A Causal Concentration-Response Function for Air Pollution: Evidence from Wildfire Smoke

A Causal Concentration-Response Function for Air Pollution: Evidence from Wildfire Smoke

David Molitor (University of Illinois)

Many countries regulate air pollution to reduce harm to human health. The health effects of marginal pollution reductions matter greatly for determining optimal environmental policy, especially for developed countries where pollution levels are relatively low and further reductions may be very costly. We study the dose-response relationship between exposure to fine particulate matter (PM2.5) and mortality caused by exposure to wildfire smoke plumes of varying intensity. Using a novel satellite-based dataset of daily smoke plume coverage for the entire US from 2007 to 2017, we find that drifting wildfire smoke generates frequent and significant variation in ground-level fine particulate matter in places far from the wildfires. We link this variation to 100% Medicare administrative data to provide the first nationwide evaluation of the mortality cost of wildfire smoke exposure among the U.S. elderly. We have three main findings. First, wildfire smoke accounts for over 20% of average PM2.5 concentrations in the US. Second, wildfire smoke exposure significantly increases 3-day mortality for the elderly. Third, we find that the dose-response relationship between PM2.5 and mortality caused by wildfire smoke is concave: large air pollution shocks have proportionally smaller mortality effects than smaller air pollution shocks. Methodologically, our study shows how different studies could find different estimated effects even for the same pollutant if they are examining pollution variation of different magnitudes. From a policy perspective, our results point to large benefits of additional air quality improvements in the US.