Home > Programs & Funding > Profiles of the Title IV-E Child Welfare Waiver Demonstration Projects > California - Intensive Service Options

California

Demonstration Type:	Intensive Service Options
Approval Date:	August 19, 1997
Implementation Date:	December 1, 1998
Completion Date:	December 31, 20051
Interim Evaluation Report Date:	May 30, 2001
Final Evaluation Report Date:	September 16, 2004

Target Population

California's title IV-E waiver demonstration targeted three groups of title IV-E-eligible children: (1) those at risk of out-of-home placement; (2) those currently in out-of-home placement with the permanency goal of family reunification, adoption, or guardianship; and (3) other children in out-of-home care who without intensive services would otherwise remain in care or move to a higher level of care.

Jurisdiction

This demonstration was implemented in seven California counties: Alameda, Fresno, Humboldt, Los Angeles, Riverside, Sacramento, and San Luis Obispo.

Intervention

California's Child Welfare Demonstration Project was approved on August 19, 1997. Originally, California proposed to implement and evaluate three new approaches to child welfare services: a Kinship Permanence Component, an Extended Voluntary Placement Component, and an Intensive Services Component. The State discontinued the Extended Voluntary Placement component in August 2000 due to slow implementation and low enrollment. In addition, California discontinued the Kinship Permanence component when the statewide program KinGap was implemented and funded through TANF savings.

The Intensive Services component was scheduled to end on September 30, 2003, but it continued to operate under short-term waiver extensions until December 31, 2005. California tested two distinct intensive service models: Wraparound services and Family Group Decision Making (FGDM). Five counties (Alameda, Humboldt, Los Angeles, Sacramento, and San Luis Obispo) implemented Wraparound programs, while two counties (Fresno and Riverside) implemented FGDM. Since many human services in California (including child welfare) are county administered, each county developed a highly individualized approach to its intensive services intervention.

Evaluation Design

The State's evaluation consisted of outcome and process components, as well as a cost analysis. Using an experimental design with random assignment, the overarching hypothesis of California's evaluation was that intensive service options would be just as cost-effective, and lead to better child welfare outcomes, as traditional child welfare services. For the process component of the evaluation, the State examined the changes required to implement the interventions and the context in which county programs operated. In addition, the process evaluation included a model fidelity assessment that explored the degree to which program implementation remained consistent within the philosophies and implementation objectives of each intervention.

The State's outcome evaluation measured several child welfare outcomes of interest, including (1) the number of children placed in group homes; (2) the number of placement changes per child; (3) length of time in out-of-home care; (4) child safety (as indicated by child abuse and neglect reports, removal from the home, child mortality, and adjudicated delinquency); (5) the number of children in out-of-home placement moved to less restrictive placement settings; and (6) child permanency, specifically, reunification with birth parents. The State also measured child well-being and satisfaction with services.

Study Sample

California originally planned to assign a total of 2,665 children to the ISC at a 5:3 ratio, with 1,666 children in the experimental group and 999 in the control group. The State subsequently reduced the initial sample size because some counties either did not implement the program model or terminated their demonstrations early. As of September 2004, a total of 664 children were enrolled in the demonstration (including both the FGDM and Wraparound Service components), with 421 children in the experimental group and 243 in the control group.

Study Limitations

California's evaluators noted several limitations specific to the FGDM impact study, including small sample sizes, the distal nature of the outcomes of interest, and contamination of the research design due to control group families receiving services similar to FGDM.

Evaluation Findings

Process Evaluation

1. Family Group Decision Making
   Fresno County opted to implement FGDM in its Voluntary Family Maintenance Unit as a means of preventing placements for at-risk families, whereas Riverside County used its waiver FDGM program to facilitate placement stability and permanence for a population of children already in placement.
   • Staff involved in the intervention were continually enthusiastic about FGDM throughout the study period, as were the families themselves. Fresno staff perceived agency managers as being "on board," despite constant concerns about the fiscal implications of FGDM. Riverside staff were less confident of agency support, especially in the latter part of the study.
   • Adequate staffing was a concern for both counties throughout the study. Fluctuations in staffing were directly related to enrollment activity; for example, the loss of a FGDM coordinator in one county temporarily brought its program to a complete halt.
   • Some contextual challenges remained intractable throughout the demonstration. Families brought with them overwhelming socio-economic issues, such as intergenerational substance abuse, poverty, and under-employment. Gaps in community resources persisted throughout the project, including an inadequate number of foster homes, the lack of rural services, and high unemployment rates.
   • Results from the model fidelity study indicated that both Fresno and Riverside Counties implemented their intended model of FGDM. Both counties were highly effective at implementing the appropriate phases of the FGDM model, including (1) referral to a trained coordinator, (2) preparation and planning, (3) the FGDM meeting, and (4) follow-up.
2. Wraparound Service Model
   Alameda County, Humboldt County, Los Angeles County, Sacramento County, and San Luis Obispo County participated in the Wraparound component of the waiver project evaluation.
   • A major issue facing all counties was the identification of a principal caregiver at the time of enrollment into the project. The Wraparound Service model was predicated on the presence of at least one caregiver, in combination with the child. However, children in the child welfare system, particularly children in the highest levels of group care, often lacked an identified caregiver. The issue of identifying a primary caregiver remained unresolved during the demonstration and called into question the appropriateness of a Wraparound Service model for a child welfare population.
   • The enrollment/intake process was crucial to the successful implementation of the Wraparound program. The county with the most successful intake process developed a specialized intake coordinator position to meet with the child and family after program referral to explain the evaluation and obtain their informed consent to participate in research. In contrast, implementation problems were much more common in counties in which the enrollment process was the responsibility of a case-carrying social worker (i.e., child welfare worker or probation worker).
   • Adequate staffing was one of the most difficult problems faced by counties implementing Wraparound services. The intensive nature of Wraparound work provided a number of disincentives, making staff recruitment difficult. Problems with staff recruitment and retention delayed or interrupted project implementation in some cases.
   • Counties reported challenges with implementing a Wraparound Service model within existing county fiscal structures. Funding streams for child welfare and mental health services are often categorical in nature, and counties' existing accounting infrastructures were not set up to accommodate the fiscal flexibility inherent in a Wraparound Service model.
   • Model fidelity was tested in Alameda County using an interview battery called the Wraparound Fidelity Index, or WFI. The WFI Overall Score indicated a statistically significant difference between the average percentage score of the Wraparound group (78 percent) and the control group (67 percent) receiving traditional child welfare services. These findings provide initial evidence that the experimental and control groups were receiving different interventions, and that the Alameda Wraparound project adhered closely to its original service model.

Outcome Evaluation

1. Family Group Decision Making
   • Maltreatment Rates: No statistically significant differences in maltreatment rates emerged between the experimental and control groups in either Fresno or Riverside County.
   • Permanency: No statistically significant differences emerged between the experimental and control groups in the likelihood of permanency (e.g., reunification) or in the average duration of out-of-home placement.
   • Child and Family Well-Being: Due to small sample sizes and low response rates, California's evaluators aggregated the samples from Fresno and Riverside Counties and used longitudinal analysis to measure changes in child and family well-being over time. Data from surveys administered to children and caregivers within 30 days of enrollment into the demonstration were compared with survey data collected 12 months later. Low response rates precluded tests of statistical significance. Some positive changes were observed in caregivers' reports of children's health status, with more children reported in "good" or "excellent" health 12 months following entry into the demonstration than at initial enrollment. Improvements were also noted in children's emotional well-being as measured by reports of how often children felt "pleased with themselves" or had a "feeling of being successful." Family functioning and parenting, as measured by the Family Unpredictability Scale and other parenting questions designed specifically for the study, showed no improvements over time in any areas.
2. Wraparound Service Model
   The State's outcome evaluation did not find statistically significant evidence of increased child safety, placement stability, or permanence for children receiving Wraparound services. However, there were some statistically significant child welfare outcome findings in specific counties:
   • Compared with the control group, a larger proportion of children in Alameda County receiving Wraparound services were living in family-based environments at the end of the study.
   • Compared with the control group, a smaller proportion of children in Sacramento County receiving Wraparound services exited from the child welfare system due to incarceration.
   In Alameda County, where assessments of child well-being were conducted, youth respondents reported improved health status and both youth and caregivers reported improved youth emotional/behavioral adjustment. Caregiver respondents reported improved satisfaction with services.

¹ California's original five-year demonstration was completed September 30, 2003. Short-term extensions were granted through December 31, 2005. Back

Back to Table of Contents