Achieving Diversity in the Parents Involved Era: Evidence for Geographic Integration Plans in Metropolitan School Districts

Julian Vasquez Heilig, University of Texas at Austin
Meredith Richard, University of Texas at Austin
Kori Stroub, University of Texas at Austin
Michael Volonnino, University of Texas at Austin

Abstract

In the wake of the Parents Involved decision, which rendered unconstitutional voluntary school integration plans using individual student race, districts have adopted a number of alternative integration strategies to combat the re-segregation of America’s schools. One promising approach, developed by Berkeley Unified School District, uses neighborhood demographic and socioeconomic characteristics as proxies for student race and ethnicity in assigning students to schools. This study provides the first empirical assessment of such “geographic integration models” by 1) modeling how accurately neighborhood demographic and socioeconomic characteristics predicted student race/ethnicity, and 2) estimating the potential increases in school diversity under such a plan.

First, we present results of an analysis of the prediction of student

race/ethnicity using data on 3.9 million students in the 10 most populous metropolitan school districts, drawn from Census and National Center for Education Statistics Common Core of Data. Second, we present the results of a student assignment simulation estimating the expected change in diversity under a geographic integration plan in Dallas Independent School District.

Our findings suggest that neighborhood characteristics are fairly strong predictors of student race/ethnicity, and that most schools would experience small to moderate increases in diversity under geographic integration plans. A student assignment model using a combination of demographic and socioeconomic predictors would be more effective than models using socioeconomic predictors, but worse than models using aggregate student race alone. Gains in diversity would be largest for elementary schools and small schools across all models, and for schools in more segregated districts with less diverse neighborhoods. We conclude by considering the legal ramifications of such geographic integration plans.