Do Better Roads Increase School Enrollment? Evidence from a Unique Road Policy in India

Abstract

In 2001, a national infrastructure development program was initiated in India to construct new all-season roads (roads that can be used in all-weather especially monsoons) in villages that previously had only had dry-season roads (roads that are difficult to use in monsoons). The eligibility rule that was used for undertaking construction of new all-season roads in these villages was that a minimum population of 500 had to be benefit from this road. This eligibility rule induced a nonlinear relationship between the population and the number of new all season roads in the villages of India today. In order to control for factors like communities collective action ability that simultaneously determine the timing of completion of roads and students enrollment in the schools, I instrument new roads with a population eligibility criteria. Exploiting the exogenous nature of the program eligibility criteria I compare student’s enrollment in schools between villages on either side of the population cutoff. The most conservative estimates show that an improved access to school by better roads increases school enrollment by 22 percent in 2009. I find no spurious effects in 2002 when the same villages did not have a road. The effect of better access to roads on students’ enrollment is heterogeneous depending on the age cohort and the caste to which they belong.

Keywords: Infrastructure, Human Capital, Regression Discontinuity, India

JEL Classification: H54, J24, I25, C14, C21, I21