Analysis of crime around a low-barrier, rotating homeless shelter

Abstract

Concerns about crime often motivate local opposition to homeless services, such as shelters with low barriers to entry. To understand whether this central tenet of “not in my back yard” (NIMBY) opposition to low-barrier homeless shelters is supported empirically, researchers leveraged the natural variation associated with the operation of a seasonal, low-barrier shelter. Specifically, for three months each winter, an emergency shelter rotates to a different host location, often a church, each week. The shelter hosts 50–60 unhoused community members each night and is considered low-barrier as there are very few restrictions to entry, such as sobriety or pet ownership. To understand whether crime increased in the vicinity of the shelter host, the authors examined both arrest records and calls for service over a two-to-three-year period, for each of the 15 shelter sites. Using fixed effects Poisson and Negative Binomial regressions, we consistently find that arrests and calls for service do not significantly increase or decrease around hosts of the emergency, low-barrier homeless shelter. This finding contrasts with much of the literature on homelessness and crime, and suggests additional studies are needed that explore shelter specific factors. This study also has policy implications, as concerns about crime often motivate local opposition to the siting of homeless shelters in neighborhoods.