The Politics of Drought Relief : Evidence from Southern India

Increases in weather extremes due to climate change pose a serious threat to the livelihoods of the rural poor and create an increasingly critical role for government intervention. In India, allocations of drought relief from state to local governments systematically deviate from national guidelines – objective environmental criteria that define drought. To assess the motives for these deviations, I develop a dynamic probabilistic voting model that identifies likely patterns of political distortions. Using an instrumental variables approach, I identify these patterns empirically: the probability of receiving relief increases with electoral competition for the state ruling party and is lower for politically aligned areas. This mistargeting of relief relative to national guidelines is costly. I show that failure to respond to drought is associated with a 4% drop in following-year agricultural production – an effect that would be more than mitigated by relief. Despite this, 34% of drought-affected areas do not receive relief and 32% of allocations instead go to areas where relief appears to have no effect on agricultural production