Association of Marijuana Legalization With Marijuana Use Among US High School Students, 1993-2019

Thirty-six states have legalized medical marijuana and 18 states have passed recreational marijuana laws (RMLs). Organizations such as the American Academy of Pediatrics are concerned that legalization will encourage youth marijuana use.¹ Marijuana use during adolescence may adversely affect areas of the prefrontal cortex, which control important cognitive processes.¹

Using data from the Youth Risk Behavior Survey (YRBS) for the period 1993-2017, Anderson et al² found that RML adoption was associated with an 8% decrease in the odds of marijuana use among high school students. These authors, however, had prelegalization and postlegalization data from only 7 states and pre– and post–recreational sales data from only 3 states, calling into question the generalizability of their results.³

Using data from the YRBS for the period 1993-2019, this study provides updated estimates of the association between legalization and adolescent marijuana use. During this extended period, pre- and post-RML data from the YRBS are available from 10 states; 7 states contributed more than one wave of post-RML data, and these same 7 states contributed data to the YRBS before and after the first dispensary sales began. Anderson et al³ had multiple waves of post-RML data from 3 states.

Following Anderson et al,² national and state YRBS data were pooled in this repeated cross-sectional study. Ethical review was not required because analyses of secondary, deidentified data are considered exempt from requiring institutional review board approval by the San Diego State University Institutional Review Board. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.

Multivariable logistic regression was used to estimate the association between legalization and marijuana use. Effective legalization dates come from Anderson and Rees.⁴ To control for time-invariant factors at the state level and common trends, all models were adjusted for 50 state and 13 survey wave indicators. Alternative models were further adjusted for individual- and state-level characteristics; event-study estimates were produced by replacing the RML indicator with a series of leads and lags. Two-sided hypothesis tests were used, and estimates were considered significant if P < .05. Analyses were conducted with STATA 16.1 (StataCorp).

The mean (SD) age of YRBS respondents was 15.9 (1.23) years, 51.4% of respondents reported as female, and 57.6% reported as non-Hispanic white. The first two columns of the Table show estimated odds ratios (ORs) of current and frequent marijuana use, adjusted for state and survey-wave indicators. In the third and fourth columns, ORs were further adjusted for individual- and state-level covariates.

Based on the pooled YRBS data, and in the fully adjusted models, RML adoption was not associated with current marijuana use (OR, 1.00; 95% CI, 0.92-1.10) or frequent marijuana use (OR, 0.98; 95% CI, 0.90-1.07). In the fully adjusted models, medical marijuana law (MML) adoption was associated with a 6% decrease (OR, 0.94; 95% CI, 0.89-0.98) in the odds of current marijuana use and a 7% decrease (OR, 0.93; 95% CI, 0.87-0.99) in the odds of frequent marijuana use. Estimates from separate analyses of the national and state YRBS and estimates of the association between the opening of the first recreational dispensary and marijuana use were qualitatively similar to those above (Table).⁵

Finally, the Figure shows RML event-study estimates (Panel A). Prior to legalization, there was no association with marijuana use, suggesting the parallel-trends assumption held. After 2 or more years, RML adoption was associated with a decrease in marijuana use (OR, 0.85; 95% CI, 0.76, 0.95). Panel B of the Figure shows an event study for MML adoption.

Consistent with estimates from prior studies, there was little evidence that RMLs or MMLs encourage youth marijuana use.^2,6 Contrary to results of the study by Anderson et al² the overall association between RML adoption and marijuana use among adolescents was statistically indistinguishable from zero. One limitation of this study is that RMLs are a relatively new phenomenon. As more postlegalization data become available, researchers will be able to draw firmer conclusions about the relationship between RMLs and adolescent marijuana use.

Accepted for Publication: July 8, 2021.

Published: September 7, 2021. doi:10.1001/jamanetworkopen.2021.24638

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Anderson DM et al. JAMA Network Open.

Corresponding Author: D. Mark Anderson, PhD, Department of Agricultural Economics and Economics, Montana State University, PO Box 172920, Bozeman, MT 59717-2920 (dwight.anderson@montana.edu).

Author Contributions: Dr Sabia had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: All authors.

Acquisition, analysis, or interpretation of data: Anderson, Sabia, Safford.

Drafting of the manuscript: Anderson, Rees, Sabia.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Rees, Sabia, Safford.

Obtained funding: Anderson, Sabia.

Supervision: Anderson, Sabia.

Conflict of Interest Disclosures: Dr Sabia reported grants from Troesh Family Foundation and grants from Charles Koch Foundation during the conduct of the study. No other disclosures were reported.

Funding/Support: Dr Anderson acknowledges partial support from a Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health (NIH) research infrastructure grant, R24 HD042828, to the Center for Studies in Demography and Ecology at the University of Washington. Dr Sabia and Mr Safford acknowledge partial support from the Center for Health Economics & Policy Studies (CHEPS) at San Diego State University, including grant funding received from the Charles Koch Foundation.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.