The Impact of Racially Matched Foster Care Placements on Placement Stability: A Population-Based Study

Background: Placement instability for children in foster care has an adverse effect on child safety, permanency, and well-being. Prior research has found evidence of racial disparities in placement instability, with Black children at higher risk for placement changes while in foster care (Font et al., 2018). Some studies have examined racial matching between caseworker-child as a way to improve outcomes for children of color in care (Barn et al., 2016; Chenot et al., 2019), but fewer have examined racial matching in the initial foster placement and subsequent child outcomes. This study examined the impact of racial matching in initial non-relative foster parents on subsequent placement stability.

Methods: This study used data from the Adoption and Foster Care Analysis System (AFCARS), an annually collected administrative database containing the children in care for each fiscal year (FY). The analytic sample included N=75,024 children aged 0-17 years who entered foster care in FY 2015 and were placed with a non-relative foster family. This sample was followed through FY 2019 to compare those with any subsequent placement changes (placement instability) to those without (placement stability). The main independent variable was racial matching, operationalized as a dichotomous variable that compared children for whom at least one of the foster parents identified with the same racial/ethnic group as the child to those in a transracial placement. The authors conducted an extended binary logistic regression model to account for potential differences in characteristics of children assigned to a racially matched placement and those placed in transracial placements. All analyses were conducted using Stata version 16.

Results: Overall, approximately three-quarters of children were in a racially matched placement when they entered foster care, and approximately two-thirds experienced placement stability. Non-Hispanic White children had the highest rates of racially matched placements. After adjusting for other factors, racial matching decreased the likelihood of placement stability (probit=-0.19, p<.001). Children who were Black, Hispanic, or Native American had lower odds of placement stability than White children. However, the treatment effect of racial matching varied significantly by child race/ethnicity; racial matching did not impact placement stability among Hispanic or Native American children, while it negatively impacted placement stability for White children (ATE=-0.08, p<.001).

Conclusion: Consistent with prior literature, this study found that children of color had lower odds of placement stability than White children. Although racial matching has been conceptualized as a way to address disparities, results from this study suggest that more research is needed. Given the heterogeneity within racial/ethnic groups, it is also possible that racial matching would yield better outcomes if other factors such as language, culture, and beliefs were included in the initial match.