Centralization and Long-Term System Performance: Evidence From State Takeovers of School Districts

Does centralization affect the performance of public institutions in the long run? Can states more successfully improve struggling bureaucracies than local governments? This paper explores these questions in the context of educational governance with a focus on state takeovers of school districts in the U.S.—a shift away from the traditional school board governance arrangement toward more centralized decision-making at the state level. In recent decades, takeovers have become a more common policy response to perceived low performance of public school systems. This paper extends an earlier study on the topic to examine the longer-run effects of this form of political centralization on system performance. Using a nationwide sample and up-to-date event study methods, the paper finds no evidence that takeovers of districts between 2010 and 2018 generated improvements in student reading and math performance, up to nearly a decade after takeover occurred. Takeovers lasting a greater number of years are not associated with differential impacts. Findings are not driven by compositional changes in student populations or bias due to variation in treatment timing. This form of political centralization from local to state levels therefore does not appear to be a consistent tool for improving the performance of public institutions.