Association for Public Policy Analysis & Management

Rural workfare programs guaranteeing work at a pre-determined wage are intended to provide security to the unemployed and underemployed during the agricultural off-season and are an increasingly used feature of labor market policy. In recent empirical work, India’s National Rural Employment Guarantee Act (NREGA), the largest rural workfare program in the world, has been attributed with crowding out work, raising private sector wages and lowering rural-urban migration. The empirical literature is agnostic about the sign and magnitude of spillovers generated by the large scale program though. This paper studies the spillover effects of NREGA on migration cost, time allocation and casual wages in areas which did not receive the program over the study period.

This analysis builds a simple time allocation model for households in non-program districts with an outside option of migrating or commuting to a neighboring NREGA district. This model predicts that in non-program districts, private wages rise and average non-labor force participation decreases with increasing exposure intensity. This model’s predictions are tested using a nationally representative employment survey and spatial data capturing distance and thus, intensity of exposure to NREGA for non-program districts.

Combined with migration information, it is demonstrated that short-term out-migration from non-program districts was driven by cost of migrating to a NREGA neighbor. Exploiting the shock to exposure intensity due the program’s staggered rollout and spatial variation in district boundaries, the spillover effects of NREGA on casual wages and individual and household level labor market outcomes in non-program districts are then estimated. Our primary results show that average time allocated to non-labor force participation decreased by 1.8 percentage points in districts surrounded by NREGA neighbors, as compared to districts with no NREGA neighbors. The real casual wage is also estimated to have increased by 6.8%, thus providing support for theoretical predictions.