Youth Cannabis Use Did Not Rise After Legalization in a Corrected Analysis of U.S. Survey Data
A corrected analysis of federal Youth Risk Behavior Survey data found no statistically significant increase in cannabis use among U.S. high school students after states adopted medical or adult-use marijuana laws.
The study, published in JAMA Network Open in 2021 and replaced with a corrected version in March 2022, examined national survey data from 1993 through 2019. Researchers analyzed responses from 191,923 students in the national survey and more than 1.4 million respondents in separate state-level surveys.
Rather than combining the national and state datasets, the researchers analyzed them separately and applied survey weights intended to make the results representative of the populations surveyed. They examined two measures of cannabis use: any use during the previous 30 days and frequent use, defined as using cannabis at least 10 times during that period.
In the fully adjusted national analysis, adult-use legalization was not associated with current cannabis use or frequent use. The estimated odds ratio was 1.01 for current use, with a 95% confidence interval of 0.85 to 1.19, and 0.99 for frequent use, with a confidence interval of 0.78 to 1.26. Because both intervals included the possibility of no change, the results do not support the conclusion that legalization increased adolescent use.
Medical cannabis laws also were not significantly associated with either outcome. The estimates pointed toward lower use—an odds ratio of 0.94 for current use and 0.90 for frequent use—but the confidence intervals included no effect. Those figures therefore should not be presented as evidence that medical legalization caused teen use to fall.
The researchers found similar results in their separate analysis of state YRBS data. A secondary analysis that examined the timing of recreational retail sales produced evidence of lower use in some state data, but the authors emphasized that the overall findings did not establish a consistent effect of legalization on adolescent consumption.
The study was observational, meaning it identified associations in survey data rather than testing whether legalization directly caused changes in behavior. It also had important limits: only 10 states supplied both pre- and post-adult-use legalization data, and the analysis ended in 2019. The article’s online record notes that the original version was retracted and replaced on March 8, 2022, to correct errors in the analysis methods and results.
National surveillance data provide broader context. The Centers for Disease Control and Prevention’s 2019 Youth Risk Behavior Survey report found that 21.7% of high school students reported current marijuana use and 36.8% reported lifetime use. Current use showed no statistically significant change from 2009 to 2019, while lifetime use rose through 2013 and then declined through 2019.
Additional research has continued to produce nuanced results. A 2024 JAMA Pediatrics analysis of state survey data through 2021 likewise found no association between recreational legalization and adolescents’ likelihood or frequency of cannabis use, although some measures indicated lower overall use after legalization. The authors also noted that the COVID-19 pandemic and differences in state policy and retail implementation make longer-term evaluation important.
Taken together, these studies do not show that legalization automatically increases cannabis use among teenagers. They also do not demonstrate that legalization reduces youth use everywhere. The most defensible conclusion is narrower: in the survey data examined, adolescent cannabis use did not rise in a statistically detectable way after states adopted medical or adult-use legalization, while the effects of specific regulations and longer-term policy changes remain subjects for continued research.
