Impacts of Medical Marijuana Policies on Initiation of Marijuana Use: Evidence from National Longitudinal Surveys of Youth 1997

Purpose: A major concern about medical marijuana (MM) policy is its impacts on the initiation of marijuana use, particularly among adolescents and young adults. Previous research typically relied on static marijuana use prevalence estimates from repeated cross-sectional surveys. From a life-course perspective, we instead focused on initiation decisions and examined the associations between MM legalization and the age of onset of marijuana use and regular use. We also evaluated the heterogeneities of the policy impacts by states’ policy durations and individuals’ interstate mobility status.

Methods: This study utilized National Longitudinal Surveys of Youth 1997 (NLSY97) 1997-2011 data. NLSY sampled a nationally representative cohort of adolescents who were born between 1980 and 1984 (N=8984), and surveyed them annually since 1997 baseline to track their transitions from adolescence to young adulthood. The enactment of MM policies were treated as a series of quasi-experiments, in which 17 states that implemented MM policies were “treatment states” and 34 states where the MM policy status was unchanged were “control states”. The difference-in-difference approach in a duration framework were used to compare hazards of marijuana use initiation before and after MM policy implementation in treatment states, controlling for contemporaneous changes in the control states. The outcomes were reported ages of marijuana use onset and regular use onset. We included a wide range of time-invariant and time-variant individual characteristics, state-level time-varying economic conditions and other policies, as well as state-specific effects and time-specific effects.

Results: Overall, 61.46% individuals out of the 8984 adults had ever used marijuana and 14.36% baseline never users had initiated regular marijuana use by their last interview. The average ages of initiating marijuana use and regular use were 16.98 and 20.72, respectively. The implementation of MM policies was associated with 19% lower rate (p<.01) of marijuana use uptake at a given age. There was no evidence that MM legalization affected regular marijuana use uptake. Findings were robust to sensitivity analyses. In terms of policy durations, the impacts of policy implementation remained significant and negative on initiation of marijuana use after 0-2 years, 2-5 years, as well as 5-9 years (HR=.72, p<.01, HR=.88, p=.042, and HR=.73, p<.01, respectively). However, having legalized MM for more than 9 years was associated with 31% higher rate of marijuana use uptake (p<.01). In terms of interstate mobility, the implementation of MM policies was associated with 23% lower hazard of marijuana use uptake (p<.01) among stayers, who never moved across states while being at risk. The impact of MM policies was insignificant among movers, who ever had interstate migration while being at risk. There was no evidence showing the differential policy impacts in stayers and movers for regular use uptake.

Conclusions: Overall, our findings did not lend support to the concern that MM policies lead to increased marijuana initiation. Instead, we found that implementing MM policies significantly reduced the hazard rates of marijuana use initiation and the magnitude of the reduction was substantial. The effects were heterogeneous in terms of states’ policy durations and individuals’ interstate mobility status.