Who lost the most? COVID-19 school closures, PISA achievement, and socioeconomic gaps

This paper examines how COVID-19 school closures were associated with student achievement and social inequalities using data from the Programme for International Student Assessment (PISA) for 2018 and 2022. The analysis examines how different remote learning methods were associated with the relationship among school closures, educational inequality, and learning outcomes in mathematics, science, and reading. To estimate both direct and indirect associations, the study employs a Generalized Structural Equation Model (GSEM) and a Triple Difference-in-Differences approach. This method enables the simultaneous estimation of correlates of remote learning modality usage and their associations with proficiency outcomes, while accounting for students' socioeconomic backgrounds. The results show clear differences between advanced and emerging/developing economies. In advanced economies, school closures were associated with moderate average learning losses, but a significant increase in social inequalities. In emerging/developing economies, school closures were primarily associated with larger average learning losses rather than with increased socioeconomic gaps. This may be due to the limited spread of digital learning technologies that could help create advantages. In summary, the results suggest that school closures were associated less with a uniform disruption to learning and more with a redistribution of educational opportunities along socioeconomic lines, with the type of remote learning available to students associated with much of who lost the most.