Do Neighborhood Structural Resources Mediate the Association Between Alcohol Outlet Density and Violence?

Background: Neighborhood characteristics such as the built environment, perceived conditions, and social processes, impact local residents. The density of alcohol outlets in a neighborhood straddles each of these domains. Alcohol establishments are built structures where individuals congregate. They may influence social norms by weakening external controls making individuals more likely to engage in violent activities. Inadequate consideration of neighborhood social processes may have led to inconsistent findings. Most prior studies have adjusted for neighborhood effects using census variables as a proxy measure for social disadvantage. However, recent literature demonstrates the complexity of these social processes suggesting the need to consider more refined measures such as collective efficacy.

Hypothesis: Using the availability of structural resources (parks, recreation centers, libraries, etc.) as a proxy for collective efficacy, we will examine the extent to which neighborhood structural resources mediate the relationship between alcohol outlet density (AOD) and violent crime. We hypothesize that the association between AOD and violence in neighborhoods will be smaller, or non-existent, in neighborhoods with access to a greater number of structural resources.

Research Strategy: A combination of geo-spatial techniques and traditional quantitative methods will be used. We calculated neighborhood level rates of violence for Pittsburgh for the years 2008-2010. All aggravated assaults and robberies reported to the police were geocoded to a Pittsburgh neighborhood, defined by census tract. AOD for each neighborhood was calculated from a database created by combining data from InfoUSA and Dun & Bradstreet and geocoded for 2009 to a Pittsburgh neighborhood. Densities were calculated separate for on and off premise establishments. All density measures were calculated with neighborhood population levels. Neighborhood structural resources include: ball fields, bus shelters, bus routes, community pools, foreclosed properties, hospitals, libraries, museums, parks, recreation centers, schools, senior centers, universities, and vacant parcels. Using spatial analysis we will assess violent crime rates and geocoded density of structural components within neighborhood. Hot-spot analysis will be applied to identify statistically significant clusters. Upon identifying clusters, we will run factor analyses to analyze the relationships between each structural resource to identify the principal resources that mediate the association between AOD and violence. We will use Poisson regression to measure between-neighborhood variability in violence rates. Interactions between AOD and structural resources factor(s) will be tested to determine if these resources are mediators.