Significant associations between the locations of alcohol outlets and rates of violent criminal events such as assaults are observed in cross-sectional geographic studies in the alcohol prevention and criminology literature. Recently, alcohol prevention researchers and criminologists have advanced complementary modes of evaluating the relationships between retail and residential community areas and crime rates. These include geostatistical methods to explore the spatial relationships of people and places to violent crime and the theoretical approaches that predict the locations of crime in community areas. Today it appears that the ecological association between alcohol outlets and violence is not an accidental one; the relationship has been replicated in a number of studies, appears on a number of different geographic levels, and has strong theoretical support. Current studies of this relationship, however, are limited to the degree that they do not address four problems: (1) the limited use of theoretically relevant covariates in analysis models, (2) the use of biased statistical models for the analysis of geographic data, (3) a mis-specification of the outlet-violence relationship, (4) the absence of longitudinal data for the evaluation of temporally ordered effects. The current proposal suggests ways to deal with each of these issues and presents methods to explore ecological and individual expectations from theoretical models relating crime potentials (i.e., probabilities of crime across geographic areas) to arrests and injuries due to violent assaults. Four studies are proposed that examine the spatiotemporal relationships between outlet densities and violence (Study number 1), ecological correlations between outlet densities, crime potentials, retail markets and violence (Study number 2), the individual level effects of extreme levels of outlet densities on drinking behaviors, problems and aggressive norms (Study number 3), and individual level risk and protective factors related to assaults (Study number 4). The study designs enable the examination of these effects over long periods of time (10 years) and across vastly different areas of the state of California (e.g., across 1493 zip code areas). The short-term goals of this study are to clarify the mechanisms relating changes in outlet densities to assault rates and establish empirical bases for exploring the relationships of environmental characteristics (e.g., outlet densities) to individual norms for aggression. The long-term goal of this study is to provide empirical data to guide environmentally based alcohol-related crime prevention programs.
