In many parts of the world, groundwater is being extracted at rates that exceed natural renewal rates, leading many experts to conclude that groundwater mining is an unsustainable strategy for rural development. However, groundwater extraction could still be considered to be ‘weakly sustainable’ if it generates investments in man-made capital that offset the loss of natural ‘groundwater capital’ and enable income to be sustained even as the resource is depleted. In this paper, we examine whether the withdrawals of groundwater for irrigation in India have resulted in household-level investments in physical and human capital. Our empirical strategy compares villages in the same administrative units that overlie aquifers with different water storage capacities and, therefore, endowed with different levels of access to the resource. We find that greater access to the resource results in more irrigation, as expected, as well as higher asset wealth and educational attainment. The results suggest that deeming India’s irrigation-based rural development as unsustainable purely because of the decline in water tables may require reconsideration, provided that human capital is used to productively shift income from farm to off-farm sources.

Can place-based policies shape the spatial distribution of human capital through selective migration? We use geographic discontinuities in irrigation infrastructure in India granular survey data to show that women living in irrigated areas have 2.3 additional months of schooling relative to their non-irrigated neighbors. With nearly 80% of women in our sample being migrants, we show that this difference is largely driven by the selective in-migration of better-educated women into irrigated areas for marriage purposes. By attracting more educated brides, irrigation infrastructure indirectly leads to a spatial sorting of human capital, potentially enhancing socioeconomic disparities directly resulting from infrastructure placement.