Association for Public Policy Analysis & Management

When choosing where to live, many U.S. parents take local school options into account. Historically, school attendance was closely coupled with residential address, but this link has weakened in recent decades. Some districts now have open enrollment policies, allowing students to attend schools outside their neighborhoods; magnet, charter, and private schools have become popular alternatives to neighborhood schools; and some districts use student assignment policies that mix students from many neighborhoods to create socioeconomic or racial integration.

Little is known about how the proliferation of schooling options has affected residential segregation by race. We hypothesize that when a wide array of school options is available and a child does not have to attend his or her neighborhood school, residential segregation may decline. As enrollment choices weaken the direct link between housing and school assignment, families may give greater weight to other preferences so long as they can still access desirable schools. If true, school choice policies may reduce neighborhood segregation while maintaining or even increasing racial or income-based segregation between schools. Our study evaluates this hypothesis using two decades of national, geographically-referenced Census and school-district data, and in so doing explores a selection dynamic with critical implications for choice-oriented educational and housing policies.

Drawing on decennial Census data, we measure residential segregation between tracts by race/ethnicity (considering black-white, Hispanic-white, and multiracial segregation among whites, blacks, Hispanics, Asians, and other race persons) in 1990, 2000, and 2010. Our modeling strategy accounts for the trend in racial segregation from 1990 to 2000 and identifies how the change in magnet, charter, and private school availability from 2000 to 2010 shapes trends in racial segregation from 2000 to 2010. We measure magnet, charter, and private school availability both as the enrollment share of students attending such schools and in terms of the average geographic proximity of neighborhoods to a magnet or charter option. We limit our analyses to the 100 largest metropolitan areas.

Our analyses proceed at multiple levels of geography. First, we document how school choice options within metropolitan areas contribute to residential segregation in the metro. Second, we examine how school choice options within school districts contribute to residential segregation in the district. Finally, we examine how school choice options in the metropolitan area contributes to segregation between school districts, rather than tracts, in the metropolitan area. We control for MSA and district demographic compositions as well as desegregation court order and open enrollment status, and a next step in the analysis is to examine heterogeneous effects across districts and metropolitan areas.

Preliminary results suggest that greater magnet and charter school availability modestly reduces residential segregation between neighborhoods. Findings from this study have implications for both educational and housing policies, emphasizing the importance of considering how both housing and educational policy arenas operate together.