Association for Public Policy Analysis & Management

Racial and economic segregation limits educational opportunities for poor and minority children in the United States (Sharkey, 2013; Sampson, Sharkey and Raudenbush, 2008). Despite consistent findings that between, rather than within, district differences account for the majority of racial and economic segregation in schools (Logan, Minca, Adar 2012; Massey and Denton 1993), court rulings have severely limited inter-district school mobility options for students (Orfield and Eaton 1996). As a result, assisted housing policies now serve as some of the only opportunities for poor minority families wishing to send their children to higher performing schools outside of their district. Mobility voucher programs provided families with the ability to drastically change their neighborhood contexts, and by consequence their zoned school.

Evidence on the effects of housing mobility programs on academic outcomes is mixed. Chicago’s Gautreaux program, while using weak measures and quasi-experimental techniques, found significant improvements in academic achievement for students who relocated to suburban neighborhoods and schools (Kaufman and Rosenbaum 1992). Decades later, the Moving to Opportunity program, using an experimental design, found that moves to lower-income neighborhoods with only limited improvements in school quality had no significant effect on math and reading ability (Sanbonmatsu et al. 2006). Subgroup analyses, however, suggest this finding is context dependent; academic outcomes improved for MTO youth in Baltimore and Chicago, cities with the highest levels of concentrated disadvantage, but not in other sites (Burdick-Will et al. 2011).

This study uses longitudinal administrative data from the Baltimore Housing Mobility Program, a voluntary housing voucher program created in 2003 as part of a court ordered settlement to the decades long Thompson v. HUD case. Administrative data on all school-aged children in participating households was merged with achievement test score data from the Maryland State Department of Education to examine changes in academic achievement over time.

Students participating in the program experienced substantial improvements in school quality, which persisted over time (DeLuca and Rosenblatt 2017). After receiving their voucher and moving to new neighborhoods, nearly three-quarters of the BHMP students were enrolled in suburban county school districts. After moving, the students attended schools with less poverty, less racial segregation, and higher academic achievement.

Fixed effects models show a small yet significant dip in academic achievement during the school year immediately following the move. Students return to baseline in year two and steadily improve their scores each year thereafter, resulting in statistically significant positive improvements five years after voucher receipt. We interpret this pattern as reflecting the challenges student face adjusting to radically different education environments, followed by their gradual accumulation of the benefits of attending highly-resourced desegregated schools. Supplemental analyses suggest that school change, rather than residential mobility, is the primary reason for the dip and that moving to a higher quality school is a protective factor, reducing the short-term negative effects of mobility.