- The best available evidence suggests that a 10% increase in the Black share of a school’s student body would reduce achievement test scores for Black students by between 0.025 and around 0.08 standard deviations, and reduce test scores for Whites by perhaps one-quarter to two-fifths as much.
- Evidence from the Moving to Opportunity Housing Project (MTO) shows that the Black-White test score gaps stopped closing during a period when neighborhood racial segregation steadily declined, and neighborhood economic segregation rose at most slightly before declining.
- Overall, the literature on school segregation and the Black-White test score gap, though clearly not unanimous, provides more substantial support than the complementary literature on the impact of neighborhood-level segregation on comparable outcomes. By several estimates, the impact of segregation is modest.
- The downward trend in residential segregation in the 1990s was not matched by a commensurate downward trend in school segregation.
- Although schools remain more integrated than the communities they serve, school segregation is clearly converging toward the degree of neighborhood segregation. Moreover, the reduction in the rate of school segregation occurred at roughly the same time that the black-white test score gap stalled. The circumstantial case linking school segregation to the test score gap is compelling.