The effects of performance-based school closure and restart on student performance

Abstract

We study the effects of closing and restarting low-performing schools as charter schools in New Orleans and Baton Rouge. Using matched difference-in-differences identification with students in Louisiana, we estimate effects for the students who attended the treated schools at the time of treatment. We find positive and precise effects of closure/restart on elementary/middle school test scores, but no clear effects on high school graduation or college entry in New Orleans. However, in Baton Rouge high schools, the interventions reduced high school graduation by 11–15 percentage points. We also provide evidence about how and why these effects emerge. The variation in test score effects within and across cities is positively related to the increase in school value-added that treated students experienced and negatively related to student age/grade and the extent of student disruption. The effects of school closure and restart therefore depend, predictably, on policy design and implementation. This work builds on prior closure/restart research and helps explain the positive effects of the post-Katrina school reforms in New Orleans.