2026-10-16T09:42:28Z https://data.mendeley.com/oai

oai:data.mendeley.com/ng9pb44nnd.3 2026-01-28T12:58:58Z

Harrison, Sharon How Does Marijuana Legislation Affect Crime? Medical and Recreational Laws Across 50 States Mendeley Data Novel policies often have unintended consequences—programs that legalize one crime may affect other crimes. Marijuana legalization is an example of such spillovers, as states have legalized use in a staggered fashion. While previous research focuses on a few states or one type of legalization, we consider all 50 states and both medical and recreational marijuana. Using a difference in differences approach we find distinct im- pacts from medical and recreational legalization, largely borne out by dynamic analysis. In our main synthetic difference in differences model, we find that medical legalization reduces property crime, while recreational legalization reduces violent crime. Our re- sults are consistent with a channel where legalization makes other crimes costly. Thus, policymakers should wait to understand cost-benefit impacts, focus on the specific case, and learn from outcomes in similar states. We had to compile our data set from a variety of sources. These data were sometimes reported at different frequencies, and made available at different periods. We therefore ultimately use data for a sample period where we could collect all of the variables and controls. The main data on crime rates come from FBI Uniformed Crime Reports (UCR) arrest data from 1995 to 2019, while the data on recreational and medical use laws come from various state level sources. Our analysis involves both violent crime and property crime. Violent crime is a summary of the following four individual crimes: aggravated assault, murder, rape, and robbery. Similarly, property crime combines the following crimes: larceny, burglary, and motor theft. In addition to these data, we utilize a host of state-level controls including: a dummy variable for political party of the governor; a dummy variable for political party in control of the state senate; poverty rate; unemployment rate; population density; aver- age number of neighboring states that have legalized recreational and medical marijuana; cigarette tax rate; liquor-related arrests; number of police per capita; and real income per capita. We select the data for the controls based on discussions in previous research such as Pacula and Kilmer (2003) and Dragone et al. (2019). These data are from the BLS, BEA Census Bureau, and other sources. We also include state and year fixed effects and use robust clustered standard errors. More information below in the readme file. Health Economics Policy Cause of Crime Cannabis Chollete, Loran Lu, Ching-Chih Mamun, Khawaja Dataset doi:10.17632/ng9pb44nnd.3 oai:data.mendeley.com/ng9pb44nnd.3 Creative Commons Attribution 4.0 International http://creativecommons.org/licenses/by/4.0 https://data.mendeley.com/datasets/ng9pb44nnd 2026-01-28T12:58:58Z