This dissertation consists of three chapters that analyze the health effects of pollution in China. The first chapter investigates the effect of air pollution on cardiovascular mortality in the urban areas of China. The second chapter estimates the effect of water pollution on infant mortality. The third chapter studies the relationship between water pollution and cancer among the elderly.
The first chapter entitled "The Effect of Air Pollution on Cardiovascular Mortality: Evidences from the Beijing Olympic Games". I explore the exogenous air pollution variations induced by the 2008 Olympic Games to estimate the effects of air pollution on cardiovascular mortality in China. I use the regulation status during the Olympic Games as an instrument for air pollution. In the fixed-effects instrumental variable model, I find that air pollution has a robust and significant effect on cardiovascular mortality. In contrast, estimates from the conventional associational models are not robust. I estimate that decreasing current PM_10 concentration by 10% will save more than 67,000 lives (from cardiovascular diseases) in the urban areas in China each year.
China's surface water system has been severely polluted in the process of rapid industrialization. The second chapter investigates how this water pollution affects infant mortality. I find that surface water pollution has a significant, nonlinear effect on infant mortality. As surface water quality deteriorates, infant mortality first increases and then decreases. Moderate levels of pollution are the most dangerous. People's avoidance behavior may explain the results: as water becomes more polluted people reduce the consumption of surface water. The ordered-probit selection model is applied to estimate the effects, and precipitation and wastewater dumping are used as the instruments for surface water quality.
China also witnessed a dramatic increase in cancer rate in the past thirty years. In the third chapter, I investigate whether this high cancer rate is caused by water pollution. The difficulty in estimating the long-run health effects of pollution is that the lifetime exposure to pollution is hard to measure. However, China provides an ideal setting to estimate the long-run health effects of pollution because the Household Registration System (Hukou) effectively stopped people from migrating for many years. I focus on the elderly people (Age>60) because their mobility is extremely restricted by the System, so their life-time exposure to water pollution is more likely captured by the water quality data in recent years. I find that water pollution has large, significant, positive effects on all cancer mortality rate, digestive cancer mortality rate, urinary cancer mortality rate, liver and stomach cancer mortality rate. I also find that water pollution has no impact on cancer mortality rates for the younger adults (Age from 20-50), which may partially justify our argument that pollution exposure for the younger people cannot be accurately measured because they migrate.