States that launched legal recreational marijuana sales experienced an approximately 9% increase in agricultural employment, according to a study prepared for presentation at the 2026 Agricultural and Applied Economics Association Annual Meeting.
Researchers from Texas Tech University examined employment and wage data from 45 states between 1990 and 2024, comparing states that opened recreational marijuana markets with states that had not done so by the end of the study period.
The researchers used data from the U.S. Bureau of Labor Statistics’ Quarterly Census of Employment and Wages, which covers approximately 95% of U.S. jobs. Four outcomes were examined: agricultural employment, agricultural wages, employment across all industries and wages across all industries.
After accounting for differences between states and the timing of retail marijuana sales, the study found that legalization was associated with a statistically significant increase of roughly 9% in agricultural employment.
However, researchers found no statistically significant effect on agricultural wages, overall employment or wages across the broader economy.
The authors said the increase in agricultural jobs without a corresponding rise in wages may indicate that employers were able to draw from a relatively flexible labor pool, including seasonal workers, underemployed rural residents or people moving from other lower-wage industries.
Results varied considerably depending on when states began recreational marijuana sales.
The strongest increase was found among the 2017 cohort, consisting of Nevada, where agricultural employment rose by an estimated 46%. The 2019 cohort, consisting of Michigan, experienced an estimated 5.3% increase.
By contrast, the 2021 cohort, consisting of Arizona, saw an estimated 6.9% reduction in agricultural employment. Several other legalization cohorts recorded no statistically significant change.
The study concludes by stating:
This study examines the labor market effects of the policy using a Synthetic Difference in-Differences framework. The results demonstrate that the policy significantly increased agricultural employment by approximately 9 percent in treated counties, while having no discernible impact on agricultural wages, all-industry employment, or economy-wide wages. The cohort-level analysis reveals substantial heterogeneity, with the employment gains concentrated among the 2017 and 2019 adoption cohorts. The null wage response, despite significant employment growth, points to an elastic agricultural labor supply in treated areas. These findings suggest that while the policy successfully expanded agricultural employment, its broader labor market spillovers remain limited. Policymakers should consider adoption timing and local labor market conditions when designing similar interventions, as the effectiveness of the policy varies considerably across cohorts.
