Prior research on public goods provision and education suggests that representative and accountable democratic governments are more likely than their non-democratic counterparts to increase school enrollment and public spending on primary and secondary education. Yet, the existing research finds no relationship between democracy and student achievement. This paper discusses the reasons for this surprising null finding and argues that the relationship between political regimes and students' learning outcomes varies by school subject. To ensure regime stability and support, autocrats might purposefully restrict the development of young minds in terms of critical thinking. In democratic settings, by contrast, critical reasoning and other skills that students need as democratic citizens and potential political actors can be encouraged by the school curriculum and teaching methods. Distinct political goals reflected in schooling, thus, can lead to cross-national achievement differences in international tests in those disciplines that promote critical thinking.

To test the argument that autocracies avoid the cultivation of critical thinking, I employ data from the Harmonized Learning Outcomes database on student performance in mathematics, science, and reading in 164 countries from 2000 to 2017. The results of the time-series cross-section regression analysis with two-way fixed effects demonstrate that democracy indexes are indeed positively and strongly associated with students' scores in reading, which is a subject that requires critical thinking, but not in mathematics and science.