Over the last two decades, hundreds of school districts in the U.S. have been released from school desegregation orders, and, in response, the districts changed the policies assigning students to schools (i.e., student assignment policies). An emerging literature has begun to examine the effects of ending school desegregation requirements on school segregation, but less attention has been paid to the details of student assignment policy changes. Changes in school assignment policies that aimed at reducing racial school segregation are expected to increase the level of racial segregation across schools, but an important policy question is how districts can minimize the negative effect on school segregation. This study extends previous studies by paying close attention to the details in student assignment policies and examines differential effects of the types of student assignment policies that have followed in districts after release from school desegregation orders. In addition, I examine the effects of these policy changes on racial residential segregation, since long-term effects on school segregation and other distributional effects of educational resources depend on residential sorting.

The sample consists of 98 school districts who adopted a desegregation plan prior to 1987. About 50 percent of school districts in this sample were released from court desegregation orders. Based on extensive archival and case studies, I identified the details of policy changes in the sample districts. I also classified the changes in school desegregation plans by focusing on the goals and priorities of student assignment policies such as: (1) increasing neighborhood assignments, (2) avoiding the isolation of socioeconomically disadvantaged students, and (3) expanding parental ability to choose schools.

Using the Common Core of Data of the National Center for Education Statistics (1987-2012) and decennial Censuses, I computed the dissimilarity index to measure school segregation across schools and residential segregation across Census tracts within a district, respectively. Based on information I have collected on policy changes in school districts, I examine the impact of policy changes on school segregation using an interrupted time series design and examine the impact on residential segregation using a difference-in-differences matching design. These designs estimate the impact of the policy changes by comparing the changes in segregation level in the districts that replaced their previous desegregation plans at one period of time to districts that have not yet adopted changes in student assignment policies.

Preliminary results suggest that overall changes in student assignment policies increase school segregation with no effects on residential segregation. Estimated effects by the policy types suggest that the increase in school segregation is driven by the districts that adopted neighborhood assignments. Districts that adopted socioeconomic balancing plans and expanded school choice plans were successful in maintaining the level of school integration. Findings of this study suggest that school districts’ assignment policy choices are crucial in maintaining school integration. In the absence of any efforts to integrate schools, schools will become more racially segregated when school desegregation policies are removed.