Spillover Effects of Localized School Choice Programs

Abstract

In this paper I examine the spillover effects of the Los Angeles Unified School District's Zones of Choice program over the past decade on the nearby sector of charter schools. Zones of Choice (ZOC) is a school choice program aimed at increasing college readiness, career readiness, and school quality for areas with high concentrations of lower-income students. It also aims to bring students back to in-district schools, as the proportion of LAUSD students enrolled in the charter sector has been steadily increasing for a decade or more. In this program incoming high school freshmen can choose the public school they attend from the schools within their geographic zone of choice. However, this program does not apply to charter schools, which are always available as an out-of-district option. I estimate the impact of the ZOC program on nearby charter schools with a two-way fixed-effects model. Treatment identification is categorized between charter schools geographically close to a ZOC school and those that are not while allowing for sufficiently large sample sizes for both groups. According to prior research, the ZOC program has been largely successful in increasing the standardized test scores of participating public schools. Based on my analysis, the competitive spillovers have also induced large gains in test scores for nearby charter schools, though have had little impact on enrollment. Finally, I compare and contrast the standard TWFE regression estimates with those from estimators unbiased to heterogeneous treatment effects as detailed in de Chaisemartin and d'Haultfœuille (2021). Due to the construction of the treatment variable, it is possible that the treatment effect is heterogeneous. Comparing the estimates serves as a robustness check for the regression results and the empirical implications of the theoretical framework.