Labor market effects of dirty air. Evidence from administrative data

We study the impact of air pollution on labor supply and wage compensations in Italy. Matching administrative data on the universe of Italian dependent employees in the private sector with $${\hbox {PM}}_{10}$$ PM 10 concentrations and weather data at monthly frequency, we exploit exogenous variation in wind speed to instrument for endogenous air pollution exposure. We find that a one standard deviation increase in $${\hbox {PM}}_{10}$$ PM 10 level, leads to a 7.3% higher probability of sick leave and to an earning loss of 0.83 euros/worker/month. These figures generated total social excess expenditures of 755 million euro during the period 2011–2016 considering a pollution threshold set by the World Health Organization and extending the effects to the total workers population. The heterogeneity analysis shows that the impacts are larger for workers in constructions and services, for white and blue collars and for females and foreign workers, while we find no impact on managers. The sick wage received by exposed workers is not always aligned to the pollution exposure actually faced by different worker categories.