Three Essays on Development and Health Economics

This dissertation consists of three essays on development and health economics. In the first chapter, I study how abortion responds to drought-induced transitory income shocks and generates unintended demographic consequences under son preference. I focus on rural Vietnam where low rainfall induces a short-run downturn through a reduction in rice yields. With widely available sex-selection technologies at a low cost under son preference, Vietnamese parents can decide the quantity and the sex of child simultaneously, and it can be directly observed from rich household-level data on abortions. Linking rich microdata on fertility with droughts defined at a fine geographic unit, I first find no effects of droughts on the number and the composition of mothers who conceive. I then find compelling evidence that affected mothers were 30% more likely to get abortions, and the effect was mainly driven by the income effect because most abortions occurred in the pre-harvest season of the next rice crop when consumption smoothing is difficult. Surprisingly, droughts are associated with disproportionately more abortions of female fetuses, which exacerbated the problem of the skewed sex ratio: the affected birth cohorts become more male-biased due to the six abortions of female fetuses to one aborted male fetus, explaining up to approximately 3% of the sex ratio imbalance in rural Vietnam from 2004-2013. While a full rebound in births in approximately two years appears more consistent with the effect on the timing of fertility, the effect on the sex ratio at birth emphasizes that even transitory income shocks can have long-run demographic consequences. Thus, this study can shed light on how the gender gap can persist during a process of economic development. This study also enhances our understanding of the mechanism through which credit-constrained mothers adjust their fertility to smooth consumption. Finally, this study can provide timely evidence to developing countries which witness demographic transitions to low infant mortality but are vulnerable to extreme weather events. In the second chapter (joint work with Anna Choi and Semee Yoon), we study how usual economic activities can harm the health of people who are living in other countries. This study investigates the adverse effect of transboundary particulate matter on fetal health. The adverse health effects at exposures to particulate matter are evident by a handful of experimental and epidemiological studies. The health effects of PM2.5, which has a diameter of less than 2.5 micrometers, are particularly alarming because those hazardous particles are so diminutive that they can easily enter the bloodstream to cause cardiovascular and respiratory diseases. Unlike the consensus in the United States on the negative impact of pollution on human health, the evidence for the relationship between pollution and health in developing countries is not straightforward to quantify due to the lack of accurate pollution and welfare measures as well as the difficulty of finding exogenous variables to purge other endogenous factors. However, this study can circumvent these endogeneity concerns by exploiting the unique meteorological settings which can trigger transboundary transport of particulate matter. The westerlies from heavily polluted eastern China carry pollutants to South Korea, thereby intermittently exposing the population to pollution above threshold levels. We find that conditional on local weather and pollution trends, one standard deviation increase in Beijing’s PM2.5 explains 1.1% of standard deviation of daily fetal mortality rates in South Korea. We hope that the results of this research can suggest the first accurate cost estimates of transboundary fine-particle to highlight the urgent need for regional cooperation. In the third chapter, the unintended consequences of economic activities on human capital in developing countries can be further emphasized by a randomized control trial study (joint with Hyuncheol Bryant Kim, Booyuel Kim and Cristian Pop-Eleches). We use a four-year long follow-up of an intervention based on a two-step randomized design within classrooms in secondary schools in Malawi to understand the impact of male circumcision on risky sexual behaviors and the role that peers play in the decision and consequences of being circumcised. Although medical male circumcision can reduce HIV infections, its preventive effects may diminish if circumcised men are more likely to engage in risky sexual behaviors. Despite a number of short-term studies of risk compensation following male circumcision, there is scant rigorous evidence on how these behavioral responses change in the longer term. This study is the first evaluation of risk compensation over such a long follow-up period. Our analysis yields three main results. First, we show that the intervention substantially increased the demand for male circumcision for the students assigned to the treatment group. Second, we find evidence of positive peer effects in the decision to get circumcised among untreated students. Third, we find evidence of risk compensation using biomarkers of sexually transmitted infection for those who got circumcised due to the intervention, but not for those induced by peer effects