Since California became the first state to legalize the cultivation, possession, and use of marijuana for medical purposes in 1996, 20 states along with the District of Columbia have enacted medical marijuana law (MML). In 2012, Washington and Colorado legalized marijuana for recreational use, while MML or legalization of marijuana for medical purposes has continued to be debated in several states.

This study considers a number of externalities related to MML, such as hard drug use and non-drug related crime. To examine externalities related to MML, I begin by investigating empirically whether or not MML affects illegal or recreational marijuana use. If MML may cause people who aren’t prescribed medical marijuana to interact with legitimate patients, there would be more opportunities to smoke illegal marijuana. To avoid systematic reporting errors from self-reported surveys of illicit-drug use, I use illegal marijuana possession arrests as a proxy for illegal marijuana consumption, using Uniform Crime Report (UCR) data (Chu, 2013). Using Diff-in-Diff estimation with UCR to cover the years from 1994 to 2011, I compare the number of illegal marijuana possession arrests in treatment states, where MML was enacted, to control states without MML. I find that the number of illegal marijuana possession arrests increases by 6-10 percent in treatment states compared with control states.

             I turn to externalities in relation to MML, such as hard drug and non-drug related crimes. First, I use UCR and National Incident-Based Reporting System (NIBRS) data to analyze the effect of MML on illegal hard drug possession arrests. Opponents of MML have argued that a more lenient policy on marijuana, such as decriminalization or enactment of MML, will increase the demand for more dangerous drugs, demonstrating the gateway effect. To test the gateway effect, using Diff-in-Diff estimation, I show that the number of other hard drug possession arrests decreases or do not increase. Secondly, I examine whether the adoption of MML is associated with non-drug related crimes, such as violent, and property crimes. According to Goldstein (1985), there are several ways to affect violent or property crimes: (1) “Psychoparmacological violence”; (2) “Economic compulsive violence”; and (3) “Systematic violence.” Taken altogether, since marijuana is believed to inhibit aggressiveness and violence, the number of violent crimes can be affected by the gateway effect rather than by marijuana consumption itself. Despite considering the gateway effect, since enactment of MML does not cause an increase in consumption of alcohol or other drugs, I find that marijuana has little effect on violent and property crimes with the exception of larceny.

This study contributes to the existing literature which investigates MML. To my best knowledge, no previous study has examined the effect of MML on crimes, even though several studies have considered the relationship between MML and illegal marijuana use or hard drug use. This paper shows that despite the increase in illegal marijuana use after enacting MML, MML does not lead to an increase in hard drug use or non-drug related crimes.