training or multistystemic therapy (Brunk et al., 1987; Carlo, 1993; Wolfe et al., 1988). However,
more research is needed in this area. For instance, it is unclear which parenting related services are
most effective when paired with didactic training and it is also unclear how well parents retain
information learned. Long-term follow-up studies that track the retention of information in
parenting education services for child welfare parents, as well as child and family outcomes, are
needed.
Additionally, few studies have addressed the role of culture and ethnicity in the effectiveness
ofparenting education services/or parents in the child welfare system. Forehand and Kotchick
(1996) note, " ... different cultures have different histories that shape cultural theories about
parenting. Without an awareness of the values ofa particular culture, attempts to implement
parenting programs with its members may fail" (Forehand & Kotchick, 1996, p.197). As such,
28
research needs to focus on how parenting education outcomes may differ for different cultural and
ethnic groups and what types ofparenting education approaches might be most successful/or
diverse families.
Substance Abuse Treatment
Parental substance use is common among the child welfare population. In 1998, the United
States General Accounting Office [U.S. GAO] conducted a study of two state CWS systems in
California and Illinois, and reported that 65 percent of open foster care cases in California and 74
percent in Illinois, involved a family where one or both parents had a substance abuse problem for
which treatment was required. Research indicates that cases involving parental substance abuse
reunify at a significantly lower rate than do cases not involving substance abuse (Smith, 2003), and
in general, children from substance abusing households experience poorer outcomes than children
from non-substance abusing households (U.S. GAO, 1997).
As a result of the widespread su~stance use among child w~lfare parents and the poor
outcomes for cases involving parental substance abuse, many family reunification case plans
include requirements for some form of substance abuse treatment. These interventions can include
various components such as drug testing, 12-step groups, outpatient treatment or inpatient
treatment. For instance, Phase 2 case record review results indicated that 40.9 percent of mothers in
Santa Clara County are ordered random drug or alcohol testing, 40.0 percent are ordered substance
abuse or alcohol treatment and 26.8 percent are ordered 12-step groups (Hines et al., 2002).
Unfortunately few studies have evaluated the impact of substance abuse services for child welfare
parents (Gregoire & Schultz, 2001). More research efforts have focused on the effectiveness of
these services for pregnant and parenting women who may not necessarily be involved in the CWS.
In one of the few studies that has evaluated substance abuse treatment among child welfare
parents, Gregoire and Schultz (2001) analyzed the treatment and placement outcomes for 167 child
welfare parents in Lancaster County Pennsylvania. Findings revealed that those participants who
had support from significant others were more likely to complete substance abuse treatment and
also experienced better sobriety outcomes. Moreover, parents who were sober at nine months
follow-up were more likely than parents who were not sober to have child custody or to have
maintained parental rights.
Ways in which child welfare workers make decisions about cases involving substanceaffected parents also impacts the likelihood of reunification. Karoll and Poertner (2002) identified
certain factors considered to be associated with readiness for reunification in their qualitative study
of judge's, caseworker's and substance abuse counselor's indicators ofreunification. Factors
associated with professional' s indication of readiness for reunification included: 1) parental
motivation, 2) evidence of substance abuse recovery, 3) evidence of competence and reliability, 4)
the presence of social support, 5) evidence of improved parenting skills and 6) legal aspects of child
loss and prospects for reunification. However there was considerable variability among respondents
in how many clean drug tests are necessary for reunification. Although the median response was 8
months of clean tests, approximately 30 percent felt 6 months was sufficient, while 32 percent felt
that 12 months of clean tests were needed before reunification should occur.
Although research on the effectiveness of substance abuse treatment for child welfare
parents is scant, more research efforts have focused on the effectiveness of these services for
29
pregnant and parenting women who may not necessarily be involved in the CWS. However, since
most child welfare agencies respond to referrals for prenatal drug exposure efforts to offset drug use
among pregnant and parenting women is an important concern for the CWS (Ondersma, Malcoe, &
Simpson, 2001).
Overall, research on the effectiveness of substance abuse services for pregnant and parenting
women suggests these services may be beneficial, however the degree of benefit appears to depend
on the type and intensity of the intervention, as well as the case characteristics. For instance, there is
some evidence that suggests that residential treatment may be more beneficial than outpatient
treatment for women with substance abuse problems. (Haller et al., 1997). Moreover, parenting
women experience better outcomes in substance abuse treatment when they are able to reside in
residential treatment with their children (Clark, 2001; Metsch et al., 2001 ). Studies also indicate that
women are more likely to complete treatment and experience better outcomes when they have the
support of significant others (Gregoire & Schultz, 2001 ), and when family and significant others are
involved in treatment (Carten, 1996). Parenting women have also identified support and
encouragement from non-judgmental staff as particularly important in their ability to successfully
complete treatment (Akin & Gregoire, 1997; Carten, 1996).
Substance Abuse Treatment:
Impact on Children and Families of Color
Research on the effectiveness ofsubstance abuse treatment for parents color in the child
welfare system is lacking. In general, studies do not describe differential outcomes for various
racial/ethnic groups in these treatments and few studies have actually examined the impact of
substance abuse treatment on reunification outcomes for children and families ofcolor in the child
welfare system. However, research does suggest that parenting women involved in substance abuse
treatment are a high-risk group. Most participants of these services are young single women of
color, with low educational levels and income problems (Carten 1996; Clark, 2001; Knight et al.,
2001; Metsch et al., 2001 ). In general, research on the effectiveness of substance abuse treatment
services for these women has yielded promising results; however the degree of treatment impact
appears to depend on the type and intensity of the intervention, as_ well as case characteristics.
Studies suggest that participants in substance abuse treatment services are more likely to maintain
child custody (Gregoire & Schultz, 2001); and maintain sobriety (Clark, 2001; Metsch et al., 2001).
However, women with less than a high school education, criminal justice system involvement and
poor social supports are less likely to complete treatment (Knight et al., 2001 ).
Overall, research suggests that high-risk parenting women benefit from substance abuse
treatment that is comprehensive; allows them to involve their children and other family members or
supports; and is delivered by supportive, nonjudgmental staff. Although research has not
specifically evaluated the role ofethnic diversity in substance abuse treatment services for
parenting women in the child welfare system, the fact that many of the participants of these services
are women ofcolor does suggest that these services may be beneficial for a diverse population.
However, more research is needed to determine the role of culture in treatment, as well as possible
differential outcomes for various ethnic/racial groups.
Domestic Violence Treatment
Child welfare professionals are becoming increasingly aware of the importance of
30
addressing issues of domestic violence for families attempting to reunify with their children. Rates
of domestic violence among families in the CWS vary from 38 percent to 54 percent (Bowen, 2000;
McGuigan & Pratt 2001). Phase 2 results from case record review data indicated that in Santa Clara
County, 45.4 percent of cases involve domestic violence (Hines et al., 2002). Research from the
child welfare field on the effectiveness of domestic violence treatments in reunifying families and
improving outcomes for families is extremely lacking. Instead, child welfare research efforts related
to domestic violence have primarily focused on describing barriers to effective collaboration
between domestic violence and child welfare professionals and attempts to overcome these barriers.
Indeed, practioners and researchers have noted that child maltreatment and spousal abuse
service systems have historically operated in isolation from one another and typically have different
service delivery approaches (Beeman, Hagemeister, & Edleson, 1999; McKay, 1994). Recent
efforts to increase collaboration between the child welfare and domestic violence fields has focused
on cross-training between the two professions, as well as simplified processes for referring families
to child welfare or domestic violence services (Findlater & Kelly, 1999; Whitney & Davis, 1999).
Additionally, the creation of specific child welfare policies for families experiencing domestic
violence and the use of multidisciplinary task forces to address ongoing issues related to
collaboration between fields has also been described as helpful in increasing coordination and
collaboration (Findlater & Kelly, 1999; Whitney & Davis, 1999).
Santa Clara County has implemented efforts to merge child welfare and domestic violence
fields in order to better serve families affected by both child maltreatment and domestic violence.
Santa Clara County is implementing the Greenbook Project, a multidisciplinary effort involving
representatives from child welfare, the domestic violence community, law enforcement and the
juvenile court. The overall goal of the project is to change the way these groups assist families who
experience both domestic violence and child maltreatment. During the first year of implementation,
the Greenbook Project placed domestic violence advocates from Next Door Solutions to Domestic
Violence at the County Department of Social Services. These advocates provide consultation to
child welfare workers on issues of domestic violence at every level of the child welfare system.
Additionally, the Greenbook Project also visited the Miami-Dade County Dependency Court
Intervention Program for Family Violence to learn key aspects of a domestic violence advocacy
program. Other activities of the Greenbook Project include cross-training between child welfare,
domestic violence, law enforcement and legal personnel, improving cultural competency,
developing guiding principles, changing agency policy and practice, and integrated parenting
programs (Lightbourne, 2002).
The collaboration of child welfare and domestic violence fields to better meet the needs of
families affected by both issues is promising. However, in general these efforts are not focused on
specific treatments for the batterer. Indeed, child welfare research has not addressed the
effectiveness of domestic violence services for child welfare parents. Yet studies from related fields
have evaluated effectiveness of domestic violence treatment for batterers who may not necessarily
be involved in the CWS. Domestic violence treatment for batters typically involves a group
treatment that includes psycho-educational and cognitive behavioral techniques designed to change
faulty beliefs, reduce anger, improve communication and increase non-violent coping skills
(Morrel, Elliot, Murphy, & Taft, 2003). Less common are psychodynamic or support groups.
Findings from studies on the effectiveness of these treatments are mixed and much of the research is
limited due to small sample sizes, lack of comparison groups and findings that rely on self-reports.
Some studies have reported that targeted domestic violence services are effective in reducing
31
violent beh~viors (Dutton, 1986; Palmer, Brown, & Barrera, 1992), yet other studies report no
benefits from these treatments (Dunford, 2000). As with many other treatments, domestic violence
treatment tends to be most effective for participants with relatively few other risk factors (Morrel et
al., 2003; Taft, Murphy, Elliot, & Keaser, 2001).
Domestic Violence Treatment: Impact on Children and Families of Color
Research on the impact of cross-system collaboration and specific domestic violence
treatment for children and families ofcolor is lacking. Although child welfare research describing
cross-system collaboration between domestic violence professionals and child welfare professionals
is promising, these studies do not address the role ofcultural issues on domestic violence, or the
treatment of batterers and battered women. Similarly, research from related fields describing
outcomes of didactic domestic violence treatment also rarely consider the role of culture or
ethnicity.
Research suggests that the domestic violence field and the child welfare field have
historically had different approaches to treatment that have hindered collaborative efforts (Beeman
et al., 1999). Recently, more concerted efforts to link the two fields have occurred. Typically these
efforts involve the use of multidisciplinary task forces to address issues related to collaboration
(Findlater & Kelly, 1999; Lighbourne, 2002); cross-training (Findlater & Kelly, 1999; Lighbourne,
2002; Whitney & Davis, 1999); simplified processes to refer battered women to child maltreatment
prevention programs and to refer battered child welfare mothers to domestic violence services
(Findlater & Kelly, 1999; Whitney & Davis, 1999); and, the creation ofspecific child welfare
policies stating procedures for families experiencing both child maltreatment and domestic violence
(Findlater & Kelly, 1999; Whitney & Davis, 1999). Although these efforts have great potential to
impact children and families of color in the child welfare system, research has not yet addressed this
topic.
The collaborative efforts between child welfare and domestic violence tend to focus on
services and assistance to women and children, with less of an emphasis on treatments for the
perpetrators ofdomestic violence. As with other didactic interventions, such as parenting education,
there is evidence to suggest that domestic violence treatment for batterers is more effective with
low-risk groups than with high-risk groups (Morrel et al., 2003). However, research also suggests
that most men who are participating in domestic violence treatment are a high risk group who often
experience multiple stressors (Brown, Werk, Caplan, & Seraganian, 1999). Moreover, the presence
of certain risk factors appears to reduce the likelihood of treatment completion (Taft et al., 2001).
Additionally, one study found that African American men were most likely to drop out of domestic
violence treatment, even after other factors were statistically controlled, suggesting that these
treatments may not be as effective for African Americans (Taft et al., 2001 ). However, in general,
more research is needed to discern the impact of domestic violence treatments for various
racial/ethnic groups in the child welfare system.
Social Network Interventions/Family to Family Initiative
Some child welfare services have focused on the role of social network interventions on
improving outcomes for children and families. Social network interventions are designed to enhance
both formal and informal social networks, including increased networking among family members,
32
friends, neighbors, as well as more structured networking involving support groups (Gaudin,
Wodarski, Arkinson, & Avery, 1990). These interventions are based on research indicating that
people living in neighborhoods in which there are high child maltreatment rates tend to describe
these communities as lacking in formal and informal support systems (Garbarino & Kostelny,
1992). Additionally, DePanfilis ( 1996) has noted that in addition to a lack of social support systems,
neglectful families may be particularly distrustful of any support systems that are available.
In an effort to address these issues, social network interventions, such as the Family to
Family Initiative have been implemented. Although the goals of Family to Family are extensive and
go beyond just social network interventions, one of the initiative's main emphases is on
"strengthening the network of families available to care for abused and neglected children in their
own communities" and "building partnerships with at-risk neighborhoods toward that end" (Annie
E. Casey Foundation, 2003b, p. 1). Indeed, many interventions provided through Family to Family
are designed to strengthen neighborhood networks for birth and foster families in the child welfare
system. These strengthened networks are considered to have the potential to assist families in
learning new skills and becoming more self-sufficient. The tactics used to increase social networks
for families, and also increase collaboration between the child welfare system and local
communities are varied. In general there is an emphasis on identifying families' naturally occurring
support network~ and involving these networks in services. Interventions can be focused on
identifying and utilizing "natural helpers" in the communities and neighborhoods where birth and
foster families reside. Natural helpers are considered to understand the culture of a particular
community better than professional helpers. Efforts are also aimed at increasing overall
collaboration between child welfare agencies and the communities in which child welfare clients
reside.
Outcome studies on the effectiveness of Family to Family interventions are not available.
However, Gaudin et al. (1990) evaluated the effectiveness of the Social Network Intervention
Project (SNIP) for 52 neglectful families in the child welfare system. The intervention, which lasted
approximately 10 months, included first an assessment of existing formal and informal support
networks in the lives of the families, and then a series of interventions designed to enhance these
networks. Direct interventions were made with informal supports, including family members,
friends or neighbors. Efforts were also made to identify and utilize informal helpers existing in the
families' neighborhoods. Formal support networks were also encouraged through the use of support
groups, parent aides and social skills training. Post-test results from the evaluation indicated that
compared to a comparison group who received normal child welfare interventions, the families
receiving the social network services fared better on a number of outcomes assessed by their
caseworkers. Intervention parents were rated as having improved parenting skills and attitudes, and
they experienced significant increases in the size of their formal and informal support networks. At
the end of the intervention, 59 percent of the families had their child welfare cases closed, while
23 .5 percent of the comparison group has their cases closed.
Although the use of social network interventions may be a promising child welfare practice,
Gaudin et al. (1990) notes that there are organizational and professional barriers to the
implementation of such interventions. For instance, results from the SNIP evaluation indicated that
it was difficult for social workers implementing the project to maintain a focus on social network
interventions within the context of a public child welfare agency. The predominate model of
services in the agency tended to focus on case management services, and barriers such as high
33
caseloads, heavy paperwork, frequent staff turnover, and lack of collaboration between human
service agencies all contributed to difficulties in implementing the social network intervention.
Family to Family/Social Network Interventions:
Impact on Children and Families of Color
Researchers have not evaluated the impact ofsocial network interventions, such as the
Family to Family Initiative on children and families ofcolor. Given that children andfamilies of
color involved in the child welfare system are likely to reside in impoverished communities, the use
ofsocial network interventions with this population is ofparticular interest. Research exploring the
assets and informal supports imbedded in impoverished communities is limited. It certainly appears
likely that strengthening support networks for vulnerable families would be beneficial, however
research on the mechanisms through which to increase support networks for families is lacking.
Indeed, the exact interventions associated with utilizing informal helpers, strengthening
neighborhood networks and increasing overall collaboration between child welfare agencies and
the communities in which child welfare clients reside are not clearly defined. As Gaudin et al.
(1990) has suggested, there may be significant barriers to effective implementation of these
interventions within the crisis-driven, resource strapped child welfare field. More research is
needed to first discern the types of interventions associated with social network practices, as well as
the ways in which outcomes from these services may differ for various racial/ethnic groups.
Mental Health Services for Children in Out-of-Home Care
Children in the child welfare system are at an increased risk for a variety of psychological
and behavioral problems (Stevenson, 1999). Studies have documented the long-term negative
consequences of both child maltreatment and lengthy stays in out-of-home care (Courtney, Piliavin,
Grogan-Kaylor, & Nesmith, 200 I; Stevenson, 1999). In order to ameliorate these negative
consequences, many children in out-of-home care are referred to mental health services. Research
suggests that mental health services are the most commonly recommended treatment for children in
the child welfare system (Kinard, 2002). Studies have also found that children in foster care are
more likely to have a mental health diagnosis and to use mental health services than comparison
groups. For instance, children in foster care are more likely than children receiving public assistance
or children receiving SSI to have a mental health diagnosis (Harman, 200 I; dosReis, 2001 );
children in foster care tend to have more mental health hospitalizations; and use more mental health
services overall than children receiving public assistance (Reiff, 2001 ).
Yet despite the relatively high rate of mental health service usage among children in the
CWS, the nature and impact of these services are not well understood. Indeed mental health services
can include a broad range of interventions, including support groups, individual counseling, family
counseling, residential services or inpatient services. One model of services, Treatment Foster Care
(TFC), has been the topic of child welfare research. TFC, also referred to as therapeutic foster care
is an intensive, family centered intervention that seeks to place youth with behavioral and emotional
disorders in community and family settings. Children are typically placed with foster parents and
the placement is intended to be both therapeutic and family-centered. Intensive services are
delivered to parents, alternate caregivers and the child. Some studies have suggested that TFC can
be an effective transitional placement for children moving from restrictive settings into less
restrictive placements (Chamberlain, 1994, as cited in Farmer, Wagner, Burns, & Richards 2003;
Clarke, Hawkins, Murphy, & Sheeber, 1993, as cited in Farmer et al., 2003). However, not all
34
studies have found TFC to be effective. For instance, Fanner et al. (2003) used administrative data
to analyze the impact ofTFC on foster youth with emotional or behavioral disorders. Findings
indicated that use ofTFC did not result in reduced use of group homes.
In addition to TFC, some child welfare interventions for children focus on a "wraparound"
process that is targeted to high risk children who may have emotional or behavioral problems and
who are currently placed in a level 12 or higher group home, or are at risk for entry into a more
restrictive placement (California Department of Social Services [COSS], 2001). Wraparound
services are described as a "family-centered, strength-based, needs-driven planning process for
creating individualized services and supports for ·children and their families" (COSS, 2001, p. I).
Wraparound interventions are collaborative in nature, often involving multiple social service
systems in the planning and delivery of a wide range of services. Few studies have systematically
investigated the efficacy of wraparound services for children and families in the child welfare
system. However, some research suggests that children in wraparound are less likely to be removed
from their community and may experience a reduction in behavioral problems (Bruns, Burchard, &
Yoe, 1995, as cited in Toffalo, 2000; Clark, Lee, Prange, & McDonald, I 996, as cited in Toffalo,
2000; VanDenBerg, 1993, as cited in Toffalo, 2000). Yet other research has reported no significant
benefit from wraparound services (Clarke, Schafer, Burchard, & Welkowitz, 1992, as cited in
Toffalo, 2000).
Mental Health Services:
Impact on Children of Color in Out-of-Home Care
Although research is lacking in identifying effective mental health practices for children in
foster care, studies have demonstrated that children ofcolor are less likely to be recommended for
these services and to use these services. In Santa Clara County, Phase 2 case record review results
indicated that 36.2 percent of children were recommended individual counseling and 19 .6 percent
were recommended other counseling services. Differences between racial/ethnic groups in
recommended counseling services were also found. For instance, 18 percent of Black children were
recommended individual counseling, 46.3 percent of White children, 31.9 percent of Latino
children, and 35.5 percent of Asian American/Pacific Islander children were recommended
individual counseling, suggesting that children of color, and in particular African American children
have a decreased likelihood of being recommended individual counseling (Hines et al., 2002).
Differences in service recommendation and provision were also noted by Garland and
Besinger (1997) in their analysis of racial/ethnic differences in court referred pathways to mental
health services for children in foster care. Results suggested a trend for White children to be more
likely than African American or Hispanic children to be court ordered counseling/psychotherapy.
White children were also significantly more likely than African American or Hispanic children to
receive counseling during their first eight months in out-of-home care. And although not significant,
White children had higher rates of service use in general than did African American or Hispanic
children. Racial/ethnic differences in mental health service recommendations and service usage
have been found to remain even after the effects of other demographic factors are controlled
(Garland, Hough, Landsverk, McCabe, Yeh, Ganger et al., 2000).
Research on the effectiveness of mental health practices for children in out-of-home care is
lacking, and even less is known about the impact of these services for children of color.
Comprehensive treatments for high-risk children, such as treatment foster care and wraparound
35
services have not been adequately researched and have in general not consistently demonstrated
positive effects. Studies on other mental health interventions for children ofcolor in the child
welfare system are lacking. However, research does suggest that children of color are less likely
than. White children to be referred to mental health services (Garland & Besinger, 1997; Hines et al.,
2002). The reduced likelihood for children ofcolor to be referred to mental health services is of
particular concern because children of color experience long stays in the child welfare system and
are at an increased risk/or a variety ofpoor outcomes (Courtney et al., 1996). As such, more
efforts should be aimed at ensuring that children of color are referred to and receive appropriate
mental health support, and more research is needed to evaluate the impact ofthese services on
outcomes for children ofcolor.
Foster Parent and Kinship Support Services
Although the majority of research on child welfare practices has focused on interventions
with birth families and children, some research has emphasized the im.portance of services and
training for alternative caregivers, including foster parents and kin caregivers (Cuddeback & Orme,
2002; Gordon, McKinley, Satterfield, & Curtis, 2003). Some studies have suggested that kin
caregivers may receive fewer services than non- kin caregivers (Gordon et al., 2003; Scannapieco,
Hegar, & McAlpine, 1997, as cited in Cuddeback, 2002). However, other research has reported no
differences between services delivered to kin and non-kin caregivers (Cuddeback & Orme, 2002).
Research suggests that the use of kin placements is common in California and in Santa Clara
County. In 2002, 36.4 percent of children in supervised foster care in California were in a kin
placement and in Santa Clara, 40.l percent were placed with kin (Needell et al., 2003).
Additionally, children and families of color are especially likely to be placed with kin. For instance,
point in time reports indicate that in California, 40.5 percent of Black children in supervised foster
care are placed with kin; 37.0 percent ofNative American; 35.8 percent ofHispanic, 29.5 percent of
White and 29.4 percent of Asian children are placed with kin. In Santa Clara County, 66.7 percent
of Native American children are placed with kin; 41.9 percent of Hispanic; 41.4 percent of Black;
35.6 percent of Asian; and 32.5 percent of White children (Needell et al., 2003).
The prevalence of kin placements has prompted some researchers to explore the service and
support needs of kin caregivers. Gordon et al. (2003) noted the unique needs of kin caregivers in
their analysis of focus groups with 37 kin caregivers. Findings revealed that unlike non-kin
caregivers, many kin caregivers did not expect to be caring for the children in their care; they had to
adjust their family roles and many described this adjustment as sudden and significant. Kin
caregivers also described having to adjust to the involvement of the child welfare agency in their
lives; many reported feeling unappreciated by the child welfare agency and excluded from the
decision-making process for their kin children. They also described a great need for additional
support services as they took on the unexpected responsibility of caring for their kin children.
Other studies have noted that non-kin caregivers also experience unique challenges. For
instance, although non-kin caregivers are expected to support reunification efforts, many may have
conflictual feelings about supporting contact between the children in their care and the birth
families (Sanchirico & Jalonka, 2000). As a result, some child welfare practices have focused on
providing specialized training and support to non-kin caregivers to assist in the process of
reunification. For instance, Sanchirico and Jalonka (2000) surveyed 650 non-kin caregivers in New
York State and noted that only 19 percent received both specialized training to be a foster parent
36
and support and guidance from the agency. Those who received both training and support also
performed the most activities related to encouraging contact between children and their birth
families. Foster parents who received either training or support but not both also exhibited some
tendency to encourage parent child contact, however those who received both training and support
were significantly more likely to support reunification efforts. These results suggest that specialized
services delivered to foster parents can have a beneficial impact on reunification.
Additionally, other research has focused on evaluating more concerted efforts to involve
foster families in the reunification process. For instance, Landy and Munro (1998) evaluated the
impact of the Shared Parenting program. The Shared Parenting program encourages foster families
to work as a team with birth parents in order to improve the parenting skills of the parents. Foster
families were encouraged to provide support and guidance to birth parents through on-going contact
and involvement. An improved relationship between foster and birth families was expected to
increase birth parents motivation and interest in learning new skills, thereby increasing
reunification. Thirteen families participated in the evaluation. Results revealed difficulties in
retaining parents in the program. Over half of the participants did not complete the program, either
because of lack of compliance or because workers discharged families due to changes in status that
decreased the possibility of reunification. Outcomes for parents who remained in the program
revealed that those with higher socioeconomic status, families who moved less, families with no
health problems and families with less risk factors were more likely to reunify.
Although research suggests that kin and non-kin caregivers have different service needs,
studies have demonstrated that overall, both kin and non-kin caregivers have numerous unmet
service needs. For instance, Cuddeback and Orme's (2002) conducted a national evaluation of the
service experiences and needs of kin and non-kin caregivers, and found that both groups reported
numerous unmet service needs. Most alternate caregivers expressed a need for day care, respite
care, liability insurance, heath care, and transportation. Additionally, the majority of caregivers in
both groups reported feeling unprepared to care for their foster children. The authors concluded "the
levels of training and services received were inadequate for both groups of caregivers" (Cuddeback
& Orme, 2002, p. 900).
Foster and Kinship Support Services:
Impact on Children and Families of Color
The impact offoster and kinship support services ofchildren and families ofcolor has not been
adequately researched. However, research clearly indicates that kin and non-kin caregivers are in
need of increased training and supportive services. Kin caregivers may need increased support as
they take on the unexpected responsibility and role of a caregiver (Gordon et al., 2003), while nonkin caregivers may need additional support and training to facilitate the reunification process.
(Sanchirico & Jalonka, 2000). In general, both kin and non-kin caregivers describe the need for
increased services, such as day care, respite care, liability insurance, heath care, transportation,
specialized training and increased participation in the child welfare decision making process
(Cuddeback & Orme, 2002; Gordon et al., 2003; Sanchirico & Jalonka, 2000).
Research on the impact offoster and kinship support services on child welfare outcomes is
lacking. However, there is some evidence to suggest that interventions with foster parents are
associated with better reunification outcomes (Sanchirico & Jalonka, 2000). Yet more research is
clearly needed to discern the impact of services to both kin and non-kin caregiver, especially for
37
different racial/ethnic groups.
Permanency Planning Practices
The federal Adoption and Safe Families Act of 1997 shortened reunification time frames
and mandated the use of concurrent planning-a process in which both a reunification plan and an
alternative permanent placement plan are created. For children who are not reunified, the permanent
placement plan with its goal of finding a suitable alternative permanent placement for the child
becomes a major focus of services. As such, permanency planning and placement practices often
impact children at the ending stages of the CWS.
Very few studies have investigated the types of practices that are used at the ending stages in
the system and how these practices impact children of color. However, the impact of child welfare
practices on children who are not reunified is particularly important for children and families of
color. The U.S. Department of Health and Human Services (2002) reports that approximately 23
percent of children in foster care have had parental rights terminated and children of color are
disproportionately represented in this group. Nationally, 43 percent of legally free children are
African American, 34 percent are White, 13 percent are Hispanic, 2 percent are Native American
and 1 percent are Asian Pacific Islander (U.S. DHHS, 2002). In California, 68.2 percent of Black
children in foster care are in permanent placement services, 48.4 percent of White children; 41.5
percent of Hispanic children; 32.5 percent of Asian children, and 53.9 percent ofNative American
children are in permanent placement services (Needell et al., 2003). Santa Clara County has similar
racial/ethnic rates of permanent placement clients; 64.4 percent of Black children are in permanent
placement services; 44.7 percent of White children; 41.3 percent of Hispanic children; 30.6 percent
of Asian children and 43.3 percent ofNative American children are in permanent placement
services (Needell et al., 2003).
Federal policy mandates that these children have safe and permanent alternative homes,
however, in reality many wait months and years for a permanent placement and some will reside in
long term foster care until they emancipate from the system at age 18. In Kemp and Bodonyi's
(2002) analysis of legally free children in the state of Washington, the median length of stay in outof-home care for these children was 50. 7 months and demographic factors were significantly related
to permanency outcomes. For instance, older children, boys, African Americans, and Asian Pacific
Islanders were all less likely to achieve a permanent placement than others groups. Interestingly,
Latino children were significantly more like than White children to achieve a permanent placement.
The interaction of demographic characteristics was also significant in predicting permanency. 0 Ider
African American boys had an even higher chance of not achieving a permanent placement. Other
research corroborates these findings. For instance, African American children have been found to be
less likely to be adopted than White or Latino children and African Americans who are freed for
adoption wait significantly longer periods of time for a permanent placement than do children from
other racial/ethnic groups (Barth, Courtney & Berry, 1994, as cited in Kemp & Bodonyi, 2002;
Kapp, McDonald, & Diamond, 2001).
The federal Multi Ethnic Placement Act (MEPA) of 1994 and the Interethnic Adoption
Provisions (IEP) of 1996 were enacted to increase the number of children who were adopted by
preventing discrimination in the placement of children on the basis of ethnicity and allowing for
increased use oftransracial adoption. The impetus behind MEPA-IEP was a belief that children of
color, especially African Americans were languishing in care, while workers searched for same race
38
foster or adoptive homes. MEPA-IEP also called on child welfare agencies to increase efforts to
recruit foster and adoptive parents who reflect the ethnic and cultural backgrounds of their children
in care. Although MEPA-IEP may have the potential to reduce the time legally free children spend
in substitute care, rates of adoption for children of color in the CWS do not appear to have improved
since its enactment. Indeed, Courtney (1997) suggests that transracial adoption policies and
practices may not actually have a large impact on children of color. Because many legally free
children of color in out-of-home care are older and may have disabilities or health problems, many
potential adoptive parents may choose not to adopt these children. Thus, even though child welfare
agencies may seek transracial adoptions, the actual pool of available adoptive parents may be too
limited to have an impact on children of color waiting for a permanent placement (Courtney, 1997).
In an effort to understand the factors that contribute to adoptive parents willingness to adopt
Black children, Brooks and James (2003) surveyed 541 participants of the California Long-Range
Adoption Study, a longitudinal study of adopters in California. Results indicated significant
differences between participants who were willing to adopt Black children and those who were not. .
Of the sample 36 percent were willing to adopt a Black child. In general, parents who were willing
to adopt Black children were also more willing to adopt children of varying ages and special needs
children; they were more likely to be motivated to adopt due to religious or humanitarian reasons
and many eventually did adopt Black children from foster care. Interestingly, approximately 50
percent of parents who indicated they were unwilling to adopt a Black child also reported that
adoption workers or agencies discouraged them from adopting Black children, while only 11
percent of those parents who were willing to adopt Black children were encouraged by their worker
or agency to do so.
Additionally, parents who adopt transracially may experience unique challenges and have
special service needs. In an effort to identify the special needs of transracial adoptive parents Vidal
de Haymes and Simon (2003) conducted interviews with 20 youth and their adoptive parents. Youth
in the sample were either African American (70%), Latino (15%) or biracial (15%), while 95
percent of adoptive parents were White and 5 percent were Latino. Findings revealed that parents
often felt that others misjudged their intentions for adopting a child of color and that their families
"deserved pity" (Vidal de Haynes & Simon, 2003, p. 257). Many families reported that child
welfare workers, especially workers of color did not adequately support their decision to adopt a
child of color. Parents felt that workers viewed their desire to adopt transracially as curious, and
many felt that they received only minimal assistance in their efforts to adopt a child of color.
Parents identified numerous needed pre and post adoptive supports and resources. The most
frequently noted needed services were formal linkages to cultural activities, products, or services
that could educate parents on cultural issues and also provide a link for the child to their own
cultural group. Other needed support services included establishing a network of parents who
adopted transracially, educating workers to support transracial adoption, providing parenting classes
to address issues of cross cultural parenting, and developing a mechanism to help White parents
address their own possible racism, as well as racism from their family and friends.
In an effort to understand how worker perceptions may impact the likelihood of transracial
adoption for children of color, Carter-Black (2002) conducted an ethnographic study often Black
child welfare workers. Although the sample size makes generalizing the results problematic,
findings revealed that workers felt Black children need to develop an ethnic identity and an
understanding of their cultural heritage and that the CWS in general does not meet the needs of
Black clients and poor clients. The recruitment, support and maintenance of Black adoptive and
39
foster parents by the agency were seen as lacking. Many also felt that MEPA-IEP could actually be
harmful to Black children because it may result in agencies being less culturally aware or sensitive
to diverse populations.
A very (2000) also explored the impact of worker perceptions on permanent placement
practices in her study on the "hardest to place children" in New York State. Eighty case files were
examined and children's workers completed a questionnaire on their perceptions and practices
related to these children. Demographics of the sample are consistent with other studies on children
who have lengthy stays in out-of-home care. Seventy three percent of the sample were male, 60
percent were Black, 30 percent White and 10 percent of another ethnicity, many had substantial
disabilities; approximately 80 percent had siblings in the CWS and they spent approximately 11. 79
years in substitute care. Responses to survey questions revealed that 41.4 percent of workers
believed that children in the sample were ultimately not adoptable. These beliefs appeared to result
in reduced efforts on behalf of the workers to recruit potential adoptive parents. Worker attitudes
toward appropriate placements revealed that 44 percent felt placement with gay or lesbian parents
was not appropriate, and 42 percent felt transracial adoption was not appropriate. When asked to
identify resources that would be helpful in placing the children, workers most often identified
respite services, intensive adoptive parent training, intensive post-placement services, more
effective recruitment techniques and higher adoption subsidies. These findings suggest that workers
negative perceptions of hard to place children appear to result in reduced efforts to recruit adoptive
parents.
Permanency Planning Practices:
Im pact on Children and Families of Color
In general, child welfare research has focused relatively little attention on practices at the
later stages ofthe child welfare system. Although child welfare services have the goal of ensuring a
permanent placement for children, research indicates that many children who have had parental
rights terminated may wait years in out-of-home care before a permanent placement is found (Kemp
& Bodonyi, 2002), and many children remain in out-of-home care until they emancipate at age 18
(Courtney et al., 2001 ). Yet the impact of permanency planning practices on outcomes for children
of color is of great importance. Research indicates that African American children are
overrepresented among children who have had parental rights terminated and are waiting/or a
permanent placement. African American and Asian Pacific Islander children both have a reduced
likelihood ofachieving a permanent placement, and older African American boys have an even
lower likelihood ofa permanent placement (Kemp & Bodonyi 2002).
Some researchers have noted that MEPA-IEP and the issue oftransracial adoption may
actually have little impact on children of color in permanent placement services. Because the pool
of adoptive parents willing to adopt legally free children of color (who are often older, and may
have disabilities or health problems) may be relatively small, policies aimed at transracial adoption
may not have their intended impact (Courtney, 1997). Yet other research suggests that the
perceptions and practices of workers may have an important influence on the likelihood of
achieving a permanent placement. Indeed, negative perceptions of hard to place children may result
in reduced efforts to recruit adoptive parents (Avery, 2000). Additionally, parents who have adopted
transracially report that workers often did not adequately support their decision to adopt
transracially (Vidal de Haymes & Simon, 2003), and Black child welfare workers have expressed
some hesitancy regarding MEPA-IEP, believing that it could be harmful to Black children because
40
it may result in agencies being less culturally aware or sensitive to diverse populations (CarterBlack, 2002). However, in general more research is needed to discern the types of permanent
placement practices that impact children and families of color.
Impact of Contextual Factors on Child Welfare Practices
for Children and Families of Color
Contextual factors can impact both the delivery of child welfare practices and the outcomes
of these practices. Although numerous contextual factors exist, this section will focus on three
areas: 1) cultural competency of child welfare workers and practices, 2) child welfare practice in
rural areas, and 3) the role of the court system. The potential impact of these contextual factors on
practices and outcomes for children and families of color will be discussed.
The Role of Cultural Competency in Child Welfare Practice
Child welfare workers often work with children and families from a wide range of cultures
other than their own. Many practioners and researchers have noted that effective child welfare
practices are those that acknowledge and incorporate the importance of culture in the delivery of
services (Miller & Gaston, 2003). Indeed, Miller and Gaston (2003) note that inherent assumptions
within the child welfare system are grounded in Anglo-Saxon values and cultural norms about child
rearing and family. Child welfare legislation and policies often follow European standards of
culture and White, middle class family values are the standard through which ethnically diverse
parents and children are compared. As such, children and families exhibiting alternative values may
be seen as deviant by the system. These conflicts in attitudes regarding acceptable parenting
behavior may contribute to ineffective or harmful child welfare practices (Miller & Gatson, 2003).
In an effort to combat ethnocentrism in the child welfare system, many agencies have placed
increasing importance on ensuring that workers, programs, policies and practices are "culturally
competent." In general, the term cultural competence refers to an ability to recognize and respect
similarities and differences in beliefs, interpersonal styles, values, norms, and behaviors of various
ethnic and cultural groups (Roberts, 1990, as cited in Schriver, 1998).
Although cultural competency is widely viewed as an important aspect of effective child
welfare practice, research on the cultural competency of child welfare systems and the impact of
this competency on children and families is lacking. Some studies describe the problems diverse
clients experience in trying to access culturally and linguistically competent services, other studies
have described ways to improve cultural competency in child welfare agencies, while other research
has addressed mechanisms for child welfare agencies to work more collaboratively with ethnically
diverse communities.
The issue of access to culturally and linguistically appropriate services for children and
families of color is of tremendous importance. Suleiman (2003) notes that for Latino clients in
particular, language barriers can exclude them from necessary services and produce significant
delays in service delivery. Indeed language barriers are implicated as a factor in the underutilization
of services among Hispanic clients (Gomez, Ruiz, & Rumbaut, 1985, as cited in Suleiman, 2003),
and many researchers have discussed the importance of language in accurate assessment and
treat~ent of clients (Suleiman, 2003). Suleiman (2003) further notes that denial of access to
41
services and programs for people with limited English proficiency is in violation of the Civil Rights
Act of 1964; yet federally funded agencies have wide discretion in the degree to which language
needs of clients are addressed.
As such, the responsibility to provide and monitor culturally competent and linguistically
appropriate services has largely fallen on child welfare systems themselves. Some studies have
proposed various models of achieving cultural competence in child welfare agencies. For instance,
Mederos and Woldeguiorguis (2003) discussed the role of child welfare managers in producing
culturally competent child welfare agencies. Recommendations for increasing cultural competence
included first and foremost, the managers' recognition that disproportionality in services and
outcomes for children of color does exist. Indeed, other researchers have noted that a prerequisite to
true cultural competence within an agency is a commitment to cultural competence principles and
practices among top level staff (McPhatter & Ganaway, 2003). Additionally, to produce agency
wide cultural competence, managers must prioritize staff diversity through recruitment and
maintenance of bilingual and bicultural staff. Child welfare managers also need to reach out to
diverse communities to form collaborations with community representatives. This collaboration can
help the child welfare system better understand cultural differences in parenting behaviors, as well
as assist in the creation of policies and practices that are relevant to those communities.
Indeed, other studies have also documented a need for improved collaboration between child
welfare agencies and culturally diverse communities. For instance, Hosley, Gensheimer, and Yang
(2003), described the implementation process of two programs aimed at improving collaborations
between the child welfare system and two Southeast Asian communities in Minnesota. Both
programs sought to improve practices and outcomes for children and parents from these
communities. The collaborations required committed involvement from numerous agencies, as well
as parents and youth from the communities. Language barriers made the use of bilingual and
bicultural staff at collaborative meetings extremely important. Effective collaboration depended on
a number of different factors, including the need for a trusting relationship between all parties, the
incorporation of different cultural styles in the structure of meetings, decision-making and
communication styles, the need to have bilingual and bicultural staff involved in the process and
that active involvement from th~ youth and parents is crucial to successful collaboration.
Cultural Competency:
Impact on Children and Families of Color
Research has not yet investigated the impact ofculturally competent child welfare practices
on children and families of color, however there is evidence to suggest that practices that are not
culturally or linguistically appropriate result in underutilization of services and ineffective
treatments and are in violation of clients civil rights (Gomez et al., 1985, as cited in Suleiman,
2003; Suleiman, 2003). Some researchers have described processes for achieving cultural
competence within child welfare agencies. A wide range of activities aimed at increasing cultural
competence have been identified, including support from top level staff, the prioritization of staff
diversity, and forming collaborations between child welfare and ethnically and culturally diverse
communities. These activities are intended to combat the child welfare system's legacy of using
European standards of culture and White, middle class family values in assessing diverse children
and families (Miller & Gatson, 2003).
42
However achieving cultural competency in child welfare can be a difficult task. For instance
it may be challenging for child welfare agencies to recruit and retain bilingual/bicultural staff and
the process of building collaborations with diverse communities may be time consuming and labor
intensive. Yet the impact of cultural competency on children and families of color is of tremendous
importance. More research is needed to discern how child welfare agencies recruit and retain
bilingual/bicultural staff, as well as how collaborations are built between child welfare and
ethnically diverse communities. Most importantly, research is needed to systematically evaluate the
impact ofcultural competence on practices and outcomes for children and families ofcolor.
Rural Child Welfare Practices
Much of the research literature on the effectiveness of child welfare practices is conducted
in urban settings. Researchers have suggested that urban child welfare research is often not relevant
to rural communities who may face uniquely different challenges than their urban counterparts
(Templeman & Mitchell, 2002). Indeed several studies have suggested that rural communities are
particularly vulnerable to a number of different risk factors. Rural communities in comparison to
urban communities have an elevated rate of poverty, there are fewer employment opportunities in
rural communities, and rural communities have been found to have lower levels of social well-being
than urban communities (Menanteau-Horta, & Yigzaw, 2002).
In addition to community differences, the characteristics of urban and rural child welfare
workers also differ. For instance, Landsman (2002) surveyed child welfare workers in 115 rural and
urban communities in Missouri and found several differences between the two groups. Findings
indicated that agency size played an important role in practice, many of the differences between
rural and urban child welfare practice were also noted between smaller and larger agencies. Rural
and small agencies possessed workers with a greater number of years spent in their current position
compared to urban and large agencies, they also experienced greater autonomy, decision-making,
growth opportunities, and agency supportiveness, as well as more job satisfaction and
organizational commitment. In contrast, urban and large agencies had more workers with a social
work degree, more work overload, degree of specialization and adequacy of resources, they also had
higher community stress and availability of jobs. These findings suggest that child welfare workers
in rural communities are generalist practioners who may be more satisfied with their job than their
urban counterparts, while urban workers tend to have more resources and a higher degree of
specialization.
Other studies have suggested that rural child welfare agencies provide more services and
intervene more often than urban child welfare systems. For instance, research suggests that rural
child welfare agencies have a higher likelihood of offering preventive services to children and
families that do urban agencies (Drake, 1996); they tend to have higher levels of substantiation of
maltreatment (English et al., 2002); and rural child welfare agencies are more likely to intervene in
cases involving drug exposed infants (Ondersma et al., 2001 ).
The differences between urban and rural communities and chil~ welfare agencies have
important implications for child welfare practice. Templeman and Mitchell (2002) have noted that
the application of urban child welfare practices to rural communities is inappropriate. In their
qualitative study of child welfare workers, they report the need for rural specific child welfare
practices. For instance, participants noted that families in rural communities are often
geographically isolated, they often have incomplete knowledge of available resources, lack of
43
access to specialized services, and restricted mobility. Thus, traditional child welfare services that
call on families to travel to and from a variety of specialized interventions are not generally feasible
with this population. However respondents also noted that rural communities possess certain assets
that should be utilized for more effective rural child welfare practice. Most importantly, rural
communities were seen by social workers as having a strong network of informal and at times,
formal supports, including linkages between neighborhoods, and the use of faith based
organizations. More effective child welfare practice for rural communities was described as using
these natural supports to build better collaboration between child welfare agencies and rural
communities. This collaboration was seen as a mechanism through which child welfare could
perform outreach to the community.
Rural Child Welfare Practices:
Impact on Children and Families of Color
The. impact of rural child welfare practices on children andfamilies ofcolor is not
sufficiently addressed in the research literature. In general, rural communities have an increased
rate of poverty, fewer employment opportunities and lower levels of social well-being than do urban
communities (Menanteau-Horta, & Yigzaw, 2002). Differences between child welfare workers in
rural and urban areas have also been found (Landsman, 2002), and there is some evidence to
suggest that rural child welfare agencies intervene at a higher rate than do urban child welfare
agencies (Drake, 1996; English et al., 2002; Ondersma et al., 2001). These differences between
urban and rural communities and child welfare agencies suggest that applying an urban model of
child welfare practice to rural communities is not justified (Templeman & Mitchell, 2002). The
unique needs of these communities and their child welfare populations call for increased research
efforts to identify effective practices for children and families of color in rural areas.
The Role of the Court System in Child Welfare Practice
For children who become dependents of the court, the judicial system may significantly
impact practices and outcomes. In child maltreatment proceedings, judges typically have authority
over decisions about services and certain case outcomes. Indeed, studies have noted that
noncompliance with court orders is a strong predictor of termination of parental rights (Atkinson &
Butler, 1996; Jellinek et al., 1992; Smith, 2003). Other studies report that the courts can have an
important influence over the services to which children are referred (Garland & Besinger, 1997).
As such, the ways in which court systems and child welfare systems interact is an important
contextual issue affecting practices and outcomes. In general there is little research on the ways in
which child welfare and court personnel collaborate and how this collaboration, or lack of
collaboration impacts children and families of color. However the few studies that have addressed
this issue report that the two professions experience numerous barriers to effective collaboration.
For instance, in Russell's (1988) survey of98 caseworkers and 69 attorneys, significant differences
in role perception were found between the two groups. Child welfare workers and attorneys
disagreed about who should perform a number of different tasks. A majority of both groups felt that
they should have primary responsibility for the same tasks, including requesting authorization for a
dependency hearing, deciding on the allegations, explaining and interpreting court proceedings to
parents, deciding if children should testify, entering into agreements with parents regarding case
disposition, and recommending dispositions to the court. Both groups expressed a need to have
attorney and child welfare worker roles clarified and a need for cross training between the two
44
groups. Additionally, ·in open-ended responses to survey questions, both attorney and child welfare
respondents expressed some irritation over their interactions with one another. Attorneys felt that
child welfare workers should "stop trying to be attorneys" (Russell, 1988, p. 212), and child welfare
workers expressed that "we know about what lawyers do, but they don't know about what we do.
Therefore training of lawyers to casework responsibilities would be helpful" (Russell, 1988, p. 212).
In an attempt to address barriers to collaboration between child welfare and the court
system, Johnson and Cahan (1992) reported on the "Children Can't Wait Project," a training
program aimed at cross training for attorneys and social workers, and overall system modifications.
The program convened a series of seminars for child welfare workers and attorneys. Seminars
included cross training and open explorations of one another's expectations. Social workers in the
seminars expressed a need for attorneys and court personnel to increase respect and understanding
of their profession, while court personnel described a need for social workers to provide more
timely and accurate case preparation and to be knowledgeable of legal definitions. Other seminar
activities included creating action plans to expedite court proceedings by improving communication
between attorneys and workers, and modifying court proceedings themselves. Although both
attorneys and social workers had good attendance at the initial seminars of the project, attorneys had
much less involvement in multidisciplinary groups working to implement the project. Outcomes of
these collaborative efforts were not described.
Additionally, Knepper and Barton (1997) explored courtroom dynamics of court personnel
and child welfare professionals during court proceedings. District judges, circuit judges, circuit
clerks, Foster Care Review Board chairs, Court Appointed Special Advocates, and attorneys from
five sites in one state completed surveys. Interviews with court personnel and caseworkers were
also conducted, as well as focus groups with foster parents and caseworkers, court observations, and
case record reviews. Findings revealed that in general, caseworkers felt that judges respected their
opinions and judgments about cases. Yet judges also demonstrated great influence on certain
decisions and actions. When caseworkers adhered to group norms, judges often rewarded them by
not ordering "unrealistic" practices (Knepper & Barton, 1997, p. 298), and providing more
reasonable time frames for mental health evaluations to be completed. Many respondents also noted
that court proceedings were often used to finalize agreements, rather than negotiate; they lasted an
average of 10 minutes with prosecutors and other attorneys doing most of the talking. Survey results
indicated that court personnel generally do not receive training on child maltreatment issues prior to
beginning their job. Approximately 77 percent of district judges and 79 percent of circuit judges
reported that there were no training requirements for attorneys working in child welfare.
The Role of the Court System in Child Welfare Practice:
Impact on Children and Families of Color
Few studies have evaluated the potential impact of the court system on children and families
ofcolor in the child welfare system. However the court system clearly does play a significant role in
case processes and outcomes (Garland & Besinger 1997; Knepper & Barton, 1997). For instance,
Garland and Besinger ( 1997) noted that White children in the child welfare system tended to be
more likely than African American or Hispanic children to be court ordered
counseling/psychotherapy, however no racial/ethnic differences were found in social worker
recommendations in counseling/psychotherapy.
45
Research also indicates that there are numerous barriers to effective collaboration between
court personnel and child welfare professionals, including lack of clarity in their different roles
(Russell, 1988). Studies suggest that collaboration between these two groups might be improved
through cross training and clarification of roles (Cahan, 1992; Russell, 1988). More research is
needed to discern how the court system impacts child welfare practices as well as outcomes for
children and families of color.
Summary of Key Findings and Implications
This section reviewed research on evidence-based child welfare practices and their
application to children and families of color, as well as the impact of contextual factors on child
welfare practices for children and families of color. Overall, research on the effectiveness of child
welfare practices across the various choice points in the CWS is somewhat limited and
unfortunately many studies fail to assess the possible differential impact of child welfare practices
on diverse ethnic groups. Nevertheless, certain practices evaluated in the literature are emerging as
promising practices for children and families of color and there is a need to continue to promote
evidence-based practice in child welfare.
Following the pathway that children may take through the child welfare system, research on
child welfare practices at the beginning choice points. including screening, investigation and
provision of voluntary services reveals mixed results on the effectiveness ofthese practices.
Research does indicate that cases involving children of color are investigated and substantiated at a
higher rate than cases involving White children (Ards, et al., 2003; Drake, 1996; Eckenrode et al.,
1988; Wells, et al., 1995). Children of color are also more likely than White children to be placed in
out-of-home care (Needell et al., 2003; U.S. DHHS, 2001).
Studies also demonstrate that there is little overall consistency in the decision-making
practices of individual child welfare workers (Rossi et al., 1999; Schuerman et al., 1999). Practices
designed to improve decision-making including the use of structured risk assessment tools, or
multidisciplinary teams/family group conferencing may have the potential to improve decisionmaking practices, yet research has not yet evaluated the effectiveness of these interventions with
ethnically diverse populations.
Perhaps the most promising beginning choice point practice to emerge from the literature is
the use of home-based prevention services. Research does demonstrate that home-based prevention
efforts are effective in preventing maltreatment and further involvement in the CWS, as well as
improving health related behaviors and outcomes for mothers and young children (Kitzman et al.,
1997; Olds et al., 1997).
Research on continuing care and ending choice point practices provide limited evidence
that these practices are successful in reunifying families or ensuring a permanent placement for
children. However, research does consistently demonstrate that Black children are reunified at a
lower rate than are White children (Courtney, 1994; Jones, 1998; McMurty & Lie, 1992; Wells &
Guo, 1999; Westat, Inc., 2001 c). Evaluations of traditional family reunification services. and
specific treatments, such as parenting education, substance abuse, domestic violence or Family to
Family/social network interventions appear to suggest that these services have limited overall
success, and tend to be most successful with lower-risk parents (Dore & Lee, 1999; Morre/ et al.,
46
2003). However, research does suggest that substance abuse services for parenting women in
particular, are most effective when they are residential, fairly intensive and allow mothers to live
with their children while in treatment (Clark, 2001; Haller et al., 1997; Metsch et al., 2001).
Research on mental health se-rvices for children in out-of-home care suggests that children
ofcolor are less likely to be recommended for and to use child welfare se-rvices while in out-ofhome care (Garland & Besinger, 1997; Garland et al., 2000). Studies on intensive interventions for
children in out-of-home care, including wraparound services and treatment foster care, have yielded
mixed results (Clarke et al., 1992: Farmer et al., 2003). Additionally, although research does
indicate that foster and kinship alternate caregivers are in need of a variety of services, research on
the effectiveness of these services is not available (Cuddeback & Orme, 2002). Similarly, research
on permanency planning practices is limited There is evidence that children of color are more
likely than White children to have lengthier stays in the CWS once parental rights have been
terminated (Barth et al., 1994; Kapp et al., 2001 ). Research also suggests that some child welfare
workers may fail to encourage transracial adoption or placement (Avery, 2000; Brooks & James,
2003).
Additionally, contextual factors such as cultural competency of child welfare workers and
practices; child welfare practice in rural areas; and the role of the court system all impact the
effectiveness of child welfare practices. Cultural competence in child welfare settings, although
widely acknowledged as important, has not been adequately researched. Studies do suggest that the
lack of linguistically appropriate se-rvices may result in the under-utilization of se-rvices, or
inaccurate or incomplete assessments and inte-rventions (Gomez et al., 1985; Suleiman, 2003).
Moreover, research also suggests a need to differentiate between rural and urban areas when
researching and discussing child welfare practices (Templeman & Mitchell, 2002). Rural areas in
comparison to 1!,rban areas have been found to have unique risk/actors and child welfare agencies
in rural and urban areas have been found to differ in a number of ways (Drake, 1996; English et
al., 2002; Menanteau-Horta & Yigzaw, 2002; Ondersma et al., 2001). As such, there is a need to
define and evaluate rural-specific child welfare practices (Templeman & Mitchell, 2002).
Although more research is needed to discern how the court system affects child welfare
practice, research does suggest that there are numerous barriers to effective collaboration between
court and child welfare systems. Cross-training and clarification of roles have been identified as
important factors in improving the relationship between the child welfare and court systems (Cahan,
1992; Russell, 1988).
In general, much of the research literature on the impact of child welfare practices is
inconclusive. Yet very few studies describe differences in outcomes by ethnic/racial groups and
most of the research fails to even mention the effect of culture and ethnicity on the delivery and
impact of child welfare practices. This gap in the research literature is striking and has important
implications for children and families of color in the CWS. The lack of adequate knowledge of the
types of practices, services, programs, policies, tools, or strategies that are effective for children and
families of color hinders our ability to remedy the disproportionate representation of children of
color in the CWS and may leave many children and families of color without adequate or
appropriate services. The next section focuses on child welfare practices in Santa Clara County's
child welfare system.
47
IV. SANTA CLARA COUNTY CHILD WELFARE PRACTICES
In Phase 3, one area of investigation focused on the impact of child welfare practices on
children and families of color in the Santa Clara County's Department of Family and Children
Services (DFCS). Several related analyses were conducted to understand the types of practices and
case characteristics that are associated with more successful outcomes for children and families
from various racial/ethnic groups; the types of child welfare practices used in Santa Clara County;
how practices in the Main County Offices differ from South County Office practices; how practices
interact with child, family and cultural characteristics of differing racial/ethnic groups to produce
various outcomes; and the role of the court system on child welfare practices. As such, this section
contains five overall analyses:
1) A successful case analysis. This analysis examined case record review data collected in
Phase 2 to determine the types of practices and case characteristics associated with
successful case outcomes.
2) A qualitative analysis of Santa Clara County child welfare practices. Focus group and
interview data from Santa Clara County workers were collected to obtain information
about current practices, how these specific practices might positively or negatively
impact children and families of color, and what practices might be created, enhanced or
maintained to better serve children and families of color.
3) A comparison of Main County Offices versus South County Offices. This analysis used
both quantitative data from CWS/CMS, as well as qualitative data from focus groups in
South County to compare case characteristics and practices between the two areas of the
county.
4) A multivariate analysis of factors associated with reunification. This analysis
examined system-related factors and case characteristics predictive of reunification of
children with their families.
5) An analysis of the role of the court system in child welfare practice. This analysis
used case record review data collected in Phase 3 to determine whether and what type of
court-ordered changes are made to the social worker case plan recommendations at the
jurisdictional/dispositional hearing. Results from agency-wide focus groups related to
the relationship between the child welfare and court systems are also provided.
Taken together, these five sections provide extensive and important information on Santa Clara
County child welfare practices and factors that are associated with successful outcomes for children
and families of color.
48
Successful Cases Analysis
In Phase 3, one aim was to define and assess successful cases and outcomes in Santa Clara
County's Child Welfare System. An analysis of factors associated with successful case outcomes
and how these factors and outcomes differ by racial/ethnic group helps shed light on the types of
practices and conditions necessary to promote enhanced outcomes for children and families of
color.
Methods
First, our task was to operationally define "successful cases and outcomes" for this
quantitative portion, which we based on the United States Department of Health and Human
Services (US DHHS) Children's Bureau standardized outcomes. These standardized outcomes are
part of the U.S. DHHS' child and family services reviews and the Title IV-E foster care eligibility
reviews (U.S. DHHS, 2003b). These levels of success and outcomes are:
Safety
• Children are first and foremost protected from abuse and neglect
• Children are safely maintained in their homes whenever possible
Permanency
• Children have permanency and stability in their living situations
• The continuity of family relationships and connections is preserved for families
Family and Child Well-Being
• Families have enhanced capacity to provide for their children's needs
• Children receive appropriate services to meet their educational needs
• Children receive adequate services to meet their physical and mental health needs
Second, from our in-depth reviews of the 403 cases sampled for Phase 2, we matched basic
information and additional data collected from the court reports during Phase 3 with the US DHHS
guidelines. Although information from every court hearing available within our sample was
recorded, we relied primarily on data from the jurisdictional/dispositional hearing and the last two
hearings of each case to evaluate "successful cases and outcomes." We then expanded our analyses
started in Phase 2 to evaluate "successful cases and outcomes" and differences by ethnicity for this
current phase.
Results
Safety
Regarding safety, referring to the US DHHS criteria mentioned above, we used three
variables to indicate "children are first and foremost protected from abuse and neglect." Number of
prior referrals and previous number of times in the CWS indicated a risk of harm affecting these
children. The number of times removed from the family indicated a family environment that was
assessed as harmful and thus warranted additional removal.
49
We also used the initial case assignment from the jurisdictional/dispositional hearing to
denote whether "children are safely maintained in their homes whenever possible." A similar
variable, case closure type, which also refers to maintenance of the child in the home, will be
discussed in the permanency section.
Number of Prior Referrals
Of the 403 cases, 48 had missing data. Of the remaining 355 cases, 118 (33.2%) had no
prior referrals, 111 (31.2%) had 1 or 2 prior referrals, and the remaining cases (35.6%) had more
than 2 prior referrals. This indicated that more than half of the cases had prior referrals, which are
used by social workers in establishing potential risk to child safety. One-third of the cases had more
than two prior referrals. Please see Table 1a.
. Retierralsa
Table 1a: Freauency of Pnor
Number of Prior
Frequency
Referrals
0
118
1
75
2
36
3
35
4
24
5
18
6
14
7
6
8
7
4
9
10
3
11
5
12
3
13
5
14
1
15
0
16
0
17
0
18
1
a
Based on 355 cases with valid information
Percentage
33.2%
21.1%
10.1%
9.9%
6.8%
5.1%
3.9%
1.7%
2.0%
1.1%
.8%
1.4%
.8%
1.4%
0.3%
0.0%
0.0%
0.0%
0.3%
There was not a statistically significant relationship between number of prior referrals and
ethnicity. An average of2.46 prior referrals (sd = 3.14) were made before the case was opened for
the current episode. Please see Table I b.
50
Table lb Num berofp nor
. Re erra1s B:y Eh
. Grouoa
t me
Total
White
African
American
Prior Referrals
2.64
Avg.
2.46
2.14
(sd)
(3.14)
(3.27)
(2.71)
Latino
Asian
American/PI
Other
2.46
(3.25)
1.62
(2.14)
2.81
(3.37)
a Based on 355 cases with valid information
Previous Times in the CWS
Of the 403 cases, 93 had missing data. Of the remaining 310 cases, 204 (65.8%) had no
prior experience with the CWS while 106 (34.2%) did have prior experience with the CWS. This
indicates that one-third of cases had an elevated risk for abuse or neglect as evidenced by prior
substantiated abuse or neglect. Please see Table 2a.
Table 2a: Previous Times in the CWS8
Previous Times
Frequency
In the CWS
0
204
1
85
2
10
3
6
4
4
5
0
6
1
a Based on 310 cases with valid information
Percentage
65.8%
27.4%
3.2%
1.9%
1.3%
0.0%
0.3%
There was not a significant relationship between the number of previous times in the CWS
and ethnicity. On average, children across ethnic groups were previously in the CWS 0.47 times (sd
= 0.83). Please see Table 2b.
Table 2b: Previous Number of Times in CWS bv Ethnicity a
White
Latino
Total
African
American
Average
(sd)
0.47
(0.83)
0.25
(0.49)
0.50
(0.92)
0.46
(0.78)
Asian
American/PI
Other
0.63
(1.06)
0.56
(0.82)
a Based on 310 cases with valid information
Number of Times Removed from Family in Current Episode
Of the 403 cases, 15 had missing data. Of the remaining 388 cases, 257 (66.2%) were
removed once and 131 (33.8%) were removed more than once, indicating that in one-third of the
cases, the safety level in the home regarding the protection from abuse and neglect was not
consistent. Please see Table 3a.
51
·1 .m Current E,p1so
Tabl e 3a: T1mes Removed firom Fam11y
· de8
Number of
Frequency
Percentage
Times Removed
1
257
66.2%
2
98
25.3%
3
24
6.2%
4
7
1.8%
5
2
0.5%
a Based on 388 cases with valid information
There was a significant relationship between the number of times removed from family in
current episode and ethnicity. Across all ethnic groups, children were removed an average of 1.45
times (sd = 0.74). African American children had the highest average number of times removed
from family in current episode with 1.67 times (sd = 0.90), and Asian American/Pacific Islander
children had the lowest average with 1.26 times (sd = 0.51 ). Please see Table 3b.
Table 3b: Number of Times Removed from Family in Current Episode by Ethnicity'a
White
Asian
Total
African
Latino
American/Pl
American
Average
(sd)
a
1.45
(0.74)
1.67
(0.90)
1.36
(0.68)
1.44
(0.68)
1.26
(0.51)
Other
1.64
(0.92)
Based on 388 cases with valid information
Case Assignment After Jurisdictional/Dispositional Hearing
Of the 403 cases, 12 had missing data. Of the remaining 391 valid cases, 88 (22.5%) were
assigned to family maintenance. This indicates that less than one-quarter of cases are safely
maintained in the home whenever possible. There was not a significant relationship between case
assignment after the jurisdictional/dispositional hearing and ethnicity. Overall, most cases were
assigned to family reunification (71.1%, 278 of391 cases). Please see Table 4.
• a
. d'1ct1ona
. 1/0"1spos1t1onaIHearmg·
Table 4 Case A"
ssuinment aft er Juns
White
Latino
Total
African
American
34
Family
27
88
7
(19.4%)
(25.8%)
(22.5%)
Maintenance
(14.9%)
Other
Asian
American/PI
9
(29.0%)
11
(26.2%)
Family
Reunification
278
(71.1%)
38
(80.9%)
91
(68.9%)
101
(72.7%)
19
(61.3%)
29
(69.0%)
Other
25
(6.4%)
2
(4.3%)
7
(5.3%)
11
(7.9%)
3
(9.7%)
2
(4.8%)
a Based
on 391 cases with valid information
52
Permanency
Regardingpermanency, referring to the US DHHS criteria mentioned above, we used three
variables to indicate whether "children have permanency and stability in their living situations."
The number of unique placement homes and the number of placements in the current episode ·
indicated the consistency of the children's living situation while in the CWS. Since permanency and
stability also imply a solution of consistency within particular settings, case closure type was also
examined. To evaluate the outcome of whether "the continuity of family relationships and
connections is preserved for families" we also examined the proportion of reunified families within
the case closure type.
Number of Unique Placement Homes in Current Episode
Of the 403 cases, 15 had missing data. Of the remaining 388 valid cases,-162 (41.8%) had
only one or two unique placement homes. 224 (58.2%) had more than two unique homes, which
indicated that less than half of the children in the CWS are in a relatively permanent living situation.
Please see Table Sa.
Table 5a: Num berofU mque
.
Pl acement Homesa
Percentage
Number of
Frequency
Placement Homes
14.2%
1
55
27.6%
2
107
16.0%
62
3
11.9%
4
46
7.0%
5
27
6.2%
24
6
2.1%
7
8
2.8%
8
11
1.8%
9
7
2.3%
10
9
0.5%
2
11
1.0%
12
4
2.1%
13
8
0.8%
14
3
0.5%
2
15
0.0%
16
0
0.8%
17
3
0.3%
18
1
0.3%
19
1
0.3%
20
1
0.0%
21
0
0.3%
22
1
0.3%
1
23
0.0%
0
24
0.3%
25
1
53
26
27
28
29
30
31
32
33
34
35
a
0
2
0
0
0
0
0
1
0
1
Based on 388 cases with valid information
0.0%
0.5%
0.0%
0.0%
0.0%
0.0%
0.0%
0.3%
0.0%
0.3%
There was not a significant relationship between number of unique placement homes in
current episode and ethnicity. Across all ethnic groups, there was an average of 4.54 (sd = 4.67)
unique placement homes in the child's current episode. Please see Table Sb.
Table 5b: Number of Unique Placement Homes by Ethnicity•a
Total
African
White
Latino
American
Average
(sd)
a
4.54
(4.67)
5.35
(6.48)
4.70
(4.60)
4.44
(4.26)
Asian
American/PI
Other
3.29
(3.22)
4.36
(4.59)
Based on 388 cases with valid information
Number of Placements in Current Episode
Of the 403 cases, 16 had missing data. Of the remaining 387 valid cases, 222 (57.4%) had
only one or two placements. However, 165 cases had more than two placements indicating that
42.6% of children in CWS are in a relatively inconsistent living situation. Please see Table 6a.
Table 6a: Num berof Umque
.
Placement Homesa
Percentage
Number of
Frequency
Placements
20.4%
1
79
37.0%
2
143
14.5%
56
3
42
10.9%
4
5.2%
5
20
2.8%
6
11
2.1%
7
8
0.8%
8
3
2.1%
8
9
1.0%
4
10
1
0.3%
11
12
0
0.0%
54
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
a
4
2
1
0
0
1
0
1
1
0
1
0
0
0
0
0
0
0
0
0
1
Based on 387 cases with valid information
1.0%
0.5%
0.3%
0.0%
0.0%
0.3%
0.0%
0.3%
0.3%
0.0%
0.3%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.3%
There was not a significant relationship between the number of placements in current
episode and ethnicity. Across ethnic groups the average number of placements in current episode
was 3.31 (sd = 3.34). Please see Table 6b.
Table 6b: Number of Placements in Current Episode By Ethnicity'a
White
Latino
Total
African
American
Average
(sd)
3.31
(3.34)
3.27
(3.23)
3.77
(4.06)
2.99
(2.78)
Asian
American/PI
Other
3.00
(3.16)
3.20
(2.83)
a Based on 387 cases with valid information
Case Closure Type
Of the 403 cases, 386 had valid information for the case closure type. Including those
adopted, reunified with family, or placed with a relative or guardian, 183 (47.4%) of the cases have
permanency and stability in their living situation. There was not a significant relationship between
case closure type and ethnicity. Please see Table 7.
55
Tb
a le 7: Case Closure Tvoe By Ethnicitv'a
Total
African
American
White
Latino
Asian
American/PI
Other
Family
Stabilized
98
(25.4%)
(27.1%)
13
40
(30.5%)
31
(23.5%)
5
(16.1%)
9
(20.5%)
Guardianship
established or
placement
with relative
44
(11.4%)
5
(10.4%)
13
(9.9%)
21
(15.9%)
2
(6.5%)
3
(6.8%)
Reunified
with parent or
guardian,
court or noncourt specified
25
(6.5%)
3
(6.3%)
11
(8.4%)
6
(4.5%)
0
(0.0%)
5
(11.4%)
Adoption
16
(4.1%)
4
(8.3%)
6
(4.6%)
2
(1.5%)
2
(6.5%)
2
(4.5%)
Emancipation
57
(14.8%)
6
(12.5%)
11
(8.4%)
25
(18.9%)
7
(22.6%)
8
(18.2%)
27
(7.0%)
6
(12.5%)
7
(5.3%)
10
(7.6%)
1
(3.2%)
3
(6.8%)
119
(30.8%)
11
(22.9%)
43
(32.8%)
37
(28.0%)
14
(45.2%)
14
(31.8%)
(FM)
Incarceration,
runaway or
medical
services
Court ordered
termination
and other
a Based on 386 cases with valid information
Reunification
Of the 403 cases, 386 had valid information to determine whether children were reunified
with their family at case closure. Of these 386 cases, 263 (68.1 %) were not reunified with family
56
and 123 (31.9%) were. This indicates the continuity of family relationships and connections are
preserved for less than one-third of these families. Please see Table 8a.
· h F amity
·1 a
T able 8a: Propo rf10n of Ch"ld
1 ren Reum"fi1ed wit
Frequency
Percentage
Not Reunified
with Family
263
68.1%
Reunified with
Family
123
31.9%
a Based
on 386 cases with valid information
There was not a significant relationship between reunification with family and ethnicity.
Please see Table 8b. However, it is important to note that in the multivariate analyses that appears
later in this report (please see page 103), significant variations in reunification by ethnicity were
found even after controlling for other demographic and system-related factors.
Table 8b Cht"Id ren Reum"fi1ed By Ethnicitya
Total
African
American
White
Latino
Asian
American/PI
Other
Not Reunified
with Family
263
(68.1%)
32
(66.7%)
80
(61.1%)
95
(72.0%)
26
(83.9%)
30
(68.2%)
Reunified
with Family
123
(31.9%)
16
(33.3%)
51
(38.9%)
37
(28.0%)
(16.1%)
5
14
(31.8%)
a Based on 386 cases with valid information
Family and Child Well-Being
Regardingfamily and child well-being, referring to the US DHHS criteria mentioned above,
we created a summary score based on 16 variables categorized in 4 domains (overall well-being,
economic status, health status, and substance abuse related to the mother and the father) evaluated
from the last two hearings to indicate if "families have enhanced capacity to provide for their
children's needs." Although we originally wanted to examine each domain separately, the
extremely sparse information available from the last two hearings of each case prohibited these
comparisons at such a detailed level.
We then evaluated whether the child's education status was positive from the last two
hearings to indicate if "children receive appropriate services to meet their educational needs.
57
Finally, we created a summary score to evaluate the status of the child. This was based on 8
variables within similar domains above (child's overall well-being, economic status, health status,
and substance abuse) to indicate whether "children receive adequate services to meet their physical
and mental health needs."
Families Enhanced Capacity to Provide For Their Children's Needs
Of the 403 cases, 156 had valid information to determine if families had an enhanced
capability to provide for their children's needs. A positive score on this quantified information
meant that this capacity was generally enhanced, and a negative score meant the capacity was
generally diminished. The possible range for this summary score was -2.0 to +1.0. The average
score was 0.01 (sd = 0.71), which indicated that on the whole, case outcomes seemed neutral in
terms of family's capacities to provide for their children's needs. There was a marginally significant
relationship with ethnicity. African American families appear to have a more enhanced capacity
while Asian American/Pacific Islander families seem to have a more diminished capacity. Please
see Table 9.
Table 9: Family's Capacity to Provide for Their Children's Needs By Ethnicity'a
White
Asian
Total
African
Latino
American/PI
American
Average
(sd)
0.01
(0.71)
0.35
(0.59)
0.00
(0.77)
-0.04
(0.69)
-0.50
(0.53)
Other
0.07
(0.62)
a Based on 156 cases with valid information
Children Receiving Appropriate Educational Services
Of the 403 cases, 76 had valid information to determine if children received appropriate
services to meet their educational needs. A positive score meant that their educational needs were
generally met, and a negative score meant that their educational needs were generally unmet. The
possible range for this summary score was -1.0 to +1.0. The average score was 0.30 (sd = 0.71)
which indicated that educational needs were being somewhat met. There was not a significant
relationship with ethnicity. Please see Table 10.
Table 10: Children Receiving Appropriate Services to Meet Educational Needs By Ethnicitya
Other
Asian
White
Latino
Total
African
American/PI
American
Average
(sd)
a Based
0.30
(0.92)
0.75
(0.71)
on 76 cases with valid information
0.32
(0.89)
0.03
(0.98)
0.50
(0.93)
1.00
(0.00)
58
Children Receiving Adequate Services to Meet Physical and Mental Health Needs
Of the 403 cases, 156 had valid information to determine if children received adequate
services to meet their physical and mental health needs. A positive score in the summary variable
meant that their needs were generally met and a negative score that their needs were generally
unmet. The possible range for this score was -1.0 to+ 1.0. The average score was 0.11 (sd = 0.71)
indicating that on the whole, the case outcomes seemed neutral in regards to meeting children's
needs. There was not a significant relationship with ethnicity. Please see Table 11.
T a ble 11 Ch.Id
I ren R ece1vmg Adequate serv1ces to Meet Needsa
Total
African
White
Latino
American
Average
(sd)
0.11
(0.71)
-0.05
(0._76)
0.21
(0.69)
0.07
(0.71)
Asian
American/PI
Other
0.13
(0.64)
0.07
(0.83)
a Based on 156 cases with valid information
Summary of Key Findings and Implications
In Phase 2, we analyzed major case characteristics on the main sample of 1720 cases
(primarily from CWS/CMS data) and additional psychosocial factors measuring child and family
well-being for the 403 cases relying on data gathered directly from the case files. We conducted
ethnic comparisons for both samples. In Phase 3, our aim was to evaluate successful cases and
outcomes using the US DHHS guidelines and assess ethnic differences. Results indicate that there
were minimal ethnic differences
Variations of Safety, Permanency, and
Family and Child Well-Being by Ethnicity
In regards to safety, there was a significant relationship between the number of times
removed from family in current episode and ethnicity where African American children had the
highest average number of times removed from family in current episode and Asian
American/Pacific Islander children had the lowest average.
In terms offamily and child well-being, there was a marginally significant relationship
between a families capacity to provide for their children's needs and ethnicity, with African
American families appearing to have a more enhanced capacity while Asian American/Pacific
Islander families seem to have a more diminished capacity. However, from this portion of our
analysis, the general absence of significant differences by ethnic group assessed from the end of the
case implies that certain aspects leading to a successful case and outcome are not related to
ethnicity.
It is important to note that major ethnic differences do exist in Santa Clara County's CWS,
but many of these variations appear at different choice points in the system (such as psychosocial
status at the beginning of the case, and assignment to voluntary family maintenance). In addition, a
combination of factors experienced through the course of the case may be related to success or
59
positive outcomes. However, our results suggest that ethnic differences at the conclusion of the
case are not evident, if we define success according to DHHS guidelines.
Implications Related to the Disproportionate Representation of
Children of Color in the CWS
Given the relatively few ethnic differences in evaluating successful cases and outcomes,
what does this imply about ethnicity and the CWS? The absence of ethnic differences in this
segment does not negate the previous findings in Phase 1 and 2 confirming that certain ethnic
groups are over-represented (i.e., Latinos, African Americans) while others are under-represented
(i.e., Asian American/Pacific Islanders) in the county's CWS. Phase 2 results also indicate that
there are significant ethnic differences in many demographic, system-related, and psychosocial
characteristics. Yet, what does the absence of ethnic differences in successful cases and outcomes
mean in context? While results suggest there is a clear disproportion of ethnic groups in the system,
once in the system, the children and families are generally faring the same. Their relatively
homogeneous experience may be a result ofthe "one-size-fits all" limited array ofservices offered.
Although this may seem equitable and logistically efficient, it appears ultimately ineffective in
serving a diverse group of children and families. Also, to determine better how ''well" these
children are actually doing while in the system, the availability ofcomparative data, clear
operational definitions for "success," and more reliable and valid data are needed.
Limitations and Recommendations
The major limitations for this successful cases and outcomes component include the lack of
information available (especially at case closure), to establish an operational definition of a
"successful case and outcome," and the need for baseline and comparative data.
Data Issues
To assess better whether safety, permanency, and family and child well-being are
maintained and enhanced, child welfare records should include more complete and measurable
information. We acknowledge that child welfare case records were not originally intended for
research. Yet, if the agencies responsible for providing effective services wish to evaluate the
performance of their programs more reliably and validly, improving information collection and
tracking should be considered.
One step would be to create a case summary check-out form used at case closure or in an
exit-interview type situation. The check-out form would indicate the child and family's status in
major areas of psychosocial well-being (i.e., health, education, and finances), in addition to the
existing case outcomes. Given the amount of missing data assessing these domains describing case
and child status, and the relative scarcity of information recorded in the last two hearings
(presumably the place where outcomes of a case would be recorded) this form would enable
measurement and evaluation of "successful cases and outcomes" in a more consistent objective
manner. The amount of missing data and the manner in which information is assessed also implies
the need for more reliable and valid methods of recording of information.
Operational Definition of Successful Case and Outcome
60
Having a clear operational definition of a "successful case and outcome" is vital to program
evaluation and research. When we used the US DHHS criteria, the initial impression maybe that a
relatively low proportion of"successful cases and outcomes" seems to exist in regards to safety,
permanency, and well-being as for example only 32% of the children are reunified with their
families. However, again availability of comparative data along with the consistent use of a welldefined operational definition of "success" is needed to understand the actual success rates for
children and youth in the CWS. Two questions arise from our results and interpretation. First, what
actually is "success?" For example, adoption for one case may be a failure if a family has improved
its ability to support and nurture a child, but was denied reunification because the system still
deemed those parents unfit. Adoption for another case may be a success if the system permanently
removes a child from a persistently dangerous or unhealthy family environment. The same
difficulty in evaluation occurs when a family is reunified (presumably a success) but the child later
falls victim to reabuse (failure). Finding an agreed-upon operational definition may be first a
theoretical or philosophical question rather than a research one, but is important for the County to
consider if their programs are to determine their effectiveness
The second question concerns a judgment of Santa Clara County's CWS performance
according to the proportion of "successful cases and outcomes." Given the above limitations and
issues, Phase 3 cannot give a final answer to how well the county's CWS is serving children and
families. This brings us to our third limitation and set of recommendations, the need for baseline
and comparative data.
Baseline and Comparative Data
It was difficult to assess the successful cases and outcomes in Santa Clara County's CWS.
Are the proportions of children reunified with their families or of those adopted high or low? And
compared to what? Implementing better record keeping and data collection over time on key
characteristics, based on an agreed-upon operational definition would better answer the question of
effectiveness and performance. Implementing these practices across counties and across states
would also facilitate evaluation.
The recent Little Hoover Commission report (2003) urges reform in the CWS, especially for
the county and state to improve the leadership and management structure guiding the system (Little
Hoover Commission, 2003). As part of this reform, an accurate evaluation of success for children
and families involved in the CWS is needed.
61
Qualitative Analysis of Santa Clara County Child Welfare Practices
The qualitative component of this study consisted of both focus groups and interviews and
was designed to obtain insights about current child welfare practices used within the DFCS, how
these specific practices might positively or negatively impact children and families of color, and
what practices might be created, enhanced or maintained to better serve children and families of
color.
Methods
For the purposes of both the focus groups and the interviews, "practices" were defined
broadly to include programs, services, strategies, policies, or tools. Qualitative methods, including
focus groups and interviews, are ideally suited for capturing and documenting common experiences
about what "works" and what might be improved from the perspective of individuals involved in
different parts of the child welfare system, including Social Workers, supervisors and clients. These
methods may also help to illuminate quantitative findings, offer information about formal and
informal local practices, and provide a context for understanding a variety of factors that might
influence practices. The methods employed for gathering qualitative data through focus groups and
through interviews are described below.
A total of 16 structured focus group interviews were conducted during the course of this
study. Thirteen of the focus groups were composed of Social Workers, specifically workers with
titles of Social Worker II or III. The focus groups were structured to explore practices at different
choice points in the child welfare system. Specific topics for focus groups included the following:
• Emergency Response/Dependent Intake (early choice points related to entry into the system and
initial assessment and investigation).
• Diversion (provision of voluntary family maintenance, informal supervision, early intervention,
and other efforts to divert families to services without court intervention).
• Case Assignment (early choice points related to cases that are substantiated and opened).
• Mandated Services (including court ordered parenting classes, substance abuse treatment,
domestic violence services, mental health counseling or other services).
• Time Limits (including local, state and federal time limits and the impact on practices with
children and families of color).
In order to ensure that concerns of clients were adequately represented in the project, three
focus groups were conducted with parents in the child welfare system. Individual interviews were
also conducted with supervisors with experience in each of the choice points described above.
Individual interviews were used to supplement and affirm findings from the focus groups.
Social Worker Focus Group Sampling and Recruitment
Prospective participants for the Social Worker focus groups were randomly selected from a
list of workers with title of Social Worker II or Social Worker III in different units/division. Lists of
randomly selected Social Workers were generated using SPSS (Statistical Package for Social
Sciences). This random selection was conducted ''without replacement," so that if Social Workers
were randomly selected for a specific focus group topic, they would be excluded from random
samples for other focus groups. Table 12 below provides a list of the 13 Social Worker focus
62
groups, the child welfare units from which focus group participants were selected, and the number
of focus group participants.
Approximately 15 to 20 Social Workers were randomly selected to receive an invitation to
attend each of the 13 Social Worker focus groups. In recognition of possible scheduling conflicts,
emergencies, or vacations of invitees, we invited more Social Workers than were expected to attend
each focus group. Prospective focus group participants were sent an initial invitation by email two
weeks to one month in advance of the focus group. Non-responding Social Workers were sent a
second invitation to participate in the focus group. Finally, prospective focus group participants who
agreed to attend a focus group were sent a confirmation email and/or phone call two to three days
prior to the scheduled group. Each focus group was comprised of approximately 5 - 8 members,
though some were smaller. In particular, the two case assignment focus groups were small, with two
and three participants in each of the groups respectively. A total of 67 Social Workers participated
in the focus group interviews.
Table 12 Emo ovee Focus Group Comoos1t1on
Focus Group
Units Included in Selection Process
Diversion
ER/DI
Case Assignment
Mandated
Services
Time Limits
Court Services
• Early Intervention/Intensive Family Services/
Home Supervision
• Informal Supervision
Family Resource Centers
• Asian Pacific Family Resource Center
• Nuestra Casa Family Resource Center
• Uiirani Familv Resource Center
Emergency Response
• Screening
• Emergency Response
Court Services
• Deoendent Intake
Emergency Response
• Emergency Response
Court Services
• Dependent Intake
Continuing and Child Welfare Services Bureau
• Court Supervision
• Domestic Violence Unit - Court Supervision
• Continuing
• Spanish Speaking Services
• Continuing Multilingual
Adoptions Family Assessment and Development
• Adoptions and Family Drug Court
Continuing and Child Welfare Services Bureau
• Court Supervision
• Domestic Violence Unit - Court Supervision
• Continuing
• Spanish Speaking Services
• Continuing Multilingual
Adoptions Family Assessment and Development
• Pre-Adoption and Drug Court
• Family Assessment and Development
Number of
Groups
Number of
Participants
2
14
2
5
2
14
2
2
11
63
South County
TOTALS
•
Family Linka~e
South County Region
• Gilroy Family Center
• Rosanna Child Welfare Unit
• Monterey Road Child Welfare Unit
• Highland Child Welfare
3
13
13
67
Social Worker Focus Group Procedures
The Social Worker focus groups were designed to examine practices that impact children
and families of color at different choice points or units within the child welfare system. To this end,
the focus group questions explored the context in which services are provided, specific practices
that might positively or negatively impact children and families of color, and recommendations
about practices that might be maintained, enhanced or initiated to better serve children and families
of color (please see Attachment 1 for the Social Worker Focus Group Interview Guide). Each of the
focus groups took place in a room located at a DFCS office. Ten focus groups were conducted at the
main administrative office in downtown San Jose and three were conducted at Gilroy Family
Resource Center in South County. The focus groups were approximately 1½ hours in length.
All focus group participants received and signed a consent form that explained the study and
confirmed their voluntary participation in the focus group. All focus group participants also
completed a brief demographic questionnaire. The summary of focus group participant
demographics is provided in Attachment 2.
Family Focus Group Sampling and Recruitment
Three focus groups were conducted with families involved in the child welfare system. Two
focus groups were conducted with parents of children from populations that are over-represented in
the child welfare system in Santa Clara County, specifically Latino and African American children
and families. Each of these focus groups was organized in partnership with one of the family
resource centers and non-random, purposive sampling methods were employed. Participants for the
Latino parents' focus group were recruited among recent graduates of a monolingual Spanish
speaking parenting class at the Nuestra Casa Resource Center. Four parents agreed to participate in
the focus group and two were able to attend. Staff ofUjirani Family Resource Center recruited
participants for the African American parents focus group. Five parents and four staff attended the
focus group. Comments from parents and staff were documented separately. Finally, a focus group
of foster parents was convened to explore the needs of children of color in out of home placement.
The six focus group participants in the foster family focus group were all women; 2 participants
identified as Latina, 1 identified as Latina/Native American, 2 identified as Caucasian/European
American, and 1 identified as African American.
Family Focus Group Procedures
The three parent focus groups were approximately I hour in length. Questions explored in
these group interviews were focused on what has worked well for clients in getting services from
Child Welfare; what has not worked well and what services should be kept, improved or created to
better meet the needs of families. Please see Attachment 3 for the Family Focus Group Interview
64
Guide. Participating parents signed a consent form and each parent received a $10 gift certificate
from Target as a token of appreciation for contributing to the study. The family focus groups are
described in Table 13 below.
Table 13 F amity
·1 F OCUS Group Compos1t1on
Focus Group
Sponsoring Organization
Latino/ Mono-Lingual
Spanish Speaking Parents
African American Parents
Foster Families
Nuestra Casa
TOTALS
Ujirani Family Resource
Foster Home Licensing,
Recruitment & Intake - Resource
Families Support Team
Number
of
Groups
I
Number of
Participants
I
I
5
6
3
13
2
Individual Interviews Sampling and Recruitment
Fourteen in-depth interviews, of approximately one hour in length, were conducted with
supervisors in different positions in the Department of Family and Children's Services. Prospective
supervisor interviewees were randomly selected using the same sampling procedures described
earlier. In several cases, because the total number of supervisors was small, all supervisors in some
practice areas (such as Emergency Response or Dependent Intake) were included on the list of
prospective invitees. Based on the assumption that some supervisors would decline or be ·unable to
complete interviews, more participants were invited than target number of interviews (15). A total
of23 supervisors were invited to participate in interviews and 14 completed interviews.
Individual Interview Procedures
The interviews used a critical incident approach that is designed to elicit descriptions of
specific behaviors that help formulate the critical requirements of a specific activity (Flanigan,
1954), such as effective child welfare practice with children and families of color. The critical
incident technique has been used to research decision-making in child welfare (Banach, 1998) and
is a valuable tool for obtaining information about specific behaviors and practices from
practitioners, who are in the best position to evaluate which of these practices "work" or "don't
work." Rather than gathering general opinions and hunches, this method allows interviewees to
recall specific experiences related to the provision of services and to delineate factors related to
success or failure in outcomes, such as outcomes related to child welfare intervention for children
and families of color.
Interviewees responded to three (3) open ended questions and several "probe" questions that
solicited detail about specific critical incidents in practice with children and families of color. In
general, the questions explored: 1) factors and practices associated with a case perceived by the
interviewee to represent a successful outcome with an African American, Latino or Asian
American/Pacific Islander family, 2) factors and practices associated with a case considered
unsuccessful, and 3) specific practices that result in or could potentially result in enhanced
65
outcomes for children and families of color (please see Attachment 4 for the Supervisor Individual
Interview Guide). Demographic information was collected at the close of the interview.
The supervisors interviewed for this study represented different practice areas of DFCS
including Emergency Response (5), Court Services (3), Continuing and Child Welfare Services
Bureau (2), Adoptions Family Assessment and Development (1), and South County (3).
Interviewees all had extensive experience working in the Department of Social Services and
possessed an average of 17.5 years working with County DFCS (range of5.5 years to 33 years). All
fourteen interviewees possessed an MSW degree and some also were LCSW's. Half of the
interviewees were men and half were women. Four of the interviewees were Latino/a, five were
Caucasian/Euro American, three were African American, and two were Asian American/Pacific
Islander. Ten of the key informants are fluent in speaking another language, and seven use these
language skills in their work. A summary of demographic information for the 14 interviewees is
provided in Table 14.
Table 14 Superv1sor Interv1ewee Demograp 1cs
Bureau/Unit
Gender
n(¾)
Emergency Response= 5 (35.7%)
Court Services
=3 (21.4%)
Continuing
= 2 (14.3%)
Adoptions
= 1 (7.1%)
South County
= 3 (21.4%)
n(¾)
Males= 7 (50.%)
Females= 7 (50.%)
Ethnicity
n (%)
= 3 (21.4%)
African American
Asian/Pacific Islander= 2 (14.3%)
Latino/Hispanic
= 4 (28.6%)
White
= 5 (35.7%)
YearsCurrent
Position
mean
(std. dev.)
3.7
(2.4)
Years County
mean
(std. dev.)
17.5
(8.9)
Qualitative Analysis Procedures
Two members of the Child Welfare Research Team, a trained focus group leader and a
research assistant, were present during each of the focus groups with the exception of three of the
smaller focus groups that were facilitated and documented by one focus group leader. Notes were
taken during both focus group and individual supervisor interviews. Audio tape recordings were
made during social worker focus groups and individual supervisor interviews to ensure accuracy of
the transcribed notes. Data for the qualitative portion of this study consist of the transcribed notes
taken during each of the focus group and individual interviews. Qualitative methods of data analysis
were employed to analyze the focus group and individual supervisor interview data. First, major
themes were summarized for each of the choice points in the social worker focus group and
supervisor interview data. Second, themes from the family focus groups were analyzed separately.
Third, a separate analysis of themes in the qualitative data from South County Offices was
conducted to provide insights and supplement quantitative analysis of possible differences between
South County Offices and Main County Offices within DFCS. Finally, common themes across all
focus groups and interviews were identified and documented. The findings of the analysis of themes
from focus groups and interviews are presented below and are followed by specific
recommendations that emerged across focus groups and interviews.
66
Qualitative Results: Themes from Focus Groups and Interviews
Participants in focus groups and interviews were asked to describe practices that may
negatively or positively impact children and families of color in the Santa Clara County's CWS, as
well as contextual factors that may affect the delivery ·ofthese services or client outcomes. This
section describes the main themes that emerged from these focus groups and interviews. Following
the pathway that children may take through the CWS, both negative and positive practices are
described at beginning choice points as well as continuing care/ending choice points. Lastly,
important contextual factors will also be discussed in relation to the delivery and impact of these
practices. 1
Beginning Choice Points:
Practices that Negatively Impact Children and Families of Color
The main themes related to practices that negatively impact children and families of color at
beginning choice points included, 1) inappropriate referrals to the CWS, 2) inconsistency in
decision-making practices, and 3) inadequate prevention of child maltreatment or entry into the
cws.
Large numbers of inappropriate referrals were noted as having a negative impact on children
and families of color. Study participants reported that, "Many mandated reporters have no clear
definition or evidence of abuse," and that low-income children and families are frequently referred
into the system because of issues more closely associated with poverty than abuse or neglect. Focus
group participants stressed the high number of referrals from schools "based on their observations
of children wearing old or dirty clothes," "because children had lice," or in one case "because a
child brought an expired can of food to a food drive, the child was thought to therefore be
neglected." Some participants commented that low-income children from wealthier school districts
were disproportionately reported to child protective services. "A child who is at a marginal
economic level will be seen differently at a school in Los Gatos as opposed to a school anywhere
else."
Although study participants stressed that inappropriate referrals are generally screened out
by the Screening Unit, repeated referrals can be damaging because they create a perception of
increased risk that may or may not be accurate. As the number of inappropriate referrals increases
for a particular family the likelihood ofDFCS opening a case may also increase. This situation was
described as having a potentially negative impact on children and families of color in particular
because they are often more likely to have socioeconomic problems. Some respondents also
suggested that children of color are more likely to be scrutinized by law enforcement. "Police are
more harsh on African American, API and Latino Families, especially if they don't speak English,"
commented another participant.
These inappropriate referrals were described as negatively impacting screening practices
because social workers often have to spend a great deal of time screening out inappropriate .
1
Please note that focus group and interview participants discussed contextual factors related to the court system. These
findings are presented in the section entitled "The Role of the Court System in Child Welfare Practice" which begins on
page 109.
67
referrals. This situation then often reduces the amount of time social workers have to identify and
serve families who are in need of voluntary services.
Additionally, study participants felt that, in some cases, inconsistency in decision-making
practices during beginning choice points negatively impact children and families of color. Although
decision-making practices occur at all stages of the CWS, many participants.discussed these
practices in relation to clients' entry into the CWS. Although staff generally noted the utility of
structured risk assessment tools used at the entry point into the CWS, they also stressed that these
tools are limited and that social workers need to develop recommendations using both assessment
tools and their own judgment. In relation to the latter, participants expressed concern that decisionmaking is often determined by the biases and perceptions of individual social workers. Social
workers named specific risk assessment tools that are used at beginning stages of the CWS as
useful, such as the decision tree used during screening and a written protocol used in Early
Intervention. At the same time, social workers suggested that application of assessment tools might
vary among social workers based on a number of factors including individual worker bias, size of
the unit and unit level, shared expectations about assessment, and level of supervisor support and
oversight.
Moreover, study participants noted that, "Decision-making all depends on the worker and
his or her attitudes and values," and "We can so easily sway the picture so it is positive or negative
by how we write up our narratives for the court reports; the report has a lot of influence and
sometimes peoples' personal issues get in the way of making informed decisions." Furthermore,
interviewees noted that social workers are not immune to stereotypes of different groups in the
dominant culture and in the media and that social workers often judge families based on their own
personal experiences and perceptions about what is "good enough" for children. Some social
workers pointed out that, for example, workers from other cultures might misinterpret
communication norms in African American families as abusive. "Sometimes African American
families have norms about speaking loudly, yelling, and threatening (like I am going to pop
you) ... but kids are rarely hurt ... an African American social worker might be less likely to
[negatively] interpret these kind of verbalizations than a social worker from a different
background."
According to study participants, decision-making practices are also influenced by
socioeconomic factors. Families experiencing unemployment, lack of access to resources such as
daycare, or problems obtaining health care for children may be initially assessed as having a higher
risk for maltreatment. For example, "older siblings are taking care of younger siblings because
parents can't afford childcare," noted one participant. "If there were more good daycare there
wouldn't be as many neglect cases, even once in the system we [workers] can't get them [clients]
child care," added another. Social workers also underscored the impact of local high housing costs
on low-income families. This problem was encapsulated by the comments of one worker: "Many
families partner up with other individuals to maintain housing and just because there are numerous
people living in the home, relatives sharing beds, doesn't mean that this is abuse or neglect. It's
poverty." The case example below typified the illustrations provided by study participants:
A Social Worker 1 was covering a case and she was concerned about the family because
there were four children living in a basement with their family. This was a Chicano family.
She reported that the children might need to be removed I went out to this social worker
and saw completely different things. The basement was an apartment. The place was small,
68
but had couches and a crib for the baby; it was cluttered but not dirty ... Maybe another
worker might have justified what this worker saw and removed the kids for perhaps no good
reason at all.
The issue of social worker attitudes toward clients was particularly salient to family focus
group participants. Several parents commented on the notable variation in how they are treated by
different social workers. A remark that typified this theme was made by one parent who said, "some
social workers do their job because they really want to help other and some just do it for the money
and don't seem to care about others." Parents also described feeling undermined by social workers
who appeared to have a negative attitude about the client. One client summarized, "some of the
social workers are disrespectful, unfriendly, they look down on the clients and make you feel
humiliated." They elaborated, "It gets to the point where it makes the client feel like 'I don't want to
go in there,' but I have to go in there for my child and when you come out you're in tears."
Inadequate prevention of child maltreatment or entry into the CWS was also identified as
having a negative impact on children and families of color at beginning choice points. Participants
commented on the need for services in the community that would help to prevent entry into the
child welfare system. Focus group members stated that there is a need for education about child
welfare laws and definitions of child abuse, particularly for cultures in which norms related to
discipline of children differ from expectations in the dominant culture. Participants suggested that
such programs could be offered through Social Service agencies but also in community based social
institutions, such as churches, and that some of the communities that have benefited from this type
of preventive education in the past include African American, Asian American/Pacific Islander
( e.g., Vietnamese, Chinese, and Filipino), Latino, and immigrant communities.
Study participants also emphasized the problems in obtaining services without opening a
case. For example, one worker stated that "to get into a wrap-around service from an outside
agency, such as schools or probation, without the child being a dependent takes months, by this
point the problem will have escalated." Many social workers stated that opening a case had both
benefits and drawbacks: families had greater access to services and support through resources such
as Family Preservation Funds but were at greater risk for having a child removed because of time
limits or establishing a history with DFCS.
Several social workers also noted that it is actually more work for them to maintain a child
in the home with Voluntary Family Maintenance (VFM) services than it is to remove a child from
the home and refer to court intervention. Additionally, for families who do make it into VFM or
Informal Supervision (IS) services, focus group participants pointed out that resources were not
available to provide follow-up support to families to prevent re-entry into the system once their
cases are closed. "The family may still have needs but we have to close it. We can only keep it in IS
for 6 months (or 3 months in VFM); it relates to funding because you lose federal and state money
when a case is closed and you are on County money-and there is none," exemplified comments
from interviews. Consequently, "clients may end up back in services." Ideally "workers would be
able to provide some additional help to them [families] to help themselves."
Time lags between case assignment to Voluntary Family Maintenance and social workers
actually receiving the case were also noted as an issue. "By the time the social workers gets the
case, three weeks have gone by and the social worker has only 2 ½ months left [out of three months
of services under Voluntary Family Maintenance]," elaborated one social worke~.
69
Beginning Choice Points:
Practices that Positively Impact Children and Families of Color
The main themes related to practices that positively impact children and families of color at
beginning choice points included, 1) orientation of clients to the CWS, 2) team decision-making, 3)
family involvement in decision-making, and 4) the use of child care and in-home services.
Providing an orientation for parents to help them understand the components of the child
welfare systems, the court process, current laws and policies related to time limits, and how to work
with social workers was named as a best practice for families first entering the CWS. Social
workers in South County portrayed this as a best practice that was formalized into an orientation
group while social workers in other areas of the county described conducting orientations with
clients individually. Social workers and supervisors advocated for strengthening and
institutionalizing the practice of providing an orientation for clients throughout the county. Some
participants suggested that an orientation workshop could be provided or co-lead by someone
"outside the system" that might be perceived by clients as more neutral.
It would be helpful for parents to go to a workshop to educate them about time lines.
Parents don't get it immediately. Sometimes by the time the client "gets it, " it may too late.
Also sometimes social workers don't explain it well. Sometimes clients are just traumatized
about what all this means and you have to walk them through the process for the first 2
months and you have 6 months time limit - it's very difficult. The workshop should also
teach the clients how to read a court report and what the recommendations mean.
Team decision-making <TOM), which is currently being initiated in Santa Clara County,
was viewed by participants as a promising tool for mediating individual bias and cultural
misunderstanding by social workers. "Even with the decision tree in place we know it hasn't been
working because there are still so many children of color and there is still a lot of the worker's
subjective perspective put into the report; hopefully TDM can reduce this." TDM was also
perceived as a vehicle to "help get everyone on the same page" with professionals in other systems,
such as schools and service providers. Staffing to discuss cases within units and with
interdisciplinary teams was identified as a positive practice for children and families of color that
could be enhanced by TDM. Some social workers expressed concern about the amount of time that
might be required by TDM's and that "it may be hard to get everyone together in a limited time."
Study participants identified family involvement in decision-making. including the use of
family conferences as a positive practice, particularly for children and families of color. Participants
suggested this was particularly important for cultural groups that place high value on extended
family. "In the Latino culture it's appropriate to extend an invitation to grandparents, godparents,
and all parental figures," illustrated one social worker. Study participants stressed the importance of
extending family involvement to include both biological and non-biological family members. For
example, some participants noted that African American families often have non-biological aunts,
uncles, or other extended family members that should be included in family conferencing. Many
social workers and supervisors viewed the developing family-to-family initiative as a promising
strategy for expanding and institutionalizing family participation in decision-making.
70
Last, the use of childcare and in-home services were also noted as a positive practice. For
instance, in one social worker's opinion, families with adequate childcare have about an 80 percent
participation rate in diversion services, compared to 50 percent participation rate for parents without
adequate childcare. Social workers, supervisors, and families discussed the importance of childcare,
in relation to the ability of families to succeed in services. "It's really helpful to have child care
through Calworks; We need to make sure that is not cut because people can't work or train if there
is no one to watch their kids -This is a huge thing."
Social workers also noted that the provision of in-home services allows for "working with
clients in their own environment" and providing opportunities for one-on-one mentoring and role
modeling new behavior. Social workers identified agencies and programs that provide in-home
parenting classes as more effective than traditional didactic intervention practices alone. Social
workers also stated that providing in-home visits and follow up services, with the support of Social
Worker l's, was an effective practice that is underemployed because of heavy caseloads and scarcity
of resources.
Beginning Choice Points:
Recommendations to Improve Practices
Recommendations to improve practices at beginning choice points in the CWS included 1)
reducing bias and increasing accountability in decision-making, 2) training professionals in other
systems, 3) providing client orientations to families first entering the CWS, and 4) investing in
prevention and early intervention. Specifically,
Reduce Bias and Increase Accountability in Decision Making
•
•
•
•
Implement team decision-making in a manner that includes consideration of client culture.
Examine ways to facilitate accountability for team decisions.
Explore mechanisms to ensure accountability in decision-making and minimize the possible
influence of individual or supervisor bias in decision-making.
Continue the practice of facilitating cultural consultations between staff, with the support of
supervisors, including consultations across units.
Continue and expand practices related to family conferencing.
Training Professionals in Other Systems
•
•
•
Provide or collaborate in providing training for mandated reporters, including schoolteachers
and health care professionals.
Provide or collaborate in cross training with other systems and agencies "so we can understand
one another's needs and views-what we need from each other." Cross training may also serve
"to increase communication" or even "help with preventing some families from coming into the
system, including families that are already being seen by the agencies."
Provide or facilitate more training for police regarding removal, working with social workers to
assess risk before removal, coordinating with social workers regarding child and family
interviews, and ensuring sufficient training for law enforcement officers assigned to conducting
sexual abuse interviews.
71
Provide Client Orientation
•
•
•
Educate parents referred into the system about child abuse and neglect laws. Provide workshops
in their own language on topics such as court processes, how to work with social workers,
timelines, and how to navigate the system.
Consider having such workshops lead or co-lead by someone outside of the system that may be
perceived by clients as more neutral.
Hire and Train Social Worker l's to assist with family education and orientation.
Invest in Prevention and Early Intervention
•
•
•
Explore mechanisms to continue providing prevention services, including education of diverse
communities about child abuse and elder abuse.
Explore mechanisms to fund counseling and intervention services for families at risk that are
less restrictive than federally funded programs such as Family Preservation, which frequently
require formally opening a case to allow families to access resources.
Continue support and explore opportunities to expand voluntary and diversion services,
including early intervention and voluntary services at the family resource centers should
continue. These efforts are critical to "determine if families should be in the system" and ensure
opportunities for less restrictive placements for children.
Continuing Care/Ending Choice Points:
Practices that Negatively Impact Children and Families of Color
Study participants noted three overall practices that negatively impact children and families
of color at continuing care/ending choice points, including 1) problems with service plan
recommendations, 2) inconsistency in assessment of substance abusing families and need for
stronger linkages with substance abuse services, 3) barriers and service gaps related to relative and
foster placements.
Problems with service plan recommendations, including a "cookie cutter approach" to
services, and unrealistic expectations of families were noted as practices that have a negative impact
on children and families of color. For instance, several study participants suggested that using a
"cookie cutter approach" or "template for recommendations" is a practice among both social
workers and district attorneys that negatively impacts children and families of color. Participants
offered a number of explanations for this dynamic including "social workers being jaded, believing
that since they had worked with a similar family before, that the family is the same and the
interventions should be the same." Similar observations from social workers in other focus groups
suggested that "sometimes workers will get challenged if they don't take the cookie-cutter
approach," and sometimes "DA's think all parents need counseling, parenting class, etc." Several
social workers called for greater flexibility in working with families and "more creativity regarding
case plans needs. For example, some parents may just need more time to spend with their children."
Others cited examples where clients were mandated to services that were not appropriate including
a mother mandated to services before an assessment was conducted and a father "who is in a 52week batters' program-who had one incident of losing his temper and hitting his child-and
72
though hitting her was not okay he would be better served in family therapy and an advanced
parenting class."
Additionally, some service plan recommendations were also described as placing unrealistic
expectations on families that can set them up for eventual failure. These practices tend to impact
low-income families and were perceived to disproportionately impact children and families of color.
First, service plans frequently require "a laundry list of mandated services" that are unrealistic for
families. "I have a lot of clients who when I sit down with them and go over the service plan with
them for the first time they say 'I'm overwhelmed, I can't do this." "Families are in a panic to get
everything done on the service plan from ER to continuation," explained one social worker focus
group participant. Other participants noted that a family might be mandated to substance abuse
services, parenting classes, and mental health services while seeking employment and new housing.
Some families "just give up," observed participants. Moreover, clients are often required to attend
services that conflict with their job schedules, particularly clients whose schedules change from
week to week. "The expectation for clients to reschedule their work schedule is difficult, especially
those that receive hourly wages, and causes even more stress for the family."
Many participants expressed concerns that high expectations imposed on families are
reflective of and "ideal" rather than "good enough" standard for families. One study participant
summarized this dynamic: "We are not required to have 'perfect parents,' we need minimal
standards of parenting ... Just because you or I would not parent that way does not mean there is a
risk to the child." "From a client's side, they are being told by a court system and a social system
that the way that they were brought up and the way that things were done in their [clients'] family is
wrong and not acceptable," noted a social worker focus group participant.
Finally, study participants cited insufficient resources to support the success of families
attending mandated services as a problem. For example, participants in several focus groups stated
that the shortage of bus passes and increased paperwork required to obtain the limited passes that
were available negatively impact families, particularly children and families of color. "Most of
these clients don't have a vehicle so for the low income families it is challenging for them to get
their children back ... how are they supposed to fulfill their required programs without
transportation?"
Inconsistency in assessment of substance abusing families and need for stronger linkages
with substance abuse services was also noted as negatively impacting children and families of color
at the continuing care/ending choice point. Participants stressed the importance of effectively
addressing the needs of families with substance abuse problems who they perceived to comprise a
majority of caseloads. Study participants described a number of success stories in which families
were reunified when parents completed substance abuse treatment and also highlighted the fact that
when appropriate substance abuse services are provided they can be an effective best practice.
At the same time, study participants suggested that protocols for decision-making with
substance abusing families and formal agreements with substance abuse treatment agencies were
inadequate. Participants also espoused different standards about when it might be appropriate to
reunify a child with a parent or place a child with a relative in recovery. Furthermore, participants
suggested that there was a need for developing protocols for addressing relapse with substance
abusers in conjunction with both treatment programs and the courts. "We [social workers] pay lip
service, we say relapse is a part of recovery, but the system does not reflect that." Finally, several
73
study participants noted a lack of clarity and consistency in standards for assessing risk in the case
of drug exposed infants, suggesting that race and class bias may influence both initial referral and
assessment of parents of drug exposed infants in child welfare.
Study participants also commented on the scarcity of substance abuse services, particularly
residential programs for women and children. Although a variety of substance abuse treatment
services are offered throughout Santa Clara County, participants pointed out that the need for
services was greater than availability, particularly for clients whose primary language is not
English. Several participants, particularly in South County, noted that a centralized intake process
for entree into treatment was a barrier for many families because of time and transportation issues.
Additionally, social workers felt that barriers and service gaps related to relative and foster
placements also negatively impacted children and families of color at continuing care/ending choice
point practices. While the use of relative placements was described as an effective practice for
children and families of color, study respondents proposed that poverty issues and timelines
converged with other factors to create substantial obstacles to relative placement. First, pressure to
move children out of the children's shelter, in the context of little staff support to locate prospective
relative placements, often results in lost opportunities to place a child with a family member. Some
social workers described feeling "in a double bind" and, in some cases, deeply frustrated because of
having to place a child in a foster care situation that was perceived as less positive for the child.
Second, even if a relative placement is identified, it is often "impossible to complete the relative
placement paperwork within 24 hours, with the required background checks and fingerprinting."
Third, although social workers can seek a Director's Exemption to place children with families
where members may have had a criminal background, many social workers expressed concern
about the fact that even a distant criminal history can create barriers to appropriate relative
placements. Finally, new standards for relative placement "are the same as a licensed foster home
and are unrealistic for many families; this causes children to not be placed with relatives."
In general, although study participants expressed commitment to ensuring appropriate
relative placements for children, the study participants suggested that the requirements represented
by the relative packet "doesn't take into account economics or cultural practices" and
"disproportionately impacts families of color." These concerns were repeated across focus groups
and interviews and were epitomized by the following observations and examples:
The only people that can flt into it [relative packet] are those with money ... the packet
doesn't take into account the families we deal with. Many families are able to put four
children in a bedroom and they 're able to do fine, each child has their own space. It
[relative packet] does not allow for any arrangement like that at all. "
There was a case with 10 and 12 year old Hispanic/Native American girls who were
abandoned and went to live with a relative. The relative contacted CPS and stated that the
mother had not returned and wanted to know what to do with the children. The worker was
unable to assess the relative 's home immediately; therefore, the children were then placed
in a foster home. The licensed foster home was horrible, they ffoster parent] placed the
older girl on the floor on a crib mattress. It was terrible and all because they couldn't
continue to stay where they had been, because I [workerJ couldn't get the clearances fast
enough.
74
Three young Mexican American girls needed a place to live. The aunt was willing to take
them in and the children were comfortable with her. The uncle had a criminal history and
was a recovering heroin addict, who had been clean/or approximately 20 years. On paper
the uncle looked really bad, the worker was unable to clear his past. The children were
placed in foster care and may be adopted ifthe mother doesn't comply with the s~rvice plan.
In addition to barriers to relative placements, social workers pointed out a need to develop
strategies for working with foster care and adoptive families. Respondents noted that these alternate
caregivers are often deeply disappointed when the children in their care are reunified. Some social
workers suggested working more closely with fost-adopt families about the probabilities for
reunification. Foster family and social worker focus groups also expressed a need to provide
training to foster families in practical strategies for caring for children across cultural groups and a
need to develop mechanisms for experienced foster parents to mentor new foster parents.
Continuing Care/Ending Choice Points:
Practices that Positively Impact Children and Families of Color
Social workers described three overall practices that positively impact children and families
of color at the continuing care/ending choice points including I) parenting education services that
involve both parents and children in services, 2) cultural matching for children in foster care, and 3)
recruitment, training and coaching for foster families.
Parenting education services that involve both parents and children in services were
described as practices that positively impact children and families of color. Social workers and
supervisors noted that parenting classes that involved children and provided opportunities for
parent-child interaction were also a promising practice in comparison to traditional parenting
classes. Some social workers suggested that it might be helpful to ''wait until children have been
returned to order parenting classes" or to arrange for children's participation in classes.
Additionally, social workers described cultural matching for children in foster care in which
workers attempt to place children in a home with foster parents of a similar background as a
preferred practice. This was perceived to be critical in relation to matching the primary language of
the child and particularly important in the context of federal and state time lines and concurrent
planning. At the same time, both social workers and foster parents stressed the importance of a safe
and caring placement for children, whether a cultural match was possible or not. Social workers
expressed concern that the particular issues and needs of children, including but not exclusive to
culture, are sacrificed to the expedience of"finding any placement." Foster parents echoed a
concern about considering both the needs of the child and the experience of the foster parents when
arranging a "match." One foster parent caring for a special needs child of a similar background
noted that she would have been overwhelmed and tempted to abandon the idea of foster parenting if
her first foster child had been particularly challenging. Foster parents stressed that they and other
foster parents often have successful experiences with children of different backgrounds and that
they would "never want to see a child wait for a permanent home or left in a shelter because of a
cultural difference."
Social workers and foster parents suggested that recruitment. training and coaching for
foster parents is critical to working effectively with children and families of color. Specifically,
foster families need to know about how to care for children placed in their home. For example,
75
foster parents caring for African American children for the first time required basic information
about hair care. Foster parents suggested that best practices for working with children of color
included obtaining coaching and training from other foster parents, learning about the child's
culture and the individual child, and collaborating with social workers to interface with birth
parents. Study participants noted that tensions between foster families and birth families were
common but that positive practices for children required minimizing these potential problems.
Foster families described "success stories" in which foster parents maintained supportive
relationships with children who were returned to their family of origin.
Social workers and foster parents both noted the need to recruit diverse foster families,
particularly Latino and African American families in Santa Clara County. One foster mother who
identified as Latina suggested that some families in her community might be reluctant to become
foster parents because they are more likely to already have children than Anglo families. Study
participants identified outreach to prospective foster families in underrepresented communities as an
important continuing practice.
Continuing Care/Ending Choice Points:
Recommendations to Improve Practices
Recommendations to improve practices at continuing care/ending choice points included, I)
providing or facilitating access to concrete support for families, 2) facilitating increased availability
and access to services, and 3) supporting and developing foster families. Specifically,
Provide or Facilitate Access to Concrete Support for Families
•
•
•
Examine options for restoring access to bus passes for clients, both in terms of increasing
number of available passes and decreasing paperwork required from social workers to obtain
passes for clients.
Explore ways to increase access to childcare for parents participating in voluntary or mandated
services.
Facilitate access to services related to transitional housing, jobs, financial assistance (e.g.,
CalWORKS), and for families as well as educational interventions for children in the child
welfare system.
Facilitate Increased Availability/Access to Services
•
•
•
Work with policy makers, courts, communities, and other service delivery systems to identify
and address issues related to insufficient services for families. Areas where services are lacking
or limited in relation to the perceived need include substance abuse treatment (particularly
residential treatment for women and children), transitional and drug free living, mental health
services for children, services for young children exposed to trauma or substance abuse, after
school programs and other resources for children.
Collaborate with other systems to protect funding of key services in the context of state budget
problems, such as services for Oto 5 programs.
Work with contract agencies to facilitate inclusion of children in parenting classes to allow for
both didactic and experiential learning opportunities.
76
•
Sustain and increase in-home services, such as intensive wrap around services and one-on-one
mentoring during home visits.
Support and Develop Foster Families
•
•
•
Continue and expand efforts to recruit foster families that are representative of diverse
populations in the child welfare system and that match the languages of different children and
families (e.g., additional Spanish speaking and Vietnamese speaking foster families).
Provide training for foster families in practical strategies for caring for children across cultural
groups.
Continue to develop mechanisms for experienced foster parents to mentor new foster parents.
Contextual Factors that Negatively Impact Practices for Children and Families of Color
Study participants noted three overall contextual factors that negatively impact child welfare
practices for children and families of color 1) agency level factors that detract from social work
practice, 2) a need for more bilingual and bicultural social workers and services, and 3) time limit
constraints.
Certain agency level factors that detract from social work practice were mentioned by many
study participants as negatively impacting practices for children and families of color. For instance,
heavy caseloads and staff shortages were identified as critical factors impeding effective practices
for children and families of color. Children and families of color, particularly families who may be
less trusting or more intimidated by the system, may require language specific services, or who may
have multiple vulnerabilities were deemed very likely to be impacted by the heavy caseloads of
social workers. One social worker who works primarily with Spanish speaking families observed
that many of her clients require "a lot of hand holding" in navigating through the system. Given
heavy caseloads, even dedicated social workers that invest substantial time with some families will
have other families "slip through the cracks." One participant explained:
The weight ofthe caseload is so heavy-there's no time to do social work. I always feel I
have to rush. Every minute ofyour day is booked. There is no time to explain court
processes to clients.
Many participants commented positively on agency plans to reduce caseloads. This was
recognized as a critical move toward improving overall effectiveness in day-to-day social work
practice. Yet some social workers commented on some of the weaknesses in the current method for
determining caseloads. "There's no equity in the way the agency does the point system for
determining workers' caseloads .... it's based on a 100 point system and you could have 70 points
that could be 35 children in permanent placement, which is a lot of children to manage, whereas
someone else could also have 70 points in Family Reunification (FR) cases, which is not the same
workload." Others suggested that there appeared to be "no rhyme or reason to how cases get
assigned-workers with a lot of points will still get assigned cases-even when a different worker
has a lower number of points." Social workers noted that the point system accounts for the number
of children in a caseload and also assigns a higher number of points to FR cases than to Permanent
Placement (PP) and Family Maintenance (FM) cases. However, they point out that the point system
77
does not necessarily account for the severity of the case and for individual differences between
cases. "Even a case with low points can be very complicated and take up a lot of your time."
Social workers also highlighted the lack of adequate support staff to work effectively with
families, including help with transportation, investigating relative placements, and follow up with
clients at home as negatively impacting children and families of color. Social workers also
commented on the emotional strain on the job and risk for burnout that is exacerbated by too little
support.
We don't have enough support staff-not enough Social Worker J's. It's a lot harder to
manage a good FR case without a SW I to document things, help with transport, etc ... it
becomes more challenging.
This leads you on the road to burnout-I always want to do my best-I want to work harder
to get the parents to be successful. There are some workers who are tremendously burned
out.
Additionally, social workers noted that substantial amounts of required paperwork can
detract from building relationships with clients. Several social workers suggested that, as an
example, writing interim court reports in addition to required reports was duplicative and used time
that would be better spent with clients. Social workers also pointed out that not having easy access
to centralized and updated information about resources and services for families impedes effective
practice. The need for centralized and current resource information was noted as important to
avoiding "reinventing the wheel" in relation to researching resources that other workers already
knew about.
Moreover, confusion about the agency's overall mission, as well as confusion about key
agency and child welfare policies was also described as negatively impacting child welfare practices
for children and families of color. A need for greater clarity about the mission of DFCS was
expressed. "Who is the client, the child or the parent?" asked one participant. Participants varied in
their perspectives about the balance between child protection and family preservation. Social
workers also expressed confusion over certain key policies including concurrent planning, home
assessment protocols and protocols for voluntary versus court placements. Social workers suggested
that clearer guidelines related to implementation of concurrent planning were needed. Additionally,
participants reported that a more detailed written protocol on home assessments would be
beneficial, "the policy is not in writing, there is no written protocol and the most recent memo has
DRAFT written on it." Consequently, it is "difficult completing assessments and explaining the
policy to relatives, when the workers aren't clear about it themselves." Last, uncertainty about
protocols for voluntary versus court placements were noted. "There are misconceptions about
voluntarily placing a child and when that child is placed into protective custody... about whether
this protocol is required for all placements, it seems to go back and forth," commented one social
worker. Another noted "I don't have the understanding about what is a voluntary placement versus
a child taken into protective custody; it seems that ifwe get involved in any kind of removal that it
generates a protective custody."
A need for more bilingual and bicultural social workers and services was also noted as a
contextual factor that negatively impacts children and families of color. Although study participants
generally commented on the diversity of DFCS staff as a strength, there was a perceived need for
78
continued growth in this area. "There are not enough workers who speak the different languages of
the families," noted focus group participants. Some social workers suggested that "the clients are
left in limbo when a worker who speaks their language is not found" and "when a worker is found
that does speak the language of the client then two months have passed by; due to time line
constraints the worker is forced to make decisions without knowing much about the family."
"English speaking workers assigned to Spanish-speaking clients is a problem-even if the parents
and child speak English, a lot of the extended family may not," noted one worker. Reliance on
interpreters was identified as problematic because it can be difficult to find interpreters within a
limited amount of time which can lead to an inability of clients to obtain a timely answer to a
question from a worker because it may take days to find an interpreters. Moreover, there can be
problems with interpreters who are culturally inappropriate, who do not translate appropriately, or
who fail to read all of the court report to the clients.
In spite of these limitations, additional qualified interpreters and translators are needed to
facilitate communication with clients and to provide materials in client's primary language. Study
participants stressed the need for increasing language capacity both through social work staff and
through access to qualified interpreters. Access to qualified interpreters remains critical to working
in a region with representation of a growing number of cultural and language groups. "Lack of
interpreters causes continuance of cases," noted one social worker. Study participants offered a
number of illustrations of the impact that language barriers had on clients. One example is offered
below.
There are only two Korean interpreters in the South Bay. In one case with a Korean family,
Korean interpreter could not be found for a court hearing. This father had to take offfrom
work five days in a row to go to court, because each day an interpreter wasn't there. This
impacted the father because of the time he had to take offofwork. Each day he became
more and more frustrated. By the time the judge saw the case, the judge saw him at his
most frustrated point.
In addition to the need for bilingual and bicultural social workers, study participants also
expressed a need for more culturally competent services in general. "There is a need for more
culturally diverse counselors and therapists," noted study participants. Barriers to successful
compliance with mandated services include lack of counselors who have knowledge of different
cultures and long waiting lists for the few culturally diverse counselors in the area, according to
study participants. Furthermore, social worker participants observed that White families appeared to
attach less stigma and experience less resistance to accepting mental health or counseling services
than families of color. "Lack of services designed for the needs of a specific ethnicity or family can
cause Dependent Intake to hold onto cases longer due to being unable to find the necessary
resources." Some social workers suggested that existing services for English speaking clients varied
in quality and that contract agencies also differed in the degree to which they prioritized access to
services for DFCS clients.
A shortage of services for monolingual non-English speaking clients was identified as a
substantial problem for many families. "Families stay in the system longer because of lack of
Spanish speaking and other language services ... this can be frustrating for families, they want to
give up." Study participants described struggling tenaciously to find appropriate services for
families. Examples included searching for domestic violence and substance abuse services in
Tagolog, finding a Vietnamese-speaking counselor, and locating a group home for a deaf child.
79
Several participants expressed concerns that services provided in English for clients whose primary
language is non-English were ineffectual and, on occasion, harmful to families such as when
"English-speaking counselors who work with monolingual families will use the children as
translators." Spanish speaking services were perceived to be an issue throughout the county and
were particularly described as pressing by South County study participants. During one focus group
discussion of this issue a participant provided an example:
A Hispanic monolingual mother living in Sunnyvale does not have transportation [to
mandated services]. There are no Spanish resources in her area. They are all in the San
Jose/Santa Clara area, which causes the family to commute farther. It is difficult finding
resources in her area and with no bus passes transportation is a big issue.
Study participants also described time limit constraints as having a negative impact on
children and families of color. Focus group participants noted that it was often difficult to work
toward family reunification in the context of federal and state time limits. Participants uniformly
affirmed the importance of expediting placement of a child into a permanent, safe and stable home
and to seeking the least restrictive placement options for children. At the same time, they suggested
that timelines made it difficult to have enough time to work with families that require services in
order to safely reunify. "Fifteen months is not enough time to work with the client's
issues/problems. The reality of what people are dealing with is not there in the law," was a typical
observation by social workers. Several social workers and interviewees offered examples of
children who were placed into adoption who would otherwise have been able to remain with their
families if time limits were less rigid: "Court time limits played a large role in the adoption of the
child ... the time limits affected the outcome of the family and did not take into account how difficult
it is to change the way one had lead their life in a matter of a few short months."
The issue of access to timely services in the context of time limits was also perceived as an
important issue. An interviewee described a successful case in which a Latina client with a
substance abuse problem was able to reunify largely because the social worker was able to get the
mother into substance abuse treatment despite waiting lists: "There are waiting lists for services and
not enough time to successfully complete treatment with the waiting lists and the time limits ...This
family was lucky; the worker was able to get the mother into treatment within 2-3 weeks, so being
at the right place at the right time accounts for this [success]." Conflicts between child welfare time
lines were perceived as particularly common in relation to addressing the needs of substance
abusing parents that represent a majority of cases. "The Federal government only gives families in
permanent planning 12 months to complete services and parents only have 6 months for children
under 3 years of age; this is especially difficult for parents who abuse substances."
Study participants suggested that time limits disproportionately impacted families of color:
"Children of color are brought into the system because parents are on drugs,.they aren't being
watched and are in an unsafe environment at that point. The children stay in the system because the
expectation is for the parents to all of a sudden to get better because their children are in the system,
'so if you [parents] don't get better we [social workers] are going to keep your children. This is
unrealistic, especially with the court timelines," exemplified comments made by many study
participants. Families whose primary language is other than English were also perceived as being at
greater risk for failing to reunify: "It is difficult to meet timelines with certain services that are not
offered in different languages; clients are placed on waiting lists or those services don't exist."
80
Contextual Factors that Positively Impact Practices for Children and Families of Color
Respondents noted four overall contextual factors that positively impact child welfare
practices for children and families of color including I) cultural matching and cultural consultation,
2) supportive supervision 3) collaboration with other agencies and systems, and 4) a strength-based
approach to services.
Cultural matching and cultural consultation was noted as a best practice for children and
families of color. Although study participants did describe a need for more bilingual and bicultural
staff, they also commended the agency's efforts to provide a cultural match between clients and
social workers. Cultural matching was viewed as an effective strategy for facilitating development
of rapport with clients and minimizing the potential for cultural bias in assessment. Specifically,
social workers and supervisors valued the culturally specific Emergency Response (ER) units and
acknowledged efforts to match clients and social workers as much possible in Dependent Intake
(DI). At the same time, respondents asserted that it is not always possible to assign to clients
workers of a similar cultural background. In this context, social workers described obtaining
cultural consultation from other child welfare professionals as a best practice.
At the same time, study participants acknowledged that clients might prefer to interface with
social workers of a different culture than their own. For example, several social workers that
specialize in working with Asian American/Pacific Islander groups commented that some families
prefer to be assigned to a social worker of a different background out of a sense of privacy, pride, or
concern about judgment. Both social worker and client focus group participants noted that cultural
matching was not a guarantee against bias. Social workers are not immune to beingjudgmental
about clients of their own culture or biased in relation to other issues such as socioeconomic status.
In some cases, clients described their experiences with social workers from a different background
as positive and that the quality of respect was more important that sharing a common background.
To this end, continuing training on cultural competence and working effectively with specific
population groups was also considered a key practice.
High quality, supportive supervision that integrates coaching in working effectively with
diverse families was viewed as critical to practice with children and families of color. "Some
supervisors want to hear about cases and are supportive-others just want to know if you got the
reports done on time," summarized one participant. In addition, supervisors were identified as key
players in the process of brokering cultural consultations with other social workers, particularly
between social workers from different units.
Collaboration with other agencies and systems concerned with the safety and welfare of
children and families of color was also highlighted as a beneficial practice for children and families
of color. For instance, developing interagency formal agreements and connections with liaisons in
other agencies was described as beneficial because it facilitates collaboration and minimizes the
potential for agencies to give conflicting messages and mandates to clients. Social workers are "able
to divert cases or get a better picture of the family when two agencies are working together to work
with the families, which helps not only us but also the client." Other agencies mentioned in this
context include probation, hospital, medical providers, family and mental health agencies,
CalWORKS, substance abuse treatment agencies, domestic violence agencies and other entities that
may be involved with clients.
81
Additionally, collaboration with other agencies to address systemic issues was described as a
best practice. Social workers noted that community meetings are held on a monthly basis for all
community agencies and are attended by program managers, but workers are encouraged to attend.
Participation in policy level discussion between systems was considered to be an important part of
collaboration. Several participants mentioned the Greenbook project, an initiative to better
coordinate domestic violence and child welfare services, as a model for interdisciplinary problem
solving and policy development. The Greenbook project successes could help inform evolving
collaboration with other key systems, such as tqe substance abuse treatment system.
Other collaborative efforts such as outstationed social workers, co-location of services and
Family Drug Court were noted as best practices for children and families of color. Participants
described having ER social workers at different sites such as the Family Violence Center and police
stations as a valued practice. Moreover, South County social workers noted "we are located at the
resource centers where there are community providers here that are not child welfare ... we've got
AA meetings, parenting classes-everything is centralized here and that makes it comfortable for
the families." Although social workers mentioned that co-location of services have been reduced
because of budget changes and other factors, it remains a positive practice for children and families,
particularly low-income families and families of color that may have less time and fewer resources
to access services. Lastly, social workers and supervisors described the family drug court as a model
for effectively working with substance abusing families in the child welfare system. Several
examples of "successful" cases involved families assigned to family drug court.
A strength-based approach to services was also mentioned as an important best practice that can
be used at all points in the CWS. Indeed, strength-based practice was lauded as a fundamental tenet
of social work that is crucial for working with children and families of color. Specific elements of
this practice named by study participants include, looking at the unique characteristics of each
family; building solutions with families and showing that they [families] do have some control;
being willing to try new things, especially when clients make the suggestions; involving the parent
in the removal of the children, a strategy that is empowering to parents and less traumatic for the
children; demonstrating a commitment to invest time with families on the "front-end" to build
rapport, facilitate client understanding of the child welfare system, and optimize chances for
maintaining or reunifying the family; working with families to identify relative placements when
out of home placements are required and coaching relatives through the application process; and
lastly, demonstrating a willingness to advocate for the child and family with the court and other
service delivery systems.
Contextual Factors:
Recommendations to Improve Practices for Children and Families of Color
Study participants described five overall recommendations to improve practices for children
and families of color as they relate to contextual factors, including I) reducing caseloads, increasing
support and streamlining service delivery, 2) maintaining and expanding multifaceted approaches to
providing culturally competent practice, 3) increasing linguistic capacity of systems and services for
families, 4) fostering an organizational culture that is strength-based and community based, 5)
enhancing collaboration, and 6) training social workers. Specifically,
Reduce Caseloads, Increase Support and Streamline Service Delivery
82
•
•
•
•
•
•
•
•
•
•
Continue efforts to reduce social worker caseloads. Reduction in caseloads was uniformly
identified as key to providing quality services to clients, particularly children and families of
color that may have complex needs and that may initially have less trust and/or less familiarity
with child welfare and court systems. "Social workers need time to do their job and to help
clients access culturally appropriate services and programs."
Provide additional Social Worker I support for in-home assessment/support, follow up with
families, and overall case support.
Provide dedicated staff, and explore other options, to expedite requirements to qualify relative
placements. For example, policies requiring relative placement standards similar to foster
families might be reevaluated (e.g., illustrations of this as a problem included an example of two
young girls were denied relative placement because they shared a queen size bed rather than
having separate beds). In addition, develop strategies for overcoming barriers to relative
placement, such as an expedited process for Director's Exemption in the case of a distant
criminal history of a family member in the prospective home.
Examine opportunities to facilitate quicker access for social workers to criminal background
checks.
Develop procedure or protocols to facilitate communication between units, particularly in
relation to transferring ER cases to voluntary family maintenance informal supervision.
Examine options for maximizing the skills of social workers with specific cultural or language
expertise. For example, social workers in ER and DI stated that support across units that
involved overtime was recently prohibited. However, social workers suggested that this practice
failed to reduce the total amount of overtime expended by social workers while limiting
flexibility in obtaining culturally and linguistically skilled consultation and assistance.
Develop centralized, and updated resources lists for a wide range of services, including service
providers with expertise in serving diverse cultural and language groups. This practice was
identified as key to avoiding duplication of effort and ensuring access to quality services for
clients. "Sometimes we make a lot of phone calls just to find out what someone two aisles from
you already knew." In addition, update Court Services Handbook, which was perceived by
several social workers to be out of date.
Examine opportunities to provide flexibility in office hours, alternate social workers to cover
cases when social workers are on vacation, or other mechanisms to reduce potential burnout.
Provide social workers access to adequate transportation (e.g., "better cars") for travel related to
working with families.
Develop protocols that reduce paperwork so "there is more time to do social work." For
example, interim court reports could be eliminated or reduced, templates for psychosocial
assessments or other guidelines for court reports could be provided, and social workers could
"get assistance especially with filing."
Maintain and Expand Multifaceted Approach to Providing Culturally Competent Practice
•
•
Continue practice of attempting to provide a cultural match for clients as a strategy for
facilitating rapport and, in some cases, minimizing cultural or class bias.
At the same time, maintain recognition that 1) some clients may prefer to see a Social worker
who is not of their background, 2) many clients value "being treated with respect" above being
assigned a social worker of the same background, and 3) the diversity of the population in Santa
Clara County may not always allow for "matching" clients and social workers in relation to
race/ethnicity or other demographics. Ensure training and support for all social workers in
83
•
•
cultural competence and addressing the needs of different population groups among all social
workers
Continue working to ensure diverse staff in units throughout the agency.
Continue using the family resource centers, including provision of services for mandated clients.
Increase Linguistic Capacity of Systems and Services for Families
•
•
•
•
•
Facilitate development of services specifically designed for families whose primary language is
not English in key areas including mental health, domestic violence, substance abuse,
counseling services, parenting without violence classes, and other services.
Continue to hire bilingual/bicultural staff. Need for greater linguistic capacity was reported for
a number of different language groups including for Spanish speaking, Vietnamese, and Filipino
clients.
Minimize need for use of translators. When needed, provide high quality translators who are
able to communicate concepts related to the law, the social work system, timelines, and court
· processes.
Provide forms and written materials in multiple languages.
Examine opportunities for establishing guidelines for contracting or collaborating agencies in
relation to linguistic services, such as procedures disallowing use of children as translators for
monolingual speaking parents.
Foster an Organizational Culture that is Strength-Based and Community-Based
•
•
•
•
Continue to identify mechanisms, such as the Family-to-Family Initiative, to build community
involvement in decision-making.
Communicate a commitment to assuring the least restrictive placement for children from
management, through supervision, and to line staff. Provide training and support for social
workers to advocate for clients in court and other settings.
Examine strategies to minimize what study participants described as internal disincentives to
ensuring the least restrictive placement for children including "County Council whose role it is
to protect the agency sometimes telling social workers to change recommendations" and the
greater paperwork and advocacy required for less restrictive options.
Ensure client access to information about complaint procedures and access to the Ombudsman.
Enhance Collaboration
•
•
Continue to develop collaborative case planning, linkages, information sharing and policy
development with other service delivery systems, including schools, service providers,
probation, CalWORKS, housing, and other systems. Collaborative planning and problem
solving was perceived as particularly crucial in an environment of sever fiscal restraints.
Develop protocols to facilitate access to mandated services for clients in the context of federal
and state timelines, such as formal agreements and protocols to expedite access to substance
abuse treatment.
Train Social Workers
84
•
•
•
•
•
•
•
•
Require cultural competence training for all new staff and mandate continuing education for all
social workers. Include, as part of the training, opportunities for self-reflection and examination
of personal bias related to culture of other issues frequently confronted in families, such as
substance abuse. In these areas, provide training that is high quality, dynamic, and delivered by
skilled trainers who are knowledgeable about both their subject area and child welfare.
Provide education within the agency about different cultural norms in parenting and family
strengths.
Provide training on specific issues that impact all families and that may differentially impact
children and families of color including policies and practices for addressing drug exposed
infants, substance abuse, domestic violence, sexual abuse, and other topics. In addition, provide
training on procedures and practical considerations for working with diverse families, such as
training on guardianship.
Provide training on immigration law and practical strategies for working with families, such as
identification of immigration status of the child at an early stage and working with international
relative placement.
Training on these topic areas should integrate consideration of cultural differences and
evidence-based practices for different populations.
Provide training for all staff including clerical staff and eligibility workers.
Facilitate forums for social workers to share resources and effective practice strategies.
Create opportunities for social worker access to training. "We can't go to trainings because we
have too much else to do-so we're denied professional development opportunities."
Summary of Key Findings and Implications
Study participants described numerous practices and contextual factors that have either a
negative or positive impact on children and families ofcolor in Santa Clara County's child welfare
system. There is some evidence in the research literature that supports certain best practices
mentioned by study participants. For instance, general research supports the effectiveness of inhome services (see literature review), however these services appear to be most effective when they
are intensive, long-term and delivered by health or social service professionals and when the
workers delivering these services receive a high level of supervision (Barth, 1991; McGuigan et al.
2003; Olds et al., 1997). In addition, some research has suggested that parenting classes for child
welfare parents are more effective when they pair a didactic approach with a more experiential
approach, such as parent and child interaction (Carlo, 1993).
Participants also identified family conferencing as an effective practice. These study
findings are consistent with limited research in this area that suggests that inclusive practices, in
which parents are engaged in placement and other decisions, may result in more placement stability
for children, although possible differential effects for diverse populations has not been researched
(Palmer, 1996). Participants generally expressed optimism about the adoption of team decision
making (l'DM) as a vehicle for improving decision-making with diverse families, although some
social workers expressed concern about time required for implementation. Research does suggest
that TOM can be labor intensive and can be difficult to implement effectively within the context of
the child welfare system (Sieppert et al., 2000). Finally, participants noted that both formal and
informal collaborative relationships with other services delivery systems, such as domestic violence
and substance abuse treatment systems, are critical for successful interventions with children and
families. Furthermore, participants recognized the importance of training and cross-training to
85
enhance the capacity ofhelping professionals in child welfare and other systems to better address
the needs ofchildren and families of color.
Interestingly, most practices and contextual factors mentioned by social workers had both
positive and negative aspects. Many of the practices and contextual that negatively impact children
and families of color were actually barriers to successful implementation of the practices that social
workers felt have a positive impact on children and families of color. Although practices such as
orientation ofnew clients; family and team decision-making; preventive services; substance abuse
services; cultural competency; collaboration with other agencies and systems; a strength-based
approach; and the use of relative placements were described as having a positive impact on
children and families ofcolor, certain negative practices and contexts impeded their effective
implementation. For instance, having to screen out large numbers of inappropriate referrals;
inconsistency in decision-making practices; a shortage ofservices, particularly substance abuse
and preventive services; lack ofclient access to services; difficult protocols for placing children
with kin; gaps in cultural competency; time limits; and agency-level factors such as heavy
caseloads, staff shortages, substantial amounts ofpaperwork, lack ofaccess to information about
resources; and confusion about the agency's overall mission and key policies all impeded
implementation ofthe best practices. These findings suggest that in addition to building on current
positive practices for children and families of color, CWS stakeholders should also actively work
both internally and in collaboration with partners to reduce barriers to these best practices.
Indeed, in order to address these barriers to effective implementation of best practices, study
participants mentioned numerous recommendations that centered around certain key themes. In
general, there was an emphasis on improving decision-making through increased accountability,
training and cross-training to reduce bias, and increased use of group decision-making or family
involvement in decision-making. Study participants also stressed a need to improve and expand
prevention, diversion and concrete services, as well as an overall need to expand culturally and
linguistically competent services, and develop ways to improve availability and access to services.
Once a case is opened, social workers also felt it valuable to provide an orientation to clients so that
they are more aware of the CWS and court processes and can better navigate the system.
Recommendations related to organizational factors were also noted, including reducing caseloads;
increasing support and streamlining service delivery; clarification ofthe mission ofDFCS, as well
as key child welfare policies; fostering an organizational culture that is strength-based and
community-based, and more overall training for social workers.
Liniitations and Conclusions
Qualitative data, such as focus groups and interviews, provide rich insights into practices
that may positively or negatively impact children and families of color. In any such endeavor, there
are possible biases that may impact the results. For example, given the likelihood of time conflicts
and emergencies, we invited more social workers to participate in the focus groups than we
expected to attend. It is not possible to determine if other factors influenced participation. Workers
with a particular interest in this topic may have made a special effort to attend. In addition, the
perspectives and opinions of the social workers, supervisors and families who participated in the
study may not be entirely reflective of the populations they represented. At the same time,
participants were representative of different units and facets of the child welfare system, different
race/ethnic groups, and varied levels of experience. The findings of this study affirm the value of
many efforts that are already in place, such as culturally specific ER response units and family
86
resource centers, and other that are in the process of implementation including efforts to reduce
caseloads, initiation ofteam decision making, and participation in the Family-to-Family initiative.
The findings of the study also point to opportunities to further strengthen services for children and
families ofcolor. Recommendations related to training, institutionalizing group orientation for
clients, and other suggestions from study participants could be used to inform practice and planning.
It would be beneficial for a team ofmanagers, line staff, family representatives, and community
members to review these findings to identify and prioritize possible practices for adoption based on
their feasibility and utility for children and families ofcolor.
87
Comparison of the Main County Offices and South County Offices
In an effort to better understand the influence of contextual factors on child welfare practices
and outcomes for children and families of color, a comparison of Santa Clara County's Main
Offices and South County Offices was conducted. This comparison included both quantitative and
qualitative analyses. Quantitative data from CWS/CMS were used to examine differences in case
characteristics between the Main County Offices and South County Offices. A qualitative
comparison of focus group data from the Offices and South County Offices was also conducted to
discern how workers perceive practices in these two geographical locations. These analyses help
shed light on the potential impact of agency and community context on case characteristics, child
welfare practices and outcomes for children and families of color in Santa Clara County's CWS.
Quantitative Methods
In Phase 2, the CWRT continued its investigation of the disproportionate representation of
children of color in the child welfare system by analyzing a sample of cases closed during an 18month period (January 2000 through June 2001). Key case characteristics were analyzed, including
ethnic comparisons. Please see the Phase 2 report for methodology details and case characteristics
by ethnicity (Hines et al., 2002).
One of the recommendations that emerged from Phase 2 of the study was to explore possible
differences in case characteristics and practices between the Main County Offices and the South
County Offices. In Phase 3, we used the basic characteristics from the 1720 cases analyzed in Phase
2 while adding the Santa Clara County location information provided via the Child Welfare System
/ Computer Management System (CWS/CMS).
Analyses
A series of exploratory, bivariate analyses were conducted with county location as the
distinguishing variable, Main County (MC) versus South County (SC). The purpose of this initial
set of analyses was to identify differences in the CWS related to county location in order to help
describe variations not only by ethnicity, but by other case and service characteristics. The
following sections will describe (1) the proportion of cases with each county location, (2) general
characteristics by county location, and (3) significant differences related to county location.
Definition of Terms
County Location - Two overall geographical areas within Santa Clara County designated as either
Main County Offices (centered primarily in the Northern part of the County, including the main
office on Julian Street in San Jose) or South County Offices (based in Gilroy).
Quantitative Results
County Location
Of the 1720 cases, 1633 (94.9%) had valid county location information with 87 missing
(5.1%). Of the 1633 valid cases, 1487 (91.1%) were served in MC, 50 in SC (3.1%) and 96 (5.9%)
88
in both county locations. Please see Table 15.
Table 15: Proportion of Cases by County Locationa
Total
Main (MC)
8
1633
1487
8
(91.1%)
South (SC)
Both
96
(5.9%)
50
(3.1%)
Based on 1633 valid cases, excluding 87 cases with missing location
For the primary ethnic comparison, we retained three county location designations, MC, SC,
and Both (cases served in MC and SC). However, given the range in the length of time children
designated as Both were served in MC and SC, and the focus on comparing the two stated
geographical areas, subsequent analyses excluded this third category.
County Location and Child Characteristics
Child's Ethnicity
There was a significant relationship between the child's ethnicity and county location. As
noted in Phase 1 and 2, Latinos compose the largest ethnic group served in the county's CWS.
When examined by location, those served in MC were more likely to be Latinos (44. 7%, 661 of
1480 MC cases). This proportion of Latinos was higher in SC (68.0%, 34 of 50 SC cases) and Both
(62.l %, 59 of95 Both cases). African American, Asian American/Pacific Islander, and Other
children's cases were least likely located in SC and Both locations. Please see Table 16.
T able 16 Eh
. Group B;y C ounty Locat1on
t me
Total
Child's Ethnicitya
South
Both
1625
1480
50
95
•
N size
African American
•
177
(10.9%)
170
(11.5%)
1
(2.0%)
6
(6.3%)
White
•
476
(29.3%)
434
(29.3%)
29
(30.5%)
Latino
754
(46.4%)
661
(44.7%)
13
(26.0%)
34
(68.0%)
Asian American
/Pacific Islander
156
(9.6%)
155
(10.5%)
1
(2.0%)
0
(0.0%)
Other ethnicity
62
(3.8%)
60
(4.1 %)
1
(2.0%)
1
(1.1%)
•
•
•
a
Main
59
(62.1%)
Based on 1625 cases with valid MC and SC county locations and county location information
Unfortunately, the South County Office S1Jbsample was too small to allow more detailed analyses by
ethnicity. However, we continued to pursue our comparison of the two primary county locations by
other key variables.
89
Child Demographics
There was not a significant relationship between gender and county location. Within MC,
52.2% was female (776 of 1486 cases) compared to SC where 50.0% was female (25 of 50 cases).
Within SC, 47.8% was male (710 of 1486 cases) compared to SC where 50.0% was male (25 of50
cases).
There was not a significant relationship between age and county location. The average age
of the MC sample was 5.95 years (sd = 5.16), while the average age of the SC sample was 6.57
years (sd = 4.19).
There was not a significant relationship between child's language and county location. The
majority of the MC and SC samples spoke English. However, in SC, Spanish was the only other
language other than English spoken. Please see Table 17.
T able 17 Ch.ld
1 D emograp h.1cs
Gender
N Size
• Female
Total
Main
South
1536
1486
50
801
(52.1%)
776
(52.2%)
25
(50.0%)
735
(47.9%)
710
(47.8%)
25
(50.0%)
5.97
(5.13)
5.95
(5.16)
6.57
(4.19)
8
•
•
Male
Average age in years (sdt
Child's Languagec
•
•
N Size
1524
1474
50
English
•
•
1308
(85.8%)
1262
(85.6%)
46
(92.0%)
Asian American/
Pacific Islander
62
(4.1%)
62
(4.2%)
0
(0.0%)
Spanish
•
144
(9.4%)
140
(9.5%)
4
(8.0%)
Other
IO
10
(0.7%)
0
(0.0%)
•
(0.7%)
a Based on 1536 cases with valid MC and SC county location and gender information
b Based on 1537 cases with valid MC and SC county location and age information
c Based
on 1524 cases with valid MC and SC county location and language information.
90
County Location and System-Related Factors
Voluntary Status and Reason for Removal
There was a significant relationship between voluntary status and county location. The
proportion of cases in voluntary services was higher in SC at 88.0% (44 of 50 cases) compared to
41 % in MC (600 of 1462cases).
There was not a significant relationship between the reason for removal and county location.
Please see Table 18.
T able 18 Vo1untary status an dReason fior Remova
Total
8
Voluntary Status
1512
• N Size
Main
South
1462
50
•
Voluntary
•
644
(42.6%)
600
(41.0%)
44
(88.0%)
Not Voluntary
868
(57.4%)
862
(59.0%)
6
(12.0%)
1206
1196
10
Reason for Removal b
•
N Size
General Neglect
•
161
(13.3%)
158
(13.2%)
3
(30.0%)
Severe Neglect
•
172
(14.3%)
172
(14.4%)
0
(0.0%)
Physical Abuse
•
300
(24.9%)
297
(24.8%)
3
(30.0%)
Sexual Abuse
•
73
(6.1%)
73
(6.1%)
0
(0.0%)
Emotional Abuse
•
40
(3.3%)
40
(3.3%)
0
(0.0%)
Other Maltreatment
460
(38.1%)
456
(38.1%)
4
(40.0%)
•
a Based on 1512 cases with valid MC and SC county location and voluntary service information
bBased on 1206 cases with valid MC and SC county location and removal information
91
Initial Out-of-Home Placement Facility Type
There was not a significant relationship between initial out-of-home placement facility type
and county location. Please see Table 19.
Table 19: Initial Out-of-Home Placement Facility Type
Total
Main
South
Initial Out of Home Placement
Facility Type8
• N Size
1206
1196
10
•
Foster Family Home
•
258
(21.6%)
258
(21.6%)
0
(0.0%)
Group Home
•
66
(5.5%)
66
(5.5%)
0
(0.0%)
Children's Shelter/
359
(29.8%)
353
(29.5%)
6
(60.0%)
Relative Home
•
343
(28.4%)
341
(28.5%)
2
(20.0%)
Foster Family Agency
•
124
(10.3%)
123
(10.3%)
1
(10.0%)
Guardian Home/Court
Specified Home
56
(4.6%)
55
(4.6%)
I
•
a
Receiving Home/ Non-EA
/AFDC
(10.0%)
Based on 1206 cases with valid MC and SC county location, and placement infonnation
Number of Times removed from Family and
Number of Unique Placement Homes in Current Episode
There was not a significant relationship between the number of times a child was removed
from their family in current episode and county location. On average, children in the sample were
removed 1.35 times (sd = 0.70).
There was not a significant relationship between the number of unique placement homes in
current episode and county location. The sample had an average of3.56 (sd = 3.86) unique
placement homes in current episode. Although MC cases had a higher average of placements (3.57,
sd = 3.88) compared to SC (1.90, sd = 1.10), this difference did not reach statistical significance,
most likely due to the small number of SC cases available for comparison. Please see Table 20.
92
Table 20: Number ofTimes removed from Family and Number of Unique Placement Homes in
Current E.
;p1sode
Average Number of Times
Removed from Family in
South
Total
1.35
(0.70)
Main
1.35
(0.70)
(0.32)
3.56
(3.86)
3.57
(3.88)
1.90
(1.10)
LIO
Current Episode (sd?
Average Number of Unique
Placement Homes in Current
Episode (sdl
a Based 1228 cases with valid MC and SC county location and removal information
b
Based on 1228 cases with valid MC and SC county location and placement information
Average Stay (in days) per Placement Facility, Number of Placements, and Total Length of
Time in Out of Home placement in Current Episode
There was not a significant relationship between the average stay (in days) per placement
facility in current episode and county location. The average number of days a child stayed in one
placement in their current episode was 244.06 (sd = 480.43). Although MC cases had a higher
average of245.73 days in placement (sd = 482.03) compared to 41.54 days (sd = 44.71) in SC, this
difference did not reach statistical significance, again most likely due to the small number of SC
cases available for comparison.
There was not a significant relationship between the number of placements in episode and
county location. On average, children in the sample had 2.84 placements (sd = 2.95). Although MC
cases had a higher average of2.85 placements (sd = 2.95) compared to 1.80 (sd = 1.32) for SC, this
difference did not reach statistical significance, again most likely due to the small number of SC
cases available for comparison.
There was not a significant relationship between the total length of time (in months) in outof-home placement in current episode and county location. On average, children in the sample spent
13.16 months (sd = 20.40) in out-of-home placement. Although MC cases had a higher average of
13.24 months in placement (sd = 20.47) compared to 3.28 months (sd = 4.24) for SC, this difference
did not reach statistical significance, again most likely due to the small number of SC cases
available for comparison. Please see Table 21.
93
Table 21: Average Stay (in days) per Placement Facility, Number of Placements, and Total Length
. 0 Ut 0 fH ome p1acement .m Current E,p1so
. de
0 fT1mem
Average stay ( in days) per
Placement Facility Type in
Current Episode (sd)8
Number of placements in
Current Episode (sd/
Total length of time (in months)
in Out of Home Placement in
Total
Main
South
244.06
(480.43)
245.73
(482.03)
41.54
(44.71)
2.84
(2.95)
2.85
(2.95)
1.80
(1.32)
13.16
(20.40)
13.24
3.28
(4.24)
(20.47)
Current Episode (sd/
a Based on 1228 cases with valid MC and SC county locations and placement information
b
Based on 1206 cases with valid MC and SC county locations and placement information
on 1200 cases with valid MC and SC county locations and time information
c Based
Number of Episodes, Total Length of Case (in years), and Number of Workers Assigned to
Case over Time in the Current Episode
There was not a significant relationship between the number of episodes and county
location. Children in the sample had an average of 1.19 (sd = 0.47) episodes.
There was a significant relationship between the total length of the case (in years) and
county location. On average, children in the sample had a case length of 1.87 years (sd = 2.78). MC
cases lasted longer with an average of 1.93 years (sd = 2.80) compared to 0.18 years (sd = 0.39) for
SC.
There was a significant relationship between the total number ofworkers assigned to case
over time in current episode and county location. Children in the sample had an average of 4.91
workers (sd = 2.75). MC cases were assigned a higher average of 5.03 different workers (sd = 2.72)
compared to 1.48 workers (sd = 0.84) for SC. Please see Table 22.
Table 22: Number of Episodes, Total Length of Case (in years), and Number of Workers Assigned
t0 Case over T1mem
. th e CurrentE,PlSO
. de
South
Total
Main
1.16
1.19
1.19
Average Number of Episodes (sd)8
(0.48)
(0.37)
(0.47)
Average Total Length of Case (in
years) (sdl
1.87
(2.78)
1.93
(2.80)
0.18
(0.39)
Average Number of Workers
Assigned to Case over Time in the
Current Episode (sdt
4.91
(2.75)
5.03
(2.72)
1.48
(0.84)
Based on 1512 cases with valid MC and SC county locations and episode information
Based on 1512 cases with valid MC and SC county locations and case time information
c Based on 1532 cases with valid MC and SC county locations and worker information
a
b
94
Last Out-of-Home Placement Type
There was not a significant relationship between last placement type and county location.
Please see Table 23.
T a ble 23 Last Out-o f.H
- ome Placement Tvoe, and Age at Time of Case Closure
Total
Main
Last Out of Home
Placement Type8
• N Size
South
1228
1218
10
•
Foster Family Home
•
260
(21.2%)
260
(21.3%)
0
(0.0%)
Group Home
•
77
(6.3%)
77
(6.3%)
0
(0.0%)
Children's Shelter
•
354
(28.8%)
348
(28.6%)
6
(60.0%)
Relative Home
•
346
(28.2%)
344
(28.2%)
2
(20.0%)
Foster Family Agency
•
133
(10.8%)
132
(10.8%)
1
(10.0%)
Guardian Home/
Court Specified Home
58
(4.7%)
57
(4.7%)
1
(10.0%)
a Based on 1228 cases with.valid MC and SC county locations and placement information
Age at Time of Case Closure and Service Type at Case Closure
There was not a significant relationship between age at time of case closure for current
episode and county location. On average, children in the sample were 8.19 years-of-age (sd = 5. 72)
at the time of case closure. Although MC cases closed with an older child's average age of 8.23
years (sd = 5.76) compared to 7.15 years (sd = 4.22) for SC children, this difference did not reach
statistical significance.
There was a significant relationship between service type at closure ofcase and county
location. Overall, the majority of cases ended in family maintenance (59.9%, 920 of 1537 cases).
However, SC had a higher proportion of cases in family maintenance (92.0%, 46 of 50 cases)
compared to MC (58.8%, 874 of 1487 cases). Please see Table 24.
95
. T ype at Case Closure
T a ble 24 A.ge at T.1meofC ase Closure an dS erv1ce
Total
Main
Age at Time of Case
Closure for the Current
8.19
South
(5.72)
8.23
(5.76)
7.15
(4.22)
1537
1487
50
Episode (sd)8
Service Type at Case
Closureb
•
•
N Size
Emergency Response
•
39
(2.5%)
29
(2.6%)
0
(0.0%)
Family Maintenance
920
46
•
874
(58.8%)
Family Reunification
•
Permanent Placement
(59.9%)
64
(92.0%)
67
(4.4%)
(4.3%)
3
(6.0%)
511
(33.2%)
510
(34.3%)
1
(2.0%)
a Based on 1482 cases with valid MC and SC county locations and child's age at case closure information
b
Based on 1537 cases with valid MC and SC county locations. and service type information
Case Closure Type
There was a significant relationship between case closure type and county location. The
most common case closure type was court ordered termination and other (32.1 %, 482 of 1502
cases), followed by family maintenance (31.6%, 475 of 1502 cases). However, 62% (31 of 50
cases) of SC closed with family maintenance compared to 30.6% (444 of 1452 cases) in MC.
Please see Table 25.
96
T a ble 25 C ase Closure T ype
Case Closure Typea
Main
1452
South
50
•
•
N Size
Adoption
•
246
(16.4%)
246
(16.9%)
0
(0.0%)
Emancipation
93
(6.2%)
93
(6.4%)
0
(0.0%)
475
(31.6%)
444
(30.6%)
31
(62.0%)
99
(6.6%)
97
(6.7%)
2
(4.0%)
Reunified with Parent
or Guardian, Court or
non-Court Specified
61
(4.1%)
59
(4.1%)
2
(4.0%)
Incarceration,
Runaway or Medical
services
46
(3.1%)
45
(3.1%)
1
(2.0%)
Court Ordered
Termination and Other
482
(32.1%)
468
(32.2%)
14
(28.0%)
•
•
•
•
•
a
Total
1502
Family Stabilized
(FM)
Guardianship
Established or
Placement with
Relative
Based on 1502 cases with valid MC and SC county locations and case closure information
97
Qualitative Comparison of the Main County Offices and South County Offices
In addition to the quantitative comparison of the Main County Offices and South County
Offices, qualitative focus group data were collected to obtain information on how workers perceive
practices in the two geographical areas, and how these practices may differ.
Qualitative Methods
Prospective participants for the Social Worker focus groups were randomly selected from a
list of workers with title of Social Worker II or Social Worker III in different units/divisions. The
random selection and recruitment process is described in detail in the section describing qualitative
methods (please see page 61 for a full description of focus group qualitative methods). As described
above, a total of 67 Social Workers participated in the 13 focus group interviews. Of this number,
13 Social Workers from various units in South County Offices participated in three South County
focus groups. All the focus group participants in the South County sample were female. Of these, 2
(15.4%) identified as African Ameri_can, 1 identified as Asian American/Pacific Islander (7.7%), 4
identified as Latina (26.9%), and 6 identified as white/Caucasian (32.8%). A majority of the
participants (n=12, 92.3%) had masters level education, primarily MSW degrees (Please see
Attachment 2 for additional information about focus group demographics)
Procedures
Procedures for all focus groups are described in detail in the Qualitative Methods section of
this report. Procedures for all focus groups, including South County Office focus groups were
identical. As described above, ten focus groups were conducted at the main administrative office in
downtown San Jose and three were conducted at Gilroy Family Resource Center in South County.
Analysis
The process for analysis of the South County Offices data compared to other focus groups
was parallel to that described above in the Qualitative Methods section. The data for the qualitative
portion of this study consist of the transcribed notes taken during each of the focus groups and
individual interviews. Qualitative methods of data analysis were employed to analyze the focus
· group and interview data. First, major themes were summarized for each of the choice points or
units in the Social Worker focus group and supervisor interview data. A separate analysis of the
qualitative data from South County Offices was conducted to identify themes that may have been
different or more prominent in these series of 3 focus groups compared to the other 10 social worker
focus groups. The purpose of this additional analysis of the qualitative data was to obtain insights
and supplement quantitative analysis of possible differences between South County Offices and
Main County Offices within DFCS. In addition to gathering information about practices that may
differ between the Main County Offices and the South County Offices, findings from the qualitative
data also provide insights that may help explain quantitative differences in the relationship between
location and other variables, such as differences in the proportion of voluntary family maintenance
cases.
98
Qualitative Results
Many of the contextual and practice issues described by social workers and supervisors in
South County Offices paralleled comments by study participants in other regions of Santa Clara
County, such as concerns related to client poverty, housing shortages, and insufficient substance
abuse and other treatment resources, particularly for non-English speaking language groups. At the
same time, a few themes emerged that were either stronger or unique to South County Offices.
These are described below.
Context of Service Provision in South County Offices
Demographics of South County
According to study participants, South County has a large Latino community that is strongly
represented among child welfare clients. In relation to these demographics, participants emphasized
the need for bilingual and bicultural child welfare professionals. They also reported a paucity of
culturally and language specific counseling services for monolingual Spanish speaking participants.
In addition, access to counseling and services for undocumented families in the child welfare
system were perceived as a problem. Participants noted that one strength of the Family Resource
Center is the bilingual and bicultural social work staff. At the same time, participants saw a need to
expand capacity to work with other communities. "Our resource center is reflective of the majority
community ofLatinos, but some clients may not feel represented," explained one South County
focus group member. Specifically, several social workers suggested that the current and evolving
demographic changes in the region called for greater representation of African American and Asian
American/Pacific Islander social workers.
Small County Dynamic
South County Offices were perceived as different from "the main" office in part because
dynamics related to the smaller size of the region. Participants described feeling more connected to
the community, other service providers and clients because of the "a small-town atmosphere."
Participants noted that, "The community is smaller and there is more interaction within the
community and with individuals from other agencies; this tends to make people work with clients
differently." For example, one social worker commented that she often sees clients and colleagues
at the grocery store and others mentioned that clients freely come into the child welfare offices and
Family Resource Centers asking for assistance.
Furthermore, study participants in South County consistently reported a high level of both
formal and informal consultation with one another. In addition to relying on other workers and
supervisors to assess families and develop case plans, social workers working in close physical
proximity are aware of one another's cases and routinely exchange informal support and advice
about working with clients and accessing resources. Some participants noted that they might be
"less likely to connect" in larger facilities or offices "built like a maze." The absence of security for
entering the agency building "changes the atmosphere because clients can just walk in, there is an
open door policy; this changes agency attitude in working with clients and removes barriers for
clients."
99
Culture of Commitment to Maintaining and Reunifying Families
South County study participants described a shared philosophy that the vast majority of
children referred to the child welfare system can and should be with their birth parents, whether it
be through family maintenance or through family reunification. Specific values and norms
embedded in the philosophy and practice of the region include emphasis on using family strengths,
including families in the process of making decisions, and aggressively seeking alternatives to
removal. Study participants underscored the importance of this approach in working with children
and families of color. Although these principles are valued throughout the child welfare system,
participants in South County stated that these principles were infused throughout the system and
were used consistently to guide practice and social workers' accountability in assessment and
decision-making. For example, social workers described substantial involvement of supervisors and
coworkers when removal of children into protective custody is required. Study participant
comments suggested that the smaller size of the agency helps to solidify a sense of shared
philosophy and reinforce practice norms.
When asked to articulate how these values and norms are sustained and passed on, workers
cited several factors. First, study participants emphasized the importance of consistency in
messages, policies, and direction from the Program Manager and supervisors. Second, social
workers suggested that these expectations related to seeking the least restrictive placement for
children are communicated beginning with the initial hiring and orientation process through
everyday practice. "People who get hired have the same philosophy as South County ... and
supervisors ingrain values into workers," noted participants. Third, social workers stated that it was
normative for workers to obtain both formal and informal support when they are making decisions.
"Workers do a lot of staffing and hash things out; we don't make decisions alone," noted one
participant. "By using co-workers and supervisors in decision-making, you gain another
perspective," observed another. Integrating the perspectives of other workers in decision-making
was identified as crucial to mediating potential bias or misunderstanding related to cultural or
socioeconomic class differences. Finally, social workers suggested that they received substantial
support and consultation from supervisors in decision-making and that supervisor values and
practices were perceived as critical to the practices of line staff. For example, supervisor
involvement in decision-making was described as a positive practice, particularly in relation to
taking children into protective custody and out of home placement. "My supervisor was on the
phone the whole time I had to remove a child," was offered as an example from one social worker.
Such oversight was perceived as a sign of support rather than solely a mechanism to assure social
worker accountability.
Practices in Service Provision in South County Offices
Vertical Case Management
Focus group participants and supervisor interviewees named the vertical case management
model, in which social workers assigned to a case work with the family throughout their case rather
than transferring cases to a different unit, as an effective practice with clients, particularly with
children and families of color. Some study participants suggested that the vertical case management
model works well in a small community. "It works because it puts social workers in the community
with the family, making social workers accountable .. .it forces social workers to have to deal with
100
that family not just in a [case] process sense but in a real social work sense of developing
relationships," summarized one supervisor interviewee. Social Workers suggested that maintaining
continuity through a specific case enhanced trust with clients and created a greater sense of
accountability. The opportunity to develop and sustain rapport with specific clients was deemed to
be particularly important for marginalized groups, including children and families of color. Families
that return to the system are also generally reassigned to the same social worker. "In this case a new
worker may not have rapport or know family strengths ... A worker who has worked with a family in
the past knows them well enough to call them on their stuff." At the same time, respondents also
described vertical case management as work intensive and requiring additional support to ensure its
success, such as the help of Social Worker I staff.
Study participants in the focus groups that took place at the main administrative office in
San Jose location appeared to be familiar with the vertical case management model. Some
mentioned that it might be a possibility as a positive practice for children and families of color in
other parts of the county while others stressed the importance of maintaining specialized units with
staff who have expertise and an opportunity to stay current with changing laws, policies and
practices in their area.
Orientation of Clients
Social Workers in South County frequently refer parents to a client orientation that helps
parents understand the child welfare system and policies related to timelines for reunification. The
orientation is offered in English and in Spanish. It is offered at different times and provides 6 hours
of information over three sessions. Three areas are addressed in the orientation: timelines and facets
of the child welfare system (ER, DI, FR, FM, adoption), how to work with the social worker, and
resources for children and families. The orientation is not mandatory and is not offered at all sites.
One focus group participant examined outcomes (identified as either reunification or closure of
case) and found that the orientation appeared to be particularly helpful for families in the Spanish
speaking orientation. In sum, the orientation for clients was recognized as an important resource for
families, particularly families of color, who might feel confused or intimidated by the child welfare
system. At the same time, workers appeared to feel that there were opportunities to further
strengthen, institutionalize, and evaluate this practice.
Case Conferencing
Study participants stressed as a best practice the formal and informal process of meeting
with co-workers and supervisors to review cases and participate in the decision-making process.
Accountability and support in decision-making was viewed as an important factor in minimizing
potential cultural or other bias about clients and in maximizing creative thinking about how best to
work with children and families. Inclusion of families in decision-making was also noted as a
positive practice that was considered particularly important for children and families from cultures
that place a high value on extended family, including both biological and fictive kin. "If you have a
voluntary family maintenance case, you have a family conference," observed one social worker.
101
Summary of Key Findings and Implications:
Quantitative and Qualitative Comparison
of Main County Offices and South County Offices
Unfortunately, the South County location subsample was too small to allow analyses by
ethnic group. However, some key case characteristics are significantly related to county location.
Specifically, in South County Offices, a higher proportion ofcases are in voluntary services, cases
are shorter in duration, fewer workers are assigned during the course of the case, and a higher
proportion of cases are in family maintenance at case closure. These findings are also statistically
noteworthy given the small subsample size available from SC. There were also other differences in
case characteristics by county location that did not attain statistical significance but are noteworthy
for further investigation. Children served in the South County Offices appear to have a fewer
number of unique placement homes in the current episode, have a shorter average stay per
placement, a shorter length of time in out-ofhome placement, and are younger at time ofcase
closure.
These findings imply that there is a difference in the style and quality of services provided in
the South County Office location. Qualitative findings are congruent with the quantitative data and
provide some insight into the dynamics that may contribute to this difference. The higher number of
cases in voluntary services, the larger proportion ofcases in family maintenance at case closure,
and the shorter duration ofcases is consistent with what South County study participants described
as a culture of commitment to maintaining and reunifying families. This commitment was described
as integral to the philosophy of managers and social workers and reflected in the expectations
communicated to social workers by most supervisors and by peers. The practice ofproviding an
orientation to the child welfare system, time limits, the courts, and how to work with social workers
may also contribute to the differences found in the quantitative analysis. Study participants
frequently described South County as similar to a small county or rural area. This geographic
difference, and the resulting sense of "connection to the community, " was identified as a factor in
the greater sense of shared philosophy and emphasis on prevention ofout-of-home placements.
Other studies suggest that region may influence practice. For example, one study (Drake, 1996) also
found that rural areas are more likely to offer preventive services than are urban areas. The finding
that fewer workers are assigned during the course ofthe case is also consistent with descriptions of
the vertical case management model. This model calls for social workers to carry the same case
from case opening to case closure. assigned to a case works with that family from the initiation to
closure of the case.
Limitations and Recommendations
Only about 3% of cases were designated from SC, and the third location designation of
"both" was dropped. A key informant from South County explained that many of the cases
categorized as "both" may have been opened by ER workers from the Main County Office but
carried in South County prior to assignment of South County staff to a specific region of the county.
Similarities in demographics between the South County Office cases and the "both" categories
would appear to be congruent with this hypothesis. However, since this explanation could not be
verified in the data that was available, the cases designated as both were not combined with other
South County data. Future research should include a larger subsample from SC and add methods to
102
understand the reasons behind these location differences. Also, a larger South County Office
location sample may enable more detailed comparisons by ethnicity.
Some practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County Offices and adapted to
other regions ofSanta Clara County, as recommended by study participants. This practice may be
particularly helpful to low-income families and families of color that may be intimidated by or
unfamiliar with child welfare and related systems. Other practices, such as vertical case
management, may hold promise for other regions ofSanta Clara County, though the success ofthis
practice may be linked to the "small county" dynamic described by South County study
participants. Some participants in focus groups that took place at the main office mentioned this
model as promising while others asserted that specialized expertise in different areas of social work
was an advantage in providing quality services that are informed by "the most current information."
Further examination ofhow the culture ofcommitment to maintaining and reunifying families is
created, communicated, and continued and how this might be adapted to other parts of the county
warrants attention.
103
Multivariate Analysis of Factors Predicting Reunification
Preliminary Models
In order to identify system-related factors and case characteristics predictive of reunification
of children with their families, two exploratory models were tested using multivariate analyses. The
first was with a larger sample of 1720 closed cases, and the second with 403 cases from a case
record review sample. Both samples were collected during Phase 2 and additional data was
collected and analyzed during Phase 3.
Results: Closed Case Sample
We chose 8 demographic and system-related variables available through CWS/CMS to
predict reunification: child's gender, ethnicity, age at time of case opening, the number of workers
assigned across the history of the case, the length of the case, the number of unique placement
homes assigned, the number of times removed from the family, and the county location (Main
County Offices versus South County Offices).
For our analysis, due to missing data across our set of variables, 1190 cases out of 1720
were available for logistic regression. Our model was significant in predicting the reunification of a
child with the family (X2 = 176.27, df= 11, p = .001, Cox & Snell K = 0.14). Specifically, Asian
American/Pacific Islanders were less likely than Whites, African Americans, and Latinos to be
reunified, and the fewer number of workers assigned, shorter length of a case, and fewer number of
unique placement homes assigned were related to reunification (while controlling for each variable
within the model). Please see Table 26.
104
. Larger sampe
. Reum'fi1cation m
Table 26 Log1s
. f1c Regress ion Pre d'1ctmg
o e1
l Md
Model
Cox & Snell Ri
xi,
0.14
176.27
11
.001
df
p
Predictors
B and Significance
Gender (female baseline)
Ethnicity
African American vs White
Latino vs White
Asian vs White
Other vs White
Latino vs African American
Asian vs African American
Other vs African American
Asian vs Latino
Other vs Latino
Other vs Asian
Age at case opening
Worker Number
Time Length of Case
Number of Unique Placement Homes
Number of Removals
County Location (Main vs South)
Using two-tailed tests* p < .05
** p < .01
Model Coding: 1 = reunified, 0 = not reunified
Odds Ratio
0.15
1.16
0.19
0.22
-0.59*
-0.33
0.02
-0.78**
-0.53
-0.81 ***
-0.55
0.26
-0.01
-0.19***
-0.02***
-0.l 0**
-0.02
0.40
*** p < .001
1.21
1.24
0.55
0.72
1.02
0.46
0.59
0.45
0~58
1.30
0.99
0.82
0.98
0.91
0.98
1.49
Summary of Key Findings and Implications:
Closed Case Sample
Our model composed of primarily demographic and system-related variables to predict
reunification (child's gender, ethnicity, age at time of case opening, the number of workers assigned
across the history of the case, the length of the case, the number of unique placement homes
assigned, the number of times removed from the family, and the county location) was significant.
Specifically, four of these variables predicted reunification: child's ethnicity, number ofworkers
assigned throughout the case, length ofthe case, and number of unique placement homes.
Asian American/Pacific Islanders were less likely than Whites, African Americans, and
Latinos to be reunified with their families. Given that Asian American/Pacific Islanders are one of
the minority groups least acculturated to mainstream society, many of their cases might be
perceived as more severe because of the differences between traditional Asian and mainstream
American cultural attitudes toward child rearing and discipline. Culturally, many traditional Asian
105
families have a high level ofparental control and tend to use corporal punishment as part oftheir
parenting (Chao &Tseng, 2002). Many Asian parents may also be unaware about the laws defining
abuse in this society. In these cases, cultural sensitivity by investigators and caseworkers would
help determine if these Asian family cases are indeed severe or ifthey are over-interpreted as such
because of a philosophical difference between the professional and the family in parenting and
discipline. Future empirical studies could help determine the extent to which abuse severity or
cultural differences explain this relatively low reunification rate among Asian American families.
However, as Phase 2 results indicated Asian American/Pacific Islanders are also the most likely to
be placed in Voluntary Family Maintenance (Hines et al., 2002), the CWS appears to value the
preservation of Asian American/Pacific Islander families when first referred to the system (perhaps
recognizing the challenges of culturally appropriate services, language issues, and other
circumstances related to the Asian American/Pacific Islander community) but once in the system,
Asian American/Paci.fie Islander family cases are not easily resolved and circumstances may persist
that place the child at risk/or further abuse of neglect.
The fewer number ofworkers assigned was also related to reunification. This finding
indicates consistency in service and the relationship between the social worker and the family may
improve the chances of reunification. Perhaps this is due to greater familiarity and thus more
effective advocacy for the child and family. This finding could also be a function of particularly
challenging family cases already less likely to be reunified needing a greater set of workers over
time (e.g., due to worker burnout in the case, or specialization needed across the case).
Shorter case length was also predictive ofreunification. Presumably, less severe cases
could be resolved in a timely fashion, usually concluding with reunification. Following a similar
logic above regarding the number of workers, a longer case may imply more problems, thus
involving more time for resolution, but also possibly a case already protracted into a situation where
reunification is unhealthy for the child.
The fewer number of unique placement homes assigned was related to reunification.
Change in placements may be due to systemic conditions (i.e. logistics and short term availability of
space) but multiple placements can also indicate a persistent problem with a child adapting to a
placement and accepting care offered. As this may again indicate a more severe case, reunification
could also become less likely.
Due to missing data across our set of variables, 1190 cases out of 1720 were available for
this multivariate analysis. More complete information would have improved confidence in our
findings. Also, as we have made references to the possibility ofseverity ofa case influencing
reunification, better measurement and a more comprehensive evaluation offamily history, abuse
and neglect circumstances, and other indicators could produce a stronger predictive model. In the
future, we may be able to include more psychosocial and case variables in a larger sample to build
upon our current findings.
Results: Case Record Review Sample
To identify better the influence of case characteristics and indicators of success (discussed in
the Successful Cases Analysis, beginning on page 48) we added 6 variables to the model with the
larger sample collected through our in-depth case record reviews and utilized the additional
106
information coded for Phase 3: a family's enhanced capacity to provide for their children's needs,
children receiving appropriate educational services, children receiving adequate services to meet
physical and mental health needs, number of referrals, number of previous times in the CWS, and
assignment to Family Maintenance or Family Reunification services. The location variable was not
included in this model as only one case coded from the South County Office area.
For our analysis, again due to missing data across our set of variables, 246 out of 403 cases
were available for logistic regression. Our expanded model was also significant in predicting the
2
reunification of a child with the family (x2 = 63.87, df= 16, p = .001, Cox & Snell R = 0.23).
Similar to the larger sample's model, Asian American/Pacific Islanders were still less likely than
Whites, African Americans, and Latinos to be reunified. Also, shorter time length of a case again
predicted reunification. However, in this expanded model with success indicators, the number of
workers assigned and number of unique placement homes assigned were not statistically related to
reunification. The child's age at time of case opening and the number of unique placement homes
did approach significance. Interestingly, the variables related to success cases and outcomes
( discussed in the Successful Cases Analysis, beginning on page 48) were not significant. Please see
Table 27.
107
. Reum"fl1cat1on
. .m C ase R ecord R ev1ew
.
Sampe
T able 27 Log1st1c
. . Regress ion P red"1ctmg
1 Model
Model
Cox& Snell If
x2
0.23
63.87
16
.001
df
p
Predictors
B and Significance
Gender (female baseline)
Ethnicity
African American vs White
Latino vs White
Asian vs White
Other vs White
Latino vs African American
Asian vs African American
Other vs African American
Asian vs Latino
Other vs Latino
Other vs Asian
Age at case opening
Worker Number
Time Length of Case
Number of Unique Placement Homes
Number of Removals
Number of Referrals
Previous Times in the CWS
Case Assignment (FM vs FR)
Child Receiving Adequate Educational
Services
Child Receiving Adequate Services to
Meet Physical and Mental Health Needs
Family's Enhanced Capacity to Provide
for Child's Needs
Using two-tailed tests* p < .05
** p < .01
+ approached significance at p < .10
Model Coding: 1 = reunified, 0 = not reunified
Odds Ratio
0.29
1.34
0.43
-0.19
-1.63*
-0.36
-0.61
-2.06**
-0.79
-1.44*
-0.17
1.27
-0.06+
-0.06
-0.03***
-0.13+
-0.05
-0.04
0.12
-0.50
0.10
1.21
1.53
0.83
0.20
0.54
0.13
0.46
0.24
0.84
3.56
0.95
0.95
0.98
0.88
1.05
0.96
1.13
0.61
1.10
-0.18
0.84
-0.13
0.88
*** p < .001
108
Summary of Key Findings and Implications:
Case Record Review Sample
These findings indicate that particular factors are important to consider when predicting
reunification, specifically ethnicity and the length oftime a case remains open. Similar to the model
in the larger closed case sample, this implies that more severe cases (those needing more time for
resolution) may be more problematic, thus reducing the chances of reunification. We saw in the
larger sample model that number of workers and unique placement homes were related to
reunification, but the number of workers was no longer significant and number of placement homes
approached significance. In addition, the younger the child at case opening the more likely
reunification would occur. This could be that an effort is made to reunify younger children with
their families given that they are in an especially crucial stage of psychological and emotional
development related to the attachment with parents. These variables should be explored in more
detail.
The finding again that Asian American/Pacific Islanders are less likely to be reunified than
Whites, African Americans and Latinos, highlights a main ethnic difference. In Phase 2 we
discovered that Asian American/Pacific Islanders were more likely than the other ethnic groups to
be enrolled in Voluntary Family Maintenance services (Hines et al., 2002). However, our results
here imply that when Asian American/Pacific Islander children are removed from the home, their
children are less likely to be reunified This may also indicate extreme outcomes where Asian
American/Pacific Islander children are either initially diverted from the system or once in the
mainstream ofthe system are less likely to be reunited with their families.
It was interesting that the successful case and outcome variables were not predictive of
reunification. This could be due to the significant impact of other variables overshadowing the
influence of success indicators. It could also be due to our use of reunification as a dependent
variable. As we discussed in the Successful Cases Analysis (see page 48) reunification is not
necessarily synonymous with success, and thus these variables may indeed be unrelated.
Again, due to missing data across our set of variables, 246 cases out of 403 were available
for this multivariate analysis. Although only a preliminary set of models, these findings are useful
for identifying areas in the CWS and family circumstances that can be studied further, especially in
conjunction with the specific goal of reunifying children with their families. For future research,
reasons as to why major system-related factors are not predictive would also be valuable.
109
The Role of the Court System in Child Welfare Practice
Few studies have examined the potential influence of the court system on child welfare
practices and outcomes, however as noted in the literature review (Section III), the court system is
likely to have a significant impact on the trajectory of child welfare cases. In order to examine this
issue further, this analysis examines the role of the court system in child welfare practice.
Specifically, two overall issues are addressed. First, an analysis of quantitative case record review
data, and qualitative data from one focus group conducted with the Court Officer's Unit will
describe court ordered changes to social worker recommendations at the initial
jurisdictional/dispositional hearing; the types of changes the court orders; if these changes differ by
child's ethnicity; and possible explanations for these changes. Secondly, a qualitative analysis from
the agency-wide focus groups (described on page 61) will focus on themes related to the
relationship between the child welfare and court systems.
Quantitative and Qualitative Methods:
Analysis of Court-Ordered Changes
Quantitative Analysis
The quantitative analysis of the role of the court system in child welfare practice included a
new case record review of the same sample of 403 closed child welfare cases that were included in
Phase 2 of the research.
Sampling
Case record reviews include the same sample of 403 closed child welfare cases that was
analyzed in Phase 2 of the research. This sample was created from 6761 total case closures over an
18-month period, and a data file containing 1753 unique cases representing one child per family and
one case opening was constructed. This data file was used to obtain our target sample of 403 cases
selected randomly, guided by stratification according to ethnicity, age and service type. Of the 403
cases reviewed for this court analysis, 12 cases were closed, but not reviewed because they were
with a worker and not at the retention center; and 13 were not reviewed because they had been reopened.
Data Collection
Data collection took place at the Santa Clara County's Social Services Record Retention
Center. A Senior Research Assistant worked with the Retention Center Manager to ensure that the
sample of case records from Phase 2 of the research were re-pulled for review. Cases were reviewed
by three research assistants, two of these research assistants participated in the Phase 2 case record
review data collection and so had already been trained in case record review methods and one
research assistant was newly trained to complete the case record reviews.
Instrumentation
A data extraction form was created and pilot tested (please see Attachment 5 for a copy of
the Court Case Record Data Extraction Form). The form contained questions related to court
110
ordered service or visitation changes for the child, mother, father, alternate caregiver, or the whole
case.
Analysis
We addressed this aim through a 3-part series of exploratory descriptive and bivariate
analyses. First, we identified whether there were any court ordered changes and if court ordered
changes in general were related to ethnicity. A single change in recommended services for a case,
regardless of magn~tude, would classify the case as one with a "court ordered change."
Second, we examined major changes and if these were related to ethnicity. We considered
any court ordered change in the three main service areas (those concerning bypass, family
maintenance, and family reunification) as a major change.
Third, we developed 4 summary scores consisting of the number of changes imposed
separately concerning the child, mother, and father, and then for the entire case. Across each case
there were 85 variables over 9 potential areas where a change could be ordered in addition to the
three main service areas. These areas were psychological evaluation, counseling services, health
services, drug/alcohol testing, substance abuse/alcohol treatment, parenting education, support
groups, domestic violence treatment, and visitation. Changes in these areas were assessed and
summed into the 4 scores for court ordered changes aforementioned: child changes, mother
changes, father changes, and total changes. These scores were also examined for ethnic differences.
Definition of Terms
Court Ordered Change - Major changes were defined as court orders to initiate bypass,
family maintenance, or family reunification services for the mother, father, and/or the family. The
remaining changes included those across 9 areas represented by 85 variables: psychological
evaluation, counseling services, health services, drug/alcohol testing, substance abuse/alcohol
treatment, parenting education, support groups, domestic violence treatment, and visitation. A court
ordered change could also include placing more restrictions or conditions on service
recommendations.
Qualitative Analysis
A separate focus group with representatives from the Court Office Unit was conducted
toward the end of the study in order to obtain feedback on the implications and meanings of the
quantitative findings. The focus group with the Court Office Unit contained 9 participants,
including two Social Workers (22.2%) Ill's, five social work supervisors (55.6%), one Program
Manager (11.1 %) and one Legal Clerk (11.1 %). The mean number of years working in child
welfare among the Court Officer's focus group was 11.6 years and ranged from 4 to 22. Focus
group participants had a mean of 3 .2 years in their current positions, with a range of .5 to 7 years.
Seven participants had MSW's (77.8%), one had a BSW (11.1%), and one reported unspecified
postgraduate education (11. 1%). Two of the participants identified as Latino (22.2%), two identified
as Asian American/Pacific Islander (22.2%), four identified as White (44.4%), and one identified as
mixed race (11.1%). Eight of the participants were female (88.9%) and one was male (11.1%).
111
Participants in the Court Officer Unit focus group were asked a series of questions designed
to provide insights into the "story behind the statistics." Specifically, Court Officers were invited to
share their experiences in relation to overall factors that contribute to court ordered changes and
comment on specific findings from quantitative data. Please see Attachment 6 for the specific
questions used to guide the Court Officer Unit focus group discussion.
Notes from the Court Officer Unit focus group were transcribed and analyzed for key
themes related to court changes. In addition, focus group respondent interpretations and insights
about results of the quantitative data were examined and summarized.
Quantitative Results:
Analysis of Court Ordered Changes
Frequency of Court Ordered Changes
Ofthe 403 cases, 178 (44.2%) recorded at least one court ordered change from the original
social worker recommendations from the jurisdictional/dispositional hearing, 200 (49 .6%) abided
by the original recommendations, and 25 (6.2%) were not available for evaluating the additional
court ordered changes.
Court Ordered Changes and Child's Ethnicity
We eliminated the cases with missing information to do our comparison by ethnicity,
leaving us with 378 valid cases. There was no significant relationship between the existence of a
court ordered change and ethnicity. Please see Table 28.
t me Groupa
Table 28 Proport1ons ofC ourt 0 rderedCh ane;es b,y Eh.
White
Latino
Total
African
American
Court Change
Asian
American/PI
Other
No
200
(52.9%)
25
(53.2%)
65
(51.6%)
68
(51.9%)
18
(62.1%)
24
(53.3%)
Yes
178
(47.1%)
22
(46.8%)
61
(48.4%)
63
(48.1%)
11
(37.9%)
21
(46.7%)
a Based
on 378 cases with valid information
Types of Service Changes
We examined the types of court changes and distinguished three to be relatively major given
their impact on the case: bypass, family maintenance, and family reunification. Ofthe 178 cases,
the most common change with these three types was the implementation of family reunification
services: 27 cases (15.2%) involved family reunification, 2 (1.1%) cases regarded bypass, and 1
case (0.6%) regarded family maintenance. Ethnic comparisons were conducted, but no relationships
were found with these major services. Please see Table 29.
112
Table 29 Bypass, Family Maintenance, and Family Reunification Court Changes by Ethnic Groupa
Total
African
White
Latino
Asian
Other
American
American/PI
Bypass
No
176
(98.9%)
22
(100.0%)
61
(100.0%)
2
0
(0.0%)
(0.0%)
No
177
(99.4%)
22
(100.0%)
60
(98.4%)
Yes
1
(0.6%)
0
(0.0%)
(1.6%)
(0.0%)
(0.0%)
(84.8%)
151
16
(72.7%)
55
(90.2%)
52
(82.5%)
(90.9%)
27
(15.2%)
(27.3%)
Yes
(1.1%)
0
62
(98.4%)
11
(100.0%)
20
(95.2%)
(1.6%)
1
0
(0.0%)
(4.8%)
63
(100.0%)
11
(100.0%)
21
(100.0%)
0
0
(0.0%)
10
18
(85.7%)
I
3
(14.3%)
1
Family
Maintenance
1
0
Family
Reunification
No
Yes
a Based
6
6
(9.8%)
11
(17.5%)
(9.1%)
on 178 cases with court change information
Court Changes Regarding Children, Mothers, and Father
Four scores were calculated to describe the number of changes imposed by the court, each
for the child, mother, and father, and finally a total score representing the entire case. The number
of changes related to the child's circumstances ranged from 0 to 4 with an average of 0.31 (sd =
0.65). Court ordered changes related to the mother were most common, ranging from 0 to 8
changes with an average of 0. 77 (sd = 1.28). The number of changes related to the father's
circumstances ranged from Oto 7 with an average of 0.67 (sd = 1.26). The number of total court
ordered changes related to a case ranged from 1 to 16 with an average of2.53 (sd = 2.22). No
significant ethnic differences were found regarding the number of court-imposed changes. Please
see Table 30.
113
Table 30: Number of Court Changes bv Ethnicity'a
White
Total
African
American
Child Related
Changes
Avg.
0.31
0.38
0.23
(0.78)
(sd)
(0.65)
(0.53)
Mother
Related
Changes
Avg.
(sd)
Latino
Asian
American/PI
Other
0.27
(0.54)
0.36
(0.50)
0.33
(0.73)
0.77
(1.28)
1.32
(1.94)
0.69
(0.99)
0.83
(1.30)
0.18
(0.40)
0.57
(1.29)
Father Related
Changes
Avg.
(sd)
0.67
(1.26)
0.82
(1.92)
0.61
(1.14)
0.62
(1.10)
0.55
(1.21)
0.90
(1.30)
Total Case
Changes
Avg.
(sd)
2.53
(2.22)
3.36
(3.36)
2.39
(1.63)
2.54
(2.39)
1.82
(1.33)
2.43
(2.01)
a Based
on 178 cases with court change information
Qualitative Results:
Analysis of Court Ordered Changes
Quantitative findings were presented to a focus group made up of 9 representatives of the
Agency's Court Officers Unit for discussion and feedback. Quantitative results suggested that
approximately halfofthe cases had indicated a court ordered change from the recommendations of
the social worker described from the jurisdictional/dispositional hearing. From our focus group
with members from the Court Officer Unit, the consensus was that the interpretation of the
measured proportion of court ordered changes found in the quantitative analysis, depended on what
we were actually measuring. The proportion is dependent on a variety of circumstances as to why
there was a change from the social worker recommendations including clerical errors needing
correction, changes due to modifications in the law, simple (albeit common) adjustments to services
such as visitation terms in addition to actual substantial changes due to conflicting opinions
concerning the direction of the case could all be related to a change between social worker
recommendations and court ordered.
Findings from the focus group with the Court Officer Unit provide additional information
regarding changes made to social worker recommendations at the jurisdictional/dispositional
hearing. First, capturing the details needed to conclude why social worker recommendations may
or may not be followed is extremely difficult. This is because variations by judge, courtroom
114
environment, social worker credibility and reputation, assigned lawyer (i.e., court appointed vs.
privately retained), as well as formal and informal case plan negotiation tactics frequently occurring
would complicate research. Second, the personal dynamics among the stakeholders (socials
workers, lawyers, judges, and families), including the strength of their positions and philosophical
standpoint can be such an influence that systemic conditions may be poor predictors of case
outcomes and success as they are related to the judicial process. Third, that despite the complex
circumstances involved, judges, guided by social workers, determine the best possible service plan,
while being aware of interpersonal dynamics and individual characteristics that may obscure
reasonable decisions. According to the focus group, social workers have a significant influence in
the court process and usually receive a majority of services they request for their clients.
Quantitative data indicated that few changes involved bypass, family maintenance, or family
reunification. The proportion of changes in these main service areas seemed reasonable to the focus
group. The low percentage of change in bypass, FM, and FR assignment should be low given the
legal criteria to qualify for such a change, as well as the intricate legal procedures.
When presented with quantitative findings that indicated no relationship between ethnicity
and court ordered changes in major service areas, or the overall number ofcourt-ordered changes,
it was the opinion of members of the focus group that ethnicity does not contribute to decisions in
court, but rather it is circumstances that drive the case.
Qualitative Methods:
Relationship Between the Child Welfare and Court Systems
Qualitative results describing the relationship between the child welfare and court systems
are gleanedfrom the agency-wide focus groups. These focus groups were structured to explore
practices at different choice points in the child welfare system. Please see page 61 for a full
description of agency-wide focus group methods and results.
Qualitative Results:
Relationship Between the Child Welfare and Court Systems
Results from agency-wide focus groups (please see page 61 for a.full description) suggest
that the relationship between DFCS workers and district attorneys and judges is problematic at
times. Many participants described difficulties in effective collaboration between the court system
and the child welfare system. Social workers suggested that problems between the court system and
the child welfare system may reflect fundamentally different perspectives on the needs and
circumstances of children and families in the CWS, and perhaps unrealistic expectations regarding
service plans. For instance, one social worker noted, "From a client's side, they are being told by a
court system and a social system that the way that they were brought up and the way that things
were done in their [clients'] family is wrong and not acceptable," noted a social worker focus group
participant. Several participants suggested that courts directly and indirectly set an unrealistic
standard for families, exemplified by one participant's comments:
"The court system just makes these orders and that's it. They don't consider anything else
except for this narrow focus on child and parent. The social worker has to look at resources
for the child, for the family, and language barriers. The court doesn't consider the impact it
115
has on the family, who has to pay for the services, shrinking or expanding resources. The
court is in an isolated bubble anything outside that bubble is not in their framework. "
Regarding the services ordered, another social worker commented, the "Court wants it all
done at once ...the court doesn't understand that for whatever reason the client may be only capable
of doing so much within the first 6 months." Additionally once the court does order a service or
case plan, social workers report that it is difficult to adjust mandated services "because the court
requires justification for removing the order; also it is difficult convincing the DA of [the need for]
new recommendations."
In general, focus group and interview participants expressed a need to bridge these gaps
between the court system and the child welfare system. Although some social workers described
having strong collaborative relationships with the court system, many described tensions with
attorneys and other court representatives in relation to recommendations for mandated services and
disposition of cases. Social workers also conveyed that it was common for court representatives to
dismiss their psychosocial family assessments and push for recommendations that the social
workers deemed disconnected from the reality and needs of families. "Legal aids rarely visit the
children or the parents, they need to be more reality-based," commented on worker. "Social workers
and the court system are both looking at the case from different views," noted another. Some
workers depicted the district attorneys as having more of a focus on "winning" a case than on
ensuring the best outcomes for the children and family.
Working with the DA is a challenge. Their role is to advocate for the child-but they often
do only one interview and they don't know the particulars of the case. There can be
antagonism between the social worker and the DA-there should be less antagonism
between the DA and the social worker and more cooperation.
Social workers also expressed concern about the fairness of the system. For example, "In court
cases, private attorneys are more likely to have their client's case dismissed compared to court
appointed attorneys." Many participants conveyed that many court representatives and social
workers exhibit a "gap between standards that common people have and that educated professional
people have," which may disadvantage low-income families and families of color. "Attorneys look
at the information provided by the social worker and make sweeping generalizations and judgments
about families," summarized one social worker. One supervisor captured concerns about failing to
recognize the context of client's lives:
When there is child abuse or neglect we raise our eyebrows to it, we go in there and want
everything to change but we 're not providing a basic foundation for change. We 're not
developing the community, we 're not providing better housing. We tell people to go out and
get suitable housing, but we 're not providing it.
Language issues also impact the experience offamilies in the court. "Families are unable to
communicate with their lawyers because of language barriers; English only attorneys can't
communicate with a monolingual Spanish speaking client," summarized one worker. Other workers
observed that "Families may not like to leave a message for an attorney because of the language
factor" and "the court process can be confusing for English speaking families, but more so to
Spanish speaking families." Furthermore, "the court reports are all in English and the clients don't
have a translated version that they can look at."
116
Additionally, social workers expressed concern that client outcomes are often dependent on
working relationships or attitudes ofDA 's. "There are some DA's that I have a great rapport
with-but a few are very judgmental and I feel like they are not supportive of the family." The
DA's are perceived as having substantial power. "The DA's word is very strong in court because
they are supposed to be speaking on behalf of the child," stated one social worker. Other social
workers suggested that some DA' s pressured social workers to change recommendations or would
attempt to discredit them in the courtroom, particularly when the DA's case was weak.
Agency-wide focus group results also provided recommendations regarding ways to enhance
collaborative relationships between the court system and the child welfare system. These
recommendations included,
•
•
•
•
•
Explore opportunities to develop standards for practice between the courts and the child welfare
system including development of shared policies, practices and procedures. Such efforts could
help address common concerns of social workers that generally called for a "switch from courtcentered practice to family/community-centered practice."
Develop a systems level mechanism for problem solving to address conflicts between social
workers and DA' s and "set a minimum standard or tone for what is appropriate behavior in the
court."
Facilitate development of role clarity between child welfare and the courts to address
widespread concern that decision-making is often not based on the social worker assessments
and, instead, is often based on perceptions ofDA's who may not have spent sufficient time with
children and who may not have evaluated family strengths.
Facilitate a dialog with both court and child welfare workers about standards for "good enough"
parenting and the dynamics ofpoverty in decision-making.
Explore opportunities for support of social workers in the court process through training,
enhanced supervision, or coaching from experienced colleagues.
Summary of Key Findings and Implications:
The Role of the Court System in Child Welfare Practice
In Phase 3, we examined the role of the court system in child welfare practice by exploring
the types of court ordered changes that are made to social worker recommendations at the initial
jurisdictional/dispositional hearing, as well as the relationship between the child welfare and court
systems. Quantitative results indicated that the court made changes to the initial social worker
recommendations at the jurisdictional/dispositional hearing in approximately halfofthe cases.
Ethnicity was not related to whether or not a change occurred. These results do not necessarily
indicate that the system and courts treat children and families from different ethnicities the same,
but rather that when the judge decides to make a change from the social worker recommendations,
these changes appear to occur equally across ethnic groups. It was the opinion of members of the
Court Officer Unit focus group that ethnicity does not contribute to decisions in court, but rather it
is circumstances that drive the case.
Findings from the focus group with the Court Officer Unit suggest that many factors
complicate an accurate measurement ofwhy social worker recommendations may or may not be
117
followed. Factors that may impede accurate assessment of the agreement between child welfare and
court personnel on service plan recommendations include the strong influence of individual
stakeholders involved, and formal and informal case plan negotiation tactics all of which paint a
much more complicated picture.
Qualitative findings from the agency-wide focus groups suggested that collaboration
between the child welfare and court systems is problematic. Agency-wide focus group participants
felt that the child welfare and court system have different perspectives on the needs and
circumstances of children and families in the CWS and that the court system may have unrealistic
expectations of families.
Additionally, many agency-wide focus group participants described the ways in which child
welfare and district attorneys and judges interact as ineffective. Some social workers felt that some
district attorneys and judges would dismiss their assessments and recommendations and may try to
pressure social workers to change their recommendations. A somewhat similar finding was noted by
Knepper and Barton, (1997) who found that although judges tended to accept social worker
recommendations, the relationship between social workers and the court system plays a significant
role in their decisions. Their study found that when social workers adhered to the group norms of
the court, judges rewarded them by not ordering "unrealistic" practices and allowing reasonable
time frames for mental health evaluations to be completed. Similarly, agency-wide focus group
findings seem to suggest that if social workers go against the implicit rules ofthe court system, then
judges may override their recommendations, but if they adhere to expectations ofthe court system
than recommendations are accepted.
Limitations and Recommendations
From these exploratory findings, issues arose regarding limitations and recommendations.
First, although quantitative and qualitative information described many of the circumstances behind
why social worker recommendations are not upheld by the court, it is very difficult to ascertain the
degree to which each circumstance influences a court ordered change. Second, better and more
exhaustive record keeping may not be the solution, as many informal and undocumented actions in
the negotiation process occur regularly, and are confidential, thus making much of the data
unavailable for research and evaluation purposes. Third, assessing other factors such as courtroom
environment, stakeholder characteristics, and other judicial issues would be advantageous in later
research. Although obtaining this information is a formidable task, these factors should be included
in the future, as measuring primary systemic factors no longer seem adequate to explain outcomes
in the CWS.
118
V.
STATEWIDE COMPARATIVE ANALYSIS
The statewide comparative analysis of effective practices for children and families of color
in the child welfare system included a statewide survey of child welfare directors, managers or
supervisors in California counties who were asked to describe effective practices for children and
families of color in their child welfare systems. In addition, this statewide comparative analysis,
included a quantitative analysis that identified the link between these practices and levels of
disproportionate representation of children of color in county child welfare systems. As such, two
overall analyses were included in the statewide comparative analysis of effective practices for
children and families of color in the child welfare system, 1) descriptive information from the
statewide survey on the types of practices, practices and practice characteristics identified by survey
respondents, and 2) multivariate results that link types of practices, practices and practice
characteristics to over or underrepresentation of children of color within counties. Additionally,
contextual information on the estimated proportions of children of color in county populations and
county child welfare systems within selected regions of California is provided, as well as
information on the estimated proportion of children within the overall county population who are in
the CWS.
Methods
The statewide survey on effective practices for children and families of color in the· child
welfare system was designed to gather information on 1) key practices in California counties that
may have an impact on children and families of color, and 2) practices considered most promising
for children and families of color. For the purposes of the survey, "practices" were defined broadly
to include programs, services, strategies, policies, or tools.
Data Collection
Child welfare directors, managers or supervisors were mailed an invitation to participate in
the survey, and a copy of the survey. Two weeks later, a follow-up email that contained the survey
as an attachment and telephone calls were made to encourage participation. Counties that failed to
respond to the survey received approximately 3 additional follow-up emails or telephone calls
requesting their participation or a referral to someone else in the agency who might be able to
participate.
Respondents were offered four options for returning the survey to the CWRT, either by 1)
completing the survey on their computer and emailing it as an attachment, 2) completing the survey
over the telephone with a member of the CWRT, 3) faxing the completed survey, or 4) mailing the
completed survey. Surveys that were completed over the telephone were tape recorded in order to
ensure the accuracy of information and once information was verified, the tapes were erased. Fortyfive counties responded to the survey, resulting in a 77.6% response rate. Of the 45 counties who
completed the survey, 23 returned it via email, 10 by telephone, 7 by fax, and 5 returned the survey
by mail.
119
Sample
The sample included child welfare directors, managers or supervisors from California
counties. Participants for the statewide survey were identified through the 2003 County Welfare and
Social Service Directory. An invitation to participate and a copy of the survey were mailed to each
child welfare director, manager or supervisor in each of the 58 counties in California. Of the 13
counties that declined to participate, 4 indicated that their child welfare population is not diverse
and they did not have any special programs for children and families of color; 3 replied that the
survey was too time consuming, 3 did not reply to the survey, 2 indicated that they would complete
the survey, but did not return it and one county indicated that they were going through a reorganization and could not complete the survey because of these changes.
Instrumentation
A survey entitled "Statewide Survey on Effective Practices for Children and Families of
Color in the Child Welfare System" was developed and tested in a pilot study involving five
counties (please see Attachment 7 for a copy of the Statewide Survey). Feedback provided from the
pilot counties resulted in a reduction in the number of questions and minor changes in wording. The
survey contains questions related to caseload characteristics of counties and the most effective
practices for children and families of color in each county's child welfare system. The survey asks
respondents to provide demographic information on children in various components of the county's
child welfare system. Respondents are also asked to identify up to five of the most effective
practices for children and families of color in their county's child welfare system, ranking them
from #1 (most effective), #2 (second most effective) and so on. Multiple choice questions on the
most effective practice (#1) are also included in the survey.
Analysis
Quantitative analyses were conducted using the Statistical Package for the Social Sciences
(SPSS) software. Descriptive results were generated by running frequencies and percentages for
variables of interest. Multivariate results were generated by conducting controlled tests of empirical
associations between types of practices, practices, and practice characteristics and levels of
disproportionate representation of children of color in county CWS's.
Qualitative data describing the most effective practices were analyzed through content
analysis methods. The types of practices within each of the four categories were identified and
descriptive information about their target population, description and reasons why these practices
are effective with children and families of color were delineated.
Contextual Tables:
County Child Populations and CWS Involvement
The following tables provide contextual information on selected county child populations
and CWS involvement by ethnic group. Please note that the counties presented in these tables are
not necessarily the counties who responded to the statewide survey. Also please note that because
the most recent information available on county child populations is for the year 2001, information
on CWS populations for the year 2001 is also provided. As such, current 2003 proportions of
120
children of color in California county CWS may vary slightly from the 2001 proportions provided
in the following tables.
This descriptive information is presented in order to provide a context through which to
view the results from the statewide survey. Because compiling information on all 58 California
counties would be cumbersome, information on selected counties is provided. Table 3 la provides
information on the estimated proportions of children of color in county populations and the CWS
within 13 Bay Area counties. Table 31 b described estimated proportions of children of color in
county populations and the CWS within selected counties by California regions. Last, Table 31 c
provides information on the estimated proportion of children within the population in the CWS
among selected counties by region.
121
. and Ch.Id
l . Popu Iat1on
We lfiare Siystem It lil ay
Table 31 a: Est1mate
.
l reno fC oorm
l
dP roport1ons o fCh.ld
Ethnic Group
Asian
Latino
County
African American
White
%in
%in
%in
%in
%in
%in
%in
Population
Population
Population
Population
24.6%
11.9%
22.3%
66.1%
33.7%
18.0%
19.0%
Alameda
20.2%
12.7%
45.2%
13.9%
38.5%
11.4%
54.1%
Contra Costa
18.7%
9.4%
6.6%
71.3%
3.2%
19.4%
Marin
35.7%
1.4%
18.5%
67.3%
62.1%
0.7%
4.9%
24.7%
Mendocino
14.2%
52.3%
5.8%
4.3%
57.6%
Monterey
32.0%
30.1%
11.8%
32.3%
17.4%
4.3%
Napa
61.2%
1.5%
67.4%
53.1%
60.7%
2.1%
32.1%
1.2%
43.9%
0.5%
San Benito
22.2%
12.3%
37.0%
27.6%
11.6%
12.9%
70.4%
San Francisco
32.0%
29.8%
33.0%
23.9%
38.5%
26.7%
4.3%
San Mateo
12.2%
32.8%
52.0%
28.1%
3.5%
Santa Clara
35.4%
27.4%
41.0%
49.5%
3.8%
53.8%
46.1%
1.1%
3.4%
Santa Cruz
42.4%
9.1%
16.0%
15.0%
18.9%
49.5%
41.2%
Solano
4.5%
2.0%
21.8%
28.8%
70.6%
57.4%
9.0%
Sonoma
cws
cws
State of
California
37.7%
30.3%
7.1%
31.3%
cws
42.9%
33.8%
11.7%
rea ount1es
%in
cws
2.9%
2.0%
3.3%
0.2%
1.7%
2.8%
1.2%
4.8%
9.4%
7.4%
0.2%
3.7%
1.0%
2.8%
Native American
% in
%in
Population
0.4%
0.7%
0.4%
0.5%
0.2%
0.3%
6.0%
12.9%
0.2%
0.4%
0.0%
0.6%
0.4%
1.2%
0.2%
0.8%
0.3%
0.1%
0.2%
0.8%
0.3%
0.2%
0.6%
0.6%
1.0%
3.2%
cws
0.5%
1.1%
Notes:
1. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in Januacy 2001 according to Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A.,
Lecy, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. & Kim, H. (2003). Child Welfare Services Reports for California. Retrieved [8/29/2003], from University of California at Berkeley Center for Social
Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>.
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org.
3. The Bay Area Counties are the 13 identified by the CWDA, and are also categorized as "big" counties in California except San Benito which is "small."
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California County Data Book summarized information on children aged 0-17
whereas the CWS data may include children as old as 19. Some totals may not equal 100% due to missing data or rounding.
122
• and Ch"ld
.
. Popu Iat1on
Ta ble 31b Est1mate
.
We lfiare Siystem w·Ith"m SI
I reno fC o lorm
I
e ectedC ount1es B,y Reg1on
dP roport1ons o fCh"ld
County
Bay Area
Alameda
San Francisco
Santa Clara
Mountain
Calaveras 5
Placer
Sacramento
Northern
Butte
Humboldt
Lassen5
Southern
Los Angeles
Orange
San Diego
Santa Barbara
Vallev
Fresno
Merced
Stanislaus
State of California
White
% in
African American
% in
% in
Population
Ethnic Group
Latino
%in
%in
Population
Asian
%in
%in
Population
Native American
%in
% inCWS
Population
%in
Population
cws
33.7%
27.6%
35.4%
18.0%
11.6%
27.4%
19.0%
12.9%
3.5%
66.1%
70.4%
12.2%
24.6%
22.2%
32.8%
11.9%
12.3%
52.0%
22.3%
37.0%
28.1%
2.9%
4.8%
7.4%
0.4%
0.2%
·0.2%
0.7%
0.8%
0.8%
86.3%
83.3%
53.7%
91.4%
79.0%
43.4%
0.4%
0.7%
12.5%
0.8%
3.6%
35.2%
10.1%
12.0%
17.9%
4.7%
12.0%
17.1%
0.8%
3.2%
14.9%
0.0%
1.8%
3.1%
2.3%
0.8%
1.0%
0.8%
2~8%
1.0%
75.1%
78.4%
79.3%
73.9%
65.6%
61.0%
1.8%
1.0%
1.1%
8.9%
2.3%
1.2%
13.8%
8.0%
15.0%
7.9%
5.1%
26.8%
7.5%
4.3%
1.1%
1.3%
0.3%
0.0%
1.7%
8.2%
3.5%
6.8%
22.4%
8.5%
19.8%
40.9%
46.0%
42.2%
15.2%
39.1%
35.1%
41.5%
9.0%
1.8%
7.1%
2.4%
42.3%
6.8%
24.5%
15.1%
60.1%
43.3%
36.6%
50.2%
38.9%
48.0%
35.0%
42.2%
10.8%
13.7%
9.9%
4.8%
2.9%
5.3%
3.0%
1.0%
0.2%
0.2%
0.5%
0.4%
0.4%
0.2%
2.2%
0.2%
31.1%
35.9%
53.4%
28.0%
35.0%
50.3%
5.7%
3.9%
2.3%
22.3%
13.9%
13.0%
48.5%
44.9%
35.1%
45.9%
44.0%
32.4%
13.8%
15.0%
8.4%
2.1%
6.4%
2.6%
0.8%
0.4%
0.8%
1.7%
0.1%
0.5%
37.7%
30.3%
7.1%
31.3%
42.9%
33.8%
11.7%
2.8%
0.5%
1.1%
cws
cws
cws
Notes:
I. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in January 2001 according to Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A.,
Lery, 8., Shaw, T., Dawson, W., Piccus, W., Magruder, J. & Kim, H. (2003). Child Welfare Services Reports for California. Retrieved (8/29/2003), from University of California at Berkeley Center for Social
Services Research website. URL: <http:/fcssr.berkeley.edu/CWSCMSreports/>.
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org.
3. Superscript "s" denotes a CWDA categorized "small" county as opposed to "big" county.
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California County Data Book summarized information on children aged 0-17
whereas the CWS data might include children as old as 19. Some totals may not equal I00% due to missing data or rounding.
123
Table 31c:
Estimated Proportion of Children Within the Population in the Child Welfare System Among
.
Select ed Counf1es B,y Reg1on
Region and County
Number of Children
Percentage of Child
inCWS
Population in the CWS
Bay Area
San Mateo
875
0.47%
Monterey
641
0.52%
San Benitos
84
0.54%
Napa
178
0.60%
Marin
361
0.70%
Solano
821
0.71%
. Sonoma
941
0.83%
Santa Cruz
588
0.88%
Santa Clara
4,401
0.94%
Contra Costa
3,061
1.29%
Alameda
6,033
1.54%
San Francisco
3,050
1.97%
Mendocino
628
2.77%
Mountain
Placer
1.28%
843
Calaverass
1.33%
128
Sacramento
8,436
2.45%
Northern
Lassens
82
1.11%
Humboldt
389
1.27%
Butte
827
1.62%
Southern
Santa Barbara
0.48%
516
Orange
7,316
0.89%
San Diego
9,935
1.19%
Los Angeles
1.80%
52,630
Valley
Stanislaus
1280
0.89%
Merced
822
1.11%
Fresno
4757
1.80%
State of California
139,823
1.41%
Notes:
1.
Children in Child Welfare System refer to those in a service component of ER. FR. FM or PP in January 2001 according to
Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A., Lery, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. &
Kim, H. (2003). Child Welfare Services Reports for California. Retrieved [8/29/2003], from University of California at Berkeley Center for
Social Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org
3. Superscript "s" denotes a CWDA categorized "small" county as opposed to "big" county
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California
County Data Book summarized information on children aged 0-17 whereas the CWS data might include children as old as t 9.
124
Descriptive Results:
Effective Practices for Children and Families of Color
Two counties did not provide information on effective practices; as such, a total of 43
effective practices for children and families of color in the child welfare system were identified.
Practices identified by the respondents as the most effective (# 1) for children and families of color
were divided into one of four categories: 1) beginning choice point practices (N = 13), 2) continuing
or ending choice point practices (N = 12), 3) organizational or structural practices (N = 14) and 4)
both beginning and continuing choice point practices (N = 4). Descriptive results for each of these
four categories are presented separately. In addition, descriptive statistics on the location, target
population, goals, history, factors related to implementation, factors that support the practice and
barriers to successful implementation of the practice are presented.
It should also be noted that many counties reported that they do not target services to any
specific racial/ethnic group and that the practices they identified as being most effective with
children and families of color, were also in general, their most effective practices for their overall
child welfare population.
Beginning· Choice Point Practices (N = 13)
Thirteen counties identified practices that take place at early choice points as their most
effective practices for children and families of color in the CWS. The 13 beginning choice point
practices identified by the counties fell into three categories of practices: 1) family group
conferencing, or other team decision-making services (n = 7, 53.8%) 2) voluntary family ·
maintenance or other prevention/family preservation services (n = 4, 30.8%), and 3) risk assessment
tools or systems to guide decision-making (n = 2, 15.4%). Please see Table 32.
Family group conference or other team decision making
ractice
Voluntary family maintenance or other prevention/family
reservation service
Risk assessment tool ors stem
7
53.8%
4
2
Table 33 provides qualitative data on the beginning choice point practices that are related to
family group conferencing or other team decision-making practices. Most respondents described the
target population as children and families who have been detained or are at risk for court-ordered
services, although some respondents described using family group conferencing for all families
involved with the CWS, not just those at risk for court ordered out-of-home placement.
Most respondents described family group conferencing or other team decision-making
practices as involving the family, extended family or the support system in the group conference.
Some respondents described staff meetings or multidisciplinary team meetings that did not involve
the family. Survey respondents described the practice as a collaborative effort with the goal of
producing a plan for the child and family. Many respondents described the process as strengths
based and focused on how best to help children and families.
125
In general, respondents felt that the practice was effective with children and families of color
because it involves clients in the decision-making process. This inclusive model of practice was
described as particularly helpful for children and families of color because they may be "more
distrustful of the system," and also because inclusive practice may be more culturally appropriate
with some families. Additionally, team meetings not involving family members were described as
being beneficial to children and families of color because more accurate decisions are expected
from bringing numerous perspectives together.
Table 34 provides qualitative data on voluntary maintenance or other prevention/family
preservation practices. The target population for these services included families at risk for courtordered out-of-home placements, as well as one program aimed at expectant mothers and parenting
women. These practices were described as intensive in nature and focused on preventing child
maltreatment and out-of-home placements.
Voluntary family maintenance or other prevention/family preservation services were
described as effective with children and families of color because their goal is to keep children with
their birth families. Additionally, these services were viewed as effective for children and families
of color because they are collaborative in nature, involving other systems or organizations such as,
Native American tribes, mental health and probation systems.
Table 35 provides qualitative data on risk assessment tools or systems. These practices are
targeted to all children reported to the child welfare system for suspected maltreatment. Risk
assessment practices were described as tools to help make decisions at critical points by assessing a
number of different risk factors.
Risk assessment tools or systems were described as effective with children and families of
color because they help to assess risk in an objective way, thereby minimizing potential bias based
on race/ethnicity.
Table 36 provides quantitative results on all of the beginning choice point practices
identified from the statewide survey. In general, beginning choice point practices are most often
delivered at the county child welfare office (n = 7, 53.8%); they tend to be targeted to children of all
ages (n = 12, 92.3%); as well as families of any ethnicity (n = 13, 100.0%). Families with substance
abuse problems (n = 9, 69.2%) are also targeted for these beginning choice point practices, as are
families with domestic violence, (n = 9, 69.2%), families with a mental health problem (n = 9,
69.2%), families with income problems (n = 8, 61.5%), and families with a disability (n = 8,
61.5%).
The most frequently identified goal of the practices was to prevent out-of-home placement
(n = 11, 84.6%); other goals included to assess level ofrisk (n = 10, 76.9%); improve psychosocial
functioning of family (n = 10, 76.9%); address substance abuse problems of family (n = 8, 61.5%);
address socioeconomic problems of family (n = 8, 61.5%); improve parenting skills, (n = 7, 53.8%);
address mental health problems of family (n = 7, 53.8%); ensure a permanent placement for the
child (n = 7, 53.8%), and reunify children with their birth families (n = 6, 46.2%)
Table 37 provides data on the history of the beginning choice point practices. Most were
first implemented in response to a particular problem (n = 7, 53.8%) and most were first initiated by
126
the director of the child welfare system (n = 8, 61.5%). The year the practices were first
implemented ranged from 1992 to 2002, and 15.4% (n = 2) of respondents reported that the practice
has changed significantly between now and the time it was first implemented.
Table 38 provides information on factors related to implementation ofbeginning choice
point practices. The majority of these practices are permanent (n = 10, 76.9%); most are funded
through state funds (n = 7, 53.8%) and 38.5% (n = 5) are mandated. Most social workers are
provided with training in how to implement the practice (n = 12, 92.3%); 15.4% (n = 2) of counties
contract with another agency or system to deliver the practice and 69.2% (n = 9) indicated that the
practice is widespread throughout the child welfare system. No respondents reported that the
practice is delivered in a specific geographic area or community. The majority of respondents (n =
8, 61.5%) reported the practice has been evaluated to determine its effectiveness and of those, 87 .5
% (n = 7) indicated that the practice is effective with the target population. Of those counties where
the practice had not been evaluated, 80.0% (n = 4) indicated that an evaluation is not needed.
Factors that support the practices and barrier to successful implementation are presented in
Table 39. The majority of respondents felt that families and children help support the successful
implementation of the practice (n = 10, 76.9%). Support from the mental heath system was also
frequently noted (n = 9, 69.2%); as was support from the substance abuse system (n = 8, 61.5%),
contract agencies (n = 7, 53.8%), county government officials (n = 7, 53.8%), and evaluations that
show practice is successful (n = 7, 53.8%). Few barriers to successful implementation were noted,
38.5% (n = 5) ofrespondents indicated that there were no barriers to successful implementation, and
38.5% (n = 5) also reported unstable funding is a barrier.
127
Table 33: Beginning choice point practices: Family group conference or other team decision-making practice (N = 7)
Name of
Target
Description of practice
Why effective with children and families of color
population
practice
Family
Available to all
The Family Group Conference is a tool for solving
I believe that this practice works well for families
Group
consumers who
family problems that is based on a simple, traditional
of color because it is based on principles shared
by those families. It involves much extended
Conferencing have family,
belief: The combination of family strengths and
extended family, community support can keep children safe and well
family, and is centered around a meal (family
friends, support cared for. A Family Conference is a gathering of family alone time). It is inclusive of all family members,
members, friends, and community specialists who join
including elders. Meetings can be held at the
together to improve the care and protection of a specific location that best fits the needs of the family.
child or children.
Family
Children of all
Family Conferencing is a program that uses strengthThis program allows families to assist in the
based methodology by allowing families to participate
Conferencing ages
decision process as to what is best for the family.
in the decision-making process as to how best help their
children. Family Conferencing is a gathering of
extended family members, other intimate support
people, and service providers in order to share
information and give the family an opportunity to
create a plan for their children.
Involving the extended family, non-related extended
Family
All families
These families are often even more distrustful of
Group
involved with
family and all service providers in a conference in
the "system" than most families; this practice
Conferencing CWS at any
which the family takes responsibility for making a plan allows the family to exert more control over the
for the child/
stage of the
direction of services and the planning for the
process can use
children.
this program
Youth Pilot
Children at risk This program targets all children (including children of Team decision-making/goal setting/evaluation
Project
of placement, or color) at risk of placement, or at risk of being moved to process which includes the family and all support
(YPP)
at risk of more
a more restrictive placement. The program uses monies persons they wish to include, as well as the
restrictive
that would be spent on foster care to provide services to various service providers working with the
family.
placement
maintain the child in the home or in a less restrictive
placement. The program uses a team decision-making
and goal-setting model, which includes the family and
supports.
128
Table 33 continued: Beginning choice point practices: Family group conference or other team decision-making practice (N = 7)
Target
Name of
Description of practice
Why effective with children and families of color
practice
population
Placement
All detained
After a child is detained, a placement meeting is held as Families and relatives are able to participate in
Meetings
children and
soon as possible to determin the "next best placement"
the decision making and are able to provide input
their families
for that child. In addition to Social Workers and other
in to the "next best placement" for a child, which
staff, parents and family members are included
can include placement with relatives or in a
whenever possible in this process, so that their input
culturally relevant home if the child cannot be
can be provided in this decision. The safety of the child safely returned to his or her parents.
and culturally relevant placements are goals ..
Detention
All children,
No child is detained by CPS or remains detained by law Many perspectives and opinions are brought to
Case Staffing CPS may
enforcement without being discussed as soon as
bear at inception, promoting more objectivity and
consider
possible in a staffing. The participants include the line balance regarding FACTS and EVIDENCE of
detaining or who ER worker in the matter, ER Supervisor, On-going
risk and best interests of the child. Less
have been
Case Supervisor, and CPS Program Manager. The ER
possibility of stereotyping, subjectivity, ignorance
detained by law worker presents the issues of protection, risk, current
of cultural factors.
enforcement in
situation, the court worker/supervisor, lends legal
an emergency
documentation evidence perspective, all others ask
questions, add comments with the ultimate goal of
consensus on whether to proceed with intervention and
what the intervention looks like.
Team
All ER referrals Consensus decision making with workers, key program This decision-making forum is used for ALL
considered for
Assessment
supervisors, mental health clinician, PHN, family
cases moving to on-going services. The core
Planning
voluntary or
decision meeting coordinator, and often other service
group attends pretty consistently and represents
court-ordered
providers. Type & level of CWS intervention
multiple perspectives, resulting in more consistent
services
determined after exploring safety/risk, family
and well thought out decision making regarding
strengths/needs, and available
case handling and service needs.
resources/services/strategies. This is a collaborative
process which may be utilized for other key case
decisions.
129
Table 34: Beginning choice point practices Voluntary family maintenance or other prevention/family preservation service (N = 4)
Target
Description of practice
Name of
Why effective with children and families of color
population
practice
Voluntary
Families at high Social worker and social work assistant are dedicated to American Indian Tribe has facilitated family
program. Services are based on strength-based and
conferencing for tribal families.
Family
risk of court
Maintenance intervention due family-focused approaches, including family
Program
to abuse and
conferencing and family decision-making. Promoting
neglect issues.
Safe and Stable Families funding used for contracting
services to include psych assessments, intensive case
mana~ement services, and a half-time family advocate.
Expectant
Program in which pregnant mothers with a CPS history Missing
High Risk
Mothers and
or involvement coupled with drug or alcohol use are
Infant
children under
targeted for intensive services and treatment to avoid
Program
age 1
placing their children into the foster care system.
Mothers of newborns are also a target population for
this program.
Family Preservation practice is an integrated,
Family
Abused and
Built a community consensus around the effective
comprehensive approach to strengthening and
service delivery model for Family Preservation in
Preservation neglected
the County; Supported public/private
families at
preserving families who are at risk of or already
collaborative planning and service delivery
imminent risk of experiencing problems in family functioning with the
efforts; Preventive-oriented; Collaboration with
placement
goal of assuring the physical, emotional, social,
educational, cultural and spiritual development of
other County Departments such as DPSS, Mental
children in a safe and nurturing environment.
Health and Probation; Strong and continued
leaderhip within the Department of Children and
Families Services.
Missing
Provide intervention and make referrals to other
We do not have foster homes in our county. We
Keep
agencies so that family can remain intact, if possible.
have seen children go through foster care system
families
together
and the majority do not do well.
130
Table 35: Beginning choice point practices Risk assessment tool or system (N = 2)
Description of practice
Target
Name of
population
practice
Research-based & validated tools for decision-making
All new intake
Structured
at critical points in child welfare and Foster Care ·
to CPS
Decision
systems developed by the Children's Research Center
(currently) will
Making
in collaboration with COSS & California Counties.
System
soon extend to
(SOM)
all active cases
The risk assessment process looks at the child factors,
Best practice All children
including age, disability, illnesses as well as the
reported
risk
parent's ability to care-take, the history of abuse, also
assessment
speaking with the child, the parent and collateral
during the
contacts and then making a decision. We do not have a
ER phase
universal risk assessment tool, the workers are trained
in this process. We have an ER team that does this
practice.
Why effective with children and families of color
Objective nature of factors assessed.
Children are assessed in an objective wayconsidering their factors that place them at risk,
rather than being targeted for their ethnicity.
131
T a ble 36 : Begmnmg c hoice point practices: Location, taniet population and goals (N = 13)
Frequency (Percentage)
Where does practice #1 take place?
7 (53.8%)
• County child welfare office
4 (30.8%)
• Throughout the child welfare system
2 (15.4%)
• Birth family's home
Contract
agency
site
2
(15.4%)
•
Foster family's home
1
(7.7%)
(7.7%)
1
• Another out-of-home placement site
Other
2 (15.4%)
Who is the target population for practice # 1?
12 (92.3%)
• Children of all ages
(7.7%)
1
• Very young children
13 (100%)
• Families of any ethnicity
9 (69.2%)
• Families with substance abuse problems
(69.2%)
9
Families
with
domestic
violence
•
9 (69.2%)
• Families with a mental health problem
8 (61.5%)
Families with income problems
8 (61.5%)
• Families with a disability
6 (46.2%)
• Other
What are the purposes and goals of practice # 1?
11 (84.6%)
• Prevent out-of-home placement
10 (76.9%)
Assess level of risk
10 (76.9%)
• Improve psychosocial functionimz of family
(61.5%)
8
• Address substance abuse problems of familv
(61.5%)
8
• Address socioeconomic problems of family
(61.5%)
8
Address housing problems of family
(53.8%)
7
Improve parenting skills
(53.8%)
7
• Address mental health problems of familv
(53.8%)
7
• Ensure a permanent placement for the child
(46.2%)
6
• Reunify children with their birth families
(7.7%)
1
• Other
•
•
•
•
•
•
132
Table 37: Beginning choice point practices: History of practice (N = 13)
Frequency (Percentage)
Why was practice # 1 first implemented?
7 (53.8%)
• In response to a particular problem
4 (30.8%)
• Funds became available
(30.8%)
New
policies
were
implemented
4
•
Who first initiated practice # 1?
8 (61.5%)
• Director of the child welfare system
4 (30.8%)
• Child welfare workers
3 (23.1 %)
Other
•
When was practice #1 first implemented?
1
(7.7%)
• 2002
3 (23.l %)
2001
1
(7.7%)
2000
•
2 (15.4%)
• 1999
1 (7.7%)
• 1998
1
(7.7%)
1995
(7.7%)
1
1992
Has practice # 1 changed significantly between the time it
was first implemented and now?
2 (15.4%)
• Yes
•
•
•
133
Table 38: Beginning choice point practices: Factors related to implementation (N = 13)
Frequency (Percentae;e)
Current implementation status of practice # 1
IO (76.9%)
• Permanent
(23.1%)
3
• Dependent on re-funding
What is the primary funding source for practice # 1?
(53.8%)
7
• State funds
5
(38.5%)
• Federal funds
(23.1%)
3
• County funds
2
(15.4%)
Other
•
Is practice # 1 mandated?
5 (38.5%)
• Yes
Are social workers provided with training in how to
implement practice #1?
12 (92.3%)
• Yes
Do you contract with another agency or system to deliver
practice # 1?
2 (15.4%)
• Yes
Is practice # 1?
9 (69.2%)
• Widespread throughout the child welfare system
4
(30.8%)
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
(0.0%)
0
• Yes
Has practice #1 been evaluated to determine its
effectiveness?
(61.5%)
8
• Yes
If yes, what were the results of the evaluation?
(87.5%)
7
• Effective with target population
(12.5%)
1
• Other
If practice # 1 has not been evaluated, why not?
(80.0%)
4
• An evaluation was not considered to be needed
(20.0%)
1
• Other
134
Table 39: Beginning choice point practices: Factors that support practice and barriers to successful
implementation (N = 13)
Frequency (Percentage)
What are the maior factors that support practice # 1
10 (76.9%)
• Suooort from families and children
(69.2%)
9
Support from the mental health system
8
(61.5%)
Support from the substance abuse system
(53.8%)
Support
from
contract
agencies
7
•
7 (53.8%)
• Support from county government officials
(53.8%)
7
• Evaluations show that practice is successful
(46.2%)
6
Support
from
the
community
•
(46.2%)
6
• Support from state government officials
(46.2%)
6
• Stable funding
(46.2%)
6
• Support from the health system
(30.8%)
4
• Support from the criminal justice system
(30.8%)
4
• Support from foster families
(23.1%)
3
• Support from other out-of-home placement providers
What have been the barriers to successful implementation of
this practice?
5 (38.5%)
• No barriers
5 (38.5%)
• Funding is not stable
(15.4%)
2
Insufficient
support
from
child
welfare
workers
•
(7.7%)
1
• Insufficient support from management
(7.7%)
1
• Insufficient support from families and children
• Evaluations are inconclusive as to the success of the
(7.7%)
1
practice
•
•
135
Continuing Care/Ending Choice Point Practices (N = 12)
Twelve counties identified continuing care/ending choice point practices as their most
effective with children and families of color in the CWS. The 12 continuing care/ending choice
point practices identified by the counties fell into three categories of practices: I) placement related
practices (n = 6, 50.0%); 2) the Family to Family Initiative (n = 4; 33.3%); and 3) intensive and
collaborative interventions for high-risk children (n = 2; 16.7%). Please see Table 40.
Table 40: Continuin care/endin
Intensive and collaborative interventions for high risk
children
4
2
16.7%
Table 41 contains qualitative data on placement related continuing care/ending choice point
practices. The target populations of these services varies, some are targeted to all children in out-ofhome care, others are targeted specifically to Native American children, one is targeted to children
in group homes {The Village Project) or at risk for entry into group homes and one is targeted to
parents with young children likely to enter out-of-home care (Shared Family Care).
Descriptions of the practices also varied. For instance, many participants noted that the first
priority is to place children with relatives; some discussed the role of culture and ethnicity in
making placement decisions. One practice (The Village Project) is described as a multidisciplinary
effort aimed at avoiding higher levels of care for children or working toward a lower level of care.
Lastly, one practice (Shared Family Care) is aimed at providing a placement for the whole family in
a mentor, foster home.
Respondents provided a variety of opinions regarding why the practice is effective with
children and families of color. Specifically, many mentioned the importance of placing children in
culturally appropriate homes-whether with relatives or with foster parents. The Shared Family
Care program is thought to be effective with children and families of color because it allows
families with few resources to be placed together and The Village Project was considered to be
effective with children and families of color because of a multicultural staff and flexibility of
funding sources.
Table 42 provides descriptive information on the Family to Family Initiative. The target
population for this practice varies somewhat; some respondents identified children and families in
foster care as the target population, others identified care providers and resource homes or all clients
and care providers. The Family to Family Initiative is described as engaging and involving
communities in preventing out-of-home placements, reunification, and providing resource homes
within children's neighborhoods. Many also described the practice as encompassing several aspects
of the child welfare system with the goal of improving the overall child welfare system.
136
Family to Family was described as effective with children and families of color because it
involves community and family support and is aimed at reducing disparities associated with gender,
and ethnicity and increasing the number of children placed in their own communities.
Table 43 provides qualitative data on continuing/ending choice point practices that are
aimed at intensive and collaborative interventions for high-risk children. The target population for
these services is children who are at risk for a more restrictive placement. These practices were
described as collaborative in nature-involving systems such as public health, community based
organizations, probation, behavioral health and faith based organizations. Interventions are
comprehensive, intensive and culturally competent.
Intensive and collaborative interventions for high-risk children were described as being
effective with children and families of color because the services are individualized to each family's
needs and are intensive in nature These practices were also described as "engaging families in a
non-threatening and respectful manner," and assessing the types of strengths and resources already
being used by the family. Children and families of color were described as benefiting from these
services because they are respectful of culture and involve a family-driven process
Quantitative results from survey responses are provided in Table 44. Most respondents
indicated that these continuing care/ending choice point practices take place throughout the child
welfare system (n = 7, 58.3%) and are targeted to children of all ages (n = 9, 75.0%), as well as
older children (11 to 18 years) (n = 5, 41.7%), young children (4 to 10 years), (n = 3, 35.0%), and
very young children (0 to 3 years) (n= 3, 25.0%). Most are targeted to families of any ethnicity (n =
10, 83.3%), with far fewer being targeted to African American families (n = 2, 16.7%), or American
Indian families (n = 2, 16.7%). Families with substance abuse problems were also identified as a
target population (n = 9, 75.0%) as were families with a mental health problem (n = 9, 75.0%),
families with domestic violence (n = 9, 75.0%), families with income problems (n = 7, 58.3%), and
families with a disability (n = 7, 58.3%).
The most frequently noted goal of the practice is to reunify children with their birth families
(n = 10, 83.3%), followed by ensuring a permanent placement for children (n = 8, 66.7%), and
preventing an out-of-home.placement (n = 8, 66.7%). Other commonly occurring goals included
improving parenting skills (n =7, 58.3%); improving psychosocial functioning of family (n = 6,
50.0%); addressing socioeconomic problems of family (n = 6, 50.0%); and addressing housing
problems of family (n = 6, 50.0%).
Table 45 provides information on the history of the continuing care/ending choice point
practices. The majority of practices were first implemented due to new policies (n = 8, 66.7%); the
director of the child welfare system was most often identified as the person who first initiated the
practice (n = 11, 91.7%), followed by child welfare workers (n = 4, 33.3%). Many practices were
first implemented in 2001 (n = 4, 33.3%), and 33.3% (n = 4) had changed significantly since first
being implemented.
Factors related to implementation are presented in Table 46. The majority of practices are
permanent (n = 11, 91.7%), and are funded by state funds (n = 7; 58.3%). Of the practices, 41.7% (n
= 5) are mandated and 91.7% (n = 11) of counties provide social workers with training on how to
implement the practices. Of the counties, 33.3% (n = 4) contract with another agency to implement
the practices, and 58.3% (n = 7) of respondents indicated that the practice is widespread throughout
137
the child welfare system. Very few respondents indicated that there is a specific geographic
community in which the practice is targeted (n = 2, 16.7%), and only 16.7% (n = 2) indicated that
the practice has been evaluated to determine its effectiveness. Of those practices that had been
evaluated, 100.0% (n = 2) reported that the practice is effective with the target population. For
practices that had not been evaluated, 33.3% (n = 4) have not been evaluated because of insufficient
resources.
Table 4 7 provides information on factors that support the practice and barriers to successful
implementation of the practice. The most commonly mentioned supportive factors included support
from child welfare workers (n = 11, 91.7%), support from management (n = 11, 91.7%), and
support from families and children (n = 11, 91.7%). Other factors that support the practice include
support from foster families (n = 9, 75.0%); contract agencies (n = 8, 66.7%); other out-of-home
placement providers (n = 7, 58.3%), the mental health system (n = 7, 58.3%), state government
officials (n = 7, 58.3%); and county government officials (n = 7, 58.3%). Many respondents did not
note barriers to successful implementation of the practice, no barriers were reported by 16.7% (n =
2) of respondents, and 16. 7% (n = 2) reported that unstable funding is a barrier to successful
implementation.
138
Table 41: Continuing care/ending choice point practices: Placement related practices (n = 6)
Description of practice
Name of practice Target population
Culturally
sensitive
placements
All children who
require out-ofhome placement
Shared Family
Care
Mothers/Fathers
with young
children likely to
enter out of home
care
Place children
with relatives
All Children
Native American
children placed
with families
through ICWA
Native Am/ICWA
In all cases, we must first consider any willing,
able and appropriate relatives for placement. If
this is not possible, we then try to place in a home
of the same ethnicity, religion, etc. If this is not
possible, then we let the foster parents know that
their culture is an important issue, and we are
expecting them to continue to allow the children
to live in a culturally sensitive environment.
This program provides placement for the family
in a mentor home. Usually one parent (mother in
most cases) is subject to intervention by child
welfare. Rather than place the child in foster
care, it is assessed that the mother is willing to
participate in Shared Family Care and will abide
with the conditions of working and living with a
mentor family.
Identify as quickly as possible family members
for placement
Attempt to place all children with families.
Counseling/assessment/service orientated.
Why effective with children and families
of color
Our county is predominately Caucasian,
so children of color are already truly a
minority. Placing them in a home that is
sensitive to that is very important.
For families with few resources it allows
for placement of child and parent
together. Most of our families with few
resources are those living in communities
impacted by poverty, drugs, and crimes.
Families of color primarily live in these
areas.
Children are kept in their communities
with relatives of same culture and
practices promote stability for children
They are treated the same as all other
cases---dignity--compassion-provided
services- acknowledgement of their
concerns and problems.
139
Table 41 continued: Continuing care/ending choice point practices: Placement related practices (n = 6)
Why effective with children and families
Name of practice Target population Description of practice
of color
SB163
Kids in group
Multidisciplinary effort to work with families and
A multicultural staff in the county and
program-"The
homes level 13 or
kids to avoid higher levels of care or get them
in the community-based providers and
flexibility of funding sources-money
14 or at risk for
village project"
into a lower level of care-like a foster home or
going into a group back into their home. The staff are bilingual and
can be used for activities outside the
home
bi-cultural and they are able to make connections group home, so it can be used for things
with multi-cultural families. They also work hand to prevent a group home placement and
and hand with a parent advocacy group-and they not just for kids already in a group home.
advocate for children and families especially kids
with educational difficulties.
ICWANative Americans When a Native American child is removed from
We are able to place many of our Native
placement of
the home of parents, we coritact any relatives in
American children in homes that comply
Native American
order to place the child in compliance with
with ICWA.
children
ICWA.
140
Table 42: Continuing care/ending choice point practices: Family to Family Initiative (n = 4)
Name of practice Target population
Description of practice
Family to Family
All children in out
ofhome
placements
Family to Family
Initiative
All clients and
care providers
Family to Family
Initiative
Children &
families in foster
care
Family to family
Resource homes
and the
community
Regional initiatives underway to work with local
communities to prevent removal and/or maintain
children in their community/school if they must
be removed. Community collaboration on child
safety and well being.
The use of the following Family to Family
strategies are applied to all aspects of child
welfare placement practices in this County: Team
decision making, Recruitment, Retention, and
Support of Care Providers, Self-Evaluation and
Community Partnership.
Community-based, family focused and designed
to meet 9 key outcomes toward improving our
foster care system including reducing disparities
associated with ethnicity, J2;ender & age.
This is an effort to engage the community in
providing resource homes for foster children who
are being placed in homes from their
neighborhoods in which they are being removed.
The community is being engaged and as well as
community based organizations.
Why effective with children and families
of color
Community and family
involvement/support.
Missing
Outcomes include: reduction in
disparities associated with sex, gender &
ethnicity, increasing kids placed in their
own neighborhoods & communities,
Missing
141
Table 43: Continuing care/ending choice point practices: Intensive and collaborative interventions for high-risk children (n = 2)
Name of practice Target population
Description of practice
Why effective with children and families
of color
Children who are
Children's
To develop a child 'who is emotionally
As with all families, intensive services,
in danger of being disturbed/behaviorally disturbed' through a
System of Care
individualized plans
placed in a higher collaborative interagency system of care with
level of care i.e.
services that are "youth-centered;" family
Group homes
involved & focused; culturally competent;
community based and comprehensive.
This practice really values families individually
Wraparound
Children who are
The practice engages with families in a
services
placed or at risk of and values their culture. It provides intensive and respectful and non-threatening manner
placement into
supportive in-home services to keep the child at
and asking them what supports, what
level 14 group
people, in your life and in your
home or return them home. It's also been a
home care
collaborative effort between probation, behavioral community that are important to you and
health, public health, CBO's, and faith-based
how can we incorporate those aspects of
organizations.
your life into this child and family plan to
help your child stay safely at home. And
by asking those questions and evaluating
what the family uses to thrive and asking
them to make that a part of the service
planning for their child is kind of a
foreign way of working with the families.
It is a family-driven, needs based process,
rather than a professionally-driven
process. There is respect for the family's
culture.
142
Table 44: Continuing care and ending choice point practices:
Location, target population and goals (N = 12)
Where does practice #1 take place?
•
Throughout the child welfare svstem
• County child welfare office
• Birth family's home
• Foster family's home
• Court
• Another out-of-home placement site
• Other
Who is the target population for practice # 1?
• Children of all ages
• Very young children (0 to 3 years)
• Young children (4 to 10 years)
• Older children (11 to 19 years)
• Families of any ethnicity
• African American families
• American Indian families
• Asian American/Pacific Islander families
• Latino families
• White families
• Families of another ethnicity
Families with substance abuse problems
Families with a mental health problem
• Families with domestic violence
• Families with income problems
• Families with a disability
• Other
What are the purposes and goals of practice # 1?
• Reunify children with their birth families
• Ensure a permanent placement for the child
• Prevent out-of-home placement
• Improve parenting skills
• Improve psychosocial functionin~ of familv
• Address socioeconomic problems of family
• Address housing problems of familv
• Address substance abuse problems of familv
• Address mental health problems of family
Assess level of risk
• Other
•
•
•
Frequency (Percentage)
7
4
4
4
4
4
5
(58.3%)
(33.3%)
(33.3%)
(33.3%)
(33.3%)
(33.3%)
(41.7%)
9
(75.0%)
(25.0%)
(35.0%)
(41.7%)
(83.3%)
(16.7%)
(16.7%)
(8.3%)
(8.3%)
(8.3%)
(16.7%)
(75.0%)
(75.0%)
(75.0%)
(58.3%)
(58.3%)
(50.0%)
3
3
5
10
2
2
1
1
1
2
9
9
9
7
7
6
10
8
8
7
6
6
6
5
5
4
4
(83.3%)
(66.7%)
(66.7%)
(58.3%)
(50.0%)
(50.0%)
(50.0%)
(41.7%)
(41.7%)
(33.3%)
(33.3%)
143
Table 45: Continuing care and ending choice point practices: History of practice (N = 12)
Frequency (Percentage)
Why was practice # I first implemented?
(66.7%)
8
• New policies were implemented
(25.0%)
3
• Funds became available
(16.7%)
2
In
response
to
a
particular
problem
•
(50.0%)
6
• Other
Who first initiated practice # 1?
11 (91.7%)
• Director of the child welfare system
4 (33.3%)
• Child welfare workers
2 (16.7%)
• Mental health system
(8.3%)
1
• Foster families
(8.3%)
1
Substance
abuse
system
•
(8.3%)
1
• Contract agencies
(8.3%)
1
• Criminal justice system
4 (33.3%)
• Other
When was practice # 1 first implemented?
•
•
•
•
2003
2002
2001
2000
Has practice # 1 changed significantly between the time it
was first implemented and now?
• Yes
1
2
4
2
(8.3%)
(16.7%)
(33.3%)
(16.7%)
4
(33.3%)
144
Table 46: Continuing care and ending choice point practices:
Factors related to implementation (N = 12)
Current implementation status of practice # 1
• Permanent
• Demonstration
What is the primary funding source for practice # 1?
• State funds
• County funds
• Federal funds
• Nongovernmental grant
• Other
Is practice # 1 mandated?
• Yes
Are social workers provided with training in how to
implement practice #1?
• Yes
Do you contract with another agency or system to deliver
practice # 1?
• Yes
• No
• Combination of internal and external resources
Is practice # 1?
• Widespread throughout the child welfare system
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
• Yes
Has practice #1 been evaluated to determine its
effectiveness?
• Yes
If yes, what were the results of the evaluation?
• Effective with target population
If practice #1 has not been evaluated, why not?
• Insufficient resources
• Insufficient time
• An evaluation was not considered to be needed
• Other
Frequency (Percentage)
11
1
(91.7%)
(8.3%)
7
6
3
2
3
(58.3%)
(50.0%)
(25.0%)
(16.7%)
(25.0%)
5
(41.7%)
11
(91.7%)
4
4
3
(33.3%)
(33.3%)
(25.0%)
7
5
(58.3%)
(41.7%)
2
(16.7%)
2
(16.7%)
2
(100.0%)
4
2
1
3
(33.3%)
(16.7%)
(8.3%)
(25.0%)
145
Table 47: Continuing care and ending choice point practices: Factors that support practice and
barriers to successful implementation (N = 12)
Frequency (Percentage)
What are the major factors that support practice # 1
11 (91.7%)
• Support from child welfare workers
11 (91.7%)
• Support from management
(91.7%)
11
• Support from families and children
(75.0%)
9
• Support from foster families
(66.7%)
8
• Support from contract agencies
(58.3%)
7
• Support from other out-of-home placement providers
(58.3%)
7
Support from the mental health system
(58.3%)
7
Support from state government officials
(58.3%)
7
Support from county government officials
(50.0%)
6
• Stable funding
(50.0%)
6
Support
from
the
community
•
(50.0%)
6
• Support from the health system
(50.0%)
6
Support
from
criminal
justice
system
•
(41.7%)
5
• Support from the substance abuse system
(25.0%)
3
• Evaluations show that practice is successful
(16.7%)
2
• Other
What have been the barriers to successful implementation of
this practice?
(16.7%)
2
• No barriers
(16.7%)
2
• Funding is not stable
(8.3%)
1
• Insufficient support from state government officials
Insufficient
support
from
county
government
•
(8.3%)
1
officials
(8.3%)
1
• Insufficient support from child welfare workers
(8.3%)
1
• Insufficient support from foster families
(8.3%)
1
Insufficient
support
from
the
community
•
(8.3%)
1
• Insufficient support from the criminal justice system
(50.0%)
6
• Other
•
•
•
146
Both Beginning and Continuing/Ending Choice Point Practices (N = 4)
Four respondents identified practices that were related to both beginning and
continuing/ending choice points as their most effective practices for children and families of
color in the CWS. The four beginning and continuing/ending choice point practices fell into two
categories: 1) community-based services (n = 3, 75.0%); and 2) strength based assessments (n =
I, 25.0%). Please see Table 48.
Table 48: Both beginning and continuing/ending choice point practices: Types of practices (N =
4
Table 49 provides qualitative information on community based services. The target
population for these services includes low-income populations, families at risk for entry into the
child welfare system, families already in the child welfare system, and the community. These
practices were described as being located in the communities in which clients reside and include
a wide range of services such as, health, substance abuse, public assistance, domestic violence
and home parenting services. Practices were also described as collaborative and aimed at
preventing an out-of-home placement or reunifying children with their families.
These practices were described as effective with children and families of color because
services are "user friendly;" clients have easy access to many services that are housed in one
location. Community based services are free and outreach for these services is performed
through advertisements into different communities. Moreover, because these services are
typically located within clients' own communities, services are "community-specific" and clients
may run into their own community members while accessing services, thus making the services
less intrusive for clients. Additionally, these practices were described as flexible and able to
address "changing community needs."
Table 50 provides information on family-centered strength based assessments. This
. practice was described as being targeted to Hispanic, African American, American Indian and
Asian American/Pacific Islander clients. It is focused on family strengths with an empowerment
perspective.
Quantitative results from survey responses are presented in Table 51. Practices that are
both beginning and continuing/ending practices typically take place at a contract agency site (n =
2, 50.0%), or the birth family's homes (n = 2, 50.0%). One hundred percent (n = 4) of these
practices are targeted to: children of all ages, families of any ethnicity, families with substance
abuse problems, families with a mental health problem, families with domestic violence, families
with income problems, or families with a disability. Seventy five percent (n = 3) ofpractices
have the following goals: preventing an out-of-home placement; reunifying children with their
birth families; assessing level of risk; improving psychosocial functioning of families; improving
parenting skills; addressing substance abuse problems; addressing mental health problems;
addressing socioeconomic problems; and addressing housing problems.
147
Information on the history of the practices is presented in Table 52. The practices were
first implemented in response to a particular problem (n = 1, 25.0%); funds becoming available
(n = 1, 25.0%); or new policies being implemented (n = 1, 25.0%). Child welfare workers (n = 2,
50.0%); the child welfare director (n = 2, 50.0%); and contract agencies (n = 2; 50.0%) were
identified as having first initiated the practices. The year the practice was first implemented
ranged from 1967 to 1997, and 66.7% (n = 2) of respondents indicated that the practice had
changed significantly since it was first implemented.
Factors related to the implementation of both beginning and continuing/ending choice
point practices are presented in Table 53. All practices (n = 4, 100.0%) are permanent and most
are primarily funded through federal funds (n = 4, 100.0%). Fifty percent of respondents (n = 2)
indicated that the practice is mandated and 33.3% contract with another agency to deliver the
practice. Seventy five percent of respondents indicated that the practice is widespread throughout
the child welfare system. All respondents (n = 4, 100.0%) reported that the practice has been
evaluated, and 50.0% (n = 2) indicated that evaluations demonstrated that the practice is effective
with the target population.
Factors that support the practice and barriers to successful implementation are presented
in Table 54. One hundred percent of respondents (n = 4) indicated that support from the
following entities are major factors that support the practices: community, county government
officials, mental health system, substance abuse system, health system, and families and
children. Seventy five percent of respondents (n = 3) indicated that unstable funding is a barrier
to successful implementation of the practices.
148
Table 49: Both beginning and continuini /ending practices: Community based services
Target population
Description of practice
Name of
practice
Community- Low-income
Federal and state funded programs that are free to
populations
low-income clients, including the health
based
department, substance abuse treatment, Aids
services
department, WIC, immunizations, planned
parenthood. Services include domestic violence
support, clothing donations, utility assistance,
HEAD Start program; sexual assault crisis, and in home parenting programs, and teenage parenting
programs.
Children and Families at There are many Family Resource centers. A
Family
Resource
risk for entry into the
collaborative of agencies applies for funds from the
CWS and children and
funding source (though RFP). The Family
Centers
families already in
Resource Centers have a number of agencies
CWS. Most clients are
working together within one site - so it provides
Hispanic, the second
just one-stop for families to receive services. The
Family Resources Centers are located in the
most are API
communities and the neighborhoods where the
clients live. Services are meant to be accessible and
less intimidating than county government services.
Family
Community, Families
Commmunity specific family based services located
in six geographical areas in the County with embed
and Children
Resource
access to public child welfare, mental health,
Centers
alcohol and other drug and physical health services.
Why effective with children and families of
color
Because services are free and they advertise
into all areas of the community, and there are
satellite offices in many locations; so clients
have access to the services, which is
extremely helpful. They also cross-refer so we
all depend on one another. We have a large
Hispanic popuiation so information is
available in Spanish. Transportation is
accessible.
Because wherever the families are they can
reach out to the resources that meet their
need-there are a multitidue of agencies in
one place-it's a one-stop shop. Clients are
more likely to run into their own community
members at the Family Resource Centers. It is
user friendly and less intrusive than other
services.
Practice is community specific with flexibility
to address changing community needs.
Subsequently,_ it offers increased accessibility
to services, intervention/prevention and
family support and advocacy.
149
Table 50: Both beginning and continuin1 /ending practices: Strength based assessments
Target population
Description of practice
Name of
practice
Family
Hispanic, AA, American Focused on family strengths with the empowerment
Indian and Asian
perspective.
centered
strength
based
assessments
Why effective with children and families of
color
Missing
150
Table 51: Both beginning and continuing care/ending choice point practices: Location, target
population and goals (N = 4)
Frequency (Percentage)
Where does practice #1 take place?
2
(50.0%)
• Contract agency site
2
(50.0%)
• Birth family's home
1
(25.0%)
County
child
welfare
office
•
(25.0%)
1
• Foster family's home
(25.0%)
1
Another
out-of-home
placement
site
•
2
(50.0%)
• Other
Who is the target population for practice # 1?
4 (100.0%)
• Children of all ages
4
(100.0%)
• Families of any ethnicity
4 (100.0%)
• Families with substance abuse problems
4 (100.0%)
• Families with a mental health problem
(100.0%)
4
Families
with
domestic
violence
•
4 (100.0%)
• Families with income problems
4 (100.0%)
• Families with a disability
(25.0%)
1
Other
What are the purposes and goals of practice # 1?
3 (75.0%)
• Prevent out-of-home placement
3 (75.0%)
• Reunify children with their birth families
3 (75.0%)
• Assess level of risk
3 (75.0%)
• Improve psychosocial functioning of family
(75.0%)
3
Improve
parenting
skills
•
3 (75.0%)
• Address substance abuse problems of family
(75.0%)
3
• Address mental health problems of family
3 (75.0%)
• Address socioeconomic problems of family
3 (75.0%)
• Address housing problems of family
2
(50.0%)
Ensure
a
permanent
placement
for
the
child
•
3 (75.0%)
• Other
•
151
Table 52: Both beginning and continuing care/ending choice point practices:
History of practice (N = 4)
Frequency (Percentage)
Why was practice #1 first implemented?
1 (25.0%)
• In response to a particular problem
1 (25.0%)
• Funds became available
1 (25.0%)
• New policies were implemented
1 (25.0%)
• Other
Who first initiated practice #1?
2 (50.0%)
• Child welfare workers
(50.0%)
2
• Director of the child welfare system
2 (50.0%)
• Contract agencies
1 (25.0%)
Clients of the CWS
1 (25.0%)
Foster families
1 (25.0%)
Out-of-home placement providers
(25.0%)
1
• Mental health system
1 (25.0%)
• Substance abuse system
1 (25.0%)
• Health system
(25.0%)
1
Criminal
justice
system
•
2 (50.0%)
• Other
When was practice #1 first implemented?
1 (33.3%)
• 1997
1 (33.3%)
• 1995
1 (33.3%)
• 1967
Has practice # 1 changed significantly between the time it
was first implemented and now?
2 (66.7%)
Yes
•
•
•
•
152
Table 53: Both beginning and continuing care/ending choice point practices: Factors related to
implementation (N = 4)
Frequency (Percentage)
Current implementation status of practice # 1
4 (100.0%)
• Permanent
What is the primary funding source for practice # 1?
4 (100.0%)
• Federal funds
(50.0%)
2
State
funds
•
(50.0%)
2
• County funds
(25.0%)
1
• Nongovernmental grant
Is practice # 1 mandated?
(50.0%)
2
• Yes
Are social workers provided with training in how to
implement practice #1?
(50.0%)
2
• Yes
Do you contract with another agency or system to deliver
practice #1?
(33.3%)
1
• Yes
(33.3%)
1
No
•
1 (33.3%)
• Combination of internal and external resources
Is practice #1?
(75.0%)
3
• Widespread throughout the child welfare system
(25.0%)
1
Targeted
to
a
relatively
small
group
of
clients
•
Is there a specific geographic area or community in which
this practice is targeted?
(50.0%)
2
• Yes
Has practice #1 been evaluated to determine its
effectiveness?
4 (100.0%)
• Yes
If yes, what were the results of the evaluation?
(50.0%)
2
Effective with target population
• Evaluation did not provide evidence that practice was
(25.0%)
1
effective
1 (25.0%)
• Other
•
153
Table 54: Both beginning and continuing care/ending choice point practices: Factors that support
practice and barriers to successful implementation (N = 4)
Frequency (Percentage)
What are the maior factors that support practice # I
4 (100.0%)
• Support from the community
4 (100.0%)
• Support from county government officials
4
(100.0%)
• Support from the mental health system
4 (100.0%)
• Support from substance abuse svstem
4 (100.0%)
• Support from health svstem
4 (100.0%)
Support
from
families
and
children
•
3 (75.0%)
• Support from child welfare workers
(75.0%)
3
• Support from management
3 (75.0%)
• Support from foster families
3 (75.0%)
• Support from contract a,zencies
3 (75.0%)
• Support from the criminal iustice system
(75.0%)
3
• Evaluations show that practice is successful
(50.0%)
2
• Support from state government officials
(25.0%)
1
• Stable funding
(25.0%)
1
• Support from other out-of-home placement providers
What have been the barriers to successful implementation of
this practice?
3 (75.0%)
• Funding is not stable
1 (25.0%)
• Insufficient suooort from state government officials
• Evaluations are inconclusive as to the success of the
1 (25.0%)
practice
154
Organizational or Structural Practices (N = 14)
Fourteen respondents identified organizational or structural practices as their most
effective practices for children and families of color in the CWS. The fourteen
organizational/structural practices fell into three categories: 1) practices related to cultural
competency of staff (n = 10, 71.3%), 2) collaboration with American Indian tribes (n = 3,
21.3%), and 3) vertical case management (n = 1, 7.4%).
Table 55: Or anization or structural
Collaboration with American Indian Tribes
Vertical case mana ement
3
1
Qualitative data on cultural competency practices are presented in Table 56. The target
population for these services varies, some are focused on the Latino population or the Spanish
speaking population, while others are focused on all children in the system and one is focused on
social workers themselves. Practices related to cultural competency involve the use of
bilingual/bicultural staff or competent translators in the delivery of services, as well as providing
information on civil rights, providing training to social workers and matching clients and
workers on ethnicity/culture.
Many respondents mentioned the importance of having bilingual/bicultural staff and the
need to be able to communicate with clients in their own language as reasons for why these
practices are effective with children and families of color. The importance of assessing and
delivering services in the clients' language is described as resulting in more accurate
assessments, more meaningful services and better connections between worker and client.
Cultural competency trainings for social workers are described as bringing awareness and
knowledge to staff.
Table 57 provides qualitative data on organizational or structural practices related to
collaboration with American Indian tribes. These practices are targeted to Native American
children and families and generally include multidisciplinary teams involving tribal
representatives, legal professionals and child welfare staff who come together to better serve
Native American children and ensure the proper implementation of the Indian Child Welfare
Act.
Practices related to collaboration with American Indian tribes were described as effective
for children and families of color because they respect cultural heritage; encourage cultural
awareness for Native American children; result in more complete assessments; better service
planning and fewer contested hearings.
· Table 58 provides qualitative information on vertical case management. This practice is
targeted toward all clients of the child welfare system and it involves having social workers carry
the same cases all the way through the system. This practice was described as effective with
children and families of color because there is consistency in the case and a better relationship
155
between worker and client. It was also noted that this practice is effective with the overall
·
population as well.
Quantitative data from survey responses is provided in Table 59. Most
organizational/structural practices take place throughout the child welfare system (n = 8, 57.1 %)
and are targeted to children of all ages (n = 12, 85.7%). Many are targeted to families of any
ethnicity (n = 7, 50.0%), while some are targeted to Latino families (n = 4, 28.6%), or American
Indian families (n = 3, 21.4%). Additionally, families with substance abuse problems are also a
focus of these practices (n = 7, 50.0%), as are families with mental health problems (n = 7,
50.0%). Goals of the practice include preventing an out-of-home placement (n = 7, 50,0%);
ensuring a permanent placement (n = 7, 50.0%); reunifying children and families (n = 6, 42.9%);
addressing substance abuse problems of families (n = 6,42.9%); and improving parenting skills,
(n = 6, 42.9%).
Information on the history of the practices is presented in Table 60. Most practices were
first implemented in response to a particular problem (n = 7, 50.0%); and the director of the child
welfare system was most frequently mentioned as the person who first initiated the practice (n =
7, 50.0%), followed by child welfare workers (n = 6, 42.9%). The year the practice was first
implemented range from 1972 to 2002 and 46.2% of counties indicated that the practice had
changed significantly since it had first been implemented.
Table 61 provides data on factors related to implementation. Most practices are
permanent (n = 12, 92.3%) and the majority are funded through federal dollars (n = 8, 57.1 %).
Most counties also indicated that these practices are mandated (n = 8, 61.5%), and 66.7% (n = 8)
of counties indicated that social workers are provided with training in how to implement the
practice. Most counties do not contract with another agency to deliver the practice (n = 9, 64.3%)
and most also indicated that the practice is widespread throughout the child welfare system (n =
12, 92.9%). Only one county (7.1 %) indicated that there is a specific geographic area in which
the practice is targeted. Four counties (28.6%) reported that the practice has been evaluated to
determine its effectiveness and of those 66.7% (n = 2) reported that the practice was found to be
effective with the target population. Of those counties in which no evaluation has been
conducted, 50.0% (n = 5) indicated that an evaluation was not needed.
Factors that support the practices and barriers to successful implementation are presented
in Table 62. Support from child welfare workers (n = 11, 78.6%) was most commonly noted as a
major factor that supports the practice, followed by support from management (n = 9, 64.3%);
county government officials (n = 8, 57.1 %); families and children (n = 7, 50.0%); and the
community (n = 7, 50.0%). No barriers to successful implementation were noted by 28.6% of
counties (n = 4), while 14.3% (n = 2) reported insufficient support from the community and
14.3% (n = 2) noted insufficient support from child welfare workers.
156
Table 56: Organizational or structural practices: Practices related to cultural competency (n = 10)
Why effective with children and families of color
Description of practice
Target
Name of practice
population
One of things that has made it particularly
SpanishWe have a very high rate of return for
The use of
Hispanic children--one of our better
effective in this county is the size of the county.
competent
speaking
practices is the use of competent translators And our translator even though they may not be
translators
clients
and always having this available when
CW social workers they are from the community
and they are familiar with the community and the
working with a family. These are used in
the juvenile court as well. And we also have dynamics of that particular culture. For the most
part this had enhanced the practice.
bilingual staff as well.
Availability of
bilingual/bicultural
staff to serve the
Latino population.
Latino
population
At least one
bicultural/bilingual
worker in each unit
Mostly
Hispanic
children
B ilingual/B icultural
social work staff.
Spanish
speaking
population
Cultural knowledge and ability to converse in same
Hire and place bilingual/bicultural social
workers at office serving majority of Latino language.
families. Provides culturally sensitive
services to families, better assuring accurate
asessments of risk and safety for children.
When you are doing social work and having to
The agency really values language and
trying to accommodate that the best we can. assess and discuss very difficult things with people
Most of our receptionists are bilingual and
having to do with emotions and feelings these
things are very difficult to do through translation
our social service aides and social worker
because a lot of what a social worker does is their
I's are bilingual. Whenever we start to
use of self and being able to read body language
implement a program we always look to
bringing in bilingual/bicultural staff. We are and intonation and understand the subtlety of
limited in our resources but we do our best. language and what someone is saying. You are less
effective through translation. The communication
We also really try to ask this of our
is much better when services are conducted in the
providers as well.
clients language. And bicultural folks understand
the local jargon.
Ability to communicate with the families in their
Bilingual/bicultural staff assigned to
native language & an ability to understand the
bilingual/monolingual cases to ensure_
appropriate communication of issues, needs, culture helps the social worker provide appropriate
services and agency function/process.
resources & guidance to the family.
157
Table 56 continued: Organizational or structural practices: Practices related to cultural competency (n = 10)
Target
Name of practice
Why effective with children and families of color
Description of practice
population
B ilingual/B icultural Families
Recruiting and hiring bilingual/cultural staff Recruitment is a priority within service delivery
system and best practice philosophy .. .it becomes
staff
whose
to :work with the diverse population we
children
serve. Staff are assigned throughout our
integrated in every day work.
have been
child welfare programs/services.
Assessment of language and cultural
abused.
staffing needs is determined on an ongoing
basis and is used as a criteria to fill vacant
positions.
Federal Division 21
Regulation-Civil
rights
All
Emphasis of
Cultural
Competence
All cultures
We inform all clients of their civil rights.
This includes that they cannot be
discriminated against on the basis of
national origin, disabilities, race, age sex,
all those areas. We are supposed to talk to
all families about their civil rights. We ask
the family if they want a special skills
worker assigned-that would specialize in a
particular language or culture. We have
Black culture workers, Hispanic cultural
workers, and Asian, Russian special skills
workers. We ask all clients if they want a
special skills worker-this is for
language-but also culturally matching the
clients.
Our agency has a cultural competence
committee. We mandate 8 hours of training
for staff in orientation and an additional 16
hrs through our leadership academy.
Matching families with a social worker of their
ethnic background is helpful in getting families
back together. If the social worker understands the
culture-like ifthere are different discipline
techniques that are seen more in one culture, the
social worker may be able to help the child and
family more. Clients are a bit more comfortable,
social workers can confront them better-work
with the better to help bring about change.
Awareness, Knowledge
158
Table 56 continued: Organizational or structural practices: Practices related to cultural competency (n = 10)
Target
Name of practice
Description of practice
Why effective with children and families of color
population
Policy is to align
Applies
Our practice is to assign staff of similar
Ability to establish trusting relationship
case assignments to across all
ethnicity and we are able to do so most
staff from like
cws
often in our African American and Mexican
ethnicity when
components American population. If unable to do so,
we consider assigning a spanish speaking,
possible
non-Mexican-American. staff.
Cultural
Social
We require all social workers to attend
Missing
competence
workers
training on cultural competency on a regular
training
basis.
Missing
Civil Rights
Missing
Training- bringing awareness to staff
Training
159
Table 57: Organizational or structural practices: Collaboration with American Indian tribes (n = 3)
Why effective with children and families of color
Name of practice
Target
Description of practice
population
It respects their cultural background and keeps
Indian Child
American
A practice in which tribal representatives,
Indian
agency staff, kids' attorneys, County
them connected with their tribal roots, it respects
Welfare Act
that people can do things differently and still have
(ICWA)
children and Counsel, and others come together on a
monthly
basis
to
talk
about
issues
related
to
good outcomes. The tribe is actively solicited to be
Roundtable
families
involved with children and families in the CWS.
the implementation of the ICWA, how we
We pay for culturally appropriate services, like
can work to together on certifying tribal
tribal dances, and other non-traditional services.
homes, how to get services for American
Indian Children and what are the most
appropriate services. The group created a
handbook ofICWA regulations. Very
effective in opening dialogue between the
parties, sensitizing staff to issues around
services that would be more culturally
appropriate to American Indian children,
identifying and providing services, and
ensuring that kids can get to tribal homes.
Native
Indian Child Welfare Act Multidisciplinary More complete assessments, better service
ICWA
planning and less contested hearings.
Team
American
Native American families-in general they are
Some of the tribes have strong cultural
Collaboration with
Native
Americans
centers with health care and counseling and still at war with both the US and with each other. It
tribes and tribal
is effective because in spite of these things we are
so we try to work hard with them. Native
courts
able to overcome the barriers and we are able to
American children do best when placed
make use of the real resources of cultural
with a Native American family. The Tribal
court makes placement decisions and so we awareness for Native American people.
work with them to make decisions that we
Collaborating with the tribes depends on having the
are all in agreement about. We work closely right social worker work with the tribes-in
with the tribes in Native American cases.
general the tribes have been cooperative with the
cws.
160
Table 58: Organizational or structural practices: Vertical case management {n = 1)
Name of practice
Target
Description of practice
Why effective with children and families of color
population
Social Workers carry the same cases all the There's consistency-a lot of it is when cases are
Social workers
All
carry a generic case
way through the system A lot is lost when a transferred from one person to another there may
load-vertical case
case is transferred from worker to worker.
be different interpretations of the situation.
management
Families always know who their social
Effective with overall population-it would be
workers are and there is ownership
unfair to say that what we are doing is more
advantageous with group than another.
throughout the life of the case. There is a
better work product. There's a relationship
between the worker and clients.
161
T abl e 59 0 rgamzattona or structural practices: Location, tan et population and goals (N = 14)
Frequency (Percentage)
Where does practice #1 take place?
(57.l %)
8
• Throughout the child welfare system
County
child
welfare
office
(50.0%)
7
•
(42.9%)
6
• Birth family's home
Foster family's home
(42.9%)
6
(42.9%)
Court
6
5
(35.7%)
• Another out-of-home placement site
(14.3%)
2
• Contract agency site
5 (35.7%)
• Other
Who is the target population for practice # 1?
(85.7%)
12
Children of all ages
(7.1%)
1
Very
young
children
(0
to
3
years)
•
(7.1%)
I
• Young children (4 to 10 years)
(50.0%)
7
Families
of
any
ethnicity
•
(28.6%)
4
Latino families
(21.4%)
3
American Indian families
(50.0%)
7
• Families with substance abuse problems
(50.0%)
7
• Families with a mental health problem
(42.9%)
6
• Families with domestic violence
(28.6%)
4
• Families with income problems
(21.4%)
3
• Families with a disability
(7.1%)
1
Other
What are the purposes and goals of practice #1?
(50.0%)
7
• Prevent out-of-home placement
(50.0%)
7
• Ensure a permanent placement for the child
(42.9%)
6
Reunify children with their birth families
(42.9%)
6
• Address substance abuse problems of family
(42.9%)
6
• Improve parenting skills
(35.7%)
5
Assess
level
of
risk
•
(35.7%)
5
• Improve psychosocial functioning of family
(35.7%)
5
Address
mental
health
problems
of
family
•
5 (35.7%)
• Address housing problems of family
(28.6%)
4
• Address socioeconomic problems of family
(57.1%)
8
Other
•
•
•
•
•
•
•
•
162
Table 60: Organizational or structural practices: History of practice (N = 14)
Frequency (Percentage)
Why was practice #1 first implemented?
7
(50.0%)
In response to a particular problem
(21.4%)
3
• New policies were implemented
1
(7.1%)
• Funds became available·
6
(42.9%)
• Other
Who first initiated practice # 1?
(50.0%)
7
• Director of the child welfare system
(42.9%)
6
• Child welfare workers
(7.1%)
1
Clients of the CWS
(7.1%)
1
• Foster families
(7.1%)
1
• Criminal justice system
1
(7.1%)
• Contract agencies
(28.6%)
4
• Other
When was practice #1 first implemented?
(7.1%)
1
• 2002
2 (14.3%)
2000
(7.1%)
1
• 1999
2
(14.3%)
• 1998
(7.1%)
1
• 1995
(7.1%)
1
• 1980
(7.1%)
1
1972
•
Has practice # 1 changed significantly between the time it
was first implemented and now?
(46.2%)
6
• Yes
•
•
•
163
Table 61: Organizational or structural practices: Factors related to implementation (N = 14)
Frequency (Percentage)
Current implementation status of practice # 1
12 (92.3%)
• Permanent
Other
1
(7.7%)
What is the primary funding source for practice # 1?
8
(57.1%)
• Federal funds
(42.9%)
6
State
funds
•
5 (35.7%)
• County funds
(28.6%)
4
• Other
Is practice #1 mandated?
8
(61.5%)
• Yes
Are social workers provided with training in how to
implement practice # 1?
(66.7%)
8
• Yes
Do you contract with another agency or system to deliver
practice # 1?
(64.3%)
9
• No
(21.4%)
3
Combination
of
internal
and
external
resources
•
(14.3%)
2
• Yes
Is practice # 1?
(92.9%)
13
• Widespread throughout the child welfare system
(7.1%)
1
Targeted
to
a
relatively
small
group
of
clients
•
Is there a specific geographic area or community in which
this practice is targeted?
(7.1%)
1
• Yes
Has practice # 1 been evaluated to determine its
effectiveness?
4
(28.6%)
• Yes
If yes, what were the results ofthe evaluation?
(66.7%)
2
• Effective with target population
• Evaluation did not provide evidence that practice was
1
(33.3%)
effective
If practice # 1 has not been evaluated, why not?
(35.7%)
5
• An evaluation was not considered to be needed
(21.4%)
3
• Insufficient resources
(7.1%)
1
Insufficient
time
•
(7.1%)
1
• Other
•
164
Table 62: Organizational or structural practices: Factors that support practice and barriers to
successful implementation (N = 14)
Frequency (Percentage)
What are the major factors that support practice # I
11 (78.6%)
• Support from child welfare workers
9 (64.3%)
Support
from
management
•
8 (57.1%)
• Support from county government officials
(50.0%)
7
• Support from families and children
7 (50.0%)
Support from the community
6 (42.9%)
• Support from foster families
6 (42.9%)
Support
from
state
government
officials
•
5 (35.7%)
• Support from the mental health system·
4
(28.6%)
• Support from the substance abuse system
4 (28.6%)
• Support from contract agencies
4 (28.6%)
• Support from the health system
(28.6%)
4
Stable
funding
•
3 (21.4%)
• Support from other out-of-home placement providers
(14.3%)
2
• Evaluations show that practice is successful
2 (14.3%)
• Support from the criminal justice system
2 (14.3%)
Other
What have been the barriers to successful implementation of
this practice?
-4 (28.6%)
• No barriers
2 (14.3%)
• Insufficient support from the community
2 (14.3%)
• Insufficient suooort from child welfare workers
(7.1%)
I
• Funding is not stable
(50.0%)
7
Other
•
•
•
165
Summary of Key Findings and Implications:
Descriptive Results
Respondents to the California statewide survey identified practices in four overall areas
as most effective for children and families of color: 1) beginning choice point practices, 2)
continuing or ending choice point practices, 3) both beginning and continuing/ending choice
point practices, and 4) organizational or structural practices.
·
Beginning Choice Point Practices
The beginning choice point practices identified by respondents fell into three categories:
1) family group conferencing or other team decision-making practices, 2) voluntary family
maintenance or other prevention/family preservation services and 3) risk assessment tools or
systems.
In general, results from the California statewide survey indicate that beginning choice
point practices are most often delivered at the county child welfare office, and they are targeted
to children ofall ages and ethnicities. Most of these practices are aimed at preventing an out-of-
home placement, assessing level of risk or improving the psychosocial functioning of the family.
Over half of these practices have been evaluated by individual counties to determine their
effectiveness, and of those that have been evaluated most are considered to be effective with the
target population.
Results ofthe California statewide survey indicate that beginning choice point practices
are considered to result in enhanced outcomes for children and families ofcolor because they
are inclusive, collaborative and objective. Yet very little empirical evidence exists to support the
effectiveness of these practices for children and families of color. No studies have evaluated the
impact offamily group conferencing or other team decision-making practices on children and
families ofcolor. However there is some evidence to suggest that inclusive practice, in which
parents are engaged in placement and other decisions, may result in more placement stability for
children, yet possible differential effects for diverse populations has not been researched
(Palmer, 1996). More research is needed to determine if family group conferencing/other team
decision-making practices and inclusive practice result in enhanced outcomes for children and
families of color.
Additionally, the effectiveness of voluntary family maintenance or other prevention/family
preservation services in preventing out-of-home placements has not been adequately
demonstrated, and the impact of these services for children and families of color has not been
sufficiently addressed in the research literature. It is interesting to note that results from the
California statewide survey indicate that only 15.4 percent of beginning choice point practices
take place in the birth family's home, yet the research literature suggests that home-based
prevention practices may be the most effective voluntary family maintenance practice, especially
for high-risk mothers. This finding may suggest that counties are not utilizing evidence-based
prevention efforts. Research on the reasons for this underutilization of home-based prevention
services is warranted.
166
Lastly, the effectiveness ofrisk assessment tools or systems in accurately assessing risk
for future maltreatment has not been consistently demonstrated in the research literature. Yet
research does indicate that the individual decision-making practices ofchild welfare workers are
inconsistent and subject to biases that may negatively impact children and families of color. As
such, the use of more objective assessments of risk in child welfare is promising, however more
research is needed to identify risk assessment tools and systems that are reliable and valid.
Continuing Care/Ending Choice Point Practices
The continuing care/ending choice point practices identified by respondents fell into
three categories: 1) placement-related practices, 2) the Family to Family Initiative, and 3)
intensive and collaborative interventions for high-risk children.
Results from the California statewide survey indicate that, in general, the majority of
continuing care/ending choice point practices take place throughout the child welfare system
and are targeted to children ofall ages and ethnicities. However, more continuing care/ending
choice point practices are targeted to specific age groups than the other practices identified by
respondents. The goal of most of these services is to reunify children with their birth families,
ensure a permanent placement or prevent and out-of-home placement. Only two respondents
reported that these practices have been evaluated to determine their effectiveness and both of
these evaluations indicated that the practices are effective with the target population.
These continuing care/ending choice point practices are described as effective with
children and families of color because they are culturally appropriate, collaborative, intensive
and they involve clients ' communities. Yet, research on the effectiveness of these services for
children and families of color is lacking. Placement-related practices that emphasize placement
of children in culturally appropriate homes have not been evaluated to determine their impact on
children of color. However, some studies have noted that social workers may be reluctant to
place children of color with White alternate caregivers (Brooks & James, 2003; Carter-Black
2002; Vidal de Haymes & Simon, 2003). Results from the California statewide survey regarding
the identification ofplacement-related practices that emphasize placement of children ofcolor in
culturally appropriate homes do suggest that culture and ethnicity may be important factors in
child welfare workers' placement decisions. This appears particularly true for Native American
children who were mentioned specifically by two counties.
Research on the effectiveness ofthe Family to Family Initiative for children andfami/ies
ofcolor has not been conducted. There is limited evidence to suggest that social network
interventions which seek to involve community members in services are potentially effective in
increasing social networks, improving parenting skills and increasing the likelihood of case
closure (Gaudin et al., 1990). However the Family to Family Initiative encompasses more than
social network interventions and research is needed to evaluate the effectiveness ofthis practice
for diverse populations.
Although children in out-of-home care are at an increased risk for a variety ofpoor
outcomes (Stevenson, 1999), the impact of intensive and collaborative interventions for high-risk
children have demonstrated only mixed results in the research literature. While some studies on
167
wraparound services report that children in these services are less likely to be removed from their
community and may experience a reduction in behavioral problems (Bruns, Burchard, & Yoe,
1995, as cited in Toffalo, 2000; Clark, Lee, Prange, & McDonald, 1996, as cited in Toffalo, ·
2000; VanDenBerg, 1993, as cited in Toffalo, 2000), other studies report no significant benefit
from wraparound services (Clarke, Schafer, Burchard, & Welkowitz, 1992, as cited in Toffalo,
2000). More research is needed to identify aspects ofthese interventions that are most effective,
especially for children ofcolor
Both Beginning and Continuing/Ending Choice Point Practices
Practices related to both beginning and continuing/ending choice points fell into two
categories: 1) community-based services, and 2) strength-based assessments.
Results from the California statewide survey suggest that many practices that are both
beginning and continuing/ending choice point practices take place at a contract agency site or
the birth family's home. These services are targeted to children of all ages and any ethnicity.
These practices have numerous goals including: preventing an out-of-home placement;
reunifying children with their birth families; assessing level of risk; improving psychosocial
functioning of families; improving pare~ting skills; addressing substance abuse problems;
addressing mental health problems; addressing socioeconomic problems; and addressing housing
problems. All four of the respondents indicated that these practices have been evaluated and 50
percent of the evaluations indicated that the practices are effective with the target population.
Respondents to the California statewide survey described these practices as effective with
children and families of color because they are easy for clients to access; they are communityspecific, and non-intrusive. These services have broad goals that focus on both prevention and
intervention for families in contact with the CWS and research on the impact of these services on
children and families of color is lacking. In general, research on prevention and intervention with
child welfare populations does not provide convincing evidence of their effectiveness. However,
because these practices are more of an approach to services, rather than services themselves,
more research is needed to determine if these approaches are effective with ethnically diverse
populations.
Organizational or Structural Practices
Organizational or structural practices/ell into three categories: 1) cultural competency
ofstaff, 2) collaboration with American Indian tribes, and 3) vertical case management.
According to responses from the California statewide survey, most
organizational/structural practices take place throughout the CWS; the majority are targeted to
children ofall ages and ethnicities, although some are specifically targeted to Latino or Native
American clients. The primary goals of these practices include preventing an out-of-home
placement and ensuring a permanent placement. Approximately one quarter of respondents
indicated that these practices have been evaluated to determine their effectiveness and of these,
approximately 66 percent reported that the practices are effective with the target population.
168
These practices are described by respondents to the California statewide survey as
effective with children and families ofcolor because they result in better communication and
relationships between worker and client; they respect cultural heritage; increase cultural
awareness and produce more accurate and complete assessments and services. Although
cultural competency appears intuitively beneficial for ethnically diverse children and families,
research on the impact of culturally competent staff and programs in producing successful
outcomes for children and families of color has not been conducted. There is evidence to suggest
that for Latino clients. in particular, language barriers can exclude them from necessary services
and produce significant delays in service delivery; and researchers have also discussed the
importance of language in accurate assessment and treatment of clients (Suleiman, 2003). As
such, culturally and linguistically appropriate services appear to increase service use and
accuracy, although their impact on client outcomes is unclear.
Research on the effectiveness ofcollaboration with American Indian tribes and vertical
case management practices on outcomes for children and families of color is limited Effective
collaboration between the CWS and American Indian tribes appears intuitively beneficial for
Native American children, however more research is needed to identify aspects ofeffective
collaboration and how this collaboration impacts outcomes for Native American children. A
vertical case management model of child welfare service delivery may result in enhanced
outcomes. For instance, in South County Offices, where a vertical case management model is
used, families are more likely to have their cases closed with family maintenance services than
they are in the Main County Office where a vertical case management model is not used (please
see Section IV for a full discussion). These results suggest that a vertical case management
model may be an effective child welfare practice.
Limitations and Conclusions
Descriptive results of the statewide survey must be considered within the context of
certain limitations. For instance, the survey instrument was designed specifically for this study
and so its reliability and validity is not known. The self-report nature of the data also presents the
possibility of personal biases in survey responses and since the professional titles of the
respondents varied from directors, to managers to supervisors, there is the possibility that
different perspectives and information may occur as a result of these differing job titles.
Additionally, although the response rate was high (77.6%), there is the possibility that the sample
may be biased due to some counties refusing to participate.
Overall, respondents described beginning choice point practices as effective with
children and families ofcolor because they are inclusive, collaborative and objective.
Continuing care/ending choice point practices were considered effective with children and
families of color because they are culturally appropriate, collaborative, intensive and involve
client's communities. Both beginning and continuing/ending choice point practices were
described as effective because they are easy for clients to access, they are community-specific
and non-intrusive. Lastly, organizational or structural practices were considered effective with
children and families ofcolor because they are thought to result in better communication and
relationships because worker and client, they respect cultural heritage, increase cultural
awareness and produce more accurate and complete assessments and services.
169
In general, results from the California statewide survey suggest that there is a significant
gap between research on child welfare practices and the best practices for children and families
of color being used in California county child welfare systems. Little systematic research has
evaluated the impact of many of the most effective practices for children and families of color
that were identified by respondents Although the research literature has not yet provided
convincing evidence of the effectiveness ofmany of the practices identified by respondents,
respondents to the statewide survey felt that these practices were their most effective with
children and families ofcolor. These findings suggest that some counties may be utilizing
promising practices for children and families ofcolor in the child welfare system, and more
rigorous evaluations of these practices is needed.
170
Multivariate Analysis:
Impact of Statewide Practices on Differential Representation
The statewide survey identified practices, which addressed disproportionate
representation of children of color and were considered effective by county child welfare
personnel. Although perceived as effective, few of the practices had been evaluated to determine
effectiveness from an independent point of view and in a controlled manner. Child welfare
personnel nominated them as being effective, but in most cases their impact had not been
established through systematic empirical research.
For present purposes, the impact of interest is how the practices identified through the
survey might affect differential representation ofchildren ofcolor. Some of the practices had
been implemented with children of color in mind and others had not, but all were considered by
child welfare personnel to be especially valuable in addressing the needs and circumstances of .
children of color. If child welfare personnel are correct in their belief that the practices serve
children of color particularly well, then it is reasonable to believe that the practices identified in
the survey will divert children of color from the child welfare system or will reduce the scope or
duration of their involvement.
By helping children of color to follow pathways around full-fledged involvement in the
child welfare system or reducing the intensity or length of child welfare participation, the
practices identified in the survey can be expected to reduce differential representation. We can
expect lower levels of overrepresentation if the practices are effective.
Statewide data are available from sources, which, in conjunction with data from our
survey, permit the degree of empirical association to be determined between practices and
disproportionate representation. The U.S. Census and CWS/CMS publish data, available for use
by researchers and others, describing California counties on the worldwide web. The web-based
data provide an opportunity to evaluate whether there is a link between disproportionate
representation and use of perceived-to-be-effective practices.
If a study is statewide in scope like the present one, then it furnishes a basis to go beyond
the experience of a single county. By comparing groups of counties and contrasting those that
implemented a practice with those that did not, our investigation encompasses child welfare
systems around the state. The results point to promising practices identified from a statewide
perspective.
In this section, we present results from several analyses linking characteristics of
practices presented previously with disproportionate representation of children of color in the
child welfare system. Our intention was to identify types of practices, practices, and practice
characteristics-distinctions to be explained later-- associated with levels of disproportionate
representation, and which therefore might be more effective in reducing disproportionate
representation of children of color.
171
Multivariate Methods
The unit of analysis for this study was the county. The method, broadly conceived, was to
conduct controlled tests of empirical associations between types of practices, practices, and
practice characteristics on the one hand, and levels of disproportionate representation on the
other hand.
For thinking about the study's approach in greater detail, it is useful to begin by
considering the meaning of disproportional representation. Disproportional representation is
said to occur when the proportion of children of color in a county's child welfare system is
higher or lower than expected. Typically, documenting the proportion of African Americans,
Latinos, or Asian American/Pacific Islander in the population at large usually establishes the
proportion that is expected. The difference between the representation of a group in the
population at large and its representation in the child welfare system defines its level of
disproportional representation.
In our study we calculated a measure of this difference using regression analysis. We
analyzed a group's representation in the child welfare system after statistically adjusting for it
representation in the population at large. Following a research design like ours, a statistic results
from regression which defines the difference between expected and actual levels of child welfare
representation.
Disproportional representation is affected by factors other than those under the direct
control ofchild welfare authorities. The stresses under which minority families live and the
resources available to them for example, can contribute to disproportionate representation.
Counties can differ in the levels ofstress, resources, and similar differentiating characteristics.
Child welfare systems too are diverse. Some are larger and more centralized than others, for
example; policies and proc~dures vary.
It is difficult to establish definitively whether practices are effective because ofthe
differences apart from the practices between counties and child welfare systems associated with
disproportionate representation. When a link is observed between a practice and
disproportionate representation, it remains possible that a characteristic other than the practice
itself is responsible for that link. The observed link between a practice and disproportionate
representation might be coincidental. Controlling/or key differences between counties and child
welfare systems reduces the chances of associations occurring through coincidence.
We examined many indicators describing counties and their child welfare systems in
order to identify differences that were especially important to take into account. Among
indicators we screened were population size and density, wages and costs of goods and services
in the county, the geographical location of the county, size of the child welfare system and
proportion of the child welfare population in kincare. We sought key indicators from a large
number we considered reflecting differences between the counties and their child welfare
systems and linked to disproportionate representation.
172
Through this process we found two indicators that were consistently and strongly
associated with disproportionate representation: urban-rural status and per-capita income.
Counties that were urban and had higher per-capita income were notably more likely to have
higher levels ofdisproportionate representation.
We used these data to conduct a controlled analysis of the association between practice
types, practices, and practice characteristics by constructing regression equations. In each
equation, the proportion of Latino, African American, or Asian American/Pacific Islander
children served as the dependent variable. The proportion of Latinos, African Americans, or
Asian American/Pacific Islander in the county was entered first as an independent variable, as
discussed above, in order to subtract an expected level of representation from the actual
representation of each group in each county's child welfare system. We then added to the
equation each county's urban-rural status and per-capita income. Finally, we added each practice
type, practice, or practice characteristic suitable for analysis.
This enabled us to determine how much each practice type, practice, or practice
characteristic was associated with greater or lower level of representation apart from the level of
representation expected from its group representation in the county, and after adjusting for the
county's urban-rural classification and its per-capita level of income. Thus, the approach
adjusted for the three key factors: group representation in the population, whether the county was
classified as urban or rural, and per-capita income of county residents.
In most cases the three control variables alone explained disproportionate representation
in a statistical sense extremely well. Success in accounting/or differences between counties in
statistical models implies that many important differences between counties were explained and
that the possibility that associations arose by coincidence was low.
We considered practice types, practices, and practice characteristics in a three-phase
analysis. Phase 1 addressed our organization ofpractices into beginning choice point practices,
continuing care/ending choice point practices, organizational or structural practices, and both
beginning and continuing/ending choice point practices. We considered disproportionate
representation according to whether practices were beginning, continuing/ending, organizational
or structural, or both beginning or ending.
In phase 2, we considered individual practices in an attempt to consider the impact of
practices on disproportionate presentation in finer-grained detail. We were hampered in our
attempt: very few counties reported instituting certain of the practices.
Ultimately, we chose to consider at the individual practice level only those where four or
more counties indicated implementing them. Four counties is a small number for this purpose
and limit our ability to detect genuine differences that might have been present at statistically
significant levels. It proved fruitful however, and provides an important basis for discussion of
individual practices. The five practices we considered were: family group conference or other
team decision making practices (n=7), voluntary family maintenance or other prevention/family
preservation services (n=4), placement related practices (n=6), family to family initiative (n=4),
and cultural competency of staff (n= 10).
173
We asked parallel questions in all phase I and phase 2 analyses. Adjusting for urbanrural status and per-capita income, for Latinos, African Americans, and Asian Americans how is
disproportionate representation affected by type of practice? What was the impact on
disproportionate representation ofwhether a practice was one offive specific practices: family
group conference or other team decision making practices, voluntary family maintenance or
other prevention/family preservation services, placement related practices, family to family
initiative, and cultural competency of staff?
In phase three of our analysis we asked questions without regard to type ofpractice. We
considered the impact on disproportionate representation-whatever the particular practice
type-- ofdifferences in practice location, target population, purposes, reasons for
implementation, initiating persons and agencies, implementation timing, implementation status,
location and scope of the practice, geographic and community targeting,. evaluation, support,
and barriers.
There are many practice characteristics describing practices. To reduce the number of
characteristics for purposes of analysis, we clustered characteristics into logically related,
empirically associated groupings using factor analysis, a statistical procedure for clustering
individual questions to which respondents have given similar answers. When a question was
discovered not to fit well within factor-analysis created clusters of questions, we eliminated it
from the cluster and considered it in subsequent analysis individually.
We then conducted regression analyses like those conducted in phases 1 and 2. The
question was: adjusting for urban-rural status ahd per-capita income, for Latinos, African
Americans, and Asian Americans, how is disproportionate representation affected by practice
characteristics? What was the impact on disproportionate representation of differences in
practice location, target population, purposes, reasons for implementation, initiating persons and
agencies, implementation timing, implementation status, location and scope of the practice,
geographic and community targeting, evaluation, support, and barriers?
There are two ways in which practice types, practices, and practice characteristics can
be linked to disproportionate representation. One is a negative association between the
variables, indicating lower than expected overrepresentation. Negative association means that
implementing counties have relatively low or nonexistent minority overrepresentation and
suggests that the practice type, practice, or practice characteristic is successful.
Another kind ofassociation between a practice type, practice, or practice characteristic
and overrepresentation is a positive association. Positive association between variables means
that there is greater than expected overrepresentation associated with the practice type, practice,
or practice characteristic-that overrepresentation is relatively high.
Practices linked to greater overrepresentation cannot be considered successful. How does
a practice come to be linked to greater overrepresentation? It is likely that positive association
between a practice and overrepresentation comes about because instead of reducing
overrepresentation, some practices are implemented reactively, that is, in response to
174
overrepresentation. These practices do not reduce overrepresentation, although they may do so in
the future if given sufficient time. At present, we note only that a significant positive association
with overrepresentation is probably reactive and does not indicate success.
Multivariate Results
Latino Overrepresentation
Latinos were overrpresented in about 60% of counties and the degree of their
overrepresentation was as high as 28 percentage points.
Practice types. For Latinos, there was a statistically significant difference according to
whether practices were beginning choice point practices, continuing care/ending choice point
practices, organizational or structural practices, or both beginning and continuing/ending choice
point practices.
Both significant differences involved beginning choice point practices. Beginning choice
point practices were associated with about 8 percentage points less overrepresentation than both
beginning and continuing/ending choice point practices and about 9 percentage points less
overrepresentation than organizational or structural practices (See Table 63).
Practices. One statistically significant difference emerged when we considered the five
individual practices that could be analyzed. Practices .involving family group conference or other
team decision making were associated with 7 percentage points less overrepresentation than were
practices involving culturally competency of staff (See Table 63).
Practice characteristics. For Latinos, there were statistically significant difference in
target population, purposes, initiating persons and agencies, and implementation timing (See
Table 64). Latino overrepresentation was greater when areas were targeted for intervention (area
where most referrals came from; with low socioeconomic status residents; with high crime; with
large ethnic representation), when clients, foster families, and out of home placement providers
were involved in the development of the practice, when service systems were involved in the
development of the practice (mental health, substance abuse, health, criminal justice, contract
agencies). Overrepresentation was less also when "other" problems were addressed as the
purpose of the practice and when it began more recently.
African American Overrepresentation
African Americans were overrepresented in about 64% of counties and their degree of
their overrepresentation was as high as 65 percentage points.
Practice types. For African Americans, there was no statistically significant difference in
overrepresentation according to whether practices were beginning choice point practices,
continuing care/ending choice point practices, organizational or structural practices, or both
beginning and continuing/ending choice point practices (See Table 63).
175
Practices. One statistically significant difference appeared when we considered the five
individual practices, which could be analyzed. Family to Family initiatives were associated with
8-percentage points greater overrepresentation than were the remaining practices considered (See
Table 63).
Practice characteristics. For African Americans, there were differences in target
population, purposes, and initiating persons and agencies, and geographic and community
targeting. African American overrepresentation was less when areas were targeted for
intervention (area where most referrals came from; with low socioeconomic status residents;
with high crime; with large ethnic representation), ethnic minority. populations (Latino, African
American, Asian American/Pacific Islander, American Indian, families of any ethnicity), and
when very young children were targeted. African American overrepresentation was less when
foster families and clients or contract agencies first initiated the practice. African American
overrepresentation was less also when the purpose the intervention was to improve the
psychosocial functioning of the birth family (See Table 64).
Asian American/Pacific Islander Underrepresentation
In no counties were Asian American/Pacific Islanders overrerpresented. Instead, they
were equally represented in a few counties and in most counties Asian American/Pacific Islander
were underrepresented.
Practice types. For Asian American/Pacific Islanders, there was no statistically
significant difference in overrepresentation according to whether practices were beginning
choice point practices, continuing care/ending choice point practices, organizational or structural
practices, and both beginning and continuing/ending choice point practices (See Table 63).
Practices. Nor was there a statistically significant difference when we considered the five
individual practices that could be analyzed (See Table 63).
Practice characteristics. For Asian American/Pacific Islanders, there were differences in
target population, implementation timing of the practice. Asian American/Pacific Islanders under
representation was greater when areas were targeted for intervention (area where most referrals
came from; with low socioeconomic status residents; with high crime; with large ethnic
representation), ethnic minority populations (Latino, African American, Asian American/Pacific
Islanders, American Indian, families of any ethnicity). Asian American/Pacific Islander under
representation was less when system management and child welfare workers expressed support
for the intervention, and when it had begun earlier (See Table 64).
176
Table 63: Practice Types, Practices, and Disproportionate Representation
Pracfice Types
Ethnic Group
Variable
B SE Beta
T
Latino
Beginning vs
.09 .04 .14
1.98*
Organizational/
Structural
Beginning vs
.08 .03 .20 2.61 **
Beginning/Continuing
Practices
Variable:
B
SE Beta T
Family Group Conference/ .07 .03 .15
2.13*
Team Decision Making vs
Cultural Competence
African American Family to Family vs. other -.08 .04 .13
2.23*
Ethnic Group:
Latino
Table 64: Practice Characteristics and Disproportionate Representation
Ethnic Group:
Latino
Variable:
Area Targeting
Target:Other Problems
Clients Involved in
Development/Evolution
Systems Involved in
Development/Evolution
Initated Earlier
African American Ethnic Targeting
Area Targeting
Targeting: Very young
Foster family-client
Initiated
Contract agency
Initiated
Goal: Improve psychoSocial Functioning Birth
Family
Asian American/ Ethnic Targeting
Pacific Islander
Area Targeting
Management-Worker
Support
B
.04
-.05
.03
SE Beta T
.01 .19
3.07**
.26 -.13 -2.06*
2.77**
.01 .17
.03
.01
.16
2.61 **
-.00
-.04
-.02
.06
-.08
.00
.01
.01
.03
.04
-.16
-.23
-.12
.11
-.11
-2.69**
-4.20**
-1.93*
1.92**
-1.94**
-.08 .04 -.13
-2.11 *
-.05
.02 -.13
-2.11 *
.00
.00 .22
1.99*
.00
-.01
.00 .19
.00 -.29
1.90*
-3.24**
Notes:
Regressions: Ordinary Least-Squares. All models R >.80 (p>.01).
Factors= Principal Components, 1 factor per analysis, Eignenmvalue> 1. Regression-based factor scores used in
subsequent regression. Variables with loadings >.50 were eliminated from analysis and analyzed separately in
subsequent regression.
177
Summary of Key Findings and Implications:
Multivariate Analysis
When considering the 43 counties responding to the survey of practices perceived
effective in working with children of color, significant associations were uncovered between
practice types, practices, and practice characteristics on the one hand, and disproportionate
representation on the other. These associations remained after adjusting for urban rural status and
per-capita income, differences between the counties strongly linked to minority child welfare
representation. Some point to effective practices-practices that bring about minority
underrepresentation. Different patterns ofassociation were found for Latinos, African
Americans, and Asian American/Paci.fie Islanders.
For Latinos, beginning choice point practices had a beneficial effect on
overrepresentation. In particular, family group conference or other team decision-making
interventions predicted lower levels ofoverrepresentation.
Bringing together interested parties early on in child welfare involvement may divert
cases from child welfare or otherwise minimize involvement, perhaps by facilitating
communication and mobilizing otherwise overlooked resources. Conceivably, the intervention is
especially valuable for Latinos in that its family focus is congruent with cultural values, and
because of the intervention's potential to overcome linguistic and cultural barriers to
understanding.
Notably, these interventions do not target at-risk neighborhoods as a special focus of
concern. Indeed, Latino overrepresentation was greater, not less, when interventions were
implemented targeting areas with high concentrations ofdistressed families.
Other findings indicated that practices targeting at-risk populations came about when
clients, foster families, and out of home placement providers were engaged in development of
the practice, along with mental health, substance abuse, health, criminal justice, contract
agencies. It appears that when they perceived a need and had an opportunity to do so,
stakeholders mobilized and focused on high need areas and populations as intervention targets.
However, they appear do this in response to Latino overrepresentation, such that
implementation is linked to greater overrepresentation. The positive association at-risk targeting
and disproportional representation indicates that these practices do not reduce disproportional
representation. In all likelihood, the positive association means instead that stakeholders
perceive Latino overrepresentation and initiate programs in response.
For African Americans, targeting ofat-risk areas and African American populations was
beneficial in reducing overrepresentation, as well as targeting of very young children. Less
overrepresentation was linked also to whether foster families and clients and contract agencies
initiated an intervention, and whether it focused on improving psychosocial outcomes of birth
families. The findings suggest a picture almost the reverse of that found for Latinos.
178
For African Americans, there appear to be lower levels of overrepresentation than
expected too when key community stakeholders help to initiate the interventions that focus on
high-risk areas and populations. When interventions have a goal ofstrengthening the well being
of birth families, then implementation is associated with lesser overrepresentation.
On the other hand, counties that implementedfamily-to-family initiatives displayed
greater overrepresentation than counties that did not. Fami/y to family programs appear to be
used reactively, in response to high perceived levels of overrepresentation.
It is important to note however that in counties where interventions were successful
levels of overrepresentation remained high. Successful interventions were those associated with
lower overrepresentation than expected. Improvement was relative: in absolute terms there
remained a high degree of African American overrepresentation.
For Asian American/Pacific Islander populations findings suggested that, as with
Latinos, targeting ofhigh need areas was less beneficial than not engaging in this kind of
intervention. Involving clients and service systems in initiation, again more common for
interventions targeting high need areas, was also relatively unsuccessful. On the other hand,
management and worker support for a practice was associated with greater success.
Asian American/Pacific Islanders are, overall, underrepresented in child welfare systems
and one must interpret our findings in light ofthis fact. There were fewer significant differences
between practice types, practices, and practice characteristics for Asian American/Pacific
Islanders than for Latinos or African Americans. This may reflect a lesser sense of urgency felt
by administrators and workers in the face of Asian American/Paci fie Islander disproportionate
representation that is proportionally low than for Latino and African American disproportionate
representation, which is proportionately high.
A need remains to conceive and mount culturally and linguistic programs for Asian
American/Pacific Islander populations. This aspect of the study, with its focus on
disproportionate representation rather than on other indicators of system performance, may not
have been as good a vehicle as other parts of the study to identify program needs for Asian
American/Pacific Islander populations as for overrepresented Latino and African American
populations.
Certain limitations must be kept in mind when considering the present findings. The
study encompasses forty-three California counties. The number of counties implementing any
practice or practice type was smaller than ideal to undertake truly powerful statistical tests
sensitive to all underlying differences. As a matter of practical necessity, the study's design was
such that some of the links of practice types, practices, and practice characteristics with
disproportionate representation might have arisen by coincidence. The possibility that some
unmeasured characteristic explained some of the associations of interest revealed in the study
was reduced, but cannot entirely be ruled out.
Despite these limitations, the study produced meaningful patterns of findings connecting
programs and disproportionate representation. The associations proved sensitive to the
179
differential experience of different ethnic groups. They furnish a useful context of fact to further
consider the problem of disproportionate representation in Santa Clara County.
180
VI. PHASE 3 SUMMARY, RECOMMENDATIONS AND
CONCLUSION
Phases 1 and 2 of this study focused on examining reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where it occurs. The
focus ofPhase 3 was on key practices that take place within the CWS that either promote or
reduce disporportiona/ity among children and families ofcolor.
Overall, research on the impaci of child welfare practices for children and families of
color is limited. In general, much of the research literature is inconclusive and fails to support the
effectiveness of child welfare practices in improving outcomes for children and families involved
in the CWS. Very few studies describe differences in outcomes by ethnic/racial groups and most
of the research fails to take into account the effect of culture and ethnicity on the delivery and
impact of child welfare practices. This gap in the research literature is striking and has important
implications for children and families of color in the CWS. More information about child welfare
practices including programs, services, strategies, policies, and/or tools that either increase or
reduce racial/ethnic disproportionality can contribute to the development of more effective
services and supports for children and families of color in the CWS.
The three overall aims addressed in Phase 3 were: 1) to assess identified key practices for
children and families of color in Santa Clara County's CWS, 2) to examine the influence of the
court system on service recommendations for children and families and, 3) to conduct a
statewide comparative analysis of identified key practices that may affect the disproportionate
representation of children ofcolor in the CWS.
In order to address Aim 1 which involved an assessment of key practices in Santa Clara
County's CWS, four related analyses were conducted. First, the CWRT conducted analyses to
understand the types of practices and case characteristics that are associated with successful
outcomes for children and families from various racial/ethnic groups. In order to assess
successful outcomes, we used data from our Phase 2 Case Record Review sample and defined
success according to US DHHS guidelines, i.e., safety, permanency and family and child wellbeing. Results indicated that there were very few ethnic differences in successful outcomes.
In regards to safety, there was a significant relationship between the number of times
removed from family in current episode and ethnicity where African American children had the
highest average number of times removed from family in current episode and Asian
American/Pacific Islander children had the lowest average. In terms of family and child wellbeing, there was a marginally significant relationship between families' capacity to provide for
their children's needs and ethnicity, with African American families appearing to have a more
enhanced capacity while Asian American/Pacific Islander families seem to have a more
diminished capacity.
It is important to note that major ethnic differences do exist in Santa Clara County's
CWS. Phase 1 and 2 confirmed that certain ethnic groups are over-represented (i.e., Latinos,
African Americans) while others are under-represented (i.e., Asian American/Pacific Islanders)
in the county's CWS. We also found that there are significant ethnic differences in many
181
demographic, system-related, and psychosocial characteristics. Many of these variations appear
at different choice points in the system (such as psychosocial status at the beginning of the case,
and assignment to voluntary family maintenance). Results ofour successful case outcomes
analysis indicate that ethnic differences at the conclusion ofthe case are not evident.
While results suggest there is a clear disproportion of ethnic groups in the system, once in
the system, the children and families are generally faring the same. Their relatively
homogeneous experience may be a result of the "one-size-fits all" limited array of services
offered. Although this may seem equitable and logistically efficient, it appears ultimately
ineffective in serving a diverse group of children and families. Also, to determine better how
"well" these children are actually doing while in the system, the availability of comparative data,
clear operational definitions for "success," and more reliable and valid data are needed.
Implications and Recommendations
•
In order to better assess whether safety, permanency, and family and child wellbeing are maintained and enhanced, records should include more complete and
measurable information. We acknowledge that child welfare case records were
not originally intended for research. Yet, if the agencies responsible for providing
effective services wish to evaluate the performance of their programs more
reliably and validly, improving information collection and tracking should be
considered.
•
The creation of a case summary check-out form used at case closure would be one
way to track child outcomes. The check-out form would indicate the child and
family's status in major areas of psychosocial well-being (i.e., health, education,
and finances), in addition to the existing case outcomes. Given the amount of
missing data assessing these domains describing case and child status, and the
relative scarcity of information recorded in the last two court hearings, this form
would enable measurement and evaluation of "successful cases and outcomes" in
a more consistent objective manner.
•
The amount of missing data and the manner in which information is assessed also
implies the need for more reliable and valid methods of recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their families or of
those adopted high or low? And compared to what? Implementing better record
keeping and data collection over time on key characteristics, based on an agreedupon operational definition would better address the question of effectiveness and
performance. Implementing these practices across counties and across states
would also facilitate evaluation.
Second, a qualitative analysis using both focus group and interview data was conducted
to obtain insights about current child welfare practices used within the DFCS, how these specific
182
practices might positively or negatively impact children and families of color, and what practices
might be created, enhanced or maintained to better serve children and families of color.
Results provided information on numerous practices and contextual factors that have
either a negative or positive impact on children andfamilies of color in Santa Clara County's
child welfare system. Study participants recommended the use of in-home services for children
and families of color. There is some evidence in the research literature that supports the
effectiveness of in-home services (see literature review), however these services appear to be
most effective when they are intensive, long-term and delivered by health or social service
professionals and when the workers delivering these services receive a high level of supervision
(Barth, 1991; McGuigan et al. 2003; Olds et al., 1997). Participants also identifiedfamily
conferencing as an effective practice. These study findings are consistent with limited research in
this area that suggests that inclusive practices, in which parents are engaged in placement and
other decisions, may result in more placement stability for children, although possible
differential effects for diverse populations has not been researched (Palmer, 1996).
Participants generally expressed optimism about the adoption ofteam decision making
(I'DM) as a vehicle for improving decision-making with diverse families, although some social
workers expressed concern about time required for implementation. Research suggests that TDM
can be labor intensive and can be difficult to implement effectively within the context of the
child welfare system (Sieppert et al., 2000). Finally, participants noted that both formal and
informal collaborative relationships with other services delivery systems, such as domestic
violence and substance abuse treatment systems, are critical for successful interventions with
children and families. Furthermore, participants recognized the importance of training and
cross-training to enhance the capacity of helping professionals in child welfare and other
systems to better address the needs of children and families of color.
Interestingly, most practices and contextual factors mentioned by social workers had both
positive and negative aspects. Many of the practices and contextual factors that negatively
impact children and families of color were actually barriers to successful implementation of the
practices that social workers felt have a positive impact on children and families of color.
Although practices such as orientation ofnew clients; family and team decision-making;
preventive services; substance abuse services; cultural competency; collaboration with other
agencies and systems; a strength-based approach; and the use ofrelative placements were
described as having a positive impact on children and families ofcolor, certain negative
practices and contexts impeded their effective implementation. For instance, having to screen out
large numbers of inappropriate referrals; inconsistency in decision~making practices; a shortage
of services, particularly substance abuse and preventive services; lack of client access to
services; difficult protocols for placing children with kin; gaps in cultural competency; time
limits; and agency-level factors such as heavy caseloads, staff shortages, substantial amounts of
paperwork, lack of access to information about resources; and confusion about the agency's
overall mission and key policies all impeded implementation of the best practices. These findings
suggest that in addition to building on current positive practices for children and families of
color, CWS stakeholders should also actively work both internally and in collaboration with
partners to reduce barriers to these best practices.
183
Indeed, in order to address these barriers to effective implementation of best practices,
study participants mentioned numerous recommendations that centered around certain key
themes. In general, there was an emphasis on improving decision-making through increased
accountability, training and cross-training to reduce bias, and increased use ofgroup decisionmaking or family involvement in decision-making. Study participants also stressed a need to
improve and expand prevention, diversion and concrete services, as well as an overall need to
expand culturally and linguistically competent services, and develop ways to improve
availability and access to services. Once a case is opened, social workers also felt it valuable to
provide an orientation to clients so that they are more aware of the CWS and court processes and
can better navigate the system. Recommendations related to organizational factors were also
noted, including reducing caseloads; increasing support and streamlining service delivery;
clarification of the mission ofDFCS, as well as key child welfare policies; fostering an
organizational culture that is strength-based and community-based, and more overall training
for social workers.
Implications and Recommendations
•
The findings from the qualitative component of this study affirm the value of
many efforts that are already in place, such as culturally specific ER response
units and family resource centers, and other that are in the process of
implementation including efforts to reduce caseloads, initiation of team decision
making, and participation in the Family-to-Family initiative.
•
Practices that are inclusive, collaborative, culturally appropriate and involve
client's communities appear to be best suited to meeting the needs of children and
families of color. Although the research literature does not provide convincing
evidence of the effectiveness of these practices, the fact that respondents reported
these practices as those that are most effective with children and families of color
suggests that they may serve as promising models for children and families of
color and that more rigorous evaluations of these practices are needed.
•
The findings also point to opportunities to further strengthen services for children
and families of color. Recommendations related to training, institutionalizing
group orientation for clients, and other suggestions from study participants could
be used to inform practice and planning.
•
It would be beneficial for a team of managers, line staff, family representatives,
and community members to review the recommendations suggested by focus
group and interview participants to identify and prioritize possible practices for
adoption based on their feasibility and utility for children and families of color.
Third, in an effort to better understand the influence of contextual factors on child
welfare practices and outcomes for children and families of color, a comparison of Santa Clara
County's Main Offices and South County office was conducted. This comparison included both
quantitative and qualitative analyses. Quantitative data from CWS/CMS were used to examine
differences in case characteristics between the Main County Offices and South County Offices.
184
A qualitative comparison of focus group data from the Offices and South County Offices was
also conducted to discern how workers perceive practices in these two geographical locations.
These analyses help shed light on the potential impact of agency and community context on case
characteristics, child welfare practices and outcomes for children and families of color in Santa
Clara County's CWS.
Unfortunately, the South County location subsample was too small to allow analyses by
ethnic group. However, some key case characteristics are significantly related to county
location. Specifically, in South County Offices, a higher proportion ofcases are in voluntary
services, cases are shorter in duration, fewer workers are assigned during the course of the case,
and a higher proportion ofcases are in family maintenance at case closure. These findings are
also statistically noteworthy given the small subsample size available from SC. There were also
other differences in case characteristics by county location that did not attain statistical
significance but are noteworthy for further investigation. Children served in the South County
Offices appear to have fewer number of unique placement homes in the current episode, have a
shorter average stay per placement, a shorter length oftime in out-ofhome placement, and are
younger at time ofcase closure.
These findings imply that there is a difference in the style and quality of services
provided in the South County Office location. Qualitative findings are congruent with the
quantitative data and provide some insight into the dynamics that may contribute to this
difference. The higher number ofcases in voluntary services, the larger proportion ofcases in
family maintenance at case close, and the shorter duration ofcases is consistent with what South
County study participants described as a culture ofcommitment to maintaining and reunifying
families. This commitment was described as integral to the philosophy of managers and social
workers and reflected in the .expectations communicated to social workers by most supervisors
and by peers. The practice of providing an orientation to the child welfare system, time limits,
the courts, and how to work with social workers may also contribute to the differences found in
the quantitative analysis. Study participants frequently described South County as similar to a
small county or rural area. This geographic difference, and the resulting sense of "connection to
the community, " was identified as a factor in the greater sense ofshared philosophy and
emphasis on prevention of out-of-home placements. Other studies suggest that region may
influence practice. For example, one study (Drake, 1996) also found that rural areas are more
likely to offer preventive services than are urban areas. The finding that fewer workers are
assigned during the course of the case is also consistent with descriptions of the vertical case
management model. This model calls for social workers to carry the same case from case
opening to case closure.
Implications and Recommendations
•
Future research should include a larger subsample from South County and include
methods to understand the reasons behind these location differences. Also, a
larger South County Office location sample may enable more detailed
comparisons by ethnicity.
185
•
Practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County
Offices and adapted to other regions of Santa Clara County, as recommended by
study participants. This practice may be particularly helpful to low-income
families and families of color that may be intimidated by or unfamiliar with child
welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for other
regions of Santa Clara County, though the success of this practice may be linked
to the "small county" dynamic described by South County study participants.
Some participants in focus groups that took place at the main office mentioned
this model as promising while others asserted that specialized expertise in
different areas of social work was an advantage in providing quality services that
are informed by "the most current information."
•
Further examination of ways in which the culture of commitment to maintaining
and reunifying families is created, communicated, and continued merits attention.
In addition, an exploration of how this culture might be adapted to other parts of
the county would be beneficial.
Fourth, in order to identify case characteristics and system-related factors predictive of
reunification of children with their families, two exploratory models were tested using
multivariate analyses. The first model was composed of primarily demographic and systemrelated variables to predict reunification (child's gender, ethnicity, age at time of case opening,
the number of workers assigned across the history of the case, the time length of the case, the
number of unique placement homes assigned, the number of times removed from the family, and
the county location). Four of these variables predicted reunification: child's ethnicity, number of
workers assigned throughout the case, length ofthe case, and number of unique placement
homes.
Asian American/Pacific Islanders were less likely than Whites, African Americans, and
Latinos to be reunified with their families. Given that Asian American/Pacific Islanders are one
of the minority groups least acculturated to mainstream society, their cases may be perceived as
more severe or problematic because of the different cultural attitudes toward child rearing and
discipline. Differences in parenting styles may reduce the likelihood of reunification compared to
the other major ethnic groups. However, as Phase 2 results indicated Asian American/Pacific
Islanders are also the most likely to be placed in Voluntary Family Maintenance (Hines et al.,
2002), the CWS appears to value the preservation of Asian American/Pacific Islander families
when first referred to the system (perhaps recognizing the challenges of culturally appropriate
services, language issues, and other circumstances related to the Asian American/Pacific Islander
community) but once in the system, Asian American/Pacific Islander family cases are not easily
resolved and circumstances may persist that place the child at risk for further abuse of neglect.
The fewer number ofworkers assigned was also related to reunification. This finding
indicates consistency in service and the relationship between the social worker and the family
186
may improve the chances ofreunification. Perhaps this is due to greater familiarity and thus more
effective advocacy for the child and family. This finding could also ·be a function of particularly
challenging family cases already less likely to be reunified needing a greater set of workers over
time (e.g., due to worker burnout in the case, or specialization needed across the case).
Shorter case length was also predictive ofreunification. Presumably, less severe cases
could be resolved in a timely fashion, usually concluding with reunification. Following a similar
logic above regarding the number of workers, a longer case may imply more problems, thus
involving more time for resolution, but also possibly a case already protracted into a situation
where reunification is unhealthy for the child.
The fewer number of unique placement homes assigned was related to reunification.
Change in placements may be due to systemic conditions (i.e. logistics and short term
availability of space) but multiple placements can also indicate a persistent problem with a child
adapting to a placement and accepting care offered. As this may again indicate a more severe
case, reunification could also become less likely.
To identify better the influence of case characteristics and indicators of success, we added
6 variables to the model collected through our in-depth case record reviews and utilized the
additional information coded for Phase 3: a family's enhanced capacity to provide for their
children's needs, children receiving appropriate educational services, children receiving adequate
services to meet physical and mental health needs, number of referrals, number of previous times
in the CWS, and assignment to Family Maintenance or Family Reunification services. The
location variable was not included in this model as only one case coded from the South County
Office area.
Similar to the first model, Asian American/Pacific Islanders were still less likely than
Whites, African Americans, and Latinos to be reunified. Also, shorter time length of a case
again predicted reunification. However, in this expanded model with success indicators, the
number of workers assigned and number of unique placement homes assigned were not
statistically related to reunification. The child's age at time of case opening and the number of
unique placement homes did approach significance. Interestingly, the variables related to
success cases and outcomes were not significant.
Implications and Recommendations
•
These findings indicate that particular factors are important to consider when
predicting reunification, specifically ethnicity and the length of time a case
remains open. More severe cases (those needing more time for resolution) may be
more problematic, thus reducing the chances of reunification.
•
The finding that Asian American/Pacific Islanders are less likely to be reunified
than Whites, African Americans and Latinos, highlights a main ethnic difference.
In Phase 2 we discovered that Asian American/Pacific Islanders were more likely
than the other ethnic groups to be enrolled in Voluntary Family Maintenance
services (Hines et al., 2002). However, our results here imply that when Asian
187
American/Pacific Islander children are removed from the home, their children are
less likely to be reunified. This may also indicate extreme outcomes where Asian
American/Pacific Islander children are either initially diverted from the system or
once in the mainstream of the system are less ~ikely to be reunited with their
families.
•
It was interesting that the successful case and outcome variables were not
predictive of reunification. This could be due to the significant impact of other
variables overshadowing the influence of success indicators. It could also be due
to our use of reunification as a dependent variable. As we discussed in the
Successful Cases Analysis section, reunification is not necessarily synonymous
with success, and thus these variables may indeed be unrelated.
•
Due to missing data across our set of variables, a number of were available for
this multivariate analysis. More complete information would have improved
confidence in our findings.
•
Better measurement and a more comprehensive evaluation of family history,
abuse and neglect circumstances, and other indicators could produce a stronger
predictive model.
•
Although only a preliminary set of models, these findings are useful for
identifying areas in the CWS and family circumstances that can be studied
further, especially in conjunction with the specific goal of reunifying children
with their families.
Our second aim focused on examining the influence of the court system on service
recommendations for children and families ofcolor. Fey,t studies have examined the potential
influence of the court system on child welfare practices and outcomes, however as noted in the
literature review (Section III), the court system is likely to have a significant impact on the
trajectory of some child welfare cases. In Phase 3, we examined the role of the court system in
child welfare practice by exploring the types of court ordered changes that are made to social
worker recommendations at the initial jurisdictional/dispositional hearing, and how changes
might vary by ethnicity of the case. In addition, we explored the relationship between the child
welfare and court systems.
Quantitative results indicated that the court made changes to the initial social worker
recommendations at the jurisdictional/dispositional hearing in approximately halfof the cases.
Ethnicity was not related to whether or not a change occurred Yet, these results may not
necessarily indicate that the system and courts treat children and families from different
ethnicities the same, but rather that when the judge decides to make a change from the social
worker recommendations, these changes appear to occur equally across ethnic groups. It was the
opinion of members of the Court Officer Unit focus group that ethnicity does not contribute to
decisions in court, but rather it is circumstances that drive the case.
188
Additionally, although quantitative results may imply that judges accept social worker
recommendations about half of the time, findings from the focus group with the Court Officer
Unit suggest that many factors would complicate an accurate measurement of why social worker
recommendations may or may not be followed. Factors that may impede accurate assessment of
the agreement between child welfare and court personnel on service plan recommendations
include the strong influence of individual stakeholders involved, and formal and informal case
plan negotiation tactics, all of which contribute to a much more complicated process.
Qualitative .findings from the agency-wide focus groups suggested that collaboration
between the child welfare and court systems is problematic. Agency-wide focus group
participants felt that the child welfare and court system have different perspectives on the needs
and circumstances of children and families in the CWS and that the court system may have
unrealistic expectations of families. Additionally, many agency-wide focus group participants
described the ways in which child welfare and court personnel interact as ineffective. Some
social workers felt that some court personnel would dismiss their assessments and
recommendations and may try to pressure social workers to change their recommendations. A
somewhat similar finding was noted by Knepper and Barton, ( 1997) who found that although
judges tended to accept social worker recommendations, the relationship between social workers
and the court plays a significant role. Their study found that when social workers adhered to the
group norms of the court, judges rewarded them by not ordering "unrealistic" practices and
allowing reasonable time frames for mental health evaluations to be completed. Similarly,
agency-wide focus group .findings seem to suggest that ifsocial workers go against the implicit
rules of the court system, thenjudges may override their recommendations, but if they adhere to
expectations of the court system than recommendations are accepted.
Implications and Recommendations
•
Although the quantitative and qualitative information described many of the
circumstances behind why social worker recommendations are not upheld by the
court, it is very difficult to ascertain the degree to which each circumstances
influences a court ordered change. More examination of this area is needed, as is
further investigation of ways in which relationships between court personnel and
child welfare workers have an impact on the experiences of children and families
of color involved in the CWS.
•
Better and more exhaustive record keeping may not be the solution, as many
informal and undocumented actions in the negotiation process occur regularly,
and are confidential, thus making much of the data unavailable for research and
evaluation purposes.
•
Assessing other factors such as courtroom environment, stakeholder
characteristics, and other judicial issues would be advantageous in later research.
Although obtaining this information is a formidable task, these factors should be
include in the future, as measuring primarily system-related factors may not be
sufficient in efforts to explain outcomes in the CWS.
189
Our third aim was to conduct a statewide comparative analysis of identified key practices
that may affect ethnic disproportionality in the CWS. A statewide survey was administered to
child welfare directors, managers or supervisors in California counties who were asked to
identify and describe effective practices for children and families of color in their child welfare
systems. In addition, a quantitative analysis that identified the link between these practices and
county levels of disproportionate representation of children of color in county child welfare
systems was conducted.
Overall findings from the statewide comparative analysis indicate that certain types of
practices f:Ire considered by county child welfare personnel as effective with diverse populations,
and that these practices are related to disproportionate representations ofchildren of color in
county child welfare systems. Descriptive results revealed four overall types of practices
considered to be most effective with children and families of color, as well as various specific
practices within these types of practices including: 1) beginning choice point practices, which
included family group conferencing or other team decision-making practices, voluntary family
maintenance or other prevention/family preservation services, and risk assessment tools or
systems, 2) continuing care/ending choice point practices, which included, placement related
practices, the Family to Family Initiative, and intensive and collaborative interventions for highrisk children, 3) both beginning and continuing/ending choice point practices, which included,
community-based services, and strength based assessments, and 4) organizational or structural
practices which included, cultural competency of staff, collaboration with American Indian tribes
and vertical case management.
In general, very little research on the effectiveness of the practices identified by
respondents has been conducted, and in cases where evaluations have been performed, results are
generally inconclusive. More research is needed to evaluate the effectiveness of these services
for diverse populations.
Multivariate analyses revealed several significant associations between types of
practices, specific practices and practice characteristics and levels ofdisproportionate
representation ofchildren of color in county child welfare systems, when other factors such as
urban-rural status and per capita income were statistically controlled
Beginning choice point practices were found to have a beneficial effect on Latino
overrepresentation. In particular, family group conference or other team decision-making
interventions predicted lower levels of Latino overrepresentation. However, Latino
overrepresentation was greater, not less, when interventions were implemented targeting areas
with high concentrations of distressed families. Latino overrepresentation was also greater when
clients, foster families, out-of-home placement providers, and other service systems were
involved in the development of the practice. As such, it appears that practices may be developed
because stakeholders perceive Latino overrepresentation and initiate programs in response.
In contrast to Latinos, targeting practices to at-risk areas for African American
populations was beneficial in reducing overrepresentation, as well as targeting of very young
children. Less overrepresentation was linked also to whether foster families and clients and
contract agencies initiated an intervention, and whether it focused on improving psychosocial
190
outcomes of birth families. The findings suggest a picture almost the reverse of that found for
Latinos.
Additionally, there appear to be lower levels ofAfrican American overrepresentation
than expected when key community stakeholders help to initiate the interventions that focus on
high-risk areas and populations. When interventions have a goal of strengthening the well being
of birth families, then implementation is associated with less African American
overrepresentation. Yet when counties implemented Family-to-Family initiatives greater African
American overrepresentation resulted. Again, the implementation of Family to Family programs
may be used reactively, in response to high perceived levels of overrepresentation.
As with Latinos, targeting ofhigh need areas was less beneficial for Asian
American/Pacific Islander children than not engaging in this kind of intervention. Involving
clients and service systems in initiation, again more common for interventions targeting high
need areas, was also relatively unsuccessful. On the other hand, management and worker
support for a practice was associated with greater success.
Implications and Recommendations
•
County directors and child welfare administrators identified practices with similar
characteristics as those identified within Santa Clara County, i.e., those that are
inclusive, collaborative, culturally appropriate and involve client's communities
appear to be best suited to meeting the needs of children and families of color.
However, further research and evaluation is necessary to determine which
practices most effectively meet the needs of individual racial/ethnic groups.
•
While multivariate analyses suggest that certain practices do predict levels of over
and underrepresentation of children of color in the child welfare system,
additional research to determine the mechanisms through which these practices
impact this disproportionate representation is needed.
•
Additional explorations into the relationship between practices, disproportionate
representation of children of color in the child welfare system and county
characteristics would also help shed light on the process through which certain
practices impact children and families of color.
Phase 3 Conclusion
Results from Phase 2 indicated that different racial/ethnic groups experience unique
pathways through the CWS (Hines et al., 2002). At every point in the system, various child
welfare practices and services impact children and families. Phase 2 results indicated that
children and families of color in Santa Clara County's CWS tend to be recommended services
that are traditional and formal in nature and that do not appear to meet the wide range of needs
experienced by these highly diverse racial/ethnic family groups. Indeed, children and families of
color in the CWS represent a high risk group who are often impacted by myriad psychosocial
191
challenges yet services recommended for these families tend to be limited to a one-size-fits all
approach.
Moreover, there exists relatively little empirical evidence that traditional child welfare
services are effective in maintaining or reunifying children and improving family functioning.
Even less research has been conducted on the impact of child welfare practices for children and
families of color. Yet because children and families of color are disproportionately represented
in the CWS, the effectiveness of child welfare practices for these groups is of particular interest.
Research suggests that children and families of color tend to ·have longer stays in the CWS, they
receive fewer and less comprehensive services, and have poorer case outcomes than White
children and families (Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996).
This bleak outlook for children and families of color in the CWS creates an urgent need to
understand child welfare practices and services that result in enhanced outcomes for these
families. It is especially important to identify the types of individual and contextual factors that
hinder or support the effectiveness of these practices.
The primary goal of Phase 3 was to assess key practices for children and families of color
in the CWS at both the county and state levels in order to provide information that might help
generate strategies aimed at reducing ethnic/racial disparities and develop more effective
services and supports for children and families of color. While little research exists that
empirically supports policies, programs and practices that attempt to reduce ethnic/racial
disproportionality, some programs and practices appear promising and merit further exploration.
In general, these practices aim to reduce the need for out-of-home placement and increase family
and community participation in decision-making. Please see Chart 1 for a summary of promising
child welfare practices from Phase 3.
In our analysis of Santa Clara County child welfare practices, results based on focus
group and interview data indicated that in-home services, family conferencing, and team decision
making {TDM) were viewed by participants as ways to improve decision-making with families
from diverse cultural and ethnic groups. Participants noted that both formal and informal
collaborative relationships with other services delivery systems, such as domestic violence and
substance abuse treatment systems, are critical for successful interventions with children and
families. Furthermore, participants recognized the importance of training and cross-training to
enhance the capacity of helping professionals in child welfare and other systems to better address
the needs of children and families of color.
Our analysis comparing South County and the Main Office enabled us to examine
contextual factors that may have an impact on child welfare practices, in this case, small vs. large
and urban vs. rural. Results indicate that children in South County appear to have better
outcomes, i.e., they have fewer number of unique placement homes in the current episode, have a
shorter average stay per placement, a shorter length of time in out-of-home placement, and are
younger at time of case closure. Study participants described South County as "a small county or
rural area." This geographic difference, and factors related to an organizational culture that
emphasize a commitment to maintaining and reunifying families, the practice of providing an
orientation to the child welfare system, as well as the finding that fewer workers are assigned
192
Chart 1: Promising Child Welfare Practices from Phase 3
Beginning Choice Points
Promis.ine Practices
Team decision making, case conferencing, family group conferencing
and/or family involvement in decision-making
Orientation for parents on child welfare and court processes
.: ; :.-~::_:,.' . ::,. ~~ :~=. ,,:--.;-:.:~ : _,~~.;;..:.::,,>!.~ :•..~·.: ~'."...,--..,_::11 ': ·:,._f~:;..-.-~.. 1" _;.:i:-:..:.~~ ', ~"',:.
4
Continuing Care/Ending
Choice Points
..... :,' 1::..;.1.~'',':": ~-j,,d•,~- ·,')·~.J-'.f;.i~·-."- : •. : · ,l ·--~~- ~~:. ,,."·'· \
,:...:;...:_.
Contextual Factors or
Organizational/
Structural Practices
-~
Voluntary Family Maintenance/Family Preservation Services
Child care and in-home services
Risk assessment tool or system
.i ~-~~~!:__··:~·. :,-~---..~:.-:,---;~· l- -:.:;.;;:_'~-·~. ~,.;.:~. ~~-: ·~~-;.• ~; .."~'-=J~-~~·•:~\"; ?:;""': ,"":"-....,._
._h• .. ,~ :):. ·:~ '·.~ 1.-::~.}-"bic;=::,-:..' :~~-~.'=-!~·..•
i~.-~~:-~~:·:-::::,:_.-··:..:'.\t'~.:. --~::
;_.,;._. :t.::_: ;.,.::.:..
.. ,••·.,.
Parenting education services involving both child and parent
Cultural matching for children in foster care
Recruitment, training and coaching for foster parents
Placement-related practices
Family-to-Family Initiative
Intensive and collaborative interventions for high risk children
Community based services (both beginning & continuing/ending)
Strength-based assessments (both beginning & continuing/ending)
,,f·J ·l 5.-.,l'""·"t:-~'\i' ':_"':. .. ,_·7~·i-a·t_:~.J""'t..:;::----~-z.:Jt:;:-,;;. ..1;.:,?;:;c-:.:• r:~-~:!~~-l. ... _" '.'~-~=.r~:t~--"":_1;.-e,~:;:v !" l"1i•,;-.,.,r...,,.~.;..; .... :··-;;-~~·~~~•-W""""· ,\. -~·---...~~-=-~;1i!;._~~~-~-'
-. -~:A
.,.,,t
of
Cultural competency staff
Collaboration with American Indian tribes
'.< ~--: -•:•. -~J:•;;.-:.-"·•:-:-,;:.:.:
~-: .:·:,,:-t:-:
.i.: .~·1•_:.:-;~;:.c...:.::::..~· ~·:.:·-::-~•~'!"r.--~:~"!'" :-:~~· ~~y;-~:~:~·:1•.,::_;-;:,-,:-:j.~.:::::
Main County Focus Groups
Main County Focus Groups
Main County Focus Groups
Statewide Survey
Statewide Survey
Statewide Survey
Statewide Survey
Statewide Survey
f_.,:,,.: ... ,J t: ~-r'-:-_:r:~_-.:.:-...~· ;.. ,-x•_.;$.!.it...,S.---!":~:?,_;,,:;_ --;=-·=•:~::::..i.,;Jtf;'°.fi-;.:r'£.,.::_-~;::·.::.-:i.1~ ~-.--e,;,".~:.1?;;: ::;_:;-:
Cultural matching between staff and clients, and cultural consultation
Supportive supervision for staff
Collaboration with other agencies or systems
Strength-based approach to services
Small county dynamic
Culture of commitment to maintaining and reunifying families
Vertical case management
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Study Component
Main County Focus Groups
South County Focus Groups
Statewide Survey
Main County Focus Groups
South County Focus Groups
Statewide Survey
Main County Focus Groups
Statewide Survey
•
•
•
•
•
•
•
•
•
•
Main County Focus Groups
Main County Focus Groups
South County Focus Groups
Main County Focus Groups
Main County Focus Groups
South County Focus Groups
South County Focus Groups
South County Focus Groups
Statewide Survey
Statewide Survey
Statewide Survey
193
during the course of the case may all be related to these positive outcomes for children in South
County. It is interesting to note, however that our multivariate analyses that examined case
characteristics and system-related factors related to reunification indicated that when taking other
factors into account, geographical location was not a significant predictor of reunification.
Rather, fewer workers, shorter case duration, and fewer placements were the strongest predictors.
More research on factors related to agency culture and geographical location as they impact child
welfare practice and outcomes for children and families of color is clearly warranted.
Research indicates that the court system is likely to have an impact on the practices that
are provided to children and families in the CWS. In an analysis conducted to determine the
extent to which court orders changed social worker recommendations at the
jurisdictional/dispositional hearing, results indicated that the court made changes to the initial
social worker in approximately half of the cases. Results further indicated that that ethnicity was
not related to whether or not a change occurred. Findings from qualitative analysis exploring the
relationship between the child welfare and court system painted suggested that the collaboration
between the child welfare system and courts is problematic. More research is this area is needed,
as is further investigation of ways in which relationships between court personnel and child
welfare workers have an impact on the experiences of children and families of color involved in
the CWS.
Results on statewide practices indicate that certain types of practices are considered by
county child welfare personnel to be effective with diverse populations, and that these practices
are related to disproportionate representations of children of color in county child welfare
systems. Similar to practices identified within Santa Clara County's DFCS, practices at the
statewide level were described as effective with children and families of color because they are
inclusive, collaborative, culturally appropriate, and involve client's communities. Although the
research literature does not provide convincing evidence of the effectiveness of these practices,
the fact that respondents reported these practices as their most effective with children and
families of color suggests that they may serve as promising models for children and families of
color and that more rigorous evaluations of these practices are needed. Phase 3 also attempted to
assess how the practices identified through the survey might affect differential representation of
children of color in the CWS. Significant associations were detected between practice types,
practices and practice characteristics and disproportionate representation and interestingly,
different patterns of association were found for Latinos, African Americans and Asian
American/Pacific Islanders.
While researchers have described the existence of ethnic/racial disparities in every aspect
of the CWS, little attention has been given to ways to address the existing disproportinality. It is
our hope that the information contained in Phase 3 of the Children of Color Study will help in the
development of more effective programs and practices for families and children of color at both
the county and state level. Overall, our findings from Phase 3 suggest that efforts to address
racial and ethnic disparities in the child welfare system should involve a more concentrated focus
on child and family well-being, involving the creation of family-centered and community-based
services that are inclusive and collaborative and specifically designed to meet the needs of the
diverse cultural and ethnic groups in the child welfare population.
194
VII. SUMMARY OF PHASES 1, 2 AND 3: OVERALL
CONCLUSIONS AND IMPLICATIONS FOR SANTA CLARA
COUNTY'S CHILD WELFARE SYSTEM
The central focus of inquiry for this project was to identify factors related to the
disproportionate representation of children of color in Santa Clara County's CWS. In order to
address the complexity of this issue, the CWRT elected to employ a multiphase/multimethod
approach that examined factors at the individual, family and system levels. The following
summarizes findings from Phases 1, 2 and 3 of the study and provides overall implications for
the County based on our results.
Phase 1, an exploratory phase, drew from three sources: research literature at the national,
state and county levels; Santa Clara County's management information system (CWS/CMS); and
focus group discussions with professionals in Santa Clara County who provide child welfare
services, as well as parents, caregivers and youth who are recipients of CWS services. Results
indicated that factors related to racial/ethnic disproportionality were multiple and complex and
necessitated investigation at the individual, family, system and community level. Based on
preliminary analyses and available literature, we argued that children from different racial/ethnic
groups were likely to be treated differently at specific decision-making or key choice points in
the system. The concern about possible differential treatment of children of color was
substantiated by the focus group results, as well as by CWS/CMS data that indicated that there
were significant differences in the number of months in placements by race/ethnicity. For
example, results indicated that African American children spent significantly more time in
placement than their White, Hispanic/Latino, Asian American/Pacific Islander, Vietnamese, and
Filipino peers. Also, according to preliminary findings from CWS/CMS, in Santa Clara County,
following placements in~ relative home, the second most frequent placement for African
American, Native American, White and Hispanic/Latino youth, was a Foster Family Agency
(FFA); for Asian American/Pacific Islander and Vietnamese youth it was a Foster Family Home
(FFH). Almost 14% of Native American and 18% of Asian American/Pacific Islander youth in
OHP were placed in the Children's Shelter in December, 2000, a percentage that was higher than
any of the other ethnic/racial groups. We also argued that little was known about specific
pathways through the system and ways in which the pathways differed for various racial and
ethnic groups. Our findings from Phase I indicated that much of the research on children in the
CWS had focused on factors related to movement in and out of the system, but that little
information existed on the actual experiences of children in care and the individual and family
characteristics that are associated with these experiences. While focus group members
frequently stated that once a child of color entered the system, it was very hard for that child to
exit, little information existed that could explain what happened once that child was in the
system.
Phase 2 was designed to focus on the actual experiences of children in care and the
individual and family-related characteristics that were associated with these experiences. Results
based on extensive, in-depth reviews of 403 closed child welfare case records, a parallel
descriptive analysis of I 720 closed cases within the CWS/CMS database,·and key informant
interviews with managers and supervisors in the County's DFCS indicated that families
195
belonging to each of four different racial/ethnic groups (Latino, African American,
Asian/Pacific Islander and White) were characterized by different constellations of risk factors,
and that once in the system, the children had different experiences that resulted in different
outcomes. Additionally, results indicated that services recommended for the families and
children were limited to a small array of traditional services - a one-size fits all approach - that
did not necessarily meet the needs of the culturally diverse families and children. (Please see
Attachment 8 for a summary chart of Phase 2 findings.) We concluded that the paucity of
recommended services indicated that children and families of color were not likely to be
provided with sufficient preventive and supportive services and that the traditional child welfare
services might not meet the particular needs of these unique and diverse groups. These findings
are supported by research literature indicating relatively little empirical evidence that traditional
child welfare services are effective in maintaining or reunifying children and improving family
functioning. Even less research has been conducted on the impact of child welfare practices for
children and families of color. Yet because children and families of color are disproportionately
represented in the CWS, we contend that the effectiveness of child welfare practices for these
groups is of particular interest.
Phase 3 focused on agency practice and ways in which current practice interacted with
child, family and cultural characteristics of different cultural/ethnic groups. The prhnary goal of
Phase 3 was to assess key practices for children and families of color in the CWS both at the
County and state levels in order to provide information that might help generate strategies aimed
at reducing ethnic/racial disparities and develop more effective services and supports for children
and families of color. Methods included semi-structured in-depth interviews and focus groups
with DFCS supervisors and managers, social workers and parents involved in the CWS; a
qualitative and quantitative comparison of child welfare practices and outcomes between the
DFCS South County and Main County offices; a reanalysis of cases from Phase 2 with
successful outcomes and an analysis of factors that contributed to those positive outcomes; a
supplemental data collection for our case record review sample so as to include information on
court ordered changes to the social worker recommendations at the jurisdictional/dispositional
hearing; and, a statewide survey of county child welfare directors, managers or supervisors in
California counties and quantitative analyses using a statewide database consisting of countylevel characteristics.
As mentioned earlier, Phases 1 and 2 confirmed that certain ethnic groups were overrepresented (i.e., Latinos, African Americans) while others were under-represented (i.e., Asian
American/Pacific Islanders) in the County's CWS. Phase 2 results also indicated that there were
significant ethnic differences in demographic, system-related, and psychosocial characteristics
and that many of these variations appeared at early choice points in the system. During Phase 3,
we were interested in examining ethnic/racial differences in outcomes at later stages of the
system. Using data from our Phase 2 Case Record Review sample and defining successful
outcomes according to US DHHS guidelines (i.e., safety, permanency, and family and child
well-being), we found that there were minimal ethnic differences in successful outcomes. Phase
3 results suggested that when success was defined according to DHHS guidelines, ethnic
differences at the conclusion of the case were not evident. One interpretation for this finding
might be that while there is a clear disproportion of ethnic groups in the system, once in the
system, children as a whole, regardless ofrace/ethnicity, are faring at about the same level.
196
Given that Phase 2 results found that children and families of color in Santa Clara County's
CWS were receiving a "one-size fits all" approach to services, an approach that may seem
equitable and logistically efficient, but perhaps, ultimately ineffective in serving a diverse group
of children and families.
In order to further examine child welfare practices at the County level and to gain a better
understanding than the one gleaned solely through the use of CWS/CMS data, we conducted a
series of semi-structured in-depth interviews with DFCS managers and supervisors and focus
groups with social workers and parents involved in the CWS. Our results indicated that in-home
services, family conferencing, and team decision-making (TOM) were viewed by participants as
ways to improve decision-making with families from diverse cultural and ethnic groups.
Participants noted that both formal and informal collaborative relationships with other service
delivery systems, such as domestic violence and substance abuse treatment systems, were critical
for successful interventions with children and families. Furthermore, participants recognized the
importance of training and cross-training to enhance the capacity of helping professionals in
child welfare and other systems to better address the needs of children and families of color.
We were also interested in gathering information from counties throughout California
regarding promising practices for families and children of color. This information was gathered
through a statewide survey administered to child welfare directors and program administrators.
Similar to findings on practices identified within Santa Clara County's DFCS, practices at the
statewide level were described as effective with children and families of color because they are
inclusive, collaborative, culturally appropriate and involve client's communities. Although the
research literature does not provide convincing evidence of the effectiveness of these practices,
the fact that respondents reported these practices as those that are most effective with children
and families of color suggests that they may serve as promising models for children and families
of color and that more rigorous evaluations of these practices are needed (Please see Chart I on
p. 192 of the Phase 3 Final Report for a summary of promising practices for children and
families of color in the CWS).
Although practices such as orientation of new clients; family and team decision-making;
preventive services; substance abuse services; cultural competency; collaboration with other
agencies and systems; a strength-based approach; and the use of relative placements were
described as having a positive impact on children and families of color, certain negative practices
and contexts were cited as possibly impeding their effective implementation. For instance,
having to screen out large numbers of inappropriate referrals; inconsistency in decision-making
practices; a shortage of services, particularly substance abuse and preventive services; lack of
client access to services; difficult protocols for placing children with kin; gaps in cultural
competency; time limits; and agency-level factors such as heavy caseloads, staff shortages,
substantial amounts of paperwork, lack of access to information about resources; and confusion
about the agency's overall mission and key policies all impeded implementation of the best
practices. These findings suggest that in addition to building on current positive practices for
children and families of color, CWS stakeholders should also actively work both internally and
in collaboration with partners to reduce barriers to these best practices.
197
Results from focus groups conducted during Phase 1 and key informant interviews held
during Phase 2, as well as current research literature suggest that agency organizational culture
and context may have an impact on child welfare practices and resulting family and child
outcomes. During Phase 3, we conducted both quantitative and qualitative analyses comparing
South County and the Main Office that enabled us to examine various contextual factors at the
organizational level, in this case, small vs. large and urban vs. rural, and their impact on
outcomes for children and families of color. Results indicated that children in South County
appeared to have better outcomes, i.e., they had fewer number of unique placement homes in the
current episode, had a shorter average stay per placement, a shorter length of time in out-ofhome placement, and were younger at time of case closure. Qualitative findings were congruent
with the quantitative data and provided some insight into the dynamics that might contribute to
this difference. South County study participants described a culture of commitment to
maintaining and reunifying families. This commitment was described as integral to the
philosophy of managers and social workers and reflected in the expectations communicated to
social workers by most supervisors and by peers. The practice of providing an orientation to the
child welfare system, time limits, the courts, and how to work with social workers might also
contribute to the differences found in the quantitative analysis. Study participants frequently
described South County as similar to a small county or rural area. This geographic difference,
and factors related to an organizational culture that emphasizes a commitment to maintaining and
reunifying families, the practice of providing an orientation to the child welfare system, as well
as the finding that fewer workers were assigned during the course of the case may all be related
to these positive outcomes for children in South County. Further, our multivariate analyses that
examined case characteristics and system-related factors related to reunification indicated that
when taking other factors into account, fewer workers, shorter case duration, and fewer
placements were the strongest predictors of family reunification.
The issue of disproportionate involvement of children of color in the CWS has long been
an issue of concern for CWS workers, clients, researchers and government and community
groups. More recently, it has been the focus of much national attention. Santa Clara County's
Children of Color Study is one of the few to examine this issue at the local level. We applaud the
County Board of Supervisors.and DFCS for their openness and willingness to undertake such a
project. We acknowledge that there are multiple stakeholders who may be interested in
identifying a "magic bullet" or a single, straightforward explanation for the system's
racial/ethnic disproportionality. Single, straightforward explanations are appealing as they might
lead to quickly implemented solutions. However, results of this study indicate that rather than
one primary causal factor, there appear to be numerous aitd interrelated factors associated with
the disproportionate involvement of children of color in Santa Clara County's CWS. It is our
contention, that factors that operate simultaneously and in complex ways at the individual,
family and system level with differing patterns across diverse racial and ethnic groups contribute
to disproportionality throughout the system. Overall, findings suggest that efforts to address
racial and ethnic disparities in the CWS should focus on the diverse needs of the different ethnic
and racial groups involved in the CWS and the design of more culturally specific and effective
prevention and intervention programs to meet those needs.
198
Implications for Santa Clara County's Child Welfare System
The following are suggestions based on results from all three phases of the Children of Color
Study. It is our hope that these recommendations may be helpful in guiding Santa Clara
County's continuing efforts to address the complex issue of racial/ethnic disparities in its CWS.
1. There is a need to expand available child welfare services to better serve families and
children ofcolor. Phase 2 results indicated that reliance on a small array of
traditional formal services does not appear to meet the needs of the highly diverse
ethnic/racial family groups involved in the County's CWS.
•
More preventive and early intervention services for vulnerable families of color
should be implemented and carefully evaluated. The findings from the qualitative
component of Phase 3 affirmed the value of many efforts that are already in place,
such as culturally specific ER response units and family resource centers, and
other that are in the process of implementation including efforts to reduce
caseloads, initiation of team decision making, and participation in the Family-toFamily initiative.
In addition, home visitation services may be exceptionally helpful in addressing
the needs of vulnerable, at-risk families from diverse racial/ethnic groups.
Research has indicated that programs of home visitation that promote positive
health-related behaviors in mothers of young children, competent care of their
children and linkage with needed health care and hu~an services, reduce rates of
criminality, problems related to substance abuse and child abuse and neglect
among young, unmarried, isolated, poor mothers.
• _The use of more non-traditional, culturally sensitive services is clearly -warranted.
Key informants in all three phases of the study underscored this point. A paucity
of social services, particularly multi-lingual services, was cited as a significant
barrier for many families of color. Interviewees discussed the shortage of
substance abuse treatment programs, particularly those geared for women with
children and people whose primary language is not English. Multi-lingual and
culturally appropriate domestic violence services, parenting classes, and other
social services were considered in need of development.
•
Piloting new, innovative services and evaluating their success with different
ethnic/racial groups could contribute to a more diverse array of culturally specific
programs for families and children.
•
Phase 3 results also indicated that practices that are inclusive, collaborative,
culturally appropriate and involve client's communities appear to be best suited to
meeting the needs of children and families of color. Although the research
literature does not provide convincing evidence of the effectiveness of these
199
practices, the fact that respondents reported these practices as those that are most
effective with children and families of color suggests that they may serve as
promising models for children and families of color.
•
More services targeted to fathers and programs that are formulated and delivered
within a family-based framework would be of benefit to families and children of
color involved in the CWS.
•
In addition to building on current positive practices, and initiating new programs
for children and families of color, CWS stakeholders should actively work
internally and in collaboration with partners to reduce barriers to implementing
such practices. Barriers such as large numbers of inappropriate referrals;
inconsistency in decision-making practices; a shortage of services, particularly
substance abuse and preventive services; lack of client access to services; difficult
protocols for placing children with kin; time limits; and agency-level factors such
as heavy caseloads, staff shortages, substantial amounts of paperwork, lack of
access to information about resources; and confusion about the agency's overall
mission and key policies were cited during Phase 3 focus groups as impediments
to the implementation of cited best practices.
2. There is a need to involve multiple social service systems in a comprehensive and
coordinated effort to meet the needs of children and families of color. Results from
Phases 2 and 3 indicated that the problems experienced by families across the
different racial/ethnic groups span multiple systems including: mental health, juvenile
justice, adult criminal justice, substance abuse, and welfare. Statistics presented in
the literature review section of the Phase 2 Final Report also indicate that families of
color are involved in systems other than child welfare in high numbers. Prevention
and intervention efforts should involve a deliberate and organized coordination of
these multiple systems. During Phase 3 interviews and focus groups, collaboration
with other agencies and systems concerned with the safety and welfare of children
and families of color was highlighted as a beneficial practice for children and families
of color.
•
Developing interagency formal agreements and connections with liaisons in
other agencies would be beneficial as it can facilitate collaboration and
minimize the potential for agencies to give conflicting messages and mandates
to clients. Phase 3 participants suggested that social workers were better able
to divert cases or get a more complete picture of the family when two agencies
are working together. Other agencies mentioned in this context include
probation, hospital, medical providers, family and mental health agencies,
CalWORKS, substance abuse treatment agencies, domestic violence agencies
and other entities that may be involved with clients.
•
The Greenbook project, an initiative to better coordinate domestic violence
and child welfare services, could serve as a model for interdisciplinary
problem solving and policy development. The Greenbook project successes
200
could help inform evolving collaboration with other key systems, such as the
substance abuse treatment system.
•
Other coll,aborative efforts such as outstationed social workers, co-location of
services and Family Drug Court could serve as models for best practices for
children and families of color. Social workers and supervisors, interviewed
during Phase 3, described the family drug court as a model for effectively
working with substance abusing families in the child welfare system.
Participants described having ER social workers at different sites such as the
Family Violence Center and police stations as a valued practice.
3. Ways in which Agency organizational context and culture impacts families and
children of color merits attention. It is possible that characteristics of a large
burea~cratic organization impede the development and facilitation of practices that
would be of benefit to families and children of color. For example, organizational
goals such as having a unified agency mission, a shared organizational culture and
personal contact with clients. may be difficult to achieve in the face of complex and
competing demands inherent in a large organization.
•
Practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County
Offices and adapted to other regions of Santa Clara County, as recommended by
study participants. This practice may be particularly helpful to low-income
families and families of color that may be intimidated by or unfamiliar with child
welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for other
regions of Santa Clara County, though the success of this practice may be linked
to the "small county" dynamic described by South County study participants.
Some participants in focus groups that took place at the main office mentioned
this model as promising while others asserted that specialized expertise in
different areas of social work was an advantage in providing quality services that
are informed by "the most current information."
•
Further examination of ways in which the culture of commitment to maintaining
and reunifying families is created, communicated, and continued merits attention.
In addition, an exploration of how this culture might be adapted to other parts of
the county would be beneficial.
4. More attention that focuses on ways in which system level changes at the federal,
state and local levels have an impact on families and children ofcolor is warranted.
With the passage of the Multi-Ethnic Placement Act-Interethnic Adoption Provision
(MEPA-IEP, 1996) and the Adoption and Safe Families Act (AFSA, 1997), safety,
permanency and expedited placements have taken precedence in the CWS. Results
based on focus group and interview data from all three phases of the study indicated
201
that the shift toward expedited placement and an emphasis on permanency has
certainly influenced the culture and ways in which social services are being delivered
in California in general, and in Santa Clara in particular and consequently raises
several major concerns for children of color in the County's CWS. In combination
with the new regulations, the characteristics of families, children and communities of
color including chronic poverty, substance abuse, lack of community social
organization or racial segregation that increase chances of entering the CWS may
create overwhelming barriers to successful reunification for children of color in the
CWS and keep them in the system longer.
•
Regular assessment of ways in which shortened timelines, early termination of
parental rights, bypass criteria, as well as changes in adoption regulations and
incentives have had an impact .on children and families of color would be
beneficial.
•
A close monitoring of the effect of new policies, implementation of new programs
or changes in the agency's organizational structure on outcomes for families and
children of color would contribute to a better understanding of the differential
impact these innovations might have on different racial and ethnic groups.
5. In order to better assess whether safety, permanency, and family and child well-being
are maintained and enhanced, records should include more complete and measurable
information. We acknowledge that child welfare case records were not originally
intended for research and that that improving the collection and condition of child
welfare data is no easy task given the issues of limitations of the data system,
compatibility among systems of data recording and storage, and training needs.
However, concerted efforts in this area are necessary if accurate and useful
information is to be obtained and used to provide feedback on the effectiveness of
programs and services in improving outcomes for families and children in the
County's CWS.
•
The amount of missing data and the manner in which information is assessed also
implies the need for more reliable and valid methods of recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their families or of
those adopted high or low? And compared to what? Implementing better record
keeping and data collection over time on key characteristics, based on an agreedupon operational definition would better address the question of effectiveness and
performance. Implementing these practices across counties and across states
would also facilitate evaluation.
6. Future research efforts should be directed toward examining neighborhood,
community, and other macro-level factors particular to Santa Clara County and ways
in which these factors interact with individual, family and system-related
characteristics to propel children into and maintain them in the CWS. Research
202
findings consistently point to a relationship between poverty and child maltreatment.
Characteristics associated with communities and neighborhoods of poverty including;
living in a high crime area, living in public housing, having larger numbers of
dependent children, and receiving welfare benefits might place children of color at an
increased risk of entering and staying in the CWS.
It was our intention in conceptualizing the original design and scope of work for the
Children of Color Study to examine community level factors and their interaction
with individual and family characteristics during Phase 3 of the study. However, due
to the interests of various stakeholders, the community component was not included
in the current study. It is our contention that in order to understand more fully the
disproportionate involvement of children and families of color in the CWS, the
processes by which individual, family and system level factors interact with
characteristics that are, perhaps unique to Santa Clara County and impact families and
children of color and their subsequent entry into the CWS need to be examined.
203
References
Akin, B. A., & Gregoire, T. K. (1997). Parents' views on child welfare's response to
addiction. Families in Society: The Journal of Contemporary Human Services, 393-404.
Alter, C. F. (1985). Decision-making factors in cases of child neglect. Child Welfare,
64(2), 99-111.
American Humane Association (2003). What is Family Group Decision Making? Online,
retrieved July 1, 2003, from:
http://www.americanhumane.org?site?PageServer?pagemane=pc_fgdm_what_is.
Ards, S. D., Myers, S. L., Malkis, A., Sugrue, E., & Zhou, L., (2003). Racial
disproportionality in reported and substantiated child abuse and neglect: An examination of
systematic bias. Children and Youth Services Review, 25(5/6) 375-392.
Atkinson, L., & Butler, S. (1996). Court-ordered assessment: Impact of maternal
noncompliance in child maltreatment cases. Children and Youth Services Review, 20(3), 185190.
Avery, R. J. (2000). Perceptions and practice: Agency e{forts for the hardest-to-place
children. Children and Youth Services Review, 22(6), 399-420.
Baird, C., & Wagner, D. (2000). The relative validity of actuarial- and consensus-based
risk assessment systems. Children and Youth Services Review, 22(11/12), 839-871.
Banach, M. ( 1998). The best interests of the child: Decision-making factors. Families in
Society: The Journal of Contemporary Human Services, 79(3), 331-341.
Barth, R. P. (1991). An experimental evaluation of in-home child abuse prevention
services. Child Abuse and Neglect, 15, 363-375.
Barth, R. P., Snowden, L. R., Broeck, E.T., Clancey, T., Jordan, C., & Barusch, A. S.
( 1987). Contributors to reunification or permanent out-of-home care for physically abused
children. Journal of Social Service Research, 9(2/3), 31-45.
Bath, H. I., Richey, C. A., Haapala, D. A. (1992). Child age and outcome correlates in
intensive family pr~servation services. Children and Youth Services Review, 14, 389-406.
Bay Area Social Services Consortium (2000). Concurrent Planning: The Influence of
Bypass and Poor Prognosis Indicators on Child Welfare Outcomes. Berkeley, CA: Author.
Beeman, S. K., Hagemeister, A. K., & Edleson, J. L. (1999). Child protection and
battered women's services: From conflict to collaboration. Child Maltreatment, 4(2), 116-126.
204
Berry, M. (1992). An evaluation of family preservation services: Fitting agency services
to family needs. Social Work. 37(4), 314-321.
Berry, M. (1991 ). The assessment of imminence of risk of placement: Lessons from a
family preservation program. Children and Youth Services Review. 13, 239-256.
Bowen, K. (2000). Child abuse and domestic violence in families of children seen for
suspected sexual abuse. Clinical Pediatrics. 39, 33-40.
Brown, T. G., Werk, A., Caplan, T., & Seraganian, P. (1999). Violent substance abusers
in domestic violence treatment. Violence and Victims. 14(2), 179-190.
Brooks, D., & James, S. (2003). Willingness to adopt Black foster children: Implications
for child welfare policy and recruitment of adoptive parents. Children and Youth Services
Review, 25(5/6), 463-489.
Brunk, M., Henggeler, S. W. & Whelan, J.P. (1987). Comparison ofmultisystemic
therapy and parent training in the brief treatment of child abuse and neglect. Journal of
Consulting and Clinical Psychology. 55(2), 171-178.
California Department of Social Services (2003). Structured Decision Making. Online,
retrieved July 14, 2003 from: http://www.childsworld.ca.gov/Structured 352.htm.
California Department of Social Services (2001). Wraparound Services Pilot. Fiscal
Questions and Answers. Online, retrieved, May 5, 2001 from:
http://childsworld.org/services/famcenwrap.htm
California Department of Social Services (1999). Panel on Race and Bias in Risk
Assessment Thirteenth National Roundtable on Child Protective Services July 7-9, San
Francisco, CA.
Camasso, M. J., & Jagannathan, R. (2000). Modeling the reliability and predictive
validity of risk assessment in child protective services. Children and Youth Services Review,
22(11/12), 873-896.
Carlo, P. (1993). Parent education vs. parent involvement: Which type of efforts work
best to reunify families? Journal of Social Service Research, 17(1/2), 135-150.
Carten, A. J. (1996). Mothers in recovery: rebuilding families in the aftermath of
addiction. Social Work 41(2) 214-223.
Carter-Black, J. (2002). Transracial adoption and foster care placement: Worker
perception and attitude. Child Welfare, 71(2), 337-370.
Casey Foundation (2003a). Team Decision Making. Online, retrieved July 13, 2003 from:
http://aecf.org/initiatives/familytofamily/tools/decisions/htm.
205
Casey Foundation (2003b). Family to Family Home Page. Online, retrieved July 13, 2003
from: http://www.aecf.org/initiatives/familytofamily/
Chaffin, M., Bonner, B. L., & Hill, R. F. (2001). Family preservation and family support
programs: Child maltreatment outcomes across client risk levels and program types. Child Abuse
and Neglect. (25), 1269-1289.
Chao, R., & Tseng, V. (2002). Parenting of Asians. In M. H. Bornstein (Ed.). Handbook
of Parenting Vol. 4 Social Conditions and Applied Parenting (pp. 59-93). Mahwah, NJ:
Lawrence Erlbaum Associates.
Cash, S. J., & Berry, M. (2002). Family characteristics and child welfare services: Does
the assessment drive service provision? Families in Society, 83(5/6), 499-507.
Children NOW. (2001). California Report Card 2001: County Information. Online,
retrieved, August 20 2003, from: http://www.childrennow.org/california/rc-2001/county-pdflist.htm#top2.
Children's Research Center (2003). Structured Decision Making. Online, retrieved
July15, 2003 from: http://www.nccd-crc.org/crc/decision.htm.
Clark, H. W. (2001). Residential substance abuse treatment for pregnant and postpartum
women and their children: Treatment and policy implications. Child Welfare, 80(2), 179-198.
Close, M. M. (1983). Child welfare and people of color: Denial of equal access. Social
Work Research and Abstracts, 19(4), 13-20.
Cohen, E. P. (2003). Framework for culturally competent decision-making in child
welfare. Child Welfare, 82(2), 143-155.
Courtney, M., Piliavin, I., Grogan-Kaylor, A., & Nesmith, A. (2001). Foster youth
transitions to adulthood: A longitudinal view of youth leaving care. Child Welfare, 80, 685-717.
Courtney, M. E. (1997). The politics and realities oftransracial adoption. Child Welfare,
76(6), 749-779.
Courtney, M. E., Barth, R. P., Berrick, J. D., Brooks, D., Needell, B., & Park, L. (1996).
Race and child welfare services: Past research and future directions. Child Welfare 75(2) 99-137.
Courtney, M. E. (1994). Factors associated with the reunification of foster children with
their families. Social Service Review, 68(1), 81-108.
Cuddeback, G. S., & Orme, J. G. (2002). Training and services for kinship and
nonkinship foster families. Child Welfare, 91(6), 879-909.
206
Dalgleish, L. I., & Drew, E. C. (1989). The relationship of child abuse ind~cators to the
assessment of perceived risk and to the court's decision to separate. Child Abuse and Neglect,
ll, 491-506.
Davidson-Arad, B. (2001). Predicted changes in children's quality of life in decisions
regarding the removal of children at risk from their homes. Children and Youth Services Review,
23(2), 127-143.
DePanfilis, D. (1996). Social isolation of neglectful families: A review of social support
assessment and intervention models. Child Maltreatment, 1(1 ), 37-52.
DePanfilis, D., & Zuravin, S. J. (2001). Assessing risk to determine the need for services.
Children and Youth Services Review, 23(1 ), 3-20.
DePanfilis, D., & Scannapieco, M. (1994). Assessing the safety of children at risk of
maltreatment: Decision-making models. Child Welfare, 73(3), 229-245.
Drake, B. (1996). Predictors of preventive services provision among unsubstantiated
cases. Child Maltreatment, 1(2), 168-175.
Dore, M. M. & Lee, J.M. (1999). The role of parent training with abusive and neglectful
parents. Family Relations, 48(3), 313-325.
dosReis, S., Zito, J.M., Safer, D. J., Soeken, K. L. (2001). Mental health services for
youths in foster care and disabled youths. American Journal of Public Health, 91(7). 10941100.
Doueck, H.J., English, D. J., DePanfilis, D., Moote, G. T. (1993). Decision-making in
child protective services: A comparison of selected risk-assessment systems. Child Welfare,
72(5), 441---452.
Downing, J. D., Wells, S. J., & Fluke, J. (1990). Gatekeeping in child protective services:
A survey of screening policies. Child Welfare, 69(4), 357-369.
Dunford, F. W. (2000). The San Diego navy experiment: An assessment of interventions
for men who assault their wives. Journal of Consulting and Clinical Psychology, 68(3), 468476.
Dutton, D. G. (1986). The outcome of court-mandated treatment for wife assault: A
quasi-experimental evaluation. Violence and Victims, 1(3), 163-175.
Eckenrode, J., Ganzel, B., Henderson, C.R., Smith, E., Olds, D. L., Powers, J., Cole, R.,
Kitzman, H., & Sidora, K. (2000). Preventing child abuse and neglect with a program of nurse
home visitation: The limiting effects of domestic violence. The Journal of the American Medical
Association, 284(11), 1385-1396.
207
Eckenrode, J., Powers, J., Doris, J., Munsch, J., & Bolger, N. (1988). Substantiation of
child abuse and neglect reports. Journal of Consulting and Clinical Psychology, 56(1 ), 9- 16.
English, D. J., Marshall, D. B., Coghlan, L., Brummel, S., & Orme, M. (2002). Causes
and consequences of the substantiation decision in Washington state child protective s~rvices.
Children and Youth Services Review, 24(11), 817-851.
Farmer, E. M., Wagner, H. R., Bums, B. J., & Richards, J. T. (2003). Treatment foster
care in a system of care: Sequences and correlates of residential placements. Journal of Child and
Family Studies, 12(1), 11-25.
Federal Interagency Forum on Child and Family Statistics. America's Children: Key
National Indicators of Well-being, 2001. Federal Interagency Forum on Child and Family
Statistics, Washington DC: U.S. Government Printing Office.
Fein, E., & Staff, I. (1993). Last best chance: Findings from a reunification services
program. Child Welfare, 72(1), 25-40.
Fetsch, R. J., Schultz, C. J., & Wahler, J. J. (1999). A preliminary evaluation of the
Colorado rethink parenting and anger management program. Child Abuse and Neglect. 23(4),
353-360.
Findlater, J.E., & Kelly, S. (1999). Reframing child safety in Michigan: Building
collaboration among domestic violence, family preservation, and child protection teams. Child
Maltreatment, 4(2), 167-174.
359.
Flanigan, J. ( 1954). The Critical Incident Technique. Psychological Bulletin. 4, 327-
Fluke, J. D., Yuan, Y. T., Hedderson, J., & Curtis, P.A. (2003). Disproportionate
representation of race and ethnicity in child maltreatment: Investigation and victimization.
Children and Youth Services Review, 25(5/6), 359-373.
Forehand, R., & Kotchick, B. A. (1996).·Cultural diversity: A wake-up call for parent
training. Behavior Therapy, 27, 187-206.
Fraser, J. A., Armstrong, K. L., Morris, J.P. & Dadds, M. R. (2000). Homes visiting
intervention for vulnerable families with newborns: Follow-up results of a randomized controlled
trial. Child Abuse and Neglect, 24(11), 1399-1429.
Garbarino, J., & Kostelny, K. (1992). Child maltreatment as a community problem. Child
Abuse and Neglect, 16, 455-464.
Garland, A. F., & Besinger, B. A. (1997). Racial/ethnic differences in court referred
pathways to mental health services for children in foster care. Children and Youth Services
Review. 19(8), 651-666.
208
Garland, A. F., Hough, R., Landsverk, J. A., McCabe, K. M., Yeh, M., Ganger, W. C., &
Reynolds, B. (2000). Racial/ethnic differences in mental health care utilization among children
in foster care. Children's Services: Social Policy, Research and Practice, 3, 133-146.
Gaudin, J.M., Wodarski, J. S., Arkinson, M. K., & Avery, L. S. (1990). Remedying child
neglect: Effectiveness of social network interventions. The Journal of Applied Sciences, 15( 1),
97-123.
Gillespie, J.M., Byrne, B., Workman, L. J. (1995). An intensive reunification program
for children in foster care. Child and Adolescent Social Work Journal, 12(3), 213-228.
Gregoire, K. A. & Schultz, D. J. (2001). Substance-abusing child welfare parents:
Treatment and child placement outcomes. Child Welfare 80(4), 433-452
Goerge, R. M. (1990). The reunification process in substitute care. Social Service
Review, 64(3), 422-457.
Gordon, A. L., McKinley, S. E., Satterfield, M. L., & Curtis, P. (2003). A first look at the
need for enhanced support services for kinship caregivers. Child Welfare, 81(1), 77-96.
Guterman, N. B. (1999). Enrollment strategies in early home visitation to prevent
physical child abuse and neglect and the "universal versus targeted" debate: A meta-analysis of
population-based and screening-based programs. Child Abuse and Neglect, 23(9), 863-890.
Haller, D. L., Knisely, J. S., Elswick, R. K., Dawson, K. S., Scholl, S. H. (1997).
Perinatal substance abusers: Factors influencing treatment retention. Journal of Substance Abuse
Treatment, 14(6), 513-519.
Harman, J. S. (2001). Mental health care utilization and expenditures by children in foster
care. The Journal of the American Medical Association, 285(5), 524.
Harris, M. S., & Courtney, M. E. (2003). The interaction of race, ethnicity, and family
structure with respect to the timing of family reunification. Children and Youth Services Review,
25(5/6), 409-429.
Hines, A. M., Lee, P.A., Drabble, L., Snowden, L. R., & Lemon, K. (2002). An
Evaluation of Factors Related to the Disproportionate Representation of Children of Color in
Santa Clara County's Child Welfare System: Phase 2 Final Report. Report submitted to Santa
Clara County, CA Department of Social Services, Family and Children's Services, August 31,
2002.
Hochstadt, N. J., Harwicke, N. J. (1985). How effective is the multidisciplinary
approach? A follow-up study. Child Abuse and Neglect, 9, 365-372.
209
Hosley, C. A., Gensheimer, L., & Yang, M. (2003). Building effective working
relationships across culturally and ethnically diverse communities. Child Welfare. 82(2), 157168.
Jellinek, M. S., Murphy, M., Poitrast, F., Quinn, D., Bishop, S. J., & Goshko, M. (1992).
Serious child mistreatment in Massachusetts: The course of 206 children through the courts.
Child Abuse and Neglect, 16, 179-185.
Johnson, P ., & Cahn, K. ( 1992). Improving child welfare practice through improvements
in attorney-social worker relationships. Child Welfare, 74(2), 383-394.
Jones, L. (1998). The social and family correlates of successful reunification of children
in foster care. Children and Youth Services Review, 20(4), 305-323.
Jones, L. (1993). Decision making in child welfare: A critical review of the literature.
Child and Adolescent Social Work Journal, 10(3), 241-262.
Kapp, S. A., McDonald, T. P., & Diamond, K. L. (2001). The path to adoption for
children of color. Child Abuse and Neglect, 25, 215-229.
Karoll, B. R. & Poertner, J. (2002). Judges' caseworkers', and substance abuse
counselors' indicators of family reunification with substance-affected parents. Child Welfare,
71(2), 249-269.
Karski, R. L. (1999). Key decisions in child protective services: Report investigation and
court referral. Children and Youth Services Review, 21(8), 643-656.
Kemp, S. P., & Bodonyi, J.M. (2002). Beyond termination: Length of stay and predictors
of permanency for legally free children. Child Welfare, 81(1), 59-86.
Kinard, E. M. (2002). Services for maltreated children: variations by maltreatment
characteristics. Child Welfare, 81(4), 617-645.
Kitzman, H., Olds, D, L., Henderson, C. R., Hanks, C., Cole, R., Tatelbaum, R.,
McConnochie, K. M., Sidora, K., Luckey, D. W., Shaver, D., Engelhardt, K., James, D., &
Barnard, K. (1997). Effect of prenatal and infancy home visitation by nurses on pregnancy
outcomes, childhood injuries, and repeated childbearing: A randomized controlled trial. The
American Medical Association, 278(8), 644-652.
Knepper, P. E., & Barton, S. M. (1997). The effect of courtroom dynamics on child
maltreatment proceedings. Social Service Review, 71(2), 288-309.
Knight, D. K., Logan, S. M., & Simpson, D. D. (2001). Predictors of program completion
for women in residential substance abuse treatment. American Journal of Drug and Alcohol
Abuse, 27(1), 1-15.
210
Kolbo, J. R., & Strong, E. (1997). Multidisciplinary team approaches to the investigation
and resolution of child abuse and neglect: A national study. Child Maltreatment, 2(1), 61-72.
Landsman, M. J. (2002). Rural child welfare practice from an organization-inenvironment perspective. Child Welfare, 81 (5), 790-819.
Landy, S., & Munro, S. (1998). Shared parenting: Assessing the success of a foster parent
program aimed at family reunification. Child Abuse and Neglect. 22(4), 305-318.
Lee, B. J., & Goerge, R. M. (1999). Poverty, early childbearing, and child maltreatment:
A multinomial analysis. Children and Youth Services Review, 21(9/10), 755-780.
Lewandowski, C. A., & GlenMaye, L. F. (2002). Teams in child welfare settings:
Interprofessional and collaborative processes. Families in Society, 83(3), 245-256.
Lewandowski, C. A., & Pierce, L. (2002). Assessing the effect of family-centered out-ofhome care on reunification outcomes. Research on Social Work Practice, 12(2), 205-221.
Lightbourne, W. (2002). Greenbook Project Update. Report submitted to the Santa Clara
County Board of Supervisors.
Lindsey, D. (1991). Factors affecting the foster care placement decision: An analysis of
national survey data. American Journal of Orthopsychiatry, 61 (2), 272-281.
Littell, J. H. (1997). Effects of the duration, intensity, and breadth of family preservation
services: A new analysis of data from the Illinois Family First Experiment. Children and Youth
Services Review, (19)2, 17-39.
Littell, J. H., & Schuennan, J. R. (1995). A synthesis of research on family preservation
and family reunification programs. Online, retrieved January 27, 2003 from:
http://aspe.hhs.gov/hsp/cyp/fp litrev .htm
Little Hoover Commission (2003). Still in Our Hands: A Review of Efforts to Refonn
Foster Care in California. Online, retrieved July 24, 2003 from:
http://www.lhc.ca.gov/lhcdir/report168.html
Maluccio, A. N. & Ainsworth, F. (2003). Drug use by parents: A challenge for family
reunification practice. Children and Youth Services Review, 25(7), 511-533.
Martin, L. M., Peters, C. L., & Glisson, C. (1998). Factors affecting case management
recommendations for children entering state custody. Social Service Review,--, 521-544.
McGuigan, W., Katzev, A. R., & Pratt, C. C. (2003). Multi-level determinants of
retention in a home-visiting child abuse prevention program. Child Abuse & Neglect, 27, 363380.
211
McGuigan, W., M., & Pratt, C. C. (2001). The predictive impact of domestic violence on
three types of child maltreatment. Child Abuse & Neglect. 25, 869-883.
McKay, M. M. (1994). The link between domestic violence and child abuse: Assessment
and treatment considerations. Child Welfare, 73(1), 29-39.
McMurty, S., & Lie, G. (1992). Differential exit rate of minority children in foster care.
Social Work Research a~d Abstracts, 28(1), 42--48.
McPhatter, A. R., & Ganaway, T. L. (2003). Beyond the rhetoric: Strategies for
implementing culturally effective practices with children, families and communities. Child
Welfare, 82(2), 103-124.
Mederos, F., & Woldeguiorguis, I. (2003). Beyond cultural competence: What child
protection managers need to know and do. Child Welfare, 82(2), 125-142.
Menanteau-Horta, D., & Yigzaw, M. (2002). Indicators of social well-being and
elements of child welfare in Minnesota rural counties. Child Welfare, 81(5), 709-735.
Metsch, L. R., Wolfe, H.P., Fewell, R., McCoy, C. B., Elwood, W. N., Wohler-Torres,
B. et al. (2001). Treating substance-using women and their children in public housing:
Preliminary evaluation findings. Child Welfare, 80(2), 199-220.
Miller, 0. A., & Gatson, J. (2003). A model of culture-centered child welfare practice.
Child Welfare, 82(2), 235-249.
Merrel, T. M., Elliot, J. D., Murphy, C. M., & Taft, C. T. (2003). Cognitive behavioral
and supportive group treatments for partner-violent men. Behavior Therapy, 34, 77-95.
Murphey, D. A., & Braner, M. (2000). Linking child maltreatment retrospectively to birth
and home visit records: An initial examination. Child Welfare, 79(6), 711-728.
Needell, B., Brookhart, M.A., Lee, S. (2003). Black children and foster care placement
in California. Children and Youth Services Review, 25(5/6) 393--408.
Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A., Lery,
B., Shaw, T., Dawson, W., Piccus, W., Magruder, J., & Kim, H. (2003). Child Welfare Services
Reports for California. Online, retrieved June 28, 2003, from: University of California at
Berkeley Center for Social Services Research website. URL:
<http://cssr .berkeley .edu/CWSCMSreports/>
Nicholson, B. C., Brenner, V., & Fox, R. A. (1999). A community-based parenting
program with low-income mothers of young children. Families in Society, 80(3), 247-253.
Olds, D. L., Eckenrode, J., Henderson, C.R., Kitzman, H., Powers, J., Cole, R., Sidora,
K., Morris, P., Pettitt, L. M., & Luckey, D. (1997). Long-term effects of home visitation on
212
material life course and child abuse and neglect: Fifteen-year follow-up of a randomized trial.
The Journal of the American Medical Association, 278(8) 637-644.
Ondersma, S. J., Malcoe, L. H., & Simpson, S. M., (2001). Child protective services'
response to prenatal drug exposure: Results from a nationwide survey. Child Abuse and Neglect.
~657-668.
Palmer, S. E. (1996). Placement stability and inclusive practice in foster care: An
empirical study. Children and Youth Services Review, 18(7), 589-601.
Palmer, S. E., Brown, R. A., & Barrera, M. E. (1992). Group treatment for aqusive
husbands: Long-term evaluation. American Journal ofOrthopsychiatry, 62(2), 276-283.
Patterson, G. R., Chamberlain, P., & Reid, J.B. (1982). A comparative evaluation of a
parent-training program. Behavior Therapy. 13, 638-650.
Pecora, P. J., Fraser, M. W., Bennett, R. B., & Haapala, D. A. (1991). Placement rates of
children and families served by the Intensive Family Preservation Services Programs. In M. W.
Fraser, P. J. Pecora, & D. A. Haapala (Eds.), Families in Crisis: The Impact oflntensive Family
Preservation Services (pp.149-180). New York: Aldine De Gruyter.
Pennell, J., & Burford, G. (2000). Family group·decision making: Protecting children and
women. Child Welfare. 79(2), 131-158.
Rossi, P.H., Schuerman, J., & Budde, S. (1999). Understanding decisions about child
maltreatment. Evaluation Review, 23(6), 579-598.
Reiff, M. I. (2001 ). Mental health care utilization and expenditures by children in foster
care. Journal of Developmental and Behavioral Pediatrics, 22(2), 151.
Reyes, B. I. (Ed.). (2001). A Portrait of Race and Ethnicity in California. San Francisco:
Public Policy Institute of California.
Russel, R. (1988). Role perceptions of attorneys and caseworkers in child abuse cases in
juvenile court. Child Welfare, 67(3), 205-216.
Sanchirico, A., & Jablonka, K. (2000). Keeping foster children connected to their
biological parents: The impact of foster parent training and support. Child and Adolescent Social
Work Journal, 17(3), 185-203.
Scannapieco, M. (1994). Home-based services program: Effectiveness with at risk
families. Children and Youth Services Review, 16(5/6), 363-377.
Schriver, J.M. (1998). Human Behavior and the Social Environment. Allyn and Bacon:
Boston.
213
Schuerman, J., Rossi, P.H., & Budde, S. (1999). Decisions on placement and family
preservation: Agreement and targeting. Evaluation Review, 23(6), 599-618.
Sheppard, D. I., & Zangrillo, P.A. (1996). Coordinating investigations of child abuse .
.Public Welfare, Winter, 21-31.
Sieppert, J. D., Hudson, J., & Unrau, Y. (2000). Family group conferencing in child
welfare: Lessons from a demonstration project. Families in Society, 81(4), 382-391.
Smith, B. D. (2003). How parental drug use and drug treatment compliance relate to
family reunification. Child Welfare, 82(3), 335-365.
Spaid, W. M., Fraser, M. W., & Lewis, R. E. (1991 ). Changes in family functioning: Is
participation in Intensive Family Preservation Services correlated with changes in attitudes or
behaviors? In M. W. Fraser, P. J. Pecora, & D. A. Haapala (Eds.), Families in Crisis: The Impact
oflntensive Family Preservation Services (pp.131-148). New York: Aldine De Gruyter.
Stevenson, J. ( 1999). The treatment and long-term sequelae of child abuse. Journal of
Child Psychology and Psychiatry, 40(1), 89-111.
Suleiman, L. P. (2003). Beyond cultural competence: Language access and Latino civil
rights. Child Welfare, 82(2). 185-200.
Taft, C. T., Murphy, C. M., Elliot, J. D., & Keaser, M. C. (2001). Race and demographic
factors in treatment attendance for domestically abusive men. Journal of Family Violence, 16(4),
385-400.
Templeman, S. B., & Mitchell, L. (2002). Challenging the one-size-fits-all myth:
Findings and solutions from a statewide focus group of rural social workers. Child Welfare,
fil.(5), 757-772.
Thieman, A. A., & Dail, P. W. (1997). Predictors of out-of-home placement in a family
preservation program: Are welfare recipients particularly vulnerable? Policy Studies Journal,
25(1) 124-139.
Toffalo, D. A. (2000). An investigation of treatment integrity and outcomes in
wraparound services. Journal of Child and Family Studies, 9(1), 351-361.
Tumlin, K. C., & Geen, R. (2000). The Decision to Investigate: Understanding State
Child Welfare Screening Policies and Practices. Online, retrieved, February 7, 2003 from:
http://www.urban.org/url.cfm?ID=309534
U.S. Census Bureau (2001). Overview of Race and Hispanic Origin, Census 2000 Brief.
Online, retrieved November 20, 2002 from: http://www.census.gov/prod/2001pubs/c2kbr01l.pdf.
214
U.S. Census Bureau (2000). State and County Quick Facts: California. Online, retrieved
March 3, 2003 from: http://quickfacts.census.gov/qfd/states/06000.html
U.S. Department of Health and Human Services, Administration for Children, Youth and
Families, Children's Bureau (2003a). Decision-making in Unsubstantiated Child Protective
Services Cases. Washington DC: U.S. Government Printing Office.
U.S. Department of Health and Human Services. (2003b). Protecting the Well-being of
Children. Online, retrieved April 16, 2003 from:
http://www.scf.hhs.gov/news/facts/chilwelfpr.htm
U.S. Department of Health and Human Services. (2002). The Adoption and Foster Care
Analysis Reporting System Report. Online, retrieved July 16, 2003 from:
·
http://www.scf.hhs.gov/programs/cb/publications/afcars/report7 .htm.
U.S. Department of Health and Human Services, Administration for Children, Youth and
Families (2001). Child Maltreatment 2001. Washington DC: U.S. Government Printing Office.
U.S. General Accounting Office (1998). Agencies Face Challenges Securing Stable
Homes for Children of Substance Abusers. GAO/HEHS 98-182. Washington DC: U.S.
Government Printing Office.
U.S. General Accounting Office (1997). Child Welfare: States' Progress in Implementing
Family Preservation and Support Services. GAO/HEHS 97-34. Washington DC: U.S.
Government Printing Office.
Vidal de Haymes, M., & Simon, S. (2003). Transracial adoption: Families identify issues
and needed support services. Child Welfare, 82(2), 251-272.
Wald, M. S., Woolverton, M. (1990). Risk assessment: The emperor's new clothes?
Child Welfare, 69(6), 483-511.
Walton, E., Fraser, M. W., Lewis, R. E., Pecora, P. J., & Walton, W. K. (1993). In-home
family-focused reunification: An experimental study. Child Welfare, 72(5), 473-487.
Wells, K., & Guo, S. (2003). Mothers' welfare and work income and reunification with
children in foster care. Children and Youth Services Review, 25(3), 203-224.
Wells, K., & Guo, S. (1999). Reunification and reenty of foster children. Children and
Youth Services Review, 21(4), 273-294.
Wells, K., & Whittington, D. (1993). Child and family functioning after intensive family
preservation services. Social Service Review, March, 55-83.
Wells, S. J., Fluke, J. D., & Brown, C.H. (1995). The decision to investigate: Child
protection practice in 12 local agencies. Children and Youth Services Review, 17(4), 523-546.
215
Westat, Inc. (2001a). Caseworker decision-making. In Assessing the context of
permanency and reunification in the foster care system. Online, retrieved January 27, 2003 from:
http://aspe.hhs.gov/hsp/fostercare-reunifO 1/chapter7 .htm
Westat, Inc. (2001b). Evaluation of family preservation and reunification programs.
Online, retrieved January 27, 2003 from: http://aspe.hhs.gov/hsp/fampres94/index.htm
Westat, Inc. (2001c). The role of race in parental reunification. In Assessing the context
of permanency and reunification in the foster care system. Online, retrieved January 27, 2003
from: http://aspe.hhs.gov/hsp/fostercare-reunito l /chapter6.htm
Whitney, P., & Davis, L. (1999). Child abuse and domestic violence in Massachusetts:
Can practice be integrated in a Public Child Welfare Setting? Child Maltreatment, 4(2), 158166.
Wolfe, D. A., Edwards, B., Manion, I., & Koverola, C. (1988). Early intervention for
parents at risk of child abuse and neglect: A preliminary investigation. Journal of Consulting and
Clinical Psychology, 56(1), 40-47.
Wolock, I. (1982). Community characteristics and staff judgments in child abuse and
neglect cases. Social Work Research and Abstracts, 18(2), 9-15.
1
I.
INTRODUCTION
An analysis of national data on the ethnic composition of children in out-of-home-placement
reveals that, compared to their presence in the general population, there is a disproportionate
involvement of children of color in the public child welfare system (CWS). Statistics indicate that
African American and Native American children are overrepresented in the CWS. African
American children represent 15 percent of the general child population, yet comprise approximately
42 per~ent of children in the CWS and Native American children constitute approximately 1 percent
of the child population and represent approximately 2 percent of the CWS. Though not all states
provide data on Hispanic/Latino children in the CWS, aggregate statistics from those that do reveal
that while Hispanic/Latino children make up approximately 16 percent of the national child
population, they comprise 15 percent of the CWS population. Asian American/Pacific Islanders and
Whites also tend to be underrepresented at the national level. Asian American/Pacific Islander
children comprise approximately 4 percent of the general child population and 1 ·percent of the
CWS, while White children who are approximately 64 percent of the general e:hild population,
constitute 36% of children in the CWS (Federal Interagency Forum on Child and Family Statistics,
2001; U.S. Census Bureau, 2001; U.S. Department of Health and Human Services, 1999; U.S.
Department of Health.and Human Services, 2001).
Further analysis at the California level provides important information on differences and
trends that may go undetected at the national level. In California, a large and ethnically diverse
state, African Americans constitute 6 percent of the general population (32 percent of whom are
below the age of 18), but represent approximately 32.9 percent of children in supervised foster care.
Hispanics/Latinos comprise approximately 32 percent of the general population in California with
· 43 percent being below the age of 18, and represent 36.2 percent of children in the state CWS.
Native Americans represent 1.0 percent of California's population and constitute approximately 1.2
percent of the children in the CWS. Asian American/Pacific Islanders are under-represented in
California's CWS, as they constitute approximately 11 percent of California's general population,
with approximately 30 percent being younger than 18, but represent just 2.1 percent of the children
in the CWS. While Whites comprise approximately 47 percent of the population in the state, with
only 20 percent being below the age of 18, they constitute 27.5 percent of the children in
California's CWS (Needell, Webster, Cuccaro-Alamin, Armijo, Lee, Brookhart et al., 2003.; Reyes,
2001; U.S. Census Bureau, 2000).
In Santa Clara County, there also exists a disproportionate representation of children of
color in the County's CWS. When compared to the general population, African American,
Hispanic/Latino and Native American children are overrepresented in Santa Clara County's CWS,
while Asian American/Pacific Islander and White children are underrepresented. African Americans
represent 4 percent of the general child population in the county, and are 13.6 percent of children in
supervised foster care. Hispanic/Latino children represent 30 percent of the general child population
in Santa Clara County and constitute 55.3 percent of the child welfare cases. Native Americans are
approximately 0.5 percent of Santa Clara County's population and represent 0.6 percent of children
in the CWS. Asian American/Pacific Islander children represent 21 percent of the general county
child population and 5.4 percent of children in the CWS; Whites constitute 45 percent of the general
child population and 25.0 percent of the child welfare population (Needell et al., 2003; U.S. Bureau
of the Census, 2000).
2
Project Description
The disproportionate involvement of children of color in the CWS has long been an issue of
concern for CWS workers, clients, researchers and government and community groups; yet few
studies have systematically investigated the factors associated with this disproportionality. In an
effort to better understand the factors related to the disproportionate number of children of color in
the CWS in Santa Clara County, the Department of Family and Children Services (DFCS)
contracted with the Child Welfare Research Team (CWRT) in the College of Social Work at San
Jose State University to conduct a three-year study on this topic.
The overall question posed by DFCS was: What are the primary reasons why children of
color are disproportionately represented in Santa Clara County's Child Welfare System? In order to
address the complexity of this question, the CWRT elected to employ a multiphase/multimethod
approach, beginning with an initial exploratory phase that was completed in May 2001. Data for
Phase I were gathered from three sources: research literature at the national, state and county levels;
Santa Clara County's management information system (CWS/CMS); and focus group discussions
with professionals in Santa Clara County who provide child welfare services, as well as parents,
caregivers and youth who are recipients ofCWS services. Four central themes emerged from Phase
I findings {please see Phase I Final Report for a complete description of the findings from which
these themes were generated). The four emerging themes from Phase I included:
I). Little is known about specific pathways through the CWS and ways in which these
pathways differ for various racial/ethnic groups. Much of the research on children in the
CWS focuses on factors relates to movement in and out of the system. The specific
pathways through the system for children of various ethnicities is largely unknown, and may
provide important insights into the factors related to the disproportionate involvement of
children of color. Focus group participants in Phase I frequently stated that once a child of
color enters the system, it is very hard for that child to exit.
2). Various racial/ethnic groups may receive different treatment at key decision-making
points in the system. Differential treatment of children of color in both the initial reporting of
child maltreatment and throughout the various stages within the CWS may exist. CWS/CMS
data for Santa Clara County corroborate the possible differential treatment of children of
color at various points in the system. Phase I analysis revealed that African American
children spent significantly more time in placement than their White, Hispanic/Latino, and
Asian American/Pacific Islander peers. Differences in placement type were also foundalthough kincare (placement with a relative) was the most frequent out of home placement
(OHP) type across ethnicities, the second most frequent OHP type varied among
racial/ethnic groups. For African American, Native American, White, and Hispanic/Latino
children, the second most common OHP type was a Foster Family Agency. For Asian
American/Pacific Islander children, the second most frequent placement was a Foster
Family Home. Asian American/Pacific Islander (18%) children were also placed at the
Children's Shelter at higher percentages than children of other racial/ethnic groups. In
addition, focus group participants in Phase I expressed concern for possible racial/ethnic
differences in ways in which services and resources were allocated and ways in which the
judicial system handled cases.
3
3). System level changes recently initiated at the federal, state and local level will
undoubtedly have an impact on children ofcolor in the CWS. Recent federal policies,
including the Multi-Ethnic Placement Act-Interethnic Adoption Provision (MEPA-IEP,
1996); the Adoption and Safe Families Act (AFSA, 1997); the Personal Responsibility and
Work Opportunity Reconciliation Act (PRWORA, 1996); as well as state and local county
policies all impact the ways in which families and children of color enter and stay in the
CWS. Recent policy shifts toward expedited permanent placements for children in out-ofhome care and shortened timelines for reunification, as well as new restrictions for
Temporary Assistance to Needy Families (TANF) recipients (more than half of children in
foster care are from welfare eligible families) may increase the likelihood of children of
color entering and staying in the CWS-although much more research is still needed to
determine this association.
4). Factors related to the disproportionate representation ofchildren of color in the CWS
are multiple and complex. No clear consensus from the research literature exists on how
families and children of color become and stay involved in the CWS. Available evidence
suggests that, rather than one primary cause, there appear to be numerous and interrelated
factors associated with disproportionate rates of children of color in the CWS. Factors found
to be associated with CWS involvement for children of all ethnicities, include but may not
be limited to: parental mental illness substance abuse, and incarceration; spousal abuse;
living in poverty; living in neighborhoods with concentrated poverty, and living in
communities with low levels of social organization. For children of color specifically,
research also indicates that race and class biases in initial reporting and subsequent CWS
service delivery do exist.
Conclusions from Phase 1 indicated that more research was needed that focused on the
actual experiences of children in care and the individual and family-related characteristics that are
associated with these experiences. Phase 1 results also indicated that CWS/CMS data alone would
not provide the information needed to address these research aims. Thus, Phase 2, which concluded
in August 2002 focused on characteristics of children and families of color in the CWS and
experiences they had as they wended their way through the system. The methodology for Phase 2
included extensive, in-depth case record reviews with a sample of 403 closed child welfare case
records, enabling us to collect information on the individual child and family, as well as systemrelated characteristics not included in the CWS/CMS database. The case record review data
provided extensive information on decisions at key points in the system and children's progress
through the system, as well as services ordered for children at the time of case disposition and at
subsequent hearings. In addition, a parallel descriptive analysis of 1720 closed cases within the
CWS/CMS database allowed us to examine the entire histories of children as they progressed
through the CWS and to conduct analyses that were not possible with the CWS/CMS crosssectional dataset that we used in Phase 1. Key informant interviews with managers and supervisors
in the county's DFCS at key decision points in the system provided additional information
pertaining to cultural and environmental aspects. of the agency. (Please see Phase 2 Final Report for
a complete description of the study methods and findings). Based on analyses conducted during
Phase 2, five overall conclusions were drawn:
1). Families belonging to each ofthe four racial/ethnic groups studied, including: African
Americans, Hispanic/Latinos, Asian American/Paci.fie Islanders and Whites are
characterized by unique constellations ofriskfactors.
4
2). The characteristics that emerged in our analyses as key characteristics of each ofthe
four racial/ethnic groups have been associated in the research literature as risk/actors/or
child abuse and neglect. These characteristics have also been associated with increased
serious behavioral and adjustment problems in children and adolescents.
3). Once in the child welfare system, children in each ofthe four racial/ethnic groups follow
different pathways and experience different outcomes.
4). The services ordered/or families ofcolor are generally limited to aone-size-fits all
approach and to a small array of available services. The reliance on traditional formal
services does not appear to meet the needs of these culturally diverse families and children.
5). There is a need to involve multiple social service systems in a comprehensive and
coordinated effort to meet the needs ofchildren and families ofcolor.
The current report provides findings from Phase 3 that was originally to be conducted from
September 2002 to August 31, 2003. However, due to contractual changes involving the Scope of
Work, the CWRTwas not able to begin work until the end of January 2003. Thus, the research
involved in Phase 3 was conducted over a shorter seven-month tfme period. This third phase of the
research project focused on agency practice and ways in which current practice interacts with child,
family and cultural characteristics of different racial/ethnic groups. Phase 3 of the study addressed
three overall aims: 1) to assess identified key practices for children and families of color in Santa
Clara County's CWS, 2) to examine the influence of the court system on service recommendations
for children and families and, 3) to conduct a statewide comparative analysis of identified key
practices that may affect the disproportionate representation of children of color in the CWS.
The primary methodology for Phase 3 included semi-structured in-depth interviews and
focus groups with DFCS supervisors and managers, social workers and parents involved in the
CWS; a qualitative and quantitative comparison of child welfare practices and outcomes between
the DFCS South County and Main County offices; a reanalysis of cases from Phase 2 with
successful outcomes and an analysis of factors that contributed to those positive outcomes; a
supplemental data collection for our case record review sample to gather information on the role of
the court system in child welfare practice, including court ordered changes to the social worker
recommendations at the jurisdictional/dispositional hearing; and, a statewide survey of county child
welfare directors, managers or supervisors in California counties and quantitative analyses using a
statewide database consisting of county-level characteristics.
Overview of the Report
The next section of this report provides an overview of the study objectives and approach for
Phase 3, as well as the specific research questions that guide the analyses presented in each section.
The third section reviews the recent research on evidence-based child welfare practices and their
application to children and families of color. Section four focuses on the impact of child welfare
practices on children and families of color in the Santa Clara County's Department of Family and
Children Services (DFCS). Several related analyses were conducted to understand the types of
practices and case characteristics that are associated with more successful outcomes for children and
families from various racial/ethnic groups; the types of child welfare practices used in Santa Clara
County; how practices in the Main County Offices differ from South County Office practices; how
5
practices interact with child, family and cultural characteristics of differing racial/ethnic groups to
produce various outcomes; and the role of the court system on child welfare practices. Section five
describes results from our statewide comparative analysis of effective practices for children and
families of color in the child welfare system. A summary of the report, as well as recommendations
and conclusions are contained in section six.
Definition of Terms used in this Report
Racial/ethnic group refers to cultural heritage and country of origin. Racial/ethnic group is used
interchangeably with the terms, children of color and ethnic group.
Disproportionate representation ofchildren ofa particular racial/ethnic group refers to the
difference between the proportion of children of a particular racial or ethnic group in the CWS and
the proportion of children of a particular racial or ethnic group in the general population.
Overrepresentation of children ofa particular racial/ethnic group exists when the proportion of
children of a certain racial or ethnic group in the CWS exceeds its proportion in the general
population.
Underrepresentation ofchildren of a particular racial/ethnic group exists when the proportion of
children of a certain racial or ethnic group in the CWS is less than its proportion in the general
population.
African American refers to children of African American heritage. The term Black is included in
this category.
Asian American/Pacific Islander refers to children of Asian American and Pacific Islander heritage.
The following population groups are included: Asian Indian, Cambodian, Chinese, Filipino,
Guamanian, Hawaiian, Hmong, Japanese, Korean, Laotian, Polynesian, Samoan and Vietnamese
and other Asian American/Pacific Islanders.
Hispanic/Latino/a refers to children of Mexican American and Latin American heritage. The terms
Hispanic and Latino/a are included in this category.
Native American refers to children of Native American heritage and includes those designated as
Alaska Natives/Aleuts. The term American Indian is included in this category.
White refers to children of European heritage. The term Caucasian is included in this category.
Other category - Those in the Other ethnic subgroup included Native Americans, Ethiopians,
White-Armenians, and White-Middle Easterners
6
II.
OVERVIEW OF STUDY OBJECTIVES AND APPROACH
FORPHASE3
While Phases 1 and 2 of this study focused on investigating reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where they occur, the
overarching goal of Phase 3 was to focus on practices that take place within the CWS that either
promote or reduce disproportionality. More information about child welfare practices including
programs, services, strategies, policies, and/or tools can help address racial/ethnic disproportionality
by guiding the development of more effective services and supports for children and families of
color in the CWS.
In an effort to examine ways to address disproportionality within the CWS, the CWRT explored
three overall aims during Phase 3. The first aim included an assessment of identified key practices
for children and families ofcolor in the Santa Clara County's CWS. Specific research questions
related to this aim include:
•
•
•
•
•
•
What are the key practices in Santa Clara County's CWS that may have an impact on
children and families of color?
What empirical evidence exists for the effectiveness of these practices?
Have these practices been evaluated regarding their efficacy with varying racial/ethnic
groups?
How do these practices interact with child, family and cultural characteristics of differing
racial/ethnic groups involved in Santa Clara County's CWS?
How do practices in South County vs. the Main Office differ? (i.e., are there differences in
worker socialization, training, values/attitudes, stated and unstated expectations norms that
contribute to differences in practice and client outcomes?)
What practice characteristics are necessary in order to promote more positive outcomes for
children and families from different racial/ethnic groups involved in Santa Clara county's
CWS?
The methodology used to explore these research questions included 1) a review of the current
research literature on evidence-based practice for children in the CWS and the application of these
practices to children and families of color, 2) semi-structured in-depth interviews with DFCS
supervisors and managers, 3) semi-structured focus groups with DFCS social workers and parents
in the CWS, 4) a qualitative and quantitative comparison of child welfare practices and outcomes
between the DFCS South County and Main County offices, and 5) a review of cases from Phase 2
of the research with successful outcomes and an analysis of factors that contributed to positive
outcomes.
The second aim ofPhase 3 included an exploration ofthe influence of the court system on
child welfare practice, including court ordered changes to the social worker service
recommendations at the jurisdictional/dispositional hearing. We were interested in whether the
court system imposed any changes to the social workers' recommendations offered at the
jurisdictional/dispositional hearing in a child's case, and if these changes differed by the child's
7
ethnicity, as well as the nature of the relationship between the child welfare and court systems.
Specific research questions related to this aim include:
•
•
•
•
To what degree does the court system impose changes to the social workers' case plan
recommendations?
What types of changes to the social workers recommendations does the court system
order?
Do court ordered changes to the social workers' recommended case plan differ by child's
ethnicity?
How does the relationship between the child welfare and court system positively or
negatively impact child welfare practices and outcomes for children and families of color?
The methodology used to explore these research questions included case record reviews of the 403
closed child welfare cases used in Phase 2 of the research. Cases were reviewed again and
information on court ordered changes to the social worker recommendations at the
jurisdictional/dispositional hearing was gathered. Additionally, agency-wide focus groups provided
information on the how the relationship between the child welfare and the court system has an
impact on practices and outcomes for children and families of color.
The third aim ofthe research involved a statewide comparative analysis of identified key
practices that may affect the disproportionate representation of children ofcolor in the CWS.
Specific research questions associated with this aim include:
•
•
•
What are key practices in California counties that may have an impact on children and
families of color?
What practices are considered most promising by individual counties for children and
families of color?
Comparing counties with large racial/ethnic representation, what factors predict
over/underrepresentation of various racial/ethnic groups?
The methodology used to explore these research questions included 1) a statewide survey of county
child welfare directors, managers or supervisors in California counties and 2) quantitative analyses
using a statewide database consisting of county-level characteristics.
Phase 3 focused on exploring ways to address the disproportionality of children and families
of color within the CWS by examining practices that may either contribute to or reduce ethnic/racial
disparities. In order to accomplish the aims outlined above, we combined qualitative and
quantitative research methods and examined both practices within the system, as well as contextual
factors that can potentially have an impact on what transpires within the system. The report begins
with a review of the literature on effective practices for children and families in the CWS, followed
by a close up view of practices within Santa Clara County and then broadens to include a wider
perspective on statewide child welfare practices aimed at children and families of color. The
following section will describe the background and literature related to effective practices for
children and families of color in the CWS.
8
III. BACKGROUND AND LITERATURE REVIEW
Different racial/ethnic groups experience unique pathways through the child welfare system
(CWS) (Hines et al., 2002). At every point in the system, various child welfare practices and
services impact children and families. Phase 2 results indicated that children and families of color in
Santa Clara County's CWS tend to be recommended services that are traditional and formal in
nature and that do not appear to meet the wide range of needs experienced by these highly diverse
racial/ethnic family groups. Indeed, children and families of color in the CWS represent a high risk
group who are often impacted by myriad psychosocial challenges (see Phase 2 report for a full
description), yet services recommended for these families tend to be limited to a one-size-fits all
approach.
Moreover, there exists relatively little empirical evidence that traditional child welfare
services are effective in maintaining or reunifying children and improving family functioning. Even
less research has been conducted on the impact of child welfare practices for children and families
of color. Yet because children and families of color are disproportionately represented in the CWS,
the effectiveness of child welfare practices for these groups is of particular interest. Research
suggests that children and families of color tend to have longer stays in the CWS, they receive
fewer and less comprehensive services, and have poorer case outcomes than White children and
families (Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996). This bleak
outlook for children and families of color in the CWS creates an urgent need to understand child
welfare practices and services that result in enhanced outcomes for these families. It is also crucial
to identify the types of individual and contextual factors that hinder or support the effectiveness of
these practices.
This section will review the current research on evidence-based child welfare practices and
their application to children and families of color. Practices will be discussed in relation to key
choice points in the CWS including, beginning, continuing care, and ending choice points. Please
see Figure 1 for a flow chart of possible pathways and key choice points in Santa Clara County's
CWS. Additionally, contextual factors including cultural competency of workers and services; the
unique needs of rural child welfare populations; and the role ofthe court system will be discussed.
Across all stages of the CWS from beginning to end, highlights of practices that appear to be most
effective will be presented and the ways in which these best practices may impact children and
families of color will be discussed.
9
Figure 1: Possible Pathways and Key Choice Points In the Child Welfare System
Referral
(Report of Abuse or Neglect)
Emergency Response
(Screening)
Case Closed
Case Opened:
Further
Investigation
Jurisdictional Hearing:
Decision Regarding Substantiation
of Abuse/Neglect
Case Dismissed:
Unsubstantiated
Allegations
Family
Maintenance
Case Opened:
Voluntary Family
Maintenance
Case Closed:
Short-Term Services
Provided and Family
Stabilized Within 6
Months
Dispositional Hearing:
Substantiated Allegations,
Further Action and Services
Family Reunification:
Placement in FFH, FFA, Group
Home, Kincare, or Other Facility
Bypass
Hearings: Including 6-Month Review
Case Closed:
Family Stabilized
Through Family
Maintenance
Case Closed:
Family Stabilized
Through Family
Reunification
Case Closed:
Permanent Placement into
Adoption, Guardianship,
or Long Term Care
10
Impact of Child Welfare Practices on Children and Families of Color
Beginning Choice Points
When a child is reported to the CWS, there are various pathways that she or he may take
through the system (see figure 1). At the screening stage, the case may be closed altogether with no
further action; it may be opened for six months of voluntary family maintenance services; or the
case may be opened for further investigation. If the case is opened for further investigation, a
jurisdictional/dispositional hearing typically occurs in which there is a judicial finding on whether
the allegation of maltreatment is substantiated. If allegations are found to be unsubstantiated, the
case may be dismissed, however, if substantiation occurs, children may be placed in out-of-home
care, with family reunification services being ordered, or children may be maintained in the home
with orders for family maintenance services. In some cases, families may be bypassed for services
and have their parental rights terminated at this early stage in the system.
Following the pathway that children may take through the CWS, the screening process is the
first entry point into the CWS. On a national level, the U.S. Department of Health and Human
services ([U.S. DHHS], 2001) reports that approximately 2.7 million referrals were received by the
CWS in 2001; 32.7 percent of these referrals were screened out and closed with no further action.
Other national studies have found similar rates of screening out referrals, ranging from 36 percent to
42 percent (Tumlin & Geen, 2000; Wells, Fluke, & Brown, 1995).
Research suggests that certain case characteristics are associated with the decision to open a
case for further investigation or to refer to voluntary family maintenance services. For instance,
reports involving more than one child, children under age two, severe injury, a minor or other type
of injury or physical condition, allegation of sexual abuse, previous reports to the CWS, (Wells et
al., 1995); as well as child welfare cases involving AFDC receipt; alleged parental drug use; a
female primary victim; reports made by a family member; and reports with direct evidence (Karski,
1999) all have a higher likelihood of investigation than cases not involving these characteristics.
Additionally, cases involving children of color are opened for investigation at a higher rate
than cases involving White children. In Wells et al.'s (1995) analysis of 12 child welfare sites,
African American children· were significantly more likely to have reports investigated than children
of other ethnicities, however when the data was dis-aggregated and examined by sites, it was found
that almost the entire sample of African American children reported to the CWS came from two
sites that exhibited particularly high investigation rates. Yet, in an analysis of data from five states,
Fluke, Yuan, Hedderson, and Curtis (2003), found that African American children were
significantly overrepresented at the stage of investigation in each of the five states, as well as in all
counties within the five states, while White children tended to be underrepresented at the stage of
investigation. Rates of disproportionality at the investigation stage for other racial/ethnic groups,
including Asian American/Pacific Islander, Hispanic and Native American children showed much
wider variation. Depending on the state, these groups were both underrespresented and
overrepresented at the investigation stage.
Factors associated with the provision of voluntary services are somewhat less well
understood. In a statewide survey in Missouri, Drake (1996) analyzed factors associated with the
provision of preventive voluntary services to 3,442 children and families in the CWS, in
11
comparison to cases that were substantiated by the court. Cases were divided into three types of
maltreatment: sexual abuse, physical abuse and neglect. Across all three types of maltreatment,
cases from rural areas were significantly more likely to receive preventive services than cases from
urban or metro areas. Additionally, predictors of preventive service provision in sexual abuse cases
and neglect cases included those that were referred by a non-anonymous source. Physical abuse
cases were more likely to receive preventive services when a mandated reporter reported them.
Neglect cases were more likely to receive preventive services when they involved White children
and families, parents who were unemployed and children and families from non-poor
neighborhoods (Drake, 1996).
For children who are not offered voluntary family maintenance services, investigations
result in a decision regarding whether to substantiate the maltreatment and take further action or to
not substantiate and dismiss the case. Nationally, 27.5 percent of investigated referrals are
substantiated (U.S. DHSS, 2001). State data suggest similar substantiation rates. One study
conducted in New York state revealed a 30 percent substantiation rate across types of maltreatment,
with a 48 percent substantiation rate for physical abuse, 39 percent for sexual abuse and 28 percent
for neglect (Eckenrode, Powers, Doris, Munsch, & Bolger, 1988). Interestingly, there appears to be
little relationship between screening rates and substantiation rates. States with higher screening in
rates do not appear to have higher rates of substantiation of child maltreatment (Tumlin & Geen,
2000).
In California, the overall rate of substantiation of all referrals is 21.2 percent (Needell et al.,
2003). Statewide rates of substantiation between various racial/ethnic groups appear similar; Black
children have a 22 percent substantiation rate, White children 21.4 percent, Hispanic children 23 .4
percent, Asian children 24.5 percent, and Native American children 25.3 percent. Across all ethnic
groups, younger children have higher rates of substantiation than do older children (Needell et al.,
2003).
Santa Clara County's overall rate of substantiation in 2002 was 17.1 percent--a rate
somewhat lower than the state average. Santa Clara County also appears to have greater disparities
between racial/ethnic groups in substantiation rates. For instance, Black children in Santa Clara
County have an overall substantiation rate of 21 percent, White children 15 .4 percent, Hispanic
children 18 percent, Asian children 15 .1 percent, and Native American children 15 .I percent. Rates
of substantiation between racial/ethnic groups in Santa Clara County also depend on the age of the
child. As is the case within the state, younger children have a higher likelihood of substantiation,
however, in Santa Clara County it appears that Black children between O and 2 years of age have a
relatively high rate of substantiation (76.8%), whereas, White children 0-2 have a substantiation rate
of 62.6 percent, Hispanic children 52.6 percent; Asian children 46 percent, and Native American
children 66.6 percent (Needell et al., 2003).
Certain factors appear to be associated with the decision to substantiate an allegation. For
instance, reports from professional reporters tend to be substantiated at a higher rate than other
reports (Eckenrode et al, 1988; English, Marshall, Coghlan, Brummel, & Orme, 2002), as well as
cases referred by law enforcement (English et al., 2002). Poverty and early child bearing have been
identified as strong predictors of substantiation (Lee & George 1999). A history of prior referrals or
involvement in the CWS are also associated with higher rates of substantiation (Eckenrode et al.,
1988; English et al., 2002), and one statewide survey in Washington state found that referrals from
rural areas were more likely to be substantiated, although this effect was small (English et al.,
12
2002). Additionally, cooperation with child welfare agencies has been found to be associated with a
lower likelihood of substantiation (English et al., 2002). Factors associated with substantiation also
depend to some degree on the type of maltreatment alleged. For instance, sexual abuse reports that
involve older, female children in large households tend to have a higher substantiation rate than
cases not involving these characteristics, and the same is true for neglect reports that involve
younger children with previous child welfare referrals (Eckenrode et al., 1988).
·
Cases involving children of color appear to have an elevated likelihood of substantiation.
One large study conducted in New York state found that for allegations of physical abuse or
neglect, African American and Hispanic cases were substantiated at a higher rate than White cases
(Eckenrode, et al., 1988). In a more recent study conducted in Minnesota all types of maltreatment
were substantiated at a higher rate for non-White groups than for White groups; this effect remained
even after factors such as type of reporter, type of maltreatment, geographic location and
characteristics of the child and perpetrator were statistically controlled (Ards, Myers, Malkis,
Sugrue, & Zhou, 2003). Similarly, in Drake's (1996) statewide analysis of child welfare cases in
Missouri, non-White racial status was a significant predictor of substantiation even when other case
characteristics were statistically controlled. In contrast, Fluke et al.'s (2003) analysis of
disproportionate representation of race/ethnicity in investigation and victimization rates in five
states found little disproportionaility in rates of substantiation of maltreatment.
In addition to substantiation of an allegation, decisions at the beginning stages of the CWS
are also focused on whether to place a child in out-of-home care. Studies have consistently found
certain factors to be associated with the decision to remove a child from home. Children of color, in
particular African American children have an elevated likelihood of out-of-home placement
(Needell, Brookhart, & Lee, 2003; U.S. DHHS, 2001). Other factors associated with the decision to
remove a child include inadequate or low income (Berry, 1991; Lindsey, 1991 ); cases involving
prior maltreatment reports (Dalgleish & Drew, 1989; U.S. DHHS, 2001); cases involving families
with many children (Berry, 1991); lack of cooperation from the family (Dalgleish & Drew, 1989;
Davidson-Arad, 2001; Karski, 1999); cases involving young children, mothers as perpetrators, and
reports made by educational personnel are all associated with court intervention and placement is
substitute care (U.S. DHHS, 2001).
Few studies have examined factors associated with the decision to bypass family reunification
services and terminate parental rights at the early stages of the CWS. In California, any one of 13
bypass criteria can be used to deny reunification services and pursue an alterative permanent
placement. These criteria include:
1. Parents' whereabouts are unknown for six months.
2. Parental mental disability that prevents the utilization of services.
3. A sibling has been removed from the home, returned and then removed from parental custody
again.
4. Parent has caused the death of a child.
5. Parent caused severe emotional damage.
6. Parent committed severe physical or sexual abuse to the child, sibling or half-sibling.
7. Reunification has been terminated on a sibling or half-sibling due to indicators# 3, 5 or 6.
8. Parent has been found guilty of rape that conceived the child.
9. Willful abandonment constituting serious danger to the child.
I 0. Parent failed to reunify with a sibling or half-sibling.
11. Parent convicted of a violent felony.
13
12. Parent has extensive history of substance abuse and resisted treatment for three
years prior, or failed to benefit twice.
13. Parent does not want family maintenance or family reunification services, or does not want
the child removed.
In the Bay Area Social Services Consortium's ([BASSC], 2000) report on the use of bypass
criteria within San Mateo and Santa Clara Counties, approximately half of the cases reviewed had
bypass criteria present. Twenty six percent of these bypass cases were due to parental substance ·
abuse, 18 percent were due to a sibling with parental rights terminated or with a permanent plan,
and 13 percent had a parental mental disability that prevented the utilization of services (BASSC,
2000). However, not all cases with a bypass criterion were actually bypassed. Of the cases with a
bypass criterion present, 73 percent of cases involving parental mental disability; 38 percent of
cases with a sibling who had parental rights terminated or who has a permanent plan; and 13 percent
of cases involving parental substance abuse were actually bypassed. Overall 25 percent of the entire
sample were bypassed and denied reunification services. Additionally, BASSC (2000) reported that
the use of bypass criteria is not empirically or clinically valid and that bypass criteria are not
predictive of a lowered likelihood for reunification. Moreover, the report suggested that that since
many bypass indicators are common within the child welfare population, their use in denying family
reunification services may not be warranted.
Beginning Choice Points: Im pact on Children and Families of Color
Research clearly indicates that certain/actors are associated with various decisions at
beginning choice points and these factors disproportionately impact children and families ofcolor.
In general, the most consistent factors found to increase the likelihood of investigation,
substantiation or out-of-home placement include cases involving more than one child, young
children, previous reports to the child welfare system, and families with low incomes (Eckenrode et
al., 1988; English et al., 2002; Karski, 1999; Lee & George, 1999; Wells et al., 1995). Children and
families of color are at an increased risk for all of these factors, but in addition, studies have
indicated that even when these factors are statistically controlled, children of color, and in particular
African American children are more likely to have an allegation investigated and substantiated and
are also more likely to be placed out of the home (Ards et al., 2003; Drake, 1996; Wells et al.,
1995). Additionally, there is evidence to suggest that in cases of neglect, children and families of
color may receive voluntary family maintenance services less frequently than White children and
families (Drake, 1996). Taken together these findings suggest that beginning choice point decisions
may be biased against children and families ofcolor. Indeed it appears from the research that once
a child ofcolor is reported, they are more likely than White children to be set on a pathway through
the child welfare system that involves court intervention and out ofhome removal.
Decision-Making Practices at Beginning Choice Points
The types of decision-making practices that child welfare workers use at these early stages
of the CWS are crucial in determining the trajectory of the pathway that children and families take
through the system. However, decisions on whether to investigate a referral, provide voluntary
services, substantiate an allegation, remove a child or bypass a case are often complex, ambiguous
and multifaceted. Child welfare workers must simultaneously consider numerous factors, including
child safety, the best interests of the child, parental rights and preserving family integrity. These
14
factors must be considered in a complicated context; children and families in the CWS often face
multiple and complex challenges and child welfare policies and mandates fluctuate frequently.
Given this difficult context, it is not surprising that studies have found a significant lack of
consistency in child welfare decision-making practices across various stages of the system (Rossi,
Schuerman, & Budde, 1999; Tumlin & Geen, 2000; Westat, 2001a). For instance, disparities in
screening out rates between states have been found to range from a low of 5 percent to a high of 78
percent (Tumlin & Geen, 2000), while disparities between counties have been found to range from a
low of 1 percent to a high of 71 percent (Wells, et al., 1995). There is some evidence to suggest that
child welfare workers may use more restrictive screening practices when resources or staff are
reduced or when the number of referrals is high (Curtis, Boyd, Liepold, & Petit, 1995 as cited in
Tumlin & Geen, 2000; Hutchinson, 1989, as cited in Jones, 1993). Moreover, screening decisionmaking practices tend to be influenced by the average severity level of cases within the agency as a
whole and characteristics of the surrounding community. In one survey, workers in 11 child welfare
agencies in the state of New Jersey rated vignettes of child maltreatment reports as less severe in
agencies that handled the more severe cases of child maltreatment. These agencies were also located
in areas that were more socially and economically disadvantaged-suggesting that characteristics
specific to agency sites or communities influe·nce screening and substantiation decision-making
practices (Wolock, 1982).
Although virtually all states and counties screen child maltreatment reports to some degree,
screening practices and policies vary greatly between states and counties. National surveys suggest
that a majority of states do have some form of written agency policies to guide screening and
investigation decisions (Downing, Wells & Fluke, 1990); yet few states have formal or explicit
procedures or instruments delineating reports that should be screened out (Tumlin & Geen, 2000).
Approximately 65 percent of states have a single review screening process in which the decision to
investigate is made by one worker (Tumlin & Geen, 2000). Moreover, the training of child welfare
workers who screen referrals has also been found to vary widely from state to state; with some
states having educational or work experience requirements for screening workers, while others do
not (Tumlin & Geen, 2000).
Studies investigating consistency between child welfare workers in their decision-making
practices indicate wide variability in these practices. For instance, Rossi et al. (1999) explored
decision-making practices among 130 child welfare workers and experts in three states. Results
indicated that although certain factors were associated with the decision to maintain a child in the
home, including families with no prior referrals and cases involving employed caretakers who
showed signs of engagement in services, the thresholds used to trigger an out-of-home placement
decision were not consistent. For instance, some participants decided on out-of-home care in cases
involving just one prior referral, while others recommended out-of~home placements in cases
involving two or more prior referrals. Using the same data set, Schuerman, Rossi and Budde (1999)
calculated the level of agreement between workers on decisions about services, as ranging from a
high of 64 percent to a low of 48 percent, indicating a considerable lack of consensus among
workers regarding out-of-home placement decisions.
Studies also suggest that child welfare workers consider numerous factors in their decisionmaking practices and that these factors are generally not considered in isolation from one another
(English et al., 2002). For instance, a recent study by the U.S. Department of Health and Human
15
Services (2003a) identified four overall factors that influence decisions regarding whether to
substantiate an allegation, including:
1) Case factors, such as child characteristics, severity of the maltreatment, chronicity of
maltreatment, caretaker characteristics, caretaker/child relationship, social and economic
factors, and perpetrator access to the child.
·
2) Decision-maker factors, such as caseworkers' individual tendencies, experience level,
relationships with co-workers, self-assessments of skills and degree of importance
placed on state policy.
3) Organizational factors, such as workload stress, and degree of supervisor support.
4) External factors, such as state policies.
All of these factors were found to impact the decision-making process of child welfare workers in
complex ways that may differ between individual workers, counties and states (U.S. DHHS, 2003).
Yet there is some evidence to suggest that decision-making practices within a particular
agency may be more consistent than between agencies. Alter ( 1985) has suggested that child
welfare workers tend to use heuristics or assumptions to guide decision-making and that workers
within a particular agency often develop a mutual, although unspoken understanding of the kinds of
cases that should be investigated. Indeed there does appear to be greater consistency in initial
decision-making practices within agencies than between agencies. For instance, Wells et al. 's
( 1995) study of 12 child welfare sites, found wide variability between the sites in their investigation
decision-making practices, yet when all other factors were held constant, the site of the report was
more significant than any other factor in predicting an investigation. These results suggest that the
type of decision-making practices common to a particular child welfare agency play an important
role in decision-making practices related at the beginning stages of the CWS.
In addition to the importance of agency norms in decision-making practices, the labels used
to describe families may also trigger certain decisions. For instance, Martin, Peters, and Glisson
(1998) surveyed 100 child welfare workers in five states and found that workers did not make
service and placement recommendations for children based on a uniform psychosocial assessment.
Instead, decisions were influenced more by the labels attached to children upon entry into care and
judicial decisions. Children whose files contained language such as "substance abuser, child of
substance abuser, sex offender, victim of sexual abuse and victim of physical abuse" (Martin et al.,
1998, p. 3), were more likely to be recommended mental health services and more restrictive
placements-regardless of their actual need as indicated in the psychosocial assessment. Similarly,
children who were adjudicated delinquent or who were placed in out-of-home care due to behavior
problems rather than maltreatment were also recommended mental health services and more
restrictive placements, regardless of their actual need.
The influence of agency norms, the use of implicit assumptions and the complexity inherent
in child welfare decisions has prompted many practioners and researchers to call for more objective
or collaborative decision-making practices. Structured risk assessment tools, multidisciplinary
teams, and family group conferencing are considered by many to offer a more objective, balanced
and inclusive method of decision-making at beginning choice points.
16
Decision-Making Practices: Risk Assessment Tools
Structured risk assessment tools are designed to assist child welfare workers in making
accurate assessments at various stages in the CWS. Risk assessment tools generally evaluate factors
that are considered to be associated with various outcomes including such items as prior referrals
and prior substantiated cases; the presence of substance abuse, mental health or economic problems,
family constellation, the type, severity and chronicity of the maltreatment and circumstances
surrounding the current incident (Camasso & Jagannathan, 2000) Risk assessment tools fall into
two general categories: 1) consensus based systems in which risk factors in the instrument are based
on the consensus ofjudgments made by experts in the field, and 2) actuarial systems in which risk
factors are identified based on empirical evidence of factors associated with future maltreatment
(Baird & Wagner 2000). There is some evidence to suggest that actuarial systems more accurately
capture levels of risk in families referred to the CWS (Baird & Wagner, 2000).
The use of risk assessment tools varies greatly from state to state and county to county, and
these tools can be used at various decision points in the CWS from intake to case closing (Wald &
Woolverton, 1990). Although risk assessment tools are widely used in CWS agencies, their ability
to accurately assess levels of risk is an area of considerable debate in the child welfare field (Wald
& Woolverton, 1990). Several researchers have noted that risk assessment instruments have only
minimal reliability and validity in predicting case outcomes (Camasso & Jagannathan 2000;
DePanfilis & Scannapieco, 1994; Doueck, English, DePanfilis, & Moote, 1993; Wald &
Wooverton, 1990).
In 1998, certain California counties, including Santa Clara County began using a
comprehensive risk assessment system, known as Structured Decision-Making (SOM). SDM has
several components including response priority, safety assessment, risk assessment, family needs
and strengths assessment, case planning and management, case reassessment, workload based
resource allocation, and the use of management information systems (Children's Research Center,
2003). These components are designed to assist workers in assessing the severity of allegations,
making structured decisions, managing workloads, minimizing the trauma associated with
maltreatment and preventing the recurrence of maltreatment (California Department of Social
Services, 2003). The risk assessment tool used in this system is an actuarial instrument that is
intended to estimate the likelihood that maltreatment will recur. In addition to the risk assessment,
SOM also calls for a family needs assessment in which family's strengths and resources are to be
considered in conjunction with the risk assessment. Although peer reviewed evaluations of the
SOM project are not available, the Children's Research Center (as cited by the California
Department of Social Services, 2003) reports that a 12-month evaluation on the SOM project was
conducted in Michigan. Results indicated that cases receiving SOM had 27 percent fewer new
referrals, 54 percent fewer new substantiated allegations, 40 percent fewer children removed to
foster care, and 42 percent fewer child injuries than children not receiving SOM. Although these
results are promising, more evaluations are needed to discern the effectiveness of SOM, specifically
for children and families of color.
Decision-Making Practices: Multidisciplinary Teams/Family Group Conferencing
The use of multidisciplinary teams to assist in child welfare decision-making has gained
popularity in recent years. Multidisciplinary teams are designed to better coordinate efforts between
child welfare, law enforcement, legal professionals and medical professionals in making decisions
17
about child welfare cases. These efforts are intended to improve the efficiency and effectiveness of
decision-making activities and also in some cases-to reduce the potential for traumatizing children
through repeated investigatory interviews (Sheppard & Zangrillo, 1996). Approximately two thirds
of states have been found to use multidisciplinary teams in making child welfare decisions (Kolbo
& Strong, 1997).
Research on the effectiveness of multidisciplinary teams in improving decision-making in
child welfare is lacking. In one of the few studies to address this issue, Hochstadt and Harwicke
(1985) evaluated the use of multidisciplinary teams for 180 children referred to the CWS for
suspected child maltreatment. The multidisciplinary team included child welfare workers, medical
professionals and psychological personnel who completed a multidisciplinary evaluation and
participated in a case conference in which specific service recommendations and treatment plans
were made. Although no comparison group was used in the study, findings suggested that the use of
multidisciplinary teams was associated with a high percentage of children and families obtaining
recommended services on their treatment plans. One hundred percent of children with a
recommendation to return home did so, and 92 percent with a recommendation to foster care
followed that recommendation. Other service recommendations that were received included visiting
nurse (76%), additional medical assessment (64%), infant program (64%) and individual
psychotherapy (59%). The authors suggest that the use of multidisciplinary teams can positively
influence the receipt of recommended services.
In an effort to discern how multidisciplinary teams in child welfare work together,
Lewandowski and Glen Maye (2002), conducted a survey of 165 child welfare and community
professionals who were a part of a multidisciplinary child welfare team in an urban Midwest
county. The sample included social workers, counselors, community members, education
professionals, law enforcement, legal professionals and medical professionals. Findings revealed
that 72 percent of respondents had team meetings at regularly scheduled times with support staff
assigned to the team, and 53 percent had a designated team chairperson. Forty eight percent of the
sample indicated that decision-making was most likely to occur around case planning, 32 percent
indicated that decision-making was least likely to occur at intake, and 28 percent reported that
decision-making occurs at all stages. Inadequate communication was identified as the greatest
barrier to effective team decision-making, followed by inadequate resources and inadequate
professional respect. Lastly, although most respondents (63%) agreed that family members should
be active and equal participants in the team decision-making process, only 35.5 percent reported
that family members attended team meetings.
Recently, more concerted attempts to involve family members and extended support systems
in decision-making have occurred. Many child welfare systems are moving toward the use of family
group conferences to improve decision-making. The use of family group conferences began in New
Zealand in 1989, when legislation mandating their use was implemented (Sieppert, Hudson, &
Unrau, 2000). Family group conferences, also referred to as family group decision-making, are
designed to bring together family members, relatives, and other support systems in order to make
decisions about a case. The family group conference is intended to identify the family's strengths
and resources; to develop a plan to ensure child safety and improve family functioning; and to foster
cooperation, collaboration and communication between families and professionals (American
Humane Association, 2003: Pennell, & Buford, 2000). These methods are based on the principle
that families themselves possess the most information about what decisions should be made; the
approach is intended to be family centered, strengths based~ and takes into consideration issues of
18
culture and community (American Humane Association, 2003).
In addition to family group conferencing, recent efforts have also been aimed at ''team
decision-making," a closely related concept that appears to widen the scope of participants in the
decision-making process. Team decision-making is one component of the Casey Foundation's
Family to Family Initiative, a broad child welfare effort aimed at improving outcomes for children
in the CWS (see page 31 for a discussion of the Family to Family Initiative). The team decisionmaking component of the initiative involves convening a broad range of participants to share
information about the family and develop a plan for· child safety, family maintenance or
reunification. The team decision-making meeting can involve any participants who can contribute to
the decision-making process, including family members, neighbors, foster parents, service
providers, child welfare workers and supervisors, other resource staff or other community
representations (Casey Foundation, 2003a).
Research on the effectiveness of family group conferencing and team decision-making on
outcomes for children and families is lacking, although some studies have investigated the process
of implementing family group conferences. For instance Sieppert, et al. (2000) investigated the
process of implementing a family group conferencing pilot project in Canada. The sample included
23 child welfare families with court involvement and one family group conference was conducted
for each family. Mothers were the most frequent family members to attend the conferences (91 %),
while 52 percent involved a father, 70 percent involved grandmothers, 57 percent involved aunts,
and ~5 percent involved grandfathers. Other members of the conferences included child welfare
workers (83%), and foster parents (30%). Findings revealed that family group conferences tended to
be resource-intensive. Workers performed a wide range of activities in preparing for and
implementing the family group conferences. An average of 17 telephone contacts, 2.4 in-person
contacts, and 3 letter contacts were made prior to each family conference. Workers also spent time
traveling to meet with prospective conference attendees. On average, workers spent 7 hours
preparing for each conference. Family conferences lasted an average of2.75 hours. At the end of
the conference, families were asked to rate their experience of the group. Seventy percent reported
being highly satisfied with conference preparation and members in attendance; 85 percent felt a
high degree of freedom to speak out in the group; 76 percent reported feeling involved in the
decision-making process and 65 percent felt highly satisfied with the decisions that resulted from
the conference. Unfortunately the ethnicity the families involved in the family group conferences
was not provided.
Decision-Making Practices at Beginning Choice Points:
Impact on Children and Families of Color
In the beginning stages of the child welfare system, the decision-making practices ofchild
welfare workers can have a substantial impact on children and families ofcolor. However in
general there appears to be considerable inconsistency in the types of decision-making practices
used in the child welfare field (Rossi et al., 1999; Schuerman et al., 1999; Tumlin & Geen, 2000;
Westat, 2001 a). Although many child welfare systems are moving toward more objective or
multidisciplinary approaches to decision-making, a majority of states still use a single review
process in which decisions to investigate are made by one worker (Tumlin & Geen, 2000); and few
states have formal policies and procedures for decision-making at early stages of the CWS (Tumlin
& Geen, 2000).
19
However, making child welfare decisions in isolation can have deleterious effects on
children and families of color. Research suggests that child welfare workers often use implicit
assumptions to guide decision-making. Ways in which each worker thinks about diverse families,
and the overall spoken and unspoken expectations and norms of a particular child welfare agency
can have an enormous impact on decision-making practices (Cohen, 2003). As Cohen (2003) notes,
The cultural and racial background of families influences the specific factors that workers
consider in assessing the severity of risk and level of intervention. Decisions are more likely
to be made on the basis of deficits in available resources, accepted agency practice, personal
values and biases, and notions of an ideal family, than by application of consistent case rules
(Chohen, 2003, p. 145).
It is speculated that in some cases, race/ethnicity may impact decision-making by resulting in a
higher level of intervention, while in other cases race/ethnicity may result in less intervention effort.
Much ofthis variation appears to depend on the worker and the context in which the worker makes
decisions (Cohen, 2003).
The use of risk assessment tools and multidisciplinary teams/family group conferencing may
have the potential to improve decision-making practices affecting children and families of color.
Although research on the effectiveness of risk assessment tools in assessing risk for future
maltreatment has not been consistently demonstrated in the research literature, the move toward
more objective ways of assessing risk at beginning stages of the child welfare system may be
promising for children of color. For instance, in one study conducted by the California Department
of Social Services (1999), 2,500 substantiated cases from counties that used the SOM system, were
evaluated to assess whether the risk assessment instrument used in SOM had any differential impact
on African American, Hispanic, or White children and families. Findings indicated that Whites had
slightly higher scores on the risk assessment instrument than did African Americans or Hispanics
and the authors concluded that the risk assessment instrument did not disparately affect African
Americans or Hispanics. While this finding is encouraging, other studies have suggested that risk
assessment instruments in general are not always able to accurately predict case outcomes Camasso
& Jagannathan 2000; DePanfilis & Scannapieco, 1994; Doueck et al., 1993; Wald & Wooverton,
1990). More research is needed on the impact of risk assessment tools in general, and their impact
on children and families of color in particular.
Additionally, more research on the effectiveness ofmultidisciplinary teams and family group
conferencing is needed The family centered, strength based and inclusive philosophy offamily
group conferences may be a particularly useful strategy for children and families ofcolor, many of
who!11 are experiencing numerous risk factors in a highly stressed environment. The active
engagement offamilies and support systems in decision-making at early stages of the child welfare
system may potentially result in enhanced outcomes for children and families ofcolor.
Yet there is some evidence to suggest that these approaches can be labor, time and resource
intensive (Sieppert et al., 2000). As such, the effective implementation of these practices may
require the use of increased support staff to assist in scheduling and arranging meetings, as well as
identifying persons to be included in the meeting. Additionally, because beginning choice point
decision-making practices are often made quickly and in the context ofcrisis situations, the effective
use ofmultidisciplinary teams and family group conferences requires coordinated efforts and
appropriate support staff. Clearly, more research is needed to determine how group decision-
20
making can be effectively integrated into child welfare agencies and how these practices impact
outcomes for children and families ofcolor.
Voluntary Family Maintenance Practices/Family Preservation Services
In cases where maltreatment is not substantiated, families may be referred to voluntary
family maintenance services. Because the overall goal of voluntary family maintenance services is
to prevent an out-of-home placement and further involvement with the CWS, these services are
considered beginning choice point practices. Much of the research literature on services and
practices aimed at preventing out-of-home placement refer to these services as "family preservation
services," (FPS) or "intensive family preservation services" (IFPS). Many researchers and
practioners have noted that services falling under the rubric of "family preservation services" can be
quite diverse, although they do tend to share the common goal of preventing child maltreatment and
out-of-home placements (United States General Accounting Office, 1997). The model of FPS that is
most often described in the research literature includes short-term, intensive and home-based
services that are aimed at family skill-building, while also including concrete services, such as
public assistance, child care, employment training or transportation assistance (Littell, 1997).
In general, there are numerous shortcomings in the research on the effectiveness of FPS;
many studies use small sample sizes, no comparison groups and often fail to adequately describe the
nature of the intervention and the characteristics of the clients. As a result, evaluations of FPS often
yield conflicting results. While some studies report that FPS do not result in lower placement rates
or improved family functioning (Chaffin, Bonner, & Hill, 2001; Littell & Schuerman, 1995; Littell,
1997; Westat, Inc. 2001b); other studies suggest that FPS do improve outcomes for children and
families (Pecora, Fraser, Bennett, & Haapala, 1991; Scannapieco, 1994; Spaid, Fraser & Lewis,
1991; Wells & Whittington, 1993).
There is also evidence to suggest that FPS outcomes vary depending on case characteristics.
A number of different studies have noted that FPS clients who experience an out-of-home
placement typically have more risk factors and less resources than families who avoid placement.
The particular risk factors affecting FPS clients who experience out-of-home placements have been
found to vary widely. For instance, FPS cases involving substance abuse, mental illness, criminal
behavior, disability, low income, or previous CWS involvement have all been linked to an increased
likelihood of an out-of-home placement (Fraser, et al., 1991; Thieman & Dail, 1997; Thieman et al.,
1990, as cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman,
1995).
Voluntary Family Maintenance Practices/Family Preservation Services:
Impact on Children and Families of Color
Research on voluntary family maintenance practices and FPS has not systemically
evaluated the efficacy of these programs for diverse children and families ofcolor. Many studies
fail to perform analyses examining the possible differential outcomes of these services for children
and families of color. In general, research on FPS has not consistently demonstrated that these
services are effective in preventing out-of-home placement or improving family functioning
(Chaffin et al., 2001; Littell & Schuerman, 1995; Littell, 1997; Westat, Inc. 2001b), and some
researchers have concluded that there is no reliable research that supports the effectiveness of these
services (Littell & Schuerman, 1995). However, some studies do suggest that FPS may be most
21
effective with lower-risk families (Fraser et al., 1991; Thieman & Dail, 1997; Thieman et al., 1990,
as cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman, 1995).
Yet because children and families of color referred to the child welfare system often possess
numerous risk factors (Hines et al., 2002), the use of traditional FPS for these diverse clients may
not be effective. Research does suggest that FPS that are most effective with high-risk clients are
those that are more intensive, and those that involve high client participation and the delivery of
concrete services in the home (Berry, 1992; Cash & Berry, 2002; Nelson & Landsman, 1992, as
cited in Littell & Schuerman, 1995; Yuan et al., 1990, as cited in Littell & Schuerman, 1995).
Home-Based Prevention Efforts
Research suggests that the use of home-based services may be effective in preventing
maltreatment and further involvement in the CWS (Olds, Eckenrode, Henderson, Kitzman, Powers,
Cole et al., 1997). Most research on the use of in-home services has focused on their efficacy for
pregnant women or mothers of young children. While the nature ofin-home services may vary, in
general these services are aimed at improving health behaviors of pregnant and parenting women,
facilitating parenting skills, increasing support systems and encouraging healthy parent child
attachments (Olds et al., 1997). Many practioners and researchers have suggested that the perinatal
period is an important window of opportunity and that services offered during this time have
increased potential to offset a number of negative outcomes, including child maltreatment
(Guterman, 1999).
For instance, in one study involving pregnant women in rural New York state, 324 women
participated in a home nurse visitation program that sought to improve health related behaviors
during pregnancy, parenting behaviors and the personal life course development of the mother
(Olds et al., 1997). Specific services included linking families with health and human services;
involving family members and friends in the care of the child; educational services to increase
parenting abilities; and assisting mothers in completing education, obtaining employment and
family planning. Participants received services for two years; outcomes assessed at a 15-year
follow-up indicated that compared to pregnant women who received no home visitation services,
mothers in the program had significantly fewer verified reports of child maltreatment in which they
were the identified perpetrator. The program appeared to have even more of a protective effect on
unmarried women and women experiencing income problems. These groups were the least likely to
have a verified child maltreatment report. Women in the program were also less likely to have
substance abuse problems, had fewer contacts with the criminal justice system, had fewer
subsequent pregnancies and were less likely to use public assistance (Olds et al., 1997).
Interestingly, a separate analysis of the same data set indicated that the protective effects of
home visitation in reducing the likelihood of child maltreatment did not hold true for mothers
experiencing domestic violence. Mothers in the home visitation program who reported more than 28
incidents of domestic violence during the 15-year follow-up period did not experience a reduced
likelihood of verified child maltreatment (Eckenrode, Ganzel, Henderson, Sinith, Olds, Powers et
al., 2000). These results suggest that different risk factors may have varying degrees of influence on
outcomes for home visitation program participants.
In addition to child maltreatment prevention, in-home services have also been identified as a
useful public health strategy to improve health related behaviors and outcomes for mothers and
young children. Kitzman, Olds, Henderson, Hanks, Cole, Tatelbaum et al. (1997) evaluated a home-
22
visiting program that involved in-home services delivered by nurses to 1,139 low-income, primarily
African American women during pregnancy and the first two years of the child's life. Results
indicated that, compared to women who had no home-visiting intervention, women who received
in:--home services experienced fewer injuries and fewer hospitalizations for their children, and also
experienced fewer pregnancies and a lower rate of pregnancy induced hypertension.
Not surprisingly, .research also suggests that the nature of the interventions provided through
in-home programs plays a role in their effectiveness. For instance, some studies on in-home services
report no increased benefit from their use (Barth, 1991; Fraser, Armstrong, Morris, & Dadds, 2000).
However it appears from the literature that in-home programs are most effective when health or
social services professionals, as opposed to paraprofessionals, deliver the services. In-home services
that are long term, generally lasting longer than 6 months are also considered to be more effective
(Barth, 1991; Olds et al., 1997).
Home-Based Prevention Efforts:
Impact on Children and Families of Color
Although differential outcomes for various racial/ethnic groups receiving home-based
prevention services are not often cited in the research, studies do generally support the effectiveness
of these services for vulnerable parenting women. Participation in home-based services has been
linked to a reduced likelihood ofchild welfare system involvement, substance abuse problems,
contact with the criminal justice system, public assistance usage, subsequent pregnancies (Olds et
al., 1997), as well as fewer child injuries and hospitalizations, and better health during pregnancy
(Kitzman et al., 1997).
Additionally, there is evidence to suggest that home-based services are effective with highrisk populations and families ofcolor. For instance Olds et al. (1997) found that home-based
interventions were associated with even better outcomes among unmarried and low SES mothers,
and Fraser et al. (2000) reported that mothers with poor attachment to their child and who had a low
sense of competence were more likely to remain in a home-based program. Kitzman et al. ( 1997)
also noted that in-home services were associated with better outcomes for a sample of primarily
African American mothers. Furthermore, McGuigan, Katzev, and Pratt (2003) reported that older
Hispanic mothers had the greatest likelihood of remaining in a home-based program, suggesting
that in-home services may be particularly useful with families of color.
However research also suggests that the effectiveness of home-based interventions depends
to some degree on the types of risk factors present and the components of the interventions. For
instance, the presence of domestic violence has been found to eliminate the protective effects of
home-based interventions (Ekenrode et al., 2000); and unmarried mothers with less than a high
school education; mothers who received no prenatal care; and mothers who have had multiple births
may be less likely to benefit from home-based programs than mothers without these challenges
(Murphey & Braner, 2000). Furthermore, mothers living in communities with a high rate of
violence have been found to have an increased likelihood of dropping out of home-based programs
(McGuigan et al., 2003); and teenage mothers and those with insecure housing also appear to have
an increased risk of dropping out of home-based programs (Fraser et al., 2000). It also appears
from the literature that home-based programs may be most effective when they last more than 6
months; when they are delivered by health or social service professionals (Barth, 1991; Olds et al.,
23
1997); and when there is a high level ofsupervision for the workers delivering the services
(McGuigan et al., 2003).
In general, home-based prevention efforts appear to be a promising best practice for
vulnerable child welfare populations. More research is needed to discern what the necessary
components of intervention are and to what degree these interventions differentially impact diverse
groups. Additionally, it is important to note that home-based interventions may be time, labor and
resource intensive and as such, child welfare agencies may have difficulty implementing these
services. However, clients of these services clearly appear to benefit and home-based programs do
tend to result in enhanced outcomes for parents and children.
Continuing Care and Ending Choice Points
If child maltreatment is substantiated, children typically enter the CWS under some form of
court supervision. Generally families who are not bypassed for services are assigned to either family
maintenance or family reunification services (see figure 1). Under the Adoption and Safe Families
Act of 1997, families generally receive up to 18 months of family reunification services, at which
point they are either reunified or parental rights are terminated and the child enters post permanency
planning services. As such, continuing care and ending choice point practices generally include
family reunification services, including specific treatments for parents; as well as services for
children in out of home care,· services for alternate caregivers, and permanency planning services
for children who are not reunified with their birth parents.
Factors Associated with Family Reunification
Studies have clearly identified certain factors that are predictive of reunification or nonreunification. Research consistently demonstrates that White children are reunified at a higher rate
than are Black children (Courtney, 1994; Jones, 1998; McMurty & Lie, 1992; Wells & Guo, 1999;
Westat, Inc., 2001c); and young children tend to be reunified more often than older children
(Courtney, 1994; Goerge, 1990; Westat, Inc., 2001c). Harris and Courtney (2003) report that single
parent African American families are particularly unlikely to reunify, while two-parent Hispanic
families have a significantly higher likelihood of reunification than two-parent Caucasian or African
American families. Socioeconomic status also plays a role in reunification; receipt of AFDC and
lower socioeconomic status is associated with a deceased likelihood of reunification (Barth,
Snowden, Broeck, Clancey, Jordan, & Barusch, 1987; Courtney, 1994; Jones, 1998), and Westat,
Inc. (2001c) found that parents with a high school education who were currently employed were
more likely to be reunified than parents who do not have these attributes. Interestingly, Wells and
Guo (2003) report that mothers who move from welfare to work actually reunify with their children
at a slower rate than other groups. Some studies have also found that children with disabilities or
health problems are less likely to be reunified than children without these problems (Courtney,
1994; McMurty & Lie, 1992; Wells & Guo, 1999). Additionally, parents who are non-compliant
with court orders and treatment are significantly more likely to lose custody of their children than
are court compliant parents (Atkinson & Butler, 1996; Jellinek, Murphy, Poitrast, Quinn, Bishop &
Goshko, 1992; Smith, 2003).
Demographic information indicates that most children in family reunification services come
from single parent homes, often headed by the mother (Fein & Staff, 1993; Gillespie, Byrne, &
24
Workman, 1995; Lewandowski & Pierce, 2002; Walton, Fraser, Lewis, Pecora, & Walton, 1993).
Financial problems are common among family reunification clients, (Fein & Staff 1993; Gillespie et
al., 1995; Walton et al., 1993); as are previous out-of-home placements (Lewandowski & Pierce,
2002; Walton et al., 1993). In California, 16. l percent of Black children are in family reunification
services; 25. l percent of White children; 24.8 percent of Hispanic children, 20.3 percent of Asian
children, and 22 percent of Native American children are in family reunification services (Needell
et al., 2003). Children in family reunification services in California also tend to be young; 43.8
percent of children under the age of one are in family reunification services; 32.9 percent between
the ages of one and two; and 27.6 percent between the ages of three and five are in family
reunification services. In Santa Clara County, 15.7 percent ofBlack children are in family
reunification services, 24.3 percent of White children, 25 percent ofHispanic children, 20 percent
of Asian children and 13.3 percent ofNative American children are in family reunification services.
In Santa Clara County, 33.3 percent of children under the age of one are in family reunification
services, 32.8 percent between the ages of one and two; and 30.7 percent between the ages of three
and five are in family reunification services (Needell et al., 2003).
Family Reunification Services
Family reunification services are generally focused on addressing the broad range of issues
that have contributed to the removal of the child, or that prevent the child from returning home
safely. These services are grounded in the notion that the parent child attachment is of primary
importance and the continuity of these attachments is critical to the well-being and development of
the child (Maluccio & Ainsworth, 2003). Family reunification services are generally offered for a
period lasting 6-18 months and they may include a wide range of practices that vary greatly from
county to county and state to state. As is the case with family preservation practices, a large number
of different interventions fall under the definition of "family reunification services;" some family
reunification interventions are modeled after FPS and are intensive, short term and family-centered;
while others focus on the delivery of specific servi_ces designed to ameliorate particular problems.
As with research on the effectiveness of family preservation services, research on the
effectiveness of FRS is lacking; most family reunification studies have small sample sizes, no
comparison groups, no clear definition of the intervention and often fail to measure family and child
well-being in addition to reunification rates (Littell & Schuerman, 1995). Studies evaluating the
impact ofFRS demonstrate mixed results, with many researchers concluding that there is no solid
evidence that FRS are effective in returning children to their birth families (Littell & Schuerman,
1995). Indeed rates of reunification for participants in FRS have been found to range all the way
from 25 percent to 100 percent (Littell & Schuerman, 1995).
For instance, Fein and Staff (1993) evaluated the effectiveness of a fairly intensive family
reunification program in which services were delivered by both a social worker and a family
support worker; workers carried small caseloads and provided a broad range of both concrete and
clinical services, such as parenting training, budgeting assistance, transportation, help with job
training, as well as counseling, or support for substance abuse treatment. Although no comparison
group was used in the study, results indicated that 38 percent of children were reunified with their
families after one year of participation in services-and of these 10 percent were returned to OHC
at the end of the second year.
25
Some researchers have concluded that family reunification services that are intensive in
nature, sometimes following a family preservation model, may be most effective. Walton et al.
( 1993 ), evaluated the effectiveness of family reunification services that were Iimited to 90 days and
were delivered in the home of the birth parents by workers carrying small caseloads. Interventions
averaged 3.1 hours a week and were focused on concrete services, as well as psychosocial training.
Findings indicated that 93 percent of families receiving FRS reunified after the 90-day treatment
period, compared to 28 percent of a co~parison group receiving regular child welfare services. At a
12-month follow-up, the FRS children group (75%) was still more likely to be living in the home,
than the comparison group (49%).
Similarly, Gillespie et al. (1995) also reported promising results from an intensive
reunification program that was modeled after family preservation services. Services were provided
at the birth parents home, for approximately 8 to 10 hours a week and included a combination of
concrete and clinical services. In addition, this program also provided support and training to the
foster parents caring for the children, increased contact between the foster parents and social
workers, as well as the facilitation ofjoint meetings between foster parents and birth parents to
improve consistency of parenting and to share· information. Birth parents were also allowed
increasing numbers of visits based on their participation in services. Although no comparison group
was used, findings indicated that 79 percent of the children were reunified, and at 12-month followup 71 percent of the sample was sill living at home.
Less encouraging results were noted by Lewandowski and Pierce (2002) in their evaluation
of the effects of Missouri's Family Centered Out Of Home Care (FCOHC) Program, a program that
incorporates family-centered principles aimed at involving families in reunification case plans as
soon as the child enters out of home care. Services were intensive, time-limited and strengths-based.
Progress was monitored through team meetings involving the families, social workers, and other
representatives from social service systems. Differences between families receiving the FCOHC
and comparison groups receiving standard child welfare services were significant although not in
the expected direction. The FCOHC group actually experienced fewer reunifications and more
returns to out of home care, than the comparison group. Yet results did indicate that, on average,
children in the FCOHC group who spent more than seven days in out of home care, also spent fewer
days overall in out of home care than the comparison group.
In addition to the impact of family reunification services themselves, the decision-making
practices of workers also play a role in whether or not families reunify. In Westat's (2001a) analysis
of reunification decision-making practices among nine workers in three public child welfare
agencies in Washington DC, workers reported they could generally tell within the first few months
of a reunification case whether reunification would actually occur. Factors identified by workers as
important in their decision to return a child home included, the parents' compliance with services
and visitations, the strong desire of parents to have their children returned, the safety of the child
and the child's wishes and well-being. Workers also indicated that parents who were noncompliant
with services or non-responsive to efforts at reunification and cases involving young children were
most likely to lead to a decision to terminate parental rights. In general workers in this study felt
that time limits for reunification were a positive factor in the decision-making process because time
restrictions allow them to terminate parental rights more quickly in cases where reunification
appears unlikely; however workers also noted that some families with more serious problems could
be reunified if they received services beyond federal time limits. However since the sample size in
this study was small, these results should be interpreted with some caution.
26
Family Reunification Services:
Impact on Children and Families of Color
The impact offamily reunification services for children and families ofcolor is not
adequately addressed in the research literature. Although we know that children and families of
color tend to reunify less frequently than White children, the types ofservices that increase the
likelihood of reunification for children andfamilies of color are not well understood. In general
FRS research does not clearly support the effectiveness of these interventions, either on family
reunification or on family functioning (Littell & Schuerman, 1995; Lewandowski & Pierce, 2002).
Furthermore, the actual interventions that are associated with "family reunification services" are
often not clearly defined.
Research does suggest that families who fail to reunify often have serious social and
economic risk factors that do not appear to be adequately addressed by current family reunification
services. For instance, families with a decreased likelihood of reunification are those that have
income problems (Barth et al., 1987; Courtney, 1994; Jones, 1998); parents with less than a high
school education; unemployed parents (Westat, Inc., 2001c); and children with health problems or
disabilities (Courtney, 1994; McMurty & Lie, 1992; Wells & Guo, 1999). Yet these characteristics
also describe the general child welfare population. Clearly more research is needed to discern the
types of FRS interventions that are effective with these high-risk groups, as well as the possible
differential outcomes for children and families ofcolor.
Specific Family Reunification Services: Parenting Education, Substance Abuse, Domestic
Violence, and Family to Family/Social Network Interventions
In addition to research on the effectiveness of family reunification services as a whole, some
studies have focused on the effects of specific services aimed at improving particular problems such
as parenting deficits, substance abuse, domestic violence and poor social networks. As with family
reunification services, studies on specific treatments to reunify parents with their children are
limited and often have methodological problems such as small sample sizes, lack of comparison
groups, lack of treatment consistency and lack of adequate long-term follow-up.
Parenting Education Services
Programs aimed at teaching parents new parenting skills are frequently included in family
reunification case plans. For instance, Phase 2 case record review data indicated that in Santa Clara
County, 70.2 percent of mothers were recommended parenting education services at the
jurisdictional/dispositional hearing (Hines et al., 2002). In general parenting education services are
didactic in nature and focus on increasing participants' knowledge of effective parenting skills.
Research on the effectiveness of parenting education programs for families in the child welfare
system is lacking. Most parenting education evaluations have examined the impact of these
programs on parents whose children have specific behavioral problems-not for families who have
had children removed as a result of maltreatment. Research on parenting education for nonmaltreating parents has yielded some promising results (Dore & Lee, 1999). However, only a
handful of studies have evaluated parenting education services for child welfare parents. Results
from these studies suggest that child welfare parents may experience increased benefits from
parenting education programs that combine a didactic approach with another parenting education
27
service (Brunk, Henggeler, & Whelan, 1987; Carlo, 1993; Dore & Lee, 1999; Wolfe, Edwards,
Manion, & Koverola, 1988).
More rigorous studies of traditional didactic parenting education services have been
performed with samples involving parents whose children have behavioral problems. Most of these
studies suggest that low risk parents may experience significant short term benefits from parenting
education services (Petsch, Schultz, & Wahler, 1999; Nicholson, Brenner, & Fox, 1999; Patterson,
Chamberlain, & Reid, 1982). However, the research also suggests that high-risk parents typically
experience little benefit from parenting education services (Dore & Lee, 1999).
Parenting Education Services:
Impact on Children and Families of Color
Research on the impact ofparenting education services for children and families ofcolor is
lacking. Studies fail to mention the possible impact ofcultural differences in parenting behaviors on
the success or failure ofparenting education services (Forehand & Kotchick, 1996). Parenting
education appears to be most effective for lower-risk families, however, numerous stressors such as
poverty, substance abuse, domestic violence, low education levels, and community violence often
impact parents involved in the child welfare system. Dore and Lee (1999) note these risk factors
may make cognitive behavioral and didactic approaches to parenting education particularly unsuited
to these populations:
For those who are highly stressed by poverty and its accompanying pressures, even
sustaining a commitment to the training process can be overwhelming. Those parents who
lack the cognitive foundation needed for more didactic and learning based approaches seem
unable to benefit fully from behaviorally based training models (Dore & Lee, 1999, p. 322).
More promising approaches to parenting education for maltreating parents appear to involve a
multifaceted approach that includes not only instructional training in parenting behaviors, but also
experiential training and interventions that address the emotional and psychosocial stressors
affecting these high-risk parents (Carlo, 1993; Brunk et al., 1987; Dore & Lee, 1999; Wolfe et al.,
1988).
Indeed, there is some limited evidence to suggest that parents in the child welfare system
may experience increased benefits from a parenting education program than combines didactic
parenting education with another parenting related service such as behavior training, experiential
training or multistystemic therapy (Brunk et al., 1987; Carlo, 1993; Wolfe et al., 1988). However,
more research is needed in this area. For instance, it is unclear which parenting related services are
most effective when paired with didactic training and it is also unclear how well parents retain
information learned. Long-term follow-up studies that track the retention of information in
parenting education services for child welfare parents, as well as child and family outcomes, are
needed.
Additionally, few studies have addressed the role of culture and ethnicity in the effectiveness
ofparenting education services/or parents in the child welfare system. Forehand and Kotchick
(1996) note, " ... different cultures have different histories that shape cultural theories about
parenting. Without an awareness of the values of a particular culture, attempts to implement
parenting programs with its members may fail" (Forehand & Kotchick, 1996, p.197). As such,
28
research needs to focus on how parenting education outcomes may differ for different cultural and
ethnic groups and what types ofparenting education approaches might be most successful for
diverse families.
Substance Abuse Treatment
Parental substance use is common among the child welfare population. In 1998, the United
States General Accounting Office [U.S. GAO] conducted a study of two state CWS systems in
California and Illinois, and reported that 65 percent of open foster care cases in California and 74
percent in Illinois, involved a family where one or both parents had a substance abuse problem for
which treatment was required. Research indicates that cases involving parental substance abuse
reunify at a significantly lower rate than do cases not involving substance abuse (Smith, 2003), and
in general, children from substance abusing households experience poorer outcomes than children
from non-substance abusing households (U.S. GAO, 1997).
As a result of the widespread substance use among child welfare parents and the poor
outcomes for cases involving parental substance abuse, many family reunification case plans
include requirements for some form of substance abuse treatment. These interventions can include
various components such as drug testing, 12-step groups, outpatient treatment or inpatient
treatment. For instance, Phase 2 case record review results indicated that 40.9 percent of mothers in
Santa Clara County are ordered random drug or alcohol testing, 40.0 percent are ordered substance
abuse or alcohol treatment and 26.8 percent are ordered 12-step groups (Hines et al., 2002).
Unfortunately few studies have evaluated the impact of substance abuse services for child welfare
parents (Gregoire & Schultz, 2001 ). More research efforts have focused on the effectiveness of
these services for pregnant and parenting women who may not necessarily be involved in the CWS.
In one of the few studies that has evaluated substance abuse treatment among child welfare
parents, Gregoire and Schultz (2001) analyzed the treatment and placement outcomes for 167 child
welfare parents in Lancaster County Pennsylvania. Findings revealed that those participants who
had support from significant others were more likely to complete substance abuse treatment and
also experienced better sobriety outcomes. Moreover, parents who were sober at nine months
follow-up were more likely than parents who were not sober to have child custody or to have
maintained parental rights.
Ways in which child welfare workers make decisions about cases involving substanceaffected parents also impacts the likelihood of reunification. Karoll and Poertner (2002) identified
certain factors considered to be associated with readiness for reunification in their qualitative study
of judge's, caseworker's and substance abuse counselor's indicators of reunification. Factors
associated with professional's indication ofreadiness for reunification included: 1) parental
motivation, 2) evidence of substance abuse recovery, 3) evidence of competence and reliability, 4)
the presence of social support, 5) evidence of improved parenting skills and 6) legal aspects of child
loss and prospects for reunification. However there was considerable variability among respondents
in how many clean drug tests are necessary for reunification. Although the median response was 8
months of clean tests, approximately 30 percent felt 6 months was sufficient, while 32 percent felt
that 12 months of clean tests were needed before reunification should occur.
Although research on the effectiveness of substance abuse treatment for child welfare
parents is scant, more research efforts have focused on the effectiveness of these services for
29
pregnant and parenting women who may not necessarily be involved in the CWS. However, since
most child welfare agencies respond to referrals for prenatal drug exposure efforts to offset drug use
among pregnant and parenting women is an important concern for the CWS (Ondersma, Malcoe, &
Simpson, 2001).
Overall, research on the effectiveness of substance abuse services for pregnant and parenting
women suggests these services may be beneficial, however the degree of benefit appears to depend
on the type and intensity of the intervention, as well as the case characteristics. For instance, there is
some evidence that suggests that residential treatment may be more beneficial than outpatient
treatment for women with substance abuse problems. (Haller et al., 1997). Moreover, parenting
women experience better outcomes in substance abuse treatment when they are able to reside in
residential treatment with their children (Clark, 2001; Metsch et al., 2001). Studies also indicate that
women are more likely to complete treatment and experience better outcomes when they have the
support of significant others (Gregoire & Schultz, 2001 ), and when family and significant others are
involved in treatment (Carten, 1996). Parenting women have also identified support and
encouragement from non-judgmental staff as particularly important in their ability to successfully
complete treatment (Akin & Gregoire, 1997; Carten, 1996).
Substance Abuse Treatment:
Impact on Children and Families of Color
Research on the effectiveness ofsubstance abuse treatment for parents color in the child
welfare system is lacking. In general, studies do not describe differential outcomes for various
racial/ethnic groups in these treatments and few studies have actually examined the impact of
substance abuse treatment on reunification outcomes for children and families of color in the child
welfare system. However, research does suggest that parenting women involved in substance abuse
treatment are a high-risk group. Most participants of these services are young single women of
color, with low educational levels and income problems (Carten 1996; Clark, 200 I; Knight et al.,
2001; Metsch et al., 2001 ). In general, research on the effectiveness of substance abuse treatment
services for these women has yielded promising results; however the degree of treatment impact
appears to depend on the type and intensity of the intervention, as well as case characteristics.
Studies suggest that participants in substance abuse treatment services are more likely to maintain
child custody (Gregoire & Schultz, 2001); and maintain sobriety (Clark, 2001; Metsch et al., 2001).
However, women with less than a high school education, criminal justice system involvement and
poor social supports are less likely to complete treatment (Knight et al., 2001 ).
Overall, research suggests that high-risk parenting women benefit from substance abuse
treatment that is comprehensive; allows them to involve their children and other family members or
supports; and is delivered by supportive, nonjudgmental staff. Although research has not
specifically evaluated the role ofethnic diversity in substance abuse treatment services for
parenting women in the child welfare system, the fact that many of the participants of these services
are women ofcolor does suggest that these services may be beneficial for a diverse population.
However, more research is needed to determine the role of culture in treatment, as well as possible
differential outcomes for various ethnic/racial groups.
Domestic Violence Treatment
Child welfare professionals are becoming increasingly aware of the importance of
30
addressing issues of domestic violence for families attempting to reunify with their children. Rates
of domestic violence among families in the CWS vary from 38 percent to 54 percent (Bowen, 2000;
McGuigan & Pratt 2001 ). Phase 2 results from case record review data indicated that in Santa Clara
County, 45.4 percent of cases involve domestic violence (Hines et al., 2002). Research from the
child welfare field on the effectiveness of domestic violence treatments in reunifying families and
improving outcomes for families is extremely lacking. Instead, child welfare research efforts related
to domestic violence have primarily focused on describing barriers to effective collaboration
between domestic violence and child welfare professionals and attempts to overcome these barriers.
Indeed, practioners and researchers have noted that child maltreatment and spousal abuse
service systems have historically operated in isolation from one another and typically have different
service delivery approaches (Beeman, Hagemeister, & Edleson, 1999; McKay, 1994). Recent
efforts to increase collaboration between the child welfare and domestic violence fields has focused
on cross-training between the two professions, as well as simplified processes for referring families
to child welfare or domestic violence services (Findlater & Kelly, 1999; Whitney & Davis, 1999).
Additionally, the creation of specific child welfare policies for families experiencing domestic
violence and the use of multidisciplinary task forces to address ongoing issues related to
collaboration between fields has also been described as helpful in increasing coordination and
collaboration (Findlater & Kelly, 1999; Whitney & Davis, 1999).
·
Santa Clara County has implemented efforts to merge child welfare and domestic violence
fields in order to better serve families affected by both child maltreatment and domestic violence.
Santa Clara County is implementing the Greenbook Project, a multidisciplinary effort involving
representatives from child welfare, the domestic violence community, law enforcement and the
juvenile court. The overall goal of the project is to change the way these groups assist families who
experience both domestic violence and child maltreatment. During the first year of implementation,
the Greenbook Project placed domestic violence advocates from Next Door Solutions to Domestic
Violence at the County Department of Social Services. These advocates provide consultation to
child welfare workers on issues of domestic violence at every level of the child welfare system.
Additionally, the Greenbook Project also visited the Miami-Dade County Dependency Court
Intervention Program for Family Violence to learn key aspects of a domestic violence advocacy
program. Other activities of the Greenbook Project include cross-training between child welfare,
domestic violence, law enforcement and legal personnel, improving cultural competency,
developing guiding principles, changing agency policy and practice, and integrated parenting
programs (Lightbourne, 2002).
The collaboration of child welfare and domestic violence fields to better meet the needs of
families affected by both issues is promising. However, in general these efforts are not focused on
specific treatments for the batterer. Indeed, child welfare research has not addressed the
effectiveness of domestic violence services for child welfare parents. Yet studies from related fields
have evaluated effectiveness of domestic violence treatment for batterers who may not necessarily
be involved in the CWS. Domestic violence treatment for batters typically involves a group
treatment that includes psycho-educational and cognitive behavioral techniques designed to change
faulty beliefs, reduce anger, improve communication and increase non-violent coping skills
(Morrel, Elliot, Murphy, & Taft, 2003). Less common are psychodynamic or support groups.
Findings from studies on the effectiveness of these treatments are mixed and much of the research is
limited due to small sample sizes, lack of comparison groups and findings that rely on self-reports.
Some studies have reported that targeted domestic violence services are effective in reducing
31
violent behaviors (Dutton, 1986; Palmer, Brown, & Barrera, 1992), yet other studies report no
benefits from these treatments (Dunford, 2000). As with many other treatments, domestic violence
treatment tends to be most effective for participants with relatively few other risk factors (Morrel et
al., 2003; Taft, Murphy, Elliot, & Keaser, 2001).
Domestic Violence Treatment: Impact on Children and Families of Color
Research on the impact ofcross-system collaboration and specific domestic violence
treatment for children and families of color is lacking. Although child welfare research describing
cross-system collaboration between domestic violence professionals and child welfare professionals
is promising, these-studies do not address the role ofcultural issues on domestic violence, or the
treatment of batterers and battered women. Similarly, research from related fields describing
outcomes of didactic domestic violence treatment also rarely consider the role of culture or
ethnicity.
Research suggests that the domestic violence field and the child welfare field have
historically had different approaches to treatment that have hindered collaborative efforts (Beeman
et al., 1999). Recently, more concerted efforts to link the two fields have occurred. Typically these
efforts involve the use of multidisciplinary task forces to address issues related to collaboration
(Findlater & Kelly, 1999; Lighbourne, 2002); cross-training (Findlater & Kelly, 1999; Lighbourne,
2002; Whitney & Davis, 1999); simplified processes to refer battered women to child maltreatment
prevention programs and to refer battered child welfare mothers to domestic violence services
(Findlater & Kelly, 1999; Whitney & Davis, 1999); and, the creation ofspecific child welfare
policies stating procedures for families experiencing both child maltreatment and domestic violence
(Findlater & Kelly, 1999; Whitney & Davis, 1999). Although these efforts have great potential to
impact children and families of color in the child welfare system, research has not yet addressed this
topic.
The collaborative efforts between child welfare and domestic violence tend to focus on
services and assistance to women and children, with less ofan emphasis on treatments for the
perpetrators of domestic violence. As with other didactic interventions, such as parenting education,
there is evidence to suggest that domestic violence treatment for batterers is more effective with
low-risk groups than with high-risk groups (Morrel et al., 2003). However, research also suggests
that most men who are participating in domestic violence treatment are a high risk group who often
experience multiple stressors (Brown, Werk, Caplan, & Seraganian, 1999). Moreover, the presence
of certain risk factors appears to reduce the likelihood of treatment completion (Taft et al., 2001).
Additionally, one study found that African American men were most likely to drop out of domestic
violence treatment, even after other factors were statistically controlled, suggesting that these
treatments may not be as effective for African Americans (Taft et al., 2001). However, in general,
more research is needed to discern the impact ofdomestic violence treatments/or various
racial/ethnic groups in the child welfare system.
Social Network Interventions/Family to Family Initiative
Some child welfare services have focused on the role of social network interventions on
improving outcomes for children and families. Social network interventions are designed to enhance
both formal and informal social networks, including increased networking among family members,
32
friends, neighbors, as well as more structured networking involving support groups (Gaudin,
Wodarski, Arkinson, & A very, 1990). These interventions are based on research indicating that
people living in neighborhoods in which there are high child maltreatment rates tend to describe
these communities as lacking in formal and informal support systems (Garbarino & Kostelny,
1992). Additionally, DePanfilis (1996) has noted that in addition to a lack of social support systems,
neglectful families may be particularly distrustful of any support systems that are available.
In an effort to address these issues, social network interventions, such as the Family to
Family Initiative have been implemented. Although the goals of Family to Family are extensive and
go beyond just social network interventions, one of the initiative's main emphases is on
"strengthening the network of families available to care for abused and neglected children in their
own communities" and "building partnerships with at-risk neighborhoods toward that end" (Annie
E. Casey Foundation, 2003b, p. 1). Indeed, many interventions provided through Family to Family
are designed to strengthen neighborhood networks for birth and foster families in the child welfare
system. These strengthened networks are considered to have the potential to assist families in
learning new skills and becoming more self-sufficient. The tactics used to increase social networks
for families, and also increase collaboration between the child welfare system and local
communities are varied. In general there is an emphasis on identifying families' naturally occurring
support networks and involving these networks in services. Interventions can be focused on
identifying and utilizing "natural helpers" in the communities and neighborhoods where birth and
foster families reside. Natural helpers are considered to understand the culture of a particular
community better than professional helpers. Efforts are also aimed at increasing overall
collaboration between child welfare agencies and the communities in which child welfare clients
reside.
Outcome studies on the effectiveness of Family to Family interventions are not available.
However, Gaudin et al. (1990) evaluated the effectiveness of the Social Network Intervention
Project (SNIP) for 52 neglectful families in the child welfare system. The intervention, which lasted
approximately 10 months, included first an assessment of existing formal and informal support
networks in the lives of the families, and then a series of interventions designed to enhance these
networks. Direct interventions were made with informal supports, including family members,
friends or neighbors. Efforts were also made to identify and utilize informal helpers existing in the
families' neighborhoods. Form~l support networks were also encouraged through the use of support
groups, parent aides and social skills training. Post-test results from the evaluation indicated that
compared to a comparison group who received normal child welfare interventions, the families
receiving the social network services fared better on a number of outcomes assessed by their
caseworkers. Intervention parents were rated as having improved parenting skills and attitudes, and
they experienced significant increases in the size of their formal and informal support networks. At
the end of the intervention, 59 percent of the families had their child welfare cases closed, while
23 .5 percent of the comparison group has their cases closed.
Although the use of social network interventions may be a promising child welfare practice,
Gaudin et al. (1990) notes that there are organizational and professional barriers to the
implementation of such interventions. For instance, results from the SNIP evaluation indicated that
it was difficult for social workers implementing the project to maintain a focus on social network
interventions within the context of a public child welfare agency. The predominate model of
services in the agency tended to focus on case management services, and barriers such as high
33
caseloads, heavy paperwork, frequent staff turnover, and lack of collaboration between human
service agencies all contributed to difficulties in implementing the social network intervention.
Family to Family/Social Network Interventions:
Impact on Children and Families of Color
Researchers have not evaluated the impact ofsocial network interventions, such as the
Family to Family Initiative on children and families ofcolor. Given that children and families of
color involved in the child welfare system are likely to reside in impoverished communities, the use
ofsocial network interventions with this population is ofparticular interest. Research exploring the
assets and informal supports imbedded in impoverished communities is limited. It certainly appears
likely that strengthening support networks for vulnerable families would be beneficial, however
research on the mechanisms through which to increase support networks for families is lacking.
Indeed, the exact interventions associated with utilizing informal helpers, strengthening
neighborhood networks and increasing overall collaboration between child welfare agencies and
the communities in which child welfare clients reside are not clearly defined. As Gaudin et al.
(1990) has suggested, there may be significant barriers to effective implementation of these
interventions within the crisis-driven, resource strapped child welfare field. More research is
needed to first discern the types of interventions associated with social network practices, as well as
the ways in which outcomes from these services may differ for various racial/ethnic groups.
Mental Health Services for Children in Out-of-Home Care
Children in the child welfare system are at an increased risk for a variety of psychological
and behavioral problems (Stevenson, 1999). Studies have documented the long-term negative
consequences of both child maltreatment and lengthy stays in out-of-home care (Courtney, Piliavin,
Grogan-Kaylor, & Nesmith, 2001; Stevenson, 1999). In order to ameliorate these negative
consequences, many children in out-of-home care are referred to mental health services. Research
suggests that mental health services are the most commonly recommended treatment for children in
the child welfare system (Kinard, 2002). Studies have also found that children in foster care are
more likely to have a mental health diagnosis and to use mental health services than comparison
groups. For instance, children in foster care are more likely than children receiving public assistance
or children receiving SSI to have a mental health diagnosis (Harman, 2001; dosReis, 2001);
children in foster care tend to have more mental health hospitalizations; and use more mental health
services overall than children receiving public assistance (Reiff, 2001 ).
Yet despite the relatively high rate of mental health service usage among children in the
CWS, the nature and impact of these services are not well understood. Indeed mental health services
can include a broad range of interventions, including support groups, individual counseling, family
counseling, residential services or inpatient services. One model of services, Treatment Foster Care
(TFC), has been the topic of child welfare research. TFC, also referred to as therapeutic foster care
is an intensive, family centered intervention that seeks to place youth with behavioral and emotional
disorders in community and family settings. Children are typically placed with foster parents and
the placement is intended to be both therapeutic and family-centered. Intensive services are
delivered to parents, alternate caregivers and the child. Some studies have suggested that TFC can
be an effective transitional placement for children moving from restrictive settings into less
restrictive placements (Chamberlain, 1994, as cited in Farmer, Wagner, Bums, & Richards 2003;
Clarke, Hawkins, Murphy, & Sheeber, 1993, as cited in Farmer et al., 2003). However, not all
34
studies have found TFC to be effective. For instance, Farmer et al. (2003) used administrative data
to analyze the impact ofTFC on foster youth with emotional or behavioral disorders. Findings
indicated that use of TFC did not result in reduced use of group homes.
In addition to TFC, some child welfare interventions for children focus on a "wraparound"
process that is targeted to high risk children who may have emotional or behavioral problems and
who are currently placed in a level 12 or higher group home, or are at risk for entry into a more
restrictive placement (California Department of Social Services [COSS], 2001). Wraparound
services are described as a "family-centered, strength-based, needs-driven planning process for
creating individualized services and supports for children and their families" (COSS, 2001, p. 1).
Wraparound interventions are collaborative in nature, often involving multiple social service
systems in the planning and delivery of a wide range of services. Few studies have systematically
investigated the efficacy of wraparound services for children and families in the child welfare
system. However, some research suggests that children in wraparound are less likely to be removed
from their community and may experience a reduction in behavioral problems (Bruns, Burchard, &
Yoe, 199S, as cited in Toffalo, 2000; Clark, Lee, Prange, & McDonald, 1996, as cited in Toffalo,
2000; VanDenBerg, 1993, as cited in Toffalo, 2000). Yet other research has reported no significant
benefit from wraparound services (Clarke, Schafer, Burchard, & Welkowitz, 1992, as cited in
Toffalo, 2000).
Mental Health Services:
Impact on Children of Color in Out-of-Home Care
Although research is lacking in identifying effective mental health practices for children in
foster care, studies have demonstrated that children ofcolor are less likely to be recommended for
these services and to use these services. In Santa Clara County, Phase 2 case record review results
indicated that 36.2 percent of children were recommended individual counseling and 19 .6 percent
were recommended other counseling services. Differences between racial/ethnic groups in
recommended counseling services were also found. For instance, 18 percent of Black children were
recommended individual counseling, 46.3 percent of White children, 31.9 percent of Latino
children; and 35.5 percent of Asian American/Pacific Islander children were recommended
individual counseling, suggesting that children of color, and in particular African American children
have a decreased likelihood of being recommended individual counseling (Hines et al., 2002).
Differences in service recommendation and provision were also noted by Garland and
Besinger ( 1997) in their analysis of racial/ethnic differences in court referred pathways to mental
health services for children in foster care. Results suggested a trend for White children to be more
likely than African American or Hispanic children to be court ordered counseling/psychotherapy.
White children were also significantly more likely than African American or Hispanic children to
receive counseling during their first eight months in out-of-home care. And although not significant,
White children had higher rates of service use in general than did African American or Hispanic
children. Racial/ethnic differences in mental health service recommendations and service usage
have been found to remain even after the effects of other demographic factors are controlled
(Garland, Hough, Landsyerk, McCabe, Yeh, Ganger et al., 2000).
Research on the effectiveness of mental health practices for children in out-of-home care is
lacking, and even less is known about the impact of these services for children of color.
Comprehensive treatments for high-risk children, such as treatment foster care and wraparound
35
services have not been adequately researched and have in-general not consistently demonstrated
positive effects. Studies on other mental health interventions for children ofcolor in the child
welfare system are lacking. However, research does suggest that children of color are less likely
than White children to be referred to mental health services (Garland & Besinger, 1997; Hines et al.,
2002). The reduced likelihood for children ofcolor to be referred to mental health services is of
particular concern because children ofcolor experience long stays in the child welfare system and
are at an increased risk for a variety ofpoor outcomes (Courtney et al., 1996). As such, more
efforts should be aimed at ensuring that children ofcolor are referred to and receive appropriate
mental health support, and more research is needed to evaluate the impact ofthese services on
outcomes for children ofcolor.
Foster Parent and Kinship Support Services
Although the majority of research on child welfare practices has focused on interventions
with birth families and children, some research has emphasized the importance of services and
training for alternative caregivers, including foster parents and kin caregivers (Cuddeback & Orme,
2002; Gordon, McKinley, Satterfield, & Curtis, 2003). Some studies have suggested that kin
caregivers may receive fewer services than non- kin caregivers (Gordon et al., 2003; Scannapieco,
Hegar, & McAlpine, 1997, as cited in Cuddeback, 2002). However, other research has reported no
differences between services delivered to kin and non-kin caregivers (Cuddeback & Orme, 2002).
Research suggests that the use of kin placements is common in California and in Santa Clara
County. In 2002, 36.4 percent of children in supervised foster care in California were in a kin
placement and in Santa Clara, 40.1 percent were placed with kin (Needell et al., 2003).
Additionally, children and families of color are especially likely to be placed with kin. For instance,
point in time reports indicate that in California, 40.5 percent of Black children in supervised foster
care are placed with kin; 37.0 percent ofNative American; 35.8 percent of Hispanic, 29.5 percent of
White and 29.4 percent of Asian children are placed with kin. In Santa Clara County, 66.7 percent
ofNative American children are placed with kin; 41.9 percent of Hispanic; 41.4 percent of Black;
35.6 percent of Asian; and 32.5 percent of White children (Needell et al., 2003).
The prevalence of kin placements has prompted some researchers to explore the service and
support needs of kin caregivers. Gordon et al. (2003) noted the unique needs of kin caregivers in
their analysis of focus groups with 3 7 kin caregivers. Findings revealed that unlike non-kin
caregivers, many kin caregivers did not expect to be caring for the children in their care; they had to
adjust their family roles and many described this adjustment as sudden and significant. Kin
caregivers also described having to adjust to the involvement of the child welfare agency in their
lives; many reported feeling unappreciated by the child welfare agency and excluded from the
decision-making process for their kin children. They also described a great need for additional
support services as they took on the unexpected responsibility of caring for their kin children.
Other studies have noted that non-kin caregivers also experience unique challenges. For
instance, although non-kin caregivers are expected to support reunification efforts, many may have
conflictual feelings about supporting contact between the children in their care and the birth
families (Sanchirico & Jalonka, 2000). As a result, some child welfare practices have focused on
providing specialized training and support to non-kin caregivers to assist in the process of
reunification. For instance, Sanchirico and Jalonka (2000) surveyed 650 non-kin caregivers in New
York State and noted that only 19 percent received both specialized training to be a foster parent
36
and support and guidance from the agency. Those who received both training and support also
performed the most activities related to encouraging contact between children and their birth
families. Foster parents who received either training or support but not both also exhibited some
tendency to encourage parent child contact, however those who received both training and support
were significantly more likely to support reunification efforts. These results suggest that specialized
services delivered to foster parents can have a beneficial impact on reunification.
Additionally, other research has focused on evaluating more concerted efforts to involve
foster families in the reunification process. For instance, Landy and Munro ( 1998) evaluated the
impact of the Shared Parenting program. The Shared Parenting program encourages foster families
to work as a team with birth parents in order to improve the parenting skills of the parents. Foster
families were encouraged to provide support and guidance to birth parents through on-going contact
and involvement. An improved relationship between foster and birth families was expected to
increase birth parents motivation and interest in learning new skills, thereby increasing
reunification. Thirteen families participated in the evaluation. Results revealed difficulties in
retaining parents in the program. Over half of the participants did not complete the program, either
because of lack of compliance or because workers discharged families due to changes in status that
decreased the possibility of reunification. Outcomes for parents who remained in the program
revealed that those with higher socioeconomic status, families who moved less, families with no
health problems and families with less risk factors were more likely to reunify.
Although research suggests that kin and non-kin caregivers have different service needs,
studies have demonstrated that overall, both kin and non-kin caregivers have numerous unmet
service needs. For instance, Cuddeback and Orme's (2002) conducted a national evaluation of the
service experiences and needs of kin and non-kin caregivers, and found that both groups reported
numerous unmet service needs. Most alternate caregivers expressed a need for day care, respite
care, liability insurance, heath care, and transportation. Additionally, the majority of caregivers in
both groups reported feeling unprepared to care for their foster children. The authors concluded "the
levels of training and services received were inadequate for both groups of caregivers" (Cuddeback
& Orme, 2002, p. 900).
Foster and Kinship Support Services:
Impact on Children and Families of Color
The impact offoster and kinship support services ofchildren and families ofcolor has not been
adequately researched. However, research clearly indicates that kin and non-kin caregivers are in
need of increased training and supportive services. Kin caregivers may need increased support as
they take on the unexpected responsibility and role of a caregiver (Gordon et al., 2003), while nonkin caregivers may need additional support and training to facilitate the reunification process.
(Sanchirico & Jalonka, 2000). In general, both kin and non-kin caregivers describe the need for
increased services, such as day care, respite care, liability insurance, heath care, transportation,
specialized training and increased participation in the child welfare decision making process
(Cuddeback & Orme, 2002; Gordon et al., 2003; Sanchirico & Jalonka, 2000).
Research on the impact offoster and kinship support services on child welfare outcomes is
lacking. However, there is some evidence to suggest that interventions with foster parents are
associated with better reunification outcomes (Sanchirico & Jalonka, 2000). Yet more research is
clearly needed to discern the impact ofservices to both kin and non-kin caregiver, especially for
37
different racial/ethnic groups.
Permanency Planning Practices
The federal Adoption and Safe Families Act of 1997 shortened reunification time frames
and mandated the use of concurrent planning-a process in which both a reunification plan and an
alternative permanent placement plan are created. For children who are not reunified, the permanent
placement plan with its goal of finding a suitable alternative permanent placement for the child
becomes a major focus of services. As such, permanency planning and placement practices often
·
impact children at the ending stages of the CWS.
Very few studies have investigated the types of practices that are used at the ending stages in
the system and how these practices impact children of color. However, the impact of child welfare
practices on children who are not reunified is particularly important for children and families of
color. The U.S. Department of Health and Human Services (2002) reports that approximately 23
percent of children in foster care have had parental rights terminated and children of color are
disproportionately represented in· this group. Nationally, 43 percent of legally free children are
African American, 34 percent are White, 13 percent are Hispanic, 2 percent are Native American
and 1 percent are Asian Pacific Islander (U.S. DHHS, 2002). In California, 68.2 percent of Black
children in foster care are in permanent placement services, 48.4 percent of White children; 41.5
percent of Hispanic children; 32.5 percent of Asian children, and 53.9 percent of Native American
children are in permanent placement services (Needell et al., 2003). Santa Clara County has similar
racial/ethnic rates of permanent placement clients; 64.4 percent of Black children are in permanent
placement services; 44.7 percent of White children; 41.3 percent ofHispanic children; 30.6 percent
of Asian children and 43.3 percent of Native American children are in permanent placement
services (Needell et al., 2003).
Federal policy mandates that these children have safe and permanent alternative homes,
however, in reality many wait months and years for a permanent placement and some will reside in
long term foster care until they emancipate from the system at age 18. In Kemp and Bodonyi's
(2002) analysis of legally free children in the state of Washington, the median length of stay in outof-home care for these children was 50. 7 months and demographic factors were significantly related
to permanency outcomes. For instance, older children, boys, African Americans, and Asian Pacific
Islanders were all less likely to achieve a permanent placement than others groups. Interestingly,
Latino children were significantly more like than White children to achieve a permanent placement.
The interaction of demographic characteristics was also significant in predicting permanency. Older
African American boys had an even higher chance of not achieving a permanent placement. Other
research corroborates these findings. For instance, African American children have been found to be
less likely to be adopted than White or Latino children and African Americans who are freed for
adoption wait significantly longer periods of time for a permanent placement than do children from
other racial/ethnic groups (Barth, Courtney & Berry, 1994, as cited in Kemp & Bodonyi, 2002;
Kapp, McDonald, & Diamond, 2001).
The federal Multi Ethnic Placement Act (MEPA) of 1994 and the Interethnic Adoption
Provisions (IEP) of 1996 were enacted to increase the number of children who were adopted by
preventing discrimination in the placement of children on the basis of ethnicity and allowing for
increased use oftransracial adoption. The impetus behind MEPA-IEP was a belief that children of
color, especially African Americans were languishing in care, while workers searched for same race
38
foster or adoptive homes. MEPA-IEP also called on child welfare agencies to increase efforts to
recruit foster and adoptive parents who reflect the ethnic and cultural backgrounds of their children
in care. Although MEPA-IEP may have the potential to reduce the time legally free children spend
in substitute care, rates of adoption for children of color in the CWS do not appear to have improved
since its enactment. Indeed, Courtney (1997) suggests that transracial adoption policies and
practices may not actually have a large impact on children of color. Because many legally free
children of color in out-of-home care are older and may have disabilities or health problems, many
potential adoptive parents may choose not to adopt these children. Thus, even though child welfare
agencies may seek transracial adoptions, the actual pool of available adoptive parents may be too
limited to have an impact on children of color waiting for a permanent placement (Courtney, 1997).
In an effort to understand the factors that contribute to adoptive parents willingness to adopt
Black children, Brooks and James (2003) surveyed 541 participants of the California Long-Range
Adoption Study, a longitudinal study of adopters in California. Results indicated significant
differences between participants who were willing to adopt Black children and those who were not.
Of the sample 36 percent were willing to adopt a Black child. In general, parents who were willing
to adopt Black children were also more willing to adopt children of varying ages and special needs
children; they were more likely to be motivated to adopt due to religious or humanitarian reasons
and many eventually did adopt Black children from foster care. Interestingly, approximately 50
percent of parents who indicated they were unwilling to adopt a Black child also reported that
adoption workers or agencies discouraged them from adopting Black children, while only 11
percent of those parents who were willing to adopt Black children were encouraged by their worker
or agency to do so.
Additionally, parents who adopt transracially may experience unique challenges and have
special service needs. In an effort to identify the special needs of transracial adoptive parents Vidal
de Haymes and Simon (2003) conducted interviews with 20 youth and their adoptive parents. Youth
in the sample were either African American (70%), Latino (15%) or biracial (15%), while 95
percent of adoptive parents were White and 5 percent were Latino. Findings revealed that parents
often felt that others misjudged their intentions for adopting a child of color and that their families
"deserved pity" (Vidal de Haynes & Simon, 2003, p. 257). Many families reported that child
welfare workers, especially workers of color did not adequately support their decision to adopt a
child of color. Parents felt that workers viewed their desire to adopt transracially as curious, and
many felt that they received only minimal assistance in their efforts to adopt a child of color.
Parents identified numerous needed pre and post adoptive supports and resources. The most
frequently noted needed services were formal linkages to cultural activities, products, or services
that could educate parents on cultural issues and also provide a link for the child to their own
cultural group. Other needed support services included establishing a network of parents who
adopted transracially, educating workers to support transracial adoption, providing parenting classes
to address issues of cross cultural parenting, and developing a mechanism to help White parents
address their own possible racism, as well as racism from their family and friends.
In an effort to understand how worker perceptions may impact the likelihood of transracial
adoption for children of color, Carter-Black (2002) conducted an ethnographic study often Black
child welfare workers. Although the sample size makes generalizing the results problematic,
findings revealed that workers felt Black children need to develop an ethnic identity and an
understanding of their cultural heritage and that the CWS in general does not meet the needs of
Black clients and poor clients. The recruitment, support and maintenance of Black adoptive and
39
foster parents by the agency were seen as lacking. Many also felt that MEPA-IEP could actually be
harmful to Black children because it may result in agencies being less culturally aware or sensitive
to diverse populations.
Avery (2000) also explored the impact of worker perceptions on permanent placement
practices in her study on the "hardest to place children" in New York State. Eighty case files were
examined and children's workers completed a questionnaire on their perceptions and practices
related to these children. Demographics of the sample are consistent with other studies on children
who have lengthy stays in out-of-home care. Seventy three percent of the sample were male, 60
percent were Black, 30 percent White and 10 percent of another ethnicity, many had substantial
disabilities; approximately 80 percent had siblings in the CWS and they spent approximately 11. 79
years in substitute care. Responses to survey questions revealed that 41.4 percent of workers
believed that children in the sample were ultimately not adoptable. These beliefs appeared to result
in reduced efforts on behalf of the workers to recruit potential adoptive parents. Worker attitudes
toward appropriate placements revealed that 44 percent felt placement with gay or lesbian parents
was not appropriate, and 42 percent felt transracial adoption was not appropriate. When asked to
identify resources that would be helpful in placing the children, workers most often identified
respite services, intensive adoptive parent training, intensive post-placement services, more
effective recruitment techniques and higher adoption subsidies. These findings suggest that workers
negative perceptions of hard to place children appear to result in reduced efforts to recruit adoptive
parents.
Permanency Planning Practices:
Impact on Children and Families of Color
In general, child welfare research has focused relatively little attention on practices at the
later stages of the child welfare system. Although child welfare services have the goal of ensuring a
permanent placement for children, research indicates that many children who have had parental
rights terminated may wait years in out-of-home care before a permanent placement is found (Kemp
& Bodonyi, 2002), and many children remain in out-of-home care until they emancipate at age 18
(Courtney et al., 2001). Yet the impact of permanency planning practices on outcomes for children
of color is of great importance. Research indicates that African American children are
overrepresented among children who have had parental rights terminated and are waiting for a
permanent placement. African American and Asian Pacific Islander children both have a reduced
likelihood ofachieving a permanent placement, and older African American boys have an even
lower likelihood ofa permanent placement (Kemp & Bodonyi 2002).
Some researchers have noted that MEPA-IEP and the issue oftransracial adoption may
actually have little impact on children of color in permanent placement services. Because the pool
of adoptive parents willing to adopt legally free children of color (who are often older, and may
have disabilities or health problems) may be relatively small, policies aimed at transracial adoption
may not have their intended impact (Courtney, 1997). Yet other research suggests that the
perceptions and practices of workers may have an important influence on the likelihood of
achieving a permanent placement. Indeed, negative perceptions of hard to place children may result
in reduced efforts to recruit adoptive parents (A very, 2000). Additionally, parents who have adopted
transracially report that workers often did not adequately support their decision to adopt
transracially (Vidal de Haymes & Simon, 2003), and Black child welfare workers have expressed
some hesitancy regarding MEPA-IEP, believing that it could be harmful to Black children because
40
it may result in agencies being less culturally aware or sensitive to diverse populations (CarterBlack, 2002). However, in general more research is needed to discern the types of permanent
placement practices that impact children and families of color.
Impact of Contextual Factors on Child Welfare Practices
for Children and Families of Color
Contextual factors can impact both the delivery of child welfare practices and the outcomes
of these.practices. Although numerous contextual factors exist, this section will focus on three
areas: 1) cultural competency of child welfare workers and practices, 2) child welfare practice in
rural areas, and 3) the role of the court system. The potential impact of these contextual factors on
practices and outcomes for children and families of color will be discussed.
The Role of Cultural Competency in Child Welfare Practice
Child welfare workers often work with children and families from a wide range of cultures
other than their own. Many practioners and researchers have noted that effective child welfare
practices are those that acknowledge and incorporate the importance of culture in the delivery of
services (Miller & Gaston, 2003). Indeed, Miller and Gaston (2003) note that inherent assumptions
within the child welfare system are grounded in Anglo-Saxon values and cultural norms about child
rearing and family. Child welfare legislation and policies often follow European standards of
culture and White, middle class family values are the standard through which ethnically diverse
parents and children are compared. As such, children and families exhibiting alternative values may
be seen as deviant by the system. These conflicts in attitudes regarding acceptable parenting
behavior may contribute to ineffective or harmful child welfare practices (Miller & Gatson, 2003).
In an effort to combat ethnocentrism in the child welfare system, many agencies have placed
increasing importance on ensuring that workers, programs, policies and practices are "culturally
competent." In general, the term cultural competence refers to an ability to recognize and respect
similarities and differences in beliefs, interpersonal styles, values, norms, and behaviors of various
ethnic and cultural groups (Roberts, 1990, as cited in Schriver, 1998).
Although cultural competency is widely viewed as an important aspect of effective child
welfare practice, research on the cultural competency of child welfare systems and the impact of
this competency on children and families is lacking. Some studies describe the problems diverse
clients experience in trying to access culturally and linguistically competent services, other studies
have described ways to improve cultural competency in child welfare agencies, while other research
has addressed mechanisms for child welfare agencies to work more collaboratively with ethnically
diverse communities.
The issue of access to culturally and linguistically appropriate services for children and
families of color is of tremendous importance. ·Suleiman (2003) notes that for Latino clients in
particular, language barriers can exclude them from necessary services and produce significant
delays in service delivery. Indeed language barriers are implicated as a factor in the underutilization
of services among Hispanic clients (Gomez, Ruiz, & Rumbaut, 1985, as cited in Suleiman,.2003),
and many researchers have discussed the importance of language in accurate assessment and
treatment of clients (Suleiman, 2003). Suleiman (2003) further notes that denial of access to
41
services and programs for people with limited English proficiency is in violation of the Civil Rights
Act of 1964; yet federally funded agencies have wide discretion in the degree to which language
needs of clients are addressed.
As such, the responsibility to provide and monitor culturally competent and linguistically
appropriate services has largely fallen on child welfare systems themselves. Some studies have
proposed various models of achieving cultural competence in child welfare agencies. For instance,
Mederos and Woldeguiorguis (2003) discussed the role of child welfare managers in producing
culturally competent child welfare agencies. Recommendations for increasing cultural competence
included first and foremost, the managers' recognition that disproportionality in services and
outcomes for children of color does exist. Indeed, other researchers have noted that a prerequisite to
true cultural competence within an agency is a commitment to cultural competence principles and
practices among top level staff (McPhatter & Ganaway, 2003). Additionally, to produce agency
wide cultural competence, managers must prioritize staff diversity through recruitment and
maintenance of bilingual and bicultural staff. Child welfare managers also need to reach out to
diverse communities to form collaborations with community representatives. This collaboration can
help the child welfare system better understand cultural differences in parenting behaviors, as well
as assist in the creation of policies and practices that are relevant to those communities.
Indeed, other studies have also documented a need for improved collaboration between child
welfare agencies and culturally diverse communities. For instance, Hosley, Gensheimer, and Yang
(2003), described the implementation process of two programs aimed at improving collaborations
between the child welfare system and two Southeast Asian communities in Minnesota. Both
programs sought to improve practices and outcomes for children and parents from these
communities. The collaborations required committed involvement from numeroll:s agencies, as well
as parents and youth from the communities. Language barriers made the use of bilingual and
bicultural staff at collaborative meetings extremely important. Effective collaboration depended on
a number of different factors, including the need for a trusting relationship between all parties, the
incorporation of different cultural styles in the structure of meetings, decision-making and
communication styles, the need to have bilingual and bicultural staff involved in the process and
that active involvement from the youth and parents is crucial to successful collaboration.
Cultural Competency:
Im pact on Children and Families of Color
Research has not yet investigated the impact of culturally competent child welfare practices
on children and families ofcolor, however there is evidence to suggest that practices that are not
culturally or linguistically appropriate result in underutilization of services and ineffective
treatments and are in violation of clients civil rights (Gomez et al., 1985, as cited in Suleiman,
2003; Suleiman, 2003). Some researchers have described processes for achieving cultural
competence within child welfare agencies. A wide range of activities aimed at increasing cultural
competence have been identified, including support from top level staff, the prioritization of staff
diversity, and forming collaborations between child welfare and ethnically and culturally diverse
communities. These activities are intended to combat the child welfare system's legacy of using
European standards of culture and White, middle class family values in assessing diverse children
and families (Miller & Gatson, 2003).
42
However achieving cultural competency in child welfare can be a difficult task. For instance
it may be challenging for child welfare agencies to recruit and retain bilingual/bicultural staff and
the process of building collaborations with diverse communities may be time consuming and labor
intensive. Yet the impact of cultural competency on children and families of color is of tremendous
importance. More research is needed to discern how child welfare agencies recruit and retain
bilingual/bicultural staff, as well as how collaborations are built between child welfare and
ethnically diverse communities. Most importantly, research is needed to systematically evaluate the
impact ofcultural competence on practices and outcomes for children and families ofcolor.
Rural Child Welfare Practices
Much of the research literature on the effectiveness of child welfare practices is conducted
in urban settings. Researchers have suggested that urban child welfare research is often not relevant
to rural communities who may face uniquely different challenges than their urban counterparts
(Templeman & Mitchell, 2002). Indeed several studies have suggested that rural communities are
particularly vulnerable to a number of different risk factors. Rural communities in comparison to
urban communities have an elevated rate of poverty, there are fewer employment opportunities in
rural communities, and rural communities have been found to have lower levels of social well-being
than urban communities (Menanteau-Horta, & Yigzaw, 2002).
In addition to community differences, the characteristics of urban and rural child welfare
workers also differ. For instance, Landsman (2002) surveyed child welfare workers in 115 rural and
urban communities in Missouri and found several differences between the two groups. Findings
indicated that agency size played an important role in practice, many of the differences between
rural and urban child welfare practice were also noted between smaller and larger agencies. Rural
and small agencies possessed workers with a greater number of years spent in their current position
compared to urban and large agencies, they also experienced greater autonomy, decision-making,
growth opportunities, and agency supportiveness, as well as more job satisfaction and
organizational commitment. In contrast, urban and large agencies had more workers with a social
work degree, more work overload, degree of specialization and adequacy of resources, they also had
higher community stress and availability ofjobs. These findings suggest that child welfare workers
in rural communities are generalist practioners who may be more satisfied with their job than their
urban counterparts, while urban workers tend to have more resources and a higher degree of
specialization.
Other studies have suggested that rural child welfare agencies provide more services and
intervene more often than urban child welfare systems. For instance, research suggests that rural
child welfare agencies have a higher likelihood of offering preventive services to children and
families that do urban agencies (Drake, 1996); they tend to have higher levels of substantiation of
maltreatment (English et al., 2002); and rural child welfare agencies are more likely to intervene in
cases involving drug exposed infants (Ondersma et al., 2001 ).
The differences between urban and rural communities and child welfare agencies have
important implications for child welfare practice. Templeman and Mitchell (2002) have noted that
the application of urban child welfare practices to rural communities is inappropriate. In their
qualitative study of child welfare workers, they report the need for rural specific child welfare
practices. For instance, participants noted that families in rural communities are often
geographically isolated, they often have incomplete knowledge of available resources, lack of
43
access to specialized services, and restricted mobility. Thus, traditional child welfare services that
call on families to travel to and from a variety of specialized interventions are not generally feasible
with this population. However respondents also noted that rural communities possess certain assets
that should be utilized for more effective rural child welfare practice. Most importantly, rural
communities were seen by social workers as having a strong network of informal and at times,
formal supports, including linkages between neighborhoods, and the use of faith based
organizations. More effective child welfare practice for rural communities was described as using
these natural supports to build better collaboration between child welfare agencies and rural
communities. This collaboration was seen as a mechanism through which child welfare could
perform outreach to the community.
Rural Child Welfare Practices:
Impact on Children and Families of Color
The impact ofrural child welfare practices on children and families ofcolor is not
sufficiently addressed in the research literature. In general, rural communities have an increased
rate of poverty, fewer employment opportunities and lower levels of social well-being than do urban
communities (Menanteau-Horta, & Yigzaw, 2002). Differences between child welfare workers in
rural and urban areas have also been found (Landsman, 2002), and there is some evidence to
suggest that rural child welfare agencies intervene at a higher rate than do urban child welfare
agencies (Drake, 1996; English et al., 2002; Ondersma et al., 2001). These differences between
urban and rural communities and child welfare agencies suggest that applying an urban model of
child welfare practice to rural communities is not justified (Templeman & Mitchell, 2002). The
unique needs of these communities and their child welfare populations call for increased research
efforts to identify effective practices for children and families of color in rural areas.
The Role of the Court System in Cbild·Welfare Practice
For children who become dependents of the court, the judicial system may significantly
impact practices and outcomes. In child maltreatment proceedings, judges typically have authority
over decisions about services and certain case outcomes. Indeed, studies have noted that
noncompliance with court orders is a strong predictor of termination of parental rights (Atkinson &
Butler, 1996; Jellinek et al., 1992; Smith, 2003). Other studies report that the courts can have an
important influence over the services to which ch-ildren are referred (Garland & Besinger, 1997).
As such, the ways in which court systems and child welfare systems interact is an important
contextual issue affecting practices and outcomes. In general there is little research on the ways in
which child welfare and court personnel collaborate and how this collaboration, or lack of
collaboration impacts children and families of color. However the few studies that have addressed
this issue report that the two professions experience numerous barriers to effective collaboration.
·For instance, in Russell's (1988) survey of 98 caseworkers and 69 attorneys, significant differences
in role perception were found between the two groups. Child welfare workers and attorneys
disagreed about who should perform a number of different tasks. A majority of both groups felt that
they should have primary responsibility for the same tasks, including requesting authorization for a
dependency hearing, deciding on the allegations, explaining and interpreting court proceedings to
parents, deciding if children should testify, entering into agreements with parents regarding case
disposition, and recommending dispositions to the court. Both groups expressed a need to have
attorney and child welfare worker roles clarified and a need for cross training between the two
44
groups. Additionally, in open-ended responses to survey questions, both attorney and child welfare
respondents expressed some irritation over their interactions with one another. Attorneys felt that
child welfare workers should "stop trying to be attorneys" (Russell, 1988, p. 212), and child welfare
workers expressed that "we know about what lawyers do, but they don't know about what we do.
Therefore training of lawyers to casework responsibilities would be helpful" (Russell, 1988, p. 212).
In an attempt to address barriers to collaboration between child welfare and the court
system, Johnson and Cahan (1992) reported on the "Children Can't Wait Project," a training
program aimed at cross training for attorneys and social workers, and overall system modifications.
The program convened a series of seminars for child welfare workers and attorneys. Seminars
included cross training and open explorations of one another's expectations. Social workers in the
seminars expressed a need for attorneys and court personnel to increase respect and understanding
of their profession, while court personnel described a need for social workers to provide more
timely and accurate case preparation and to be knowledgeable of legal definitions. Other seminar
activities included creating action plans to expedite court proceedings by improving communication
between attorneys and workers, and modifying court proceedings themselves. Although both
attorneys and social workers had good attendance at the initial seminars of the project, attorneys had
much less involvement in multidisciplinary groups working to implement the project. Outcomes of
these collaborative efforts were not described.
Additionally, Knepper and Barton (1997) explored courtroom dynamics of court personnel
and child welfare professionals during court proceedings. District judges, circuit judges, circuit
clerks, Foster Care Review Board chairs, Court Appointed Special Advocates, and attorneys from
five sites in one state completed surveys. Interviews with court personnel and caseworkers were
also conducted, as well as focus groups with foster parents and caseworkers, court observations, and
case record reviews. Findings revealed that in general, caseworkers felt that judges respected their
opinions and judgments about cases. Yet judges also demonstrated great influence on certain
decisions and actions. When caseworkers adhered to group norms, judges often rewarded them by
not ordering "unrealistic" practices (Knepper & Barton, 1997, p. 298), and providing more
reasonable time frames for mental health evaluations to be completed. Many respondents also noted
that court proceedings were often used to finalize agreements, rather than negotiate; they lasted an
average of 10 minutes with prosecutors and other attorneys doing most of the talking. Survey results
indicated that court personnel generally do not receive training on child maltreatment issues prior to
beginning their job. Approximately 77 percent of district judges and 79 percent of circuit judges
reported that there were no training requirements for attorneys working in child welfare.
The Role of the Court System in Child Welfare Practice:
Impact on Children and Families of Color
Few studies have evaluated the potential impact of the court system on children andfami/ies
ofcolor in the child welfare system. However the court system clearly does play a significant role in
case processes and outcomes (Garland & Besinger 1997; Knepper & Barton, 1997). For instance,
Garland and Besinger ( 1997) noted that White children in the child welfare system tended to be
more likely than African American or Hispanic children to be court ordered
counseling/psychotherapy, however no racial/ethnic differences were found in social worker
recommendations in counseling/psychotherapy.
45
Research also indicates that there are numerous barriers to effective collaboration between
court personnel and child welfare professionals, including lack ofclarity in their different roles
(Russell, 1988). Studies suggest that collaboration between these two groups might be improved
through cross training and clarification of roles (Cahan, 1992; Russell, 1988). More research is
needed to discern how the court system impacts child welfare practices as well as outcomes for
children and families of color.
Summary of Key Findings and Implications
This section reviewed research on evidence-based child welfare practices and their
application to children and families of color, as well as the impact of contextual factors on child
welfare practices for children and families of color. Overall, research on the effectiveness ofchild
welfare practices across the various choice points in the CWS is somewhat limited and
unfortunately many studies fail to assess the possible differential impact ofchild welfare practices
on diverse ethnic groups. Nevertheless, certain practices evaluated in the literature are emerging as
promising practices for children and families of color and there is a need to continue to promote
evidence-based practice in child welfare.
Following the pathway that children may take through the child welfare system, research on
child welfare practices at the beginning choice points. including screening, investigation and
provision ofvoluntary services reveals mixed results on the effectiveness of these practices.
Research does indicate that cases involving children of color are investigated and substantiated at a
higher rate than cases involving White children (Ards, et al., 2003; Drake, 1996; Eckenrode et al.,
1988; Wells, et al., 1995). Children of color are also more likely than White children to be placed in
out-of-home care (Needell et al., 2003; U.S. DHHS, 2001).
Studies also demonstrate that there is little overall consistency in the decision-making
practices of individual child welfare workers (Rossi et al., 1999; Schuerman et al., 1999). Practices
designed to improve decision-making including the use of structured risk assessment tools, or
multidisciplinary teams/family group conferencing may have the potential to improve decisionmaking practices, yet research has not yet evaluated the effectiveness of these interventions with
ethnically diverse populations.
Perhaps the most promising beginning choice point practice to emerge from the literature is
the use of home-based prevention services. Research does demonstrate that home-based prevention
efforts are effective in preventing maltreatment and further involvement in the CWS, as well as
improving health related behaviors and outcomes for mothers and young children (Kitzman et al.,
1997; Olds et al., 1997).
Research on continuing care and ending choice point practices provide limited evidence
that these practices are successful in reunifyingfamilies or ensuring a permanent placement/or
children. However, research does consistently demonstrate that Black children are reunified at a
lower rate than are White children (Courtney, 1994; Jones, 1998; McMurty & Lie, 1992; Wells &
Guo, 1999; Westat, Inc., 2001c). Evaluations oftraditionalfamily reunification services, and
specific treatments, such as parenting education, substance abuse. domestic violence or Family to
Family/social network interventions appear to suggest that these services have limited overall
success, and tend to be most successful with lower-risk parents (Dore & Lee, 1999; Morre! et al.,
46
2003). However, research does suggest that substance abuse services for parenting women in
particular, are most effective when they are residential, fairly intensive and allow mothers to live
with their children while in treatment (Clark, 2001; Haller et al., 1997; Metsch et al., 2001 ).
Research on mental health services for children in out-of-home care suggests that children
ofcolor are less likely to be recommended for and to use child welfare services while in out-ofhome care (Garland & Besinger, 1997; Garland et al., 2000). Studies on intensive interventions for
children in out-of-home care, including wraparound services and treatment foster care, have yielded
mixed results (Clarke et al., 1992: Farmer et al., 2003). Additionally, although research does
indicate that foster and kinship alternate caregivers are in need of a variety of services, research on
the effectiveness of these services is not available (Cuddeback & Orme, 2002). Similarly, research
on permanency planning practices is limited. There is evidence that children of color are more
likely than White children to have lengthier stays in the CWS once parental rights have been
terminated (Barth et al., 1994; Kapp et al., 2001). Research also suggests that some child welfare
workers may fail to encourage transracial adoption or placement (Avery, 2000; Brooks & James,
2003).
Additionally, contextual factors such as cultural competency of child welfare workers and
practices; child welfare practice in rural areas; and the role of the court system all impact the
effectiveness of child welfare practices. Cultural competence in child welfare settings, although
widely acknowledged as important, has not been adequately researched. Studies do suggest that the
lack of linguistically appropriate services may result in the under-utilization of services, or
inaccurate or incomplete assessments and interventions (Gomez et al., 1985; Suleiman, 2003).
Moreover, research also suggests a need to differentiate between rural and urban areas when
researching and discussing child welfare practices (Templeman & Mitchell, 2002). Rural areas in
comparison to urban areas have been found to have unique risk factors and child welfare agencies
in rural and urban areas have been found to differ in a number ofways (Drake, 1996; English et
al., 2002; Menanteau-Horta & Yigzaw, 2002; Ondersma et al., 2001). As such, there is a need to
define and evaluate rural-specific child welfare practices (Templeman & Mitchell, 2002).
Although more research is needed to discern how the court system affects child welfare
practice, research does suggest thaf there are numerous barriers to effective collaboration between
court and child welfare systems. Cross-training and clarification of roles have been identified as
important factors in improving the relationship between the child welfare and court systems (Ca~an,
1992; Russell, 1988).
In general, much of the research literature on the impact of child welfare practices is
inconclusive. Yet very few studies describe differences in outcomes by ethnic/racial groups and
most of the research fails to even mention the effect of culture and ethnicity on the delivery and
impact of child welfare practices. This gap in the research literature is striking and has important
implications for children and families of color in the CWS. The lack of adequate knowledge of the
types of practices, services, programs, policies, tools, or strategies that are effective for children and
families of color hinders our ability to remedy the disproportionate representation of children of
color in the CWS and may leave many children and families of color without adequate or
appropriate services. The next section focuses on child welfare practices in Santa Clara County's
child welfare system.
47
IV. SANTA CLARA COUNTY CHILD WELFARE PRACTICES
In Phase 3, one area of investigation focused on the impact of child welfare practices on
children and families of color in the Santa Clara County's Department of Family and Children
Services (DFCS). Several related analyses were conducted to understand the types of practices and
case characteristics that are associated with more successful outcomes for children and families
from various racial/ethnic groups; the types of child welfare practices used in Santa Clara County;
how practices in the Main County Offices differ from South County Office practices; how practices
interact with child, family and cultural characteristics of differing racial/ethnic groups to produce
various outcomes; and the role of the court system on child welfare practices. As such, this section
contains five overall analyses:
I) A successful case analysis. This analysis examined case record review data collected in
Phase 2 to determine the types of practices and case characteristics associated with
successful case outcomes.
2) A qualitative analysis of Santa Clara County child welfare practices. Focus group and
interview data from Santa Clara County workers were collected to obtain information
about current practices, how these specific practices might positively or negatively
impact children and families of color, and what practices might be created, enhanced or
maintained to better serve children and families of color.
3) A comparison of Main County Offices versus South County Offices. This analysis used
both quantitative data from CWS/CMS, as well as qualitative data from focus groups in
South County to compare case characteristics and practices between the two areas of the
county.
4) A multivariate analysis of factors associated with reunification. This analysis
examined system-related factors and case characteristics predictive of reunification of
children with their families.
5) An analysis of the role of the court system in child welfare practice. This analysis
used case record review data collected in Phase 3 to determine whether and what type of
court-ordered changes are made to the social worker case plan recommendations at the
jurisdictional/dispositional hearing. Results from agency-wide focus groups related to
the relationship between the child welfare and court systems are also provided.
Taken together, these five sections provide extensive and important information on Santa Clara
County child welfare practices and factors that are associated with successful outcomes for children
and families of color.
48
Successful Cases Analysis
In Phase 3, one aim was to define and assess successful cases and outcomes in Santa Clara
County's Child Welfare System. An analysis of factors associated with successful case outcomes
and how these factors and outcomes differ by racial/ethnic group helps shed light on the types of
practices and conditi~ns necessary to promote enhanced outcomes for children and families of
color.
Methods
First, our task was to operationally define "successful cases and outcomes" for this
quantitative portion, which we based on the United States Department of Health and Human
Services (US DHHS) Children's Bureau standardized outcomes. These standardized outcomes are
part of the U.S. DHHS' child and family services reviews and the Title IV-E foster care eligibility
reviews (U.S. DHHS, 2003b). These levels of success and outcomes are:
Safety
• Children are first and foremost protected from abuse and neglect
• Children are safely maintained in their homes whenever possible
Permanency
• Children have permanency and stability in their living situations
• The continuity of family relationships and connections is preserved for families
Family and Child Well-Being
• Families have enhanced capacity to provide for their children's needs
• Children receive appropriate services to meet their educational needs
• Children receive adequate services to meet their physical and mental health needs
Second, from our in-depth reviews of the 403 cases sampled for Phase 2, we matched basic
information and additional data collected from the court reports during Phase 3 with the US DHHS
guidelines.· Although information from every court hearing available within our sample was
recorded, we relied primarily on data from the jurisdictional/dispositional hearing and the last two
hearings of each case to evaluate "successful cases and outcomes." We then expanded our analyses
started in Phase 2 to evaluate "successful cases and outcomes" and differences by ethnicity for this
current phase.
Results
Regarding safety, referring to the US DHHS criteria mentioned above, we used three
variables to indicate "children are first and foremost protected from abuse and neglect." Number of
prior referrals and previous number of times in the CWS indicated a risk of harm affecting these
children. The number of times removed from the family indicated a family environment that was
assessed as harmful and thus warranted additional removal.
49
We also used the initial case assignment from the jurisdictional/dispositional hearing to
denote whether "children are safely maintained in their homes whenever possible." A similar
variable, case closure type, which also refers to maintenance of the child in the home, will be
discussed in the permanency section.
Number of Prior Referrals
Of the 403 cases, 48 had missing data. Of the remaining 355 cases, 118 (33.2%) had no
prior referrals, 111 (31.2%) had 1 or 2 prior referrals, and the remaining cases (35.6%) had more
than 2 prior referrals. This indicated that more than half of the cases had prior referrals, which are
used by social workers in establishing potential risk to child safety. One-third of the cases had more
than two prior referrals. Please see Table la.
. Re erralsa
Table 1a: Frequency of Pnor
Number of Prior
Frequency
Referrals
0
118
I
75
2
36
3
35
4
24
5
18
14
6
7
6
8
7
9
4
3
10
11
5
12
3
5
13
14
1
15
0
16
0
17
0
18
1
a Based on 355 cases with valid information
Percentage
33.2%
21.1%
10.1%
9.9%
6.8%
5.1%
3.9%
1.7%
2.0%
1.1%
.8%
1.4%
.8%
1.4%
0.3%
0.0%
0.0%
0.0%
0.3%
There was not a statistically significant relationship between number of prior referrals and
ethnicity. An average of2.46 prior referrals (sd = 3.14) were made before the case was opened for
the current episode. Please see Table 1b.
50
T able lb N umberof Pnor
. Re£erra ls B~Y Ethme
. Groupa
Total
African
White
American
Prior Referrals
Avg.
2.46
2.14
2.64
(sd)
(3.14)
(2.71)
(3.27)
Latino
Asian
American/PI
Other
2.46
(3.25)
1.62
(2.14)
2.81
(3.37)
a Based on 355 cases with valid information
Previous Times in the CWS
Of the 403 cases, 93 had missing data. Of the remaining 310 cases, 204 (65.8%) had no
prior experience with the CWS while 106 (34.2%) did have prior experience with the CWS. This
indicates that one-third of cases had an elevated risk for abuse or neglect as evidenced by prior
substantiated abuse or neglect. Please see Table 2a.
Table 2a: Previous Times in the CWS 8
Previous Times
Frequency
In the CWS
0
204
85
1
2
10
3
6
4
4
5
0
1
6
a
Based on 310 cases with valid information
Percentage
65.8%
27.4%
3.2%
1.9%
1.3%
0.0%
0.3%
There was not a significant relationship between the number of previous times in the CWS
and ethnicity. On average, children across ethnic groups were previously in the CWS 0.47 times (sd
= 0.83). Please see Table 2b.
Table 2b: Previous Number of Times in CWS by Ethnicity a
White
Latino
Total
African
American
Average
(sd)
a
0.47
(0.83)
0.25
(0.49)
0.50
(0.92)
0.46
(0.78)
Asian
American/Pl
Other
0.63
(1.06)
0.56
(0.82)
Based on 310 cases with valid information
Number of Times Removed from Family in Current Episode
Ofthe 403 cases, 15 had missing data. Of the remaining 388 cases, 257 (66.2%) were
removed once and 131 (33.8%) were removed more than once, indicating that in one-third of the
cases, the safety level in the home regarding the protection from abuse and neglect was not
consistent. Please see Table 3a.
51
·1 .m Current ,p1so
Table 3a: T"1mes Removed fr om Fam11y
Number of
Frequency
Percentage
Times Removed
1
257
66.2%
2
98
25.3%
3
24
6.2%
4
7
1.8%
5
2
0.5%
a Based on 388 cases with valid infonnation
There was a significant relationship between the number of times removed from family in
current episode and ethnicity. Across all ethnic groups, children were removed an average of 1.45
times (sd = 0.74). African American children had the highest average number of times removed
from family in current episode with 1.67 times (sd = 0.90), and Asian American/Pacific Islander
children had the lowest average with 1.26 times (sd = 0.51). Please see Table 3b.
Table 3b: Number of Times Removed from Family in Current Episode bv Ethnicitva
White
Total
African
Latino
Asian
American
American/PI
Other
Average
(sd)
1.64
(0.92)
1.45
(0.74)
1.67
(0.90)
1.36
(0.68)
1.44
(0.68)
1.26
(0.51)
a Based on 388 cases with valid infonnation
Case Assignment After Jurisdictional/Dispositional Hearing
Of the 403 cases, 12 had missing data. Of the remaining 391 valid cases, 88 (22.5%) were
assigned to family maintenance. This indicates that less than one-quarter of cases are safely
maintained in the home whenever possible. There was not a significant relationship between case
assignment after the jurisdictional/dispositional hearing and ethnicity. Overall, most cases were
assigned to family reunification (71.1 %, 278 of 391 cases). Please see Table 4.
• a
Table 4: Case Assijl;nment aft er Jurisd1ctiona 1spos1t1ona Hearmg
White
Latino
Total
African
American
Family
34
27
88
7
(22.5%)
Maintenance
(14.9%)
(25.8%)
(19.4%)
Asian
American/PI
9
(29.0%)
Other
11
(26.2%)
Family
Reunification
278
(71.1 %)
38
(80.9%)
91
(68.9%)
101
(72.7%)
19
(61.3%)
29
(69.0%)
Other
25
(6.4%)
2
(4.3%)
7
(5.3%)
11
3
(9.7%)
2
(4.8%)
a Based on 391 cases with valid infonnation
(7.9%)
52
Permanency
Regarding permanency, referring to the US DHHS criteria mentioned above, we used three
·variables to indicate whether "children have permanency and stability in their living situations."
The number of unique placement homes and the number of placements in the current episode
indicated the consistency of the children's living situation while in the CWS. Since permanency and
stability also imply a solution of consistency within particular settings, case closure type was also
examined. To evaluate the outcome of whether "the continuity of family relationships and
connections is preserved for families" we also examined the proportion of reunified families within
the case closure type.
Number of Unique Placement Homes in Current Episode
Of the 403 cases, 15 had missing data. Of the remaining 388 valid cases, 162 (41.8%) had
only one or two unique placement homes. 224 (58.2%) had more than two unique homes, which
indicated that less than half of the children in the CWS are in a relatively permanent living situation.
Please see Table 5a.
T a bl e 5a: N um berofU ntQUe
.
Placement H omesa
Percentage
Number of
Frequency
Placement Homes
55
14.2%
1
2
27.6%
107
16.0%
62
3
11.9%
46
4
7.0%
5
27
6.2%
24
6
2.1%
8
7
2.8%
8
11
1.8%
9
7
2.3%
10
9
0.5%
11
2
1.0%
12
4
2.1%
13
8
0.8%
14
3
0.5%
15
2
0.0%
16
0
0.8%
17
3
0.3%
18
I
0.3%
1
19
0.3%
20
1
0.0%
21
0
22
0.3%
1
0.3%
23
1
0.0%
24
0
0.3%
25
1
53
26
27
28
29
30
31
32
33
34
35
0
2
0
0
0
0
0
1
0
1
a Based on 388 cases with valid information
0.0%
0.5%
0.0%
0.0%
0.0%
0.0%
0.0%
0.3%
0.0%
0.3%
There was not a significant relationship between number of unique placement homes in
current episode and ethnicity. Across all ethnic groups, there was an average of 4.54 (sd = 4.67)
unique placement homes in the child's current episode. Please see Table 5b.
. a
Table5 b: Number of Umque Placement Homes biy Ethn'ICity'
White
Total
African
Latino
American
Average
(sd)
4.54
(4.67)
5.35
(6.48)
4.70
(4.60)
4.44
(4.26)
Asian
American/Pl
Other
3.29
(3.22)
4.36
(4.59)
a Based on 388 cases with valid information
Number of Placements in Current Episode
Of the 403 cases, 16 had missing data. Of the remaining 387 valid cases, 222 (57.4%) had
only one or two placements. However, 165 cases had more than two placements indicating that
42.6% of children in CWS are in a relatively inconsistent living situation. Please see Table 6a.
.
Placement Homesa
Table 6a: Num bero f Umque
Percentage
Frequency
Number of
Placements
20.4%
1
79
37.0%
143
2
14.5%
56
3
10.9%
42
4
5.2%
20
5
2.8%
11
6
2.1%
7
8
0.8%
8
3
2.1%
8
9
1.0%
4
10
0.3%
11
1
0.0%
12
0
54
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
a
4
2
1
0
0
1
0
1
1
0
1
0
0
0
0
0
0
0
0
0
1
Based on 387 cases with valid information
1.0%
0.5%
0.3%
0.0%
0.0%
0.3%
0.0%
0.3%
0.3%
0.0%
0.3%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.3%
There was not a significant relationship between the number of placements in current
episode and ethnicity. Across ethnic groups the average number of placements in current episode
was 3.31 (sd = 3.34). Please see Table 6b.
Table 6b: Number of Placements in Current Episod e By Eth nicitya
White
Latino
African
Total
American
Average
(sd)
a Based
3.31
(3.34)
3.27
(3.23)
3.77
(4.06)
2.99
(2.78)
Asian
American/PI
Other
3.00
(3.16)
3.20
(2.83)
on 387 cases with valid information
Case Closure Type
Of the 403 cases, 386 had valid information for the case closure type. Including those
adopted, reunified with family, or placed with a relative or guardian, 183 (47.4%) of the cases have
permanency and stability in their living situation. There was not a significant relationship between
case closure type and ethnicity. Please see Table 7.
55
.. a
T a ble 7 Case Closure Type By Ethmc1ty'
Total
African
American
Family
Stabilized
(FM)
White
Latino
Asian
American/PI
Other
98
(25.4%)
13
(27.1%)
40
(30.5%)
31
(23.5%)
5
(16.1%)
9
(20.5%)
44
(11.4%)
5
(10.4%)
(9.9%)
13
21
(15.9%)
2
(6.5%)
3
(6.8%)
Reunified
with parent or
guardian,
court or noncourt specified
25
(6.5%)
3
(6.3%)
11
(8.4%)
6
(4.5%)
0
(0.0%)
5
(11.4%)
Adoption
16
(4.1%)
4
(8.3%)
6
(4.6%)
2
(1.5%)
2
(6.5%)
2
(4.5%)
Emancipation
57
(14.8%)
6
(12.5%)
11
(8.4%)
25
(18.9%)
7
(22.6%)
8
(18.2%)
27
(7.0%)
6
(12.5%)
7
(5.3%)
10
(7.6%)
(3.2%)
I
3
(6.8%)
119
(30.8%)
11
(22.9%)
43
(32.8%)
37
(28.0%)
14
(45.2%)
14
(31.8%)
Guardianship
established or
placement
with relative
Incarceration,
runaway or
medical
services
Court ordered
termination
and other
a Based
on 386 cases with valid information
Reunification
Of the 403 cases, 386 had valid information to determine whether children were reunified
with their family at case closure. Of these 386 cases, 263 (68.1 %) were not reunified with family
56
and 123 (31.9%) were. This indicates the continuity of family relationships and connections are
preserved for less than one-third of these families. Please see Table 8a.
. h F amllY
·1 a
T able 8a: Propo rf10nof Ch"ld
1 ren R eum"fl1ed wit
Frequency
Percentage
Not Reunified
with Family
263
68.1%
Reunified with
Family
123
31.9%
a Based on 386 cases with valid information
There was not a significant relationship between reunification with family and ethnicity.
Please see Table 8b. However, it is important to note that in the multivariate analyses that appears
later in this report (please see page 103), significant variations in reunification by ethnicity were
found even after controlling for other demographic and system-related factors.
• •
a
T a ble 8b Ch"ld
1 ren Reum"fi1ed By Ethmc1ty·
Total
African
American
White
Latino
Asian
American/Pl
Other
Not Reunified
with Family
263
(68.1%)
32
(66.7%)
80
(61.1 %)
95
(72.0%)
26
(83.9%)
30
(68.2%)
Reunified
with Family
123
(31.9%)
16
(33.3%)
51
(38.9%)
37
(28.0%)
5
(16.1%)
14
(31.8%)
a Based on
386 cases with valid information
Family and Child Well-Being
Regardingfamily and child well-being, referring to the US DHHS criteria mentioned above,
we created a summary score based on 16 variables categorized in 4 domains ( overall well-being,
economic status, health status, and substance abuse related to the mother and the father) evaluated
from the last two hearings to indicate if "families have enhanced capacity to provide for their
children's needs." Although we originally wanted to examine each domain separately, the
extremely sparse information available from the last two hearings of each case prohibited these
comparisons at such a detailed level.
We then evaluated whether the child's education status was positive from the last two
hearings to indicate if "children receive appropriate services to meet their educational needs.
57
Finally, we created a summary score to evaluate the status of the child. This was based on 8
variables within similar domains above (child's overall well-being, economic status, health status,
and substance abuse) to indicate whether "children receive adequate services to meet their physical
and mental health needs."
Families Enhanced Capacity to Provide For Their Children's Needs
Of the 403 cases, 156 had valid information to determine if families had an enhanced
capability to provide for their children's needs. A positive score on this quantified information
meant that this capacity was generally enhanced, and a negative score meant the capacity was
generally diminished. The possible range for this summary score was -2.0 to +1.0. The average
score was 0.01 (sd = 0.71), which indicated that on the whole, case outcomes seemed neutral in
terms of family's capacities to provide for their children's needs. There was a marginally significant
relationship with ethnicity. African American families appear to have a more enhanced capacity
while Asian American/Pacific Islander families seem to have a more diminished capacity. Please
see Table 9.
Table 9: Family's Capacity to Provide for Their Children's Needs By Ethnicitya
Total
African
White
Latino
Asian
American
American/Pl
Other·
Average
(sd)
0.07
(0.62)
0.01
(0.71)
0.35
(0.59)
0.00
(0.77)
-0.04
(0.69)
-0.50
(0.53)
a Based on 156 cases with valid information
Children Receiving Appropriate Educational Services
Of the 403 cases, 76 had valid information to determine if children received appropriate
services to meet their educational needs. A positive score meant that their educational needs were
generally met, and a negative score meant that their educational needs were generally unmet. The
possible range for this summary score was -1.0 to +1.0. The average score was 0.30 (sd = 0.71)
which indicated that educational needs were being somewhat met. There was not a significant
relationship with ethnicity. Please see Table 10.
Table 10: Children Receivin~ Appropriate Services to Meet Educational Needs By Ethnicitya
White
Asian
Other
Total
African
Latino
American/PI
American
Average
(sd)
a Based
0.30
(0.92)
0.75
(0.71)
on 76 cases with valid information
0.32
(0.89)
0.03
(0.98)
0.50
(0.93)
1.00
(0.00)
58
Children Receiving Adequate Services to Meet Physical and Mental Health Needs
Of the 403 cases, 156 had valid information to determine if children received adequate
services to meet their physical and mental health needs. A positive score in the summary variable
meant that their needs were generally met and a negative score that their needs were generally
unmet. The possible range for this score was -1.0 to +1.0. The average score was 0.11 (sd = 0.71)
indicating that on the whole, the case outcomes seemed neutral in regards to meeting children's
needs. There was not a significant relationship with ethnicity. Please see Table 11.
s
T a ble 11 Ch"ld
1 ren R ece1vmii Ad equate erv1ces t o Meet Needsa
White
Total
African
Latino
American
Average
(sd)
0.11
(0.71)
-0.05
(0.76)
0.21
(0.69)
0.07
(0.71)
Asian
American/PI
Other
0.13
(0.64)
0.07
(0.83)
a Based on 156 cases with valid information
Summary of Key Findings and Implications
In Phase 2, we analyzed major case characteristics on the main sample of 1720 cases
(primarily from CWS/CMS data) and additional psychosocial factors measuring child and family
well-being for the 403 cases relying on data gathered directly from the case files. We conducted
ethnic comparisons for both samples. In Phase 3, our aim was to evaluate successful cases and
outcomes using the US DHHS guidelines and assess ethnic differences. Results indicate that there
were minimal ethnic differences
Variations of Safety, Permanency, and
Family and Child Well-Being by Ethnicity
In regards to safety, there was a significant relationship between the number oftimes
removed from family in current episode and ethnicity where African American children had the
highest average number of times removed from family in current episode and Asian
American/Pacific Islander children had the lowest average.
In terms offamily and child well-being, there was a marginally significant relationship
between a families capacity to provide for their children's needs and ethnicity, with African
American families appearing to have a more enhanced capacity while Asian American/Pacific
Islander families seem to have a more diminished capacity. However, from this portion of our
analysis, the general absence of significant differences by ethnic group assessed from the end of the
case implies that certain aspects leading to a successful case and outcome are not related to
ethnicity.
It is important to note that major ethnic differences do exist in Santa Clara County's CWS,
but many of these variations appear at different choice points in the system (such as psychosocial
status at the beginning of the case, and assignment to voluntary family maintenance). In addition, a
combination of factors experienced through the course of the case may be related to success or
59
positive outcomes. However, our results suggest that ethnic differences at the conclusion of the
case are not evident, if we define success according to DHHS guidelines.
Implications Related to the Disproportionate Representation of
Children of Color in the CWS
Given the relatively few ethnic differences in evaluating successful cases and outcomes,
what does this imply about ethnicity and the CWS? The absence of ethnic differences in this
segment does not negate the previous findings in Phase 1 and 2 confirming that certain ethnic
groups are over-represented (i.e., Latinos, African Americans) while others are under-represented
(i.e., Asian American/Pacific Islanders) in the county's CWS. Phase 2 results also indicate that
there are significant ethnic differences in many demographic, system-related, and psychosocial
characteristics. Yet, what does the absence of ethnic differences in successful cases and outcomes
mean in context? While results suggest there is a clear disproportion of ethnic groups in the system,
once in the system, the children and families are generally faring the same. Their relatively
homogeneous experience may be a result ofthe "one-size-fits all" limited array of services offered
Although this may seem equitable and logistically efficient, it appears ultimately ineffective in
serving a diverse group of children and families. Also, to determine better how "well" these
children are actually doing while in the system, the availability of comparative data, clear
operational definitions for "success," and more reliable and valid data are needed.
Limitations and Recommendations
The major limitations for this successful cases and outcomes component include the lack of
information available (especially at case closure), to establish an operational definition of a
"successful case and outcome," and the need for baseline and comparative data.
Data Issues
To assess better whether safety, permanency, and family and child well-being are
maintained and enhanced, child welfare records should include more complete and measurable
information. We acknowledge that child welfare case records were not originally intended for
research. Yet, if the agencies responsible for providing effective services wish to evaluate the
performance of their programs more reliably and validly, improving information collection and
tracking should be considered.
One step would be to create a case summary check-out form used at case closure or in an
exit-interview type situation. The check-out form would indicate the child and family's status in
major areas of psychosocial well-being (i.e., health, education, and finances), in addition to the
existing case outcomes. Given the amount of missing data assessing these domains describing case
and child status, and the relative scarcity of information recorded in the last two hearings
(presumably the place where outcomes of a case would be recorded) this form would enable
measurement and evaluation of "successful cases and outcomes" in a more consistent objective
manner. The amount of missing data and the manner in which information is assessed also implies
the need for more reliable and valid methods of recording of information.
Operational Definition of Successful Case and Outcome
60
Having a clear operational definition of a "successful case and outcome" is vital to program
evaluation and research. When we used the US DHHS criteria, the initial impression maybe that a
relatively low proportion of"successful cases and outcomes" seems to exist in regards to safety,
permanency, and well-being as for example only 32% of the children are reunified with their
families. However, again availability of comparative data along with the consistent use of a welldefined operational definition of "success" is needed to understand the actual success rates for
children and youth in the CWS; Two questions arise from our results and interpretation. First, what
actually is "success?" For example, adoption for one case may be a failure if a family has improved
its ability to support and nurture a child, but was denied reunification because the system still
deemed those parents unfit. Adoption for another case may be a success if the system permanently
removes a child from a persistently dangerous or unhealthy family environment. The same
difficulty in evaluation occurs when a family is reunified (presumably a success) but the child later
falls victim to reabuse (failure). Finding an agreed-upon operational definition may be first a
theoretical or philosophical question rather than a research one, but is important for the County to
consider if their programs are to determine their effectiveness
The second question concerns a judgment of Santa Clara County's CWS performance
according to the proportion of "successful cases and outcomes." Given the above limitations and
issues, Phase 3 cannot give a final answer to how well the county's CWS is serving children and
families. This brings us to our third limitation and set of recommendations, the need for baseline
and comparative data.
Baseline and Comparative Data
It was difficult to assess the successful cases and outcomes in Santa Clara County's CWS.
Are the proportions of children reunified with their families or of those ado'pted high or low? And
compared to what? Implementing better record keeping and data collection over time on key
characteristics, based on an agreed-upon operational definition would better answer the question of
effectiveness and performance. Implementing these practices across counties and across states
would also facilitate evaluation.
The recent Little Hoover Commission report (2003) urges reform in the CWS, especially for
the county and state to improve the leadership and management structure guiding the system (Little
Hoover Commission, 2003). As part of this reform, an accurate evaluation of success for children
and families involved in the CWS is needed.
61
Qualitative Analysis of Santa Clara County Child Welfare Practices
The qualitative component of this study consisted of both focus groups and interviews and
was designed to obtain insights about current child welfare practices used within the DFCS, how
these specific practices might positively or negatively impact children and families of color, and
what practices might be created, enhanced or maintained to better serve children and families of
color.
Methods
For the purposes of both the focus groups and the interviews, "practices" were defined
broadly to include programs, services, strategies, policies, or tools. Qualitative methods, including
focus groups and interviews, are ideally suited for capturing and documenting common experiences
about what "works" and what might be improved from the perspective of individuals involved in
different parts of the child welfare system, including Social Workers, supervisors and clients. These
methods may also help to illuminate quantitative findings, offer information about formal and
informal local practices, and provide a context for understanding a variety of factors that might
influence practices. The methods employed for gathering qualitative data through focus groups and
through interviews are described below.
A total of 16 structured focus group interviews were conducted during the course of this
study. Thirteen of the fucus groups were composed of Social Workers, specifically workers with
titles of Social Worker II or III. The focus groups were structured to explore practices at different
choice points in the child welfare system. Specific topics for focus groups included the following:
• Emergency Response/Dependent Intake (early choice points related to entry into the system and
initial assessment and investigation).
• Diversion (provision of voluntary family maintenance, informal supervision, early intervention,
and other efforts to divert families to services without court intervention).
• Case Assignment (early choice points related to cases that are substantiated and opened).
• Mandated Services (including court ordered parenting classes, substance abuse treatment,
domestic violence services, mental health counseling or other services).
• Time Limits (including local, state and federal time limits and the impact on practices with
children and families of color).
In order to ensure that concerns of clients were adequately represented in the project, three
focus groups were conducted with parents in the child welfare system. Individual interviews were
also conducted with supervisors with experience in each of the choice points described above.
Individual interviews were used to supplement and affirm findings from the focus groups.
Social Worker Focus Group Sampling and Recruitment
Prospective participants for the Social Worker focus groups were randomly selected from a
list of workers with title of Social Worker II or Social Worker III in different units/division. Lists of
randomly selected Social Workers were generated using SPSS (Statistical Package for Social
Sciences). This random selection was conducted "without replacement," so that if Social Workers
were randomly selected for a specific focus group topic, they would be excluded from random
samples for other focus groups. Table 12 below provides a list of the 13 Social Worker focus
62
groups, the child welfare units from which focus group participants were selected, and the number
of focus group participants.
Approximately 15 to 20 Social Workers were randomly selected to receive an invitation to
attend each of the 13 Social Worker focus groups. In recognition of possible scheduling conflicts,
emergencies, or vacations of invitees, we invited more Social Workers than were expected to attend
each focus group. Prospective focus group participants were sent an initial invitation by email two
weeks to one month in advance of the focus group. Non-responding Social Workers were sent a
second invitation to participate in the focus group. Finally, prospective focus group participants who
agreed to attend a focus group were sent a confirmation email and/or phone call two to three days
prior to the scheduled group. Each focus group was comprised of approximately 5 - 8 members,
though some were smaller. In particular, the two case assignment focus groups were small, with two
and three participants in each of the groups respectively. A total of 67 Social Workers participated
in the focus group interviews.
Table 12 Emp oyee Focus Group Compos1tion
Focus Group
Units Included in Selection Process
Diversion
ER/DI
Case Assignment
Mandated
Services
Time Limits
Court Services
• Early Intervention/Intensive Family Services/
Home Supervision
• Informal Supervision
Family Resource Centers
• Asian Pacific Family Resource Center
• Nuestra Casa Family Resource Center
• Uiirani Family Resource Center
Emergency Response
• Screening
• Emergency Response
Court Services
• Dependent Intake
Emergency Response
• Emergency Response
Court Services
• Dependent Intake
Continuing and Child Welfare Services Bureau
• Court Supervision
• Domestic Violence Unit - Court Supervision
• Continuing
• Spanish Speaking Services
• Continuing Multilingual
Adoptions Family Assessment and Development
• Adoptions and Family Drug Court
Continuing and Child Welfare Services Bureau
• Court Supervision
• Domestic Violence Unit - Court Supervision
• Continuing
• Spanish Speaking Services
• Continuing Multilingual
Adoptions Family Assessment and Development
• Pre-Adoption and Drug Court
• Family Assessment and Development
Number of
Groups
Number of
Particioants
2
14
2
5
2
10
2
11
2
14
63
South County
TOTALS
•
Familv Linkage
South County Region
• Gilroy Family Center
• Rosanna Child Welfare Unit
• Monterey Road Child Welfare Unit
• Highland Child Welfare
3
13
13
67
Social Worker Focus Group Procedures
The Social Worker focus groups were de.signed to examine practices that impact children
and families of color at different choice points or units within the child welfare system. To this end,
the focus group questions explored the context in which services are provided, specific practices
that might positively or negatively impact children and families of color, and recommendations
about practices that might be maintained, enhanced or initiated to better serve children and families
of color (please see Attachment 1 for the Social Worker Focus Group Interview Guide). Each of the
focus groups took place in a room located at a DFCS office. Ten focus groups were conducted at the
main administrative office in downtown San Jose and three were conducted at Gilroy Family
Resource Center in South County. The focus groups were approximately 1½ hours in length.
All focus group participants received and signed a consent form that explained the study and
confirmed their voluntary participation in the focus. group. All focus group participants also
completed a brief demographic questionnaire. The summary of focus group participant
demographics is provided in Attachment 2.
Family Focus Group Sampling and Recruitment
Three focus groups were conducted with families involved in the child welfare system. Two
focus groups were conducted with parents of children from populations that are over-represented in
the child welfare system in Santa Clara County, specifically Latino and African American children
and families. Each of these focus groups was organized in partnership with one of the family
resource centers and non-random, purposive sampling methods were employed. Participants for the
Latino parents' focus group were recruited among recent graduates of a monolingual Spanish
speaking parenting class at the Nuestra Casa Resource Center. Four parents agreed to participate in
the focus group and two were able to attend. Staff ofUjirani Family Resource Center recruited
participants for the African American parents focus group. Five parents and four staff attended the
focus group. Comments from parents and staff were documented separately. Finally, a focus group
of foster parents was convened to explore the needs of children of color in out of home placement.
The six focus group participants in the foster family focus group were all women; 2 participants
identified as Latina, 1 identified as Latina/Native American, 2 identified as Caucasian/European
American, and 1 identified as African American.
Family Focus Group Procedures
The three parent focus groups were approximately I hour in length. Questions explored in
these group interviews were focused on what has worked well for clients in getting services from
Child Welfare; what has not worked well and what services should be kept, improved or created to
better meet the needs of families. Please see Attachment 3 for the Family Focus Group Interview
64
Guide. Participating parents signed a consent form and each parent received a $10 gift certificate
from Target as a token of appreciation for contributing to the study. The family focus groups are
described in Table 13 below.
Ta ble 13 Famity
·1 FOCUS Group Compos1t10n
Focus Group
Sponsoring Organization
Latino/ Mono-Lingual
Spanish Speaking Parents
African American Parents
Foster Families
TOTALS
Nuestra Casa
Ujirani Family Resource
Foster Home Licensing,
Recruitment & Intake - Resource
Families Support Team
Number
of
Groups
1
Number of
Participants
1
1
5
6
3
13
2
Individual Interviews Sampling and Recruitment
Fourteen in-depth interviews, of approximately one hour in length, were conducted with
supervisors in different positions in the Department of Family and Children's Services. Prospective
supervisor interviewees were randomly selected using the same sampling procedures described
earlier. In several cases, because the total number of supervisors was small, all supervisors in some
practice areas (such as Emergency Response or Dependent Intake) were included on the list of
prospective invitees. Based on the assumption that some supervisors would decline or be unable to
complete interviews, more participants were invited than target number of interviews ( 15). A total
of23 supervisors were invited to participate in interviews and 14 completed interviews.
Individual Interview Procedures
The interviews used a critical incident approach that is designed to elicit descriptions of
specific behaviors that help formulate the critical requirements of a specific activity (Flanigan,
1954), such as effective child welfare practice with children and families of color. The critical
incident technique has been used to research decision-making in child welfare (Banach, 1998) and
is a valuable tool for obtaining information about specific behaviors and practices from
practitioners, who are in the best position to evaluate which of these practices "work" or "don't
work." Rather than gathering general opinions and hunches, this method allows interviewees to
recall specific experiences related to the provision of services and to delineate factors related to
success or failure in outcomes, such as outcomes related to child welfare intervention for children
and families of color.
Interviewees responded to three (3) open ended questions and several "probe" questions that
solicited detail about specific critical incidents in practice with children and families of color. In
general, the questions explored: 1) factors and practices associated with a case perceived by the
interviewee to represent a successful outcome with an African American, Latino or Asian
American/Pacific Islander family, 2) factors and practices associated with a case considered
unsuccessful, and 3) specific practices that result in or could potentially result in enhanced
65
outcomes for children and families of color (please see Attachment 4 for the Supervisor Individual
Interview Guide). Demographic information was collected at the close of the interview.
The supervisors interviewed for this study represented different practice areas ofDFCS
including Emergency Response (5), Court Services (3), Continuing and Child Welfare Services
Bureau (2), Adoptions Family Assessment and Development (1 ), and South County (3).
Interviewees all had extensive experience working in the Department of Social Services and
possessed an average of 17.5 years working with County DFCS (range of 5.5 years to 33 years). All
fourteen interviewees possessed an MSW degree and some also were LCSW's. Half of the
interviewees were men and half were women. Four of the interviewees were Latino/a, five were
Caucasian/Euro American, three were African American, and two were Asian American/Pacific
Islander. Ten of the key informants are fluent in speaking another language, and seven use these
language skills in their work. A summary of demographic information for the 14 fnterviewees is
provided in Table 14.
T a ble 14 Superv1sor Interv1ewee D emograp. 1cs
Bureau/Unit
Gender
n(¾)
Emergency Response= 5 (35.7%)
Court Services
=3 (21.4%)
Continuing
= 2 (14.3%)
Adoptions
= I (7.1%)
South County
= 3 (21.4%)
n(¾)
Males= 7 (50.%)
Females= 7 (50.%)
Ethnicity
n(¾)
= 3 (21.4%)
African American
Asian/Pacific Islander= 2 (14.3%)
Latino/Hispanic
=4 (28.6%)
White
= 5 (35.7%)
YearsCurrent
Position
mean
(std. dev.)
3.7
(2.4)
YearsCounty
mean
(std. dev.)
17.5
(8.9)
Qualitative Analysis Procedures
Two members of the Child Welfare Research Team, a trained focus group leader and a
research assistant, were present during each of the focus groups with the exception of three of the
smaller focus groups that were facilitated and documented by one focus group leader. Notes were
taken during both focus group and individual supervisor interviews. Audio tape recordings were
made during social worker focus groups and individual supervisor interviews to ensure accuracy of
the transcribed notes. Data for the qualitative portion of this study consist of the transcribed notes
taken during each of the focus group and individual interviews. Qualitative methods of data analysis
were employed to analyze the focus group and individual supervisor interview data. First, major
themes were summarized for each of the choice points in the social worker focus group and
supervisor interview data. Second, themes from the family focus groups were analyzed separately.
Third, a separate analysis of themes in the qualitative data from South County Offices was
conducted to provide insights and supplement quantitative analysis of possible differences between
South County Offices and Main County Offices within DFCS. Finally, common themes across all
focus groups and interviews were identified and documented. The findings of the analysis of themes
from focus groups and interviews are presented below and are followed by specific
recommendations that emerged across focus groups and interviews.
.66
Qualitative Results: Themes from Focus Groups and Interviews
Participants in focus groups and interviews were asked to describe practices that may
negatively or positively impact children and families of color in the Santa Clara County's CWS, as
well as contextual factors that may affect the delivery of these services or client outcomes. This
section describes the main themes that emerged from these focus groups and interviews. Following
the pathway that children may take through the CWS, both negative and positive practices are
described at beginning choice points as well as continuing care/ending choice points. Lastly,
important contextual factors will also be discussed in relation to the delivery and impact of these
practices. 1
Beginning Choice Points:
Practices that Negatively Impact Children and Families of Color
The main themes related to practices that negatively impact children and families of color at
beginning choice points included, I) inappropriate referrals to the CWS, 2) inconsistency in
decision-making practices, and 3) inadequate prevention of child maltreatment or entry into the
cws.
Large numbers of inappropriate referrals were noted as having a negative impact on children
and families of color. Study participants reported that, "Many mandated reporters have no clear
definition or evidence of abuse," and that low-income children and families are frequently referred
into the system because of issues more closely associated with poverty than abuse or neglect. Focus
group participants stressed the high number of referrals from schools "based on their observations
of children wearing old or dirty clothes," "because children had lice," or in one case "because a
child brought an expired can of food to a food drive, the child was thought to therefore be
neglected." Some participants commented that low-income children from wealthier school districts
were disproportionately reported to child protective services. "A child who is at a marginal
economic level will be seen differently at a school in Los Gatos as opposed to a school anywhere
else."
Although study participants stressed that inappropriate referrals are generally screened out
by the Screening Unit, repeated referrals can be damaging because they create a perception of
increased risk that may or may not be accurate. As the number of inappropriate referrals increases
for a particular family the likelihood ofDFCS opening a case may also increase. This situation was
described as having a potentially negative impact on children and families of color in particular
because they are often more likely to have socioeconomic problems. Some respondents also
suggested that children of color are more likely to be scrutinized by law enforcement. "Police are
more harsh on African American, API and Latino Families, especially if they don't speak English,"
commented another participant.
These inappropriate referrals were described as negatively impacting screening practices
because social workers often have to spend a great deal of time screening out inappropriate
1
Please note that focus group and interview participants discussed contextual factors related to the court system. These
findings are presented in the section entitled "The Role of the Court System in Child Welfare Practice" which begins on
page 109.
67
referrals. This situation then often reduces the amount of time social workers have to identify and
serve families who are in need of voluntary services.
Additionally, study participants felt that, in some cases, inconsistency in decision-making
practices during beginning choice points negatively impact children and families of color. Although
decision-making practices occur at all stages of the CWS, many participants discussed these
practices in relation to clients' entry into the CWS. Although staff generally noted the utility of
structured risk assessment tools used at the entry point into the CWS, they also stressed that these
tools are limited and that social workers need to develop recommendations using both assessment
tools and their own judgment. In relation to the latter, participants expressed concern that decisionmaking is often determined by the biases and perceptions of individual social workers. Social
workers named specific risk assessment tools that are used at beginning stages of the CWS as
useful, such as the decision tree use4 during screening and a written protocol used in Early
Intervention. At the same time, social workers suggested that application of assessment tools might
vary among social workers based on a number of factors including individual worker bias, size of
the unit and unit level, shared expectations about assessment, and level of supervisor support and
oversight.
Moreover, study participants noted that, "Decision-making all depends on the worker and
his or her attitudes and values," and "We can so easily sway the picture so it is positive or negative
by how we write up our narratives for the court reports; the report has a lot of influence and
sometimes peoples' personal issues get in the way of making informed decisions." Furthermore,
interviewees noted that social workers are not immune to stereotypes of different groups in the
dominant culture and in the media and that social workers often judge families based on their own
pers_onal experiences and perceptions about what is "good enough" for children. Some· social
workers pointed out that, for example, workers from other cultures might misinterpret
communication norms in African American families as abusive. "Sometimes African American
families have norms about speaking loudly, yelling, and threatening (like I am going to pop
you) ... but kids are rarely hurt ... an African American social worker might be less likely to
[negatively] interpret these kind of verbalizations than a social worker from a different
background."
According to study participants, decision-making practices are also influenced by
socioeconomic factors. Families experiencing unemployment, lack of access to resources such as
daycare, or problems obtaining health care for children may be initially assessed as having a higher
risk for maltreatment. For example, "older siblings are taking care of younger siblings because
parents can't afford childcare," noted one participant. "If there were more good daycare there
wouldn't be as many neglect cases, even once in the system we [workers] can't get them [clients]
child care," added another. Social workers also underscored the impact of local high housing costs
on low-income families. This problem was encapsulated by the comments of one worker: "Many
families partner up with other individuals to maintain housing and just because there are numerous
people living in the home, relatives sharing beds, doesn't mean that this is abuse or neglect. It's
poverty." The case example below typified the illustrations provided by study participants:
A Social Worker 1 was covering a case and she was concerned about the family because
there were four children living in a basement with their family. This was a Chicano family.
She reported that the children might need to be removed. I went out to this social worker
and saw completely different things. The basement was an apartment. The place was small,
68
but had couches and a crib for the baby; it was cluttered but not dirty ... Maybe another
worker might have justified what this worker saw and removed the kids for perhaps no good
reason at all.
The issue of social worker attitudes toward clients was particularly salient to family focus
group participants. Several parents commented on the notable variation in how they are treated by
different social workers. A remark that typified this theme was made by one parent who said, "some
social workers do their job because they really want to help other and some just do it for the money
and don't seem to care about others." Parents also described feeling undermined by social workers
who appeared to have a negative attitude about the client. One client summarized, "some of the
social workers are disrespectful, unfriendly, they look down on the clients and make you feel
humiliated." They elaborated, "It gets to the point where it makes the client feel like 'I don't want to
go in there,' but I have to go in there for my child and when you come out you're in tears."
Inadequate prevention of child maltreatment or entry into the CWS was also identified as
having a negative impact on children and families of color at beginning choice points. Participants
commented on the need for services in the community that would help to prevent entry into the
child welfare system. Focus group members stated that there is a need for education about child
welfare laws and definitions of child abuse, particularly for cultures in which norms related to
discipline of children differ from expectations in the dominant culture. Participants suggested that
such programs could be offered through Social Service agencies but also in community based social
institutions, such as churches, and that some of the communities that have benefited from this type
of preventive education in the past include African American, Asian American/Pacific Islander
(e.g., Vietnamese, Chinese, and Filipino), Latino, and immigrant communities.
Study participants also emphasized the problems in obtaining services without opening a
case. For example, one worker stated that "to get into a wrap-around service from an outside
agency, such as schools or probation, without the child being a dependent takes months, by this
point the problem will have escalated." Many social workers stated that opening a case had both
benefits and drawbacks: families had greater access to services and support through resources such
as Family Preservation Funds but were at greater risk for having a child removed because of time
limits or establishing a history with DFCS.
Several social workers also noted that it is actually more work for them to maintain a child
in the home with Voluntary Family Maintenance (VFM) services than it is to remove a child from
the home and refer to court intervention. Additionally, for families who do make it into VFM or
Informal Supervision (IS) services, focus group participants pointed out that resources were not
available to provide follow-up support to families to prevent re-entry into the system once their
cases are closed. "The family may still have needs but we have to close it. We can only keep it in IS
for 6 months (or 3 months in VFM); it relates to funding because you lose federal and state money
when a case is closed and you are on County money-and there is none," exemplified comments
from interviews. Consequently, "clients may end up back in services." Ideally "workers would be
able to provide some additional help to them [families] to help themselves."
Time lags between case assignment to Voluntary Family Maintenance and social workers
actually receiving the case were also noted as an issue. "By the time the social workers gets the
case, three weeks have gone by and the social worker has only 2 ½ months left [out of three months
of services under Voluntary Family Maintenance]," elaborated one social worker.
69
Beginning Choice Points:
Practices that Positively Impact Children and Families of Color
The main themes related to practices that positively impact children and families of color at
beginning choice points included, I) orientation of clients to the CWS, 2) team decision-making, 3)
family involvement in decision-making, and 4) the use of child care and in-home services.
Providing an orientation for parents to help them understand the components of the child
welfare systems, the court process, current laws and policies related to time limits, and how to work
with social workers was named as a best practice for families first entering the CWS. Social
workers in South County portrayed this as a best practice that was formalized into an orientation
group while social workers in other areas of the county described conducting orientations with
clients individually. Social workers and supervisors advocated for strengthening and
institutionalizing the practice of providing an orientation for clients throughout the county. Some
participants suggested that an orientation workshop could be provided or co-lead by someone
"outside the system" that might be perceived by clients as more neutral.
It would be helpful for parents to go to a workshop to educate them about time lines.
Parents don't get it immediately. Sometimes by the time the client "gets it, " it may too late.
Also sometimes social workers don 't explain it well. Sometimes clients are just traumatized
about what all this means and you have to walk them through the process for the first 2
months and you have 6 months time limit - it's very difficult. The workshop should also
teach the clients how to read a court report and what the recommendations mean.
Team decision-making <TDM), which is currently being initiated in Santa Clara County,
was viewed by participants as a promising tool for mediating individual bias and cultural
misunderstanding by social workers. "Even with the decision tree in place we know it hasn't been
working because there are still so many children of color and there is still a lot of the worker's
subjective perspective put into the report; hopefully TDM can reduce this." TDM was also
perceived as a vehicle to "help get everyone on the same page" with professionals in other systems,
such as schools and service providers. Staffing to discuss cases within units and with
interdisciplinary teams was identified as a positive practice for children and families of color that
could be enhanced by TDM. Some social workers expressed concern about the amount of time that
might be required by TDM's and that "it may be hard to get everyone together in a limited time."
Study participants identified family involvement in decision-making. including the use of
family conferences as a positive practice, particularly for children and families of color. Participants
suggested this was particularly important for cultural groups that place high value on extended
family. "In the Latino culture it's appropriate to extend an invitation to grandparents, godparents,
and all parental figures," illustrated one social worker. Study participants stressed the importance of
extending family involvement to include both biological and non-biological family members. For
example, some participants noted that African American families often have non-biological aunts,
uncles, or other extended family members that should be included in family conferencing. Many
social workers and supervisors viewed the developing family-to-family initiative as a promising
strategy for expanding and institutionalizing family participation in decision-making.
70
Last, the use of childcare and in-home services were also noted as a positive practice. For
instance, in one social worker's opinion, families with adequate childcare have about an 80 percent
participation rate in diversion services, compared to 50 percent participation rate for parents without
adequate childcare. Social workers, supervisors, and families discussed the importance of childcare,
in relation to the ability of families to succeed in services. "It's really helpful to have child care
through Calworks; We need to make sure that is not cut because people can't work or train if there
is no one to watch their kids -This is a huge thing."
Social workers also noted that the provision of in-home services allows for "working with
clients in their own environment" and providing opportunities for one-on-one mentoring and role
modeling new behavior. Social workers identified agencies and programs that provide in-home
parenting classes as more effective than traditional didactic intervention practices alone. Social
workers also stated that providing in-home visits and follow up services, with the support of Social
Worker l's, was an effective practice that is underemployed because of heavy caseloads and scarcity
of resources.
Beginning Choice Points:
Recommendations to Improve Practices
Recommendations to improve practices at beginning choice points in the CWS included 1)
reducing bias and increasing accountability in decision-making, 2) training professionals in other
systems, 3) providing client orientations to families first entering the CWS, and 4) investing in
prevention and early intervention. Specifically,
Reduce Bias and Increase Accountability in Decision Making
•
•
•
•
Implement team decision-making in a manner that includes consideration of client culture.
Examine ways to facilitate accountability for team decisions.
Explore mechanisms to ensure accountability in decision-making and minimize the possible
influence of individual or supervisor bias in decision-making.
Continue the practice of facilitating cultural consultations between staff, with the support of
supervisors, including consultations across units.
Continue and expand practices related to family conferencing.
Training Professionals in Other Systems
•
•
•
Provide or collaborate in providing training for mandated reporters, including schoolteachers
and health care professionals.
Provide or collaborate in cross training with other systems and agencies "so we can understand
one another's needs and views-what we need from each other." Cross training may also serve
"to increase communication" or even "help with preventing some families from coming into the
system, including families that are already being seen by the agencies."
Provide or facilitate more training for police regarding removal, working with social workers to
assess risk before removal, coordinating with social workers regarding child and family
interviews, and ensuring sufficient training for law enforcement officers assigned to conducting
sexual abuse interviews.
71
Provide Client Orientation
•
•
•
Educate parents referred into the system about child abuse and neglect laws. Provide workshops
in their own language on topics such as court processes, how to work with social workers,
timelines, and how to navigate the system.
Consider having such workshops lead or co-lead by someone outside of the system that may be
perceived by clients as more neutral.
Hire and Train Social Worker 1's to assist with family education and orientation.
Invest in Prevention and Early Intervention
•
•
•
Explore mechanisms to continue providing prevention services, including education of diverse
communities about child abuse and elder abuse.
Explore mechanisms to fund counseling and intervention services for families at risk that are
less restrictive than federally funded programs such as Family Preservation, which frequently
require formally opening a case to allow families to access resources.
Continue support and explore opportunities to expand voluntary and diversion services,
including early intervention and voluntary services at the family resource centers should
continue. These efforts are critical to "determine if families should be in the system" and ensure
opportunities for less restrictive placements for children.
Continuing Care/Ending Choice Points:
Practices that Negatively Impact Children and Families of Color
Study participants noted three overall practices that negatively impact children and families
of color at continuing care/ending choice points, including 1) problems with service plan
recommendations, 2) inconsistency in assessment of substance abusing families and need for
stronger linkages with substance abuse services, 3) barriers and service gaps related to relative and
foster placements.
Problems with service plan recommendations, including a "cookie cutter approach" to
services, and unrealistic expectations of families were noted as practices that have a negative impact
on children and families of color. For instance, several study participants suggested that using a
"cookie cutter approach" or "template for recommendations" is a practice among both social
workers and district attorneys that negatively impacts children and families of color. Participants
offered a number of explanations for this dynamic including "social workers being jaded, believing
that since they had worked with a similar family before, that the family is the same and the
interventions should be the same." Similar observations from social workers in other focus groups
suggested that "sometimes workers will get challenged if they don't take the cookie-cutter
approach," and sometimes "DA's think all parents need counseling, parenting class, etc." Several
social workers called for greater flexibility in working with families and "more creativity regarding
case plans needs. For example, some parents may just need more time to spend with their children."
Others cited examples where clients were mandated to services that were not appropriate including
a mother mandated to services before an assessment was conducted and a father "who is in a 52week batters' program-who had one incident of losing his temper and hitting his child-and
72
though hitting her was not okay he would be better served in family therapy and an advanced
parenting class."
Additionally, some service plan recommendations were also described as placing unrealistic
expectations on families that can set them up for eventual failure. These practices tend to impact
low-income families and were perceived to disproportionately impact children and families of color.
First, service plans frequently require "a laundry list of mandated services" that are unrealistic for
families. "I have a lot of clients who when I sit down with them and go over the service plan with
them for the first time they say 'I'm overwhelmed, I can't do this." "Families are in a panic to get
everything done on the service plan from ER to continuation," explained one social worker focus
group participant. Other participants noted that a family might be mandated to substance abuse
services, parenting classes,.and mental health services while seeking employment and new housing.
Some families "just give up," observed participants. Moreover, clients are often required to attend
services that conflict with their job schedules, particularly clients whose schedules change from
week to week. "The expectation for clients to reschedule their work schedule is difficult, especially
those that receive hourly wages, and causes even more stress for the family."
Many participants expressed concerns that high expectations imposed on families are
reflective of and "ideal" rather than "good enough" standard for families. One study participant
summarized this dynamic: "We are not required to have 'perfect parents,' we need minimal
standards of parenting .. .Just because you or I would not parent that way does not mean there is a
risk to the child." "From a client's side, they are being told by a court system and a social system
that the way that they were brought up and the way that things were done in their [clients'] family is
wrong and not acceptable," noted a social worker focus group participant.
Finally, study participants cited insufficient resources to support the success of families
attending mandated services as a problem. For example, participants in several focus groups stated
that the shortage of bus passes and increased paperwork required to obtain the limited passes that
were available negatively impact families, particularly children and families of color. "Most of
these clients don't have a vehicle so for the low income families it is challenging for them to get
their children back ... how are they supposed to fulfill their required programs without
transportation?"
Inconsistency in assessment of substance abusing families and need for stronger linkages
with substance abuse services was also noted as negatively impacting children and families of color
at the continuing care/ending choice point. Participants stressed the importance of effectively
addressing the needs of families with substance abuse problems who they perceived to comprise a
majority of caseloads. Study participants described a number of success stories in which families
were reunified when parents completed substance abuse treatment and also highlighted the fact that
when appropriate substance abuse services are provided they can be an effective best practice.
At the same time, study participants suggested that protocols for decision-making with
substance abusing families and formal agreements with substance abuse treatment agencies were
inadequate. Participants also espoused different standards about when it might be appropriate to
reunify a child with a parent or place a child with a relative in recovery. Furthermore, participants
suggested that there was a need for developing protocols for addressing relapse with substance
abusers in conjunction with both treatment programs and the courts. "We [social workers] pay lip
service, we say relapse is a part of recovery, but the system does not reflect that." Finally, several
73
study participants noted a lack of clarity and consistency in standards for assessing risk in the case
of drug exposed infants, suggesting that race and class bias may influence both initial referral and
assessment of parents of drug exposed infants in child welfare.
Study participants also commented on the scarcity of substance abuse services, particularly
residential programs for women and children. Although a variety of substance abuse treatment
services are offered throughout Santa Clara County, participants pointed out that the need fo·r
services was greater than availability, particularly for clients whose primary language is not
English. Several participants, particularly in South County, noted that a centralized intake process
for entree into treatment was a barrier for many families because of time and transportation issues.
Additionally, social workers felt that barriers and service gaps related to relative and foster
placements also negatively impacted children and families of color at continuing care/ending choice
point practices. While the use of relative placements was described as an effective practice for
children and families of color, study respondents proposed that poverty issu~s and timelines
converged with other factors to create substantial obstacles to relative placement. First, pressure to
move children out of the children's shelter, in the context of little staff support to locate prospective
relative placements, often results in lost opportunities to place a child with a family member. Some
social workers described feeling "in a double bind" and, in some cases, deeply frustrated because of
having to place a child in a foster care situation that was perceived as less positive for the child.
Second, even if a relative placement is identified, it is often "impossible to complete the relative
placement paperwork within 24 hours, with the required background checks and fingerprinting."
Third, although social workers can seek a Director's Exemption to place children with families
where members may have had a criminal background, many social workers expressed concern
about the fact that even a distant criminal history can create barriers to appropriate relative
placements. Finally, new standards for relative placement ''are the same as a licensed foster home
and are unrealistic for many families; this causes children to not be placed with relatives."
In general, although study participants expressed commitment to ensuring appropriate
relative placements for children, the study participants suggested that the requirements represented
by the relative packet "doesn't take into account economics or cultural practices" and
"disproportionately impacts families of color." These concerns were repeated across focus groups
and interviews and were epitomized by the following observations and examples:
The only people that canfit into it [relative packet] are those with money ... the packet
doesn't take into account the families we deal with. Many families are able to put four
children in a bedroom and they 're able to do fine, each child has their own space. It
[relative packet] does not allow for any arrangement like that at all. "
There was a case with 10 and 12 year old Hispanic/Native American girls who were
abandoned and went to live with a relative. The relative contacted CPS and stated that the
mother had not returned and wanted to know what to do with the children. The worker was
unable to assess the relative 's home immediately; therefore, the children were then placed
in a foster home. The licensed foster home was horrible, they ffoster parent] placed the
older girl on the floor on a crib mattress. It was terrible and all because they couldn't
continue to stay where they hadbeen, because I [worker] couldn't get the clearances fast
enough.
74
Three young Mexican American girls needed a place to live. The aunt was willing to take
them in and the children were comfortable with her. The uncle had a criminal history and
was a recovering heroin addict, who had been clean for approximately 20 years. On paper
the uncle looked really bad, the worker was unable to clear his past. The children were
placed in foster care and may be adopted if the mother doesn't comply with the service plan.
In addition to barriers to relative placements, social workers pointed out a need to develop
strategies for working with foster care and adoptive families. Respondents noted that these alternate
caregivers are often deeply disappointed when the children in their care are reunified. Some social
workers suggested working more closely with fost-adopt families about the probabilities for
reunification. Foster family and social worker focus groups also expressed a need to provide
training to foster families in practical strategies for caring for children across cultural groups and a
need to develop mechanisms for experienced foster parents to mentor new foster parents.
Continuing Care/Ending Choice Points:
Practices that Positively Impact Children and Families of Color
Social workers described three overall practices that positively impact children and families
of color at the continuing care/ending choice points including 1) parenting education services that
involve both parents and children in services, 2) cultural matching for children in foster care, and 3)
recruitment, training and coaching for foster families.
Parenting education services that involve both parents and children in services were
described as practices that positively impact children and families of color. Social workers and
supervisors noted that parenting classes that involved children and provided opportunities for
parent-child interaction were also a promising practice in comparison to traditional parenting
classes. Some social workers suggested that it might be helpful to "wait until children have been
returned to order parenting classes" or to arrange for children's participation in classes.
Additionally, social workers described cultural matching for children in foster care in which
workers attempt to place children in a home with foster parents of a similar background as a
preferred practice. This was perceived to be critical in relation to matching the primary language of
the child and particularly important in the context of federal and state time lines and concurrent
planning. At the same time, both social workers and foster parents stressed the importance of a safe
and caring placement for children, whether a cultural match was possible or not. Social workers
expressed concern that the particular issues and needs of children, including but not exclusive to
culture, are sacrificed to the expedience of "finding any placement." Foster parents echoed a
concern about considering both the needs of the child and the experience of the foster parents when
arranging a "match." One foster parent caring for a special needs child of a similar background
noted that she would have been overwhelmed and tempted to abandon the idea of foster parenting if
her first foster child had been particularly challenging. Foster parents stressed that they and other
foster parents often have successful experiences with children of different backgrounds and that
they would "never want to see a child wait for a permanent home or left in a shelter because of a
cultural difference."
Social workers and foster parents suggested that recruitment, training and coaching for
foster parents is critical to working effectively with children and families of color. Specifically,
foster families need to know about how to care for children placed in their home. For example,
75
foster parents caring for African American children for the first time required basic information
about hair care. Foster parents suggested that best practices for working with children of color
included obtaining coaching and training from other foster parents, learning about the child's
culture and the individual child, and collaborating with social workers to interface with birth
parents. Study participants noted that tensions between foster families and birth families were
common but that positive practices for children required minimizing these potential problems.
Foster families described "success stories" in which foster parents maintained supportive
relationships with children who were returned to their family of origin.
Social workers and foster parents both noted the need to recruit diverse foster families,
particularly Latino and African American families in Santa Clara County. One foster mother who
identified as Latina suggested that some families in her community might be reluctant to become
foster parents because they are more likely to already have children than Anglo families. Study
participants identified outreach to prospective foster families in underrepresented communities as an
important continuing practice.
Continuing Care/Ending Choice Points:
Recommendations to Improve Practices
Recommendations to improve practices at continuing care/ending choice points included, 1)
providing or facilitating access to concrete support for families, 2) facilitating increased availability
and access to services, and 3) supporting and developing foster families. Specifically,
Provide or Facilitate Access to Concrete Sugport for Families
•
•
•
Examine options for restoring access to bus passes for clients, both in terms of increasing
number of available passes and decreasing paperwork required from social workers to obtain
passes for clients.
Explore ways to increase access to childcare for parents participating in voluntary or mandated
services.
Facilitate access to services related to transitional housing, jobs, financial assistance (e.g.,
CalWORKS), and for families as well as educational interventions for children in the child
welfare system.
Facilitate Increased Availability/Access to Services
•
•
•
Work with policy makers, courts, communities, and other service delivery systems to identify
and address issues related to insufficient services for families. Areas where services are lacking
or limited in relation to the perceived need include substance abuse treatment (particularly
residential treatment for women and children), transitional and drug free living, mental health
services for children, services for young children exposed to trauma or substance abuse, after
school programs and other resources for children.
Collaborate with other systems to protect funding of key services in the context of state budget
problems, such as services for Oto 5 programs.
Work with contract agencies to facilitate inclusion of children in parenting classes to allow for
both didactic and experiential learning opportunities.
76
•
Sustain and increase in-home services, such as intensive wrap around services and one-on-one
mentoring du.ring home visits.
Support and Develop Foster Families
•
•
•
Continue and expand efforts to recruit foster families that are representative of diverse
populations in the child welfare system and that match the languages of different children and
families (e.g., additional Spanish speaking and Vietnamese speaking foster families).
Provide training for foster families in practical strategies for caring for children across cultural
groups.
Continue to develop mechanisms for experienced foster parents to mentor new foster parents.
Contextual Factors that Negatively Impact Practices for Children and Families of Color
Study participants noted three overall contextual factors that negatively impact child welfare
practices for children and families of color 1) agency level factors that detract from social work
practice, 2) a need for more bilingual and bicultural social workers and services, and 3) time limit
constraints.
Certain agency level factors that detract from social work practice were mentioned by many
study participants as negatively impacting practices for children and families of color. For instance,
heavy caseloads and staff shortages were identified as critical factors impeding effective practices
for children and families of color. Children and families of color, particularly famili~s who may be
less trusting or more intimidated by the system, may require language specific services, or who may
have multiple vulnerabilities were deemed very likely to be impacted by the heavy caseloads of
social workers. One social worker who works primarily with Spanish speaking families observed
that many of her clients require "a lot of hand holding" in navigating through the system. Given
heavy caseloads, even dedicated social workers that invest substantial time with some families will
have other families "slip through the cracks." One participant explained:
The weight ofthe caseload is so heavy-there 's no time to do social work. I always feel I
have to rush. Every minute ofyour day is booked. There is no time to explain court
processes to clients.
Many participants commented positively on agency plans to reduce caseloads. This was
recognized as a critical move toward improving overall effectiveness in day-to-day social work
practice. Yet some social workers commented on some of the weaknesses in the current method for
determining caseloads. "There's no equity in the way the agency does the point system for
determining workers' caseloads .... it's based on a 100 point system and you could have 70 points
that could be 35 children in permanent placement, which is a lot of children to manage, whereas
someone else could also have 70 points in Family Reunification (FR) cases, which is not the same
workload." Others suggested that there appeared to be "no rhyme or reason to how cases get
assigned-workers with a lot of points will still get assigned cases-even when a different worker
has a lower number of points." Social workers noted that the point system accounts for the number
of children in a caseload and also assigns a higher number of points to FR cases than to Permanent
Placement (PP) and Family Maintenance (FM) cases. However, they point out that the point system
77
does not necessarily account for the severity of the case and for individual differences between
cases. "Even a case with low points can be very complicated and take up a lot of your time."
Social workers also highlighted the lack of adequate support staff to work effectively with
families, including help with transportation, investigating relative placements, and follow up with
clients at home as negatively impacting children and families of color. Social workers also
commented on the emotional strain on the job and risk for burnout that is exacerbated by too little
support.
We don't have enough support staff-not enough Social Worker J's. It's a lot harder to
manage a good FR case without a SW I to document things, help with transport, etc ... it
becomes more challenging.
This leads you on the road to burnout-I always want to do my best-I want to work harder
to get the parents to be successful. There are some workers who are tremendously burned
out.
Additionally, social workers noted that substantial amounts of required paperwork can
detract from building relationships with clients. Several social workers suggested that, as an
example, writing interim court reports in addition to required reports was duplicative and used time
that would be better spent with clients. Social workers also pointed out that not having easy access
to centralized and updated information about resources and services for families impedes effective
practice. The need for centralized and current resource information was noted as important to
avoiding "reinventing the wheel" in relation to researching resources that other workers already
knew about.
Moreover, confusion about the agency's overall mission, as well as confusion about key
agency and child welfare policies was also described as negatively impacting child welfare practices
for children and families of color. A need for greater clarity about the mission ofDFCS was
expressed. "Who is the client, the child or the parent?" asked one participant. Participants varied in
their perspectives about the balance between child protection and family preservation. Social
workers also expressed confusion over certain key policies including concurrent planning, home
assessment protocols and protocols for voluntary versus court placements. Social workers suggested
that clearer guidelines related to implementation of concurrent planning were needed. Additionally,
participants reported that a more detailed written protocol on home assessments would be
beneficial, ''the policy is not in writing, there is no written protocol and the most recent memo has
DRAFT written on it." Consequently, it is "difficult completing assessments and explaining the
policy to relatives, when the workers aren't clear about it themselves." Last, uncertainty about
protocols for voluntary versus court placements were noted. "There are misconceptions about
voluntarily placing a child and when that child is placed into protective custody... about whether
this protocol is required for all placements, it seems to go back and forth," commented one social
worker. Another noted "I don't have the understanding about what is a voluntary placement versus
a child taken into protective custody; it seems that ifwe get involved in any kind of removal that it
generates a protective custody."
A need for more bilingual and bicultural social workers and services was also noted as a
contextual factor that negatively impacts children and families of color. Although study participants
generally commented on the diversity ofDFCS staff as a strength, there was a perceived need for
78
continued growth in this area. "There are not enough workers who speak the different languages of
the families," noted focus group participants. Some social workers suggested that "the clients are
left in limbo when a worker who speaks their language is not found" and "when a worker is found
that does speak the language of the client then two months have passed by; due to time line
constraints the worker is forced to make decisions without knowing much about the family."
"English speaking workers assigned to Spanish-speaking clients is a problem-even if the parents
and child speak English, a lot of the extended family may not," noted one worker. Reliance on
interpreters was identified as problematic because it can be difficult to find interpreters within a
limited amount of time which can lead to an inability of clients to obtain a timely answer to a
question from a worker because it may take days to find an interpreters. Moreover, there can be
problems with interpreters who are culturally inappropriate, who do not translate appropriately, or
who fail to read all of the court report to the clients.
In spite of these limitations, additional qualified interpreters and translators are needed to
facilitate communication with clients and to provide materials in client's primary language. Study
participants stressed the need for increasing language capacity both through social work staff and
through access to qualified interpreters. Access to qualified interpreters remains critical to working
in a region with representation of a growing number of cultural and language groups. "Lack of
interpreters causes continuance of cases," noted one social worker. Study participants offered a
number of illustrations of the impact that language barriers had on clients. One example is offered
below.
There are only two Korean interpreters in the South Bay. In one case with a Koreanfamily,
Korean interpreter could not be found for a court hearing. This father had to take offfrom
work five days in a row to go to court, because each day an interpreter wasn't there. This
impacted the father because ofthe time he had to take offofwork. Each day he became
more and more frustrated By the time the judge saw the case, the judge saw him at his
most frustrated point.
In addition to the need for bilingual and bicultural social workers, study participants also
expressed a need for more culturally competent services in general. "There is a need for more
culturally diverse counselors and therapists," noted study participants. Barriers to successful
compliance with mandated services include lack of counselors who have knowledge of different
cultures and long waiting lists for the few culturally diverse counselors in the area, according to
study participants. Furthermore, social worker participants observed that White families appeared to
attach less stigma and experience less resistance to accepting mental health or counseling services
than families of color. "Lack of services designed for the needs of a specific ethnicity or family can
cause Dependent Intake to hold onto cases longer due to being unable to find the necessary
resources." Some social workers suggested that existing services for English speaking clients varied
in quality and that contract agencies also differed in the degree to which they prioritized access to
services for DFCS clients.
A shortage of services for monolingual non-English speaking clients was identified as a
substantial problem for many families. "Families stay in the system longer because of lack of
Spanish speaking and other language services ... this can be frustrating for families, they want to
give up." Study participants described struggling tenaciously to find appropriate services for
families. Examples included searching for domestic violence and substance abuse services in
Tagolog, finding a Vietnamese-speaking counselor, and locating a group home for a deaf child.
79
Several participants expressed concerns that services provided in English for clients whose primary
language is non-English were ineffectual and, on occasion, harmful to families such as when
"English-speaking counselors who work with monolingual families will use the children as
translators." Spanish speaking services were perceived to be an issue throughout the county and
were particularly described as pressing by South County study participants. During one focus group
discussion of this issue a participant provided an example:
A Hispanic monolingual mother living in Sunnyvale does not have transportation [to
mandated services]. There are no Spanish resources in her area. They are all in the San
Jose/Santa Clara area, which causes the family to commute farther. It is difficult finding
resources in her area and with no bus passes transportation is a big issue.
Study participants also described time limit constraints as having a negative impact on
children and families of color. Focus group participants noted that it was often difficult to work
toward family reunification in the context of federal and state time limits. Participants uniformly
affirmed the importance of expediting placement of a child into a permanent, safe and stable home
and to seeking the least restrictive placement options for children. At the same time, they suggested
that timelines made it difficult to have enough time to work with families that require services in
order to safely reunify. "Fifteen months is not enough time to work with the client's
issues/problems. The reality of what people are dealing with is not there in the law," was a typical
observation by social workers. Several social workers and interviewees offered examples of
children who were placed into adoption who would otherwise have been able to remain with their
families if time limits were less rigid: "Court time limits played a large role in the adoption of the
child ... the time limits affected the outcome of the family and did not take into account how difficult
it is to change the way one had lead their life in a matter of a few short months."
The issue of access to timely services in the context of time limits was also perceived as an
important issue. An interviewee described a successful case in which a Latina client with a
substance abuse problem was able to reunify largely because the social worker was able to get the
mother into substance abuse treatment despite waiting lists: "There are waiting lists for services and ·
not enough time to successfully complete treatment with the waiting lists and the time limits ...This
family was lucky; the worker was able to get the mother into treatment within 2-3 weeks, so being
at the right place at the right time accounts for this [success]." Conflicts between child welfare time
lines were perceived as particularly common in relation to addressing the needs of substance
abusing parents that represent a majority of cases. "The Federal government only gives families in
permanent planning 12 months to complete services and parents only have 6 months for children
under 3 years of age; this is especially difficult for parents who abuse substances."
Study participants suggested that time limits disproportionately impacted families of color:
"Children of color are brought into the system because parents are on drugs, they aren't being
watched and are in an unsafe environment at that point. The children stay in the system because the
expectation is for the parents to all of a sudden to get better because their children are in the system,
'so if you [parents] don't get better we [social workers] are going to keep your children. This is
unrealistic, especially with the court timelines," exemplified comments made by many study
participants. Families whose primary language is other than English were also perceived as being at
greater risk for failing to reunify: "It is difficult to meet timelines with certain services that are not
offered in different languages; clients are placed on waiting lists or those services don't exist."
80
Contextual Factors that Positively Impact Practices for Children and Families of Color
Respondents noted four overall contextual factors that positively impact child welfare
practices for children and families of color including 1) cultural matching and cultural consultation,
2) supportive supervision 3) collaboration with other agencies and systems, and 4) a strength-based
approach to services.
Cultural matching and cultural consultation was noted as a best practice for children and
families of color. Although study participants did describe a need for more bilingual and bicultural
staff, they also commended the agency's efforts to provide a cultural match between clients and
social workers. Cultural matching was viewed as an effective strategy for facilitating development
of rapport with clients and minimizing the potential for cultural bias in assessment. Specifically,
social workers and supervisors valued the culturally specific Emergency Response (ER) units and
acknowledged efforts to match clients and social workers as much possible in Dependent Intake
(DI). At the same time, respondents asserted that it is not always possible to assign to clients
workers of a similar cultural background. In this context, social workers described obtaining
cultural consultation from other child welfare professionals as a best practice.
At the same time, study participants acknowledged that clients might prefer to interface with
social workers of a different culture than their own. For example, several social workers that
specialize in working with Asian American/Pacific Islander groups commented that some families
prefer to be assigned to a social worker of a different background out of a sense of privacy, pride, or
concern about judgment. Both social worker and client focus group participants noted that cultural
matching was not a guarantee against bias. Social workers are not immune to being judgmental
about clients of their own culture or biased in relation to other issues such as socioeconomic status.
In some cases, clients described their experiences with social workers from a different background
as positive and that the quality of respect was more important that sharing a common background.
To this end, continuing training on cultural competence and working effectively with specific
population groups was also considered a key practice.
High quality, supportive supervision that integrates coaching in working effectively with
diverse families was viewed as critical to practice with children and families of color. "Some
supervisors want to hear about cases and are supportive-others just want to know if you got the
reports done on time," summarized one participant. In addition, supervisors were identified as key
players in the process of brokering cultural consultations with other social workers, particularly
between social workers from different units.
Collaboration with other agencies and systems concerned with the safety and welfare of
children and families of color was also highlighted as a beneficial practice for children and families
of color. For instance, developing interagency formal agreements and connections with liaisons in
other agencies was described as beneficial because it facilitates collaboration and minimizes the
potential for agencies to give conflicting messages and mandates to clients. Social workers are "able
to divert cases or get a better picture of the family when two agencies are working together to work
with the families, which helps not only us but also the client." Other agencies mentioned in this
context include probation, hospital, medical providers, family and mental health agencies,
CalWORKS, substance abuse treatment agencies, domestic violence agencies and other entities that
may be involved with clients.
81
Additionally, collaboration with other agencies to address systemic issues was described as a
best practice. Social workers noted that community meetings are held on a monthly basis for all
community agencies and are attended by program managers, but workers are encouraged to attend.
Participation in policy level discussion between systems was considered to be an important part of
collaboration. Several participants mentioned the Greenbook project, an initiative to better
coordinate domestic violence and child welfare services, as a model for interdisciplinary problem
solving and policy development. The Greenbook project successes could help inform evolving
collaboration with other key systems, such as the substance abuse treatment system.
Other collaborative efforts such as outstationed social workers, co-location of services and
Family Drug Court were noted as best practices for children and families of color. Participants
described having ER social workers at different sites such as the Family Violence Center and police
stations as a valued practice. Moreover, South County social workers noted "we are located at the
resource centers where there are community providers here that are not child welfare ... we've got
AA meetings, parenting classes-everything is centralized here and that makes it comfortable for
the families." Although social workers mentioned that co-location of services have been reduced
because of budget changes and other factors, it remains a positive practice for children and families,
particularly low-income families and families of color that may have less time and fewer resources
to access services. Lastly, social workers and supervisors described the family drug court as a model
for effectively working with substance abusing families in the child welfare system. Several
examples of "successful" cases involved families assigned to family drug court.
A strength-based approach to services was also mentioned as an important best practice that can
be used at all points in the CWS. Indeed, strength-based practice was lauded as a fundamental tenet
of social work that is crucial for working with children and families of color. Specific elements of
this practice named by study participants include, looking at the unique characteristics of each
family; building solutions with families and showing that they [families] do have some control;
being willing to try new things, especially when clients make the suggestions; involving the parent
in the removal of the children, a strategy that is empowering to parents and less traumatic for the
children; demonstrating a commitment to invest time with families on the "front-end" to build
rapport, facilitate client understanding of the child welfare system, and optimize chances for
maintaining or reunifying the family; working with families to identify relative placements when
out of home placements are required and coaching relatives through the application process; and
lastly, demonstrating a willingness to advocate for the child and family with the court and other
service delivery systems.
Contextual Factors:
Recommendations to Improve Practices for Children and Families of Color
Study participants described five overall recommendations to improve practices for children
and families of color as they relate to contextual factors, including 1) reducing caseloads, increasing
support and streamlining service delivery, 2) maintaining and expanding multifaceted approaches to
providing culturally competent practice, 3) increasing linguistic capacity of systems and services for
families, 4) fostering an organizational culture that is strength-based and community based, 5)
enhancing collaboration, and 6) training social workers. Specifically,
Reduce Caseloads, Increase Support and Streamline Service Delivery
82
•
•
•
•
•
•
•
•
•
•
Continue efforts to reduce social worker caseloads. Reduction in caseloads was uniformly
identified as key to providing quality services to clients, particularly children and families of
color that may have complex needs and that may initially have less trust and/or less'familiarity
with child welfare and court systems. "Social workers need time to do their job and to help
clients access culturally appropriate services and programs."
Provide additional Social Worker I support for in-home assessment/support, follow up with
families, and overall cas~ support.
Provide dedicated staff, and explore other options, to expedite requirements to qualify relative
placements. For example, policies requiring relative placement standards similar to foster
families might be reevaluated (e.g., illustrations of this as a problem included an example of two
young girls were denied relative placement because they shared a queen size bed rather than
having separate beds). In addition, develop strategies for overcoming barriers to relative
placement, such as an expedited process for Director's Exemption in the case of a distant
criminal history of a family member in the prospective home.
Examine opportunities to facilitate quicker access for social workers to criminal background
checks.
Develop procedure or protocols to facilitate communication between units, particularly in
relation to transferring ER cases to voluntary family maintenance informal supervision.
Examine options for maximizing the skills of social workers with specific cultural or language
expertise. For example, social workers in ER and DI stated that support across units that
involved overtime was recently prohibited. However, social workers suggested that this practice
failed to reduce the total amount of overtime expended by social workers while limiting
flexibility in obtaining culturally and linguistically skilled consultation and assistance.
Develop centralized, and updated resources lists for a wide range of services, including service
providers with expertise in serving diverse cultural and language groups. This practice was
identified as key to avoiding duplication of effort and ensuring access to quality services for
clients. "Sometimes we make a lot of phone calls just to find out what someone two aisles from
you already knew." In addition, update Court Services Handbook, which was perceived by
several social workers to be out of date.
Examine opportunities to provide flexibility in office hours, alternate social workers to cover
cases when social workers are on vacation, or other mechanisms to reduce potential burnout.
Provide social workers access to adequate transportation (e.g., "better cars") for trav~l related to
working with families.
Develop protocols that reduce paperwork so "there is more time to do social work." For
example, interim court reports could be eliminated or reduced, templates for psychosocial
assessments or other guidelines for court reports could be provided, and social workers could
"get assistance especially with filing."
Maintain and Expand Multifaceted Approach to Providing Culturally Competent Practice
•
•
Continue practice of attempting to provide a cultural match for clients as a strategy for
facilitating rapport and, in some cases, minimizing cultural or class bias.
At the same time, maintain recognition that 1) some clients may prefer to see a Social worker
who is not of their background, 2) many clients value "being treated with respect" above being
assigned a social worker of the same background, and 3) the diversity of the population in Santa
Clara County may not always allow for "matching" clients and social workers in relation to
race/ethnicity or other demographics. Ensure training and support for all social workers in
83
•
•
cultural competence and addressing the needs of different population groups among all social
workers
Continue working to ensure diverse staff in units throughout the agency.
Continue using the family resource centers, including provision of services for mandated clients.
Increase Linguistic Capacity of Systems and Services for Families
•
•
•
•
•
Facilitate development of services specifically designed for families whose primary language is
not English in key areas including mental health, domestic violence, substance. abuse,
counseling services, parenting without violence classes, and other services.
Continue to hire bilingual/bicultural staff. Need for greater linguistic capacity was reported for
a number of different language groups including for Spanish speaking, Vietnamese, and Filipino
clients.
Minimize need for use of translators. When needed, provide high quality translators who are
able to communicate concepts related to the law, the social work system, timelines, and court
processes.
Provide forms and written materials in multiple languages.
Examine opportunities for establishing guidelines for contracting or collaborating agencies in
relation to linguistic services, such as procedures disallowing use of children as translators for
monolingual speaking parents.
Foster an Organizational Cult~re that is Strength-Based and Community-Based
•
•
•
•
Continue to identify mechanisms, such as the Family-to-Family Initiative, to build community
involvement in decision-making.
Communicate a commitment to assuring the least restrictive placement for children from
management, through supervision, and to line staff. Provide training and support for social
workers to advocate for clients in court and other settings.
Examine strategies to minimize what study participants described as internal disincentives to
ensuring the least restrictive placement for children including "County Council whose role it is
to protect the agency sometimes telling social workers to change recommendations" and the
greater paperwork and advocacy required for less restrictive options.
Ensure client access to information about complaint procedures and access to the Ombudsman.
Enhance Collaboration
•
•
Continue to develop collaborative case planning, linkages, information sharing and policy
development with other service delivery systems, including schools, service providers,
probation, CalWORKS, housing, and other systems. Collaborative planning and problem
solving was perceived as particularly crucial in an environment of sever fiscal restraints.
Develop protocols to facilitate access to mandated services for clients in the context of federal
and state timelines, such as formal agreements and protocols to expedite access to substance
abuse treatment.
Train Social Workers
84
•
•
•
•
•
•
•
•
Require cultural competence training for all new staff and mandate continuing education for all
social workers. Include, as part of the training, opportunities for self-reflection and examination
of personal bias related to culture of other issues frequently confronted in families, such as
substance abuse. In these areas, provide training that is high quality, dynamic, and delivered by
skilled trainers who are knowledgeable about both their subject area and child welfare.
Provide education within the agency about different cultural norms in parenting and family
strengths.
Provide training on specific issues that impact all families and that may differentially impact
children and families of color including policies and practices for addressing drug exposed
infants, substance abuse, domestic violence, sexual abuse, and other topics. In addition, provide
training on procedures and practical considerations for working with diverse families, such as
training on guardianship.
Provide training on immigration law and practical strategies for working with families, such as
identification of immigration status of the child at an early stage and working with international
relative placement.
Training on these topic areas should integrate consideration of cultural differences and
evidence-based practices for different populations.
Provide training for all staff including clerical staff and eligibility workers.
Facilitate forums for social workers to share resources and effective practice strategies.
Create opportunities for social worker access to training. "We can't go to trainings because we
have too much else to do-so we're denied professional development opportunities."
Summary of Key Findings and Implications
Study participants described numerous practices and contextual factors that have either a
negative or positive impact on children and families ofcolor in Santa Clara County's child welfare
system. There is some evidence in the research literature that supports certain best practices
mentioned by study participants. For instance, general research supports the effectiveness of inhome services (see literature review), however these services appear to be most effective when they
are intensive, long-term and delivered by health or social service professionals and when the
workers delivering these services receive a high level of supervision (Barth, 1991; McGuigan et al.
2003; Olds et al., 1997). In addition, some research has suggested that parenting classes for child
welfare parents are more effective when they pair a didactic approach with a more experiential
approach, such as parent and child interaction (Carlo, 1993).
Participants also identifiedfamily conferencing as an effective practice. These study
findings are consistent with limited research in this area that suggests that inclusive practices, in
which parents are engaged in placement and other decisions, may result in more placement stability
for children, although possible differential effects for diverse populations has not been researched
(Palmer, 1996). Participants generally expressed optimism about the adoption of team decision
making (J'DM) as a vehicle for improving decision-making with diverse families, although some
social workers expressed concern about time required for implementation. Research does suggest
that TOM can be labor intensive and can be difficult to implement effectively within the context of
the child welfare system (Sieppert et al., 2000). Finally, participants noted that both formal and
informal collaborative relationships with other services delivery systems, such as domestic violence
and substance abuse treatment systems, are critical for successful interventions with children and
families. Furthermore, participants recognized the importance oftraining and cross-training to
85
enhance the capacity ofhelping professionals in child welfare and other systems to better address
the needs ofchildren and families ofcolor.
Interestingly, most practices and contextual factors mentioned by social workers had both
positive and negative aspects. Many of the practices and contextual that negatively impact children
and families of color were actually barriers to successful implementation of the practices that social
workers felt have a positive impact on children and families of color. Although practices such as
orientation ofnew clients; family and team decision-making; preventive services; substance abuse
services; cultural competency; collaboration with other agencies and systems; a strength-based
approach; and the use ofrelative placements were described as having a positive impact on
children and families ofcolor, certain negative practices and contexts impeded their effective
implementation. For instance, having to screen out large numbers of inappropriate referrals;
inconsistency in decision-making practices; a shortage ofservices, particularly substance abuse
and preventive services; lack ofclient access to services; difficult protocols for placing children
with kin; gaps in cultural competency; time limits; and agency-level factors such as heavy ·
caseloads, staffshortages, substantial amounts ofpaperwork, lack of access to information about
resources; and confusion about the .agency's overall mission and key policies all impeded
implementation ofthe best practices. These findings suggest that in addition to building on current
positive practices for children and families of color, CWS stakeholders should also actively work
both internally and in collaboration with partners to reduce barriers to these best practices.
Indeed, in order to address these barriers to effective implementation of best practices, study
participants mentioned numerous recommendations that centered around certain key themes. In
general, there was an emphasis on improving decision-making through increased accountability,
training and cross-training to reduce bias, and increased use of group decision-making or family
involvement in decision-making. Study participants also stressed a need to improve and expand
prevention, diversion and concrete services, as well as an overall need to expand culturally and
linguistically competent services, and develop ways to improve availability and access to services.
Once a case is opened, social workers also felt it valuable to provide an orientation to clients so that
they are more aware of the CWS and court processes and can better navigate the system.
Recommendations related to organizational factors were also noted, including reducing caseloads;
increasing support and streamlining service delivery; clarification .ofthe mission ofDFCS, as well
as key child welfare policies; fostering an organizational culture that is strength-based and
community-based, and more overall training for social workers.
Limitations and Conclusions
Qualitative data, such as focus groups and interviews, provide rich insights into practices
that may positively or negatively impact children and families of color. In any such endeavor, there
are possible biases that may impact the results. For example, given the likelihood of time conflicts
and emergencies, we invited more social workers to participate in the focus groups than we
expected to attend. It is not_possible to determine if other factors influenced participation. Workers
with a particular interest in this topic may have made a special effort to attend. In addition, the
perspectives and opinions of the social workers, supervisors and families who participated in the
study may not be entirely reflective of the populations they represented. At the same time,
participants were representative of different units and facets of the child welfare system, different
race/ethnic groups, and varied levels of experience. The findings ofthis study affirm the value of
many efforts that are already in place, such as culturally specific ER response units and family
86
resource centers, and other that are in the process of implementation including efforts to reduce
caseloads, initiation ofteam decision making, and participation in the Family-to-Family initiative.
The findings ofthe study also point to opportunities to further strengthen services for children and
families ofcolor. Recommendations related to training, institutionalizing group orientation for
clients, and other suggestions from study participants could be used to inform practice and planning.
It would be beneficial for a team ofmanagers, line staff, family representatives, and community
members to review these findings to identify and prioritize possible practices for adoption based on
their feasibility and utility for children and families ofcolor.
87
Comparison of the Main County Offices and South County Offices
In an effort to better understand the influence of contextual factors on child welfare practices
and outcomes for children and families of color, a comparison of Santa Clara County's Main
Offices and South County Offices was conducted. This comparison included both quantitative and
qualitative analyses. Quantitative data from CWS/CMS were used to examine differences in case
characteristics between the Main County Offices and South County Offices. A qualitative
comparison of focus group data from the Offices and South County Offices was also conducted to
discern how workers perceive practices in these two geographical locations. These analyses help
shed light on the potential impact of agency and community context on case characteristics, child
welfare practices and outcomes for children and families of color in Santa Clara County's CWS.
Quantitative Methods
In Phase 2, the CWRT continued its investigation of the disproportionate representation of
children of color in the child welfare system by analyzing a sample of cases closed during an 18month period (January 2000 through June 2001 ). Key case characteristics were analyzed, including
ethnic comparisons. Please see the Phase 2 report for methodology details and case characteristics
by ethnicity (Hines et al., 2002).
One of the recommendations that emerged from Phase 2 of the study was to explore possible
differences in case characteristics and practices between the Main County Offices and the South
County Offices. In Phase 3, we used the basic characteristics from the 1720 cases analyzed in Phase
2 while adding the Santa Clara County location information provided via the Child Welfare System
/ Computer Management System (CWS/CMS).
Analyses
A series of exploratory, bivariate analyses were conducted with county location as the
distinguishing variable, Main County (MC) versus South County (SC). The purpose of this initial
set of analyses was to identify differences in the CWS related to county location in order to help
describe variations not only by ethnicity, but by other case and service characteristics. The
following sections will describe (1) the proportion of cases with each county location, (2) general
characteristics by county location, and (3) significant differences related to county location.
Definition of Terms
County Location - Two overall geographical areas within Santa Clara County designated as either
Main County Offices (centered primarily in the Northern part of the County, including the main
office on Julian Street in San Jose) or South County Offices (based in Gilroy).
Quantitative Results
County Location
Of the 1720 cases, 1633 (94.9%) had valid county location information with 87 missing
(5.1%). Of the 1633 valid cases, 1487 (91.1%) were served in MC, 50 in SC (3.1%) and 96 (5.9%)
88
in both county locations. Please see Table 15.
•
a
T able 15 : P roport1on o fC ases biy County Location
Total
Main (MC)
8
1633
1487
(91.1%)
South (SC)
Both
96
(5.9%)
50
(3.1%)
a Based on 1633 valid cases, excluding 87 cases with missing location
For the primary ethnic comparison, we retained three county location designations, MC, SC,
and Both (cases served in MC and SC). However, given the range in the length of time children
designated as Both were served in MC and SC, and the focus on comparing the two stated
geographical areas, subsequent analyses excluded this third category.
County Locatio~ and Child. Characteristics
Child's Ethnicity
There was a significant relationship between the child's ethnicity and county location. As
noted in Phase 1 and 2, Latinos compose the largest ethnic group served in the county's CWS.
When examined by location, those served in MC were more likely to be Latinos (44.7%, 661 of
1480 MC cases). This proportion of Latinos was higher in SC (68.0%, 34 of 50 SC cases) and Both
(62.1 %, 59 of 95 Both cases). African American, Asian American/Pacific Islander, and Other
children's cases were least likely located in SC and Both locations. Please see Table 16.
Table 16: Ethnic Group By County Location
Total
Child's Ethnicitya
•
•
N size
•
White
•
476
(29.3%)
Latino
754
(46.4%)
•
•
African American
1625
177
(10.9%)
Asian American
/Pacific Islander
156
(9.6%)
Other ethnicity
62
(3.8%)
Main
South
Both
1480
50
1
(2.0%)
95
6
(6.3%)
13
(26.0%)
29
(30.5%)
34
(68.0%)
59
(62.1%)
1
(2.0%)
0
(0.0%)
1
(2.0%)
1
(1.1%)
170
(11.5%)
434
(29.3%)
661
(44.7%)
155
(10.5%)
60
(4.1%)
a Based on 1625 cases with valid MC and SC county locations and county location infonnation
Unfortunately, the South County Office subsample was too small to allow more detailed analyses by
ethnicity. However, we continued to pursue our comparison of the two primary county locations by
other key variables.
89
Child Demographics
There was not a significant relationship between gender and county location. Within MC,
52.2% was female (776 of 1486 cases) compared to SC where 50.0% was female (25 of50 cases).
Within SC, 47.8% was male (710 of 1486 cases) compared to SC where 50.0% was male (25 of 50
cases).
There was not a significant relationship between age and county location. The average age
of the MC sample was 5.95 years (sd = 5.16), while the average age of the SC sample was 6.57
years (sd = 4.19).
There was not a significant relationship between child's language and county location. The
majority of the MC and SC samples spoke English. However, in SC, Spanish was the only other
language other than English spoken. Please see Table 17.
T able 17 Ch.Id
1 Demograph.1cs
Gender
Total
Main
South
1536
1486
801
(52.1%)
776
(52.2%)
50
25
(50.0%)
735
(47.9%)
710
(47.8%)
25
(50.0%)
5.97
(5.13)
5.95
(5.16)
6.57
(4.19)
8
•
•
N Size
Female
•
Male
Average age in years (sd/
Child's Languagec
•
•
N Size
1524
1474
50
English
•
•
1308
(85.8%)
1262
(85.6%)
46
(92.0%)
Asian American/
Pacific Islander
62
(4.1%)
62
(4.2%)
0
(0.0%)
Spanish
•
144
(9.4%)
4
(8.0%)
Other
140
(9.5%)
10
(0.7%)
•
IO
(0.7%)
a Based on 1536 cases with valid MC and SC county location and gender information
b Based on 1537 cases with valid MC and SC county location and age information
c Based on 1524 cases with valid MC and SC county location and language information.
0
(0.0%)
90
County Location and System-Related Factors
Voluntary Status and Reason for Removal
There was a significant relationship between voluntary status and county location. The
proportion of cases in voluntary services was higher in SC at 88.0% (44 of50 cases) compared to
41% in MC (600 of 1462cases).
There was not a significant relationship between the reason for removal and county location.
Please see Table 18.
T a ble 18 Vo luntary status an dReason or Remova
Total
8
Voluntary Status
N Size
1512
Main
South
1462
50
•
•
Voluntary
•
644
(42.6%)
600
(41.0%)
44
(88.0%)
Not Voluntary
868
(57.4%)
862
(59.0%)
6
(12.0%)
1206
1196
10
Reason for Removal b
•
N Size
General Neglect
•
161
(13.3%)
158
(13.2%)
3
(30.0%)
Severe Neglect
•
172
(14.3%)
172
(14.4%)
0
(0.0%)
Physical Abuse
•
300
(24.9%)
297
(24.8%)
3
(30.0%)
Sexual Abuse
•
73
(6.1%)
73
(6.1%)
0
(0.0%)
Emotional Abuse
•
40
(3.3%)
40
(3.3%)
0
(0.0%)
Other Maltreatment
460
(38.1%)
456
(38.1%)
4
(40.0%)
•
a Based on 1512 cases with valid MC and SC county location and voluntary service information
b Based
on 1206 cases with valid MC and SC county location and removal information
91
Initial Out-of-Home Placement Facility Type
There was not a significant relationship between initial out-of-home placement facility type
and county location. Please see Table 19.
T a bl e 19 : Imt1a
.. 1Out-o f.-Home Placement Facility Type
Total
Main
South
1206
1196
10
Initial Out of Home Placement
Facility Type8
N Size
•
•
Foster Family Home
•
258
(21.6%)
258
(21.6%)
0
(0.0%)
Group Home
•
66
(5.5%)
66
(5.5%)
0
(0.0%)
Children's Shelter/
359
(29.8%)
353
(29.5%)
6
(60.0%)
Relative Home
•
343
(28.4%)
341
(28.5%)
2
(20.0%)
Foster Family Agency
123
(10.3%)
I
•
124
(10.3%)
Guardian Home/Court
56
(4.6%)
55
(4.6%)
•
a
Receiving Home/ Non-EA
/AFDC
Specified Home
(10.0%)
1
(10.0%)
Based on 1206 cases with valid MC and SC county location, and placement information
Number of Times removed from Family and
Number of Unique Placement Homes in Current Episode
There was not a significant relationship between the number of times a child was removed
from their family in current episode and county location. On average, children in the sample were
removed 1.35 times (sd = 0.70).
There was not a significant relationship between the number of unique placement homes in
current episode and county location. The sample had an average of 3.56 (sd = 3.86) unique
placement homes in current episode. Although MC cases had a higher average of placements (3.57,
sd = 3.88) compared to SC (1.90, sd = 1.10), this difference did not reach statistical significance,
most likely due to the small number of SC cases available for comparison. Please see Table 20.
92
Table 20: Number of Times removed from Family and Number of Unique Placement Homes in
C urrentE.
,p1sode
Average Number of Times
Removed from Family in
Current Episode (sd)8
Average Number of Unique
Placement Homes in Current
Episode (sd)b
Total
1.35
(0.70)
Main
1.35
(0.70)
South
1.10
(0.32)
3.56
(3.86)
3.57
(3.88)
1.90
(1.10)
a Based 1228 cases with valid MC and SC county location and removal information
b
Based on 1228 cases with valid MC and SC county location and placement information
Average Stay (in days) per Placement Facility, Number of Placements, and Total Length of
Time in Out of Home placement in Current Episode
There was not a significant relationship between the average stay (in days) per placement
facility in current episode and county location. The average number of days a child stayed in one
placement in their current episode was 244.06 (sd = 480.43). Although MC cases had a higher
average of245.73 days in placement (sd = 482.03) compared to 41.54 days (sd = 44.71) in SC, this
difference did not reach statistical significance, again most likely due to the small number of SC
cases available for comparison.
There was not a significant relationship between the number of placements in episode and
county location. On average, children in the sample had 2.84 placements (sd = 2.95). Although MC
cases had a higher average of2.85 placements (sd = 2.95) compared to 1.80 (sd = 1.32) for SC, this
difference did not reach statistical significance, again most likely due to the small number of SC
cases available for comparison.
There was not a significant relationship between the total length of time (in months) in outof-home placement in current episode and county location. On average, children in the sample spent
13.16 months (sd = 20.40) in out-of-home placement. Although MC cases had a higher average of
13.24 months in placement (sd = 20.47) compared to 3.28 months (sd = 4.24) for SC, this difference
did not reach statistical significance, again most likely due to the small number of SC cases
available for comparison. Please see Table 21.
93
Table 21: Average Stay (in days) per Placement Facility, Number of Placements, and Total Length
. 0 ut o f H ome p1acement m
. Current E,ptso
. de
0 fT1mem
Average stay (in days) per
Placement Facility Type in
Current Episode (sd)8
Number of placements in
Total
Main
South
244.06
(480.43)
245.73
(482.03)
41.54
(44.71)
2.84
1.80
(1.32)
3.28
(4.24)
Current Episode (sdl
(2.95)
2.85
(2.95)
Total length of time (in months)
in Out of Home Placement in
Current Episode (sdt
13.16
(20.40)
13.24
(20.47)
a Based on 1228 cases with valid MC and SC county locations and placement information
Based on 1206 cases with valid MC and SC county locations and placement information
c Based on 1200 cases with valid MC and SC county locations and time information
b
Number of Episodes, Total Length of Case (in years), and Number of Workers Assigned to
Case over Time in the Current Episode
There was not a significant relationship between the number of episodes and county
location. Children in the sample had an average of 1.19 (sd = 0.47) episodes.
There was a significant relationship between the total length of the case (in years) and
county location. On average, children in the sample had a case length of 1.87 years (sd = 2. 78). MC
cases lasted longer with an average of 1.93 years (sd
SC.
= 2.80) compared to 0.18 years (sd = 0.39) for
There was a significant relationship between the total number ofworkers assigned to case
over time in current episode and county location. Children in the sample had an average of 4.91
workers (sd = 2.75). MC cases were assigned a higher average of 5.03 different workers (sd = 2.72)
compared to 1.48 workers (sd = 0.84) for SC. Please see Table 22.
Table 22: Number of Episodes, Total Length of Case (in years), and Number of Workers Assigned
tC
heeurrent E.d
,p1so e
mt
0 ase over T 1me ·
South
Main
Total
1.16
1.19
1.19
Average Number of Episodes (sdt
(0.37)
(0.48)
(0.47)
0.18
Average Total Length of Case (in
1.93
1.87
(0.39)
(2.78)
(2.80)
years) (sdl
Average Number of Workers
Assigned to Case over Time in the
Current Episode (sdt
4.91
(2.75)
5.03
(2.72)
Based on 1512 cases with valid MC and SC county locations and episode information
Based on 1512 cases with valid MC and SC county locations and case time information
c Based on 1532 cases with valid MC and SC county locations and worker information
a
b
1.48
(0.84)
94
Last Out-of-Home Placement Type
There was not a significant relationship between last placement type and county location.
Please see Table 23.
T a ble 23 Last O ut-o f.H
- ome Placement Type, andA.2e at T1me of Case Closure
Total
Main
Last Out of Home
Placement Type8
1218
1228
• N Size
South
10
•
Foster Family Home
•
260
(21.2%)
260
(21.3%)
0
(0.0%)
Group Home
•
77
(6.3%)
77
(6.3%)
0
(0.0%)
Children's Shelter
•
354
(28.8%)
6
(60.0%)
Relative Home
•
346
(28.2%)
348
(28.6%)
344
(28.2%)
2
(20.0%)
Foster Family Agency
•
133
(10.8%)
132
(10.8%)
1
(10.0%)
Guardian Home/
Court Specified Home
58
(4.7%)
57
(4.7%)
I
a
(10.0%)
Based on 1228 cases with valid MC and SC county locations and placement information
Age at Time of Case Closure and Service Type at Case Closure
There was not a significant relationship between age at time of case closure for current
episode an.d county location. On average, children in the sample were 8.19 years-of-age (sd = 5. 72)
at the time of case closure. Although MC cases closed with an older child's average age of 8.23
years (sd = 5.76) compared to 7.15 years (sd = 4.22) for SC children, this difference did not reach
statistical significance.
There was a significant relationship between service type at closure of case and county
location. Overall, the majority of cases ended in family maintenance (59.9%, 920 of 1537 cases).
However, SC had a higher proportion of cases in family maintenance (92.0%, 46 of 50 cases)
compared to MC (58.8%, 874 of 1487 cases). Please see Table 24.
95
. T ype at Case Closure
T a ble 24 A.ge at T1meofC ase Closure andS erv1ce
Main
Total
Age at Time of Case
8.19
8.23
South
Closure for the Current
(5.72)
(5.76)
7.15
(4.22)
Episode (sdt
Service Type at Case
Closureb
1537
1487
50
•
•
N Size
Emergency Response
•
39
(2.5%)
29
(2.6%)
0
(0.0%)
Family Maintenance
•
920
(59.9%)
874
(58.8%)
46
(92.0%)
Family Reunification
•
67
(4.4%)
64
(4.3%)
3
(6.0%)
Permanent Placement
511
(33.2%)
510
(34.3%)
1
(2.0%)
a Based on 1482 cases with valid MC and SC county locations and child's age at case closure information
b Based on 1537 cases with valid MC and SC county locations and service type information
Case Closure Type
There was a significant relationship between case closure type and county location. The
most common case closure type was court ordered termination and other (32.1 %, 482 of 1502
cases), followed by family maintenance (31.6%, 475 of 1502 cases). However, 62% (31 of 50
cases) ~fSC closed with family maintenance compared to 30.6% (444 of 1452 cases) in MC.
Please see Table 25.
96
T able 25 Case Closure T ype
Case Closure Type
• N Size
8
Total
1502
Main
1452
South
50
•
Adoption
•
246
(16.4%)
246
(16.9%)
0
(0.0%)
Emancipation
93
(6.2%)
0
(0.0%)
31
(62.0%)
•
•
•
•
•
a
Family Stabilized
(FM)
475
(31.6%)
93
(6.4%)
444
(30.6%)
Guardianship
99
(6.6%)
97
(6.7%)
2
(4.0%)
61
(4.1%)
59
(4.1%)
2
(4.0%)
Incarceration,
Runaway or Medical
services
46
(3.1%)
45
(3.1%)
1
(2.0%)
Court Ordered
Termination and Other
482
(32.1%)
468
(32.2%)
14
(28.0%)
Established or
Placement with
Relative
Reunified.with Parent
or Guardian, Court or
non-Court Specified
Based on 1502 cases with valid MC and SC county locations and case closure information
97
Qualitative Comparison of the Main County Offices and South County Offices
In addition to the quantitative comparison of the Main County Offices and South County
Offices, qualitative focus group data were collected to obtain information on how workers perceive
practices in the two geographical areas, and how these practices may differ.
Qualitative Methods
Prospective participants for the Social Worker focus groups were randomly selected from a
list of workers with title of Social Worker II or Social Worker III in different units/divisions. The
random selection and recruitment process is described in detail in the section describing qualitative
methods (please see page 61 for a full description of focus group qualitative methods). As described
above, a total of 67 Social Workers participated in the 13 focus group interviews. Of this number,
13 Social Workers from various units in South County Offices participated in three South County
focus groups. All the focus group participants in the South County sample were female. Of these, 2
(15.4%) identified as African American, 1 identified as Asian American/Pacific Islander (7.7%), 4
identified as Latina (26.9%), and 6 identified as white/Caucasian (32.8%). A majority of the
participants (n=l2, 92.3%) had masters level education, primarily MSW degrees (Please see
Attachment 2 for additional information about focus group demographics)
Procedures
Procedures for all focus groups are described in detail in the Qualitative Methods section of
this report. Procedures for all focus groups, including South County Office focus groups were
identical. As described above, ten focus groups were conducted at the main administrative office in
downtown San Jose and three were conducted at Gilroy Family Resource Center in South County.
Analysis
The process for analysis of the South County Offices data compared to other focus groups
was parallel to that described above in the Qualitative Methods section. The data for the qualitative
portion of this study consist of the transcribed notes taken during each of the focus groups and
individual interviews. Qualitative methods of data analysis were employed to analyze the focus
group and interview data. First, major themes were summarized for each of the choice points or
units in the Social Worker focus group and supervisor interview data. A separate analysis of the
qualitative data from South County Offices was conducted to identify themes that may have been
different or more.prominent in these series of 3 focus groups compared to the other 10 social worker
focus groups. The purpose of this additional analysis of the qualitative data was to obtain insights
and supplement quantitative analysis of possible differences between South County Offices and
Main County Offices within DFCS. In addition to gathering information about practices that may
differ between the Main County Offices and the South County Offices, findings from the qualitative
data also provide insights that may help explain quantitative differences in the relationship between
location and other variables, such as differences in the proportion of voluntary family maintenance
cases.
98
Qualitative Results
Many of the contextual and practice issues described by social workers and supervisors in
South County Offices paralleled comments by study participants in other regions of Santa Clara
County, such as concerns related to client poverty, housing shortages, and insufficient substance
abuse and other treatment resources, particularly for non-English speaking language groups. At the
same time, a few themes emerged that were either stronger or unique to South County Offices.
These are described below.
Context of Service Provision in South County Offices
Demographics of South County
According to study participants, South County has a large Latino community that is strongly
represented among child welfare clients. In relation to these demographics, participants emphasized
the need for bilingual and bicultural child welfare professionals. They also reported a paucity of
culturally and language specific counseling services for monolingual Spanish speaking participants.
In addition, access to counseling and services for undocumented families in the child welfare
system were perceived as a problem. Participants noted that one strength of the Family Resource
Center is the bilingual and bicultural social work staff. At the same time, participants saw a need to
expand capacity to work with other communities. "Our resource center is reflective of the majority
community of Latinos, but some clients may not feel represented," explained one South County
focus group member. Specifically, several social workers suggested that the current and evolving
demographic changes in the region called for greater representation of African American and Asian
American/Pacific Islander social workers.
Small County Dynamic
South County Offices were perceived as different from "the main" office in part because
dynamics related to the smaller size of the region. Participants described feeling more connected to
the community, other service providers and clients because of the "a small-town atmosphere."
Participants noted that, "The community is smaller and there is more interaction within the
community and with individuals from other agencies; this tends to make people work with clients
differently." For example, one social worker commented that she often sees clients and colleagues
at the grocery store and others mentioned that clients freely come into the child welfare offices and
Family Resource Centers asking for assistance.
Furthermore, study participants in South County consistently reported a high level of both
formal and informal consultation with one another. In addition to relying on other workers and
supervisors to assess families and develop case plans, social workers working in close physical
proximity are aware of one another's cases and routinely exchange informal support and advice
about working with clients and accessing resources. Some participants noted that they might be
"less likely to connect" in larger facilities or offices "built like a maze." The absence of security for
entering the agency building "changes the atmosphere because clients can just walk in, there is an
open door policy; this changes agency attitude in working with clients and removes barriers for
clients."
99
Culture of Commitment to Maintaining and Reunifying Families
South County study participants described a shared philosophy that the vast majority of
children referred to the child welfare system can and should be with their birth parents, whether it
be through family maintenance or through family reunification. Specific values and norms
embedded in the philosophy and practice of the region include emphasis on using family strengths,
including families in the process of making decisions, and aggressively seeking alternatives to
removal. Study participants underscored the importance of this approach in working with children
and families of color. Although these principles are valued throughout the child welfare system,
participants in South County stated that these principles were infused throughout the system and
were used consistently to guide practice and social workers' accountability in assessment and
decision-making. For example, social workers described substantial involvement of supervisors and
coworkers when removal of children into protective custody is required. Study participant
comments suggested that the smaller size of the agency helps to solidify a sense of shared
philosophy and reinforce practice norms.
When asked to articulate how these values and norms are sustained and passed on, workers
cited several factors. First, study participants emphasized the importance of consistency in
messages, policies, and direction from the Program Manager and supervisors. Second, social
workers suggested that these expectations related to seeking the least restrictive placement for
children are communicated beginning with the initial hiring and orientation process through
everyday practice. "People who get hired have the same philosophy as South County ... and
supervisors ingrain values into workers," noted participants. Third, social workers stated that it was
normative for workers to obtain both formal and informal support when they are making decisions.
"Workers do a lot of staffing and hash things out; we don't make decisions alone," noted one
participant. "By using co-workers and supervisors in decision-making, you gain another
perspective," observed another. Integrating the perspectives of other workers in decision-making
was identified as crucial to mediating potential bias or misunderstanding related to cultural or
socioeconomic class differences. Finally, social workers suggested that they received substantial
support and consultation from supervisors in decision-making and that supervisor values and
practices were perceived as critical to the practices of line staff. For example, supervisor
involvement in decision-making was described as a positive practice, particularly in relation to
taking children into protective custody and out of home placement. "My supervisor was on the
phone the whole time I had to remove a child," was offered as an example from one social worker.
Such oversight was perceived as a sign of support rather than solely a mechanism to assure social
worker accountability.
Practices in Service Provision in South County Offices
Vertical Case Management
Focus group participants and supervisor interviewees named the vertical case management
model, in which social workers assigned to a case work with the family throughout their case rather
than transferring cases to a different unit, as an effective practice with clients, particularly with
children and families of color. Some study participants suggested that the vertical case management
model works well in a small community. "It works because it puts social workers in the community
with the family, making social workers accountable ... it forces social workers to have to deal with
100
that family not just in a [case] process sense but in a real social work sense of developing
relationships," summarized one supervisor interviewee. Social Workers suggested that maintaining
continuity through a specific case enhanced trust with clients and created a greater sense of
accountability. The opportunity to develop and sustain rapport with specific clients was deemed to
be particularly important for marginalized groups, including children and families of color. Families
that return to the system are also generally reassigned to the same social worker. "In this case a new
worker may not have rapport or know family strengths ... A worker who has worked with a family in
the past knows them well enough to call them on their stuff." At the same time, respondents also
described vertical case management as work intensive and requiring additional support to ensure its
success, such as the help of Social Worker I staff.
Study participants in the focus groups that took place at the main administrative office in
San Jose location appeared to be familiar with the vertical case management model. Some
mentioned that it might be a possibility as a positive practice for children and families of color in
other parts of the county while others stressed the importance of maintaining specialized units with
staff who have expertise and an opportunity to stay current with changing laws, policies and
practices in their area.
Orientation of Clients
Social Workers in South County frequently refer parents to a client orientation that helps
parents understand the child welfare system and policies related to timelines for reunification. The
orientation is offered in English and in Spanish. It is offered at different times and provides 6 hours
of information over three sessions. Three areas are addressed in the orientation: timelines and facets
of the child welfare system (ER, DI, FR, FM, adoption), how to work with the social worker, and
resources for children and families. The orientation is not mandatory and is not offered at all sites.
One focus group participant examined outcomes (identified as either reunification or closure of
case) and found that the orientation appeared to be particularly helpful for families in the Spanish
speaking orientation. In sum, the orientation for clients was recognized as an important resource for
families, particularly families of color, who might feel confused or intimidated by the child welfare
system. At the same time, workers appeared to feel that there were opportunities to further
strengthen, institutionalize, and evaluate this practice.
Case Conferencing
Study participants stressed as a best practice the formal and informal process of meeting
with co-workers and supervisors to review cases and participate in the decision-making process.
Accountability and support in decision-making was viewed as an important factor in minimizing
potential cultural or other bias about clients and in maximizing creative thinking about how best to
work with children and families. Inclusion of families in decision-making was also noted as
positive practice that was considered particularly important for children and families from cultures
that place a high value on extended family, including both biological and fictive kin. "If you have a
voluntary family maintenance case, you have a family conference," observed one social worker.
a
101
Summary of Key Findings and Implications:
Quantitative and Qualitative Comparison
of Main County Offices and South County Offices
Unfortunately, the South County location subsample was too small to allow analyses by
ethnic group. However, some key case characteristics are significantly related to county location.
Specifically, in South County Offices, a higher proportion ofcases are in voluntary services, cases
are shorter in duration, fewer workers are assigned during the course of the case, and a higher
proportion ofcases are in family maintenance at case closure. These findings are also statistically
noteworthy given the small subsample size available from SC. There were also other differences in
case characteristics by county location that did not attain statistical significance but are noteworthy
for further investigation. Children served in the South County Offices appear to have a fewer
number of unique placement homes in the current episode, have a shorter average stay per
placement, a shorter length oftime in out-of-home placement, and are younger at time ofcase
closure.
These findings imply that there is a difference in the style and quality of services provided in
the South County Office location. Qualitative findings are congruent with the quantitative data and
provide some insight into the dynamics that may contribute to this difference. The higher number of
cases in voluntary services, the larger proportion ofcases in family maintenance at case closure,
and the shorter duration ofcases is consistent with what South County study participants described
as a culture ofcommitment to maintaining and reunifying families. This commitment was described
as integral to the philosophy of managers and social workers and reflected in the expectations
communicated to social workers by most supervisors and by peers. The practice ofproviding an
orientation to the child welfare system, time limits, the courts, and how to work with social workers
may also contribute to the differences found in the quantitative analysis. Study participants
frequently described South County as similar to a small county or rural area. This geographic
difference, and the resulting sense of "connection to the community,,, was identified as a/actor in
the greater sense ofshared philosophy and emphasis on prevention ofout-of-home placements.
Other studies suggest that region may influence practice. For example, one study (Drake, 1996) also
found that rural areas are more likely to offer preventive services than are urban areas. The finding
that fewer workers are assigned during the course of the case is also consistent with descriptions of
the vertical case management model. This model calls for social workers to carry the same case
from case opening to case closure. assigned to a case works with that family from the initiation to
closure of the case.
Limitations and Recommendations
Only about 3 % of cases were designated from SC, and the third location designation of
"both" was dropped. A key informant from South County explained that many of the cases
categorized as "both" may have been opened by ER workers from the Main County Office but
carried in South County prior to assignment of South County staff to a specific region of the county.
Similarities in demographics between the South County Office cases and the "both" categories
would appear to be congruent with this hypothesis. However, since this explanation could not be
verified in the data that was available, the cases designated as both were not combined with other
South County data. Future research should include a larger subsample from SC and add methods to
102
understand the reasons behind these location differences. Also, a larger South County Office
location sample may enable more detailed comparisons by ethnicity.
Some.practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County Offices and adapted to
other regions ofSanta Clara County, as recommended by study participants. This practice may be
particularly helpful to low-income families and families of color that may be intimidated by or
unfamiliar with child welfare and related systems. Other practices, such as vertical case
management, may hold promise for other regions ofSanta Clara County, though the success ofthis
practice may be linked to the "small county'' dynamic described by South County study
participants. Some participants in focus groups that took place at the main office mentioned this
model as promising while others asserted that specialized expertise in different areas of social work
was an advantage in providing quality services that are informed by "the most current information."
Further examination of how the culture ofcommitment to maintaining and reunifying families is
created, communicated, and continued and how this might be adapted to other parts of the county
warrants attention.
103
Multivariate Analysis of Fae.tors Predicting Reunification
Preliminary Models
In order to identify system-related factors and case characteristics predictive of reunification
of children with their families, two exploratory models were tested using multivariate analyses. The
first was with a larger sample of 1720 closed cases, and the second with 403 cases from a case
record review sample. Both samples were collected during Phase 2 and additional data was
collected and analyzed during Phase 3.
Results: Closed Case Sample
We chose 8 demographic and system-related variables available through CWS/CMS to
predict reunification: child's gender, ethnicity, age at time of case opening, the number of workers
assigned across the history of the case, the length of the case, the number of unique placement
homes assigned, the number of times removed from the family, and the county location (Main
County Offices versus South County Offices).
For our analysis, due to missing data across our set of variables, 1190 cases out of I 720
were available for logistic regression. Our model was significant in predicting the reunification of a
child with the family (X2 = 176.27, df = 11, p = .001, Cox & Snell R = 0.14). Specifically, Asian
American/Pacific Islanders were less likely than Whites, African Americans, and Latinos to be
reunified, and the fewer number of workers assigned, shorter length of a case, and fewer number of
unique placement homes assigned were related to reunification (while controlling for each variable
within the model). Please see Table 26.
104
. . Reum.fi1cation in Larger Sample Model
T a ble 26 Log1st1c
. . R egress1on Pred1ctmg
Model
Cox & Snell R:
0.14
176.27
11
.001
xl.
df
p
Predictors
B and Significance
Gender (female baseline)
Ethnicity
African American vs White
Latino vs White
Asian vs White
Other vs White
Latino vs African American
Asian vs African American
Other vs African American
Asian vs Latino
Other vs Latino
Other vs Asian
Age at case opening
Worker Number
Time LenITTh of Case
Number of Unique Placement Homes
Number of Removals
County Location (Main vs South)
Using two-tailed tests * p < .05
** p < .01
Model Coding: 1 = reunified, 0 = not reunified
Odds Ratio
0.15
1.16
0.19
0.22
-0.59*
-0.33
0.02
-0.78**
-0.53
-0.81 ***
-0.55
0.26
-0.01
-0.19***
-0.02***
-0.10**
-0.02
0.40
*** p < .001
1.21
1.24
0.55
0.72
1.02
0.46
0.59
0.45
0.58
1.30
0.99
0.82
0.98
0.91
0.98
1.49
Summary of Key Findings and Implications:
Closed Case Sample
Our model composed of primarily demographic and system-related variables to predict
reunification (child's gender, ethnicity, age at time of case opening, the number of workers assigned
across the history of the case, the length of the case, the number of unique placement homes
assigned, the number of times removed from the family, and the county location) was significant.
Specifically, four of these variables predicted reunification: child's ethnicity, number ofworkers
assigned throughout the case, length of the case, and number of unique placement homes.
Asian American/Pacific Islanders were less likely than Whites, African Americans, and
Latinos to be reunified with their families. Given that Asian American/Pacific Islanders are one of
the minority groups least acculturated to mainstream society, many oftheir cases might be
perceived as more severe because ofthe differences between traditional Asian and mainstream
American cultural attitudes toward child rearing and discipline. Culturally, many traditional Asian
105
families have a high level ofparental control and tend to use corporal punishment as part oftheir
parenting (Chao & Tseng, 2002). Many Asian parents may also be unaware about the laws defining
abuse in this society. In these cases, cultural sensitivity by investigators and caseworkers would
help determine if these Asianfamily cases are indeed severe or if they are over-interpreted as such
because ofa philosophical difference between the professional and the family in parenting and
discipline. Future empirical studies could help determine the extent to which abuse severity or
cultural differences explain this relatively low reunification rate among Asian American families.
However, as Phase 2 results indicated Asian American/Pacific Islanders are also the most likely to
be placed in Voluntary Family Maintenance (Hines et al., 2002), the CWS appears to value the
preservation of Asian American/Pacific Islander families when first referred to the system (perhaps
recognizing the challenges of culturally appropriate services, language issues, and other
circumstances related to the Asian American/Pacific Islander community) but once in the system,
Asian American/Pacific Islander family cases are not easily resolved and circumstances may persist
that place the child at risk for further abuse ofneglect.
The fewer number ofworkers assigned was also related to reunification. This finding
indicates consistency in service and the relationship between the social worker and the family may
improve the chances ofreunification. Perhaps this is due to greater familiarity and thus more
effective advocacy for the child and family. This finding could also be a function of particularly
challenging family cases already less likely to be reunified needing a greater set of workers over
time (e.g., due to worker burnout in the case, or specialization needed across the case).
Shorter case length was also predictive of reunification. Presumably, less severe cases
could be resolved in a timely fashion, usually concluding with reunification. Following a similar
logic above regarding the number of workers, a longer case may imply more problems, thus
involving more time for resolution, but also possibly a case already protracted into a situation where
reunification is unhealthy for the child.
The fewer number of unique placement homes assigned was related to reunification.
Change in placements may be due to systemic conditions (i.e. logistics and short term availability of
space) but multiple placements can also indicate a persistent problem with a child adapting to a
placement and accepting care offered. As this may again indicate a more severe case, reunification
could also become less likely.
Due to missing data across our set of variables, 1190 cases out of 1720 were available for
this multivariate analysis. More complete information would have improved confidence in our
findings. Also, as we have made references to the possibility of severity ofa case influencing
reunification, better measurement and a more comprehensive evaluation offamily history, abuse
and neglect circumstances, and other indicators could produce a stronger predictive model. In the
future, we may be able to include more psychosocial and case variables in a larger sample to build
upon our current findings.
Results: Case Record Review Sample
To identify better the influence of case characteristics and indicators of success (discussed in
the Successful Cases Analysis, beginning on page 48) we added 6 variables to the model with the
larger sample collected through our in-depth case record reviews and utilized the additional
106
information coded for Phase 3: a family's enhanced capacity to provide for their children's needs,
children receiving appropriate educational services, children receiving adequate services to meet
physical and mental health needs, number of referrals, number of previous times in the CWS, and
assignment to Family Maintenance or Family Reunification services. The location variable was not
included in this model as only one case coded from the South County Office area.
For our analysis, again due to missing data across our set of variables, 246 out of 403 cases
were available for logistic regression. Our expanded model was also significant in predicting the
reunification of a child with the family (X2 = 63.87, df= 16, p = .001, Cox & Snell R2 = 0.23).
Similar to the larger sample's model, Asian American/Pacific Islanders were still less likely than
Whites, African Americans, and Latinos to be reunified. Also, shorter time length of a case again
predicted reunification. However, in this expanded model with success indicators, the number of
workers assigned and number of unique placement homes assigned were not statistically related to
reunification. The child's age at time of case opening and the riumber of unique placement homes
did approach significance. Interestingly, the variables related to success cases and outcomes
( discussed in the Successful Cases Analysis, beginning on page 48) were not significant. Please see
Table 27.
107
. Reum.fi1caf10n .m C ase R ecord Rev1ew
. Sampe
Table 27 Log1st1c
. . R egress1on Pre d.1ctmg
1 Model
Model
Cox & Snell R'-
0.23
63.87
16
.001
xi.
df
p
Predictors
B and Significance
Gender (female baseline)
Ethnicity
African American vs White
Latino vs White
Asian vs White
Other vs White
Latino vs African American
Asian vs African American
Other vs African American
Asian vs Latino
Other vs Latino
Other vs Asian
Age at case opening
Worker Number
Time Length of Case
Number of Unique Placement Homes
Number of Removals
Number of Referrals
Previous Times in the CWS
Case Assignment (FM vs FR)
Child Receiving Adequate Educational
Services
Child Receiving Adequate Services to
Meet Physical and Mental Health Needs
Family's Enhanced Capacity to Provide
for Child's Needs
Using two-tailed tests* p < .05
** p < .01
+ approached significance at p < . l 0
Model Coding: I = reunified, 0 = not reunified
Odds Ratio
0.29
1.34
0.43
-0.19
-1.63*
-0.36
-0.61
-2.06**
-0.79
-1.44*
-0.17
1.27
-0.06+
-0.06
-0.03***
-0.ll..
-0.05
-0.04
0.12
-0.50
0.10
1.21
1.53
0.83
0.20
0.54
0.13
0.46
0.24
0.84
3.56
0.95
0.95
0.98
0.88
1.05
0.96
1.13
0.61
1.10
-0.18
0.84
-0.13
0.88
*** p < .001
108
Summary of Key Findings and Implications:
Case Record Review Sa~ple
These findings indicate that particular factors are important to consider when predicting
reunification, specifically ethnicity and the length oftime a case remains open. Similar to the model
in the larger closed case sample, this implies that more severe cases (those needing more time for
resolution) may be more problematic, thus reducing the chances ofreunification. We saw in the
larger sample model that number of workers and unique placement homes were related to
reunification, but the number of workers was no longer significant and number of placement homes
approached significance. In addition, the younger the child at case opening the more likely
reunification would occur. This could be that an effort is made to reunify younger children with
their families given that they are in an especially crucial stage of psychological and emotional
development related to the attachment with parents. These variables should be explored in more
detail.
The finding again that Asian American/Pacific Islanders are less likely to be reunified than
Whites, African Americans and Latinos, highlights a main ethnic difference. In Phase 2 we
discovered that Asian American/Pacific Islanders were more likely than the other ethnic groups to
be enrolled in Voluntary Family Maintenance services (Hines et al., 2002). However, our results
here imply that when Asian American/Pacific Islander children are removed from the home, their
children are less likely to be reunified. This may also indicate extreme outcomes where Asian
American/Pacific Islander children are either initially diverted from the system or once in the
mainstream ofthe system are less likely to be reunited with their families.
It was interesting that the successful case and outcome variables were not predictive of
reunification. This could be due to the significant impact of other variables overshadowing the
influence of success indicators. It could also be due to our use of reunification as a dependent
variable. As we discussed in the Successful Cases Analysis (see page 48) reunification is not
necessarily synonymous with success, and thus these variables may indeed be unrelated.
Again, due to missing data across our set of variables, 246 cases out of 403 were available
for this multivariate analysis. Although only a preliminary set of models, these findings are useful
for identifying areas in the CWS and family circumstances that can be studied further, especially in
conjunction with the specific goal of reunifying children with their families. For future research,
reasons as to why major system-related factors are not predictive would also be valuable.
109
The Role of the Court System in Child Welfare Practice
Few studies have examined the potential influence of the court system on child welfare
practices and outcomes, however as noted in the literature review (Section III), the court system is
likely to have a significant impact on the trajectory of child welfare cases. In order to examine this
issue further, this analysis examines the role of the court system in child welfare practice.
Specifically, two overall issues are addressed. First, an analysis of quantitative case record review
data, and qualitative data from one focus group conducted with the Court Officer's Unit will
describe court ordered changes to social worker recommendations at the initial
jurisdictional/dispositional hearing; the types of changes the court orders; if these changes differ by
child's ethnicity; and possible explanations for these changes. Secondly, a qualitative analysis from
the agency-wide focus groups (described on page 61) will focus on themes related to the
relationship between the child welfare and court systems.
Quantitative and Qualitative Methods:
Analysis of Court-Ordered Changes
Quantitative Analysis
- The quantitative analysis of the role of the court system in child welfare practice included a
new case record review of the same sample of 403 closed child welfare cases that were included in
Phase 2 of the research.
Sampling
Case record reviews include the same sample of 403 closed child welfare cases that was
analyzed in Phase 2 of the research. This sample was created from 6761 total case closures over an
18-month period, and a data file containing 1753 unique cases representing one child per family and
one case opening was constructed. This data file was used to obtain our target sample of 403 cases
selected randomly, guided by stratification according to ethnicity, age and service type. Of the 403
cases reviewed for this court analysis, 12 cases were closed, but not reviewed because they were
with a worker and not at the retention center; and 13 were not reviewed because they had been reopened.
Data Collection
Data collection took place at the Santa Clara County's Social Services Record Retention
Center. A Senior Research Assistant worked with the Retention Center Manager to ensure that the
sample of case records from Phase 2 of the research were re-pulled for review. Cases were reviewed
by three research assistants, two of these research assistants participated in the Phase 2 case record
review data collection and so had already been trained in case record review methods and one
research assistant was newly trained to complete the case record reviews.
Instrumentation
A data extraction form was created and pilot tested (please see Attachment 5 for a copy of
the Court Cas~ Record Data Extraction Form). The form contained questions related to court
110
ordered service or visitation changes for the child, mother, father, alternate caregiver, or the whole
case.
Analysis
We addressed this aim through a 3-part series of exploratory descriptive and bivariate
analyses. First, we identified whether there were any court ordered changes and if court ordered
changes in general were related to ethnicity. A single change in recommended services for a case,
regardless of magnitude, would classify the case as one with a "court ordered change."
Second, we examined major changes and if these were related to ethnicity. We considered
any court ordered change in the three main service areas (those concerning bypass, family
maintenance, and family reunification) as a major change.
Third, we developed 4 summary scores consisting of the number of changes imposed
separately concerning the child, mother, and father, and then for the entire case. Across each case
there were 85 variables over 9 potential areas where a change could be ordered in addition to the
three main service areas. These areas were psychological evaluation, counseling services, health
services, drug/alcohol testing, substance abuse/alcohol treatment, parenting education, support
groups, domestic violence treatment, and visitation. Changes in these areas were assessed and
summed into the 4 scores for court ordered changes aforementioned: child changes, mother
changes, father changes, and total changes. These scores were also examined for ethnic differences.
Definition of Terms
Court Ordered Change - Major changes were defined as court orders to initiate bypass,
family maintenance, or family reunification services for the mother, father, and/or the family. The
remaining changes included those across 9 areas represented by 85 variables: psychological
evaluation, counseling services, health services, drug/alcohol testing, substance abuse/alcohol
treatment, parenting education, support groups, domestic violence treatment, and visitation. A court
ordered change could also include placing more restrictions or conditions on service
recommendations.
Qualitative Analysis
A separate focus group with representatives from the Court Office Unit was conducted
toward the end of the study in order to obtain feedback on the implications and meanings of the
quantitative findings. The focus group with the Court Office Unit contained 9 participants,
including two Social Workers (22.2%) Ill's, five social work supervisors (55.6%), one Program
Manager (I 1. 1%) and one Legal Clerk (1 1. 1%). The mean number of years working in child
welfare among the Court Officer's focus group was 11.6 years and ranged from 4 to 22. Focus
group participants had a mean of 3.2 years in their current positions, with a range of .5 to 7 years.
Seven participants had MS W's (77.8%), one had a BSW (11.1 %), and one reported unspecified
postgraduate education (I 1.1 %). Two of the participants identified as Latino (22.2%), two identified
as Asian American/Pacific Islander (22.2%), four identified as White (44.4%), and one identified as
mixed race (11. I%). Eight of the participants were female (88.9%) and one was male (11.1 %).
111
Participants in the Court Officer Unit focus group were asked a series of questions designed
to provide insights into the "story behind the statistics." Specifically, Court Officers were invited to
share their experiences in relation to overall factors that contribute to court ordered changes and
comment on specific findings from quantitative data. Please see Attachment 6 for the specific
questions used to guide the-Court Officer Unit focus group discussion.
Notes from the Court Officer Unit focus group were transcribed and analyzed for key
themes related to court changes. In addition, focus group respondent interpretations and insights
about results of the quantitative data were examined and summarized.
Quantitative Results:
Analysis of Court Ordered Changes
Frequency of Court Ordered Changes
Of the 403 cases, 178 (44.2%) recorded at least one court ordered change from the original
social worker recommendations from the j urisdictionaVd_ispositional hearing, 200 (49 .6%) abided
by the original recommendations, and 25 (6.2%) were not available for evaluating the additional
court ordered changes.
Court Ordered Changes and Child's Ethnicity
We eliminated the cases with missing information to do our comparison by ethnicity,
leaving us with 378 valid cases. There was no significant relationship between the existence of a
court ordered change and ethnicity. Please see Table 28.
t me Groupa
Table 28 Proport1ons ofC ourt 0 rderedChanges b1y Eh.
White
Latino
Total
African
American
Court Change
8
Asian
American/PI
Other
No
200
(52.9%)
25
(53.2%)
65
(51.6%)
68
(51.9%)
18
(62.1%)
24
(53.3%)
Yes
178
(47.1%)
22
(46.8%)
61
(48.4%)
63
(48.1%)
11
(37.9%)
21
(46.7%)
Based on 378 cases with valid infonnation
Types of Service Changes
We examined the types of court changes and distinguished three to be relatively major given
their impact on the case: bypass, family maintenance, and family reunification. Of the 178 cases,
the most common change with these three types was the implementation of family reunification
services: 27 cases (152%) involved family reunification, 2 (1.1 %) cases regarded bypass, and 1
case (0.6%) regarded family maintenance. Ethnic comparisons were conducted, but no relation~hips
were found with these major services. Please see Table 29.
112
Table 29 Bypass, Family Maintenance, and Family Reunification Court Changes by Ethnic Group8
Total
African
White
Latino
Asian
Other
American
American/PI
Bypass
No
Yes
11
20
(95.2%)
0
1
(4.8%)
176
(98.9%)
22
(100.0%)
61
(100.0%)
62
(98.4%)
(100.0%)
2
0
(0.0%)
0
(0.0%)
1
(1.6%)
(0.0%)
(1.1%)
Family
Maintenance
(99.4%)
(100.0%)
22
60
(98.4%)
63
(100.0%)
(100.0%)
11
21
(100.0%)
1
(0.6%)
0
(0.0%)
1
(1.6%)
0
(0.0%)
0
(0.0%)
0
(0.0%)
No
151
(84.8%)
16
(72.7%)
55
(90.2%)
52
(82.5%)
10
(90.9%)
18
(85.7%)
Yes
27
(15.2%)
6
(27.3%)
6
(9.8%)
11
(17.5%)
1
(9.1%)
3
(14.3%)
No
Yes
177
Family
Reunification
a Based on 178 cases with court change information
Court Changes Regarding Children, Mothers, and Father
Four scores were calculated to describe the number of changes imposed by the court, each
for the child, mother, and father, and finally a total score representing the entire case. The number
of changes related to the child's circumstances ranged from Oto 4 with an average of 0.31 (sd =
0.65). Court ordered changes related to the mother were most common, ranging from Oto 8
changes with an average of 0. 77 (sd = 1.28). The number of changes related to the father's
circumstances ranged from Oto 7 with an average of0.67 (sd = 1.26). The number of total court
ordered changes related to a case ranged from 1 to 16 with an average of2.53 (sd = 2.22). No
significant ethnic differences were found regarding the number of court-imposed changes. Please
see Table 30.
113
Table 30: Number of Court Chan9:es by Ethnicity'a
Total
African
White
American
Child Related
Changes
Avg.
0.31
0.23
0.38
(sd)
(0.65)
(0.53)
(0.78)
Mother
Related
Changes
Avg.
(sd)
Latino
Asian
American/PI
Other
0.27
(0.54)
0.36
(0.50)
0.33
(0.73)
0.77
(1.28)
1.32
(l.94)
0.69
(0.99)
0.83
(1.30)
0.18
(0.40)
0.57
(1.29)
Father Related
Changes
Avg.
(sd)
0.67
(1.26)
0.82
(1.92)
0.61
(1.14)
0.62
(1.10)
0.55
(1.21)
0.90
(1.30)
Total Case
Changes
Avg.
(sd)
2.53
(2.22)
3.36
(3.36)
2.39
(1.63)
2.54
(2.39)
1.82
(1.33)
2.43
(2.01)
a Based
on 178 cases with court change information
Qualitative Results:
Analysis of Court Ordered Changes
Quantitative findings were presented to a focus group made up of9 representatives of the
Agency's Court Officers Unit for discussion and feedback. Quantitative results suggested that
approximately half of the cases had indicated a court ordered change from the recommendations of
the social worker describedfrom the jurisdictional/dispositional hearing. From our focus group
with members from the Court Officer Unit, the consensus was that the interpretation of the
measured proportion of court ordered changes found in the quantitative analysis, depended on what
we were actually measuring. The proportion is ·dependent on a variety of circumstances as to why
there was a change from the social worker recommendations including clerical errors needing
correction, changes due to modifications in the law, simple (albeit common) adjustments to services
such as visitation terms in addition to actual substantial changes due to conflicting opinions
concerning the direction of the case could all be related to a change between social worker
recommendations and court ordered.
Findings from the focus group with the Court Officer Unit provide additional information
regarding changes made to social worker recommendations at the jurisdictional/dispositional
hearing. First, capturing the details needed to conclude why social worker recommendations may
or may not be followed is extremely difficult. This is because variations by judge, courtroom
114
environment, social worker credibility and reputation, assigned lawyer (i.e., court appointed vs.
privately retained), as well as formal and informal case plan negotiation tactics frequently occurring
would complicate research. Second, the personal dynamics among the stakeholders (socials
workers, lawyers, judges, and families), including the strength of their positions and philosophical
standpoint can be such an influence that systemic conditions may be poor predictors of case
outcomes and success as they are related to the judicial process. Third, that despite the complex
circumstances involved, judges, guided by social workers, determine the best possible service plan,
while being aware of interpersonal dynamics and individual characteristics that may obscure
reasonable decisions. According to the focus group, social workers have a significant influence in
the court process and usually receive a majority of services they request for their clients.
Quantitative data indicated that few changes involved bypass, family maintenance, or family
reunification. The proportion of changes in these main service areas seemed reasonable to the focus
group. The low percentage of change in bypass, FM, and FR assignment should be low given the
legal criteria to qualify for such a change, as well as the intricate legal procedures.
When presented with quantitative findings that indicated no relationship between ethnicity
and court ordered changes in major service areas, or the overall number of court-ordered changes,
it was the opinion ofmembers of the focus group that ethnicity does not contribute to decisions in
court, but rather it is circumstances that drive the case.
Qualitative Methods:
Relationship Between the Child Welfare and Court Systems
Qualitative results describing the relationship between the child welfare and court systems
are gleaned from the agency-wide focus groups. These focus groups were structured to explore
practices at different choice points in the child welfare system. Please see page 61 for a full
description of agency-wide focus group methods and results.
Qualitative Results:
Relationship Between the Child Welfare and Court Systems
Results from agency-wide focus groups (please see page 61 for a full description) suggest
that the relationship between DFCS workers and district attorneys andjudges is problematic at
times. Many participants described difficulties in effective collaboration between the court system
and the child welfare system. Social workers suggested that problems between the court system and
the child welfare system may reflect fundamentally different perspectives on the needs and
circumstances of children and families in the CWS, and perhaps unrealistic expectations regarding
service plans. For instance, one social worker noted, "From a client's side, they are being told by a
court system and a social system that the way that they were brought up and the way that things
were done in their [clients'] family is wrong and not acceptable," noted a social worker focus group
participant. Several participants suggested that courts directly and indirectly set an unrealistic
standard for families, exemplified by one participant's comments:
"The court system just makes these orders and that's it. They don't consider anything else
except for this narrow focus on child and parent. The social worker has to look at resources
for the child, for the family, and language barriers. The court doesn't consider the impact it
115
has on the family, who has to pay for the services, shrinking or expanding resources. The
court is in an isolated bubble anything outside that bubble is not in their framework "
Regarding the services ordered, another social worker commented, the "Court wants it all
done at once ...the court doesn't understand that for whatever reason the client may be only capable
of doing so much within the first 6 months." Additionally once the court does order a service or
case plan, social workers report that it is difficult to adjust mandated services "because the court
requires justification for removing the order; also it is difficult convincing the DA of [the need for]
new recommendations."
In general, focus group and interview participants expressed a need to bridge these gaps
between the court system and the child welfare system. Although some social workers described
having strong collaborative relationships with the court system, many described tensions with
attorneys and other court representatives in relation to recommendations for mandated services and
disposition of cases. Social workers also conveyed that it was common for court representatives to
dismiss their psychosocial family assessments and push for recommendations that the social
workers deemed disconnected from the reality and needs of families. "Legal aids rarely visit the
children or the parents, they need to be more reality-based," commented on worker. "Social workers
and the court system are both looking at the case from different views," noted another. Some
workers depicted the district attorneys as having more of a focus on "winning" a case than on
ensuring the best outcomes for the children and family.
Working with the DA is a challenge. Their role is to advocate for the child-but they often
do only one interview and they don't know the particulars ofthe case. There can be
antagonism between the social worker and the DA-there should be less antagonism
between the DA and the social worker and more cooperation.
Social workers also expressed concern about the fairness ofthe system. For example, "In court
cases, private attorneys are more likely to have their client's case dismissed compared to court
appointed attorneys." Many participants conveyed that many court representatives and social
workers exhibit a "gap between standards that common people have and that educated professional
people have," which may disadvantage low-income families and families of color. "Attorneys.look
at the information provided by the social worker and make sweeping generalizations and judgments
about families," summarized one social worker. One supervisor captured concerns about failing to
recognize the context of client's lives:
When there is child abuse or neglect we raise our eyebrows to it, we go in there and want
everything to change but we 're not providing a basicfoundationfor change. We 're not
developing the community, we 're not providing better housing. We tell people to go out and
get suitable housing, but we 're not providing it.
Language issues also impact the experience offamilies in the court. "Families are unable to
communicate with their lawyers because of language barriers; English only attorneys can't
communicate with a monolingual Spanish speaking client," summarized one worker. Other workers
observed that "Families may not like to leave a message for an attorney because of the language
factor" and "the court process can be confusing for English speaking families, but more so to
Spanish speaking families." Furthermore, "the court reports are all in English and the clients don't
have a translated version that they can look at."
116
Additionally, social workers expressed concern that client outcomes are often dependent on
working relationships or attitudes ofDA 's. "There are some DA's that I have a great rapport
with-but a few are very judgmental and I feel like they are not supportive of the family." The
DA's are perceived as having substantial power. "The DA's word is very strong in court because
they are supposed to be speaking on behalf of the child," stated one social worker. Other social
workers suggested that some DA's pressured social workers to change recommendations or would
attempt to discredit them in the courtroom, particularly when the DA's case was weak.
Agency-wide focus group results also provided recommendations regarding ways to enhance
collaborative relationships between the court system and the child welfare system. These
recommendations included,
•
•
•
•
•
Explore opportunities to develop standards for practice between the courts and the child welfare
system including development of shared policies, practices and procedures. Such efforts could
help address common concerns of social workers that generally called for a "switch from courtcentered practice to family/community-centered practice."
Develop a systems level mechanism for problem solving to address conflicts between social
workers and DA's and "set a minimum standard or tone for what is appropriate behavior in the
court."
Facilitate development of role clarity between child welfare and the courts to address
widespread concern that decision-making is often not based on the social worker assessments
and, instead, is often based on perceptions ofDA's who may not have spent sufficient time with
children and who may not have evaluated family strengths.
Facilitate a dialog with both court and child welfare workers about standards for "good enough"
parenting and the dynamics of poverty in decision-making.
Explore opportunities for support of social workers in the court process through training,
enhanced supervision, or coaching from experienced colleagues.
Summary of Key Findings and Implications:
The Role of the Court System in Child Welfare Practice
In Phase 3, we examined the role of the court system in child welfare practice by exploring
the types of court ordered changes that are made to social worker recommendations at the initial
jurisdictional/dispositional hearing, as well as the relationship between the child welfare and court
systems. Quantitative results indicated that the court made changes to the initial social worker
recommendations at the jurisdictional/dispositional hearing in approximately halfof the cases.
Ethnicity was not related to whether or not a change occurred. These results do not necessarily
indicate that the system and courts treat children and families from different ethnicities the same,
but rather that when the judge decides to make a change from the social worker recommendations,
these changes appear to occur equally across ethnic groups. It was the opinion of members of the
Court Officer Unit focus group that ethnicity does not contribute to decisions in court, but rather it
is circumstances that drive the case.
Findings from the focus group with the Court Officer Unit suggest that many factors
complicate an accurate measurement ofwhy social worker recommendations may or may not be
117
followed. Factors that may impede accurate assessment of the agreement between child welfare and
court personnel on service plan recommendations include the strong influence of individual
stakeholders involved, and formal and informal case plan negotiation tactics all of which paint a
much more complicated picture.
Qualitative findings from the agency-wide focus groups suggested that collaboration
between the child welfare and court systems is problematic. Agency-wide focus group participants
felt that the child welfare and court system have different perspectives on the needs and
circumstances of children and families in the CWS and that the court system may have unrealistic
expectations of families.
Additionally, many agency-wide focus group participants described the ways in which child
welfare and district attorneys and judges interact as ineffective. Some social workers felt that some
district attorneys and judges would dismiss their assessments and recommendations and may try to
pressure social workers to change their recommendations. A somewhat similar finding was noted by
Knepper and Barton, (1997) who found that although judges tended to accept social worker
recommendations, the relationship between social workers and the court system plays a significant
role in their decisions. Their study found that when social workers adhered to the group norms of
the court, judges rewarded them by not ordering "unrealistic" practices and allowing reasonable
time frames for mental health evaluations to be completed. Similarly, agency-wide focus group
findings seem to suggest that if social workers go against the implicit rules ofthe court system, .then
judges may override their recommendations, but ifthey adhere to expectations ofthe court system
than recommendations are accepted.
Limitations and Recommendations
From these exploratory findings, issues arose regarding limitations and recommendations.
First, although quantitative and qualitative information described many of the circumstances behind
why social worker recommendations are not upheld by the court, it is very difficult to ascertain the
degree to which each circumstance influences a court ordered change. Second, better and more
exhaustive record keeping may not be the solution, as many informal and undocumented actions in
the negotiation process occur regularly, and are confidential, thus making much of the data
unavailable for research and evaluation purposes. Third, assessing other factors such as courtroom
environment, stakeholder characteristics, and other judicial issues would be advantageous in later
research. Although obtaining this information is a formidable task, these factors should be included
in the future, as measuring primary systemic factors no longer seem adequate to explain outcomes
in the CWS.
118
V.
STATEWIDE COMPARATIVE ANALYSIS
The statewide comparative analysis of effective practices for children and families of color
in the child welfare system included a statewide survey of child welfare directors, managers or
supervisors in California counties who were asked to describe effective practices for children and
families of color in their child welfare systems. In addition, this statewide comparative analysis,
included a quantitative analysis that identified the link between these practices and levels of
disproportionate representation of children of color in county child welfare systems. As such, two
overall analyses were included in the statewide comparative analysis of effective practices for
children and families of color in the child welfare system, 1) descriptive information from the
statewide survey on the types of practices, practices and practice characteristics identified by survey
respondents, and 2) multivariate results that link types of practices, practices and practice
characteristics to over or underrepresentation of children of color within counties. Additionally,
contextual information on the estimated proportions of children of color in county populations and
county child welfare systems within selected regions of California is provided, as well as
information on the estimated proportion of children within the overall county population who are in
the CWS.
Methods
The statewide survey on effective practices for children and families of color in the child
welfare system was designed to gather information on 1) key practices in California counties that
may have an impact on children and families of color, and 2) practices considered most promising
for children and families of color. For the purposes of the survey, "practices" were defined broadly
to include programs, services, strategies, policies, or tools.
Data Collection
Child welfare directors, managers or supervisors were mailed an invitation to participate in
the survey, and a copy of the survey. Two weeks later, a follow-up email that contained the survey
as an attachment and telephone calls were made to encourage participation. Counties that failed to
respond to the survey received approximately 3 additional follow-up emails or telephone calls
requesting their participation or a referral to someone else in the agency who might be able to
participate.
Respondents were offered four options for returning the survey to the CWRT, either by 1)
completing the survey on their computer and emailing it as an attachment, 2) completing the survey
over the telephone with a member of the CWRT, 3) faxing the completed survey, or 4) mailing the
completed survey. Surveys that were completed over the telephone were tape recorded in order to
ensure the accuracy of information and once information was verified, the tapes were erased. Fortyfive counties responded to the survey, resulting in a 77.6% response rate. Of the 45 counties who
completed the survey, 23 returned it via email, 10 by telephone, 7 by fax, and 5 returned the survey ·
by mail.
119
Sample
The sample included child welfare directors, managers or supervisors from California
counties. Participants for the statewide survey were identified through the 2003 County Welfare and
Social Service Directory. An invitation to participate and a copy of the survey were mailed to each
child welfare director, manager or supervisor in each of the 58 counties in California. Of the 13
counties that declined to participate, 4 indicated that their child welfare population is not diverse
and they did not have any special programs for children and families of color; 3 replied that the
survey was too time consuming, 3 did not reply to the survey, 2 indicated that they would complete
the survey, but did not return it and one county indicated that they were goirig through a reorganization and could not complete the survey because of these changes.
Instrumentation
A survey entitled "Statewide Survey on Effective Practices for Children and Families of
Color in the Child Welfare System" was developed and tested in a pilot study involving five
counties (please see Attachment 7 for a copy of the Statewide Survey). Feedback provided from the
pilot counties resulted in a reduction in the number of questions and minor changes in wording. The
survey contains questions related to caseload characteristics of counties and the most effective
practices for children and families of color in each county's child welfare system. The survey asks
respondents to provide demographic information on children in various components of the county's
child welfare system. Respondents are also asked to identify up to five of the most effective
practices for children and families of color in their county's child welfare system, ranking them
from #1 (most effective), #2 (second most effective) and so on. Multiple choice questions on the
most effective practice (#1) are also included in the survey.
Analysis
Quantitative analyses were conducted using the Statistical Package for the Social Sciences
(SPSS) software. Descriptive results were generated by running frequencies and percentages for
variables of interest. Multivariate results were generated by conducting controlled tests of empirical
associations between types of practices, practices, and practice characteristics and levels of
disproportionate representation of children of color in county CWS's.
Qualitative data describing the most effective practices were analyzed through content
analysis methods. The types of practices within each of the four categories were identified and
descriptive information about their target population, description and reasons why_these practices
are effective with children and families of color were delineated.
Contextual Tables:
County Child Populations and CWS Involvement·
The following tables provide contextual information on selected county child populations
and CWS involvement by ethnic group. Please note that the counties presented in these tables are
not necessarily the counties who responded to the statewide survey. Also please note that because
the most recent information available on county child populations is for the year 2001, information
on CWS populations for the year 2001 is also provided. As such, current 2003 proportions of
120
children of color in California county CWS may vary slightly from the 2001 proportions provided
in the following tables.
This descriptive information is presented in order to provide a context through which to
view the results from the statewide survey. Because compiling information on all 58 California
counties would be cumbersome, information on selected counties is provided. Table 31 a provides
information on the estimated proportions of children of color in county populations and the CWS
within 13 Bay Area counties. Table 31 b described estimated proportions of children of color in
county populations and the CWS within selected counties by California regions. Last, Table 31 c
provides information on the estimated proportion of children within the population in the CWS
·
among selected counties by region.
121
I . Popu laf10n and Ch"ld
We lfiare S1ysem
t
W"th·
T able 31 a: E sf1mat ed Propo rf10ns o fCh"ld
I
I m Bay A rea C ounf1es
I reno fC oorm
Ethnic Group
County
White
African American
Latino
Asian
Native American
%in
%in
%in
%in
%in
%in
%in
%in
%in
%in
Population
cws Population
cws
Population
cws
Population
cws
Population
cws
Alameda
33.7%
18.0%
19.0%
66.1%
24.6%
11.9%
22.3%
2.9%
0.4%
0.7%
54.1%
38.5%
11.4%
Contra Costa
45.2%
20.2%
12.7%
13.9%
2.0%
0.4%
0.5%
Marin
71.3%
35.7%
3.2%
19.4%
18.7%
9.4%
6.6%
3.3%
0.2%
0.3%
Mendocino
67.3%
62.1%
0.7%
4.9%
24.7%
18.5%
1.4%
0.2%
6.0%
12.9%
Monterey
30.1%
4.3%
14.2%
52.3%
5.8%
32.0%
57.6%
1.7%
0.4%
0.2%
Napa
67.4%
1.5%
11.8%
61.2%
32.3%
17.4%
4.3%
2.8%
0.6%
0.0%
San Benito
43.9%
32.1%
0.5%
1.2%
53.1%
60.7%
2.1%
1.2%
0.4%
1.2%
San Francisco
27.6%
11.6%
12.9%
70.4%
22.2%
12.3%
37.0%
4.8%
0.2%
0.8%
San Mateo
38.5%
26.7%
4.3%
29.8%
33.0%
32.0%
23.9%
9.4%
0.3%
0.1%
Santa Clara
35.4%
27.4%
3.5%
12.2%
32.8%
52.0%
28.1%
7.4%
0.2%
0.8%
Santa Cruz
53.8%
46.1%
1.1%
3.4%
41.0%
49.5%
3.8%
0.2%
0.3%
0.2%
41.2%
42.4%
49.5%
15.0%
18.9%
9.1%
16.0%
Solano
3.7%
0.6%
0.6%
Sonoma
70.6%
57.4%
2.0%
9.0%
21.8%
28.8%
4.5%
1.0%
1.0%
3.2%
State of
California
37.7%
30.3%
7.1%
31.3%
42.9%
33.8%
11.7%
2.8%
0.5%
1.1%
Notes:
I. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in January 2001 according to Needell, B.• Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A.,
Lery, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. & Kim, H. (2003). Child Welfare Services Reports for California. Retrieved (8/29/2003), from University of California at Berkeley Center for Social
Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>.
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org.
3. The Bay Area Counties are the 13 identified by the CWDA, and are also categorized as "big" counties in California except San Benito which is "small."
.
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California County Data Book summarized information on children aged 0-17
whereas the CWS data may include children as old as 19. Some totals may not equal l 00% due to missing data or rounding.
122
• and Ch.Id
.
. Popu Iat1on
.
Ta ble 31b Est1mate
dP roport1ons o fCh.ld
1 We lfiare S,ystem w·Ith.m Se lectedC ount1es B,y Reg1on
1 reno fC o lorm
Ethnic Group
County
White
African American
Latino
Asian
Native American
%in
%in
%in
%in
%in
% in
%in
%in
% inCWS
% in
Population
cws Population cws Population cws Population cws Population
Bay Area
33.7%
18.0%
19.0%
24.6%
11.9%
22.3%
2.9%
0.4%
Alameda
66.1%
0.7%
27.6%
12.9%
22.2%
12.3%
37.0%
4.8%
0.2%
0.8%
San Francisco
11.6%
70.4%
52.0%
28.1%
Santa Clara
35.4%
27.4%
3.5%
12.2%
32.8%
7.4%
0.2%
0.8%
Mountain
0.8%
10.1%
4.7%
0.8%
0.0%
86.3%
91.4%
0.4%
2.3%
Calaveras5
0.8%
3.6%
12.0%
12.0%
3.2%
1.8%
Placer
83.3%
79.0%
0.7%
0.8%
2.8%
17.9%
17.1%
14.9%
3.1%
Sacramento
53.7%
43.4%
12.5%
35.2%
1.0%
1.0%
Northern
8.9%
13.8%
7.9%
7.5%
1.3%
1.7%
6.8%
Butte
75.1%
73.9%
1.8%
78.4%
1.0%
2.3%
8.0%
5.1%
4.3%
0.3%
8.2%
Humboldt
65.6%
22.4%
79.3%
1.1%
1.2%
15.0%
26.8%
1.1%
0.0%
3.5%
61.0%
8.5%
Lassen5
Southern
10.8%
2.9%
0.2%
60.1%
38.9%
0.4%
Los Angeles
19.8%
15.2%
9.0%
42.3%
Oran,ie
40.9%
39.1%
1.8%
43.3%
48.0%
13.7%
0.2%
6.8%
5.3%
0.2%
35.0%
San Die,io
46.0%
35.1%
7.1%
24.5%
36.6%
9.9%
3.0%
0.5%
2.2%
42.2%
4.8%
42.2%
41.5%
2.4%
15.1%
50.2%
1.0%
0.4%
0.2%
Santa Barbara
Valley
31.1%
28.0%
5.7%
22.3%
48.5%
45.9%
13.8%
2.1%
0.8%
1.7%
Fresno
44.9%
44.0%
15.0%
35.9%
35.0%
3.9%
13.9%
6.4%
0.4%
0.1%
Merced
0.8%
0.5%
53.4%
50.3%
2.3%
13.0%
35.1%
32.4%
8.4%
2.6%
Stanislaus
State of California
37.7%
30.3%
7.1%
31.3%
42.9%
33.8%
11.7%
2.8%
0.5%
1.1%
Notes:
I. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in January 2001 according to Needell, B., Webster, D., Cuccaro-Alamin, S., Annijo, M., Lee, S., Brookhart, A.,
Lery, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. & Kim, H. (2003). Child Welfare Services Reports for California. Retrieved [8/29/2003), from University of California at Berkeley Center for Social
Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>.
2. Children in the population refer to those aged Oto 17 according to the California County Data Book 2001, Children Now, www.childrennow.org.
3. Superscript "s" denotes a CWDA categorized "small" county as opposed to "big" county.
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California County Data Book summarized infonnation on children aged 0-17
whereas the CWS data might include children as old as 19. Some totals may not equal I00% due to missing data or rounding.
123
Table 31c:
Estimated Proportion of Children Within the Population in the Child Welfare System Among
.
Selecte d Count.1es B R eg1on
Region and County
Number of Children
Percentage of Child
inCWS
Pooulation in the CWS
Bay Area
San Mateo
875
0.47%
Monterey
641
0.52%
San Benitos
84
0.54%
Napa
178
0.60%
Marin
361
0.70%
Solano
821
0.71%
Sonoma
941
0.83%
Santa Cruz
588
0.88%
Santa Clara
4,401
0.94%
Contra Costa
3,061
1.29%
Alameda
6,033
1.54%
San Francisco
3,050
1.97%
Mendocino
628
2.77%
Mountain
Placer
843
1.28%
Calaverass
128
1.33%
Sacramento
8,436
2.45%
Northern
Lassens
82
1.11%
Humboldt
1.27%
389
Butte
827
1.62%
Southern
Santa Barbara
516
0.48%
Orange
7,316
0.89%
9,935
1.19%
San Die20
Los Ane;eles
52,630
1.80%
Valley
Stanislaus
1280
0.89%
Merced
822
1.11%
Fresno
4757
1.80%
State of California
139,823
1.41%
Notes:
1. Children in Child Welfare System refer to those in a service component of ER, FR, FM or PP in January 2001 according to
Needell, B., Webster, D., Cuccaro-Alamin, S., Armijo, M., Lee, S., Brookhart, A., Lery, B., Shaw, T., Dawson, W., Piccus, W., Magruder, J. &
Kim, H. (2003). Child Welfare Services Reports for California. Retrieved [8/29/2003], from University of California at Berkeley Center for
Social Services Research website. URL: <http://cssr.berkeley.edu/CWSCMSreports/>
2. Children in the population refer to those aged 0 to 17 according to the California County Data Book 2001, Children Now, www.childrennow.org
3. Superscript "s" denotes a CWDA categorized "small" county as opposed to "big" county
4. Estimated proportions were calculated combining data from the two above sources. Proportions are also estimated given that the California
County Data Book summarized information on children aged 0-17 whereas the CWS data might include children as old as 19.
124
Descriptive Results:
Effective Practices for Children and Families of Color
Two counties did not provide information on effective practices; as such, a total of 43
effective practices for children and families of color in the child welfare system were identified.
Practices identified by the respondents as the most effective (# 1) for children and families of color
were divided into one of four categories: 1) beginning choice point practices (N = 13), 2) continuing
or ending choice point practices (N = 12), 3) organizational or structural practices (N = 14) and 4)
both beginning and continuing choice point practices (N = 4). Descriptive results for each of these
four categories are presented separately. In addition, descriptive statistics on the location, target
population, goals, history, factors related to implementation, factors that support the practice and
barriers to successful implementation of the practice are presented.
It should also be noted that many counties reported that they do not target services to any
specific racial/ethnic group and that the practices they identified as being most effective with
children and families of color, were also in general, their most effective practices for their overall
child welfare population.
Beginning Choice Point Practices (N = 13)
Thirteen counties identified practices that take place at early choice points as their most
effective practices for children and families of color in the CWS. The 13 beginning choice point
practices identified by the counties fell into three categories of practices: 1) family group
conferencing, or other team decision-making services (n = 7, 53.8%) 2) voluntary family
maintenance or other prevention/family preservation services (n = 4, 30.8%), and 3) risk assessment
tools or systems to guide decision-making (n = 2, 15.4%). Please see Table 32.
Family group conference or other team decision making
ractice
Voluntary family maintenance or other prevention/family
reservation service
Risk assessment tool ors stem
7
53.8%
4
2
30.8%
15.4%
Table 33 provides qualitative data on the beginning choice point practices that are related to
family group conferencing or other team decision-making practices. Most respondents described the
target population as children and families who have been detained or are at risk for court-ordered
services, although some respondents described using family group conferencing for all families
involved with the CWS, not just those at risk for court ordered out-of-home placement.
Most respondents described family group conferencing or other team decision-making
practices as involving the family, extended family or the support system in the group conference.
Some respondents described staff meetings or multidisciplinary team meetings that did not involve
the family. Survey respondents described the practice as a collaborative effort with the goal of
producing a plan for the child and family. Many respondents described the process as strengths
based and focused on how best to help children and families.
125
In general, respondents felt that the practice was effective with children and families of color
because it involves clients in the decision-making process. This inclusive model of practice was
described as particularly helpful for children and families of color because they may be "more
distrustful of the system," and also because inclusive practice may be more culturally appropriate
with some families. Additionally, team meetings not involving family members were described as
being beneficial to children and families of color because more accurate decisions are expected
from bringing numerous perspectives together.
Table 34 provides qualitative data on voluntary maintenance or other prevention/family
preservation practices. The target population for these services included families at risk for courtordered out-of-home placements, as well as one program aimed at expectant mothers and parenting
women. These practices were described as intensive in nature and focused on preventing child
maltreatment and out-of-home placements.
Voluntary family maintenance or other prevention/family preservation services were
described as effective with children and families of color because their goal is to keep children with
their birth families. Additionally, these services were viewed as effective for children and families
of color because they are collaborative in nature, involving other systems or organizations such as,
Native American tribes, mental health and probation systems.
Table 35 provides qualitative data on risk assessment tools or systems. These practices are
targeted to all.children reported to the child welfare system for suspected maltreatment. Risk
assessment practices were described as tools to help make decisions at critical points by assessing a
number of different risk factors.
Risk assessment tools or systems were described as effective with children and families of
color because they help to assess risk in an objective way, thereby minimizing potential bias based
on race/ethnicity.
Table 36 provides quantitative results on all of the beginning choice point practices
identified from the statewide survey. In general, beginning choice point practices are most often
delivered at the county child welfare office (n = 7, 53.8%); they tend to be targeted to children of all
ages (n = 12, 92.3%); as well as families of any ethnicity (n = 13, 100.0%). Families with substance
abuse problems (n = 9, 69.2%) are also targeted for these beginning choice point practices, as are
families with domestic violence, (n = 9, 69.2%), families with a mental health problem (n = 9,
69.2%), families with income problems (n = 8, 61.5%), and families with a disability (n = 8,
61.5%).
The most frequently identified goal of the practices was to prevent out-of-home placement
(n = 11, 84.6%); other goals included to assess level ofrisk (n = 10, 76.9%); improve psychosocial
functioning of family (n = 10, 76.9%); address substance abuse problems of family (n = 8, 61.5%);
address socioeconomic problems of family (n = 8, 61.5%); improve parenting skills, (n = 7, 53.8%);
address mental health problems of family (n = 7, 53.8%); ensure a permanent placement for the
child (n = 7, 53.8%), and reunify children with their birth families (n = 6, 46.2%)
Table 37 provides data on the history of the beginning choice point practices. Most were
first implemented in response to a particular problem (n = 7, 53.8%) and most were first initiated by
126
the director of the child welfare system (n = 8, 61.5%). The year the practices were first
implemented ranged from 1992 to 2002, and 15.4% (n = 2) of respondents reported that the practice
has changed significantly between now and the time it was first implemented.
Table 38 provides information on factors related to implementation of beginning choice
point practices. The majority of these practices are permanent (n = 10, 76.9%); most are funded
through state funds (n = 7, 53.8%) and 38.5% (n = 5) are mandated. Most social workers are
provided with training in how to implement the practice (n = 12, 92.3%); 15.4% (n = 2) of counties
contract with another agency or system to deliver the practice and 69.2% (n = 9) indicated that the
practice is widespread throughout the child welfare system. No respondents reported that the
practice is delivered in a specific geographic area or community. The majority of respondents (n =
8, 61.5%) reported the practice has been evaluated to determine its effectiveness and of those, 87 .5
% (n = 7) indicated that the practice is effective with the target population. Of those counties where
the practice had not been evaluated, 80.0% (n = 4) indicated that an evaluation is not needed.
Factors that support the practices and barrier to successful implementation are presented in
Table 39. The majority ofrespondents felt that families and children help support the successful
implementation of the practice (n = 10, 76.9%). Support from the mental heath system was also
frequently noted (n = 9, 69.2%); as was support from the substance abuse system (n = 8, 61.5%),
contract agencies (n = 7, 53.8%), county government officials (n = 7, 53.8%), and evaluations that
show practice is successful (n = 7, 53.8%). Few barriers to successful implementation were noted,
38.5% (n = 5) of respondents indicated that there were no barriers to successful implementation, and
38.5% (n = 5) also reported unstable funding is a barrier.
127
Table 33: Beginning choice point practices: Family group conference or other team decision-making practice (N = 7)
Target
Description of practice
Name of
Why effective with children and families of color
practice
population
Family
Available to all
The Family Group Conference is a tool for solving
I believe that this practice works well for families
Group
consumers who
family problems that is based on a simple, traditional
of color because it is based on principles shared
Conferencing have family,
belief: The combination of family strengths and
by those families. It involves much extended
extended family, community support can keep children safe and well
family, and is centered around a meal (family
friends, support cared for. A Family Conference is a gathering of family alone time) .. It is inclusive of all family members,
members, friends, and community specialists who join
including elders. Meetings can be held at the
together to improve the care and protection of a specific location that best fits the needs of the family.
child or children.
Children of all
Family
Family Conferencing is a program that uses strengthThis program allows families to assist in the
Conferencing ages
based methodology by allowing families to participate
decision process as to what is best for the family.
in the decision-making process as to how best help their
children. Family Conferencing is a gathering of
extended family members, other intimate support
people, and service providers in order to share
information and give the family an opportunity to
create a plan for their children.
Involving the extended family, non-relateq extended
Family
All families
These families are often even more distrustful of
Group
involved with
family and all service providers in a conference in
the "system" than most families; this practice
Conferencing CWS at any
which the family takes responsibility for making a plan allows the family to exert more control over the
stage of the
for the child/
direction of services and the planning for the
children.
process can use
this program
Children at risk This program targets all children (including children of Team decision-making/goal setting/evaluation
Youth Pilot
of placement, or color) at risk of placement, or at risk of being moved to process which includes the family and all support
Project
(YPP)
at risk of more
a more restrictive placement. The program uses monies persons they wish to include, as well as the
restrictive
that would be spent on foster care to provide services to various service providers working with the
family.
placement
maintain the child in the home or in a less restrictive
placement. The program uses a team decision-making
and goal-setting model, which includes the family and
supports.
128
Table 33 continued: BeJ?;inning
Target
Name of
practice
population
Placement
All detained
Meetings
children and
their families
choice point practices: Family group conference or other team decision-making practice (N = 7)
Description of practice
Why effective with children and families of color
After a child is detained, a placement meeting is held as
soon as possible to determin the "next best placement"
for that child. In addition to Social Workers and other
staff, parents and family members are included
whenever possible in this process, so that their input
can be provided in this decision. The safety of the child
and culturally relevant placements are goals ..
All children,
Detention
No child is detained by CPS or remains detained by law
Case Staffing CPS may
enforcement without being discussed as soon as
consider
possible in a staffing. The participants include the line
detaining or who ER worker in the matter, ER Supervisor, On-going
have been
Case Supervisor, and CPS Program Manager. The ER
detained by law worker presents the issues of protection, risk, current
enforcement in
situation, the court worker/supervisor, lends legal
an emergency
documentation evidence perspective, all others ask
questions, add comments with the ultimate goal of
consensus on whether to proceed with intervention and
what the intervention looks like.
All ER referrals Consensus decision making with workers, key program
Team
considered for
supervisors, mental health clinician, PI-IN, family
Assessment
voluntary or
Planning
decision meeting coordinator, and often other service
court-ordered
providers. Type & level of CWS intervention
services
determined after exploring safety/risk, family
strengths/needs, and available
resources/services/strategies. This is a collaborative
process which may be utilized for other key case
decisions.
Families and relatives are able to participate in
the decision making and are able to provide input
in to the "next best placement" for a child, which
can include placement with relatives or in a
culturally relevant home if the child cannot be
safely returned to his or her parents.
Many perspectives and opinions are brought to
bear at inception, promoting more objectivity and
balance regarding FACTS and EVIDENCE of
risk and best interests of the child. Less
possibility of stereotyping, subjectivity, ignorance
of cultural factors.
This decision-making forum is used for ALL
cases moving to on-going services. The core
group attends pretty consistently and represents
multiple perspectives, resulting in more consistent
and well thought out decision making regarding
case handling and service needs.
129
Table 34: Bee;inning choice point practices Voluntary family maintenance or other prevention/family preservation service (N = 4)
Target
Description of practice
Why effective with children and families of color
Name of
population
practice
Families at high Social worker and social work assistant are dedicated to American Indian Tribe has facilitated family
Voluntary
conferencing for tribal families.
risk of court
program. Services are based on strength-based and
Family
Maintenance intervention due family-focused approaches, including family
conferencing and family de_cision-making. Promoting
Program
to abuse and
Safe and Stable Families funding used for contracting
neglect issues.
services to include psych assessments, intensive case
management services, and a half-time family advocate.
High Risk
Expectant
Program in which pregnant mothers with a CPS history Missing
Mothers and
or involvement coupled with drug or alcohol use are
Infant
children under
targeted for intensive services and treatment to avoid
Program
placing their children into the foster care system.
age 1
Mothers of newborns are also a target population for
this program.
Family Preservation practice is an integrated,
Built a community consensus around the effective
Family
Abused and
comprehensive approach to strengthening and
service delivery model for Family Preservation in
Preservation neglected
the County; Supported public/private
families at
preserving families who are at risk of or already
collaborative planning and service delivery
imminent risk of experiencing problems in family functioning with the
goal of assuring the physical, emotional, social,
placement
efforts; Preventive-oriented; Collaboration with
other County Departments such as DPSS, Mental
educational, cultural and spiritual development of
Health and Probation; Strong and continued
children in a safe and nurturing environment.
leaderhip within the Department of Children and
Families Services.
Provide intervention and make referrals to other
We do not have foster homes in our county. We
Missing
Keep
agencies so that family can remain intact, if possible.
have seen children go through foster care system
families
and the majority do not do well.
together
130
Table 35: Beginning choice point practices Risk assessment tool or system (N = 2)
Description of practice
Name of
Target
population
practice
Research-based & validated tools for decision-making
All new intake
Structured
at critical points in child welf~re and Foster Care
to CPS
Decision
systems developed by the Children's Research Center
(currently) will
Making
in collaboration with COSS & California Counties.
System
soon extend to
(SDM)
all active cases
The risk assessment process looks at the child factors,
Best practice All children
including age, disability, illnesses as well as the
reported
risk
parent's ability to care-take, the history of abuse, also
assessment
speaking with the child, the parent and collateral
during the
contacts and then making a decision. We do not have a
ER phase
universal risk assessment tool, the workers are trained
in this process. We have an ER team that does this
practice.
Why effective with children and families of color
Objective nature of factors assessed.
Children are assessed in an objective wayconsidering their factors that place them at risk,
rather than being targeted for their ethnicity.
131
Table 36: Beginning choice point practices: Location, target population and goals (N = 13)
Frequency (Percentage)
Where does practice #1 take place?
7 (53.8%)
• County child welfare office
(30.8%)
4
• Throughout the child welfare system
2 (15.4%)
• Birth family's home
2 (15.4%)
• Contract agency site
1 (7.7%)
Foster
family's
home
•
I (7.7%)
• Another out-of-home placement site
2 (15.4%)
• Other
Who is the target population for practice # 1?
12 (92.3%)
• Children of all ages
(7.7%)
I
• Very young children
(100%)
13
• Families of any ethnicity
9 (69.2%)
• Families with substance abuse problems
9 (69.2%)
Families with domestic violence
(69.2%)
9
• Families with a mental health problem
8 (61.5%)
• Families with income problems
8 (61.5%)
• Families with a disabilitv
6 (46.2%)
• Other
What are the purposes and goals of practice #1?
11 (84.6%)
• Prevent out-of-home placement
10 (76.9%)
• Assess level of risk
10 (76.9%)
• Improve psychosocial functioning of family
(61.5%)
8
Address substance abuse problems of family
(61.5%)
8
Address
socioeconomic
problems
of
familv
•
(61.5%)
8
• Address housing problems of family
(53.8%)
7
• Improve parenting skills
(53.8%)
7
• Address mental health problems of family
(53.8%)
7
• Ensure a permanent placement for the child
(46.2%)
6
Reunify
children
with
their
birth
families
•
(7.7%)
1
• Other
•
•
132
Table 37: Beginning choice point practices: History of practice (N = 13)
Frequency (Percentage)
Why was practice# 1 first implemented?
7 (53.8%)
• In response to a particular problem
4
(30.8%)
• Funds became available
4 (30.8%)
• New policies were implemented
Who first initiated practice # 1?
8 (61.5%)
• Director of the child welfare system
4 (30.8%)
• Child welfare workers
3 (23.l %)
Other
•
When was practice #1 first implemented?
1
(7.7%)
• 2002
3 (23.1%)
• 2001
(7.7%)
1
2000
•
2 (15.4%)
• 1999
1
(7.7%)
• 1998
(7.7%)
1
• 1995
1
(7.7%)
• 1992
Has practice # 1 changed significantly between the time it
was first implemented and now?
2 (15.4%)
• Yes
133
Table 38: Beginning choice point practices: Factors related to implementation (N = 13)
Frequency (Percentage)
Current implementation status of practice # 1
IO (76.9%)
• Permanent
(23.1%)
3
Dependent
on
re-funding
•
What is the primary funding source for practice #1?
(53.8%)
7
• State funds
(38.5%)
5
• Federal funds
(23.1%)
3
County
funds
•
(15.4%)
2
• Other
Is practice # 1 mandated?
(38.5%)
5
• Yes
Are social workers provided with training in how to
implement practice # 1?
12 (92.3%)
• Yes
Do you contract with another agency or system to deliver
practice # 1?
2 (15.4%)
• Yes
Is practice # 1?
9 (69.2%)
• Widespread throughout the child welfare system
4
(30.8%)
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
(0.0%)
0
• Yes
Has practice # 1 been evaluated to determine its
effectiveness? .
(61.5%)
8
• Yes
If yes, what were the results of the evaluation?
(87.5%)
7
• Effective with target population
(12.5%)
1
Other
•
If practice #1 has not been evaluated, why not?
(80.0%)
4
• An evaluation was not considered to be needed
(20.0%)
1
• Other
134
Table 39: Beginning choice point practices: Factors that support practice and barriers to successful
implementation (N = 13)
Frequency (Percentage)
What are the major factors that support practice # 1
10 (76.9%)
• Support from families and children
(69.2%)
9
• Support from the mental health system
(61.5%)
8
• Support from the substance abuse system
(53.8%)
7
• Support from contract agencies
(53.8%)
7
• Support from county government officials
(53.8%)
7
Evaluations
show
that
practice
is
successful
•
(46.2%)
6
• Support from the community
(46.2%)
6
Support from state government officials
(46.2%)
6
Stable funding
(46.2%)
6
Support from the health system
(30.8%)
4
Support
from
the
criminal
justice
system
•
(30.8%)
4
• Support from foster families
3
(23.1%)
• Support from other out-of-home placement providers
What have been the barriers to successful implementation of
this practice?
(38.5%)
5
• No barriers
(38.5%)
5
Funding
is
not
stable
•
(15.4%)
2
• Insufficient support from child welfare workers
(7.7%)
1
• Insufficient support from management
1
(7.7%)
• Insufficient support from families and children
• Evaluations are inconclusive as to the success of the
(7.7%)
1
practice
•
•
•
135
Continuing Care/Ending Choice Point Practices (N = 12)
Twelve counties identified continuing care/ending choice point practices as their most
effective with children and families of color in the CWS. The 12 continuing care/ending choice
point practices identified by the counties fell into three categories of practices: 1) placement related
practices (n = 6, 50.0%); 2) the Family to Family Initiative (n = 4; 33.3%); and 3) intensive and
collaborative interventions for high-risk children (n = 2; 16.7%). Please see Table 40.
Table 40: Continuin care/endin
Intensive and collaborative interventions for high risk
children
2
16.7%
Table 41 contains qualitative data on placement related continuing care/ending choice point
practices. The target populations of these services varies, some are targeted to all children in out-ofhome care, others are targeted specifically to Native American children, one is targeted to children
in group homes (The Village Project) or at risk for entry into group homes and one is targeted to
parents with young children likely to enter out-of-home care (Shared Family Care).
Descriptions of the practices also varied. For instance, many participants noted that the first
priority is to place children with relatives; some discussed the role of culture and ethnicity in
making placement decisions. One practice {The Village Project) is described as a multidisciplinary
effort aimed at avoiding higher levels of care for children or working toward a lower level of care.
Lastly, one practice (Shared Family Care) is aimed at providing a placement for the whole family in
a mentor, foster home.
Respondents provided a variety of opinions regarding why the practice is effective with
children and families of color. Specifically, many mentioned the importance of placing children in
culturally appropriate homes-whether with relatives or with foster parents. The Shared Family
Care program is thought to be effective with children and families of color because it allows
families with few resources to be placed together and The Village Project was considered to be
effective with children and families of color because of a multicultural staff and flexibility of
funding sources.
Table 42 provides descriptive information on the Family to Family Initiative. The target
population for this practice varies somewhat; some respondents identified children and families in
foster care as the target population, others identified care providers and resource homes or all clients
and care providers. The Family to Family Initiative is described as engaging and involving
communities in preventing out-of-home placements, reunification, and providing resource homes
within children's neighborhoods. Many also described the practice as encompassing several aspects
of the child welfare system with the goal of improving the overall child welfare system.
136
Family to Family was described as effective with children and families of color because it
involves community and family support and is aimed at reducing disparities associated with gender,
and ethnicity and increasing the number of children placed in their own communities.
Table 43 provides qualitative data on continuing/ending choice point practices that are
aimed at intensive and collaborative interventions for high-risk children. The target population for
these services is children who are at risk for a more restrictive placement. These practices were
described as collaborative in nature-involving systems such as public health, community based
organizations, probation, behavioral health and faith based organizations. Interventions are
comprehensive, intensive and culturally competent.
Intensive and collaborative interventions for high-risk children were described as being
effective with children and families of color because the services are individualized to each family's
needs and are intensive in nature These practices were also described as "engaging families in a
non-threatening and respectful manner," and assessing the types of strengths and resources already
being used by the family. Children and families of color were described as benefiting from these
services because they are respectful of culture and involve a family-driven process
Quantitative results from survey responses are provided in Table 44. Most respondents
indicated that these continuing care/ending choice point practices take place throughout the child
welfare system (n = 7, 58.3%) and are targeted to children of all ages (n = 9, 75.0%), as well as
older children (11 to 18 years) (n = 5, 41.7%), young children (4 to 10 years), (n = 3, 35.0%), and
very young children (0 to 3 years) (n= 3, 25.0%). Most are targeted to families of any ethnicity (n =
10, 83.3%), with far fewer being targeted to African American families (n = 2, 16.7%), or American
Indian families (n = 2, 16.7%). Families with substance abuse problems were also identified as a
target population (n = 9, 75.0%) as were families with a mental health problem (n = 9, 75.0%),
families with domestic violence (n = 9, 75.0%), families with income problems (n = 7, 58.3%), and
families with a disability (n = 7, 58.3%).
The most frequently noted goal of the practice is to reunify children with their birth families
(n = 10, 83.3%), followed by ensuring a permanent placement for children (n = 8, 66.7%), and
preventing an out-of-home placement (n = 8, 66.7%). Other commonly occurring goals included
improving parenting skills (n =7, 58.3%); improving psychosocial functioning of family (n = 6,
50.0%); addressing socioeconomic problems of family (n = 6, 50.0%); and addressing housing
problems of family (n = 6, 50.0%).
Table 45 provides information on the history of the continuing care/ending choice point
practices. The majority of practices were first implemented due to new policies (n = 8, 66.7%); the
director of the child welfare system was most often identified as the person who first initiated the
practice (n = 11, 91.7%), followed by child welfare workers (n = 4, 33.3%). Many practices were
first implemented in 2001 (n = 4, 33.3%), and 33.3% (n = 4) had changed significantly since first
being implemented.
Factors related to implementation are presented in Table 46. The majority of practices are
permanent (n = 11, 91.7%), and are funded by state funds (n = 7; 58.3%). Of the practices, 41.7% (n
= 5) are mandated and 91. 7% (n = 11) of counties provide social workers with training on how to
implement the practices. Of the counties, 33.3% (n = 4) contract with another agency to implement
the practices, and 58.3% (n = 7) of respondents indicated that the practice is widespread throughout
137
the child welfare system. Very few respondents indicated that there is a specific geographic
community in which the practice is targeted (n = 2, 16.7%), and only 16.7% (n = 2) indicated that
the practice has been evaluated to determine its effectiveness. Of those practices that had been
evaluated, 100.0% (n = 2) reported that the practice is effective with the target population. For
practices that had not been evaluated, 33.3% (n = 4) have not been evaluated because of insufficient
resources.
Table 47 provides information on factors that support the practice and barriers to successful
implementation of the practice. The most commonly mentioned supportive factors included support
from child welfare workers (n = 11, 91.7%), support from management (n = 11, 91.7%), and
support from families and children (n = 11, 91.7%). Other factors that support the practice include
support from foster families (n = 9, 75.0%); contract agencies (n = 8, 66.7%); other out-of-home
placement providers (n = 7, 58.3%), the mental health system (n = 7, 58.3%), state government
officials (n = 7, 58.3%); and county government officials (n = 7, 58.3%). Many respondents did not
note barriers to successful implementation of the practice, no barriers were reported by 16.7% (n =
2) of respondents, and 16.7% (n = 2) reported that unstable funding is a barrier to successful
implementation.
138
Table 41: Continuing care/ending choice point practices: Placement related practices (n = 6)
Why effective with children and families
Description of practice
Name of practice Target population
of color
Our county is predominately Caucasian,
All children who
In all cases, we must first consider any willing,
Culturally
so children of color are already truly a
able and appropriate relatives for placement. If
require out-ofsensitive
this is not possible, we then try to place in a home minority. Placing them in a home that is
placements
home placement
sensitive to that is very important.
of the same ethnicity, religion, etc. If this is not
possible, then we let the foster parents know that
their culture is an important issue, and we are
expecting them to continue to allow the children
to live in a culturally sensitive environment.
For families with few resources it allows
Mothers/Fathers
This program provides placement for the family
Shared Family
Care
with young
in a mentor home. Usually one parent (mother in for placement of child and parent
together. Most of our families with few
children likely to
most cases) is subject to intervention by child
resources are those living in communities
welfare. Rather than place the child in foster
enter out of home
impacted by poverty, drugs, and crimes.
care, it is assessed that the mother is willing to
care
Families of color primarily live in these
participate in Shared Family Care and will abide
with the conditions of working and living with a
areas.
mentor family.
Children are kept in their communities
All Children
Identify as quickly as possible family members
Place children
with relatives of same culture and
for
placement
with relatives
practices promote stability for children
They are treated the same as all other
Native American Native Am/lCWA Attempt to place all children with families.
cases-dignity-compassion-provided
Counseling/assessment/service orientated.
children placed
services- acknowledgement of their
with families
concerns and problems.
through ICWA
139
Table 41 continued: Continuing care/ending cho.ice point practices: Placement related practices (n = 6)
Name of practice Target population Description of practice
Why effective with children and families
of color
Multidisciplinary effort to work with families and
S8163
Kids in group
A multicultural staff in the county and
program-"The
kids to avoid higher levels of care or get them
homes level 13 or
in the community-based providers and
village project"
14 or at risk for
flexibility of funding sources-money
into a lower level of care-like a foster home or
going into a group back into their home. The staff are bilingual and
can be used for activities outside the
home
bi-cultural and they are able to make connections group home, so it can be used for things
with multi-cultural families. They also work hand to prevent a group home placement and
and hand with a parent advocacy group-and they not just for kids already in a group home.
advocate for children and families especially kids
with educational difficulties.
ICWA-·
Native Americans When a Native American child is removed from
We are able to place many of our Native
placement of
the home of parents, we contact any relatives in
American children in homes that comply
Native American
with ICWA.
order to place the child in compliance with
children
ICWA.
140
Table 42: Continuing care/ending choice point practices: Family to Family Initiative (n = 4)
Name of practice Target population
Description of practice
Family to Family
All children in out
of home
placements
Family to Family
Initiative
All clients and
care providers
Family to Family
Initiative
Children &
families in foster
care
Family to family
Resource homes
and the
community
Regional initiatives underway to work with local
communities to prevent removal and/or maintain
children in their community/school if they must
be removed. Community collaboration on child
safety and well being.
The use of the following Family to Family
strategies are applied to all aspects of child
welfare placement practices in this County: Team
decision making, Recruitment, Retention, and
Support of Care Providers, Self-Evaluation and
Community Partnership.
Community-based, family focused and designed
to meet 9 key outcomes toward improving our
foster care system including reducing disparities
associated with ethnicity, gender & age.
This is an effort to engage the community in
providing resource homes for foster children who
are being placed in homes from their
neighborhoods in which they are being removed.
The community is being engaged and as well as
community based organizations.
Why effective with children and families
of color
Community and family
involvement/support.
Missing
Outcomes include: reduction in
disparities associated with sex, gender &
ethnicity, increasing kids placed in their
own neighborhoods & communities,
Missing
141
Table 43: Continuing care/ending choice point practices: Intensive and collaborative interventions for high-risk children (n = 2)
Name of practice Target population Description of practice
Why effective with children and families
of color
Children's
Children who are
To develop a child 'who is emotionally
As with all families, intensive services,
System of Care
in danger of being disturbed/behaviorally disturbed' through a
individualized plans
placed in a higher collaborative interagency system of care with
services that are "youth-centered;" family
level of care i.e.
Group homes
involved & focused; culturally competent;
community based and comprehensive.
Wraparound
Children who are
This practice really values families individually
The practice engages with families in a
services
placed or at risk of and values their culture. It provides intensive and respectful and non-threatening manner
placement into
supportive in-home services to keep the child at
and asking them what supports, what
home or return them home. It's also been a
level 14 group
people, in your life and in your
home care
collaborative effort between probation, behavioral community that are important to you and
health, public health, CBO's, and faith-based
how can we incorporate those aspects of
organizations.
your life into this child and family plan to
help your child stay safely at home. And
by asking those questions and evaluating
what the family uses to thrive and asking
them to make that a part of the service
planning for their child is kind of a
foreign way of working with the families.
It is a family-driven, needs based process,
rather than a professionally-driven
process. There is respect for the family's
culture.
142
Table 44: Continuing care and ending choice point practices:
Location, target population and goals (N = 12)
Where does practice # 1 take place?
Throughout the child welfare system
• County child welfare office
• Birth family's home
• Foster family's home
• Court
• Another out-of-home placement site
• Other
Who is the target population for practice # 1?
• Children of all ages
• Very young children (0 to 3 years)
• Young children (4 to 10 years)
• Older children (11 to 19 years)
Families of any ethnicity
African American families
• American Indian families
• Asian American/Pacific Islander families
• Latino families
• White families
Families of another ethnicity
• Families with substance abuse problems
• Families with a mental health problem
• Families with domestic violence
• Families with income problems
Families with a disability
• Other
What are the purposes and goals of practice #I?
• Reunify children with their birth families
• Ensure a permanent placement for the child
• Prevent out-of-home placement·
• Improve parenting skills
• Improve psychosocial functioning of family
• Address socioeconomic problems of family
• Address housing problems of family
• Address substance abuse problems of family
• Address mental health problems of family
• Assess level of risk
• Other
•
•
•
•
•
Frequency (Percentage)
7
4
4
4
4
4
(58.3%)
(33.3%)
(33.3%)
(33.3%)
(33.3%)
(33.3%)
(4L7%)
9
3
3
(75.0%)
(25.0%)
(35.0%)
(41.7%)
(83.3%)
(16.7%)
(16.7%)
(8.3%)
(8.3%)
(8.3%)
(16.7%)
(75.0%)
(75.0%)
(75.0%)
(58.3%)
(58.3%)
(50.0%)
5
5
IO
2
2
1
I
I
2
9
9
9
7
7
6
IO
8
8
7
6
6
6
5
5
4
4
(83.3%)
(66.7%)
(66.7%)
(58.3%)
(50.0%)
(50.0%)
(50.0%)
(41.7%)
(41.7%)
(33.3%)
(33.3%)
143
Table 45: Continuing care and endin2 choice point practices: History of practice (N = 12)
Frequency (Percentage)
Why was practice # 1 first implemented?
8
(66.7%)
• New policies were implemented
(25.0%)
3
• Funds became available
2
(16.7%)
• In response to a particular problem
(50.0%)
6
• Other
Who first initiated practice #1?
11 (91.7%)
• Director of the child welfare system
4 (33.3%)
• Child welfare workers
2 (16.7%)
• Mental health system
1
(8.3%)
• Foster families
1
(8.3%)
Substance
abuse
system
•
(8.3%)
1
• Contract a2encies
(8.3%)
1
• Criminal justice system
4 (33.3%)
Other
When was practice #1 first implemented?
1
(8.3%)
• 2003
2 (16.7%)
• 2002
4 (33.3%)
2001
•
2 (16.7%)
• 2000
Has practice # 1 changed significantly between the time it
was first implemented and now?
4 (33.3%)
• Yes
•
144
Table 46: Continuing care and ending choice point practices:
Factors related to implementation (N = 12)
Current implementation status of practice # 1
• Permanent
• Demonstration
What is the primary funding source for practice # 1?
• State funds
• County funds
• Federal funds
• Nongovernmental grant
• Other
Is practice #1 mandated?
• Yes
Are social workers provided with training in how to
implement practice # 1?
• Yes
Do you contract with another agency or system to deliver
practice # 1?
• Yes
• No
Combination of internal and external resources
-Is practice # 1?
• Widespread throughout the child welfare system
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
• Yes
Has practice # 1 been evaluated to determine its
effectiveness?
• Yes
If yes, what were the results of the evaluation?
• Effective with target population
If practice # 1 has not been evaluated, why not?
• Insufficient resources
• Insufficient time
• An evaluation was not considered to be needed
• Other
.-
Frequency (Percentage)
11
1
(91.7%)
(8.3%)
7
6
3
2
3
(58.3%)
(50.0%)
(25.0%)
(16.7%)
(25.0%)
5
(41.7%)
11
(91.7%)
4
4
3
(33.3%)
(33.3%)
(25.0%)
7
5
(58.3%)
(41.7%)
2
(16.7%)
2
(16.7%)
2
(100.0%)
4
2
1
3
(33.3%)
(16.7%)
(8.3%)
(25.0%)
145
Table 47: Continuing care and ending choice point practices: Factors that support practice and
· barriers to successful implementation (N = 12)
Frequency (Percentage)
What are the maior factors that support practice # 1
11 (91.7%)
• Support from child welfare workers
11 (91.7%)
• Support from management
11 (91.7%)
• Support from families and children
(75.0%)
9
• Support from foster families
8 (66.7%)
• Support from contract agencies
(58.3%)
7
Support
from
other
out-of-home
placement
providers
•
(58.3%)
7
• Support from the mental health system
(58.3%)
7
• Support from state government officials
(58.3%)
7
• Support from county govemmenfofficials
(50.0%)
6
• ·Stable funding
(50.0%)
6
Support
from
the
community
•
(50.0%)
6
• Support from the health system
6.
(50.0%)
• Support from criminal justice system
5 (41.7%)
• Support from the substance abuse system
(25.0%)
3
• Evaluations show that practice is successful
(16.7%)
2
• Other
What have been the barriers to successful implementation of
this practice?
(16.7%)
2
• No barriers
(16.7%)
2
• Funding is not stable
1
(8.3%)
• Insufficient support from state government officials
Insufficient
support
from
county
government
•
(8.3%)
I
officials
(8.3%)
1
• Insufficient support from child welfare workers
(8.3%)
I
• Insufficient support from foster families
(8.3%)
I
Insufficient
support
from
the
community
•
(8.3%)
I
• Insufficient support from the criminal justice system
(50.0%)
6
• Other
146
Both Beginning and Continuing/Ending Choice Point Practices (N = 4)
Four respondents identified practices that were related to both beginning and
continuing/ending choice points as their most effective practices for children and families of
color in the CWS. The four beginning and continuing/ending choice point practices fell into two
categories: 1) community-based services (n = 3, 75.0%); and 2) strength based assessments (n =
1, 25.0%). Please see Table 48.
Table 48: Both beginning and continuing/ending choice point practices: Types of practices (N =
4
Table 49 provides qualitative information on community based services. The target
population for these services includes low-income populations, families at risk for entry into the
child welfare system, families already in the child welfare system, and the community. These
practices were described as being located in the communities in which clients reside and include
a wide range of services such as, health, substance abuse, public assistance, domestic violence
and home parenting services. Practices were also described as collaborative and aimed at
preventing an out-of-home placement or reunifying children with their families.
These practices were described as effective with children and families of color because
services are "user friendly;" clients have easy access to many services that are housed in one
location. Community based services are free and outreach for these services is performed
through advertisements into different communities. Moreover, because these services are
typically located within clients' own communities, services are "community-specific" and clients
may run into their own community members while accessing services, thus making the services
less intrusive for clients. Additionally, these practices were described as flexible and able to
address "changing community needs."
Table 50 provides information on family-centered strength based assessments. This
practice was described as being targeted to Hispanic, African American, American Indian and
Asian American/Pacific Islander clients. It is focused on family strengths with an empowerment
perspective.
Quantitative results from survey responses are presented in Table 51. Practices that are
both beginning and continuing/ending practices typically take place at a contract agency site (n =
2, 50.0%), or the birth family's homes (n = 2, 50.0%). One hundred percent (n = 4) of these
practices are targeted to: children of all ages, families of any ethnicity, families with substance
abuse problems, families with a mental health problem, families with domestic violence, families
with income problems, or families with a disability. Seventy five percent (n = 3) of practices
have the following goals: preventing an out-of-home placement; reunifying children with their
birth families; assessing level of risk; improving psychosocial functioning of families; improving
parenting skills; addressing substance abuse problems; addressing mental health problems;
addressing socioeconomic problems; and addressing housing problems.
147
Information on the history of the practices is presented in Table 52. The practices were
first implemented in response to a particular problem (n = 1, 25.0%); funds becoming available
(n = 1, 25.0%); or new policies being implemented (n = 1, 25.0%). Child welfare workers (n = 2,
50.0%); the child welfare director (n = 2, 50.0%); and contract agencies (n = 2; 50.0%) were
identified as having first initiated the practices. The year the practice was first implemented
ranged from 1967 to 1997, and 66. 7% (n = 2) of respondents indicated that the practice had
changed significantly since it was first implemented.
Factors related to the implementation of both beginning and continuing/ending choice
point practices are presented in Table 53. All practices (n = 4, 100.0%) are permanent and most
are primarily funded through federal funds (n = 4, 100.0%). Fifty percent of respondents (n = 2)
indicated that the practice is mandated and 33.3% contract with another agency to deliver the
practice. Seventy five percent of respondents indicated that the practice is widespread throughout
the child welfare system. All respondents (n = 4, 100.0%) reported that the practice has been
evaluated, and 50.0% (n = 2) indicated that evaluations demonstrated that the practice is effective
with the target population.
Factors that support the practice and barriers to successful implementation are presented
in Table 54. One hundred percent of respondents (n = 4) indicated that support from the
following entities are major factors that support the practices: community, county government
officials, mental health system, substance abuse system, health system, and families and
children. Seventy five percent of respondents (n = 3) indicated that unstable funding is a barrier
to successful implementation of the practices.
148
Table 49: Both be_ginnin_g and continuini /ending practices: Community based services
Name of
Target population
Description of practice
practice
Community- Low-income
Federal and state funded programs that are free to
based
populations
low-income clients, including the health
services
department, substance abuse treatment, Aids
department, WIC, immunizations, planned
parenthood. Services include domestic violence
support, clothing donations, utility assistance,
HEAD Start program; sexual assault crisis, and in home parenting programs, and teenage parenting
programs.
Family
Children and Families at There are many Family Resource centers. A
Resource
risk for entry into the
collaborative of agencies applies for funds from the
Centers
CWS and children and
funding source (though RFP). The Family
families already in
Resource Centers have a nun;iber of agencies
CWS. Most clients are
working together within one site - so it provides
just one-stop for families to receive services. The
Hispanic, the second
most are API
Family Resources Centers are located in the
communities and the neighborhoods where the
clients live. Services are meant to be accessible and
less intimidating than county _government services.
Community, Families
Commmunity specific family based services located
Family
and Children
in six geographical areas in the County with embed
Resource
access to public child welfare, mental health,
Centers
alcohol and other drug and physical health services.
Why effective with children and families of
color
Because services are free and they advertise
into all areas of the community, and there are
satellite offices in many locations; so clients
have access to the services, which is
extremely helpful. They also cross-refer so we
all depend on one.another. We have a large
Hispanic population so information is
available in Spanish. Transportation is
accessible.
Because wherever the families are they can
reach out to the resources that meet their
need-there are a multitidue of agencies in
one place-it's a one-stop shop. Clients are
more likely to run into their own community
members at the Family Resource Centers. It is
user friendly and less intrusive than other
services.
Practice is community specific with flexibility
to address changing community needs.
Subsequently, it offers increased accessibility
to services, intervention/prevention and
family support and advocacy.
149
Table 50: Both beginning and continuin, lendine practices: Streneth based assessments
Description of practice
Target population
Name of
practice
Family
Hispanic, AA, American Focused on family strengths with the empowerment
perspective.
Indian and Asian
centered
strength
based
assessments
Why effective with children and families of
color
Missing
150
Table 51: Both beginning and continuing care/ending choice point practices: Location, target
population and goals (N = 4)
Frequency (Percentage)
Where does practice # 1 take place?
2
(50.0%)
• Contract agency site
2
(50.0%)
• Birth family's home
1
(25.0%)
County
child
welfare
office
•
1 · (25.0%)
• Foster family's home
(25.0%)
1
• Another out-of-home placement site
2
(50.0%)
• Other
Who is the target population for practice #1?
4 (100.0%)
• Children of all ages
4 (100.0%)
• Families of any ethnicity
(100.0%)
4
Families
with
substance
abuse
problems
•
4 (100.0%)
Families with a mental health problem
4 (100.0%)
Families with domestic violence
4 (100.0%)
• Families with income problems
4 (100.0%)
• Families with a disability
I
(25.0%)
Other
•
What are the purposes and goals of practice #1?
3 (75.0%)
• Prevent out-of-home placement
3 (75.0%)
• Reunify children with their birth families
3 (75.0%)
Assess
level
of
risk
•
3 (75.0%)
• Improve psychosocial functioning of family
(75.0%)
3
• Improve parenting skills
3 (75.0%)
• Address substance abuse problems of family
3 (75.0%)
• Address mental health problems of family
3 (75.0%)
• Address socioeconomic problems of family
3 (75.0%)
• Address housing problems of family
(50.0%)
2
• Ensure a permanent placement for the child
3 (75.0%)
• Other
•
•
151
Table 52: Both beginning and continuing care/ending choice point practices:
History of practice (N = 4)
Frequency (Percentage)
Why was practice #1 first implemented?
1 (25.0%)
• In response to a particular problem
1 (25.0%)
• Funds became available
1
(25.0%)
New
policies
were
implemented
•
1 (25.0%)
• Other
Who first initiated practice # 1?
2 (50.0%)
• Child welfare workers
2 (50.0%)
Director
of
the
child
welfare
system
•
2 (50.0%)
• Contract agencies
1 (25.0%)
• Clients of the CWS
1 (25.0%)
• Foster families
1 (25.0%)
• Out-of-home placement providers
1
(25.0%)
Mental
health
system
•
1 (25.0%)
• Substance abuse system
1 (25.0%)
• Health system
1 (25.0%)
• Criminal justice system
2 (50.0%)
• Other
When was practice # 1 first implemented?
1 (33.3%)
• 1997
(33.3%)
1
1995
•
1 (33.3%)
• 1967
Has practice # 1 changed significantly between the time it
was first implemented and now?
2 (66.7%)
• Yes
152
Table 53: Both beginning and continuing care/ending choice point practices: Factors related to
implementation (N = 4)
Frequency (Percentage)
Current implementation status of practice # 1
4 (100.0%)
Permanent
What is the primary funding source for practice # 1?
4 (100.0%)
• Federal funds
2
(50.0%)
• State funds
(50.0%)
2
County funds
(25.0%)
1
Nongovernmental grant
Is practice # 1 mandated?
2
(50.0%)
• Yes
Are social workers provided with training in how to
implement practice #1?
(50.0%)
2
• Yes
Do you contract with another agency or system to deliver
practice #I?
(33.3%)
I
• Yes
(33.3%)
1
• No
(33.3%)
1
Combination of internal and external resources
Is practice # 1?
(75.0%)
3
• Widespread throughout the child welfare system
(25.0%)
1
• Targeted to a relatively small group of clients
Is there a specific geographic area or community in which
this practice is targeted?
2
(50.0%)
• Yes
Has practice # 1 been evaluated to determine its
effectiveness?
4 (100.0%)
Yes
If yes, what were the results of the evaluation?
2
(50.0%)
• Effective with target population
• Evaluation did not provide evidence that practice was
1
(25.0%)
effective
(25.0%)
1
Other
•
•
•
•
•
•
153
Table 54: Both beginning and continuing care/ending choice point practices: Factors that support
practice and barriers to successful implementation (N = 4)
Frequency (Percentage)
What are the major factors that support practice # I
4 (100.0%)
• Support from the community
4 (100.0%)
• Support from county government officials
4
(100.0%)
• Support from the mental health system
4 (100.0%)
• Support from substance abuse system
4 (100.0%)
• Support from health system
4 (100.0%)
Support
from
families
and
children
•
(75.0%)
3
• Support from child welfare workers
(75.0%)
3
• Support from management
(75.0%)
3
• Support from foster families
(75.0%)
3
• Support from contract agencies
(75.0%)
3
Support
from
the
criminal
justice
system
•
(75.0%)
3
• Evaluations show that practice is successful
(50.0%)
2
• Support from state government officials
(25.0%)
1
• Stable funding
(25.0%)
1
• Support from other out-of-home placement providers
What have been the barriers to successful implementation of
this practice?
(75~0%)
3
• Funding is not stable
1
(25.0%)
• Insufficient support from state government officials
• Evaluations are inconclusive as to the success of the
(25.0%)
1
practice
154
Organizational or Structural Practices (N = 14)
Fourteen respondents identified organizational or structural practices as their most
effective practices for children and families of color in the CWS. The fourteen
organizational/structural practices fell into three categories: I) practices related to cultural
competency of staff (n = 10, 71.3%), 2) collaboration with American Indian tribes (n = 3,
21.3%), and 3) vertical case management (n = I, 7.4%).
Table 55: Or anization or structural
Collaboration with American Indian Tribes
Vertical case mana ement
3
I
Qualitative data on cultural competency practices are presented in Table 56. The target
population for these services varies, some are focused on the Latino population or the Spanish
speaking population, while others are focused on all children in the system and one is focused on
social workers themselves. Practices related to cultural competency involve the use of
bilingual/bicultural staff or competent translators in the delivery of services, as well as providing
information on civil rights, providing training to social workers and matching clients and
workers on ethnicity/culture.
Many respondents mentioned the importance of having bilingual/bicultural staff and the
need to be able to communicate with clients in their own language as reasons for why these
practices are effective with children and families of color. The importance of assessing and
delivering services in the clients' language is described as resulting in more accurate
assessments, more meaningful services and better connections between worker and client.
Cultural competency trainings for social workers are described as bringing awareness and
knowledge to staff.
Table 57 provides qualitative data on organizational or structural practices related to
collaboration with American Indian tribes. These practices are targeted to Native American
children and families and generally include multidisciplinary teams involving tribal
representatives, legal professionals ~nd child welfare staff who come together to better serve
Native American children and ensure the proper implementation of the Indian Child Welfare
Act.
Practices related to collaboration with American Indian tribes were described as effective
for children and families of color because they respect cultural heritage; encourage cultural
awareness for Native American children; result in more complete assessments; better service
planning and fewer contested hearings.
Table 58 provides qualitative information on vertical case management. This practice is
targeted toward all clients of the child welfare system and it involves having social workers carry
the same cases all the way through the system. This practice was described as effective with
children and families of color because there is consistency in the case and a better relationship
155
between worker and client. It was also noted that this practice is effective with the overall
population as well.
Quantitative data from survey responses is provided in Table 59. Most
organizational/structural practices take place throughout the child welfare system (n = 8, 57.1 %)
and are targeted to children of all ages (n = 12, 85.7%). Many are targeted to families of any
ethnicity (n = 7, 50.0%), while some are targeted to Latino families (n = 4, 28.6%), or American
Indian families (n = 3, 21.4%). Additionally, families with substance abuse problems are also a
focus of these practices (n = 7, 50.0%), as are families with mental health problems (n = 7,
50.0%). Goals of the practice include preventing an out-of-home placement (n = 7, 50,0%);
ensuring a permanent placement (n = 7, 50.0%); reunifying children and families (n = 6, 42.9%);
addressing substance abuse problems of families (n = 6,42.9%); and improving parenting skills,
(n = 6, 42.9%).
Information on the history of the practices is presented in Table 60. Most practices were
first implemented in response to a particular problem (n = 7, 50.0%); and the director of the child
welfare system was most frequently mentioned as the person who first initiated the practice (n =
7, 50.0%), followed by child welfare workers (n = 6, 42.9%). The year the practice was first
implemented range from 1972 to 2002 and 46.2% of counties indicated that the practice had
changed significantly since it had first been implemented.
Table 61 provides data on factors related to implementation. Most practices are
permanent (n = 12, 92.3%) and the majority are funded through federal dollars (n = 8, 57.1 %).
Most counties also indicated that these practices are mandated (n = 8, 61.5%), and 66.7% (n = 8)
of counties indicated that social workers are provided with training in how to implement the
practice. Most counties do not contract with another agency to deliver the practice (n = 9, 64.3%)
and most also indicated that the practice is widespread throughout the child welfare system (n =
12, 92.9%). Only one county (7.1 %) indicated that there is a specific geographic area in which
the practice is targeted. Four counties (28.6%) reported that the practice has been evaluated to
determine its effectiveness and of those 66.7% (n = 2) reported that the practice was found to be
effective with the target population. Of those counties in which no evaluation has been
conducted, 50.0% (n = 5) indicated that an evaluation was not needed.
Factors that sugport the gractices and barriers to successful imglementation are presented
in Table 62. Support from child welfare workers (n = 11, 78.6%) was most commonly noted as a
major factor that supports the practice, followed by support from management (n = 9, 64.3%);
county government officials (n = 8, 57.1 %); families and children (n = 7, 50.0%); and the
community (n = 7, 50.0%). No barriers to successful implementation were noted by 28.6% of
counties (n = 4), while 14.3% (n = 2) reported insufficient support from the community and
14.3% (n = 2) noted insufficient support from child welfare workers.
156
Table 56: Organizational or structural practices: Practices related to cultural competency (n = 10)
Why effective with children and families of color
Target
Description of practice
Name of practice
population
We have a very high rate of return for
One of things that has made it particularly
SpanishThe use of
effective in this county is the size of the county.
speaking
Hispanic children--one of our better
competent
clients
practices is the use of competent translators And our translator even though they may not be
translators
CW social workers they are from the community
and always having this available when
and they are familiar with the community and the
working with a family. These are used in
the juvenile court as well. And we also have dynamics of that particular culture. For the most
part this had enhanced the practice.
bilingual staff as well.
Availability of
bilingual/bicultural
staff to serve the
Latino population.
Latino
population
At least one
bicultural/bilingual
worker in each unit
Mostly
Hispanic
children
B ilingual/B icultural
social work staff.
Spanish
speaking
population
Hire and place bilingual/bicultural social
Cultural knowledge and ability to converse in same
workers at office serving majc;>rity of Latino language.
families. Provides culturally sensitive
services to families, better assuring accurate
asessments of risk and safety for children.
The agency really values language and
When you are doing social work and having to
trying to accommodate that the best we can. assess and discuss very difficult things with people
having to do with emotions and feelings these
Most of our receptionists are bilingual and
our social service aides and social worker
things are very difficult to do through translation
I's are bilingual. Whenever we start to
because a lot of what a social worker does is their
use of self and being able to read body language
implement a program we always look to
bringing in bilingual/bicultural staff. We are and intonation and understand the subtlety of
limited in our resources but we do our best. language and what someone is saying. You are less
effective through translation. The communication
We also really try to ask this of our
is much better when services are conducted in the
providers as well.
clients language. And bicultural folks understand
the local jargon.
Bilingual/bicultural staff assigned to
Ability to communicate with the families in their
bilingual/monolingual cases to ensure
native language & an ability to understand the
appropriate communication of issues, needs, culture helps the social worker provide appropriate
resources & guidance to the family.
services and agency function/process.
157
Table 56 continued: Organizational or structural practices: Practices related to cultural competency (n = 10)
Why effective with children and families of color
Target
Description of practice
Name of practice
population
Recruiting and hiring bilingual/cultural staff Recruitment is a priority within service delivery
B ilingual/B icultural Families
to work with the diverse population we
system and best practice philosophy .. .it becomes
whose
staff
integrated in every day work.
children
serve. Staff are assigned throughout our
have been
child welfare programs/services.
Assessment of language and cultural
abused.
staffing needs is determined on an ongoing
basis and is used as a criteria to fill vacant
positions.
Federal Division 21
Regulation-Civil
rights
All
Emphasis of
Cultural
Competence
All cultures
We inform all clients of their civil rights.
This includes that they cannot be .
discriminated against on the basis of
national origin, disabilities, race, age sex,
all those areas. We are supposed to talk to
all families about their civil rights. We ask
the family if they want a special skills
worker assigned-that would specialize in a
particular language or culture. We have
Black culture workers, Hispanic cultural
workers, and Asian, Russian special skills
workers. We ask all clients if they want a
special skills worker-this is for
language-but also culturally matching the
clients.
Our agency has a cultural competence
committee. We mandate 8 hours of training
for staff in orientation and an additional 16
hrs through our leadership academy.
Matching families with a social worker of their
ethnic background is helpful in getting families
back together. If the social worker understands the
culture-like if there are different discipline
techniques that are seen more in one culture, the
social worker may be able to help the child and
family more. Clients are a bit more comfortable,
social workers can confront them better-work
with the better to help bring about change.
Awareness, Knowledge
158
Table 56 continued: Organizational or structural practices: Practices related to cultural competency (n = 10)
Name of practice
Target
Description of practice
Why effective with children and families of color
population
Policy is to align
Applies
Our practice is to assign staff of similar
Ability to establish trusting relationship
case assignments to across all
ethnicity and we are able to do so most
cws
often in our African American and Mexican
staff from like
ethnicity when
components American population. If unable to do so,
possible
we consider assigning a spanish speaking,
non-Mexican-American. staff.
Cultural
Social
We require all social workers to attend
Missing
competence
workers
training on cultural competency on a regular
training
basis.
Civil Rights
Missing
Missing
Training- bringing awareness to staff
Training
159
Table 57: Organizational or structural practices: Collaboration with American Indian tribes (n = 3)
Target
Description of practice
Name of practice
Why effective with children and families of color
population
Indian Child
American
A practice in which tribal representatives,
It respects their cultural background and keeps
Indian
agency staff, kids' attorneys, County
Welfare Act
them connected with their tribal roots, it respects
(ICWA)
children and Counsel, and others come together on a
that people can do things differently and still have
families
Roundtable
monthly basis to talk about issues related to good outcomes. The tribe is actively solicited to be
the implementation of the ICWA, how we
involved with children and families in the CWS.
can work to together on certifying tribal
We pay for culturally appropriate services, like
tribal dances, and other non-traditional services.
homes, how to get services for American
Indian Children and what are the most
appropriate services. The group created a
handbook of ICWA regulations. Very
effective in opening dialogue between the
parties, sensitizing staff to issues around
services that would be more culturally
appropriate to American Indian children,
identifying and providing services, and
ensuring that kids can get to tribal homes.
ICWA
Native
Indian Child Welfare Act Multidisciplinary More complete assessments, better service
Team
planning and less contested hearings.
American
Some of the tribes have strong cultural
Collaboration with
Native
Native American families-in general they are
centers
with
health
care
and
counseling
and
tribes and tribal
Americans
still at war with both the US and with each other. It
so we try to work hard with them. Native
is effective because in spite of these things we are
courts
American children do best when placed
able to overcome the barriers and we are able to
with a Native American family. The Tribal
make use of the real resources of cultural
court makes placement decisions and so we awareness for Native American people.
work with them to make decisions that we
Collaborating with the tribes depends on having the
are all in agreement about. We work closely right social worker work with the tribes-in
with the tribes in Native American cases.
general the tribes have been cooperative with the
cws.
160
Table 58: Organizational or structural practices: Vertical case management (n = 1)
Name of practice
Target
Description of practice
Why effective with children and families of color
population
Social workers
All
Social Workers carry the same cases all the There's consistency-a lot of it is when cases are
carry a generic case
way through the system A lot is lost when a transferred from one person to another there may
load-vertical case
case is transferred from worker to worker.
be different interpretations of the situation.
management
Families always know who their social
Effective with overall population-it would be
workers are and there is ownership
unfair to say that what we are doing is more
throughout the life of the case. There is a
advantageous with group than another.
better work product. There's a relationship
between the worker and clients.
161
T a bl e 59 0 rgamzat1ona or structural practices: Location, tan et population and goals (N = 14)
Frequency (Percentage)
Where does practice #1 take place?
Throughout the child welfare system
(57.1%)
8
County child welfare office
7
(50.0%)
(42.9%)
6
• Birth family's home
Foster
family's
home
(42.9%)
6
•
(42.9%)
6
• Court
(35.7%)
5
• Another out-of-home placement site
(14.3%)
2
• Contract agency site
5 (35.7%)
• Other
Who is the target population for practice #I?
12
(85.7%)
• Children of all ages
(7.1%)
I
• Very young children (0 to 3 vears)
(7.1%)
I
• Young children (4 to IO years)
(50.0%)
7
Families
of
any
ethnicity
•
(28.6%)
4
• Latino families
(21.4%)
3
• American Indian families
(50.0%)
7
• Families with substance abuse problems
(50.0%)
7
• Families with a mental health problem
(42.9%)
6
Families
with
domestic
violence
•
4
(28.6%)
• Families with income problems
(21.4%)
3
• Families with a disability
(7.1%)
I
• Other
What are the purposes and goals of practice #I?
(50.0%)
7
• Prevent out-of-home placement
(50.0%)
7
• Ensure a permanent placement for the child
(42.9%)
6
Reunify
children
with
their
birth
families
•
(42.9%)
6
• Address substance abuse problems of family
(42.9%)
6
• Improve parenting skills
(35.7%)
5
• Assess level of risk
5 (35.7%)
• Improve psychosocial functioning of family
(35.7%)
5
Address
mental
health
problems
of
family
•
(35.7%)
5
• Address housing problems of familv
(28.6%)
4
Address
socioeconomic
problems
of
family
•
(57.1%)
8
• Other
•
•
162
Table 60: Organizational or structural practices: History of practice (N = 14)
Frequency (Percentage)
Why was practice #1 first implemented?
(50.0%)
7
• In response to a particular problem
(21.4%)
3
• New policies were implemented
1
(7.1%)
• Funds became available
(42.9%)
6
• Other
Who first initiated practice # 1?
(50.0%)
7
• Director of the child welfare system
6
(42.9%)
Child
welfare
workers
•
(7.1%)
1
• Clients of the CWS
(7.1%)
1
• Foster families
1
(7.1%)
• Criminal justice system
(7.1%)
1
• Contract agencies
(28.6%)
4
Other
When was practice # 1 first implemented?
(7.1%)
1
• 2002
2 (14.3%)
• 2000
(7.1%)
1
1999
•
2 (14.3%)
• 1998
(7.1%)
1
• 1995
1
(7.1%)
• 1980
1 (7.1%)
1972
Has practice #1 changed significantly between the time it
was first implemented and now?
(46.2%)
6
• Yes
•
•
163
T able 6 1: 0 rgamzational or structural practices: Factors related to implementation (N = 14)
Frequency (Percentage)
Current implementation status of practice # 1
12 (92.3%)
• Permanent
(7.7%)
1
• Other
What is the primary funding source for practice # 1?
(57.1%)
8
• Federal funds
(42.9%)
State
funds
6
•
(35.7%)
County funds
5
(28.6%)
4
Other
•
Is practice # 1 mandated?
(61.5%)
8
Yes
Are social workers provided with training in how to
implement practice # 1?
(66.7%)
8
• Yes
Do you contract with another agency or system to deliver
practice #1?
(64.3%)
9
• No
(21.4%)
3
Combination
of
internal
and
external
resources
•
(14.3%)
2
• Yes
Is practice #1?
(92.9%)
13
• Widespread throughout the child welfare system
1
(7.1%)
Targeted
to
a
relatively
small
group
of
clients
•
Is there a specific geographic area or community in which
this practice is targeted?
(7.1%)
1
• Yes
Has practice # 1 been evaluated to determine its
effectiveness?
(28.6%)
4
• Yes
If yes, what were the results of the evaluation?
(66.7%)
2
• Effective with target population
Evaluation did not provide evidence that practice was
(33.3%)
1
effective
If practice # 1 has not been evaluated, why not?
(35.7%)
5
• An evaluation was not considered to be needed
(21.4%)
3
• Insufficient resources
(7.1%)
I
Insufficient
time
•
(7.1%)
I
• Other
•
•
•
164
Table 62: Organizational or structural practices: Factors that support practice and barriers to
successful implementation (N = 14)
Frequency (Percentage)
What are the major factors that support practice # 1
11 (78.6%)
• Support from child welfare workers
9 (64.3%)
• Support from manag;ement
8 (57.1%)
• Support from county government officials
(50.0%)
Support
from
families
and
children
7
•
7 (50.0%)
• Support from the community
6 (42.9%)
• Support from foster families
6 (42.9%)
• Support from state R;ovemment officials
5 (35.7%)
• Support from the mental health system
4
(28.6%)
Support
from
the
substance
abuse
system
•
4 (28.6%)
• Support from contract agencies
(28.6%)
4
Support
from
the
health
system
•
4 (28.6%)
• Stable funding
3 (21.4%)
• Support from other out-of-home placement providers
2 (14.3%)
• Evaluations show that practice is successful
2 (14.3%)
• Support from the criminal justice system
2 (14.3%)
• Other
What have been the barriers to successful implementation of
this practice?
4 (28.6%)
• No barriers
2 (14.3%)
• Insufficient support from the community
2 (14.3%)
• Insufficient support from child welfare workers
(7.1%)
1
• Funding is not stable
7 (50.0%)
• Other
165
Summary of Key Findings and Implications:
Descriptive Results
Respondents to the California statewide survey identified practices in four overall areas
as most effective for children and families of color: 1) beginning choice point practices, 2)
continuing or ending choice point practices, 3) both beginning and continuing/ending choice
point practices, and 4) organizational or structural practices.
Beginning Choice Point Practices
The beginning choice point practices identified by respondents fell into three categories:
1) family group conferencing or other team decision-making practices, 2) voluntary family
maintenance or other prevention/family preservation services and 3) risk assessment tools or
systems.
In general, results from the California statewide survey indicate that beginning choice
point practices are most often delivered at the county child welfare office, and they are targeted
to children of all ages and ethnicities. Most of these practices are aimed at preventing an out-of-
home placement, assessing level of risk or improving the psychosocial functioning of the family.
Over half of these practices have been evaluated by individual counties to determine their
effectiveness, and of those that have been evaluated most are considered to be effective with the
target population.
Results of the California statewide survey indicate that beginning choice point practices
are considered to result in enhanced outcomes for children and families of color because they
are inclusive, collaborative and objective. Yet very little empirical evidence exists to support the
effectiveness of these practices for children and families of color. No studies have evaluated the
impact offamily group conferencing or other team decision-making practices on children and
families ofcolor. However there is some evidence to suggest that inclusive practice, in which
parents are engaged in placement and other decisions, may result in more placement stability for
children, yet possible differential effects for diverse populations has not been researched
(Palmer, 1996). More res~arch is needed to determine if family group conferencing/other team
decision-making practices and inclusive practice result in enhanced outcomes for children and
families of color.
Additionally, the effectiveness of voluntary family maintenance or other prevention/family
preservation services in preventing out-of-home placements has not been adequately
demonstrated, and the impact ofthese services for children and families of color has not been
sufficiently addressed in the research literature. It is interesting to note that results from the
California statewide survey indicate that only 15.4 percent of beginning choice point practices
take place in the birth family's home, yet the research literature suggests that home-based
prevention practices may be the most effective voluntary family maintenance practice, especially
for high-risk mothers. This finding may suggest that counties are not utilizing evidence-based
prevention efforts. Research on the reasons for this underutilization of home-based prevention
services is warranted.
166
Lastly, the effectiveness ofrisk assessment tools or systems in accurately assessing risk
for future maltreatment has not been consistently demonstrated in the research literature. Yet
research does indicate that the individual decision-making practices ofchild welfare workers are
inconsistent and subject to biases that may negatively impact children and families ofcolor. As
such, the use of more objective assessments of risk in child welfare is promising, however more
research is needed to identify risk assessment tools and systems that are reliable and valid.
Continuing Care/Ending Choice Point Practices
The continuing care/ending choice point practices identified by respondents fell into
three categories: 1) placement-related practices, 2) the Family to Family Initiative, and 3)
intensive and collaborative interventions for high-risk children.
Results from the California statewide survey indicate that, in general, the majority of
continuing care/ending choice point practices take place throughout the child welfare system
and are targeted to children of all ages and ethnicities. However, more continuing care/ending
choice point practices are targeted to specific age groups than the other practices identified by
respondents. The goal of most of these services is to reunify children with their birth families,
ensure a permanent placement or prevent and out-of-home placement. Only two respondents
reported that these practices have been evaluated to determine their effectiveness and both of
. these evaluations indicated that the practices are effective with the target population.
These continuing care/ending choice point practices are described as effective with
children and families of color because they are culturally appropriate, collaborative, intensive
and they involve clients' communities. Yet, research on the effectiveness of these services for
children and families of color is lacking. Placement-related practices that emphasize placement
of children in culturally appropriate homes have not been evaluated to determine their impact on
children of color. However, some studies have noted that social workers may be reluctant to
place children of color with White alternate caregivers (Brooks & James, 2003; Carter-Black
2002; Vidal de Haymes & Simon, 2003). Results from the California statewide survey regarding
the identification ofplacement-related practices that emphasize placement ofchildren of color in
culturally appropriate homes do suggest that culture and ethnicity may be important factors in
child welfare workers 'placement decisions. This appears particularly true for Native American
children who were mentioned specifically by two counties.
Research on the effectiveness ofthe Family to Family Initiative for children andfamilies
ofcolor has not been conducted. There is limited evidence to suggest that social network
interventions which seek to involve community members in services are potentially effective in
increasing social networks, improving parenting skills and increasing the likelihood of case
closure (Gaudin et al., 1990). However the Family to Family Initiative encompasses more than
social network interventions and research is needed to evaluate the effectiveness ofthis practice
for diverse populations.
Although children in out-of-home care are at an increased risk for a variety ofpoor
outcomes (Stevenson, 1999), the impact of intensive and collaborative interventions for high-risk
children have demonstrated only mixed results in the research literature. While some studies on
167
wraparound services report that children in these services are less likely to be removed from their
community and may experience a reduction in behavioral problems (Bruns, Burchard, & Yoe,
1995, as cited in Toffalo, 2000; Clark, Lee, Prange, & McDonald, 1996, as cited in Toffalo,
2000; VanDenBerg, 1993, as cited in Toffalo, 2000), other studies report no significant benefit
from wraparound services (Clarke, Schafer, Burchard, & Welkowitz, 1992, as cited in Toffalo,
2000). More research is needed to identify aspects ofthese interventions that are most effective,
especially for children ofcolor
Both Beginning and Continuing/Ending Choice Point Practices
Practices related to both beginning and continuing/ending choice points fell into two
categories: 1) community-based services, and 2) strength-based assessments.
Results from the California statewide survey suggest that many practices that are both
beginning and continuing/ending choice point practices take place at a contract agency site or
the birth family's home. These services are targeted to children of all ages and any ethnicity.
These practices have numerous goals including: preventing an out-of-home placement;
reunifying children with their birth families; assessing level of risk; improving psychosocial
functioning of families; improving parenting skills; addressing substance abuse problems;
addressing mental health problems; addressing socioeconomic problems; and addressing housing
problems. All four of the respondents indicated that these practices have been evaluated and 50
percent of the evaluations indicated that the practices are effective with the target population.
Respondents to the California statewide survey described these practices as effective with
children and families ofcolor because they are easy for clients to access; they are communityspecific, and non-intrusive. These services have broad goals that focus on both prevention and
intervention for families in contact with the CWS and research on the impact of these services on
children and families of color is lacking. In general, research on prevention and intervention with
child welfare populations does not provide convincing evidence of their effectiveness. However,
because these practices are more of an approach to services, rather than services themselves,
more research is needed to determine if these approaches are effective with ethnically diverse
populations.
Organizational or Structural Practices
Organizational or structural practices fell into three categories: 1) cultural competency
of staff, 2) collaboration with American Indian tribes, and 3) vertical case management.
According to responses from the California statewide survey, most
organizational/structural practices take place throughout the CWS; the majority are targeted to
children ofall ages and ethnicities, although some are specifically targeted to Latino or Native
American clients. The primary goals of these practices include preventing an out-of-home
placement and ensuring a permanent placement. Approximately one quarter of respondents
indicated that these practices have been evaluated to determine their effectiveness and of these,
approximately 66 percent reported that the practices are effective with the target population.
168
These practices are described by respondents to the California statewide survey as
effective with children and families ofcolor because they result in better communication and
relationships between worker and client; they respect cultural heritage; increase cultural
awareness and produce more accurate and complete assessments and services. Although
cultural competency appears intuitively beneficial for ethnically diverse children and families,
research on the impact of culturally competent staff and programs in producing successful
outcomes for children and families of color has not been conducted. There is evidence to suggest
that for Latino clients in particular, language barriers can exclude them from necessary services
and produce significant delays in service delivery; and researchers have also discussed the
importance of language in accurate assessment and treatment of clients (Suleiman, 2003). As
such, culturally and linguistically appropriate services appear to increase service use and
accuracy, although their impact on client outcomes is unclear.
Research on the effectiveness ofcollaboration with American Indian tribes and vertical
case management practices on outcomes for children and families of color is limited. Effective
collaboration between the CWS and American Indian tribes appears intuitively beneficial for
Native American children, however more research is needed to identify aspects ofeffective
collaboration and how this collaboration impacts outcomes for Native American children. A
vertical case management model of child welfare service delivery may result in enhanced
outcomes. For instance, in South County Offices, where a vertical case ·management model is
used, families are more likely to have their cases closed with family maintenance services than
they are in the Main County Office where a vertical case management model is not used (please
see Section IV for a full discussion). These results suggest that a vertfoal case management
model may be an effective child welfare practice.
Limitations and Conclusions
Descriptive results of the statewide survey must be considered within the context of
certain limitations. For instance, the survey instrument was designed specifically for this study
and so its reliability and validity is not known. The self-report nature of the data also presents the
possibility of personal biases in survey responses and since the professional titles of the
respondents varied from directors, to managers to supervisors, there is the possibility that
different perspectives and information may occur as a result of these differing job titles.
Additionally, although the response rate was high (77.6%), there is the possibility that the sample
may be biased due to some counties refusing to participate.
Overall, respondents described beginning choice point practices as effective with
children and families ofcolor because they are inclusive, collaborative and objective.
Continuing care/ending choice point practices were considered effective with children and
families of color because they are culturally appropriate, collaborative, intensive and involve
client's communities. Both beginning and continuing/ending choice point practices were
described as effective because they are easy for clients to access, they are community-specific
and non-intrusive. Lastly, organizational or structural practices were considered effective with
children and families of color because they are thought to result in better communication and
relationships because worker and client, they respect cultural heritage, increase cultural
awareness and produce more accurate and complete assessments and services.
169
In general, results from the California statewide survey suggest that there is a significant
gap between research on child welfare practices and the best practices for children and families
of color being used in California county child welfare systems. Little systematic research has
evaluated the impact of many of the most effective practices for children and families of color
that were identified by respondents Although the research literature has not yet provided
convincing evidence of the effectiveness of many ofthe practices identified by respondents,
respondents to the statewide survey felt that these practices were their most effective with
children and families ofcolor. These findings suggest that some counties may be utilizing
promising practices for children and families ofcolor in the child welfare system, and more
rigorous evaluations ofthese practices is needed.
·
170
Multivariate Analysis:
Impact of Statewide Practices on Differential Representation
The statewide survey identified practices, which addressed disproportionate
representation of children of color and were considered effective by county child welfare
personnel. Although perceived as effective, few of the practices had been evaluated to determine
effectiveness from an independent point of view and in a controlled manner. Child welfare
personnel nominated them as being effective, but in most cases their impact had not been
established through systematic empirical research.
For present purposes, the impact of interest is how the practices identified through the
survey might affect differential representation of children ofcolor. Some of the practices had
been implemented with children of color in mind and others had not, but all were considered by
child welfare personnel to be especially valuable in addressing the needs and circumstances of
children of color. If child welfare personnel are correct in their belief that the practices serve
children of color particularly well, then it is reasonable to believe that the practices identified in
the survey will divert children of color from the child welfare system or will reduce the scope or
duration of their involvement.
By helping children of color to follow pathways around full-fledged involvement in the
child welfare system or reducing the intensity or length of child welfare participation, the
practices identified in the survey can be expected to reduce differential representation. We can
expect lower levels of overrepresentation if the practices are effective.
Statewide data are available from sources, which, in conjunction with data from our
survey, permit the degree of empirical association to be determined between practices and
disproportionate representation. The U.S. Census and CWS/CMS publish data, available for use
by researchers and others, describing California counties on the worldwide web. The web-based
data provide an opportunity to evaluate whether there is a link between disproportionate
representation and use of perceived-to-be-effective practices.
If a study is statewide in scope like the present one, then it furnishes a basis to go beyond
the experience of a single county. By comparing groups of counties and contrasting those that
implemented a practice with those that did not, our investigation encompasses child welfare
systems around the state. The results point to promising practices identified from a statewide
perspective.
In this section, we present results from several analyses linking characteristics of
practices presented previously with disproportionate representation of children of color in the
child welfare system. Our intention was to identify types of practices, practices, and practice
characteristics-distinctions to be explained later-- associated with levels of disproportionate
representation, and which therefore.might be more effective in reducing disproportionate
representation of children of color.
171
Multivariate Methods
The unit of analysis for this study was the county. The method, broadly conceived, was to
conduct controlled tests of empirical associations between types of practices, practices, and
practice characteristics on the one hand, and levels of disproportionate representation on the
other hand.
For thinking about the study's approach in greater detail, it is useful to begin by
considering the meaning of disproportional representation. Disproportional representation is
said to occur when the proportion of children of color in a county's child welfare system is
higher or lower than expected. Typically, documenting the proportion of African Americans,
Latinos, or Asian A~erican/Pacific Islander in the population at large usually establishes the
proportion that is expected. The difference between the representation of a group in the
population at large and its representation in the child welfare system defines its level of
disproportional representation.
In our study we calculated a measure of this difference using regression analysis. We
analyzed a group's representation in the child welfare system after statistically adjusting for it
representation in the population at large. Following a research design like ours, a statistic results
from regression which defines the difference between expected and actual levels of child welfare
representation.
Disproportional representation is affected by factors other than those under the direct
control ofchild welfare authorities. The stresses under which minority families live and the
resources available to them for example, can contribute to disproportionate representation.
Counties can differ in the levels ofstress, resources, and similar differentiating characteristics.
Child welfare systems too are diverse. Some are larger and more centralized than others, for
example; policies and procedures vary.
It is difficult to establish definitively whether practices are effective because of the
differences apart from the practices between counties and child welfare systems associated with
disproportionate representation. When a link is observed between a practice and
disproportionate representation, it remains possible that a characteristic other than the practice
itself is responsible for that link. The observed link between a practice and disproportionate
representation might be coincidental. Controlling/or key differences between counties and child
welfare systems reduces the chances ofassociations occurring through coincidence.
We examined many indicators describing counties and their child welfare systems in
order to identify differences that were especially important to take into account. Among
indicators we screened were population size and density, wages and costs of goods and services
in the county, the geographical location of the county, size of the child welfare system and
proportion of the child welfare population in kincare. We sought key indicators from a large
number we considered reflecting differences between the counties and their child welfare
systems and linked to disproportionate representation.
172
Through this process we found two indicators that were consistently and strongly
associated with disproportionate representation: urban-rural status and per-capita income.
Counties that were urban and had higher per-capita income were notably more likely to have
higher levels ofdisproportionate representation.
We used these data to conduct a controlled analysis of the association between practice
types, practices, and practice characteristics by constructing regression equations. In each
equation, the proportion of Latino, African American, or Asian American/Pacific Islander
children served as the dependent variable. The proportion of Latinos, African Americans, or
Asian American/Pacific Islander in the county was entered first as an independent variable, as
discussed above, in order to subtract an expected level of representation from the actual
representation of each group in each county's child welfare system. We then added to the
equation e~ch county's urban-rural status and per-capita income. Finally, we added each practice
type, practice, or practice characteristic suitable for analysis.
This enabled us to determine how much each practice type, practice, or practice
characteristic was associated with greater or lower level of representation apart from the level of
representation expected from its group representation in the county, and after adjusting for the
county's urban-rural classification and its per-capita level of income. Thus, the approach
adjusted for the three key factors: group representation in the population, whether the county was
classified as urban or rural, and per-capita income of county residents.
In most cases the three control variables alone explained disproportionate representation
in a statistical sense extremely well. Success in accounting for differences between counties in
statistical models implies that many important differences between counties were explained and
that the possibility that associations arose by coincidence was low.
We considered practice types, practices, and practice characteristics in a three-phase
analysis. Phase 1 addressed our organization ofpractices into beginning choice point practices,
continuing care/ending choice point practices, organizational or structural practices, and both
beginning and continuing/ending choice point practices. We considered disproportionate
representation according to whether practices were beginning, continuing/ending, organizational
or structural, or both beginning or ending.
In phase 2, we considered individual practices in an attempt to consider the impact of
practices on disproportionate presentation in finer-grained detail. We were hampered in our
attempt: very few counties reported instituting certain of the practices.
Ultimately, we chose to consider at the individual practice level only those where four or
more counties indicated implementing them. Four counties is a small number for this purpose
and limit our ability to detect genuine differences that might have been present at statistically
significant levels. It proved fruitful however, and provides an important basis for di~cussion of
individual practices. The five practices we considered were: family group conference or other
team decision making practices (n=7), voluntary family maintenance or other prevention/family
preservation services (n=4), placement related practices (n=6), family to family initiative (n=4),
and cultural competency of staff (n=l 0).
173
We asked parallel questions in all phase 1 and phase 2 analyses. Adjusting for urbanrural status and per-capita income, for Latinos, African Americans, and Asian Americans how is
disproportionate representation affected by type of practice? What was the impact on
disproportionate representation ofwhether a practice was one of.five specific practices:family
group conference or other team decision making practices, voluntary family maintenance or
other prevention/family preservation services, placement related practices, family to family
initiative, and cultural competency ofstaff?
In phase three ofour analysis we asked questions without regard to type ofpractice. We
considered the impact on disproportionate representation-whatever the particular practice
type-- of differences in practice location, target population, purposes, reasons for
implementation, initiating persons and agencies, implementation timing, implementation status,
location and scope of the practice, geographic and community targeting, evaluation, support,
and barriers.
There are many practice characteristics describing practices. To reduce the number of
characteristics for purposes of analysis, we clustered characteristics into logically related,
empirically associated groupings using factor analysis, a statistical procedure for clustering
individual questions to which respondents have given similar answers. When a question was
discovered not to fit well within factor-analysis created clusters of questions, we eliminated it
from the cluster and considered it in subsequent analysis individually.
We then conducted regression analyses like those conducted in phases I and 2. The
question was: adjusting for urban-rural status and per-capita income, for Latinos, African
Americans, and Asian Americans, how is disproportionate representation affected by practice
characteristics? What was the impact on disproportionate representation of differences in
practice location, target population, purposes, reasons for implementation, initiating persons and
agencies, implementation timing, implementation status, location and scope of the practice,
geographic and community targeting, evaluation, support, and barriers?
There are two ways in which practice types, practices, and practice characteristics can
be linked to disproportionate representation. One is a negative association between the
variables, indicating lower than expected overrepresentation. Negative association means that
implementing counties have relatively low or nonexistent minority overrepresentation and
suggests that the practice type, practice; or practice characteristic is successful.
Another kind of association between a practice type, practice, or practice characteristic
and overrepresentation is a positive association. Positive association between variables means
that there is greater than expected overrepresentation associated with the practice type, practice,
or practice characteristic-that overrepresentation is relatively high.
Practices linked to greater overrepresentation cannot be considered successful. How does
a practice come to be linked to greater overrepresentation? It is likely that positive association
between a practice and overrepresentation comes about because instead of reducing
overrepresentation, some practices are implemented reactively, that is, in response to
174
overrepresentation. These practices do not reduce overrepresentation, although they may do so in
the future if given sufficient time. At present; we note only that a significant positive association
with overrepresentation is probably reactive and does not indicate success.
Multivariate Results
Latino Overrepresentation
Latinos were overrpresented in about 60% of counties and the degree of their
overrepresentation was as high as 28 percentage points.
Practice types. For Latinos, there was a statistically significant difference according to
whether practices were beginning choice point practices, continuing care/ending choice point
practices, organizational or structural practices, or both beginning and continuing/ending choice
point practices.
Both significant differences involved beginning choice point practices. Beginning choice
point practices were associated with about 8 percentage points less overrepresentation than both
beginning and continuing/ending choice point practices and about 9 percentage points less
overrepresentation than organizational or structural practices (See Table 63).
Practices. One statistically significant difference emerged when we considered the five
individual practices that could be analyzed. Practices involving family group conference or other
team decision making were associated with 7 percentage points less overrepresentation than were
practices involving culturally competency of staff (See Table 63).
Practice characteristics. For Latinos, there were statistically significant difference in
target population, purposes, initiating persons and agencies, and implementation timing (See
Table 64). Latino overrepresentation was greater when areas were targeted for intervention (area
where most referrals came from; with low socioeconomic status residents; with high crime; with
large ethnic representation), when clients, foster families, and out of home placement providers
were involved in the development of the practice, when service systems were involved in the
development of the practice (mental health, substance abuse, health, criminal justice, contract
agencies). Overrepresentation was less also when "other" problems were addressed as the
purpose of the practice and when it began more recently.
African American Overrepresentation
African Americans were overrepresented in about 64% of counties and their degree of
their overrepresentation was as high as 65 percentage points.
Practice types. For African Americans, there was no statistically significant difference in
overrepresentation according to whether practices were beginning choice point practices,
continuing care/ending choice point practices, organizational or structural practices, or both
beginning and continuing/ending choice point practices (See Table 63).
175
Practices. One statistically significant difference appeared when we considered the five
individual practices, which could be analyzed. Family to Family initiatives were associated with
8-percentage points greater overrepresentation than were the remaining practices considered (See
Table 63).
Practice characteristics. For African Americans, there were differences in target
population, purposes, and initiating persons and agencies, and geographic and community
targeting. African American overrepresentation was less when areas were targeted for
intervention (area where most referrals came from; with low socioeconomic status residents;
with high crime; with large ethnic representation), ethnic minority populations (Latino, African
American, Asian American/Pacific Islander, American Indian, families of any ethnicity), and
when very young children were targeted. African American overrepresentation was less when
foster families and clients or contract agencies first initiated the practice. African American
overrepresentation was less also when the purpose the intervention was to improve the
psychosocial functioning of the birth family (See Table 64).
Asian American/Pacific Islander Underrepresentation
In no counties were Asian American/Pacific Islanders overrerpresented. Instead, they
were equally represented in a few counties and in most counties Asian American/Pacific Islander
were underrepresented.
Practice types. For Asian American/Pacific Islanders, there was no statistically
significant difference in overrepresentation according to whether practices were beginning
choice point practices, continuing care/ending choice point practices, organizational or structural
practices, and both beginning and continuing/ending choice point practices (See Table 63).
Practices. Nor was there a statistically significant difference when we considered the five
individual practices that could be analyzed (See Table 63).
Practice characteristics. For Asian American/Pacific Islanders, there were differences in
target population, implementation timing of the practice. Asian American/Pacific Islanders under
representation was greater when areas were targeted for intervention (area where most referrals
came from; with low socioeconomic status residents; with high crime; with large ethnic
representation), ethnic minority populations (Latino, African American, Asian American/Pacific
Islanders, American Indian, families of any ethnicity). Asian American/Pacific Islander under
representation was less when system management and child welfare workers expressed support
for the intervention, and when it had begun earlier (See Table 64).
176
Table 63: Practice Types, Practices, and Disproportionate Representation
pract1ce
. Types
Ethnic Group
Variable
B SE Beta
T
Latino
Beginning vs
.09 .04 .14
1.98*
Organizational/
Structural
Beginning vs
.08 .03 .20 2.61 **
Beginning/Continuing
Practices
Variable:
B
Family Group Conference/ .07
Team Decision Making vs
Cultural Competence
African American Family to Family vs. other -.08
Ethnic Group:
Latino
SE
.03
Beta T
.15
2.13*
.04
.13
2.23*
Table 64: Practice Characteristics and Disproportionate Representation
Ethnic Group:
Latino
Variable:
Area Targeting
Target:Other Problems
Clients Involved in
Development/Evolution
Systems Involved in
Development/Evolution
Initated Earlier
African American Ethnic Targeting
Area Targeting
Targeting: Very young
Foster family-client
Initiated
Contract agency
Initiated
Goal: Improve psychoSocial Functioning Birth
Family
Asian American/ Ethnic Targeting
Pacific Islander
Area Targeting
Management-Worker
Support
B
.04
-.05
.03
SE Beta T
.01 .19
3.07**
.26 -.13 -2.06*
2.77**
.01 .17
.03
.01
.16
2.61 **
-.00 .00 -.16
-.04 .01 -.23
-.02 .01 -.12
.06 .03 .11
-.08 .04 -.11
-2.69**
-4.20**
-1.93*
1.92**
-1.94**
-.08 .04 -.13
-2.11 *
-.05
.02 -.13
-2.11 *
.00
.00
.00
-.01
.00 .19
.00 -.29
.22
1.99*
1.90*
-3.24**
Notes:
Regressions: Ordinary Least-Squares. All models R >.80 (p>.01).
Factors = Principal Components, 1 factor per analysis, Eignenmvalue> 1. Regression-based factor scores used in
subsequent regression. Variables with loadings >.50 were eliminated from analysis and analyzed separately in
subsequent regression.
177
Summary of Key Findings and Implications:
Multivariate Analysis
When considering the 43 counties responding to the survey of practices perceived
effective in working with children of color, significant associations were uncovered between
practice types, practices, and practice characteristics on the one hand, and disproportionate
representation on the other. These associations remained after adjusting for urban rural status and
per-capita income, differences between the counties strongly linked to minority child welfare
representation. Some point to effective practices-practices that bring about minority
underrepresentation. Different patterns ofassociation were found for Latinos, African
Americans, and Asian American/Pacific Islanders.
For Latinos, beginning choice point practices had a beneficial effect on
overrepresentation. In particular, family group conference or other team decision-making
interventions predicted lower levels ofoverrepresentation.
Bringing together interested parties early on in child welfare involvement may divert
cases from child welfare or otherwise minimize involvement, perhaps by facilitating
communication and mobilizing otherwise overlooked resources. Conceivably, the intervention is
especially valuable for Latinos in that its family focus is congruent with cultural values, and
because of the intervention's potential to overcome linguistic and cultural barriers to
understanding.
Notably, these interventions do not target at-risk neighborhoods as a special focus of
concern. Indeed, Latino overrepresentation was greater, not less, when interventions were
implemented targeting areas with high concentrations ofdistressed families.
Other findings indicated that practices targeting at-risk populations came about when
clients, foster families, and out of home placement providers were engaged in development of
the practice, along with mental health, substance abuse, health, criminal justice, contract
agencies. It appears that when they perceived a need and had an opportunity to do so,
stakeholders mobilized and focused on high need areas and populations as intervention targets.
However, they appear do this in response to Latino ove"epresentation, such that
implementation is linked to greater overrepresentation. The positive association at-risk targeting
and disproportional representation indicates that these practices do not reduce disproportional
representation. In all likelihood, the positive association means instead that stakeholders
perceive Latino overrepresentation and initiate programs in response.
For African Americans, targeting ofat-risk areas and African American populations was
beneficial in reducing overrepresentation, as well as targeting ofvery young children. Less
overrepresentation was linked also to whether foster families and clients and contract agencies
initiated an intervention, and whether it focused on improving psychosocial outcomes of birth
families. The findings suggest a picture almost the·reverse of that found for Latinos.
178
For African Americans, there appear to be lower levels ofoverrepresentation than
expected too when key community stakeholders help to initiate the interventions that focus on
high-risk areas and populations. When interventions have a goal of strengthening the well being
of birth families, then implementation is associated with lesser overrepresentation.
On the other hand, counties that implemented family-to-family initiatives displayed
greater overrepresentation than counties that did not. Family to family programs appear to be
used reactively, in response to high perceived levels of overrepresentation.
It is important to note however that in counties where interventions were successful
levels of overrepresentation remained high. Successful interventions were those associated with
lower overrepresentation than expected. Improvement was relative: in absolute terms there
remained a high degree of African American overrepresentation.
For Asian American/Pacific Islander populations findings suggested that, as with
Latinos, targeting of high need areas was less beneficial than not engaging in this kind of
intervention. Involving clients and service systems in initiation, again more common for
interventions targeting high need areas, was also relatively unsuccessful. On the other hand,
management and worker support for a practice was associated with greater success.
Asian American/Pacific Islanders are, overall, underrepresented in child welfare systems
and one must interpret our findings in light ofthis fact. There were fewer significant differences
between practice types, practices, and practice characteristics for Asian American/Pacific
Islanders than for Latinos or African Americans. This may reflect a lesser sense of urgency felt
by administrators and workers in the face of Asian American/Pacific Islander disproportionate
representation that is proportionally low than for Latino and African American disproportionate
representation, which is proportionately high.
A need remains to conceive and mount culturally and linguistic programs for Asian
American/Pacific Islander populations. This aspect of the study, with its focus on
disproportionate representation rather than on other indicators of system performance, may not
have been as good a vehicle as other parts of the study to identify program needs for Asian
American/Pacific Islander populations as for overrepresented Latino and African American
populations.
Certain limitations must be kept in mind when considering the present findings. The
study encompasses forty-three California counties. The number of counties implementing any
practice or practice type was smaller than ideal to undertake truly powerful statistical tests
sensitive to all underlying differences. As a matter of practical necessity, the study's design was
such that some of the links of practice types, practices, and practice characteristics with
disproportionate representation might have arisen by coincidence. The possibility that some
unmeasured characteristic explained some of the associations of interest revealed in the study
was reduced, but cannot entirely be ruled out.
Despite these limitations, the study produced meaningful patterns of findings connecting
programs and disproportionate representation. The associations proved sensitive to the
179
differential experience of different ethnic groups. They furnish a useful context of fact to further
consider the problem of disproportionate representation in Santa Clara County.
180
VI. PHASE 3 SUMMARY, RECOMMENDATIONS AND
CONCLUSION
Phases 1 and 2 of this study focused on examining reasons for the racial/ethnic
disproportionality in the County's CWS and identifying the decision points where it occurs. The
focus ofPhase 3 was on key practices that take place within the CWS that either promote or
reduce disporportiona/ity among children and families ofcolor.
Overall, research on the impact of child welfare practices for children and families of
color is limited. In general, much of the research literature is inconclusive and fails to support the
effectiveness of child welfare practices i'n improving outcomes for children and families involved
in the CWS. Very few studies describe differences in outcomes by ethnic/racial groups and most
of the research fails to take into account the effect of culture and ethnicity on the delivery and
impact of child welfare practices. This gap in the research literature is striking and has important
implications for children and families of color in the CWS. More information about child welfare
practices _including programs, services, strategies, policies, and/or tools that either increase or
reduce racial/ethnic disproportionality can contribute to the development of more effective
services and supports for children and families of color in the CWS.
The three overall aims addressed in Phase 3 were: 1) to assess identified key practices for
children and families ofcolor in Santa Clara County's CWS, 2) to examine the influence ofthe
court system on service recommendations for children and families and, 3) to conduct a
statewide comparative analysis of identified key practices that may affect the disproportionate
representation of children ofcolor in the CWS.
In order to address Aim 1 which involved an assessment of key practices in Santa Clara
County's CWS, four related analyses were conducted. First, the CWRT conducted analyses to
understand the types of practices and case characteristics that are associated with successful
outcomes for children and families from various racial/ethnic groups. In order to assess
successful outcomes, we used data from our Phase 2 Case Record Review sample and defined
success according to US DHHS guidelines, i.e., safety, permanency and family and child wellbeing. Results indicated that there were very few ethnic differences in successful outcomes.
In regards to safety, there was a significant relationship between the number of times
removed from family in current episode and ethnicity where African American children had the
highest average number of times removed from family in current episode and Asian
American/Pacific Islander children had the lowest average. In terms of family and child wellbeing, there was a marginally significant relationship between families' capacity to provide for
their children's needs and ethnicity, with African American families appearing to have a more
enhanced capacity while Asian American/Pacific Islander families seem to have a more
diminished capacity.
It is important to note that major ethnic differences do exist in Santa Clara County's
CWS. Phase 1 and 2 confirmed that certain ethnic groups are over-represented (i.e., Latinos,
African Americans) while others are under-represented (i.e., Asian American/Pacific Islanders)
in the county's CWS. We also found that there are significant ethnic differences in many
181
demographic, system-related, and psychosocial characteristics. Many of these variations appear
at different choice points in the system (such as psychosocial status at the beginning of the case,
and assignment to voluntary family maintenance). Results ofour successful case outcomes
analysis indicate that ethnic differences at the conclusion ofthe case are not evident.
While results suggest there is a clear disproportion of ethnic groups in the system, once in
the system, the children and families are generally faring the same. Their relatively
homogeneous experience may be a result of the "one-size-fits all" limited array of services
offered. Although this may seem equitable and logistically efficient, it appears ultimately
ineffective in serving a diverse group of children and families. Also, to determine better how
"well" these children are actually doing while in the system, the availability ofcomparative data,
clear operational definitions for "success," and more reiiable and valid data are needed.
Implications and Recommendations
•
In order to better assess whether safety, permanency, and family and child wellbeing are maintained and enhanced, records should include more complete and
measurable information. We acknowledge that child welfare case records were
not originally intended for research. Yet, if the agencies responsible for providing
effective services wish to evaluate the performance of their programs more
reliably and validly, improving information collection and tracking should be
considered.
•
The creation of a case summary check-out form used at case closure would be one
way to track child outcomes. The check-out form would indicate the child and
family's status in major areas of psychosocial well-being (i.e., health, education,
and finances), in addition to the existing case outcomes. Given the amount of
missing data assessing these domains describing case and child status, and the
relative scarcity of information recorded in the last two court hearings, this form
would enable measurement and evaluation of "successful cases and outcomes" in
a more consistent objective manner.
•
The amount of missing data and the manner in which information is assessed also
implies the need for more reliable and valid methods of recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their families or of
those adopted high or low? And compared to what? Implementing better record
keeping and data collection over time on key characteristics, based on an agreedupon operational definition would better address the question of effectiveness and
performance. Implementing these practices across counties and across states
would also facilitate evaluation.
Second, a qualitative analysis using both focus group and interview data was conducted
to obtain insights about current child welfare practices used within the DFCS, how these specific
182
practices might positively or negatively impact children and families of color, and what practices
might be created, enhanced or maintained to better serve children and families of color.
Results provided information on numerous practices and contextual factors that have
either a negative or positive impact on children and families ofcolor in Santa Clara County's
child welfare system. Study participants recommended the use of in-home services for children
and families of color. There is some evidence in the research literature that supports the
effectiveness of in-home services (see literature review), however these services appear to be
most effective when they are intensive, long-term and delivered by health or social service
professionals and when the workers delivering these services receive a high level of supervision
(Barth, 1991; McGuigan et al. 2003; Olds et al., 1997). Participants also identifiedfamily
conferencing as an effective practice. These study findings are consistent with limited research in
this area that suggests that inclusive practices, in which parents are engaged in placement and
other decisions, may result in more placement stability for children, although possible
differential effects for diverse populations has not been researched (Palmer, 1996).
Participants generally expressed optimism about the adoption ofteam decision making
(/'DM) as a vehicle for improving decision-making with diverse families, although some social
workers expressed concern about time required for implementation. Research suggests that TOM
can be labor intensive and can be difficult to implement effectively within the context of the
child welfare system (Sieppert et al., 2000). Finally, participants noted that both formal and
informal collaborative relationships with other services delivery systems, such as domestic
violence and substance abuse treatment systems, are critical for successful interventions with
children and families. Furthermore, participants recognized the importance of training and
cross-training to enhance the capacity of helping professionals in child welfare and other
systems to better address the needs ofchildren and families ofcolor.
Interestingly, most practices and contextual factors mentioned by social workers had both
positive and negative aspects. Many of the practices and contextual factors that negatively
impact children and families of color were actually barriers to successful implementation of the
practices that social workers felt have a positive impact on children and families of color.
Although practices such as orientation ofnew clients; family and team decision-making;
preventive services; substance abuse services; cultural competency; collaboration with other
agencies and systems; a strength-based approach; and the use ofrelative placements were
described as having a positive impact on children and families ofcolor, certain negative
practices and contexts impeded their effective implementation. For instance, having to screen out
large numbers of inappropriate referrals; inconsistency in decision-making practices; a shortage
of services, particularly substance abuse and preventive services; lack of client access to
services; difficult protocols for placing children with kin; gaps in cultural competency; time
limits; and agency-level factors such as heavy caseloads, staff shortages, substantial amounts of
paperwork, lack of access to information about resources; and confusion about the agency's
overall mission and key policies all impeded implementation of the best practices. These findings
suggest that in addition to building on current positive practices for children and families of
color, CWS stakeholders should also actively work both internally and in collaboration with
partners to reduce barriers to these best practices.
183
Indeed, in order to address these barriers to effective implementation of best practices,
study participants mentioned numerous recommendations that centered around certain key
themes. In general, there was an emphasis on improving decision-making through increased
accountability, training and cross-training to reduce bias, and increased use ofgroup decisionmaking or family involvement in decision-making. Study participants also stressed a need to
improve and expand prevention, diversion and concrete services, as well as an overall need to
expand culturally and linguistically competent services, and develop ways to improve
availability and access to services. Once a case is opened, social workers also felt it valuable to
provide an orientation to clients so that they are more aware of the CWS and court processes and
can better navigate the system. Recommendations related to organizational factors were also
noted, including reducing caseloads; increasing support and streamlining service delivery;
clarification of the mission ofDFCS, as well as key child welfare policies; fostering an
organizational culture that is strength-based and community-based, and more overall training
for social workers.
Implications and Recommendations
•
The findings from the qualitative component of this study affirm the value of
many efforts that are already in place, such as culturally specific ER response
units and family resource centers, and other that are in the process of
implementation including efforts to reduce caseloads, initiation of team decision
making, and participation in the Family-to-Family initiative.
•
Pract~ces that are inclusive, collaborative, culturally appropriate and involve
client's communities appear to be best suited to meeting the needs of children and
families of color. Although the research literature does not provide convincing
evidence of the effectiveness of these practices, the fact that respondents reported
these practices as those that are most effective with children and families of color
suggests that they may serve as promising models for children and families of
color and that more rigorous evaluations of these practices are needed.
•
The findings also point to opportunities to further strengthen services for children
and families of color. Recommendations related to training, institutionalizing
group orientation for clients, and other suggestions from study participants could
be used to inform practice and planning.
•
It would be beneficial for a team of managers, line staff, family representatives,
and community members to review the recommendations suggested by focus
group and interview participants to identify and prioritize possible practices for
adoption based on their feasibility and utility for children and families of color.
Third, in an effort to better understand the influence of contextual factors on child
welfare practices and outcomes for children and families of color, a comparison of Santa Clara
County's Main Offices and South County office was conducted. This comparison included both
quantitative and qualitative analyses. Quantitative data from CWS/CMS were used to examine
differences in case characteristics between the Main County Offices and South County Offices.
184
A qualitative comparison of focus group data from the Offices and South County Offices was
also conducted to discern how workers perceive practices in these two geographical locations.
These analyses help shed light on the potential impact of agency and community context on case
characteristics, child welfare practices and outcomes for children and families of color in Santa
Clara County's CWS.
Unfortunately, the South County location subsample was too small to allow analyses by
ethnic group. However, some key case characteristics are significantly related to county
location. Specifically, in South County Offices, a higher proportion ofcases are in voluntary
services, cases are shorter in duration, fewer workers are assigned during the course ofthe case,
and a higher proportion ofcases are in family maintenance at case closure. These findings are
also statistically noteworthy given the small subsample size available from SC. There were also
other differences in case characteristics by county location that did not attain statistical
significance but are noteworthy for further investigation. Children served in the South County
Offices appear to have fewer number of unique placement homes in the current episode, have a
shorter average stay per placement, a shorter length oftime in out-ofhome placement, and are
younger at time of case closure.
These findings imply that there is a difference in the style and quality of services
provided in the South County Office location. Qualitative findings are congruent with the
quantitative data and provide some insight into the dynamics that may contribute to this
difference. The higher number ofcases in voluntary services, the larger proportion ofcases in
family maintenance at case close, and the shorter duration ofcases is consistent with what South
County study participants described as a culture ofcommitment to maintaining and reunifying
families. This commitment was described as integral to the philosophy of managers and social
workers and reflected in the expectations communicated to social workers by most supervisors
and by peers. The practice of providing an orientation to the child welfare system, time limits,
the courts, and how to work with social workers may also contribute to the differences found in
the quantitative analysis. Study participants frequently described South County as similar to a
small county or rural area. This geographic difference, and the resulting sense of "connection to
the community, " was identified as a factor in the greater sense ofshared philosophy and
emphasis on prevention ofout-ofhome placements. Other studies suggest that region may
influence practice. For example, one study (Drake, 1996) also found that rural areas are more
likely to offer preventive services than are urban areas. The finding that fewer workers are
assigned during the course ofthe case is also consistent with descriptions of the vertical case
management model. This model calls for social workers to carry the same case from case
opening to case closure.
Implications and Recommendations
•
Future research should include a larger subsample from South County and include
methods to understand the reasons behind these location differences. Also, a
larger South County Office location sample may enable more detailed
comparisons by ethnicity.
185
•
Practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County
Offices and adapted to other regions of Santa Clara County, as recommended by
study participants. This practice may be particularly helpful to low-income
families and families of color that may be intimidated by or unfamiliar with child
welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for other
regions of Santa Clara County, though the success of this practice may be linked
to the "small county" dynamic described by South County study participants.
Some participants in focus groups that took place at the main office mentioned
this model as promising while others asserted that specialized expertise in
different areas of social work was an advantage in providing quality services that
are informed by "the most current information."
•
Further examination of ways in which the culture of commitment to maintaining
and reunifying families is created, communicated, and continued merits attention.
In addition, an exploration of how this culture might be adapted to other parts of
the county would be beneficial.
Fourth, in order to identify case characteristics and system-related factors predictive of
reunification of children with their families, two exploratory models were tested using
multivariate analyses. The first model was composed of primarily demographic and systemrelated variables to predict reunification (child's gender, ethnicity, age at time of case opening,
the number of workers assigned across the history of the case, the time length of the case, the
number of unique placement homes assigned, the number of times removed from the family, and
the county location). Four ofthese variables predicted reunification: child's ethnicity, number of
workers assigned throughout the case, length ofthe case, and number of unique placement
homes.
Asian American/Pacific Islanders were less likely than Whites, African Americans, and
Latinos to be reunified with their families. Given that Asian American/Pacific Islanders are one
of the minority groups least acculturated to mainstream society, their cases may be perceived as
more severe or problematic because of the different cultural attitudes toward child rearing and
discipline. Differences in parenting styles may reduce the likelihood of reunification compared to
the other major ethnic groups. However, as Phase 2 results indicated Asian American/Pacific
Islanders are also the most likely to be placed in Voluntary Family Maintenance (Hines et al.,
2002), the CWS appears to value the preservation of Asian American/Pacific Islander families
when first referred to the system (perhaps recognizing the challenges of culturally appropriate
services, language issues, and other circumstances related to the Asian American/Pacific Islander
community) but once in the system, Asian American/Pacific Islander family cases are not easily
resolved and circumstances may persist that place the child at risk for further abuse of neglect.
The fewer number ofworkers assigned was also related to reunification. This finding
indicates consistency in service and the relationship between the social worker and the family
186
may improve the chances of reunification. Perhaps this is due to greater familiarity and thus more
effective advocacy for the child and family. This finding could also be a function of particularly
challenging family cases already less likely to be reunified needing a greater set of workers over
time (e.g., due to worker burnout in the case, or specialization needed across the case).
Shorter case length was also predictive ofreunification. Presumably, less severe cases
could be resolved in a timely fashion, usually concluding with reunification. Following a similar
logic above regarding the number of workers, a longer case may imply more problems, thus
involving more time for resolution, but also possibly a case already protracted into a situation
where reunification is unhealthy for the child.
The fewer number of unique placement homes assigned was related to reunification.
Change in placements may be due to systemic conditions (i.e. logistics and short term
availability ofspace) but multiple placements can also indicate a persistent problem with a child
adapting to a placement and accepting care offered. As this may again indicate a more severe
case, reunification could also become less likely.
To identify better the influence of case characteristics and indicators of success, we added
6 variables to the model collected through our in-depth case record reviews and utilized the
additional information coded for Phase 3: a family's enhanced capacity to provide for their
children's needs, children receiving appropriate educational services, children receiving adequate
services to meet physical and mental health needs, number of referrals, number of previous times
in the CWS, and assignment to Family Maintenance or Family Reunification services. The
location variable was not included in this model as only one case coded from the South County
Office area.
Similar to the first model, Asian American/Pacific Islanders were still less likely than
Whites, African Americans, and Latinos to be reunified. Also, shorter time length ofa case
again predicted reunification. However, in this expanded model with success indicators, the
number of workers assigned and number of unique placement homes assigned were not
statistically related to reunification. The child's age at time of case opening and the number of
unique placement homes did approach significance. Interestingly, the variables related to
success cases and outcomes were not significant.
Implications and Recommendations
•
These findings indicate that particular factors are important to consider when
predicting reunification, specifically ethnicity and the length of time a case
remains open. More severe cases (those needing more time for resolution) may be
more problematic, thus reducing the chances of reunification.
•
The finding that Asian American/Pacific Islanders are less likely to be reunified
than Whites, African Americans and Latinos, highlights a main ethnic difference.
In Phase 2 we discovered that Asian American/Pacific Islanders were more likely
than the other ethnic groups to be enrolled in Voluntary Family Maintenance
services (Hines et al., 2002). However, our results here imply that when Asian
187
American/Pacific Islander children are removed from the home, their children are
less likely to be reunified. This may also indicate extreme outcomes where Asian
American/Pacific Islander children are either initially diverted from the system or
once in the mainstream of the system are less likely to be reunited with their
families.
•
It was interesting that the successful case and outcome variables were not
predictive of reunification. This could be due to the significant impact of other
variables overshadowing the influence of success indicators. It could also be due
to our use of reunification as a dependent variable. As we discussed in the
Successful Cases Analysis section, reunification is not necessarily synonymous
with success, and thus these variables may indeed be unrelated.
•
Due to missing data across our set of variables, a number of were available for
this multivariate analysis. More complete information would have improved
confidence in our findings.
•
Better measurement and a more comprehensive evaluation of family history,
abuse and neglect circumstances, and other indicators could produce a stronger
predictive model.
•
Although only a preliminary set of models, these findings are useful for
identifying areas in the CWS and family circumstances that can be studied
further, especially in conjunction with the specific goal of reunifying children
with their families.
Our second aim focused on examining the influence of the court system on service
recommendations for children and families ofcolor. Few studies have examined the potential
influence of the court system on child welfare ·practices and outcomes, however as noted in the
literature review (Section III), the court system is likely to have a significant impact on the
trajectory of some child welfare cases. In Phase 3, we examined the role of the court system in
child welfare practice by exploring the types of court ordered changes that are made to social
worker recommendations at the initial jurisdictional/dispositional hearing, and how changes
might vary by ethnicity of the case. In addition, we explored the relationship between the child
welfare and court systems.
Quantitative results indicated that the court made changes to the initial social worker
recommendations at the jurisdictional/dispositional hearing in approximately halfof the cases.
Ethnicity was not related to whether or not a change occurred Yet, these results may not
necessarily indicate that the system and courts treat children and families from different
ethnicities the same, but rather that when the judge decides to make a change from the social
worker recommendations, these changes appear to occur equally across ethnic groups. It was the
opinion of members of the Court Officer Unit focus group that ethnicity does not contribute to
decisions in court, but rather it is circumstances that drive the case.
188
Additionally, although quantitative results may imply that judges accept social worker
recommendations about half of the time, findings from the focus group with the Court Officer
Unit suggest that many factors would complicate an accurate measurement of why social worker
recommendations may or may not be followed. Factors that may impede accurate assessment of
the agreement between child welfare and court personnel on service plan recommendations
include the strong influence of individual stakeholders involved, and formal and informal case
plan negotiation tactics, all of which contribute to a much more complicated process.
Qualitative findings from the agency-wide focus groups suggested that collaboration
between the child welfare and court systems is problematic. Agency-wide focus group
participants felt that the child welfare and court system have different perspectives on the needs
and circumstances of children and families in the CWS and that the court system may have
unrealistic expectations of families. Additionally, many agency-wide focus group participants
described the ways in which child welfare and court personnel interact as ineffective. Some
social workers felt that some court personnel would dismiss their assessments and
recommendations and may try to pressure social workers to change their recommendations. A
somewhat similar finding was noted by Knepper and Barton, ( 1997) who found that although
judges tended to accept social worker recommendations, the relationship between social workers
and the court plays a significant role. Their study found that when social workers adhered to the
group norms of the court, judges rewarded them by not ordering "unrealistic" practices and
allowing reasonable time frames for mental health evaluations to be completed. Similarly,
agency-wide focus group findings seem to suggest that if social workers go against the implicit
rules ofthe court system, then judges may override their recommendations, but if they adhere to
expectations ofthe court system than recommendations are accepted
Implications and Recommendations
•
Although the quantitative and qualitative information described many of the
circumstances behind why social worker recommendations are not upheld by the
court, it is very difficult to ascertain the degree to which each circumstances
influences a court ordered change. More examination of this area is needed, as is
further investigation of ways in which relationships between court personnel and
child welfare workers have an impact on the experiences of children and families
of color involved in the CWS.
•
Better and more exhaustive record keeping may not be the solution, as many
informal and undocumented actions in the negotiation process occur regularly,
and are confidential, thus making much of the data unavailable for research and
evaluation purposes.
•
Assessing other factors such as courtroom environment, stakeholder
characteristics, and other judicial issues would be advantageous in later research.
Although obtaining this information is a formidable task, these factors should be
include in the future, as measuring primarily system-related factors may not be
sufficient in efforts to explain outcomes in the CWS.
189
Our third aim was to conduct a statewide comparative analysis of identified key practices
that may affect ethnic disproportiona/ity in the CWS. A statewide survey was administered to
child welfare directors, managers or supervisors in California counties who were asked to
identify and describe effective practices for children and families of color in their child welfare
systems. In addition, a quantitative analysis that identified the link between these practices and
county levels of disproportionate representation of children of color in county child welfare
systems was conducted.
Overall findings from the statewide comparative analysis indicate that certain types of
practices are considered by county child welfare personnel as effective with diverse populations,
and that these practices are related to disproportionate representations ofchildren of color in
county child welfare systems. Descriptive results revealed four overall types of practices
considered to be most effective with children and families of color, as well as various specific
practices within these types of practices including: 1) beginning choice point practices, which
included family group conferencing or other team decision-making practices, voluntary family
maintenance or other prevention/family preservation services, and risk assessment tools or
systems, 2) continuing care/ending choice point practices, which included, placement related
practices, the Family to Family Initiative, and intensive and collaborative interventions for highrisk children, 3) both beginning and continuing/ending choice point practices, which included,
community-based services, and strength based assessments, and 4) organizational or structural
practices which included, cultural competency of staff, collaboration with American India~ tribes
and vertical case management.
In general, very little research on the effectiveness of the practices identified by
respondents has been conducted, and in cases where evaluations have been performed, results are
generally inconclusive. More research is needed to evaluate the effectiveness of these services
for diverse populations.
Multivariate analyses revealed several significant associations between types of
practices, specific practices and practice characteristics and levels of disproportionate
representation ofchildren ofcolor in county child welfare systems, when other factors such as
urban-rural status and per capita income were statistically controlled
Beginning choice point practices were found to have a beneficial effect on Latino
overrepresentation. In particular, family group conference or other team decision-making
interventions predicted lower levels of Latino overrepresentation. However, Latino
overrepresentation was greater, not less, when interventions were implemented targeting areas
with high concentrations of distressed families. Latino overrepresentation was also greater when
clients, foster families, out-of-home placement providers, and other service systems were
involved in the development of the practice. As such, it appears that practices may be developed
because stakeholders perceive Latino overrepresentation and initiate programs in response.
In contrast to Latinos, targeting practices to at-risk areas for African American
populations was beneficial in reducing overrepresentation, as well as targeting ofvery young
children. Less overrepresentation was linked also to whether foster families and clients and
contract agencies initiated an intervention, and whether it focused on improving psychosocial
190
outcomes of birth families. The findings suggest a picture almost the reverse of that found for
Latinos.
Additionally, there appear to be lower levels ofAfrican American overrepresentation
than expected when key community stakeholders help to initiate the interventions that focus on
high-risk areas and populations. When interventions have a goal of strengthening the well being
of birth families, then implementation is associated with less African American
overrepresentation. Yet when counties implemented Family-to-Family initiatives greater African
American overrepresentation resulted. Again, the implementation of Family to Family programs
may be used reactively, in response to high perceived levels of overrepresentation.
As with Latinos, targeting ofhigh need areas was less beneficial for Asian
American/Pacific Islander children than not engaging in this kind of intervention. Involving
clients and service systems in initiation, again more common for interventions targeting high
need areas, was also relatively unsuccessful. On the other hand, management and worker
support for a practice was associated with greater success.
Implications and Recommendations
•
County directors and child welfare administrators identified practices with similar
characteristics as those identified within Santa Clara County, i.e., those that are
inclusive, collaborative, culturally appropriate and involve client's communities
appear to be best suited to meeting the needs of children and families of color.
However, further research and evaluation is necessary to determine which
practices most effectively meet the needs of individual racial/ethnic groups.
•
While multivariate analyses suggest that certain practices do predict levels of over
and underrepresentation of children of color in the child welfare system,
additional research to determine the mechanisms through which these practices
impact this disproportionate representation is needed.
•
Additional explorations into the relationship between practices, disproportionate
representation of children of color in the child welfare system and county
characteristics would also help shed light on the process through which certain
practices impact children and families of color.
Phase 3 Conclusion
Results from Phase 2 indicated that different racial/ethnic groups experience unique
pathways through the CWS (Hines et al., 2"002). At every point in the system, various child
welfare practices and services i'mpact children and families. Phase 2 results indicated that
children and families of color in Santa Clara County's CWS tend to be recommended services
that are traditional and formal in nature and that do not appear to meet the wide range of needs
experienced by these highly diverse racial/ethnic family groups. Indeed, children and families of
color in the CWS represent a high risk group who are often impacted by myriad psychosocial
191
challenges yet services recommended for these families tend to be limited to a one-size-fits all
approach.
Moreover, there exists relatively little empirical evidence that traditional child welfare
services are effective in maintaining or reunifying children and improving family functioning.
Even less research has been conducted on the impact of child welfare practices for children and
families of color. Yet because children and families of color are disproportionately represented
in the CWS, the effectiveness of child welfare practices for these groups is of particular interest.
Research suggests that children and families of color tend to have longer stays in the CWS, they
receive fewer and less comprehensive services, and have poorer case outcomes than White
children and families (Close, 1983; Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996).
This bleak outlook for children and families of color in the CWS creates an urgent need to
understand child welfare practices and services that result in enhanced outcomes for these
families. It is especially important to identify the types of individual and contextual factors that
hinder or support the effectiveness of these practices.
The primary goal of Phase 3 was to assess key practices for children and families of color
in the CWS at both the county and state levels in order to provide information that might help
generate strategies aimed at reducing ethnic/racial disparities and develop more effective
services and supports for children and families of color. While little research exists that
empirically supports policies, programs and practices that attempt to reduce ethnic/racial
disproportionality, some programs and practices appear promising and merit further exploration.
In general, these practices aim to reduce the need for out-of-home placement and increase family
and community participation in decision-making. Please see Chart 1 for a summary of promising
child welfare practices from Phase 3.
In our analysis of Santa Clara County child welfare practices, results based on focus
group and interview data indicated that in-home services, family conferencing, and team decision
making (TDM) were viewed by participants as ways to improve decision-making with families
from diverse cultural and ethnic groups. Participants noted that both formal and informal
collaborative relationships with other services delivery systems, such as domestic violence and
substance abuse treatment systems, are critical for successful interventions with children and
families. Furthermore, participants recognized the importance of training and cross-training to
enhance the capacity of helping professionals in child welfare and other systems to better address
the needs of children and families of color.
Our analysis comparing South County and the Main Office enabled us to examine
contextual factors that may have an impact on child welfare practices, in this case, small vs. large
and urban vs. rural. Results indicate that children in South County appear to have better
outcomes, i.e., they have fewer number of unique placement homes in the current episode, have a
shorter average stay per placement, a shorter length of time in out-of-home placement, and are
younger at time of case closure. Study participants described South County as "a small county or
rural area." This geographic difference, and factors related to an organizational culture that
emphasize a commitment to maintaining and reunifying families, the practice of providing an
orientation to the child welfare system, as well as the finding that fewer workers are assigned
192
Chart 1: Promising Child Welfare Practices from Phase 3
Beginning Choice Points
Promisine Practices
Team decision making, case conferencing, family group conferencing
and/or family involvement in decision-making
Orientation for parents on child welfare and court processes
.;_:; ~-. ,. ~t.-t~~- -~~;:~ ~· .. "-\;.-::.:
:'.~-:_-:t,.~-~ ':. <' /,:._:- !-<. ,.t;>..:.-. :~. <-:: ·.: _~- i:-.::':
Continuing Care/Ending
Choice Points
;:.: . .,,.-,1;;":•.;_, :.t_;~.,~.:.,.;:. ..:;.\:;~ ·-- 1~~-~·::..:.-.... ·tt'"~~.::-.=i~-~ 'f.-tr./'~_1 :.·::::;...~ 7~:-~ ...
Contextual Factors or
Organizational/
Structural Practices
-~:c:i!:_;-:,:~
Voluntary Family Maintenance/Family Preservation Services
Child care and in-home services
Risk assessment tool or system
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
cr.··~~~~~:;,::~::;:-,.:.~:s.-~- ·· ~.i:..;;:,:t-:·. •·-::-:~.:;;.:-:;, t: .:~::::·'.:\,:·:.:o1,h'!:~t1!~-.:;:.~.=::.:.r:·-~.:_::~t· .! ~,-·= /'- ~:!,i.,-;-:_~"::f,.;~;-:i·,:::-:::;.:,S'.:..~-~-.;.1~:-~""E~1..:;:::-:-;.:..J-k_ 1:~-.;:':j:':~.:..-.:": ·J,:~ --,.·/:; -~··":.~..-fC?::._~
Parenting education services involving both child and parent
Cultural matching for children in foster care
Recruitment, training and coaching for foster parents
Placement-related practices
Family-to-Family Initiative
Intensive and collaborative interventions for high risk children
Community based services (both beginning & continuing/ending)
Strength-based assessments (both beginning & continuing/ending)
•
•
Study Component
Main County Focus Groups
South County Focus Groups
Statewide Survey
Main County Focus Groups
South County Focus Groups
Statewide Survey
Main County Focus Groups
Statewide Survey
"!'" -.. ...
~r:;-::::tJ;.~.--,-:-7~---,.~~2'.--...... -• _
...... .:il ,.._.: .. c'.
Main County Focus Groups
Main County Focus Groups
Main County Focus Groups
Statewide Survey
Statewide Survey
Statewide Survey
Statewide Survey
Statewide Survey
r.t: . - ·-":"!=i.~"·";;;t:".:'!..._...~,..~~~~~!.'::.~~;,;,-.,t;:.:i:!i\.~-:~....-:~:-::.~-.1;,~-::~~1,~h=-~~L!.:$:"."~~~1::;.;;~.:..~~r.,,;~~}.~?'i~~·~t-~Y5~·LGt:~'T~;;.h--I~~~~;;:-..-~ :i:~-,~:tr.:.3.'~=-n..~-z:~~~~..11....-~..,17;~;~::tr~
Cultural matching between staff and clients, and cultural consultation
Supportive supervision for staff
Collaboration with other agencies or systems
Strength-based approach to services
Small county dynamic
Culture of commitment to maintaining and reunifying families
Vertical case management
Cultural competency of staff
Collaboration with American Indian tribes
•
•
•
•
•
•
•
•
•
•
'"'.-;";i,.-~·-~·7••.. •• ..........-.,...,,r•• '",
~~~~;.-:·~~M..i:~"~;:-,.1.-:-z.:.;s;-~'-!':1
Main County Focus Groups
Main County Focus Groups
South County Focus Groups
Main County Focus Groups
Main County Focus Groups
South County Focus Groups
South County Focus Groups
South County Focus Groups
Statewide Survey
Statewide Survey
Statewide Survey
193
during the course of the case may all be related to these positive outcomes for children in South
County. It is interesting to note, however that our multivariate analyses that examined case
characteristics and system-related factors related to reunification indicated that when taking other
factors into account, geographical location was not a significant predictor of reunification.
Rather, fewer workers, shorter case duration, and fewer placements were the strongest predictors.
More research on factors related to agency culture and geographical location as they impact child
welfare practice and outcomes for children and families of color is clearly warranted.
Research indicates that the court system is likely to have an impact on the practices that
are provided to children and families in the CWS. In an analysis conducted to determine the
extent to which court orders changed social worker recommendations at the
jurisdictional/dispositional hearing, results indicated that the court made changes to the initial
social worker in approximately half of the cases. Results further indicated that that ethnicity was
not related to whether or not a change occurred. Findings from qualitative analysis exploring the
relationship between the child welfare and court system painted suggested that the collaboration
between the child welfare system and courts is problematic. More research is this area is needed,
as is further investigation of ways in which relationships between court personnel and child
welfare workers have an impact on the experiences of children and families of color involved in
theCWS.
Results on statewide practices indicate that certain types of practices are considered by
county child welfare personnel to be effective with diverse populations, and that these practices
are related to disproportionate representations of children of color in county child welfare
systems. Similar to practices identified within Santa Clara County's DFCS, practices at the
statewide level were described as effective with children and families of color because they are
inclusive, collaborative, culturally appropriate, and involve client's communities. Although the
research literature does not provide convincing evidence of the effectiveness of these practices,
the fact that respondents reported these practices as their _most effective with children and
families of color suggests that they may serve as promising models for children and families of
color and that more rigorous evaluations of these practices are needed. Phase 3 also attempted to
assess how the practices identified through the survey might affect differential representation of
children of color in the CWS. Significant associations were detected between practice types,
practices and practice characteristics and disproportionate representation and interestingly,
different patterns of association were found for Latinos, African Americans and Asian
American/Pacific Islanders.
While researchers have described the existence of ethnic/racial disparities in every aspect
of the CWS, little attention has been given to ways to address the existing disproportinality. It is
our hope that the information contained in Phase 3 of the Children of Color Study will help in the
development of more effective programs and practices for families and children of color at both .
the county and state level. Overall, our findings from Phase 3 suggest that efforts to address
racial and ethnic disparities in the child welfare system should involve a more concentrated focus
on child and family well-being, involving the creation of family-centered and community-based
services that are inclusive and collaborative and specifically designed to meet the needs of the
·
diverse cultural and ethnic groups in the child welfare population.
194
VII. SUMMARY OF PHASES 1, 2 AND 3: OVERALL
CONCLUSIONS AND IMPLICATIONS FOR SANTA CLARA
COUNTY'S CHILD WELFARE SYSTEM
The central focus of inquiry for this project was to identify factors related to the
disproportionate representation of children of color in Santa Clara County's CWS. In order to
address the complexity of this issue, the CWRT elected to employ a multiphase/multimethod
approach that examined factors at the individual, family and system levels. The following
summarizes findings from Phases 1, 2 and 3 of the study and provides overall implications for
the County based on our results.
Phase 1, an exploratory phase, drew from three sources: research literature at the national,
state and county levels; Santa Clara County's management information system (CWS/CMS); and
focus group discussions with professionals in Santa Clara County who provide child welfare
services, as well as parents, caregivers and youth who are recipients of CWS services. Results
indicated that factors related to racial/ethnic disproportionality were multiple and complex and
necessitated investigation at the individual, family, system and community level. Based on
preliminary analyses and available literature, we argued that children from different racial/ethnic
groups were likely to be treated differently at specific decision-making or key choice points in
the system. The concern about possible differential treatment of children of color was
substantiated by the focus group results, as well as by CWS/CMS data that indicated that there
were significant differences in the number of months in placements by race/ethnicity. For
example, results indicated that African American children spent significantly more time in
placement than their White, Hispanic/Latino, Asian American/Pacific Islander, Vietnamese, and
Filipino peers. Also, according to preliminary findings from CWS/CMS, in Santa Clara County,
following placements in a relative home, the second most frequent placement for African
American, Native American, White and Hispanic/Latino youth, was a Foster Family Agency
(FFA); for Asian American/Pacific Islander and Vietnamese youth it was a Foster Family Home
(FFH). Almost 14% of Native American and 18% of Asian American/Pacific Islander youth in
OHP were placed in the Children's Shelter in December, 2000, a percentage that was higher than
any of the other ethnic/racial groups. We also argued that little was known about specific
pathways through the system and ways in which the pathways differed for various racial and
ethnic groups. Our findings from Phase 1 indicated that much of the research on children in the
CWS had focused on factors related to movement in and out of the system, but that little
information existed on the actual experiences of children in care and the individual and family
characteristics that are associated with these experiences. While focus group members
frequently stated that once a child of color entered the system, it was very hard for that child to
exit, little information existed that could explain what happened once that child was in the
system.
Phase 2 was designed to focus on the actual experiences of children in care and the
individual and family-related characteristics that were associated with these experiences. Results
based on extensive, in-depth reviews of 403 closed child welfare case records, a parallel
descriptive analysis of 1720 closed cases within the CWS/CMS database, and key informant
interviews with managers and supervisors in the County's DFCS indicated that families
195
belonging to each of four different racial/ethnic groups (Latino, African American,
Asian/Pacific Islander and White) were characterized by different constellations of risk factors,
and that once in the system, the children had different experiences that resulted in different
outcomes. Additionally, results indicated that services recommended for the families and
children were limited to a small array of traditional services - a one-size fits all approach - that
did not necessarily meet the needs of the culturally diverse families and children. (Please see
Attachment 8 for a summary chart of Phase 2 findings.) We concluded that the paucity of
recommended services indicated that children and families of color were not likely to be
provided with sufficient preventive and supportive services and that the traditional child welfare
services might not meet the particular needs of these unique and diverse groups. These findings
are supported by research literature indicating relatively little empirical evidence that traditional
child welfare services are effective in maintaining or reunifying children and improving famUy
functioning. Even less research has been conducted on the impact of child welfare practices for
children and families of color. Yet because children and families of color are disproportionately
represented in the CWS, we contend that the effectiveness ofchild welfare practices for these
groups is of particular interest.
Phase 3 focused on agency practice and ways in which current practice interacted with
child, family and cultural characteristics of different cultural/ethnic groups. The primary goal of
Phase 3 was to assess key practices for children and families of color in the CWS both at the
County and state levels in order to provide information that might help generate strategies aimed
at reducing ethnic/racial disparities and develop more effective services and supports for children
and families of color. Methods included semi-structured in-depth interviews and focus groups
with DFCS supervisors and managers, social workers and parents involved in the CWS; a
qualitative and quantitative comparison of child welfare practices and outcomes between the
DFCS South County and Main County offices; a reanalysis of cases from Phase 2 with
successful outcomes and an analysis of factors that contributed to those positive outcomes; a
supplemental data collection for our case record review sample so as to include information on
court ordered changes to the social worker recommendations at the jurisdictional/dispositional
hearing; and, a statewide survey of county child welfare directors, managers or supervisors in
California counties and quantitative analyses using a statewide database consisting of countylevel characteristics.
As mentioned earlier, Phases 1 and 2 confirmed that certain ethnic groups were overrepresented· (i.e., Latinos, African Americans) while others were under-represented (i.e., Asian
American/Pacific Islanders) in the County's CWS. Phase 2 results also indicated that there were
significant ethnic differences in demographic, system-related, and psychosocial characteristics
and that many of these variations appeared at early choice points in the system. During Phase 3,
we were interested in examining ethnic/racial differences in outcomes at later stages of the
system. Using data from our Phase 2 Case Record Review sample and defining successful
outcomes according to US DHHS guidelines (i.e., safety, permanency, and family and child
well-being), we found that there were minimal ethnic differences in successful outcomes. Phase
3 results suggested that when success was defined according to DHHS guidelines, ethnic
differences at the conclusion of the case were not evident. One interpretation for this finding
might be that while there is a clear disproportion of ethnic groups in the system, once in the
system, children as a whole, regardless of race/ethnicity, are faring at about the same level.
196
Given that Phase 2 results found that children and families of color in Santa Clara County's
CWS were receiving a "one-size fits all" approach to services, an approach that may seem
equitable and logistically efficient, but perhaps, ultimately ineffective in serving a diverse group
of children and families.
In order to further examine child welfare practices at the County level and to gain a better
understanding than the one gleaned solely through the use ofCWS/CMS data, we conducted a
series of semi-structured in-depth interviews with DFCS managers and supervisors and focus
groups with social workers and parents involved in the CWS. Our results indicated that in-home
services, family conferencing, and team decision-making (TOM) were viewed by participants as
ways to improve decision-making with families from diverse cultural and ethnic groups.
Participants noted that both formal and informal collaborative relationships with other service
delivery systems, such as domestic violence and substance abuse treatment systems, were critical
for successful interventions with children and families. Furthermore, participants recognized the
importance of training and cross-training to enhance the capacity of helping professionals in
child welfare and other systems to better address the needs of children and families of color.
We were also interested in gathering information from counties throughout California
regarding promising practices for families and children of color. This information was gathered
through a statewide survey administered to child welfare directors and program administrators.
Similar to findings on practices identified within Santa Clara County's DFCS, practices at the
statewide level were described as effective with children and families of color because they are
inclusive, collaborative, culturally appropriate and involve client's communities. Although the
research literature does not provide convincing evidence of the effectiveness of these practices,
the fact that respondents reported these practices as those that are most effective with children
and families of color suggests that they may serve as promising models for children and families
of color and that more rigorous evaluations of these practices are needed (Please see Chart 1 on
p. 192 of the Phase 3 Final Report for a summary of promising practices for children and
families of color in the CWS).
Although practices such as orientation of new clients; family and team decision-making;
preventive services; substance abuse services; cultural competency; collaboration with other
agencies and systems; a strength-based approach; and the use of relative placements were
described as having a positive impact on children and families of color, certain negative practices
and contexts were cited as possibly impeding their effective implementation. For instance,
having to screen out large numbers of inappropriate referrals; inconsistency in decision-making
practices; a shortage of services, particularly substance abuse and preventive services; lack of
client access to services; difficult protocols for placing children with kin; gaps in cultural
competency; time limits; and agency-level factors such as heavy caseloads, staff shortages,
substantial amounts of paperwork, lack of access to information about resources; and confusion
about the agency's overall mission and key policies all impeded implementation of the best
practices. These findings suggest that in addition to building on current positive practices for
children and families of color, CWS stakeholders should also actively work both internally and
in collaboration with partners to reduce barriers to these best practices.
197
Results from focus groups conducted during Phase 1 and key informant interviews held
during Phase 2, as well as current research literature suggest that agency organizational culture
and context may have an impact on child welfare practices and resulting family and child
outcomes. During Phase 3, we conducted both quantitative and qualitative analyses comparing
South County and the Main Office that enabled us to examine various contextual factors at the
organizational level, in this case, small vs. large and urban vs. rural, and their impact on
outcomes for children and families of color. Results indicated that children in South County
appeared to have better outcomes, i.e., they had fewer number of unique placement homes in the
current episode, had a shorter average stay per placement, a shorter length of time in out-ofhome placement, and were younger at time of case closure. Qualitative findings were congruent
with the quantitative data and provided some insight into the dynamics that might contribute to
this difference. South County study participants described a culture of commitment to
maintaining and reunifying families. This commitment was described as integral to the
philosophy of managers and social workers and reflected in the expectations communicated to
social workers by most supervisors and by peers. The practice of providing an orientation to the
child welfare system, time limits, the courts, and how to work with social workers might also
contribute to the differences found in the quantitative analysis. Study participants frequently
described South County as similar to a small county or rural area. This geographic difference,
and factors related to an organizational culture that emphasizes a commitment to maintaining and
reunifying families, the practice of providing an orientation to the child welfare system, as well
as the finding that fewer workers were assigned during the course of the case may all be related
to these positive outcomes for children in South County. Further, our multivariate analyses that
examined case characteristics and system-related factors related to reunification indicated that
when taking other factors into account, fewer workers, shorter case duration, and fewer
placements were the strongest predictors of family reunification.
The issue of disproportionate involvement of children of color in the CWS has long been
an issue of concern for CWS workers, clients, researchers and government and community
groups. More recently, it has been the focus of much national attention. Santa Clara County's
Children of Color Study is one of the few to examine this issue at the local level. We applaud the
County Board of Supervisors and DFCS for their openness and willingness to undertake such a
project. We acknowledge that there are multiple stakeholders who may be interested in
identifying a "magic bullet" or a single, straightforward explanation for the system's
racial/ethnic disproportionality. Single, straightforward explanations are appealing as they might
lead to quickly implemented solutions. However, results of this study indicate that rather than
one primary causal factor, there appear to be numerous and interrelated factors associated with
the disproportionate involvement of children of color in Santa Clara County's CWS. It is our
contention, that factors that operate simultaneously and in complex ways at the individual,
family and system level with differing patterns across diverse racial and ethnic groups contribute
to disproportionality throughout the system. Overall, findings suggest that efforts to address
racial and ethnic disparities in the CWS should focus on the diverse needs of the different ethnic
and racial groups involved in the CWS and the design of more culturally specific and effective
prevention and intervention programs to meet those needs.
198
Implications for Santa Clara County's Child Welfare System
The following are suggestions based on results from all three phases of the Children of Color
Study. It is our hope that these recommendations may be helpful in guiding Santa Clara
County's continuing efforts to address the complex issue of racial/ethnic disparities in its CWS.
l. There is a need to expand available child welfare services to better serve families and
children ofcolor. Phase 2 results indicated that reliance on a small array of
traditional formal services does not appear to meet the needs of the highly diverse
ethnic/racial family groups involved in the County's CWS.
•
More preventive and early intervention services for vulnerable families of color
should be implemented and carefully evaluated. The findings from the qualitative
component of Phase 3 affirmed the value of many efforts that are already in place,
such as culturally specific ER response units and family resource centers, and
other that are in the process of implementation including efforts to reduce
caseloads, initiation of team decision making, and participation in the Family-toFamily initiative.
In addition, home visitation services may be exceptionally helpful in addressing
the needs of vulnerable, at-risk families from diverse racial/ethnic groups.
Research has indicated that programs of home visitation that promote positive
health-related behaviors in mothers of young children, competent care of their
children and linkage with needed health care and human services, reduce rates of
criminality, problems related to substance abuse and child abuse and neglect
among young, unmarried, isolated, poor mothers.
•
The use of more non-traditional, culturally sensitive services is clearly -warranted.
Key informants in all three phases of the study underscored this point. A paucity
of social services, particularly multi-lingual services, was cited as a significant
barrier for many families of color. Interviewees discussed the shortage of
substance abuse treatment programs, particularly those geared for women with
children and people whose primary language is not English. Multi-lingual and
culturally appropriate domestic violence services, parenting classes, and other
social services were considered in need of development.
•
Piloting new, innovative services and evaluating their success with different
ethnic/racial groups could contribute to a more diverse array of culturally specific
programs for families and children.
•
Phase 3 results also indicated that practices that are inclusive, collaborative,
culturally appropriate and involve client's communities appear to be best suited to
meeting the needs of children and families of color. Although the research
literature does not provide convincing evidence of the effectiveness of these
199
practices, the fact that respondents reported these practices as those that are most
effective with children and families of color suggests that they may serve as
promising models for children and families of color.
•
More services targeted to fathers and programs that are formulated and delivered
within a family-based framework would be of benefit to families and children of
color involved in the CWS.
•
In addition to building on current positive practices, and initiating new programs
for children and families of color, CWS stakeholders should actively work
internally and in collaboration with partners to reduce barriers to implementing
such practices. Barriers such as large numbers of inappropriate referrals;
inconsistency in decision-making practices; a shortage of services, particularly
substance abuse and preventive services; lack of client access to services; difficult
protocols for placing children with kin; time limits; and agency-level factors such
as heavy caseloads, staff shortages, substantial amounts of paperwork, lack of
access to information about resources; and confusion about the agency's overall
mission and key policies were cited during Phase 3 focus groups as impediments
to the implementation of cited best practices.
2. There is a need to involve multiple social service systems in a comprehensive and
coordinated effort to meet the needs ofchildren and families ofcolor. Results from
Phases 2 and 3 indicated that the problems experienced by families across the
different racial/ethnic groups span multiple systems including: mental health, juvenile
justice, adult criminal justice, substance abuse, and welfare. Statistics presented in
the literature review section of the Phase 2 Final Report also indicate that families of
color are· involved in systems other than child welfare in high numbers. Prevention
and intervention efforts should involve a deliberate and organized coordination of
these multiple systems. During Phase 3 interviews and focus groups, collaboration
with other agencies and systems concerned with the safety and welfare of children
and families of color was highlighted as a beneficial practice for children and families
of color.
•
Developing interagency formal agreements and connections with liaisons in
other agencies would be beneficial as it can facilitate collaboration and
minimize the potential for agencies to give conflicting messages and mandates
to clients. Phase 3 participants suggested that social workers were better able
to divert cases or get a more complete picture of the family when two agencies
are working together. Other agencies mentioned in this context include
probation, hospital, medical providers, family and mental health agencies,
CalWORKS, substance abuse treatment agencies, domestic violence agencies
and other entities thatrriay be involved with clients.
•
The Greenbook project, an initiative to better coordinate domestic violence
and child welfare services, could serve as a model for interdisciplinary
problem solving and policy development. The Greenbook project successes
200
could help inform evolving collaboration with other key systems, such as the
substance abuse treatment system.
•
Other collaborative efforts such as outstationed social workers, co-location of
services and Family Drug Court could serve as models for best practices for
children and families of color. Social workers and supervisors, interviewed
during Phase 3, described the family drug court as a model for effectively
working with substance abusing families in the child welfare system.
Participants described having ER social workers at different sites such as the
Family Violence Center and police stations as a valued practice.
3. Ways in which Agency organizational context and culture impacts families and
children ofcolor merits attention. It is possible that characteristics of a large
bureaucratic organization impede the development and facilitation of practices that
would be of benefit to families and children of color. For example, organizational
goals such as having a unified agency mission, a shared organizational culture and
personal contact with clients may be difficult to achieve in the face of complex and
competing demands inherent in a large organization.
•
Practices, such as providing a group orientation to families that may assist them in
navigating through the system may be easily strengthened in South County
Offices and adapted to other regions of Santa Clara County, as recommended by
study participants. This practice may be particularly helpful to low-income
families and families of color that ·may be intimidated by or unfamiliar with child
welfare and related systems.
•
Other practices, such as vertical case management, may hold promise for other
regions of Santa Clara County, though the success of this practice may be linked
to the "small county" dynamic described by South County study participants.
Some participants in focus groups that took place at the main·office mentioned
this model as promising while others asserted that specialized expertise in
different areas of social work was an advantage in providing quality services that
are informed by "the most current information."
•
Further examination of ways in which the culture of commitment to maintaining
and reunifying families is created, communicated, and continued merits attention.
In addition, an exploration of how this culture might be adapted to other parts of
the county would be beneficial.
4. More attention thatfocuses on ways in which system level changes at the federal,
state and local levels have an impact on families and children ofcolor is warranted
With the passage of the Multi-Ethnic Placement Act-Interethnic Adoption Provision
(MEPA-IEP, 1996) and the Adoption and Safe Families Act (AFSA, 1997), safety,
permanency and expedited placements have taken precedence in the CWS. Results
based on focus group and interview data from all three phases of the study indicated
201
that the shift toward expedited placement and an emphasis on permanency has
certainly influenced the culture and ways in which social services are being delivered
in California in general, and in Santa Clara in particular and consequently raises
several major concerns for children of color in the County's CWS. In combination
with the new regulations, the characteristics of families, children and communities of
color including chronic poverty, substance abuse, lack of community social
organization or racial segregation that increase chances of entering the CWS may
create overwhelming barriers to successful reunification for children of color in the
CWS and keep them in the system longer.
•
Regular assessment of ways in which shortened timelines, early termination of
parental rights, bypass criteria, as well as changes in adoption regulations and
incentives have had an impact on children and families of color would be
beneficial.
•
A close monitoring of the effect of new policies, implementation of new programs
or changes in the agency's organizational structure on outcomes for families and
children of color would contribute to a better understanding of the differential
impact these innovations might have on different racial and ethnic groups.
5. In order to better assess whether safety, permanency, and family and child well-being
are maintained and enhanced, records should include more complete and measurable
information. We acknowledge that child welfare case records were not originally
intended for research and that that improving the collection and condition of child
welfare data is no easy task given the issues of limitations of the data system,
compatibility among systems of data recording and storage, and training needs.
However, concerted efforts in this area are necessary if accurate and useful
information is to be obtained and used to provide feedback on the effectiveness of
programs and services in improving outcomes for families and children in the
County's CWS.
•
The amount of missing data and the manner in which information is assessed also
implies the need for more reliable and valid methods of recording of information.
•
It was difficult to assess the successful cases and outcomes in Santa Clara
County's CWS. Are the proportions of children reunified with their families or of
those adopted high or low? And compared to what? Implementing better record
keeping and data collection over time on key characteristics, based on an agreedupon operational definition would better address the question of effectiveness and
performance. Implementing these practices across counties and across states
would also facilitate evaluation.
6. Future research efforts should be directed toward examining neighborhood,
community, and other macro-level factors particular to Santa Clara County and ways
in which these factors interact with individual, family and system-related
characteristics to propel children into and maintain them in the CWS. Research
202
findings consistently point to a relationship between poverty and child maltreatment.
. Characteristics associated with communities and neighborhoods of poverty including;
living in a high crime area, living in public housing, having larger numbers of
dependent children, and receiving welfare benefits might place children of color at an
increased risk of entering and staying in the CWS.
It was our intention in conceptualizing the original design and scope of work for the
Children of Color Study to examine community level factors and their interaction
with individual and family characteristics during Phase 3 of the study. However, due
to the interests of various stakeholders, the community component was not included
in the current study. It is our contention that in order to understand more fully the
disproportionate involvement of children and families of color in the CWS, the
processes by which individual, family and system level factors interact with
characteristics that are, perhaps unique to Santa Clara County and impact families and
children of color and their subsequent entry into the CWS need to be examined.
203
References
Akin, B. A., & Gregoire, T. K. (1997). Parents' views on child welfare's response to
addiction. Families in Society: The Journal of Contemporary Human Services, 393-404.
Alter, C. F. ( 1985). Decision-making factors in cases of child neglect. Child Welfare,
64(2), 99-111.
American Humane Association (2003). What is Family Group Decision Making? Online,
retrieved July 1, 2003, from:
http://www.americanhumane.org?site?PageServer?pagemane=pc_ fgdm _ what_is.
Ards, S. D., Myers, S. L., Malkis, A., Sugrue, E., & Zhou, L., (2003). Racial
disproportionality in reported and substantiated child abuse and neglect: An examination of
systematic bias. Children and Youth Services Review, 25(5/6) 375-392.
Atkinson, L., & Butler, S. (1996). Court-ordered assessment: Impact of maternal
noncompliance in child maltreatment cases. Children and Youth Services Review, 20(3), 185190.
Avery, R. J. (2000). Perceptions and practice: Agency efforts for the hardest-to-place
children. Children and Youth Services Review, 22(6), 399--420.
Baird, C., & Wagner, D. (2000). The relative validity of actuarial- and consensus-based
risk assessment systems. Children and Youth Services Review, 22(11/12), 839-871.
Banach, M. (1998). The best interests of the child: Decision-making factors. Families in
Society: The Journal of Contemporary Human Services, 79(3), 331-341.
Barth, R. P. (1991). An experimental evaluation of in-home child abuse prevention
services. Child Abuse and Neglect 15, 363-375.
Barth, R. P., Snowden, L. R., Broeck, E.T., Clancey, T., Jordan, C., & Barusch, A. S.
(1987). Contributors to reunification or permanent out-of-home care for physically abused
children. Journal of Social Service Research, 9(2/3), 31--45.
Bath, H. I., Richey, C. A., Haapala, D. A. (1992). Child age and outcome correlates in
intensive family preservation services. Children and Youth Services Review, 14, 389--406.
Bay Area Social Services Consortium (2000). Concurrent Planning: The Influence of
Bypass and Poor Prognosis Indicators on Child Welfare Outcomes. Berkeley, CA: Author.
Beeman, S. K., Hagemeister, A. K., & Edleson, J. L. (1999). Child protection and
battered women's services: From conflict to collaboration. Child Maltreatment, 4(2), 116-126.
204
Berry, M. (1992). An evaluation of family preservation services: Fitting agency services
to family needs. Social Work, 37(4), 314-321.
Berry, M. (1991). The assessment of imminence of risk of placement: Lessons from a
family preservation program. Children and Youth Services Review, 13, 239-256.
Bowen, K. (2000). Child abuse and domestic violence in families of children seen for
suspected sexual abuse. Clinical Pediatrics, 39, 33-40.
Brown, T. G., Werk, A., Caplan, T., & Seraganian, P. (1999). Violent substance abusers
in domestic violence treatment. Violence and Victims, 14(2), 179-190.
Brooks, D., & James, S. (2003). Willingness to adopt Black foster children: Implications
for child welfare policy and recruitment of adoptive parents. Children and Youth Services
Review, 25(5/6), 463-489.
Brunk, M., Henggeler, S. W. & Whelan, J.P. (1987). Comparison ofmultisystemic
therapy and parent training in the brief treatment of child abuse and neglect. Journal of
Consulting and Clinical Psychology. 55(2), 171-178.
California Department of Social Services (2003). Structured Decision Making. Online,
retrieved July 14, 2003 from: http://www.childsworld.ca.gov/Structured 352.htm.
California Department of Social Services (2001). Wraparound Services Pilot, Fiscal
Questions and Answers. Online, retrieved, May 5, 2001 from:
http://childsworld.org/services/famcenwrap.htm
California Department of Social Services (1999). Panel on Race and Bias in Risk
Assessment Thirteenth National Roundtable on Child Protective Services July 7-9, San
Francisco, CA.
Camasso, M. J., & Jagannathan, R. (2000). Modeling the reliability and predictive
validity of risk assessment in child protective services. Children and Youth Services Review,
22(11/12), 873-896.
Carlo, P. ( 1993). Parent education vs. parent involvement: Which type of efforts work
best to reunify families? Journal of Social Service Research, 17(1/2), 135-1°50.
Carten, A. J. (1996). Mothers in recovery: rebuilding families in the aftermath of
addiction. Social Work 41(2) 214-223.
Carter-Black, J. (2002). Transracial adoption and foster care placement: Worker
perception and attitude. Child Welfare, 71(2), 337-370.
Casey Foundation (2003a). Team Decision Making. Online, retrieved July 13, 2003 from:
http://aecf.org/initiatives/familytofamily/tools/decisions/htm.
205
Casey Foundation (2003b). Family to Family Home Page. Online, retrieved July 13, 2003
from: http://www.aecf.org/initiatives/familytofamily/
Chaffin, M., Bonner, B. L., & Hill, R. F. (2001). Family preservation and family support
programs: Child maltreatment outcomes across client risk levels and program types. Child Abuse
and Neglect, (25), 1269-1289.
Chao, R., & Tseng, V. (2002). Parenting of Asians. In M. H. Bornstein (Ed.). Handbook
of Parenting Vol. 4 Social Conditions and Applied Parenting (pp. 59-93). Mahwah, NJ:
Lawrence Erlbaum Associates.
Cash, S. J., & Berry, M. (2002). Family characteristics and child welfare services: Does
the assessment drive service provision? Families in Society, 83(5/6), 499-507.
Children NOW. (2001). California Report Card 2001: County Information. Online,
retrieved, August 20 2003, from: http://www.childrennow.org/califomia/rc-2001/county-pdflist.htm#top2.
Children's Research Center (2003). Structured Decision Making. Online, retrieved
July 15, 2003 from: http://www.nccd-crc.org/crc/decision.htm.
Clark, H. W. (2001). Residential substance abuse treatment for pregnant and postpartum
women and their children: Treatment and policy implications. Child Welfare, 80(2), 179-198.
Close, M. M. ( 1983). Child welfare and people of color: Denial of equal access. Social
Work Research and Abstracts, 19(4), 13-20.
Cohen, E. P. (2003). Framework for culturally competent decision-making in child
welfare. Child Welfare, 82(2), 143-155.
Courtney, M., Piliavin, I., Grogan-Kaylor, A., & Nesmith, A. (2001). Foster youth
transitions to adulthood: A longitudinal view of youth leaving care. Child Welfare, ,80, 685-717.
Courtney, M. E. (1997). The politics and realities oftransracial adoption. Child Welfare,
76(6), 749-779.
Courtney, M. E., Barth, R. P., Berrick, J. D., Brooks, D., Needell, B., & Park, L. (1996).
Race and child welfare services: Past research and future directions. Child Welfare 75(2) 99-137.
Courtney, M. E. (1994). Factors associated with the reunification of foster children with
their families. Social Service Review, 68(1), 81-108.
Cuddeback, G. S., & Orme, J. G. (2002). Training and services for kinship and
nonkinship foster families. Child Welfare, 91(6), 879-909.
206
Dalgleish, L. I., & Drew, E. C. (1989). The relationship of child abuse indicators to the
assessment of perceived risk and to the court's decision to separate. Child Abuse and Neglect,
.U,491-506.
Davidson-Arad, B. (2001). Predicted changes in children's quality of life in decisions
regarding the removal of children at risk from their homes. Children and Youth Services Review,
23(2), 127-143.
DePanfilis, D. (1996). Social isolation of neglectful families: A review of social support
assessment and intervention models. Child Maltreatment, 1(1), 37-52.
DePanfilis, D., & Zuravin, S. J. (2001). Assessing risk to determine the need for services.
Children and Youth Services Review, 23(1 ), 3-20.
DePanfilis, D., & Scannapieco, M. (1994). Assessing the safety of children at risk of
maltreatment: Decision-making models. Child Welfare, 73(3), 229-245.
Drake, B. (1996). Predictors of preventive services provision among unsubstantiated
cases. Child Maltreatment, 1(2), 168-175.
Dore, M. M. & Lee, J.M. (1999). The role of parent training with abusive and neglectful
parents. Family Relations, 48(3), 313-325.
dosReis, S., Zito, J.M., Safer, D. J., Soeken, K. L. (2001). Mental health services for
youths in foster care and disabled youths. American Journal of Public Health, 91 (7). 10941100.
Doueck, H.J., English, D. J., DePanfilis, D., Moote, G. T. (1993). Decision-making in
child protective services: A comparison of selected risk-assessment systems. Child Welfare,
72(5), 441-452.
Downing, J. D., Wells, S. J., & Fluke, J. (1990). Gatekeeping in child protective services:
A survey of screening policies. Child Welfare, 69(4), 357-369.
Dunford, F. W. (2000). The San Diego navy experiment: An assessment of interventions
for men who assault their wives. Journal of Consulting and Clinical Psychology, 68(3), 468476.
Dutton, D. G. (1986). The outcome of court-mandated treatment for wife assault: A
quasi-experimental evaluation. Violence and Victims, 1(3), 163-175.
Eckenrode, J., Ganzel, B., Henderson, C.R., Smith, E., Olds, D. L., Powers, J., Cole, R.,
Kitzman, H., & Sidora, K. (2000). Preventing child abuse and neglect with a program of nurse
home visitation: The limiting effects of domestic violence. The Journal of the American Medical
Association, 284(11), 1385-1396.
207
Eckenrode, J., Powers, J., Doris, J., Munsch, J., & Bolger, N. (1988). Substantiation of
child abuse and neglect reports. Journal of Consulting and Clinical Psychology, 56(1), 9-16.
English, D. J., Marshall, D. B., Coghlan, L., Brummel, S., & Orme, M. (2002). Causes
and consequences of the substantiation decision in Washington state-child protective services.
Children and Youth Services Review, 24(11), 817-851.
Farmer, E. M., Wagner, H. R., Bums, B. J., & Richards, J. T. (2003). Treatment foster
care in a system of care: Sequences and correlates of residential placements. Journal of Child and
Family Studies, 12(1), 11-25.
Federal Interagency Forum on Child and Family Statistics. America's Children: Key
National Indicators of Well-being, 2001. Federal Interagency Forum on Child and Family
Statistics, Washington DC: U.S. Government Printing Office.
Fein, E., & Staff, I. (1993). Last best chance: Findings from a reunification services
program. Child Welfare, 72(1), 25-40.
Fetsch, R. J., Schultz, C. J., & Wahler, J. J. (1999). A preliminary evaluation of the
Colorado rethink parenting and anger management program. Child Abuse and Neglect, 23(4),
353-360.
.
Findlater, J.E., & Kelly, S. (1999). Reframing child safety in Michigan: Building
collaboration among domestic violence, family preservation, and child protection teams. Child
Maltreatment, 4(2), 167-174.
359.
Flanigan, J. (1954). The Critical Incident Technique. Psychological Bulletin. 4, 327-
Fluke, J. D., Yuan, Y. T., Hedderson, J., & Curtis, P.A. (2003). Disproportionate
representation of race and ethnicity in child maltreatment: Investigation and victimization.
Children and Youth Services Review, 25(5/6), 359-373.
Forehand, R., & Kotchick, B. A. (1996). Cultural diversity: A wake-up call for parent
training. Behavior Therapy, 27, 187-206.
Fraser, J. A., Armstrong, K. L., Morris, J.P. & Dadds, M. R. (2000). Homes visiting
intervention for vulnerable families with newborns: Follow-up results of a randomized controlled
trial. Child Abuse and Neglect, 24(11), 1399-1429.
Garbarino, J., & Kostelny, K. (1992). Child maltreatment as a community problem. Child
Abuse and Neglect 16, 45 5-464.
Garland. A. F., & Besinger, B. A. (1997). Racial/ethnic differences in court referred
pathways to mental health services for children in foster care. Children and Youth Services
Review, 19(8), 651-666.
208
Garland, A. F., Hough, R., Landsverk, J. A., McCabe, K. M., Yeh, M., Ganger, W. C., &
Reynolds, B. (2000). Racial/ethnic differences in mental health care utilization among children
in foster care. Children's Services: Social Policy, Rese·arch and Practice, 3, 133-146.
Gaudin, J.M., Wodarski, J. S., Arkinson, M. K., & Avery, L. S. (1990). Remedying child
neglect: Effectiveness of social network interventions. The Journal of Applied Sciences, 15(1 ),
97-123.
Gillespie, J.M., Byrne, B., Workman, L. J. (1995). An intensive reunification program
for children in foster care. Child and Adolescent Social Work Journal, 12(3), 213-228.
Gregoire, K. A. & Schultz, D. J. (2001). Substance-abusing child welfare parents:
Treatment and child placement outcomes. Child Welfare 80(4), 433-452
Goerge, R. M. (1990). The reunification process in substitute care. Social Service
Review, 64(3), 422--457.
Gordon, A. L., McKinley, S. E., Satterfield, M. L., & Curtis, P. (2003). A first look at the
need for enhanced support services for kinship caregivers. Child Welfare, 81 ( 1), 77-96.
Guterman, N. B. (1999). Enrollment strategies in early home visitation to prevent
physical child abuse and neglect and the "universal versus targeted" debate: A meta-analysis of
population-based and screening-based programs. Child Abuse and Neglect, 23(9), 863-890.
Haller, D. L., Knisely, J. S., Elswick, R. K., Dawson, K. S., Scholl, S. H. (1997).
Perinatal substance abusers: Factors influencing treatment retention. Journal of Substance Abuse
Treatment, 14(6), 513-519.
Harman, J. S. (2001). Mental health care utilization and expenditures by children in foster
care. The Journal of the American Medical Association, 285(5), 524.
Harris, M. S., & Courtney, M. E. (2003). The interaction of race, ethnicity, and family
structure with respect to the timing of family reunification. Children and Youth Services Review,
25(5/6), 409--429.
Hines, A. M., Lee, P.A., Drabble, L., Snowden, L. R., & Lemon, K. (2002). An
Evaluation of Factors Related to the Disproportionate Representation of Children of Color in
Santa Clara County's Child Welfare System: Phase 2 Final Report. Report submitted to Santa
Clara County, CA Department of Social Services, Family and Children's Services, August 31,
2002.
Hochstadt, N. J ., Harwicke, N. J. ( 1985). How effective is the multidisciplinary
approach? A follow-up study. Child Abuse and Neglect, 9, 365-372.
209
Hosley, C. A., Gensheimer, L., & Yang, M. (2003). Building effective working
relationships across culturally and ethnically diverse communities. Child Welfare, 82(2), 157168.
Jellinek, M. S., Murphy, M., Poitrast, F., Quinn, D., Bishop, S. J., & Goshko, M. (1992).
Serious child mistreatment in Massachusetts: The course of 206 children through the courts.
Child Abuse and Neglect, 16, 179-185.
Johnson, P., & Cahn, K. (1992). Improving child welfare practice through improvements
in attorney-social worker relationships. Child Welfare, 74(2), 383-394.
Jones, L. (1998). The social and family correlates of successful reunification of children
in foster care. Children and Youth Services Review, 20(4), 305-323.
Jones, L. (1993). Decision making in child welfare: A critical review of the literature.
Child and Adolescent Social Work Journal, 10(3), 241-262.
Kapp, S. A., McDonald, T. P., & Diamond, K. L. (2001). The path to adoption for
children of color. Child Abuse and Neglect, 25, 215-229.
Karoll, B. R. & Poertner, J. (2002). Judges' caseworkers', and substance abuse
counselors' indicators of family reunification with substance-affected parents. Child Welfare,
71(2), 249-269.
· Karski, R. L. (1999). Key decisions in child protective services: Report investigation and
court referral. Children and Youth Services Review, 21(8), 643-656.
Kemp, S. P., & Bodonyi, J.M. (2002). Beyond termination: Length of stay and predictors
of permanency for legally free children. Child Welfare, 81(1), 59-86.
Kinard, E. M. (2002). Services for maltreated children: variations by maltreatment
characteristics. Child Welfare, 81(4), 617-645.
Kitzman, H., Olds, D, L., Henderson, C. R., Hanks, C., Cole, R., Tatelbaum, R.,
McConnochie, K. M., Sidora, K., Luckey, D. W., Shaver, D., Engelhardt, K., James, D., &
Barnard, K. (1997). Effect of prenatal and infancy home visitation by nurses on pregnancy
outcomes, childhood injuries, and repeated childbearing: A randomized controlled trial. The
American Medical Association, 278(8), 644-652.
Knepper, P. E., & Barton, S. M. (1997). The effect of courtroom dynamics on child
maltreatment proceedings. Social Service Review, 71(2), 288-309.
Knight, D. K., Logan, S. M., & Simpson, D. D. (2001). Predictors of program completion
for women in residential substance abuse treatment. American Journal ofD.rug and Alcohol
Abuse, 27(1), 1-15.
210
Kolbo, J. R., & Strong, E. (1997). Multidisciplinary team approaches to the investigation
and resolution of child abuse and neglect: A national study. Child Maltreatment, 2(1 ), 61-72.
Landsman, M. J. (2002). Rural child welfare practice from an organization-inenvironment perspective. Child Welfare, 81 (5), 790-819.
Landy, S., & Munro, S. (1998). Shared parenting: Assessing the success of a foster parent
program aimed at family reunification. Child Abuse and Neglect, 22(4), 305-318.
Lee, B. J., & Goerge, R. M. (1999). Poverty, early childbearing, and child maltreatment:
A multinomial analysis. Children and Youth Services Review, 21(9/10), 755-780.
Lewandowski, C. A., & GlenMaye, L. F. (2002). Teams in child welfare settings:
lnterprofessional and collaborative processes. Families in Society, 83(3), 245-256.
Lewandowski, C. A., & Pierce, L. (2002). Assessing the effect of family-centered out-ofhome care on reunification outcomes. Research on Social Work Practice, 12(2), 205-221.
Lightbourne, W. (2002). Greenbook Project Update. Report submitted to the Santa Clara
County Board of Supervisors.
Lindsey, D. (1991). Factors affecting the foster care placement decision: An analysis of
national survey data. American Journal of Orthopsychiatry, 61 (2), 272-281.
Littell, J. H. (1997). Effects of the duration, intensity, and breadth of family preservation
services: A new analysis of data from the Illinois Family First Experiment. Children and Youth
Services Review, (19)2, 17-39.
Littell, J. H., & Schuerman, J. R. (1995). A synthesis of research on family preservation
and family reunification programs. Online, retrieved January 27, 2003 from:
http://aspe.hhs.gov/hsp/cyp/fp litrev .htm
Little Hoover Commission (2003). Still in Our Hands: A Review of Efforts to Reform
Foster Care in California. Online, retrieved July 24, 2003 from:
http://www.lhc.ca.gov/lhcdir/report 168 .html
Maluccio, A. N. & Ainsworth, F. (2003). Drug use by parents: A challenge for family
reunification practice. Children and Youth Services Review, 25(7), 511-533.
Martin, L. M., Peters, C. L., & Glisson, C. (1998). Factors affecting case management
recommendations for children entering state custody. Social Service Review,--, 521-544.
McGuigan, W., Katzev, A. R., & Pratt, C. C. (2003). Multi-level determinants of
retention in a home-visiting child abuse prevention program. Child Abuse & Neglect. 27, 363380.
211
McGuigan, W., M., & Pratt, C. C. (2001 ). The predictive impact of domestic violence on
three types of child maltreatment. Child Abuse & Neglect, 25, 869-883.
McKay, M. M. (1994). The link between domestic violence and child abuse: Assessment
and treatment considerations. Child Welfare, 73(1), 29-39.
McMurty, S., & Lie, G. (1992). Differential exit rate of minority children in foster care.
Social Work Research and Abstracts, 28(1 ), 42-48.
McPhatter, A. R., & Ganaway, T. L. (2003). Beyond the rhetoric: Strategies for
implementing culturally effective practices with children, families and communities. Child
Welfare, 82(2), 103-124.
Mederos, F., & Woldeguiorguis, I. (2003). Beyond cultural competence: What child
protection managers need to know and do. Child Welfare, 82(2), 125-142.
Menanteau-Horta, D., & Yigzaw, M. (2002). Indicators of social well-being and
elements of child welfare in Minnesota rural counties. Child Welfare, 81 (5), 709-73 5.
Metsch, L. R., Wolfe, H.P., Fewell, R., McCoy, C. B., Elwood, W. N., Wohler-Torres,
B. et al. (2001 ). Treating substance-using women and their children in public housing:
Preliminary evaluation findings. Child Welfare, 80(2), 199-220.
Miller, 0. A., & Gatson, J. (2003). A model of culture-centered child welfare practice.
Child Welfare, 82(2), 235-.249.
Morrel, T. M., Elliot, J. D., Murphy, C. M., & Taft, C. T. (2003). Cognitive behavioral
and supportive group treatments for partner-violent men. Behavior Therapy, 34, 77-95.
Murphey, D. A., & Braner, M. (2000). Linking child maltreatment retrospectively to birth
and home visit records: An initial examination. Child Welfare, 79(6), 711-728.
Needell, B., Brookhart, M.A., Lee, S. (2003). Black children and foster care placement
in California. Children and Youth Services Review, 25(5/6) 393-408.
Needell, B., Webster, D., Cuccaro~Alamin, S., Armijo, M., Lee, S., Brookhart, A., Lery,
B., Shaw, T., Dawson, W., Piccus, W., Magruder, J., & Kim, H. (2003). Child Welfare Services
Reports for California. Online, retrieved June 28, 2003, from: University of California at
Berkeley Center for Social Services Research website. URL:
<http://cssr.berkeley.edu/CWSCMSreports/>
Nicholson, B. C., Brenner, V., & Fox, R. A. (1999). A community-based parenting
program with low-income mothers of young children. Families in Society, 80(3), 247-253.
Olds, D. L., Eckenrode, J., Henderson, C.R., Kitzman, H., Powers, J., Cole, R., Sidora,
K., Morris, P., Pettitt, L. M., & Luckey, D. (1997). Long-term effects of home visitation on
212
material life course and child abuse and neglect: Fifteen-year follow-up of a randomized trial.
The Journal of the American Medical Association, 278(8) 637-644.
Ondersma, S. J., Malcoe, L. H., & Simpson, S. M., (2001). Child protective services'
response to prenatal drug exposure: Results from a nationwide survey. Child Abuse and Neglect,
25, 657-668.
Palmer, S. E. (1996). Placement stability and inclusive practice in foster care: An
empirical study. Children and Youth Services Review, 18(7), 589-601.
Palmer, S. E., Brown, R. A., & Barrera, M. E. (1992). Group treatment for abusive
husbands: Long-term evaluation. American Journal of Orthopsychiatry, 62(2), 276-283.
Patterson, G. R., Chamberlain, P., & Reid, J.B. (1982). A comparative evaluation of a
parent-training program. Behavior Therapy, 13, 638-650.
Pecora, P. J., Fraser, M. W., Bennett, R. B., & Haapala, D. A. (1991). Placement rates of
children and families served by the Intensive Family Preservation Services Programs. In M. W.
Fraser, P. J. Pecora, & D. A. Haapala (Eds.), Families in Crisis: The Impact oflntensive Family
Preservation Services (pp.149-180). New York: Aldine De Gruyter.
Pennell, J., & Burford, G. (2000). Family group decision making: Protecting children and
women. Child Welfare, 79(2), 131-158.
Rossi, P. H., Schuerman, J., & Budde, S. (1999). Understanding decisions about child
maltreatment. Evaluation Review, 23(6), 579-598.
Reiff, M. I. (2001). Mental health care utilization and expenditures by children in foster
care. Journal of Developmental and Behavioral Pediatrics. 22(2), 151.
Reyes, B. I. (Ed.). (2001). A Portrait of Race and Ethnicity in California. San Francisco:
Public Policy Institute of California.
Russel, R. (1988). Role perceptions of attorneys and caseworkers in child abuse cases in
juvenile court. Child Welfare, 67(3), 205-216.
Sanchirico, A., & Jablonka, K. (2000). Keeping foster children connected to their
biological parents: The impact of foster parent training and support. Child and Adolescent Social
Work Journal, 17(3), 185-203.
Scannapieco, M. (1994). Home-based services program: Effectiveness with at risk
families. Children and Youth Services Review, 16(5/6), 363-377.
Schriver, J. M. ( 1998). Human Behavior and the Social Environment. Allyn and Bacon:
Boston.
213
Schuerman, J., Rossi, P.H., & Budde, S. (1999). Decisions on placement ·and family
preservation: Agreement and targeting. Evaluation Review, 23(6), 599-618.
Sheppard, D. I., & Zangrillo, P.A. (1996). Coordinating investigations of child abuse.
Public Welfare, Winter, 21-31.
Sieppert, J. D., Hudson, J., & Unrau, Y. (2000). Family group conferencing in child
welfare: Lessons from a demonstration project. Families in Society. 81(4), 382-391.
Smith, B. D. (2003). How parental drug use and drug treatment compliance relate to
family reunification. Child Welfare, 82(3), 335-365.
Spaid, W. M., Fraser, M. W., & Lewis, R. E. (1991). Changes in family functioning: Is
participation in Intensive Family Preservation Services correlated with changes in attitudes or
behaviors? In M. W. Fraser, P. J. Pecora, & D. A. Haapala (Eds.), Families in Crisis: The Impact
oflntensive Family Preservation Services (pp.131-148). New York: Airline De Gruyter.
Stevenson, J. (1999). The treatment and long-term sequelae of child abuse. Journal of
Child Psychology and Psychiatry, 40(1 ), 89-11 I.
Suleiman, L. P. (2003). Beyond cultural competence: Language access and Latino civil
rights. Child Welfare, 82(2). 185-200.
Taft, C. T., Murphy, C. M., Elliot, J. D., & Keaser, M. C. (2001). Race and demographic
factors in treatment attendance for domestically abusive men. Journal of Family Violence, 16(4),
385-400.
Templeman, S~ B., & Mitchell, L. (2002). Challenging the one-size-fits-all myth:
Findings and solutions from a statewide focus group of rural social workers. Child Welfare,
fil.(5}, 757-772.
Thieman, A. A., & Dail, P. W. (1997). Predictors of out-of-home placement in a family
preservation program: Are welfare recipients particularly vulnerable? Policy Studies Journal,
25(1) 124-139.
Toffalo, D. A. (2000). An investigation of treatment integrity and outcomes in
wraparound services. Journal of Child and Family Studies, 9(1), 351-361.
Tumlin, K. C., & Geen, R. (2000). The Decision to Investigate: Understanding State
Child Welfare Screening Policies and Practices. Online, retrieved, February 7, 2003 from:
http://www.urban.org/url.cfm?ID=309534
U.S. Census Bureau (2001). Overview of Race and Hispanic Origin, Census 2000 Brief.
Online, retrieved November 20, 2002 from: http://www.census.gov/prod/2001pubs/c2kbr011.pdf.
214
U.S. Census Bureau (2000). State and County Quick Facts: California. Online, retrieved
March 3, 2003 from: http://quickfacts.census.gov/qfd/states/06000.html
U.S. Department of Health and Human Services, Administration for Children, Youth and
Families, Children's Bureau (2003a). Decision-making in Unsubstantiated Child Protective
Services Cases. Washington DC: U.S. Government Printing Office.
U.S. Department of Health and Human Services. (2003b). Protecting the Well-being of
Children. Online, retrieved April 16, 2003 from:
http://www.scf.hhs.gov/news/facts/chilwel fpr .htm
U.S. Department of Health and Human Services. (2002). The Adoption and Foster Care
Analysis Reporting System Report. Online, retrieved July 16, 2003 from:
http://www.scf.hhs.gov/programs/cb/publications/afcars/report7 .htm.
U.S. Department of Health and Human Services, Administration for Children, Youth and
Families (2001). Child Maltreatment 2001. Washington DC: U.S. Government Printing Office.
U. S. General Accounting Office (1998). Agencies Face Challenges Securing Stable
Homes for Children of Substance Abusers. GAO/HEHS 98..;182. Washington DC: U.S.
Government Printing Office.
U.S. General Accounting Office (1997). Child Welfare: States' Progress in Implementing
Family Preservation and Support Services. GAO/HEHS 97-34. Washington DC: U.S.
Government Printing Office.
Vidal de Haymes, M., & Simon, S. (2003). Transracial adoption: Families identify issues·
and needed support services. Child Welfare, 82(2), 251-272.
Wald, M. S., Woolverton, M. (1990). Risk assessment: The emperor's new clothes?
Child Welfare, 69(6), 483-511.
Walton, E., Fraser, M. W., Lewis, R. E., Pecora, P. J., & Walton, W. K. (1993). In-home
family-focused reunification: An experimental study. Child Welfare, 72(5), 473--487.
Wells, K., & Guo, S. (2003). Mothers' welfare and work income and reunification with
children in foster care. Children and Youth Servic.es Review, 25(3), 203-224.
Wells, K., & Guo, S. (1999). Reunification and reenty of foster children. Children and
Youth Services Review, 21(4), 273-294.
Wells, K., & Whittington, D. (1993). Child and family functioning after intensive family
preservation services. Social Service Review, March, 55-83.
Wells, S. J., Fluke, J. D., & Brown, C.H. (1995). The decision to investigate: Child
protection practice in 12 local agencies. Children and Youth Services Review, 17(4), 523-546.
215
Westat, Inc. (2001a). Caseworker decision-making. In Assessing the context of
permanency and reunification in the foster care system. Online, retrieved January 27, 2003 from:
http://aspe.hhs.gov/hsp/fostercare-reunifO 1/chapter? .htm
Westat, Inc. (2001b). Evaluation of family preservation and reunification programs.
Online, retrieved January 27, 2003 from: http://aspe.hhs.gov/hsp/fampres94/index.htm
Westat, Inc. (2001c). The role of race in parental reunification. In Assessing the context
of permanency and reunification in the foster care system. Online, retrieved January 27, 2003
from: http://aspe.hhs.gov/hsp/fostercare-reunifO 1/chapter6.htm
Whitney, P., & Davis, L. (1999). Child abuse and domestic violence in Massachusetts:
Can practice be integrated in a Public Child Welfare Setting? Child Maltreatment, 4(2), 158166.
Wolfe, D. A., Edwards, B., Manion, I., & Koverola, C. (1988). Early intervention for
parents at risk of child abuse and neglect: A preliminary investigation. Journal of Consulting and
Clinical Psychology, 56(1), 40--47.
Wolock, I. (1982). Community characteristics and _staff judgments in child abuse and
neglect cases. Social Work Research and Abstracts, 18(2), 9-15.
Document
Report provides findings from Phase 3 that was conducted from September 2002 to January 2003 focusing on agency practice and ways in which current practice interacts with child, family and cultural characteristics of different racial / ethnic groups. The study addressed three overall aims: 1) to assess identified key practices for children and families of color in Santa Clara County's CWS, 2) to examine the influence of the court system on service recommendations for children and families and, 3) to conduct a statewide comparative analysis of identified key practices that may affect the disproportionate representation of children of color in the CWS
Initiative
Collection
James T. Beall, Jr.
Content Type
Report
Resource Type
Document
Date
09/02/2003
District
District 4
Creator
The Child Welfare Research Team, College of Social Work, San Jose State University
Language
English
Rights
No Copyright: http://rightsstatements.org/vocab/NoC-US/1.0/