The size of the middle class and educational outcomes: Theory and evidence from the Indian subcontinent

This paper proposes a stylised model to derive the effect of a sizeable middle class on average educational outcomes. Under reasonable assumptions, the model predicts that the spending share on education increases if the middle class becomes larger such that the size of the middle class has a positive impact on education. We test the relationship empirically by using village/neighbourhood level data from Indian household surveys. To tackle the issue of potential endogeneity of the middle class share of the population, we propose a novel instrument that relies on the fraction of the population belonging to the third (middle) caste ("sudra"). Using this IV strategy, our empirical results support a positive effect of a larger middle class on educational outcomes. Furthermore, we show that the share of the middle class is a more important determinant of female education than male education and that the effect of the middle class on education is more pronounced in rural areas.