Soutenance de thèse : Claire Lepault

Abstract

This dissertation comprises three empirical essays on environmental and health economics in the context of India. Two chapters focus on the effect of water pollution on infant mortality, while the other investigates the effect of heat stress on anemia prevalence in children and adults. A common objective of the chapters is to estimate the effect of environmental damage on health, relying on new datasets, and to provide a cost evaluation of these effects.

The first chapter investigates the effects of urban sewage treatment on water quality and downstream infant mortality in India. I exploit the staggered introduction of urban sewage treatment plants over the period 2010-2020 using the difference-in-differences method. I show that after starting wastewater treatment, levels of fecal coliforms – a commonly used measure of fecal contamination in water – decreased in average by 50%. Mortality under the age of six months declined by 20% downstream of the plants, with larger effects for boys and children from the bottom wealth quintiles. The results are consistent across several estimators robust to heterogeneous treatment effects, are not driven by selective migration, and are only found downstream of the plants, which rules out confounding effects from other local policies. Applying my results to a back-of-the-envelope calculation, the cost per life saved is INR 6 million and the cost per disability-adjusted life year (DALY) averted is INR 85k ($1,330 in 2015).

The second chapter analyzes the relationship between heat exposure and anemia in India over the past decade. We match blood hemoglobin measures from nearly 259,000 children, 901,000 women and 138,000 men to two different measures of exposure to heat, based respectively on air temperature and wet-bulb temperature. We document an inverted U-shaped relationship between temperature exposure over the past 30 days and hemoglobin levels. Being exposed to a mean air temperature above 33°C, relative to 17-21°C, reduces hemoglobin, and increases anemia prevalence significantly, for children, women, and men. The mechanisms responsible do not seem to be related to reduced food availability, but malaria and direct effects, such as hyperthermia, could explain part of the relationship between temperature and anemia. Our cost estimation of an additional degree of warming, in the most severely affected bin, suggests that the morbidity costs of extreme temperature exposure could be higher than previously measured using data from medical visits.

The third chapter explores the relationship between fertilizer runoff, nitrogen concentration in water, and perinatal health. We document that nitrate levels in many groundwater sources surpass established standards, in contrast to rivers. We find that nitrate concentrations exceeding the Indian government threshold correlate with an increase in neonatal mortality and in child mortality. Finally, we provide evidence that, on average, exposure to nitrogen from fertilizers applied within the district of birth during the first trimester of pregnancy increases neonatal mortality by around 2%. Heterogeneity analysis suggests that this effect is higher for children born in households primarily drinking groundwater, as well as in household that do not treat drinking water and live in rural areas. A back-of-the-envelope calculation suggests that the cost associated with neonatal mortality may surpass benefit on crop yield.