Does public sector control reduce variance in school quality?

Does the government control of school systems facilitate equality in school quality? Whether centralized or localized control produces more equality depends not only on what 'could' happen in principle, but also on what does happen in practice. We use the Programme for International Student Assessment (PISA) database to examine the association between school sector and the variance in school fixed effects. We find, on average, the same inequality in adjusted learning achievement across the private and public schools. However, in some countries, such as Denmark, there is more equality across the public sector schools, while in others, such as Mexico, there is more equality across the private schools. Among the 18 non-OECD countries, the standard deviation across schools in adjusted quality is, on average, 36% higher in government schools. Our findings suggest that top-down educational systems in weak states can be lose-lose relative to localized systems relying on bottom-up control, producing both worse average performance and higher inequality.