The HomeRebuilders project was an ambitious effort to test a major reform of the foster care system in New York City. In 1993, the New York State Department of Social Services (DSS) and the New York City Child Welfare Administration began testing a new approach to the financing of services to foster children and their birth families based on concepts from managed care. Based on the premise that paying for each day a child is in foster care (per diem payment) is a disincentive to return children home, the demonstration tested an alternative method of agency reimbursement. Instead of paying for each day in care, agencies were paid a flat amount of money or capitation payments for an identified group of children in foster care. The funds were to be used to serve the children for 3 years. A projection of care day utilization was calculated based on each agency's historic care day usage. A strategy was agreed upon to pay agencies a percentage of the 3-year capitation each year. Funding was front loaded in the first year to encourage early discharge.
It was hypothesized that the change in the payment system would achieve earlier permanency for children through intensified discharge planning and aftercare services. Funds could be used for foster care costs and any services the agencies believed would achieve earlier permanency. Although the same fiscal methodology was designed for all agencies, a particular program model was not mandated. Instead, the demonstration emphasized common themes, and it was incumbent upon the agencies to establish programs to achieve permanency in a more timely manner, and each agency developed and implemented unique HomeRebuilders programs. If the programs were successful, children in the HomeRebuilders program would spend less time in foster care than a comparable group of children not in the program. Besides reducing care days, it was also expected that participants would experience a reduced rate of reentry to foster care.
Evaluating the Demonstration
The initiative included six agencies, using an experimental design involving comparison groups in five of the agencies, with random assignment in three. The new financing mechanism was to be used in the experimental groups while services to the control group were financed under the old system.
Six agencies participated in the HomeRebuilders program: Harlem Dowling, Little Flower, Miracle Makers, New York Foundling, St. Christopher-Ottilie, and St. Christopher-Jennie Clarkson. Three agencies, Harlem Dowling, Little Flower, and Miracle Makers, used random assignment procedures. New York Foundling and St. Christopher-Ottilie used different offices for their experimental and comparison groups. At St. Christopher- Jennie Clarkson, all of the foster boarding home children were selected to participate in the HomeRebuilders project.
Based on certain criteria, the state selected approximately 3000 children for participation in the experimental and comparison groups. New York State originally planned to collect follow-up information on families, but resource limitations prohibited this from happening. The Evaluation of Family Preservation and Reunification Services supplemented the evaluation efforts of New York State. The evaluation began approximately two and one-half years after the HomeRebuilders program was initiated, close to the time that the demonstration was facing a precipitous ending, 6 months early.
The purpose of the evaluation was to document the implementation of the program, the services provided, and the outcomes of those services. The sources for this information were New York State administrative data, follow-up interviews with a sample of workers and client families, and interviews with administrative, supervisory, and front line staff about the implementation of the demonstration. Caretakers and caseworkers were interviewed between July and December 1996, some three to three and one-half years after the implementation of the experiment. A total of 433 caretaker interviews and interviews with caseworkers of 407 of these cases were completed. Administrative data on some 3000 children participating in the experiment were analyzed. From the administrative and interview data, the characteristics of children and families involved in the demonstration, the services provided, and the outcomes for children and their families were examined. Outcomes include: percentage of case closings, days in foster care subsequent to the beginning of the experiment on July 1, 1993, reentry rates, and child and family functioning.
Capitation payments were intended as a catalyst to change case planning decisions and service provision for children in foster care. The purpose was to accelerate decision-making to achieve earlier permanency for children. Changes in the focus and approach to case planning decisions and service provision occurred in varying degrees across the agencies. Little Flower and St. Christopher-Jennie Clarkson exhibited the most comprehensive changes in both these areas. They developed strategies to prioritize cases by identifying cases that could be returned home quickly and making efforts up front to achieve additional discharges during the first year. Aftercare services were then used to maintain the savings accrued from the "early" discharge of children. New York Foundling did not make extensive service provision modifications, but they did change decision-making processes. They reported a new approach to handling adoption cases and extensive case review procedures that placed a new focus on permanency. By contrast, Harlem Dowling, Miracle Makers, and St. Christopher-Ottilie did not systematically implement new decision-making processes in the experimental group. Instead, they said they focused on enhancing strategies and procedures that were already in place.
Other programmatic changes made by agencies included reducing caseload size to facilitate increased contacts. Four of the six agencies recruited casework staff to work specifically for the experimental group. All agencies hired a number of specialists. Little Flower and St Christopher-Jennie Clarkson made the greatest investments in the number of new staff. All agencies, except New York Foundling, reduced the caseload size for the experimental group.
Characteristics of Children in the Demonstration
Children in the HomeRebuilders experiment were almost entirely African American or Hispanic. Over one-half were less than 6 years of age. Most were in their first placements in nonrelative foster homes and had been in placement a fairly long period of time, on average a little less than 3 years. Some characteristics of children varied among agencies and between experimental and comparison groups within agencies. Of particular importance to the findings are the shorter length of time in care at the beginning of the experiment in the Little Flower experimental group and the larger proportion of African American and smaller proportion of Hispanic children in the Miracle Makers experimental group.
Discharge and Days in Care
The rapid achievement of permanency and reduction in days of care were central objectives of the HomeRebuilders program. Two measures were examined: the percentage of children whose cases were closed and the number of days in care subsequent to July 1, 1993. The analysis focused on differences between experimental and comparison groups within agencies. Aggregating results across agencies is not warranted because of considerable differences among agencies. The data indicate that the HomeRebuilders program had the effect of speeding up discharge and reducing days in care in one of the three agencies using random assignment (Little Flower). In the other two random assignment agencies, Harlem Dowling and Miracle Makers, discharge proportions were not significantly higher and days in care were not significantly lower for the experimental group. In one of the non-random-assignment agencies (New York Foundling), discharge rates were higher and days in care lower for the HomeRebuilders group compared with the comparison group. The results for Little Flower and New York Foundling were confirmed in a regression analysis controlling for certain characteristics of the children. St. Christopher-Jennie Clarkson had high proportions of cases closed and low average days in care, but definitive conclusions cannot be drawn about those results because there was no comparison group in that agency.
About two-thirds of the children began the experiment with a permanency goal of return home; nearly all of the remainder had a goal of adoption. Many of the original goals of return home were subsequently changed to adoption, so that by the end of the experiment, in December 1995, the proportions had substantially reversed, with 55% of the children having been adopted or having a goal of adoption.
Rates of reentry to foster care were very low in both the experimental and comparison groups.
Child and Family Functioning
Parent and family functioning was explored for respondents who were birth parents. Caretakers were asked a series of questions to measure depression, caretaking behaviors, punishment, positive and negative life events, difficulty paying for things, and overall assessment of improvement. On scales derived to measure these variables, few significant differences were found between the experimental and comparison groups within agencies. Similar questions were asked of the caseworkers and again there was little difference between experimental and comparison groups.
Caretakers were also asked about children's negative behaviors, positive behaviors, aggression, and school problems. Responses of all caretakers and specifically of birth parents were analyzed. No significant differences were found for all respondents, and for birth parents only child aggression scores were lower in the experimental group than in the comparison group for only one agency. Caseworkers were asked a number of questions about the behavior of each child in a family. A scale based on these questions revealed no significant differences between the experimental and comparison groups.
Many factors may have contributed to the finding of no difference for child and family functioning. Two are particularly relevant to this evaluation: 1) interviews were conducted retrospectively, 6 to 12 months after the end of the program and caseworkers and caretakers experienced recall problems, and 2) a lack of statistical power to detect effects because of low sample sizes within agency experimental and comparison groups.
In those agencies with lower care days in the experimental group, one might interpret the lack of effects on child and family functioning as indicating that care days can be reduced without adversely affecting functioning. But lack of a statistically significant difference between the experimental and comparison groups in functioning does not prove that there was no detrimental effect. There are other explanations for lack of statistical significance as indicated above. We cannot conclude there was harm, nor can we conclude there was not.
Both caretakers and caseworkers were asked about service receipt. Caretakers were asked a number of questions about social program involvement, services received and caseworker activities. Overall, caretakers in the experimental and comparison groups reported little difference in the receipt of services, although the experimental group may have received slightly more services and more intensive services. The experimental groups reported more day care in Harlem Dowling and Little Flower, more medical and dental care in Harlem Dowling and Miracle Makers, and more educational services in St. Christopher-Ottilie.
Caseworkers indicated a slightly higher receipt of services for the experimental groups. Scales were constructed to identify the number of services provided and number of different kinds of caseworker activities. The results indicate that in the perception of caseworkers in Little Flower, experimental group families received more service. There is little evidence for higher levels of service in the experimental groups of the other agencies, although results were in the same direction, but were not statistically significant for Miracle Makers and Harlem Dowling.
It is thought that one of the most important aspects of reunification work is aftercare services, that is, services to prevent reentry to foster care following the return of children (and possibly to adoptive families after the adoption). Neither the caseworker or caretaker interviews uncovered differences between experimental and comparison groups in level of aftercare services. The agencies did stress that the availability of aftercare services enabled HomeRebuilders workers to take greater risks in returning children home.
A primary goal of initiatives like HomeRebuilders is to reduce days in foster care. The principal mechanism by which this is to be achieved is a shift in the fiscal incentives provided to agencies. The findings of this evaluation suggest that the goal is likely to be achieved only under certain circumstances: in agencies with particular characteristics or with particular approaches to the work. Among necessary preconditions, the data suggest the importance of clear decision-making protocols, triaging strategies, and data systems that can be used for case and program management. Hence, changes in fiscal incentives do not appear to be sufficient causes of major reform in child welfare. Although the flexibility in funding provided the opportunity for agencies to develop creative approaches to service delivery, one might also ask whether such funding changes are necessary for reform, or whether other approaches might achieve the desired benefits. There are current in the field proposals for major change that do not place fiscal reform at the center of concern, most notably the shifting of responsibility for response to child maltreatment to communities. At the moment, there is no dominant approach to change in the child welfare system that has been empirically demonstrated to be superior. More experimentation is needed, with both managed care and other approaches.