The Central Role of Noise in Evaluating Interventions That Use Test Scores to Rank Schools

Abstract

Many programs reward or penalize schools based on students' average performance. Mean reversion is a potentially serious hindrance to the evaluation of such interventions. Chile's 900 Schools Program (P-900) allocated resources based on cutoffs in schools' mean test scores. This paper shows that transitory noise in average scores and mean reversion lead conventional estimation approaches to overstate the impacts of such programs. It further shows how a regressiondiscontinuity design can be used to control for reversion biases. It concludes that P-900 had significant effects on test score gains, albeit much smaller than is widely believed