GOAL 2:
Inform: Improve the Collection of Data and the Use of Information to Support Affected Communities and Inform Policy Makers

For Goal 2, the Strategy Group identified actions that would help key organizations start immediately to collect, report, and analyze relevant data on housing instability, thereby allowing data-driven decisions on targeted interventions to be made during and after the pandemic as soon as possible. Having this information would enable actors to implement the other interventions recommended in this report to make more sustained improvements and better reach those most in need. Particularly important to this goal is that organizations working to reduce housing instability be aware of and understand the full complexity of the disparate and disproportionate harm suffered by BIPOC communities as a result of evictions and housing insecurity.

Specifically, to improve and inform policy decisions around aid, outreach, and tenant protections, federal, state, and local stakeholders need accurate, timely, easily accessible, and detailed data on eviction filings, eviction actions, characteristics of renter households in distress, and health outcomes due to eviction and the aftereffects of the pandemic. Policy makers, social service providers, tenant and landlord associations, and by extension tenants and landlords themselves need data on housing instability that extend beyond the basic metrics of eviction filings so they can understand the outcomes, drivers, and people involved in those filings. An effective data collection strategy for this more comprehensive picture would include data collection from both government records and the community itself. For community-focused data collection, community-based organizations would be valuable partners. Many such organizations possess the know-how, familiarity, and trust that are essential in gathering and ultimately disseminating to disinvested communities both quantitative and qualitative information, and any external initiative will be less effective without community-based partners.

Across the four scenarios explored, the Strategy Group found that filling data gaps and providing data on housing instability are robust needs; that is, these actions would be needed under a range of outcomes to support affected communities, both tenants and landlords, and to inform policy makers. Note that given the interconnected needs addressed under Goal 2, several of the recommended actions overlap. This overlap will better allow interventions reliant on critical eviction data to proceed even if some of the actions are not taken promptly. It also reflects the fact that, given the wide jurisdictional variability of laws, rules, and procedures in housing courts around the country, a broad compendium of strategies is preferred over a one-size-fits-all approach.

STRATEGY 2A: Improve Data Collection, Reporting, and Access to Build Actionable Understanding of Eviction

The lack of a comprehensive and granular eviction database makes it difficult to assess the scale of housing instability in the United States and, in particular, how different populations and geographies have been affected. Having such data would help jurisdictions determine to whom resources should be directed in immediate response to the pandemic-driven housing crisis; it would also inform the development of policy to mitigate evictions in future outbreaks. Additionally, a national eviction database could reduce rental housing instability generally by facilitating the enforcement of fair housing laws in community-centric ways that are not possible today. While some aggregate data regarding the demographics of housing instability are publicly available,¹ there currently exists no platform or data infrastructure to support a nationwide eviction notice, filing, and execution dashboard that can provide sufficiently comprehensive, timely, and granular data to identify populations at risk of eviction. Existing public data consolidation and aggregation efforts track eviction filings in a limited number of cities² or focus on issues in specific localities.³ Indeed, approximately one-third of U.S. counties have no available annual eviction figures.⁴ This gap in information on evictions, their root causes, and the needs for assistance have led to underinformed decision making and administrative challenges for rental assistance organizations, and may result in program design flaws.

Rationale for Actions 2A-1 and 2A-2: These two actions focus on collecting state and local jurisdiction eviction data to inform rapid response that can prevent a wave of pandemic-related evictions while simultaneously supporting the development of a standardized and modernized data collection system for future use in monitoring ongoing eviction trends.

___________________

¹ See HUD’s Picture of Subsidized Households (https://www.huduser.gov/portal/datasets/assthsg.html). Additionally, see Princeton University’s Eviction Lab work (https://www.evictionlab.org) focused on trying to remedy the information gaps in American housing insecurity.

² See https://evictionlab.org.

³ See https://antievictionmap.com; https://evictions.study.

⁴ See https://www.newamerica.org/future-land-housing/reports/displaced-america/housing-loss-and-poor-data.

The most easily accessible data on evictions comes mainly from three points in the eviction process: notices (where required), filings/unlawful detainers, and sheriff writs of restitution. Loss of housing can occur at any of these points. To understand the state of evictions in the United States during the pandemic and beyond, it is important to obtain data from all three points, each of which provides critical details on where interventions can best be implemented and on the nature of evictions for households in a given area.

State and local courts, housing authorities, and law enforcement agencies are important sources for such raw data on the legal process of eviction. However, the majority of eviction data is hidden within unsearchable digital images and physical copies of court records (Leung et al., 2020). Different jurisdictions have very different procedures and resources, and so have different capacities to record and report on these data. Some do not possess the technology required for electronic record keeping. Scaling up the collection of national eviction data will therefore require modernization of record keeping in localities nationwide. Data science tools can be used to mine records⁵ and could fill important gaps in knowledge about who is being evicted and where evictions most commonly occur.

Modernizing and establishing minimum standards for electronic records would provide a local benefit by improving court and other information systems. This benefit and possibly federal funding could serve as an incentive for jurisdictions to use these new tools and potential funds in part to report past and current eviction case data to HUD and the U.S. Census Bureau for curation and analysis. Such a project would have the additional advantage of making it easier for certain jurisdictions to identify themselves for additional federal housing support.

Action 2A-1: The U.S. Congress, in partnership with HUD, the U.S. Census Bureau, and experts in housing instability, should consider establishing a temporary (2- to 3-year) Eviction Data-Collection Assistance Program that would (1) create a task force to assess the current state of data and existing databases and inform data collection strategies, and (2) fund underresourced state and local jurisdictions to collect data, modernize their electronic filing and records systems, and overcome administrative barriers to fulfilling reporting requirements discussed under Action 2A-2.

Action 2A-2: HUD, public housing authorities, and local and state governments should establish a federal reporting standard for eviction data, as well as a program akin to the Uniform Crime Reporting (UCR) program through which local police departments voluntarily submit their crime data to a state UCR program or directly to the Federal Bureau of Investigation. HUD could offer an online system for free to promote a basic level of uniformity, information security, and federal access. In addition, state and local jurisdictions should require

___________________

⁵ Examples include natural language processing and Bayesian demographic estimation.

• Landlords and their legal representatives to notify local authorities whenever eviction notices or unlawful detainer notices are sent,⁶ and those authorities should record the actions and their outcomes (e.g., abandonment).
• Sheriff’s departments (or other responsible law enforcement agencies) to record and report when they are asked to carry out or execute a writ of restitution.
• State and local courts to file electronic copies of all documents associated with eviction cases, and authorities should report these cases through the reporting program recommended here.
• Public housing authorities, as organizations that serve largely populations most vulnerable to eviction, to file independent reports on any eviction actions within the public housing system, including the circumstances and outcomes of any notices, filings, and writs and the stage in the process if and when the tenant was removed from the unit.

Rationale for Actions 2A-3 and 2A-4: The preceding two actions, if taken, would resolve the most pressing issues of (1) compiling the most critical and readily available pandemic-related eviction data, and (2) durably improving reporting systems in state and local jurisdictions. Once these first steps have been accomplished, Actions 2A-3 and 2A-4 cast a broader data collection net and include a retrospective analysis of the causes and consequences of evictions. This deeper dive into eviction data includes much more granular and comparative quantitative data beyond the basic needs described in Actions 2A-1 and 2A-2, as well as qualitative data that can help yield new understanding of eviction and housing instability more broadly.

While data from state and local courts, housing authorities, and law enforcement agencies are useful for understanding issues broadly, they often miss the nuanced experiences of specific communities. Court filings alone may overestimate forcible removals and underestimate unofficial losses of housing as fractions of eviction actions. For example, many renters vacate housing after being notified by a landlord of pending eviction but before the actual court filing itself (so-called “voluntary eviction,” or abandonment).⁷ Cases of people who are asked informally to vacate housing for which they have no lease also are not captured in court records. Some existing quantitative data collection instruments and initiatives (see examples under Action 2A-3 below) do extend into more nuanced territory but provide only annual estimates of the numbers, demographics, and circumstances of evicted households and of those at risk of eviction. More frequent ongoing data collection that includes more granular estimates of the scope and magnitude of the problem at the local level could help generate solutions.⁸ It is particularly important to ensure that the focus extends beyond coastal urban areas to include other urban areas, small cities, suburban communities, and rural communities, where eviction rates can be substantially higher than the national average.⁹

___________________

⁶ Some states require that notice be sent to tenants prior to the filing of an eviction action; some do not. Requiring prefiling notice is another valuable protection for renters given the short timelines for evictions.

⁷ Cases of “voluntary” or “informal” eviction are legally considered to be abandonment.

⁸ While a census tract level of granularity would be ideal, this may not be feasible in a reasonable timeframe in many locations or when critical data privacy is taken into consideration.

⁹ See the Eviction Lab’s rankings here: https://evictionlab.org/rankings/#/evictions?r=United%20States&a=2&d=evictionRate.

Furthermore, while quantitative data are important, they are insufficient on their own. Additional insights come from experience dealing directly with tenants, landlords, and the issues that arise, including the actual experience of processes, practices, rules, laws, and outcomes. Examining these qualitative, community-based data can provide opportunities to identify more people at risk of informal eviction or displacement and ways to improve services and procedures. Similarly, documenting the degree of unmet need within federal housing programs would allow policy makers to see the true scope of housing instability in the United States. Community-based organizations are, in many cases, better equipped and positioned to provide and collect such information within their communities relative to government agencies.

To devise effective solutions for the eviction and housing instability crisis, policy makers and community-based organizations need to understand all factors leading to these phenomena and their consequences—including those that can be tracked easily by numbers and those that are too complex to be captured by simple quantitative metrics. They also need ongoing updates of those indicators to consistently track the scope of the problem, the landscape of housing regulations, and the local institutional contexts in which the crisis is unfolding. Only then will policy impacts and the consequences of inaction become fully apparent.

Action 2A-3: HUD, in collaboration with academics and other researchers, the Census Bureau, philanthropies, landlords and property managers, and community-based organizations, should launch and coordinate a comprehensive quantitative research program that provides an ongoing accurate picture of housing insecurity in the United States based on existing data and data to be collected in the future. Types and sources of data used for this research should include

• data from existing collection instruments, such as the U.S. Census Household Pulse Survey,¹⁰ the Rental Housing Finance Survey,¹¹ and the American Housing Survey,¹² and local homelessness management information systems;¹³
• information on eviction cases as stated in the November 2020 update to HUD’s Research Roadmap;¹⁴
• a compendium of relevant state and local laws and regulations germane to affordable housing, housing habitability, eviction, and tenant discrimination;

___________________

¹⁰ See https://www.census.gov/data-tools/demo/hhp/#.

¹¹ See https://www.census.gov/programs-surveys/rhfs.html.

¹² See https://www.census.gov/programs-surveys/ahs.html.

¹³ See https://www.hudexchange.info/programs/hmis.

¹⁴ See https://www.huduser.gov/portal/sites/default/files/pdf/Research-Roadmap-2020.pdf (p. 52, accessed May 19, 2021). While HUD has already called for a feasibility study and clearly recognizes the complexity of this monumental task, it must be acted upon rapidly and treated as an immediate priority since many subsequent actions will depend on its quick and effective implementation.

• eviction data collected by private firms for the purpose of tenant screening (such information may have to be licensed or purchased outright);
• near-real-time information about rental applicants, including source of income, demographic information consistent with the protected classes under the Fair Housing Act,¹⁵ and the application rates and waiting lists for federal housing assistance programs; and
• near-real-time information about the locations and quantities of outstanding tenant rental arrears nationwide so emergency funds and other assistance can be targeted effectively and distributed rapidly—information that will be particularly important for preventing major disruptions for both tenants and landlords during recovery from the COVID-19 pandemic and future crises.

Action 2A-4: HUD, in collaboration with community-based organizations, public housing authorities, and academic and other research institutes, should launch and coordinate a comprehensive qualitative research program to identify, record, and analyze both short-term predictors of housing instability and more robust, long-term metrics to inform future policy. These indicators will be essential to monitoring eviction risks and inequities and responding rapidly to unmet housing needs in targeted, equitable, and community-centric ways.¹⁶ Such qualitative indicators may include tenant and landlord characteristics related to history, identity, or behavior and circumstances surrounding the initiation of eviction proceedings.

Rationale for Action 2A-5: Once all of the relevant data streams have been effectively compiled and standardized, they can be translated into an actionable format, such as an online dashboard. It is imperative that the insights gained from this data collection effort be available and understandable to whoever can benefit from them. Analyzing and operationalizing these data and insights will require an accessible public platform where policy makers, landlords, community-based organizations, and individual renters can engage with the data and generate new, context-specific understandings. Equally important is recognizing that while data are essential for providing support for renters (Kirchner and Goldstein, 2020), data can also be used to increase housing instability for renters in that information about an individual’s current or past housing insecurity can increase barriers to future housing. Strong privacy protections are therefore needed to ensure that data collected for policy and outreach purposes are not used for purposes harmful to renters’ future prospects. Action 2A-5 requires that public data be aggregated and anonymized while still enabling an analysis of race, gender, age, ethnicity, and disability status. Institutions and scholars interested in accessing the data at a more granular level to serve or study these vulnerable populations will need to adhere to strong protocols protecting personally identifiable information.

Another important consideration in implementing the dashboard will be outreach. The capability for a targeted response to the housing crisis is useless if that capability is

___________________

¹⁵ Public Law 90-284, 42 U.S.C. § 3601.

¹⁶ For more information on a potential methodology for this action, explore the Urban Institute’s 2020 paper Assessing Options for Federal Rental Assistance during the Pandemic (Galvez et al., 2020).

not advertised to the actors who will put it to use. Partnerships with social media entities and communication strategists will need to be considered to support the development of a coordinated public information campaign that raises awareness about both the housing crisis and the solutions available to those facing housing instability (see Goal 1).

Action 2A-5: HUD, in consultation with data scientists in the private, academic, and nongovernmental organization sectors, should build a publicly accessible National Housing Security Dashboard combining, collating, translating, and visualizing all the information sought under Actions 2A-1, 2A-2, 2A-3, and 2A-4 into a usable online interface. Actionable and geographically granular information on the dashboard should grow as new types of information become available and should progressively include

• national, state, county, and tract aggregated counts of eviction notices, filings, settlements, executions, and outcomes of tenants who vacated (either voluntarily or forcibly), by characteristics of affected tenants and their landlords;
• interactive breakouts explaining specific protection protocols, such as whether particular tenants are receiving federal housing assistance and under which programs, the reasons for eviction, and statistics on illegal evictions (e.g., those conducted in violation of COVID-19–related moratoriums); and
• a reporting of housing assistance needs containing national, state, county, and tract aggregated counts of households that have applied for or are receiving low-income housing assistance, sortable by characteristics of affected tenants and programs.

STRATEGY 2B: Develop a Compendium of Promising Strategies for Housing and Public Health Authorities to Mitigate Evictions, Housing Instability, and Their Downstream Health Effects

Rationale: Although the COVID-19 pandemic increasingly appears to be coming under control in the United States, public health experts anticipate occasional outbreaks and possible emergence of additional, more dangerous variants of SARS-CoV-2. Both poor health and the need to care for others, such as family members with medical concerns or children during school closures, can increase housing instability through loss of employment and inability to pay rent. In particular, as discussed earlier, research indicates

that eviction leads many displaced families to “double up” with another household,¹⁷ which increases the risk of health problems (Bush and Shinn, 2018; Evans et al., 2002), including the spread of infectious disease. Multifamily dwellings present more opportunities for exposure and therefore higher risks in a pandemic relative to single-family dwellings (Harlem, 2020), and people at greater risk of housing instability are more likely to reside in multifamily settings in much of the country. However, guidelines are lacking on how to devise effective strategies at the local, state, and national levels that take into account pandemic-related health data and metrics to inform housing policies. This impediment has led to piecemeal and inadequate interventions. Given the wide variation in salient laws, rules, and procedures, few solutions would work in all jurisdictions; therefore, the nation needs a compendium of strategies that may be applicable in different circumstances. The variability in policy responses across different jurisdictions presents a unique opportunity for rapid-response research to identify and understand the tools that can be useful for reducing housing instability, but capitalizing on this irreplaceable opportunity will require that resources be recruited quickly. The National Science Foundation’s Rapid Response Research (RAPID) (National Science Foundation, 2020) funding mechanism, the National Institutes of Health’s Funding Opportunities Specific to COVID-19,¹⁸ and similar sources may be particularly well suited to supporting this critical strategic science.

Action 2B-1: HUD and HHS (specifically the CDC), in partnership with state agencies administering health care services, public health, and Medicare/Medicaid, as well as organizations and businesses in the health care sector, should fund research to compile strategies and ordinances aimed at mitigating the risk of COVID-19 in housing-unstable populations across the nation, assess their effectiveness, and develop models or best practices that can be applied by local and state housing and public health authorities and organizations to mitigate the health effects of evictions throughout the COVID-19 pandemic and its aftermath, including strategies for tracking risks and determining appropriate indicators for enacting and lifting such measures.

Action 2B-2: HUD, HHS, and USDOL, in partnership with housing services, state agencies administering health care services, public health, Medicare/Medicaid, and unemployment insurance, as well as nonprofit organizations and the private health care sector, should publish guidelines on the conditions under which it is safe to remove such short-term COVID-19 protections as eviction moratoriums based on local health metrics and employment statistics. Specifically, in partnership with state public health authorities,

___________________

¹⁷ Sixty-five percent of evicted families reported having doubled up, according to data from the Fragile Families and Child Wellbeing Study (https://fragilefamilies.princeton.edu/documentation), a large survey administered starting in 1998 (Nande et al., 2021). Forty percent of parents doubled up after being forced to move for such reasons as eviction, disrepair, or safety concerns or because they were unable to afford their unit, according to a large interview study of American families (Harvey, 2018).

¹⁸ See https://grants.nih.gov/grants/guide/COVID-Related.cfm.

these actors should link COVID-19 ordinances, cases, hospitalizations, intensive care unit (ICU) visits, mortality, testing availability, vaccination coverage, and employment statistics in the national eviction dashboard proposed under Action 2A-5. These guidelines could also inform understanding of the accessibility and availability of safe in-person pre-K care and K-12 education.

STRATEGY 2C: Commission Research to Understand Continuing Housing Instability Problems and Deploy Efforts to Address Them

Rationale: Housing instability can result from a variety of factors, including poverty, weak finances, involvement with the criminal justice system, unemployment, limited access to health care, housing market dynamics, systemic racism, and others (Bhutta et al., 2020; Hsu et al., 2018; Purtle et al., 2020). However, the root cause in most of these cases is the sheer burden of rental costs faced by vulnerable households. According to the latest data from the American Housing Survey, the majority (61.0 percent) of renting households below the poverty line spend at least half their income on housing, with one in four of those households (50.6 percent) spending more than 70 percent of their income on shelter costs alone (U.S. Census Bureau, 2020). The COVID-19 pandemic provoked an economic crisis that resulted in millions of workers losing their jobs or dropping out of the labor force altogether, pushed households that were already in a precarious position over the edge, and increased poverty. These economic impacts may linger for years to come. The accumulation of rental arrears during this period means that housing instability may affect millions of people in the United States who were not previously at risk. The scope and complexity of housing instability have prompted many organizations to rethink their work and establish new types of services to provide in response to the needs of tenants and landlords and the available housing data. The need remains, however, for evidence-based interventions and strategies for reducing housing instability, as well as better methods for evaluating the effectiveness of such solutions.

Action 2C-1: Academic or other research institutions should perform a gap analysis of all programs that provide housing support to identify needs, overlaps, inefficiencies, and gaps in outreach during the pandemic or other catastrophic events.

Action 2C-2: Academic or other research institutions should use the data available on the proposed eviction dashboard (see Strategy 2A) to develop metrics for the evaluation of

housing support programs for both tenants and property owners (this action is in support of Action 2A-4).

Action 2C-3: Academic or other research institutions should create interdisciplinary research agendas for the evaluation of health, social, and economic interventions used to reduce housing instability in the context of such major stressors as a pandemic, and for research on additional needs in this area that reflects the diversity and lived experiences of the most affected and at-risk communities.

Action 2C-4: Academic or other research institutions should support efforts to collect the evidence needed to fill gaps in knowledge about differential outcomes of eviction by race and ethnicity, such as impacts and costs of eviction for BIPOC, and the potential for racial inequities in outcomes due to policy design and implementation in the context of such major stressors as a pandemic.