Hot spots of violence in Boston: Spatial considerations for alcohol outlets and drug markets

Purpose: We examine the relationship between alcohol outlets, drug markets and violence in the city of Boston, for 2006. We determine what geographic/environmental and individual factors are related to the production of violence (including relationships between target and adjacent geographic areas). Further, we identify model-based hot spots of violence. Methods: Data from the Boston Police Department, the census, survey data on neighborhood collective efficacy and Massachusetts state data on alcohol outlet type and location are used. Spatial modeling is employed at the block group level (n=544), and maps of hot spots for violence and drug markets are produced. Results: Using spatial regression, we found negative relationships between percentage of the population younger than 16 and population over 25 with at least an associate's degree were associated with a decrease in assaults. Positive relationships were found for increased minority presence, increased percentage of families below the poverty line, percentage vacant housing, percentage renting. For environmental measures, restaurants selling any/all types of alcohol, package (liquor) stores and drug distribution arrests were positively related to violent incidents. Hot spots, (n=17) had higher percentages of population younger than 18, vacant housing units, percentage renters, more restaurants selling all types of alcohol, drug possession arrests, adjacent area vacant housing units, most types of alcohol outlets. Conclusions: Controlling for a variety of socio-demographic and areal characteristics including drug arrests and alcohol outlet type, we found that there are distinct hot spots for violence and that adjacent/lagged area characteristics were related to target area violence.