The theory developed in Chapter 5 makes predictions about how the quality and extent of private and public schooling are determined at the aggregate level, and about how schooling and fertility choices vary across households within a given political entity. In this chapter,we compare these predictions to data.We start by focusing on state-level variation in the extent and quality of public education in the United States. This setting is well suited to examining the predictions of our theory for democratic countries, since all US states operate within the same overall political framework, while exhibiting considerable variation in schooling policies as well as the distribution of income. Moreover, we are able to link state-level evidence to household data from the US Census to assess the micro implications of our theory.We then extend the analysis to cross-country data, which allows us to probe the theory's predictions for non-democratic countries. Here we use data from the OECD and the World Bank on public and private education spending, as well as micro data from the OECD Program for International Student Assessment (PISA).

Inequality, fertility, and schooling across US states

Our model predicts that in a democracy, the choice of public versus private schooling and the level of funding of public schooling are driven by income inequality (see Proposition 5.3). In particular, a state with higher income inequality should exhibit a higher share of private schooling, lower overall spending on public schooling, but higher public education spending per student.