Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purpose: An untested, yet often cited argument in favor of transracial adoption (TRA) from substitute care has been the assertion that white foster families are more likely to adopt than black foster families and, as a result, transracial adoptive placements could serve as an effective means to accelerate the rate of exit from care among black children. Perhaps as a result of the divisive nature of the debate around TRA or the methodological difficulties created by non-random assignment to foster homes, a rigorous empirical examination of the relationship between foster parent race and adoption has yet to occur. Thus, it is with the goal of addressing this shortcoming that the current study was undertaken. Methods: Administrative data from the Illinois Department of Children and Family Services regarding all placements between 1995 and 2002 (N=42,567) were matched with 2000 Census records, Illinois state wage data, and Medicaid paid claims records. Acknowledging both the importance of timing vis-à-vis adoption outcomes, and the fact that the relationship between placement selection and subsequent adoption is likely confounded by endogeneity and sample selection, the likelihood of adoption was analyzed using (a) multistate hazard models and (b) recursive, simultaneous-equation models for discrete outcome data (Greene, 2002). Control variables included data pertaining to the characteristics of children (e.g. demographics; health, disability, and mental health diagnoses; substitute care history), foster parents (e.g. demographics, wage income, household size, license type), and child welfare agencies (e.g. organizational structure and resource environment, characteristics of case workers). Results: The results of the multivariate analyses reveal a complex relationship between child race, foster parent race, and adoption. Children who were found to be more readily adopted (e.g., younger children and children without siblings) were also found to be more likely to be placed with white foster parents than with non-white parents. However, after controlling for this endogenously-determined placement, placements in white foster homes were found to be more likely to result in an adoption than placements in non-white foster homes. Interestingly, child race was not found to be a significant predictor of adoption after the inclusion of foster parent race. Conversely, although child race was found to be significantly associated with reunification, foster parent race was not. Implications: The primary implication of this study is that the locus from which race exerts an influence on the timing of permanence may differ across permanency outcomes. Although child race appears to predict the likelihood of reunification, the timeliness of adoption appears to be a function of mechanisms associated with foster parent race. It is important to note, however, that these findings do not necessarily support the contention that white foster families are more inclined to adopt that black foster families. Differential treatment of black foster homes by the juvenile courts or child welfare agencies could also account for the observed difference across foster parent race.