The Effect of Shelter Availability on Arrests of People Experiencing Homelessness

• In Event: Homelessness and the Criminal Justice System

Abstract

In the United States, thousands of people experiencing homelessness are arrested and charged with violating quality-of-life ordinances each year. Often, the charges stem from violations of laws and ordinances that criminalize acts and behaviors associated with homelessness, such as camping, sleeping, and storing property in public places. In recent years, a major Supreme Court decision—Grants Pass v. Johnson—has allowed communities to further escalate criminalization by allowing cities to arrest people experiencing homelessness for sleeping in public, even when they have nowhere else to go. This follows the rise in these laws nationally, with 98 of the 100 largest cities passing at least one ordinance criminalizing these behaviors since 2000, despite research finding it to not decrease homelessness in a community (e.g., Herring 2019; Lebovits and Sullivan 2025). The use of quality-of-life ordinances as a tool to manage people experiencing homelessness has negative consequences, exacerbating barriers to employment, increasing financial strain through criminal fines and fees, reducing access to mental health and addiction resources, and creating a homelessness jail cycle. Moreover, these policies create costs for cities and states by expending scarce resources on managing the homeless population through the criminal justice system and other safety net services rather than through services. 

Yet, almost no causal research exists studying the relationship between shelter availability and arrests. In this study, we identify the causal connection between shelter availability and arrests of people experiencing homelessness for quality-of-life ordinances. We implement a quasi-experimental design—regression discontinuity—around like-random variation in the precise high or low temperatures that activate extreme weather plans, which create a sudden, temporary increase in shelter availability. To do so, we take advantage of access to restricted administrative, individual-level data on shelter use and capacity, and daily public records on jail incarcerations and extreme weather plans in the largest 100 counties in the United States. We hypothesize that an increase in shelter availability during extreme weather emergency plan activations reduces the number of arrests of people experiencing homelessness for quality-of-life ordinances. Our research contributes to literature on unsheltered homelessness, shelter resistance, and the effect of extreme weather conditions on the experience of homelessness, specifically by providing knowledge on an alternative solution to unsheltered homelessness than criminalization, namely, shelter. Our study will also provide an elasticity of shelter on unsheltered homelessness particularly in cases of unfavorable unsheltered conditions.

Author

• Stephanie Casey Pierce, University of Tennessee (Knoxville)
  Presenting Author