Transboundary Wildfire Smoke and Expressed Sentiment: Evidence from Twitter

Rui Du,Ajkel Mino,Jianghao Wang,Siqi Zheng

DOI: https://doi.org/10.2139/ssrn.4066034

2022-01-01

Abstract:This paper exploits the exogenous variation in wind directions to identify the causal effect of transboundary wildfire smoke on the real-time sentiment of Twitter users in Southeast Asia. We find that a one-standard-deviation increase in the difference between upwind and downwind fires reduces the sentiment score by roughly 2 percent of a standard deviation. The adverse sentiment effects vary substantially across countries and increase in size with average income, average sentiment level, proximity to fires, adaptability and flexibility, and daily temperature. We show that cross-boundary air pollution is the underlying channel at work and cross-border pollution has much greater sentiment impacts than domestically produced wildfire smoke, highlighting the frequent geopolitical tensions and public outcry about the cross-border air pollution spillover. These findings have important implications for mitigating cross-boundary externalities and psychosocial costs from rising wildfire risks.