Airless democracy: Air pollution and voter turnout

Air pollution is a major externality whose consequences extend beyond health and productivity. This paper shows that short-run pollution shocks also reduce democratic participation. We combine official, municipality-level election results from 32 national, European, regional, and municipal elections in Italy (2013-2022) with newly assembled daily measures of PM2.5, PM10, and NO2 for all Italian municipalities. Our identification strategy exploits quasi-random election-day deviations in local pollution relative to recent conditions, and we corroborate the results using wind speed as an instrument for particulate matter. Higher pollution on election day substantially depresses turnout: a 10 μg/m3 increase in PM2.5 (roughly doubling typical exposure) lowers participation by 2-3 percentage points, corresponding to about one million fewer votes. The estimates are similar for PM10 and NO2, and when pollution exceeds WHO guideline thresholds. Using post-election survey data from the 2013, 2018, and 2022 national elections coupled with survey-date exposure, we find consistent individual-level declines in reported voting intentions, with larger effects among citizens who report higher political interest. These findings identify the political-economy cost of air pollution, which not only reduces turnout but distorts the democratic representation by altering who turns out, not just how many. Our results suggest that environmental regulation can strengthen the democratic process by improving political participation and representation, in addition to its health and welfare benefits.