Research That Matters (January 17 - 20, 2008)


Hampton Ballroom (Omni Shoreham)

Permanency Outcomes of Children in Kinship and Non-Kinship Foster Care: Minimizing the Effects of Selection Bias with Propensity Score Matching

Eun Koh, PhD, University of Illinois at Urbana-Champaign.

Purpose: Kinship foster care has become a major part of the formal child welfare system since the 1990s. With increasing acceptance and use of kinship placements, many studies have been conducted to examine the comparative outcomes for children in kinship and non-kinship foster care, especially in the area of stability and permanence. However, previous research has not successfully controlled the problem of selection bias, heavily relying on statistical regression models. The study compares the permanency outcomes of kin children to matched samples of non-kin children, addressing the issue of selection bias by using propensity score matching (PSM) to balance mean differences in the characteristics of children in kin and non-kin placements.

Method: The data come from the March 2000 to September 2005 6-month files submitted by the states of Arizona, Illinois, and Ohio to the federal Adoption and Foster Care Analysis Reporting Systems (AFCARS). A sample of 20,545 kin and 24,055 non-kin children who were initially placed in kinship or non-kinship foster homes was created, and children in kinship care were matched to the non-kinship sample using PSM. The permanency outcomes of children in kin and non-kin foster care are compared, using both bivariate and survival analysis.

Results: The findings based on the unmatched samples are consistent with prior studies. Children in kinship placements are more likely to be discharged to legal guardianship while they have lower rates of reunification and adoption. Children in kinship care also have greater placement stability. However, mixed findings are reported when the analysis is based on the matched samples. While non-kinship placements are still advantageous in achieving reunification in the samples from Arizona and Illinois, no significant group differences are observed for the samples from Ohio. Children in kinship and non-kinship care have similar likelihood of adoption in the samples from Arizona and Illinois, but the rates of adoption are still higher for non-kin children in the samples from Ohio. The advantage of kinship foster care holds in the matched samples of all three states in the outcome measures for legal guardianship and placement stability.

Implications: The results of the study suggest that permanency differences may be less attributable to the type of placement itself than what has been thought before, especially in the outcome areas of reunification and adoption. Rather, it is more likely that the different characteristics of children and caregivers in kin and non-kin placements as well as states' specific child welfare policies and practices account for the observed differences in permanency outcomes for children in both types of family foster care. However, the findings that kinship care is advantageous in achieving legal guardianship and placement stability in all three states imply that some factors inherent to the type of placement directly influence certain outcome measures. Child welfare professionals should continue to explore the conditions under which placements with kin advance the permanence of children in foster care as well as to identify risk and protective factors that are independent from the differences among various placement types.