| dc.description.abstract | I present three essays on the interaction between environmental policy, ambient air quality and human health. The first essay explores how trees in urban areas benefit human health. I provide causal evidence on the elasticity of air pollution and mortality to urban forest loss from the exogenous introduction of the emerald ash borer insect to the continental United States. Trees primarily benefit urban health by reducing pollution; dieback that affects up to 4.2% of city forests is associated with increases in mean PM2.5 that reach 4.4%. Damage to city forests ultimately leads to excess deaths of up to 1.8%; much of this increase is driven by increases in cardiovascular and respiratory disease mortality. If the estimated median elasticity of all-cause mortality to tree damage of 0.42 is extrapolated to all forests in the urban continental United States, my results imply urban forests reduced all-cause mortality by 29.3%, or 299000 deaths total, in 2014. In the second essay, I study if carbon prices improve local air quality in addition to reducing carbon emissions. Using variation in distances from coal power plants to local authority districts in Great Britain, I find significant and spatially extensive reductions in ambient nitrogen oxide levels, of up to 0.9 ugm^{-3} (4% at the mean) per gigawatt (GW) of nearby coal capacity following the introduction of a carbon price. These reductions are associated with significant (but less extensive) reductions in all-cause mortality, of up to 2 deaths per 100000 per month per GW of nearby coal capacity, and mortality from circulatory disease near these power plants. There is no evidence for increases in mortality rates near the closest competitors to coal power plants. These findings suggest that improvements in health are an additional reason to implement carbon prices more widely. In the third essay, I consider the effect of omitted variable bias from cross-sectional differences in the time the outbreak started, a bias which is not eliminated by using panel models with cross-sectional and time fixed effects. Using polynomial models of disease spread over time to correct for this bias, I find that historical levels of PM2.5 have increased mortality rates from COVID-19, with an elasticity of .541, 95% CI (.479, .603) as of May 18. My estimates are about half of prior cross-sectional estimates. However, morbidity rates are unaffected by the same variable, suggesting that historical exposure to PM2.5 increases the severity of COVID-19 infection. | |
