Do cannabis stores matter? Analysis of Toronto’s districts and traffic injuries

Author(s): Nazif-Munoz, Dominguez, Brown, Ouimet

Slidedeck Presentation:

Slide deck not available



The consequences of cannabis recreational use policies (CRUP) in traffic outcomes are mixed. Some studies focusing on cannabis-involved fatal crashes have also found significant increases after the enactment of CRUP. In Colorado it was found an increase by 63% in this outcome. These changes, maybe explained by the opening and consolidation of commercial retailers. Similarly in Uruguay, research has pointed out that CRUP may increase traffic fatalities in urban settings when access to commercial retailers increases over time. However, absence of significant changes has also been reported in United States and Canada. In Oregon and Washington, three years after their respective CRUP was enacted no changes in all traffic fatalities were observed. In the provinces of Ontario and Alberta, the enactment of its CRUP was not associated with concomitant changes in traffic injuries in all drivers nor youth drivers.


This study examines the association between the number of cannabis stores (NCS) and road traffic injuries in Toronto 150 days after the enactment of Canada's CRUP.


We explore the association of NCS and all injured road per 1 000 000 population, applying a nave and a fuzzy difference-in-difference study designs. We use generalized linear models with an interaction between time-period and number of NCS per district per population, comparing a period before the introduction of NCS to the 150 days after Canada legalized cannabis recreational use. We adjust for precipitation, temperature, and snow, and apply models to four districts: Toronto-East York, Scarborough, North York, and Etobicoke. Information is gathered from Toronto Police Service, Alcohol and Gaming Commission of Ontario, and Environment Canada.


After 150 days, NCS were associated with increases of 4% (Incidence Rate Ratio [IRR]: 1.04, 95% Confidence Interval (CI): 1.02, 1.04) in Toronto-East York, and 2% (IRR 1.02 95% CI: 1.01, 1.03) in Scarborough; and decreases of 11% (IRR: 0.89, 95% CI: 0.79, 1.00) in North York, and 27% (IRR: 0.89, 95% CI: 0.63, 0.84) in Etobicoke in all injured road users per 1000000 population. Overall, NCS across Toronto were not associated with traffic injury rates (IRR: 0.98, 95% CI: 0.95, 1.02).


Results are sensitive to methods applied. While with a nave DID standard errors are wide with the fuzzy DID design these are more precise. However results suggests that in Toronto taken as a city the CRUP is not associated with changes in traffic injuries 150 days after its enactment. If the spatial distribution and the quantity of cannabis store per capital matter, then results indicate a complex association at the district level. In other words we observe that proportional increases in the number of cannabis stores at the district level brings contradictory outcomes.


More research is needed to understand the extent under which district level difference are explained by, changes in acute cannabis consumption, spill over effects across Toronto, travel patterns explaining where individual actually buy cannabis, and last if differences between these districts exist in terms of road safety initiatives