Fostering Innovation in Permanency Planning: The Subsidized Guardianship Waiver Demonstrations

Background: On October 7, 2008, President Bush signed into law the bipartisan Fostering Connections to Success and Increasing Adoptions (FCSIA) Act. A key provision, the kinship Guardianship Assistance Program (GAP), was based on findings from a series of randomized controlled trials (RCTs) that were conducted by states under federal IV-E waiver authority to test the efficacy and effectiveness of subsidized guardianship in increasing rates of permanence for children in kinship and traditional foster care. The purpose of this proposed presentation is to report the results of these RCTs from the initial efficacy trial in Illinois in the late 1990s and the effectiveness replications in Tennessee and Milwaukee, Wisconsin. The presentation will describe how these findings influenced the drafting of federal legislation and helped to cement support for passage of the final bill. It will also discuss how these findings are currently being used in states like New York and Virginia to persuade local stakeholders to sign on to the new federal program.

Methods: The proposed presentation uses administrative and survey data collected from three RCTs of subsidized guardianship in the states of Illinois, Tennessee, and Wisconsin. Findings from these evaluations are supplemented with published findings from the nine other guardianship waiver demonstrations approved by the U.S. Administration of Children and Families prior to 2006. The evaluations of the demonstrations in Illinois, Tennessee, and Wisconsin were designed by the same evaluation team and involved the random assignment of sibling groups to intervention and comparison groups. Children assigned to the comparison groups remained eligible for all existing permanency options in effect prior to the waiver‘s implementation, including reunification and adoption assistance, while children assigned to the intervention groups were also eligible for the supplementary permanency option of subsidized guardianship. Sample sizes are 2,245 children in Illinois, 1,122 in Tennessee, and 576 in Milwaukee, Wisconsin. Analyses use logistic regression to estimate the impact of the intervention on permanency outcomes and two-stage least squares to adjust for incomplete treatment compliance in estimating the effects of the intervention on foster care expenditures and length of stay. Replicate weights were used to account for survey non-response and clustering within sibling group.

Results: Intention-to-treat (ITT) results suggest that assignment to the intervention group boosted overall permanency rates by 6.6 percentage points in Illinois, 11.2 percentage points in Tennessee, and 18.8% in Milwaukee, Wisconsin. Treatment-on –treated (TOT) analyses imply that the offer of subsidized guardianship reduced the average length of stay in foster care by between 22% and 30%.

Conclusions: Replications of the Illinois findings in Tennessee and Wisconsin were compelling enough to persuade Congress to include kinship guardianship assistance in its FCSIA Act. The experience reinforces the importance of well-designed RCTs in assuring stakeholders of the benefits and cost-effectiveness of policy innovations. At the same time, experimental findings must be translated to local contexts in order to convince local stakeholders that adoption of the innovation will also be effective in their particular jurisdictions.