The effects of PM2.5 concentrations on traffic violations and accident severity in Guangdong China

Abstract

How does air pollution affect road safety? This study examines the impact of local PM_2.5 concentrations on traffic violations and accident severity in Guangdong Province, China, from 2006 to 2014. Using instrumental variable methods to address endogeneity, we find a significant relationship between PM_2.5 concentrations and the occurrence of high-penalty traffic violations and severe accidents. Robustness checks confirm these results hold when using alternative pollutants (PM₁₀, SO₂, and SO₄). To account for the bias caused by nonlinear relationships between air pollution and weather, we employ a random forest model to isolate the weather-independent component of PM_2.5 concentrations. Further analysis reveals significant heterogeneity in these effects across driver characteristics and geographic areas. This paper also explores potential visual and psychological mechanisms through which PM_2.5 influences traffic safety. Our findings offer valuable policy implications for environmental and traffic management strategies.

Keywords: Air pollution; Instrumental variable; Machine learning; Road traffic accidents.

Fig. 1

Distribution of PM_2.5 Concentration, Traffic Violations and Severity in Guangdong Province. The figure was created using the software QGIS Desktop, version 3.40.2.

Fig. 2

The Trend of PM_2.5, Point and Severity. The figure was created using the software Stata, version 18.

Fig. 3

Correlation Heatmap of Variables. The figure was created using the software Stata, version 18.