Underage tobacco sales violations and neighborhood crime arrest in Philadephia: a multiscale GIS-based analysis

Neighborhood crime and frequency of police arrest are important indicators for determining neighborhood safety and well-being. Although extant literature has examined crime and neighborhood characteristics and other macro-level determinants, virtually no study has considered the link between neighborhood crime arrest and underage tobacco sales violation (TOB-VIO). This study linked crime arrest data from the Philadelphia Police Department, underage tobacco sales violations from the Philadelphia city database, and sociodemographic/economic data from the Census Bureau to the county shapefile. The principal component analysis (PCA) technique was used to reduce 23 selected variables to four factors: Racial Composition and Neighborhood Disadvantage; Gender Composition; Youth Composition and Poverty; and Ethnic Composition. To account for the effect of spatial dependence in the data, spatial autoregressive and multiscale geographically weighted regression or MGWR techniques were chosen over the nonspatial ordinary least squares regression. Spatial regressions show that crime arrest positively correlates with TOB-VIO and factors from PCA across Philadelphia County. Findings from this study may serve as an area-specific guide for neighborhood policies to improve residents’ safety and well-being and intensify enforcement on the underage tobacco sale violation.