| Popis: |
1 billion people worldwide live over 2 km from a paved road. Consequently, I investi-gate medium-run impacts of rural road construction on structural transformation in India- identifying how responsive such benefits are based on a) external market condi¬tions and b) in-village electrification. I leverage a regression discontinuity design and triple difference strategy, exploiting discontinuities in population-based eligibility and staggered rollout of the Indian PMGSY rural road program- which aimed to provided all-weather road (AWR) connectivity to 115, 000 villages nationwide. I combine the program with a unique natural experiment induced by the US fracking boom, which created a parallel agricultural commodity boom in the price of guar, a crop provid¬ing a necessary fracking input. I compare heterogeneous impacts of AWRs in villages with high and low-intensity exposure to the fracking boom, and separately investigate heterogeneity of roads by village electrification access, exploiting variable implemen-tation intensity of the nationwide RGGVY electrification program. My results im¬ply structural transformation benefits of AWRs are relatively unresponsive to village electrification, whereas external economic conditions can drastically influence these impacts. RD analysis showcases labor reallocation gains from AWRs were entirely concentrated in non-Boom villages- where roads caused a 12.1-7 percentage-point reduction in share of workers employed in agriculture, and 9.2-8 percentage-point in-creased share employed in non-agricultural manual labour. Conversely, AWRs caused significantly reduced (net zero) structural transformation benefits in boom villages. My findings are robust to multiple specification tests, varying electrification levels, and suggest substantial within-village heterogeneity, with largest discrepancies in new labor market entrants. A plausible mechanism is reduced out-migration impacts of AWRs in boom-villages. These results confirm theoretical predictions that local eco¬nomic conditions can drastically influence the impact of infrastructure investments suggesting the need for effective spatial and temporal targeting. |
