Foster Care Experience and Future Child Serving System Involvement

Background and Purpose:  Permanency is a key outcome in the federal child welfare measures and reunification is the preferred exit, any analysis of reunification is incomplete without looking at the children who reenter child welfare. This study examines the risk of reentry into the child welfare system of children who exit to reunification within 18 months of entry as a means of developing a predictive model of foster care reentries.  Of particular interest is to understand the differences that exist between children who are reunifying after their initial removal compared to youth who have had prior child welfare experience.

 Method: All children who exited foster care to reunification between federal fiscal year 2011 through federal fiscal year 2013 (n=3,941) are examined.  Instances of reentry into foster care within 18 months were documented for these children (n=796). Using logistic regression, reentry into foster care controlling for relevant covariates suggested in the literature is examined.  Novel variables have been added to account for the number of risk factors present, the number of siblings at home, and whether the return home was court ordered. This study corrects for the presence of sibling groups in the data using statistical software allowing for all observations to be utilized.  The analysis is broken out by groups of children in their initial removal (n=3,088) and those having prior child welfare experience (n=853).

 Results:  This analysis suggest the most powerful predictors of reentry into foster care after reunification are: the presence of a child behavioral indicator; having siblings in care at the same time; and having the child returned home by the court without agency consent.  Comparing initial removals and other removals shows significant differences across these groups.  First children have a much higher rate of reentry when they have prior child welfare experience and females have higher odds of reentry in the prior child welfare experience group (OR=1.6), while gender is not a significant predictor in the initial child welfare group.  Additionally, this analysis is in the process of examining whether there is movement into other child serving systems (Juvenile Justice) and what differences might exist between these groups.

 Conclusion and Implications:  Examining the characteristics of children who have exited to reunification in order to predict reentry has suggested a number of characteristics that need to be examined and considered at the point of reunification.  First, having prior child welfare experience increases the odds of having future child welfare experience suggesting that there are a core group of families that need more supports than they are currently receiving.  Steps taken to achieve reunification for foster children should avoid situations where there is a high likelihood of reentering foster care.  The use of analytic models will help us develop a predictive model in order to prioritize post-reunification services and offer opportunities to engage families in order to effectively manage and ensure successful reunifications.