Abstract:
This paper investigates the effects of air pollution on public mood using sentiment analysis of geolocated social media data. Analyzing approximately 7 million twitter posts from the United States in July 2015, we examine how fluctuations in air quality caused by Canadian wildfires influence sentiment. We find robust evidence that higher exposure to particulate matter leads to decreased positive sentiment and increased negative sentiment. Given the importance of mood as a factor in labor productivity, our results suggest that the short-term psychological effects of air pollution, alongside its well-documented physical health impacts, should be considered in policy discussions, as negative shifts in public mood due to poor air quality could have far-reaching economic consequences.