Long-Term Mortality Effects of US Air Pollution

Exposure to air pollution is an established cause of increased human morbidity and mortality but quantifying these effects is difficult. Observational studies that examine the correlation between life expectancy and air pollution across space and over time may not account fully for many factors other than pollution levels that contribute to mortality. Quasi-experimental studies can address these challenges, but they usually measure pollution exposure and health outcomes over relatively short time periods and do not shed light on the long-term effects associated with pollution exposure over years or decades.

In The Long-Run Effect of Air Pollution on Survival (NBER Working Paper 31858), Tatyana Deryugina and Julian Reif measure the health effects of exposure to sulfur dioxide (SO₂), a byproduct of the combustion of coal and oil that is a major source of fine particulates. They combine data on death records with measures of air pollution from the Environmental Protection Agency’s Air Quality System database over the period 1972–1988. They exploit changes in wind direction, measured on a grid with a resolution of about 86 miles, along with data on SO₂ levels to calculate the variation in county-level SO₂ levels that results from wind-blown regional, but not local, pollution.

The researchers uncover two distinct mortality effects of acute exposure to SO₂: a mortality displacement effect that manifests as an immediate increase in the mortality rate among frail individuals who had a short life expectancy, and an accelerated aging effect among healthier individuals that also speeds up mortality but does not result in immediate death. Following a 1 part-per-billion (ppb) rise in SO₂ levels, the research finds a statistically significant increase in same-day mortality rates for all age groups older than 20. The effects range from 0.017 deaths per million individuals for 20-to-44-year-olds to 2.3 additional deaths per million for those over age 85. On the day of the SO₂ spike, death records show increased mortality from many causes, including cardiovascular disease, cancer, and other diseases. This suggests that frail individuals with many health conditions are affected by the higher pollution level. Over time, however, only the mortality rate from cardiovascular and other diseases remains elevated: it rises by a factor of four at a horizon of one month. Over the same horizon, the rise in pollution has no effect on cancer mortality, a finding consistent with the increase in short-run cancer mortality being driven by individuals who were very likely to die in the next month.

The researchers use their estimates to assess the lifetime effect of a permanent decline of 1 ppb in SO₂ levels, roughly a 10 percent drop in average pollution levels during their sample period. They estimate that this would increase life expectancy by about 1.2 years. This effect is more than 7 times larger than the change in life expectancy that would be implied by the study’s short-run estimates alone, suggesting that accounting for the long-run health effects of pollution exposure is essential. Three-quarters of the improvements in life expectancy occur after age 65, which implies that cohorts born after the passage of the 1970 Clean Air Act have yet to experience the survival benefits of the drastic reductions in US air pollution over the past 50 years.

—Abigail Hiller

This research was supported by the National Institutes of Health under Award Numbers R01AG053350 and R01AG073365.