Marijuana-related traffic fatalities have increased dramatically in recent years. This increase is nationwide and is not strictly limited to states that enact medical or recreational laws, suggesting the perceived risk of cannabis use in consumers is abating. As of this paper, 19 states have not enacted medical or recreational cannabis laws and the drug remains illegal. Municipalities within those 19 states, however, have passed their own measures reducing the punishment and de-prioritizing the enforcement of cannabis laws. In this paper, we exploit the temporal and geographic variation in city cannabis laws to estimate the effect cannabis decriminalization has on fatal traffic crashes. Using a difference-in-differences approach, we find that these laws lead to differing effects among various demographic subgroups. Our results suggest that laws prompt an overall increase in traffic fatalities in young adults between 15 and 24 and a decrease in the likelihood of fatal accidents among retirement age group. Additionally, the positive effect is most pronounced on weekend nights for young males, while there is a negative effect on weekdays traffic fatalities for older males, including prime-working (aged 25-44), experienced (aged 45-61) and retried age (aged 62+) groups. Thus, we argue the decrease in the legal penalty of cannabis use is affecting consumer choices on the extensive margin. Anecdotally, with a lower cost of marijuana consumption, there may be a potential increase of marijuana use among young adults in the weekend for leisure reasons with a non-negative travelling incentives. Meanwhile, such laws may induce marijuana use at home among working and retired age males for medicinal purposes such as chronic pain relief and sleep issue treatments. These results differ from previous work examining the impact of marijuana legalization laws (Anderson et al., 2013; Santaella-Tenorio et al., 2017; Hansen et al., 2018) and provide insight into the heterogeneous effects of the various levels of cannabis liberalization on traffic safety.
We depart from the existing literature in two key dimensions. First, we focus on the decriminalization of cannabis in states that otherwise oppose the legality of the drug. Thus, our results will have policy implications in non-medical marijuana states and will contribute to the literature by shedding light on the impact of black market cannabis consumption on traffic safety. Second, where prior literature aggregates outcomes to state-year observations, we focus on policy changes at the city level. This more granular approach allows us to control for local level characteristics to better isolate the impact of cannabis liberalization on traffic fatalities.
To estimate the effect of cannabis decriminalization on traffic fatalities, we pair a self-constructed dataset which documents city and municipality decriminalization laws with traffic fatality data from the Fatality Analysis Reporting System. The FARS is a census of fatal motor vehicle crashes that provide detailed information on all individuals involved in the crash. This information includes driver information, crash location, and whether drugs or alcohol were detected in the drivers involved. Our outcomes of interest include total traffic fatalities by demographics such as age and gender during weekdays versus weekends.
