The effect of place on voting behavior: The case of the Arizona proposition to legalize recreational marijuana

Abstract

In 2020, Arizonans approved Proposition 207, the Smart and Safe Arizona Act, which legalized recreational marijuana sales. Previous research has typically used non‐spatial survey data to understand marijuana legalization voting patterns. However, voting behavior can, in part, be shaped by geographic context, or place, which is unaccounted for in aspatial survey data. We use multiscale geographically weighted regression to analyze how place shaped Proposition 207 voting behavior, independently of demographic variations across space. We find significant spatial variability in the sensitivity of voting for Proposition 207 to changes in several of the predictor variables of opposition and support for recreational marijuana legalization. We argue that local statistical modeling approaches provide a more in‐depth understanding of ballot measure voting behavior than the current use of global models.Related ArticlesBranton, Regina, and Ronald J. McGauvran. 2018. “Mary Jane Rocks the Vote: The Impact of Climate Context on Support for Cannabis Initiatives.” Politics & Policy 46(2): 209–32. https://doi.org/10.1111/polp.12248.Brekken, Katheryn C., and Vanessa M. Fenley. 2020. “Part of the Narrative: Generic News Frames in the U.S. Recreational Marijuana Policy Subsystem.” Politics & Policy 49(1): 6–32. https://doi.org/10.1111/polp.12388.Fisk, Jonathan M., Joseph A. Vonasek, and Elvis Davis. 2018. “‘Pot'reneurial Politics: The Budgetary Highs and Lows of Recreational Marijuana Policy Innovation.” Politics & Policy 46(2): 189–208. https://doi.org/10.1111/polp.12246.

Summary of articles in the February 2024 issue of Politics & Policy and a few editor's tips

“…Contributing strongly to P&P's growing repository of articles on the legalization of recreational marijuana in the United States (see, e.g., Branton & McGauvran, 2018;Brekken & Fenley, 2020;Fisk et al, 2018), Mathews and Fotheringham (2024) examine how local context adds nuance when accounting for local variations in voting behavior. They show that spatial modeling assists analyses of the covariate effects on ballot measure voting behavior in some rather significant ways.…”