Crime levels near federally legal cannabis stores in two Canadian cities.

Armstrong M. · The International journal on drug policy · 2026

**Background & Methods** This was a retrospective comparative interrupted time series analysis examining police-reported crime data from Toronto and Ottawa, Canada during January 2018–February 2020 (15 months pre-intervention and 11 months post-intervention). The study evaluated whether crime levels changed following the opening of eight initial licensed recreational cannabis stores in April 2019, using geographic information system software to identify crimes within 3,000 meters of store locations stratified into four distance categories, with fixed effect panel data regressions to assess associations between store proximity and crime trends. **Key Findings** • Police reported 31,095 total crimes within 3,000 meters of the initial cannabis stores during the study period, with monthly crime levels slightly elevated after April 2019 compared to baseline; however, this increase was not statistically associated with cannabis store proximity. • Increased crime levels showed significant associations with seasonal factors (warmer winter weather) and pre-existing temporal trends rather than store opening, with findings remaining robust across multiple sensitivity analyses. • The study found no significant associations between legal cannabis store openings and changes in either personal crime (one category examined) or property crime (four categories examined) in the geographic areas studied. **Dosage & Administration** Not reported. **Safety & Adverse Effects** Not reported. **Evidence Quality** This study provides moderate-quality evidence with several important limitations. The analysis was restricted to two Canadian cities and only eight initial stores, limiting generalizability. The 11-month post-intervention observation period may be insufficient to detect longer-term crime pattern changes. The study is descriptive in nature (review article) and relies on police-reported crime data, which may be subject to reporting bias and does not capture unreported criminal activity. Attribution of causality is limited by numerous confounding variables (weather, socioeconomic factors, policing practices) that could influence crime trends independently of cannabis legalization. The fixed effect panel regression approach strengthens causal inference but cannot definitively establish causation. Results may not generalize to jurisdictions with different cannabis regulatory frameworks, enforcement patterns, or baseline crime characteristics.

More recent cannabis research