(En-) 'lightening' Children: Assessing the Impacts of Access to Electricity on Children's Achievement Levels. 2023. [PDF] [Online Appendix]. Review of Development Economics. With Somdeep Chatterjee and Jai Kamal. 

Abstract: The welfare impacts of electrification are well documented in the literature, including the effects of electricity on school enrollment. However, the spillover effects of electrification on children’s achievement levels are scarce. We use three complementary but distinct econometric models to establish a causal relationship between electrification and test scores using nationally representative household panel data from India. We find positive results irrespective of the choice of econometric model, and these results seem to be mediated by changing time-use patterns of children with access to electricity. We first exploit the plausibly exogenous variation in access to electricity due to a universal electrification program in the state of West Bengal in India and we find positive effects of electrification on children’s test scores. By age group, we find that younger cohorts benefit more in terms of their reading scores than older cohorts. Then, to ascertain external validity of these results, we replicate them over a nationally representative sample using fixed effects and instrumental variables estimation and find similar results. At the intensive margin, we find that access to more hours of electricity positively affects test scores. We identify an increase in time spent by children on study-related activities as the potential channel for these results.

working papers

Abstract: This paper assesses the impact of bank access on household credit sources in India. Using a nationally representative household survey and a regression discontinuity design based on a 2005 branch expansion policy of the central bank, I find no evidence that the policy induced borrowing households to secure their largest loans from banks instead of informal sources. This is true for different categories of loans, the only exception being education loans. The overall null effects hold across various household characteristics, except some significant effects for a few loan categories. The null effects are driven by the rural sub-sample which constitutes the majority of Indian households. I provide two potential mechanisms for these results: more urban branches in treated districts, and the persistence of informal borrowing. My findings imply that enhancing bank presence may not suffice for changing households’ borrowing behavior, given the widespread preference for borrowing from informal sources in this context. The findings are robust to alternative specifications of the bandwidth, the bandwidth multiplier, and the polynomial degree.

Abstract: This paper exploits random assignment of female quotas for leadership positions on Indian village councils to assess its causal effect on corruption. Since the mid- 1990s, India has mandated that one third of Village Council Chief positions be randomly reserved for women. Using data from the Rural Economic and Demo- graphic Survey (REDS) 2006, we find that an additional term reserved for a female head as compared to just a single one reduces both the occurrence and amount of bribes paid by households. This occurrence also reduces statistically significantly across five different types of officials. We provide two potential mechanisms for these results: women political leaders are more likely to be (i) efficient at solv- ing households’ problems and (ii) transparent in terms of selecting households as beneficiaries for government programs. To the best of our knowledge, this is the first statistically significant evidence regarding the effect of women leaders on corruption in general without regard to any single government project.