The Social and Individual Effects of Homeless Shelter : Evidence from Temporary Shelter Provision

What does homeless shelter achieve? We leverage administrative records of homeless services in Los Angeles County to construct a novel dataset of daily, site-level counts of shelter beds and occupants from 2014 to 2019. We pair this with daily, block-level crime incident data and daily, hospital-level data on ER visits to assess the relationship between shelter and area crime and health. We exploit variation from shocks to shelter availability from Los Angeles County's winter shelters program to study the effects of providing temporary shelter. We find that reducing unsheltered homelessness significantly reduces crime and ER visits for psychiatric conditions. We conclude with evidence that entering shelter also reduces short-run mortality but find no evidence that temporary shelter reduces future homelessness more than street outreach or other non-shelter services. Our findings suggest that shelter functions as a public good with high social benefits. When agents charged with provision of homeless services are evaluated on their ability to reduce overall homelessness, they are unlikely to internalize these benefits and may under provide shelter