Property crime fell sharply in Massachusetts communities that were among the first in the state to open recreational marijuana stores, while violent crime also declined at roughly the same rate as Massachusetts overall, according to an analysis of federal crime records and state licensing data.
The analysis examined 18 municipalities that opened recreational marijuana retailers during Massachusetts’ first year of legal sales and for which complete, directly comparable federal crime data were available for both 2017, the last full year before recreational sales began, and 2025, the most recent full year available.
Across those communities, the population-weighted average property crime rate fell 34.9%, from approximately 1,798 offenses per 100,000 residents in 2017 to 1,171 per 100,000 in 2025. The calculation used each municipality’s 2017 population as a fixed weight in both years.
Property crime declined in all 18 communities examined.
The decline was greater than the statewide decrease over the same period. Massachusetts’ property crime rate fell from about 1,428 per 100,000 residents in 2017 to 1,019 in 2025, a decrease of approximately 28.6%.
Violent crime also declined among the early-retail communities, though the change closely tracked the statewide trend. The population-weighted average violent-crime rate for the 18 communities fell 17.2%, from approximately 546 violent crimes per 100,000 residents to 452.
Massachusetts as a whole experienced a 16.4% decline, from about 354 violent crimes per 100,000 residents in 2017 to 296 in 2025.
The findings do not establish that marijuana stores caused crime to fall. Store locations were not randomly assigned, communities differ substantially in population and economic conditions, and numerous factors unrelated to marijuana policy can influence crime over an eight-year period.
What the results do show is that communities that were among the earliest to host recreational marijuana retailers did not experience a broad increase in violent or property crime relative to their pre-sales levels, despite concerns frequently raised during legalization debates that retail marijuana businesses could contribute to crime.
Property Crime Declined in Every Community Examined
The Marijuana Herald identified municipalities with stores operating during the first year of Massachusetts’ adult-use market using Cannabis Control Commission licensing records.
Legal recreational sales began November 20, 2018. By November 2019, 33 marijuana retailers had opened statewide, according to the Cannabis Control Commission.
The 18 early-retail municipalities with complete comparable 2017 and 2025 federal crime records included Brockton, Brookline, Chicopee, Easthampton, Fall River, Gardner, Greenfield, Hudson, Lowell, Lynn, Marlborough, Nantucket, Newton, Northampton, Pittsfield, Salem, Somerset and Worcester.
The size of the property-crime decreases varied considerably.
Pittsfield’s property crime rate fell by approximately 62.7%, while Easthampton saw a roughly 54.6% decline and Gardner saw a decrease of approximately 53.8%.
Nantucket and Somerset each recorded declines of more than 50%.
Even the smallest decline among the 18 communities was a decrease: Fall River’s property crime rate was approximately 11.8% lower in 2025 than in 2017.
The FBI’s 2017 city-level records show substantial differences among the communities before recreational sales began. For example, Pittsfield recorded 1,092 property crimes among a population of 42,546 in 2017, while Salem recorded 951 among 43,385 residents. Northampton reported 555 property crimes among 28,458 residents.
Those differences are one reason the analysis focused on changes within each community rather than simply comparing which municipalities had the highest or lowest crime rates.
Violent Crime Picture Is More Mixed
The violent-crime results were less uniform.
Of the 18 communities, 12 had lower violent-crime rates in 2025 than in 2017, while six had increases.
Among the largest declines, Salem’s reported violent-crime rate dropped about 74%, Nantucket’s fell approximately 57%, and Easthampton’s declined roughly 56%.
Brockton, which recorded 917 violent crimes among 96,016 residents in 2017, had a violent-crime rate approximately 44% lower by 2025.
Pittsfield declined by roughly 35%, Greenfield by 26%, Gardner by 24%, Fall River by 22%, Lynn by 22% and Worcester by 17%.
Other communities moved in the opposite direction. Hudson had an unusually low 2017 baseline of just seven violent crimes among nearly 20,000 residents, making its percentage increase particularly large. Lowell, Brookline and Marlborough also recorded higher rates in 2025 than in 2017.
Taken together and weighted by the communities’ baseline populations, however, the overall violent-crime rate declined 17.2%.
That is almost identical to Massachusetts’ statewide 16.4% decline, suggesting the early marijuana-retail communities largely followed the broader statewide violent-crime trajectory rather than diverging sharply from it.
Why 2025 Matters
The comparison uses 2025 because it is the newest complete annual federal crime dataset available.
The FBI released its 2025 Reported Crimes in the Nation statistics in August, based on submissions from more than 17,000 law enforcement agencies covering 96.2% of the U.S. population. The agency says its 2025 city crime tables include agencies that submitted 12 months of complete offense data.
For Massachusetts, FBI-derived 2025 data show approximately 21,092 violent crimes, or 296.2 per 100,000 residents, and 72,594 property crimes, or 1,019.3 per 100,000.
The Marijuana Herald excluded early-retail municipalities where a clean 2017-to-2025 comparison could not be established rather than filling gaps with estimates or substituting a different year.
The FBI defines violent crime as murder and nonnegligent manslaughter, rape, robbery and aggravated assault. Property crime consists of burglary, larceny-theft and motor vehicle theft.
What the Analysis Can and Cannot Show
The results are an association, not proof of causation.
Massachusetts municipalities made their own decisions about whether to accommodate marijuana businesses, and retailers themselves selected particular markets. Communities willing to host marijuana stores may differ from communities that did not in ways that also affect crime.
The analysis also does not claim that crime fell specifically around dispensary locations. It measures municipal crime rates, meaning an incident on the opposite side of a city is counted the same as one occurring next to a marijuana retailer.
Population, policing practices, economic conditions, the COVID-19 pandemic and changes in crime reporting can also affect long-term comparisons.
Still, the consistency of the property-crime result is notable: every one of the 18 early-retail communities with complete comparable federal data had less property crime in 2025 than before recreational sales began.
And on a population-weighted basis, neither of the two major crime categories moved in the direction predicted by warnings that legal marijuana storefronts would broadly make their host communities less safe.
Instead, property crime declined substantially more than it did statewide, while violent crime fell almost exactly in line with the Massachusetts trend.
Methodology
The Marijuana Herald identified Massachusetts municipalities with recreational marijuana retailers operating during the first year of adult-use sales using Cannabis Control Commission licensing and commencement records. The Commission reports that 33 retailers were operating by the end of the market’s first year.
Crime figures were drawn from the Federal Bureau of Investigation’s Uniform Crime Reporting data. The analysis used 2017 as the baseline, because it was the final full calendar year before Massachusetts recreational marijuana sales began in November 2018, and 2025 as the endpoint, the latest complete annual dataset.
Municipalities without complete comparable data in both years were excluded. Rates were calculated per 100,000 residents. To calculate the combined change across the 18 municipalities, each municipality’s crime rate was weighted using its 2017 population, with the same fixed population weights applied to both years. This prevents changes in municipal population over time from altering each community’s relative influence on the comparison.
The analysis excludes marijuana possession and other drug offenses because legalization itself changes which marijuana activity constitutes a crime.
The FBI cautions that Uniform Crime Reporting statistics measure offenses reported to and recorded by law enforcement and should not, by themselves, be used as rankings of community safety.
