Foster Care and High School Mobility: General Estimates of Mobility Rates and the Effects of the Fostering Connections and Increasing Adoptions Act

Background/Purpose:Elevated school mobility for foster children is something that has been previously noted in non-peer-reviewed literature. At some level, this trend is precisely what would be expected. In the absence of a policy seeking to actively prevent school mobility for foster children, removing a child from one home and placing him in another would necessarily place him at increased risk of a school change. While such school changes would be expected to contribute to decreased educational achievement for any child, the combination of such changes in conjunction with the potential social and emotional barriers faced by a child that has been placed in out-of-home care appears to exacerbate the effects on academic performance for foster children compared with school mobility in the general population. Understanding the phenomenon of foster child school mobility and policies that can be adopted to combat this mobility is thus of importance to the fields of education and social work.

In spite of the importance of foster child school mobility, the peer-reviewed literature is nearly silent on this topic. In this paper, I seek to provide a comprehensive analysis of differences in school mobility as a function of foster care status. Additionally, I examine the impact of the Fostering Connections and Increasing Adoptions Act (FCIA) (PL 110-351) on foster child school mobility – an act which was partially designed to provide more expansive school mobility protections for foster children.

Methods:I use probabilistically linked data from the educational and child welfare agencies in a state in the western US. I specifically focus the current analysis on high-school children (grades 9 through 12). I utilize a three-level multilevel Poisson regression model (academic enrollment years, nested in children, nested in districts) to calculate the expected count of annual school transitions for all children enrolled in high-school from 2004 through 2011 (1,103,614 academic enrollment years nested in 303,889 children). Within this model, I utilize a difference-in-differences approach to compare the change in school mobility for foster children with the school mobility for non-foster children and how the expected count has changed since the passage of FCIA.

Results:The results of the model indicate that, on average, academic enrollment years which coincide with a foster placement tend to have transitions at 5.58 times the rate of those years which do not coincide with a foster placement (p < 0.001). The difference-in-differences estimator indicates that academic enrollment years taking place after the passage of FCIA have transitions at rates of 0.79 times the rate of those years taking place prior to FCIA (p < 0.001). In other words, FCIA appears to have reduced annual mobility rates for foster children by 21 percent.

Conclusions and Implications: Within the assumptions of the difference-in-differences approach, my results appear to indicate that FCIA has been successful at reducing the rate of school mobility for foster children. Additional research is needed to examine local variation in the application of FCIA and how this variation may increase or decrease the effects observed here.