What Happens After Cannabis Legalization? A Review of 176 Studies Finds a Mixed Picture

Debates about cannabis legalization often ask for a single verdict: did it help or hurt? A new systematic review shows why the honest answer depends on which policy, population and outcome is being measured. Researchers synthesized 176 quasi-experimental reports published over three decades. The most consistent signals were more cannabis use among adults and young adults and more cannabis-related health care use. The picture for youth and for alcohol, tobacco and opioids was less uniform.

What the Review Actually Studied

The open-access review was published July 28 in The Milbank Quarterly and highlighted by McGill University on September 8. The authors searched five academic databases plus gray literature for studies evaluating a change in medical, recreational or decriminalization policy. They limited the synthesis to quasi-experimental designs intended to estimate a policy effect more credibly than a simple cross-sectional comparison.

The 176 included reports appeared from 1993 through 2024. About two-thirds were published after 2019, 148 were peer reviewed and 141 used U.S. data. Difference-in-differences was the most common method, used in 129 reports, while 42 used interrupted time-series designs.

Adult Cannabis Use Rose More Consistently Than Youth Use

Medical and recreational legalization may increase cannabis initiation across age groups, the review concluded. Increases in the prevalence of use were clearer among young adults and adults. Evidence among youth was weaker and inconsistent. That distinction matters because a headline saying legalization "increased use" can hide large differences by age.

The review also found evidence of increased cannabis-related health care utilization among adults, including emergency department visits and poison center reports. Those outcomes are not interchangeable with the number of people who use cannabis. They measure a more serious subset of events, and increases can reflect changes in products, availability, reporting or care-seeking as well as policy itself.

The weaker youth signal does not make youth exposure irrelevant. Product potency, advertising and frequency still matter, as our reports on teen cannabis use and cognitive development and cannabis advertising safeguards explain.

Alcohol Effects Depended on the Policy and Population

Alcohol did not move in one direction. Medical cannabis legalization was associated in some studies with reduced alcohol use among youth and adults, which is consistent with substitution. Recreational legalization produced a more complicated pattern. Among young adults, it was linked in the review to increased alcohol use and binge drinking, a pattern more consistent with co-use.

Our separate review of THC beverages and alcohol substitution reaches a similar caution: substitution is possible in some settings, but mixing and complementary use remain part of the public-health picture.

That split is a warning against treating all legalization policies as the same exposure. A medical registry, possession reform and a commercial adult-use market change different things. Age limits, store density, product rules, advertising and price can also shape what happens after the law changes.

Tobacco and Opioid Findings Were Promising but Not a Cure Claim

A limited group of studies suggested tobacco use stayed the same or decreased across age groups after cannabis policy changes. Most opioid studies found no change or decreases in outcomes such as prescriptions, sales and health care visits. Evidence for opioid mortality was less consistent.

Those population signals do not establish cannabis as a treatment for nicotine dependence or opioid use disorder. A state-level decline can have many contributors, and the review evaluated policy effects rather than assigning individuals to use cannabis. Our opioid-withdrawal evidence review explains why early cannabis signals should not displace evidence-based medications for opioid use disorder.

Why 176 Studies Still Do Not Produce a Universal Answer

Quasi-experimental research uses policy timing and comparison groups to approximate a natural experiment. It can be far more informative than asking whether two things happened at the same time, but it remains observational. States that legalize can differ from states that do not. Policies are implemented unevenly, outcomes are defined differently and several reforms may occur close together.

The authors therefore synthesized results descriptively rather than combining every outcome into one pooled number. They also found few studies of decriminalization and little clear evidence about changes involving other illicit drugs. Newer product markets and policy changes after 2024 are outside this evidence set.

What Readers and Policymakers Can Take From It

The review does not validate the most sweeping claims from either side. It supports planning for increased adult use and cannabis-related health care demand after legalization. It also suggests that youth effects may be smaller than adult effects and that spillovers to alcohol, tobacco and opioids depend on the form of reform and the population studied.

For future evaluations, the useful question is not simply whether a state legalized. It is what the state legalized, how the market was designed, when retail access began, which products became common and which group experienced the outcome. That is where a mixed evidence base becomes actionable.