Policy spillover effects on student achievement: evidence from PISA

National education reforms do not occur in isolation. Countries look towards each other to identify ways that improve the quality of their education systems. When evaluating the effect of an education policy, it is worth considering both local effects of the policy and its spillover effects on other countries. Ignoring spillover effects between countries can lead to biased estimates of policy effects and suboptimal decision making. This paper examines spillover effects of one widespread education policy, school autonomy, on student achievement using three waves of data from the Programme for International Student Assessment (PISA). The spatial autoregressive model is applied to capture both spillover and local effects of school autonomy. Overall, school autonomy raises student achievement in Reading, Mathematics, and Science. We confirm the existence of positive and statistically significant average spillover effects; thus, estimates based on linear regression underestimate the impact of school autonomy. Our findings indicate that there is spatial dependence in student achievement across countries linked to the geographic proximity between countries. Possible extensions of this work are discussed.