To examine the influence of political economy and sociodemographic factors on the probability that a state decriminalizes marijuana in a given year. Prior literature has looked at the factors affecting the adoption of state medical marijuana statutes, but not decriminalization. Marijuana has been effectively prohibited by federal law in the United States since the Marijuana Tax Act of 1937. Furthermore, with the passage of the Controlled Substances Act (CSA) of 1970, marijuana was classified as a Schedule I drug, among those considered to have the “highest potential for abuse and no accepted medical use.” However, state-level resistance to this classification of the drug came to fruition in 1973 as Oregon became the first state to "decriminalize" marijuana, which is to remove criminal penalties for possession of small, personal-use amounts of the drug. Alaska, Colorado, California, and Ohio followed suit in 1975, and as of now, 21 states have decriminalized, the most recent being Delaware in 2015.
Data/Methods
To build our database, we utilize data from a variety of sources, including the Census Bureau, the Bureau of Economic Analysis, the House of Representatives, the "My Campaigns" website, Dave Leip's Election Atlas, the Marijuana Policy Project, and the National Organization for the Reform of Marijuana Laws (NORML). We further plan to add data on state citizen and government ideologies from Richard Fording and state economic and election data from Carl Klarner, and to supplement missing values of Census data with weighted state-level means from supplements of the Current Population Survey.
We will use event history analysis (specifically, conditional logit models since the failure time data are discrete) to estimate the probability that a state decriminalizes based on the factors mentioned above, as well as policy diffusion factors such as the laws in border states and ideological differences with adopting states. We will adjust for the spatial correlation in laws among states using various methods recommended in the literature.
Results/Expected Results
The results, which are preliminary at this stage, suggest that states with a higher percentage of bachelor’s degree holders, higher unemployment rates, and higher percentages of residents under 20 are more likely to decriminalize, while states with higher percentages of Republican votes in federal and gubernatorial elections are less likely to do so. Surprisingly, policy diffusion effects do not appear influential, as the percentage of bordering states that have decriminalized has an insignificant effect on the likelihood that a state decriminalizes.