Exploring ‘near’: Characterizing the spatial extent of drinking place influence on crime

Abstract

The important role of facilities in understanding crime patterns is widely recognized. Studies have demonstrated a connection between the presence of facilities such as bars, parks, schools and fast food restaurants and higher crime rates. Typically, these studies use a single distance threshold. Areas within the threshold are assumed to be related to the facility and those outside the threshold unrelated. But the choice of threshold in each study is usually an ad hoc decision based on the expertise of the researcher. Until recently, there has been no systematic evaluation of the methodology used to define those thresholds. This paper evaluates two methods for determining an empirically-based answer to the question ‘How close is “near”?’ The results of an example analysis testing the association of drinking places and crime in Seattle, Washington are reported. The two most common facility-based measures, Euclidean distance buffers and Street distance buffers are compared across two levels of aggregation and 18 separate distances. Findings indicate the geographic extent of increased crime around drinking places varies based on the type of buffer (Euclidean vs. Street distance) and the width of the distance bands (street block vs. 402 meter (quarter mile) increments). The geographic extent of the influence of drinking places on crime is best captured by street distance measures across street block distances (122 meter bands). Implications of these findings for theory and practice are discussed.