Research

Land Institutions, Agricultural Productivity, and Climate Shocks                           (Job Market Paper)

Land institutions shape agricultural productivity in the developing world, and are often restrictive. This paper studies the interaction between climate shocks and land institutions, emphasizing the differential effect of drought on agricultural productivity across villages with varying degrees of land rental barriers.  I develop a model featuring occupational choice, risky farm investment, and frictions in land rental markets to quantify efficiency losses from drought under different institutional settings. Calibrated with micro panel data from India, the model shows that drought-induced efficiency losses are three times larger in villages with high rental barriers compared to those with no barriers. Two mechanisms drive this result. First, low-ability farmers reduce fertilizer investment to limit exposure to uninsurable drought shocks, at the cost of productive efficiency. These are precisely the farmers that operate an inefficiently large share of village land because of rental barriers, amplifying the negative effects of land misallocation. Second, high rental barriers weaken incentives for low-ability farmers to select into wage labor, as supplemental income from renting out land is limited. Both channels reduce aggregate productivity during drought. Projecting forward, I show that stronger land institutions substantially mitigate the negative impact of climate change on agriculture, highlighting the importance of institutional reform in adapting to an increasingly uncertain climate.

Trade Liberalization, Structural Change, and Skill Premium Growth in India (submitted) - Draft, Slides

This paper examines the impact of changing trade costs on skill premium growth and sectoral value-added shares in India between 1995 and 2005. I develop a quantitative trade model featuring multi-stage production and two types of labor, in which structural change and skill premium growth arise through relative price effects and shifting comparative advantage. I calibrate the model to India and two trading partners, then simulate trade cost reductions and sectoral productivity growth over time. Counterfactual exercises show that trade cost reductions alone can explain much of India’s service sector expansion, manufacturing decline, and skill premium growth, whereas productivity growth in isolation plays a limited role. The interaction between changing trade costs and productivity growth, however, is quantitatively important. The two-stage structure is important, as it generates an amplified elasticity of relative wages with respect to trade costs, and allows the model to match moments that a one-stage model cannot. 

The Sovereign Spread Compressing Effect of Fiscal Rules During Global Crises, with Ergys Islamaj and Agustin Samano - World Bank Version

This paper studies whether fiscal rules can signal fiscal responsibility and compress borrowing costs for emerging economies during periods of global crisis. Using daily data on sovereign spreads for 58 emerging market economies from 2019-2022 and 26 countries from 2007-2009, this paper shows that the compressing effect of fiscal rules on sovereign spreads is stronger during global crises. We find that the existence of a fiscal rule reduces sovereign spreads with a high degree of statistical significance, regardless of the extent to which enforcement of the rule occurred during the global crisis. In our baseline test covering the COVID-19 timeframe, estimates of the average spread compressing effect of fiscal rules range from 319 to 378 basis points. We also find that for countries that deviated from a fiscal rule during a global crisis, the median duration to return to the baseline fiscal balance is 3.5 years. This fact explains why the spread compressing effect is independent of the enforcement of the rule during a global crisis, as lenders expect countries to return to compliance with the fiscal rule in the aftermath of a crisis. Our results suggest that second-generation rules have increased not only the flexibility but also the credibility of fiscal rules, even during crisis periods. 

Barriers to Adoption: Evidence from BT-Corn, with Honey Batra and Jason Hall - Draft

How well suited to the developing world are developed world technologies?  This question is central to understanding differences in technology adoption across the development spectrum, but has been challenging for the literature to address empirically.  We make progress on this question by exploiting unique characteristics of one of the most important agricultural innovations in recent history - BT corn - to construct a new dataset.  This dataset matches BT crops with the set of insects each kills.  We then leverage recent advances in large language models to extract information from the scientific literature on the country-level importance of each of those insects.  Our approach allows us to directly construct counterfactual yield gains under adoption across the global income distribution.  We find that the gradient of potential yield gains with respect to development is flat.  The level of predicted gains is also substantial in much of the developing world, with our baseline estimates averaging 53% in Sub-Saharan Africa.