Does Greater School Autonomy Make a Difference? Evidence from a Randomized Natural Experiment in South Korea

We study the effects of school autonomy using a randomized natural experiment in Seoul. Private and public schools subject to the equalization policy in Seoul admit students assigned randomly to them, receive equal government funding, charge identical fees, and use similar curricula, while private schools have greater flexibility in personnel decisions, and their principals and teachers face stronger incentives to perform. We find that private high schools have on average fewer violent incidents per student, a higher four-year college entrance rate, and better test scores. The effects appear to channel through the within-school dispersions of teacher salary and types.