One model of distribution that has proliferated in recent years is the medical cannabis dispensary

California legalized medical cannabis through a 1996 ballot initiative known as Proposition 215 . Prop 215 protects physicians who recommend medical cannabis as well as those qualified patients and caregivers who cultivate, possess, and use cannabis with a valid recommendation. The word “dispensary”, however, does not appear in the law. To this day, dispensaries in California hold a precarious legal existence. The uncertainty created by federal cannabis prohibition is major contributing factor. Although Prop 215 extends its protections to designated individuals who may cultivate cannabis on behalf of qualified patients, the original law is silent as to how medical cannabis might be distributed outside of the patient-caregiver relationship. In 2003 the California legislature passed Senate Bill 420, the “Medical Marijuana Program Act”, which established an optional statewide registry system for patients and caregivers. Although S.B. 420 was intended to clarify the provisions of Prop 215, it largely ignored the question of distribution. Beyond the statutory restrictions of Prop 215 and S.B. 420, the State of California imposes very few regulations on medical cannabis. Jerry Brown, then Attorney General of California, issued a directive to law enforcement in 2008 regarding the state’s medical cannabis laws. He noted that “dispensaries” are not defined by law but posited that storefront dispensaries that are properly organized as cooperatives or collectives may be legal if they meet certain requirements .

He listed requirements similar to those already found in some municipal ordinances, including: non-profit status, how to cure cannabis restriction of membership to qualified patients and caregivers , and prohibition of sales to non-members. The California State Board of Equalization has issued seller’s permits to dispensaries since a 2005 ruling that sales tax must be assessed on all medical cannabis transactions, regardless of their legality . In 2010 Governor Arnold Schwarzenegger signed A.B. 2650, prohibiting dispensaries within 600 feet of schools. Beyond these scant regulations at the statewide level, responsibility for MCDs has largely been left to the discretion of municipal governments. This study examines the effect on crime of 26 medical cannabis dispensaries operating in San Francisco during the year 2010. San Francisco was one of the first municipalities to formally authorize MCDs, and I argue that it does so effectively—compared to other jurisdictions—by setting up a permitting process for MCDs and placing reasonable restrictions on their location and operation. Thus San Francisco represents an important manifestation of this study’s primary independent variable, which is not medical cannabis per se but the set of policies governing medical cannabis dispensaries. Specifically, San Francisco presents an opportunity to test the effect of locally regulated MCDs on surrounding neighborhoods. Crime data for this study were obtained from the San Francisco Police Department . A formal request was made under the California Public Records Act for lists of serious crimes at the citywide level for the year 2010 including date, category of crime, and location by city block. Lists were also compiled of the addresses of all MCDs operating in the city in 2010, using data from the Public Health Department.

Borrowing from a research design presented by Williams, Freisthler, and Sims , all data were aggregated to census tract boundaries. Census tracts are convenient units of analysis because they have fairly consistent populations and their boundaries tend to align with the physical environment. Regression analyses are conducted to test the criminogenic effect of locally regulated MCDs, controlling for socioeconomic disadvantage, family disruption, and residential instability. Findings indicate a weak but statistically significant relationship between MCD density and crime. It appears that, in San Francisco, neighborhoods containing MCDs have slightly higher rates of crime than neighborhoods that do not contain MCDs. Due to the limited number of cases—26 dispensaries across 16 census tracts, compared to 173 non-MCD-containing tracts—I am cautious about speculating as to the causal nature of this link. Stronger links are found between crime and some of the other variables examined by this study, which are derived from the criminological literature on social disorganization theory. According to social disorganization theory as put forward by Sampson and Groves , high crime rates can be explained by the presence of certain “exogenous sources of social disorganization” including socioeconomic disadvantage, family disruption, and residential instability. Measures of each of these concepts are significantly associated with crime, and with much greater explanatory power than is found for MCD density. Based on these findings, I conclude that prohibitory MCD enforcement policies are unlikely to be an effective use of scarce law enforcement resources, insofar as such policies are rooted in the assumption that MCDs attract crime.

In chapter II, I review the existing literature on cannabis, MCDs, and crime, and discuss important theoretical and empirical issues that arise in spatial analyses of this sort. Routine activities theory provides a useful lens through which to analyze competing claims about the potentially criminogenic effects of certain land uses . In chapter III, I present a conceptual model for explaining the relationship between crime, MCDs, and selected “exogenous sources of social disorganization”. The research design employed by this study is described in greater detail in chapter IV along with a presentation of results and a discussion of findings. I conclude in chapter V by discussing some important implications for policymakers.The State of California has recognized cannabis as a legitimate medical treatment since 1996. In this paper I argue that the important policy question, then, is not whether but how to allow for the drug’s distribution. In the political arena, proponents and critics of MCDs have made antithetical claims about the relationship between MCDs and crime. Proponents claim that MCDs decrease crime, while critics allege that they are magnets for criminal activity . In this paper I explore these competing claims and their empirical implications, which have received little attention to date from the academic community. Specifically, this study explores the spatial relationship between MCDs and crime in San Francisco. In this chapter I review the relevant literature on cannabis, crime, and urban sociology. I first discuss the alleged link between cannabis use and crime, which has been challenged by research in economics and criminology. The drug-crime link seems especially weak with respect to cannabis, compared to alcohol and other drugs . In the second section I present an ecological framework known as routine activities theory, which attempts to explain crime based on the circumstances in which it occurs. Specifically, routine activities theory states that three convergent factors are required in order for a direct-contact predatory crime to occur: motivated offenders, suitable targets, and lack of capable guardianship . The third and final section of this chapter discusses social disorganization theory, which attempts to explain the relationship between crime and certain neighborhood characteristics. Building on the work of urban sociologists Henry Shaw and Clifford McKay, social disorganization theorists have identified economic deprivation, residential instability, population heterogeneity, and family disruption as categories of neighborhood characteristics that explain crime . I incorporate these insights into the model of urban crime that I present and test in subsequent chapters.In the next two sections I aim to clarify the important distinction that exists between cannabis and medical cannabis dispensaries , and how these different concepts might affect crime. This paper is concerned with the latter. In the empirical analysis that I present in the following chapters, trimming cannabis the key independent variable under review is not a drug per se but a land use that distributes a drug to certain qualified individuals. This study is not designed to address the relationship between cannabis, as a substance, and criminal behavior—a question that would require controlled behavioral experiments or field research. Instead, I explore the criminological impact of MCDs as retail businesses that operate in complex urban environments.

The policy question is therefore one of land use, not substance control per se. Nonetheless, it is possible that MCDs could have an indirect effect on neighborhood crime rates via the underlying effects of cannabis , if the latter were independently associated with crime. Toward that end, I review in this section the relevant literature on cannabis and crime. The alleged link between cannabis and socially undesirable behavior dates almost as far as back any other mention of the drug in United States history. In fact, a compelling argument can be made that modern cannabis prohibition has its roots in the historical perception, fueled by racist and xenophobic stereotypes of Mexican immigrants, that use of the drug led to violent criminal behavior. In 1914, an article appeared in the Los Angeles Times linking a mysterious plant with Mexican immigrants and “many revolting crimes.”9 Years later, a Montana lawmaker would say of the drug: “When some beet field peon takes a few rares of this stuff, he thinks he has just been elected president of Mexico so he starts out to execute all his political enemies.” These stories from the west were relayed nationally by outlets like the New York Times, who quickly caught on to the stereotype of the violent, marijuana-fueled Mexican criminal. One piece in 1925 ran with the headline, “Mexican, Crazed by Marijuana, Runs Amuck With Butcher Knife.” It went on to describe the violent slaying of six hospital patients by a veteran of the Mexican Army, who was allegedly intoxicated at the time of the murders. When the United States Congress first discussed a proposal to prohibit the drug federally in 1937, its members likely had these violent images in mind. Fast forward to present day. The link between marijuana and violent criminal behavior has been challenged by research in economics and criminology. In 2000 the economist Andrew Resignato reviewed existing drugs-crime research and concluded that—despite the assumption made by national drug control policies that there is a clear and causal relationship between drugs and crime—greater evidence may actually exist to suggest a link between “drug enforcement/control/prohibition” and crime. Benson and other scholars have also found that drug enforcement policies have effects on crime rates, independently of drug use . These findings highlight a core challenge faced by drug policy researchers. It is impossible, in most cases, to study the social consequences of a drug without considering, at the same time, the legal status of that drug and what conditions that might impose on users of the drug. Much of the relevant drugs-crime literature employs a tripartite theoretical framework put forward by Goldstein . Goldstein identified three mechanisms through which drugs might contribute to criminal behavior: psychopharmacological factors, economic compulsion, and systemic violence. “Psychopharmacological risk” of criminal behavior refers to the possibility that the psychoactive components of a drug might make a user more likely to engage in criminal behavior, strictly in terms of his or her neurobiology. “Economic compulsion” refers to the process by which drug users might engage in criminal activity in order to fund their potentially expensive drug habits. “Systemic violence” occurs when drug markets are enmeshed in broader networks of crime and violence, in which drug users become involved. Pacula and Kilmer report a positive association between marijuana use and reports of nonviolent crime, but did not find a causal link between marijuana use and violent crime. They concluded that marijuana use might play a greater role in a criminal getting caught than in the original decision to commit the crime. Indeed, a large number of arrestees test positive for cannabis in the United States . But correlation does not mean causation. Several explanations exist for the high rate of marijuana use among criminal offenders that do not rely on a causal link between the drug and criminal behavior. First, there is the simple observation that cannabis use is very prevalent in the United States. In 2010 an estimated 22.6 million Americans above the age of 12 were “current users” of illicit drugs; of these, 17.4 million—or 6.9% of the population —were current users of cannabis . Given these high rates of use, one would expect significant portions of any sample of Americans to test positive for marijuana; that a majority of criminal offenders test positive for marijuana is not, in itself, a very interesting finding. Taylor and colleagues report that between 34% and 61% of adult males used cannabis at least once in the year 2000, including between 17% and 63% of violent criminal offenders and between 34% and 74% of offenders in property crimes .