Association for Public Policy Analysis & Management

Background: To date, 23 states plus the District of Columbia have implemented a medical marijuana law (MML). Opponents of MMLs fear that opening storefront dispensaries and allowing patients to cultivate marijuana at home may create prime targets for individuals looking for cash or drugs, resulting in civil disruptions and increased criminal activity.

Objective: This study aims to determine whether MMLs impact crime by evaluating their impact on arrest rates for a number of violent and non-violent crimes. Specifically, the analysis considers whether specific characteristics of MMLs (i.e. storefront dispensaries, in-home cultivation, and required registration of medical marijuana patients) have a differential effect on arrest rates.

Approach: Using state-level data from the 1994-2012 Uniform Crime Reports, this study uses fixed effects ordinary least squares regression models (i.e. difference-in-difference) to estimate the effect of MMLs on arrest rates.  Arrest rates for a total of 15 property, violent, and non-violent crimes are considered (e.g., robbery, theft, burglary, driving under the influence, among others).  These estimates rely on the natural variation in state MML implementation over time and also control for economic and demographic heterogeneity at the state level. The models are also run controlling for the presence of dispensaries, the allowance of in-home cultivation, and the implementation of a required registry for marijuana users.  Given that MML rules and criminal justice policies differ for youth versus adults, I estimate the effect of MML characteristics on arrest rates separately for juvenile (aged 10-17) and adult (aged 18 and older) populations. 

Results: Preliminary results suggest MMLs lead to a significant decrease in arrest rates for violent crimes among both juveniles and adults.  Initial estimates also point to a reduction in arrest rates for property crime, which is likely driven by a decrease in burglary and theft arrests among juveniles.  Adults living in states with MMLs experience a significant decline in drug abuse violation arrests. Potential mechanisms to explain the decline in arrest rates include increased security at dispensaries and homes, the decreased level of alcohol consumption that accompanies the implementation of MMLs, and the role of law enforcement. 

Conclusions: The results of this study point to an overall reduction in arrest rates as a result of MMLs.  As more states consider medical marijuana laws and marijuana legalization, this study provides estimates of the potential impact of increasing access to marijuana on criminal activity.