We estimate the impact of air pollution on labor supply among workers affiliated to Spain’s universal sickness-leave insurance. A 10% reduction in high-pollution days reduces the sick-leave rate by 0.8% of the mean (2.79%). Based on this estimate, we causally attribute 5.6 million avoided sickness days to air quality improvements that occurred in urban Spain between 2005 and 2014, saving €0.5 billion in foregone production. Across workers, treatment effects increase with age but decrease with overall health and with individual unemployment risk. This is consistent with vulnerable workers sorting into jobs with better employment protection. It also rationalizes higher observed absence rates in the public sector and among workers with permanent contracts vs. those with temporary contracts. Within workers, we estimate that reductions in unemployment risk increase absenteeism when pollution is low. This suggests that measuring the pollution-absence gradient under high unemployment risk yields a closer approximation to the underlying (unobserved) pollution-health gradient.
(Joint with Laura Hospido (Banco de España) and Ulrich Wagner (University of Mannheim))
