As all Montana residents are well aware, the summer of 2017 has been a brutal season for wildfires. As detailed in this recent Wired article, more than 80 major wildfires are burning across the West. Fueled by the summer drought, Montana has been particularly hard-hit, with 26 large fires burning more than one million acres of forest and sending a nasty haze across the entire state.
Beyond the direct economic costs of the loss of homes and forest, the poor air quality resulting from these fires has consequences for public health. Quantifying the impact of particulate matter on health is difficult because places with high levels of pollution (ie cities) tend to have different health outcomes from other regions for a variety of reasons besides pollution. But a recent paper in American Economic Review: Papers and Proceedings uses the prevalence of Indonesian forest fires from 1990-2015 as a natural experiment to assess the impact of the resulting deterioration of air quality in nearby Singapore.
The authors use a “two-stage” statistical approach in which they first find the impact of fires on air quality, and then the effect of air quality on health. They find that a one standard deviation increase in annual fires in Indonesia caused a 0.7 standard deviation increase in clinic attendances for respiratory problems in Singapore.* This result is likely to be somewhat comparable to the health impacts of the current round of fires in Montana, as the average area of forests burned annually in Indonesia over the study period is in the same ballpark as the one million-plus acres burned in Montana this summer.
The paper’s result is compelling, although it’s unclear what the effects of fires are on longer-term health outcomes like chronic diseases and mortality. Still, these health impacts are a less obvious but still substantial additional cost of the drought that has been discussed extensively on this blog. As droughts are likely to become more common as a result of climate change, much more research needs to be done on the overall impacts of large-scale forest fires.
*A one standard deviation increase above the average roughly corresponds to a level of fires that would occur in about 1/6 of years. A 0.7 standard deviation increase above the average occurs in about 1/4 years.