Eviction as a social determinant of sexual health outcomes

The United States (U.S.) is in the midst of a housing affordability crisis and an eviction epidemic, which is particularly impacting vulnerable populations: low-income households, people of color, single mothers, and children.^1–5 Two-thirds of low-income U.S. families live in rental housing,⁶ a proportion that has grown in the wake of the 2008 Great Recession.⁷ Since the Recession, low-income households have experienced rising housing costs while their incomes have remained stagnant.^7,8 As a result, most poor renting households spend over half of their incomes on housing, while nearly one in four spends over 70%.⁸ As more and more families struggle to pay rent, the rate of evictions has increased. Recent estimates suggest that 2-3% of renters nationwide experience eviction, with rates of eviction as high as 16% (i.e. 16 eviction judgments per 100 renter-occupied homes) in some parts of the country.⁹

While the links between housing instability and population health more broadly are well-established,¹⁰ there is a dearth of research specifically examining the connections between eviction and health including sexual health outcomes.¹¹ This is an important gap for two key reasons. First, evictions are considered by experts to be among the most deleterious sources of housing instability¹² in that they often come about suddenly, create extreme financial strain and stress, and carry lasting legal consequences that can preclude families from accessing quality, affordable housing in the future. This suggests that eviction may have profound health effects and may operate both as a fundamental cause of disease,¹³ operating through poverty, and a pathway through which poverty becomes embodied in health outcomes.¹⁴ Second, eviction as a source of housing instability is both specific and policy-sensitive. Levers at the city and state for eviction may include reforming eviction law, changing rent court processes, or expanding existing housing voucher programs.

In this issue of Sexually Transmitted Diseases, Linda Niccolai and colleagues investigate the ecological (county-level) relationship between eviction and two sexually transmitted infections (STIs, i.e. chlamydia and gonorrhea),¹⁵ an interesting foray joining a long line of research demonstrating that STI incidence is highly sensitive to social processes.^16–18 The authors leverage a novel data source, the Eviction Lab National Database, to answer this research question. The Eviction Lab data represents the first attempt to compile court records of eviction filings and judgments on a national scale.⁹ After controlling for a number of county-level demographic, social, and geographic covariates, the authors found that, on average, counties with high eviction rates (>1.9 eviction judgments per 100 renter-occupied households) had 63.8 (95% CI 45.1, 82.5) more chlamydia cases and 20.4 (95% CI 13.5, 27.4) more gonorrhea cases per 100,000 population one year later, when compared to counties with low eviction rates (<0.59 judgments per 100 renter-occupied households). These findings were statistically significant and proved robust to a number of thoughtful sensitivity analyses.

The authors present several plausible explanations for the identified associations between eviction rates and STIs one year later at the county-level. In general, the authors make the case that eviction may increase sexual and/or social vulnerabilities in ways that increase community levels of STIs. For example, individuals may change their behavior including engaging in sex work or increasing risky sexual behaviors in exchange for housing or other material needs (e.g. food, transportation). They may also change their behavior in ways that increase risks for STIs (e.g. alcohol or drug use) in an effort to cope with the stress that housing instability and the associated material hardship presents. Evictions may also cause changes to relationships (i.e. disruptions in monogamous relationships, initiation of new relationships) resulting in changes to sexual network structures and increasing STI transmission potential within communities. Eviction and the resultant displacement may also disrupt healthcare access and decrease opportunities for STI testing and treatment, resulting in longer durations of infection. These mechanisms may impact vulnerable populations in particular who have little or no buffer to safeguard against the more deleterious outcomes related to eviction.

While we find this a well-conceptualized study, we offer one comment and two potential limitations. The authors suggest that the study’s ecological design is a limitation. While we agree that the results cannot be interpreted as the effect of individual-level eviction on individual STI risk, it is also true that solutions to the eviction epidemic will likely be implemented and evaluated at a community-level (e.g. county or state). Thus, we encourage researchers to continue to focus on policy-relevant geographies to aid in the identification of policy-based solutions to eviction and its downstream health effects. One limitation that bears mentioning is the potential for selection bias. Approximately 17% of U.S. counties had missing eviction rates and were excluded from the analyses. While it could be the case that the excluded counties do not differ systematically from the analytic set, this is difficult to assess based on the data presented and suggests that the generalizability of the findings might be limited to those counties with eviction data. It should also be noted that even among counties with available data, Eviction Lab data quality may vary by state.¹⁹ An additional limitation in this study is that it is unclear whether highly advantaged U.S. counties (i.e. largely white, with high income, high education, and low unemployment) ever experience high rates of eviction. If not, effect estimates for this subgroup of counties might be “off-support,” relying heavily on extrapolation.^20,21 Moreover, these estimates may have limited utility,^20,22 begging the question: How do we interpret the effect of exposure to high eviction rates in counties that are very unlikely to experience high eviction? In order to judge how much this issue, referred to in the epidemiology literature as “non-positivity,” might threaten causal inference in this study, it would be useful for the authors to present data illustrating the degree to which combinations of the various covariates included in the multivariable regression models are predictive of exposure to eviction, as is common practice in studies that use propensity score-based approaches to address confounding.^21,23

Limitations aside, this publication contributes to a small but growing body of literature linking eviction to negative health outcomes.^24–30 We look forward to more research focused on eviction and sexual health outcomes to confirm and strengthen these findings. Perhaps more importantly, we look forward to research pointing to specific interventions to prevent eviction and its downstream effects on sexual health. Some interesting work evaluating the impact of housing programs on health is already charting a course. For instance, a randomized trial evaluating the Housing Opportunities for People with AIDS (HOPWA) program in three U.S. sites (Baltimore, MD, Chicago, IL, and Los Angeles, CA) demonstrated that provision of rental assistance could effectively address the mental and physical health needs in unstably housed and homeless people living with HIV/AIDS.³¹ In an observational study among public housing residents in Atlanta, GA, relocation from public housing to other neighborhoods resulted in a reduced odds of testing positive for an STI (chlamydia, gonorrhea and trichomonas).³² Given the health promoting potential of high-quality, stable housing, researchers and public health advocates should lend their expertise to the planning, implementation, and evaluation of interventions to address the root problem of affordable housing in the U.S. Adopting a “Health in All Policies” approach, we can address this important social determinant of health, advance health equity, and promote population health.^33,34