Association for Public Policy Analysis & Management

In 2016, at least one in 50 renter households in major metropolitan communities throughout the United States received an eviction judgment (Eviction Lab, 2018). Despite the prevalence of evictions, there exists little academic research on how evictions affect families, individuals, and neighborhoods (Desmond & Shollenberger, 2015). Much of what we know stems either from literature on residential mobility—which often does not distinguish evictions from voluntary moves or concentrates on moves precipitated by public housing closure or natural disasters. From the limited body of academic literature, we know that an eviction deepens the experience of poverty by driving evicted families into more distressed neighborhoods (Desmond & Shollenberger, 2015), substandard housing (Desmond, 2012; Desmond, Gershenson, & Kiviat, 2015), increased material hardship (Desmond & Kimbro, 2015), increased risk of unemployment (Desmond & Gershenson, 2017), and poor physical and mental health (Desmond & Kimbro, 2015; Manzo, Kleit, & Couch, 2008). Evictions are often concentrated in low-income neighborhoods and disproportionately affect women of color, especially those with children (Desmond, 2012; Desmond, An, Winkler, & Ferriss, 2013).

Desmond (2015) has established that evictions can increase future housing instability in the form of more frequent moves, but despite the substantial resources dedicated to homelessness prevention, little empirical research has investigated evictions as a possible pathway to homelessness. The present research addresses this gap by drawing upon a database of over 142,000 evictions filed in the Franklin County, Ohio municipal court between 2010 and 2016. By linking the eviction data to other local administrative sources, including jail data, homelessness data, code enforcement, legal aid, mediation services, and Housing Choice Voucher (HCV) applications, I explore the relationships between eviction and homelessness. My research design exploits two sources of randomization, allowing the opportunity to establish a causal link between evictions and homelessness. My first identification strategy exploits the random assignment of magistrates to eviction cases. My second identification strategy takes advantage of the lottery system used by the Columbus Metropolitan Housing Authority to select HCV applicants for HCVs. This exogenous source of randomization allows me to compare the eviction and homelessness outcomes of tenants who rent on the subsidized market to similar tenants who rent on the private market.

I expect to find increased instances of homelessness among tenants who experience an eviction. Moreover, I expect an increased risk of homelessness among tenants with histories of incarceration and among tenants who rent on the private market. Further, I expect to find a decreased risk of eviction among tenants who receive some form of intervention from Legal Aid or a local nonprofit that provides landlord-tenant mediation and an increased risk of eviction associated with tenants residing in properties with histories of code violations. This research provides important evidence for practitioners and policymakers, who have begun to reexamine tenant-landlord law and legal procedures in light of increasing public scrutiny on the role of evictions in contributing to housing instability and exacerbating poverty.