Whether school choice mechanisms can reduce racial and ethnic segregation is among the most contested questions in education policy. Centralized ranked-choice systems — which use algorithms to randomly assign students to oversubscribed schools based on ranked preferences — have been adopted by a growing number of large urban districts on the premise that standardizing enrollment can equalize access and disrupt the segregative sorting that characterizes decentralized choice systems. Prior evidence on the segregation effects of these systems, however, has been largely confined to single cities and descriptive methods.
Francisco Lagos, Jason Saltmarsh, and Jing Liu provide the first causal, nationwide evidence on this question by constructing an original panel dataset covering the 300 largest U.S. school districts between 2000 and 2022. The dataset documents whether and when each district implemented a centralized-admission lottery, enabling a difference-in-differences design that exploits the staggered rollout of these systems across districts.
Lottery adoption is associated with a 2.0 percentage point increase in the share of White students enrolled in district schools, and with modest improvements in cross-racial exposure: Black-White exposure rose by 1.6 percentage points and students-of-color-to-White exposure by 1.8 percentage points. At the same time, White students became less exposed to all non-White groups, and the dissimilarity index — which measures evenness of student distribution across schools — showed no significant improvement. These patterns are robust across specifications and estimators. The authors interpret these results as evidence that centralized lotteries produce modest compositional shifts but fall short of reducing overall segregation, complicating claims that administrative streamlining alone can expand equitable access. They call for complementary interventions — including weighted lotteries, transportation support, and targeted rollout in high-segregation districts — to address the structural drivers of school segregation.
- Working Paper: https://edworkingpapers.com/ai...