Measuring and Explaining Management in Schools: New Approaches Using Public Data

Abstract: Learning in schools across the world has stagnated: children are in school, but often not learning. A shift in thinking about how to improve learning is emphasizing ''system-level'' reform. Management practices can be an important link in education systems, with evidence from the small set of countries sampled in the World Management Survey pointing to a strong positive relationship between measures of management and student outcomes. Yet the mechanisms behind this finding remain a black box. In this paper, we develop a simple theoretical framework that shows how management practices can affect school performance through channels traditionally examined in the personnel economics literature---the selection and incentive margins of agents. We then take the predictions of this framework to the data. Specifically, we build a new cross-country measure of management structures using the Programme for International Student Assessment (PISA), the largest education dataset currently available. We find that the positive relationship between management and student outcomes holds in 5 of the 6 PISA regions, and present evidence that is consistent with both the selection and incentive mechanisms proposed in the theory.