Poster #142 - Does Direct Democracy Accelerate Policy Diffusion? Evidence from Recreational Marijuana Legalization

• In Event: Saturday Poster Luncheon

Abstract

Recreational marijuana legalization (RML) provides a useful case for examining how institutional design shapes policy diffusion under conditions of moral conflict, fiscal incentives, and federal-state tension. Although prior diffusion research has emphasized mechanisms such as learning, imitation, and geographic spillover, it has paid less attention to whether the pathway of adoption influences subsequent diffusion. This paper asks whether early RML adoptions achieved through direct democracy exert a stronger influence on later state adoptions than those achieved through legislative enactment, and whether diffusion is more likely in regional contexts where direct-democracy access is more prevalent.  

The paper develops a framework that treats direct democracy not simply as an institutional arrangement, but as a potential diffusion signal. Because ballot-based adoption may confer greater public legitimacy and visibility than legislative adoption, early direct-democracy adoptions may serve as more influential precedents for later adopters. At the same time, states located in regions with broader access to direct-democracy institutions could benefit from a political environment that is more conducive to policy innovation and reform.  

To evaluate these expectations, I construct a state-year panel covering 49 U.S. states (except Nebraska) from 2010 to 2024 and estimate discrete-time event history models of RML adoption. The analysis uses pathway-specific neighboring adoption measures to distinguish prior adoptions achieved through direct democracy from those achieved through legislative action. The models also incorporate regional institutional context, fiscal conditions, legislative composition, unified government, and annual time trends.  

Preliminary findings show limited support for the claim that direct-democracy-origin adoptions exert a stronger influence on later state adoption than legislative-origin adoptions. The results instead suggest that broader regional institutional context matters more than pathway-specific influence: states located in areas with greater access to direct-democracy institutions appear more likely to adopt RML, even though the hypothesized amplification effect is not statistically significant.  

These findings contribute to policy diffusion research by highlighting the importance of institutional opportunity structures in the spread of contentious state policies. For policymakers, the results suggest that the broader regional environment surrounding reform may matter more than whether early adopters used ballot initiatives or legislative enactment.

Author

• Nan Zhu, University of Illinois, Chicago
  Presenting Author