A new peer-reviewed study published in the journal Variance found no statistically significant connection between marijuana legalization and increased vehicle accident fatalities, insurance claim frequency, or average claim cost.
The research was conducted by the University of Maryland Center for Environmental Science and analyzed data from both Canada and the United States using a variety of machine learning and statistical tools.
The study used data from 2016 to 2019 to avoid pandemic-related distortions and found that factors such as weather conditions and weekly travel patterns were far more predictive of accident rates than whether or not marijuana had been legalized. In Canada, researchers found that pre-legalization trends in insurance claims remained stable after legalization. In the U.S., no consistent increase in fatal accident rates was found when comparing states with legal marijuana to matched states without it.
Researchers applied advanced causal inference techniques, including random forest modeling and matrix completion, to ensure accurate comparisons. They also used propensity score matching to align states based on demographic and traffic variables. Across all methods, the results remained consistent: marijuana legalization did not significantly affect fatality rates or crash-related insurance costs.
While the study did identify a single statistically significant decline in accident fatalities when comparing Massachusetts with Florida, no other pairing showed similar results. The authors concluded that marijuana legalization, when viewed alongside other variables, does not appear to be a driving factor behind accident trends, and that broader patterns of human behavior and environmental factors are stronger indicators.
The study provides a data-driven framework for evaluating future policy shifts and their potential impact on public safety and insurance metrics.
