The Human Impacts of Air Pollution: Three Studies Using Internet Metrics

Abstract

Chapter 1: We provide first evidence of a link from daily air pollution exposure to sleep loss in a panel of Chinese cities. We develop a social media-based, city-level metric for sleeplessness, and bolster causal claims by instrumenting for pollution with plausibly exogenous variations in wind patterns. Estimates of effect sizes are substantial and robust. In our preferred specification, a one standard deviation increase in AQI causes an 11.6% increase in sleeplessness. The results sustain qualitatively under OLS estimation but are attenuated. The analysis provides a previously unaccounted-for benefit of more stringent air quality regulation. It also offers a candidate mechanism in support of recent research that links daily air quality to diminished workplace productivity, cognitive performance, school absence, traffic accidents, and other detrimental outcomes.

Chapter 2: We provide linear and non-parametric estimates of the causal impact of short-term exposure to polluted air on the prevalence of cough in a panel of a hundred Chinese cities. In our central estimate, which exploits plausibly-exogenous variations in the number of agricultural fires burning in the vicinity as an instrument, we find that a one standard deviation increase in airborne pollution causes a roughly 5% increase in the prevalence of cough in the affected city. Amongst pollutants the effect can be tied specifically to particulate matter (PM2.5). The results prove resilient in a series of robustness tests and falsification exercises.

Chapter 3: We provide the first study of the relationship between air pollution and students' migration intentions for higher education. Young people's interest in local study is proxied by their Baidu search index for local universities. The IV method is supplemented to identify the causal link by instrumenting for particular matter with plausibly exogenous variations in temperature inversion strength. The estimates of effect sizes are substantial and robust. When air quality in Beijing moves from good-day level to moderately-polluted level, people's search for local education decreases by 3.8% under OLS and 11.8% under IV. The results release the signal that people lost their interest in local universities due to the elevated air pollution. There could be future out-migration to cleaner cities for higher education.