Racial Segregation and the Black-White Test Score Gap

The impact of school segregation on student outcomes is a central concern of education policy. We compare within-city black-white test score gaps across metropolitan areas that differ in the extent of school segregation. More segregated cities have robustly larger black-white gaps in average SAT scores and in other measures of educational achievement. Using an estimator derived from “control function” approaches to instrumental variables models, we decompose the effects of school segregation into components attributable to differences in residential sorting patterns, to courtordered school desegregation, and a residual. The residential component has a large effect on test score gaps, but the remaining components do not. These results do not appear attributable either to endogeneity of residential segregation or to differences in school resources. Our results therefore suggest that the composition of neighborhoods matters, but not that of schools. We uncover evidence, however, that non-residential school segregation is negatively associated with black-white differences in assignment to advanced-track courses, suggesting that within-school segregation may offset much of the variation in non-residential, across-school segregation.