Association for Public Policy Analysis & Management

Large increases in rents since the Great Recession have increased attention to the consequences of housing instability for low- and moderate-income households, with particular attention focused on residential evictions. Evictions are a national problem, with more than 700,000 renter households being threatened with an eviction notice each year. Recent research suggests that evictions have cascading disruptive effects leading to job loss, adverse health effects, and negative consequences for children Desmond (2016). As rents have increased, many large cities have introduced new measures to stem evictions, such as expanded legal aid for low-income tenants facing eviction, short-term rental assistance, and improved tenant protections.

We investigate the causes and consequences of court-ordered residential evictions for low-income tenants in New York City using a novel administrative data set we have assembled from various sources. We merge the universe of New York City housing court cases from 2006 to 2016 with outcomes from various administrative sources: public benefits (cash assistance, food stamps and Medicaid); applications and stays in homeless shelters; quarterly employment and earnings data; and hospitalizations from the State’s SPARCS data set.

We test whether evictions have a causal effect on these outcomes using a rigorous quasi-experimental research design that leverages the random assignment of housing court cases to courtrooms in New York City’s Housing Court. We estimate instrumental variables models that isolate the causal effect of eviction due to being randomly assigned to a courtroom with higher ex ante risk of eviction.

We then explore the extent to which adverse events might also be a cause of housing court appearances and eviction. We do this using an event study design that leverages the high frequency of many of our outcomes and the exact timing of court appearances, resolutions, and evictions. This design will inform whether adverse events such as job loss or health events might precipitate appearances and evictions.

Our results will provide new causal evidence on the dynamics of housing instability and disadvantage and inform how policies can be designed to maintain functioning housing markets while avoiding the most adverse effects of eviction on the well-being and opportunities of disadvantaged families.