The final class of studies that dealt with alcohol examined aggregate data of neighborhood or city crime rates and various measures of alcohol availability, usually the density of alcohol outlets in a given geographic area. These studies find that alcohol availability is positively related to violent crime in 29 out of 42 models (69%), and positively related to nonviolent crime in 20 out of 23 models (87%). The apparent disparity in those proportions is not evident in the study-level tally (see Table 11.2) where 80% of the studies for both categories, violent and nonviolent, reported a PoC that was statistically significant. Comparing those proportions, we conclude that alcohol availability is associated with violence, but not a differential predictor.
The illicit drugs are a more heterogeneous group to assess. We chose to summarize the findings under several different categories in Tables 11.1 and 11.2. We summarized tests of marijuana, cocaine, crack-cocaine, and heroin/opiate consumption because these drugs were either separately tested in the set of reviewed studies, or they are drugs that have received particular attention in the debate about drugs and crime. We group models that tested other types of illicit drugs, such as amphetamines, sedatives, barbiturates, PCP, etc. These have received less individualized attention by policymakers and the public. They have also received little attention by researchers so there are too few studies to examine each separately. The category “undifferentiated illegal drug consumption” summarizes findings from studies that grouped together all different illicit drugs into a single “drug use” variable. The “substance use” category reports findings from the handful of studies that created a single variable combining both alcohol and illicit drug consumption.