Working Papers

“The Effect of Export Activity on Domestic Prices: Evidence from India's Rice Sector"

[Latest Draft]

Coverage: World Bank Development Impact Blog

How does export activity affect prices in domestic consumer markets? To explore this question, I exploit a natural experiment provided by India’s rice export restrictions during 2007-2011. I first document that the binding restrictions had a considerable negative impact on producers. However, there is little evidence to suggest that the lower prices transmitted to consumers. To estimate the causal impact of export activity on domestic consumer prices, I use a difference-in-differences framework. I show that, following the imposition of export restrictions, Indian districts with higher exposure to export trade experienced a greater increase in prices paid by consumers in local markets. To measure the intensity of exposure to trade, I use a novel strategy exploiting spatial variation in districts’ proximities to export trade routes along India’s road network. The estimated price effects are substantial – prices increased by an additional 5 to 6.5 percent in districts exposed to export activity. Further, the impact of export activity on prices is most pronounced for higher-quality products. I illustrate that the presence of strong complementarities between exports and intra-national trade is the main mechanism driving my results. By exploiting synergies with export activity, intra-national trade encounters lower domestic trade costs, which are then translated to lower prices faced by consumers in domestic markets. My findings suggest that promoting export activity could be a means to reduce intra-national trade barriers for large developing economies.

“Political (Mis)selection: The Role of Historical Institutions"


Is there a persistent impact of historical institutions on the selection of “bad" politicians? In this paper, I document the nuanced consequences of history on modern adverse political selection using variation in colonial land tenure systems during British rule in India. Politicians with criminal charges are significantly more likely to run for office, and win elections, in areas where extractive landlords historically held substantial property rights. Corrupt politicians who experience disproportionately large private financial gains in office, however, are significantly more likely to be in areas where individual cultivators historically held property rights. My analysis establishes these estimates are not biased by endogeneity concerns. Rather, a plausible mechanism indicates a comparative advantage for criminal politicians in the conflictual electoral environment of landlord areas. The relatively less competitive elections in non-landlord areas, instead, are susceptible to elite capture and subsequent wealth accumulation by politicians.

“Financial Inclusion for Smallholder Farmers using Digital Credit: Experimental Evidence from Ghana", with Dean Karlan, Monica Lambon-Quayefio, and Christopher Udry

AEA RCT Registry ID 0005829 [Pre Analysis Plan]

Coverage: The Center for Effective Global Action Blog

Financial market imperfections pose major barriers to low income farmers throughout developing countries. If smallholder farmers' credit constraints can be relaxed through adequate and timely credit, there could be substantial welfare gains through greater agricultural investment and downstream outcomes. To examine this, we design a randomized controlled trial where we offer credit to farmers in southern Ghana through a fully digitized process, incorporating a promising technology for financial inclusion. We find that the treatment increases farmers' farm input expenditures but has limited impact on crop production, sales, and profits. However, our intervention faced loan delivery delays, and for a plausibly endogenous subset of farmers who received their loans on time, there is a large increase in these outcomes. We also document suggestive evidence for resource reallocation by women farmers, who seem to use their loans not on their farms but on their economic enterprises. Working paper available soon.

“Digital Networks and the Diffusion of Political Movements: Evidence from Mobile Internet in Africa", with Ricardo Dahis and Joris Mueller

[Most Recent Slides]

We study the diffusion of political movements via digital networks, exploiting the rapid rollout of 3G mobile internet across the African continent over the last decade. Our main result illustrates that political protests are significantly more likely to spill over between places that share a language after they get connected to each other through 3G. To arrive at our results, we use an innovative empirical design motivated by a gravity model estimated on a dyadic dataset. We document that the spillovers between places connected to 3G are practically independent of spatial distance. This is in sharp contrast to how spillovers between places without 3G decay with spatial distance. The results suggest that communication networks may contribute to the spread of political movements beyond the geographic neighborhood traditionally studied in the literature. We confirm that our results are resilient to a battery of robustness and placebo checks. We then show evidence that social media usage and group identities can play important roles in explaining our results. We are currently exploring these mechanisms using novel data on social networks provided to us by Facebook.

“Technology, Information and Integration: Evidence from India's National Agriculture Market"

Agricultural markets in developing countries are spatially fragmented and lack adequate use of technology. Recent literature highlighting benefits of information and technology adoption in agriculture have focused on farmer level interventions. In contrast, there is relatively little evidence on the effects of market level interventions. In this paper, I analyze the effects of one such intervention: India's National Agriculture Market (eNAM), an electronic trading portal designed to increase transparency in trading and creating a unified market. Exploiting the staggered adoption of eNAM across India and a dataset comprising 5.5 million observations, I show that prices received by farmers increase by 2 to 4 percent as a result of eNAM. Further, the increases are larger for higher quality produce. At the same time, there is an increase in the volume of trade in these markets. Taken together, these mechanisms suggest that the technology increases farmers' bargaining power and enables them to receive better prices. Working Paper available soon.

Other Works in Progress

“Does Land Use Consolidation Work? Evidence from Rwanda", with Jeremy Magruder

“The Effects of Digital Advertising on Small and Medium Businesses: Evidence from a Large Online Experiment", with Chris Hooton

“Mothers Bearing the Cost of Daughters: Evidence from Ghana"