Do Legal Cannabis Stores Increase Nearby Crime?

When a cannabis store opens, neighbourhood concerns often arrive before its first customers. Residents may worry about theft, intoxication, or whether a once-quiet commercial strip will become less safe. These are understandable questions, but answering them requires more than comparing crime totals before and after a storefront appears.

A Canadian study published in the International Journal of Drug Policy offers a measured response. In “Crime levels near federally legal cannabis stores in two Canadian cities,” researcher Michael J. Armstrong found no statistically significant association between licensed store openings and nearby crime in Toronto and Ottawa.¹

The finding does not establish that every cannabis retailer is harmless in every setting. It does suggest that municipalities should examine specific evidence before assuming legal stores create neighbourhood crime problems. More broadly, it demonstrates why evaluating legalization requires separating what happened after a policy change from what happened because of it.

What the Canadian Cannabis Store Study Examined

The research covered eight early licensed cannabis stores: five in Toronto and three in Ottawa. Their surrounding neighbourhoods were observed from January 2018 through February 2020, providing 15 months before April 2019 and 11 months afterward. Most stores opened that April, although two Toronto locations followed in May and June.

Researchers examined assault, automobile theft, bicycle theft, breaking and entering, and theft over $5,000. These were the five crime categories available in both cities’ published datasets. Altogether, police recorded 31,095 incidents within three kilometres of the stores during the study period.

Using geographic information software, Armstrong divided surrounding areas into four walking-distance bands. The analysis compared changes closer to stores with changes farther away, while accounting for existing crime trends and weather.

Ontario’s initial licensing lottery partly randomized store locations, strengthening the research design. However, applicants had already arranged their proposed locations, so this was not a fully randomized experiment assigning cannabis stores to otherwise identical neighbourhoods.

Why Rising Crime Did Not Implicate Cannabis Stores

The raw numbers could easily have produced an alarming headline. During the matched 11-month comparison periods, crime was 10.1 percent higher in Toronto and 5.9 percent higher in Ottawa after stores began opening.

Yet those increases did not show a significant association with proximity to cannabis retailers. Instead, the analysis identified warmer weather and pre-existing trends as relevant factors. Toronto already had a statistically significant upward crime trend before the stores opened.

Weather was particularly important. Heating degree days, a measure reflecting how cold conditions are, accounted for approximately 54 percent of the variation in the model. Crime tended to rise in warmer periods and fall in colder ones.

The practical lesson extends beyond cannabis. Imagine a store opening while crime is increasing across the surrounding city. A nearby increase cannot automatically be attributed to that business. The more useful question is whether its immediate surroundings changed differently from comparison areas after accounting for other influences.

Here, neither the estimated immediate change nor the subsequent trend change near stores was statistically significant. Alternative analyses also failed to detect significant opening-related changes.

Legal Retail Growth Does Not Necessarily Mean More Consumption

One explanation proposed in the paper is that early licensed stores primarily attracted people who already used cannabis. Customers may have switched from illicit dealers or unlicensed dispensaries to regulated retailers without substantially changing their consumption.

This mechanism was not directly tested, but it helps explain why store openings might produce substantial legal sales without a corresponding change in neighbourhood crime.

The distinction remains relevant as Canada’s retail market matures. MyCannabis coverage of new Canadian cannabis store openings illustrates continued retail expansion. However, adding storefronts does not reveal whether customers are new consumers, existing consumers changing suppliers, or shoppers moving between licensed competitors.

Separate MyCannabis reporting on cannabis demand after legalization reinforces the importance of distinguishing access, purchasing motivation, and consumption frequency. Those measures can follow different trajectories. More visible retail activity should not be treated as proof of greater intoxication or criminal behaviour.

Health Canada’s Canadian Cannabis Survey 2024 provides broader context: 72 percent of respondents who used cannabis reported usually obtaining it from a legal source, compared with 37 percent in 2019. That shift supports the plausibility of supplier substitution, although it cannot explain the crime patterns around these particular stores.

What Municipalities Can Learn About Cannabis Retail Policy

The strongest policy implication is that restrictions should match the problem they are intended to solve. If a municipality opposes cannabis retailers because of anticipated crime, it should examine evidence about crime rather than treating discomfort with cannabis as an equivalent measure.

That does not remove the need for oversight. It makes oversight more precise. Traffic, litter, public consumption, youth access, and criminal incidents are different outcomes requiring different information and responses.

Measure Specific Neighbourhood Outcomes

A practical local assessment could track:

• Changes in defined crime categories near stores and comparison areas.
• Documented complaints about public consumption, traffic, or nuisance.
• Retail compliance and access to authorized purchasing channels.

This approach gives residents a clearer basis for raising concerns and businesses a clearer understanding of expectations. A parking problem can be addressed through parking management. A compliance failure can be investigated through the relevant regulator. Neither automatically establishes that legal retail increases neighbourhood crime.

There is also a broader precedent for revisiting cannabis rules as experience accumulates. Health Canada’s 2025 cannabis regulatory changes adjusted requirements across areas including licensing and packaging. Federal production rules differ from municipal retail decisions, but both benefit from reviewing whether a requirement serves its intended purpose.

Why These Findings Cannot Settle Every Safety Question

The study’s limitations matter. Eight stores represent a small sample, and 11 months of operation cannot establish long-term effects. Ending the analysis in February 2020 avoided pandemic disruptions, but also excluded the much larger retail network that developed afterward.

Statistical uncertainty was substantial. For example, the estimated immediate change specific to nearby areas had a confidence interval extending roughly 14 percentage points in either direction. A non-significant result therefore does not prove the true effect was exactly zero.

The research also examined police-reported incidents rather than every crime committed, and it did not measure every community concern. Its findings cannot resolve questions about impaired driving, health emergencies, or public consumption.

Geography creates another boundary. Downtown Toronto and Ottawa are different from smaller communities with fewer businesses and different travel patterns. MyCannabis coverage of small-town cannabis retail offers useful context for those differences. The urban findings should inform rural discussions, while leaving room for local evidence.

A Better Way to Judge Cannabis Store Safety

The encouraging result is that early licensed stores did not produce a detectable crime increase in this analysis. That provides grounds for questioning predictions of inevitable neighbourhood harm, while preserving the need to monitor outcomes as markets evolve.

For retailers, the opportunity is to pair evidence with accountable operations and constructive community relationships. For municipalities, it is to establish measurable expectations and respond proportionately when problems occur.

Future research should examine more stores over longer periods and distinguish among neighbourhood types. Until then, this paper supports a careful conclusion: crime rose during the observation period, but the analysis did not connect that rise to licensed cannabis store openings.

Legalization debates become more productive when they ask which outcomes changed, where they changed, and what might explain them. That shift allows communities to protect public safety while giving regulated businesses a fair assessment based on observed results.

References:

¹ Armstrong, M. J. (2026). Crime levels near federally legal cannabis stores in two Canadian cities. International Journal of Drug Policy, 157, Article 105529. https://doi.org/10.1016/j.drugpo.2026.105529