Recent years have seen a proliferation of policy proposals aimed at reducing evictions, from legal aid to rental assistance. However, designing effective policies requires understanding whether evictions primarily result from persistent tenant nonpayment or from landlords’ unwillingness to tolerate temporary payment difficulties. Despite extensive research documenting the prevalence and consequences of evictions in the United States, the underlying drivers of eviction decisions have remained poorly understood.
The authors address this question using unprecedented access to detailed lease-level ledger data from landlords operating in high-eviction neighborhoods across multiple US cities. Their dataset provides month-by-month records of payments, tenancy duration, vacancies, and eviction timing—allowing the first comprehensive examination of how nonpayment patterns evolve before eviction decisions. This granular data enables the researchers to track both the frequency of missed payments and landlords’ responses, revealing patterns that previous studies could not observe.
Analysis of these data reveals two key patterns. First, nonpayment is common, with roughly half of tenants missing at least one payment, but many recover financially without facing eviction. While eviction is almost always preceded by nonpayment, nonpayment does not invariably lead to eviction. Second, landlords often tolerate multiple months of nonpayment before pursuing eviction, suggesting they face a meaningful tradeoff between evicting and waiting to see if payment resumes. This forbearance indicates that landlords carefully weigh the costs and benefits of eviction rather than automatically initiating proceedings after missed payments.
To understand whether high eviction rates stem from persistent nonpayment risk or low eviction costs, the authors develop a model of landlord decision-making. The model reveals that eviction decisions are shaped by two key forces: significant costs of eviction (2-3 months’ rent, or 13-20% of the average tenancy length) and uncertainty about tenants’ future ability to pay. As a result, most evictions involve tenants who have become persistent non-payers rather than those facing temporary difficulties. This finding suggests that landlords generally wait to learn about tenants’ longer-term payment prospects before pursuing eviction.
The authors then analyze three policy proposals: eviction taxes, procedural delays, and rental assistance for tenants in arrears. Each policy reduces eviction rates by only about 5%, primarily by delaying rather than preventing evictions. While some tenants avoid eviction entirely, most affected tenants are simply evicted 4-7 months later than they would have been without the policy. The limited impact of these interventions reflects the underlying reality that most evictions occur only after landlords have determined that tenants are unlikely to resume regular payment.
These findings have important implications for policy design. While modest interventions in the eviction process may not prevent most evictions, they can still improve welfare in two ways. First, they provide valuable insurance to tenants during periods of financial vulnerability by extending the time they have to recover economically. Second, if the social costs of eviction are substantial and not captured by private decision-making, even small reductions in eviction rates could justify policy intervention. Additional protections that prolong the eviction process could also help tenants by lowering their housing costs when they are most financially vulnerable.
The key insight for policymakers is that effective policies must address the underlying causes of persistent nonpayment, not just modify the eviction process itself. While expanding eviction protections may serve important social goals, achieving large reductions in eviction rates likely requires addressing the economic conditions that lead to sustained inability to pay rent.