research

We estimate the causal effect of FEMA aid and SBA loans (federal aid) on economic recovery post disaster in the United States from 2003-2021. For the identification, we use a novel instrument based on the political competition within county and it’s effect on the volume of federal aid distribution, especially when disaster occurs closer to the election. Due to the winner-takes-all nature of the US election, we argue that in non-swing states, politicians in the swing counties have no incentive to perform any better, and thus political competitiveness affects GDP only through federal aid. We find that a 1% increase in disaster aid results in a 0.03% increase in local GDP for the next year. With respect to the literature on political competition, we found evidence that political competition within a county is positively associated with 1) higher volume of disaster aid in non-swing states and 2) higher GDP growth rate in swing states.

We examine the effect of affordable high-speed mobile internet on children’s educational outcomes using the nationwide rollout of Reliance Jio’s 4G network in India in 2016 as a natural experiment. Jio’s entry led to an abrupt and unanticipated decline in mobile data prices and a rapid expansion of 4G coverage. Exploiting cross-district variation in early exposure to Jio’s network, we implement a difference-in-differences design using district-level test score data from the Annual Status of Education Report (ASER). We find that districts with greater initial access to low-cost 4G internet experienced statistically significant improvements of approximately 0.18 standard deviations in reading scores and 0.15 standard deviations in mathematics scores between 2016 and 2018. These effects are robust to alternative treatment definitions, a continuous treatment-intensity specification, and controls for local economic activity and demographic composition. Our results suggest that widespread access to affordable, mobile-first internet can complement traditional education inputs and improve learning outcomes in developing-country contexts.