Unraveling the Nexus: how street network morphology influences crime in Detroit

Urban streets are primary settings for criminal activities. Although prior studies have examined various environmental factors influencing criminal behavior, insufficient attention has been paid to street configurational types within street network morphology. To address this gap, this study employs kernel density estimation and spatial autocorrelation to analyze the spatial distribution patterns of assault, robbery, larceny, and motor vehicle theft in Detroit in 2019. Based on these spatial patterns, a comprehensive street environmental indicator system was constructed, incorporating three dimensions: street network morphology, nodal characteristics, and socioeconomic attributes. Negative binomial regression models were subsequently employed to analyze the effects of these factors on the spatial distribution of the four crime types. The findings reveal significant spatial clustering of all four types of crimes in Detroit. Neighborhoods with a higher proportion of community roads, increased street permeability for pedestrians, a higher density of dining establishments, and elevated rental rates were more likely to experience criminal activities. Conversely, neighborhoods with a greater number of intersections exhibited lower crime frequencies. Areas with a higher percentage of residents holding bachelor’s degrees or above were more prone to property crimes. In terms of street network morphology, the proportion of ring roads (cell-ratio) had a positive impact on assault, robbery, and larceny. This study contributes to the literature by expanding the indicator system for analyzing the relationship between street environments and crimes, enriching existing crime pattern theories, and providing new perspectives and practical guidance for street design aimed at crime prevention.