Development & Health

Abstract: "Over the past two decades, genetically modified (GM) crops have transformed global agriculture, yet their broader health consequences remain contested. This paper examines the medium- and long-term impact of Bt cotton adoption, a pest-resistant variety introduced in India in 2002, on infant mortality. Using geo-referenced data from the National Family and Health Surveys and spatial variation in agro-climatic cotton suitability across Haryana, Punjab, and Rajasthan, I implement a difference-in-differences design that exploits the staggered diffusion of Bt cotton. I find that adoption is associated with a significant increase in infant mortality: locations in the top two quintiles of cotton suitability experienced increases of about 0.77 percentage points relative to low-suitability areas. The results are robust to alternative estimators that account for staggered adoption, state-year shocks, and local confounders, as well as to placebo tests and triple-difference specifications. Evidence on mechanisms suggests that rising seed costs and increased maternal labor participation contributed to the observed health effects. These findings highlight unintended consequences of technological change in smallholder agriculture, underscoring the need to evaluate agricultural innovations not only in terms of productivity but also in terms of human development and child well-being."

Data: National Family and Health Surveys, FAO-GAEZ, Administrative data from the Indian Ministries of Agriculture and Textiles, 2001 Census of India.

 Presented in (and scheduled *) 2026 Bristol Applied Economic Meetings (University of Bristol), European Economic Association 2026 Meeting (University College Dublin)

Abstract: "Over the past two decades, genetically modified (GM) crops have transformed global agriculture, yet their broader health consequences remain contested. This paper examines the medium- and long-term impact of Bt cotton adoption, a pest-resistant variety introduced in India in 2002, on infant mortality. Using geo-referenced data from the National Family and Health Surveys and spatial variation in agro-climatic cotton suitability across Haryana, Punjab, and Rajasthan, I implement a difference-in-differences design that exploits the staggered diffusion of Bt cotton. I find that adoption is associated with a significant increase in infant mortality: locations in the top two quintiles of cotton suitability experienced increases of about 0.77 percentage points relative to low-suitability areas. The results are robust to alternative estimators that account for staggered adoption, state-year shocks, and local confounders, as well as to placebo tests and triple-difference specifications. Evidence on mechanisms suggests that rising seed costs and increased maternal labor participation contributed to the observed health effects. These findings highlight unintended consequences of technological change in smallholder agriculture, underscoring the need to evaluate agricultural innovations not only in terms of productivity but also in terms of human development and child well-being."

Data: National Family and Health Surveys, FAO-GAEZ, Administrative data from the Indian Ministries of Agriculture and Textiles, 2001 Census of India.

 Presented in (and scheduled *) 2026 Bristol Applied Economic Meetings (University of Bristol), European Economic Association 2026 Meeting (University College Dublin)