The impact of street lights on spatial-temporal patterns of crime in Detroit, Michigan

Abstract

The importance of understanding crime in the United States assumed enhanced protrusion in the wake of the increased crime rates year by year in certain cities. Neighborhood social demographic variables have been largely used to measure their associations with crime. Other than those social factors, street lighting is a feature of urban and suburban settlement which is widely thought to be a necessary element in preventing crime. Previous research has drawn mixed conclusions about the relationship between street lighting and crime, and the effect of streetlights on neighborhood crime is not entirely definitive. To address this challenge, we examined the spatial associations between street light density, neighborhood social disorganization characteristics and crime (e.g., burglary, vehicle theft, weapons offenses, etc.) in Detroit, Michigan in 2014. Using the street lighting data from the Detroit Public Lighting Authority, crime data from the City of Detroit, supplemented with Census 2010 data, we conducted a Generalized Least Squares model of neighborhood crime in 879 census block groups to test the random effects of the spatial variables and different hours of day on crime. The results show an inverse relationship between street light density and crime rates across census block groups in Detroit and the effects of time period of a day vary according to different types of crime. These findings provided more credible evidence for researchers and policy makers to effectively optimize scarce public safety resources, such as improving street lighting in disadvantaged neighborhoods.