On the distribution of education and democracy

This paper empirically analyzes the influence of education on democracy by controlling for unobservable heterogeneity and by taking into account the persistency of some of the variables. The most novel finding is that an increase in the education attained by the majority of the population is what matters for the implementation and sustainability of democracy, rather than the average years of schooling. We show this result is robust to issues pertaining omitted variables, outliers, sample selection, or a narrow definition of the variables used to measure democracy.