Millions of homeowners have fallen behind on mortgage payments during the COVID-19 crisis. While the impact has been widespread, Black and Hispanic and Latino borrowers have been disproportionately affected.1 In response, the federal government has provided a number of protections for borrowers, including offering forbearance—a lender’s agreement to not take action against a borrower for delinquency or default for a limited period of time. The US Treasury is also distributing about $10 billion through the Homeowner Assistance Fund (HAF) program, intended to help homeowners avoid defaults, delinquencies, and displacement through assistance with, for instance, mortgage, insurance, and utility payments.2
But using these funds is complicated. Like Emergency Rental Assistance, HAF could require many states to establish new programs from scratch or to scale programs that might have been much smaller prior to COVID-19. States may also struggle to determine where, and for whom, funds are most needed or what channels to best distribute them. States can accelerate and enhance the impact of HAF funding, in part, by taking a granular approach to designing their programs, including using data to target benefits and to guide important decisions about how to distribute funds, as well as by establishing appropriate channels to distribute relief. Here we outline several practices and ideas states may consider in order to help homeowners through HAF.
Using data to understand the challenge
States do not have access to perfect information, and it can be difficult to procure some of the data that would be most helpful for program design—such as the total amount of debt owed by eligible homeowners who are behind on their mortgages or the number of homeowners who can no longer afford their monthly payments. However, states can learn critical lessons from the data they can access, which can affect how they design their HAF programs.
Magnitude of need
A useful source of information for states is data from the Mortgage Bankers Association (MBA), specifically on delinquency rates over time. MBA data can help states answer a critical question: Are homeowners who are in distress in a state relatively new to experiencing delinquency, or have they been behind on payments for a long time? The answer to this question affects how the HAF program can be designed to best serve them.
For example, a national analysis of delinquencies on FHA loans (designed for low- to moderate-income borrowers) and conventional loans reveals a concerning scenario: the surge in delinquencies of more than 90 days in early 2020 had not yet returned to prepandemic levels by the third quarter of 2021 (exhibit). This trend indicates that many people who initially experienced distress due to the COVID-19 pandemic had already been delinquent for an extended period of time and were likely still in need of assistance. Rates of short-term delinquencies, however, had already returned to prepandemic levels or below. This means that the homeowners who require assistance from HAF are more likely to be in longer-term delinquency rather than to have recently become delinquent and, consequently, that they may require many months of payments.
States can perform this same analysis for their own populations to understand the specific characteristics of households in distress. Understanding that the typical homeowner in need might have accumulated many months of arrearages, for example, could inform states’ policy discussions around issues such as setting maximum household benefits. If states can perform more detailed analytics, it is also possible to use the millions of public loan-level records in Freddie Mac’s and Fannie Mae’s data service to estimate the implied monthly payments for the average loan that has been in delinquency for more than 90 days in a given jurisdiction. These data could help states make decisions about the HAF budget and options for benefit levels, as well as inform conversations about the role of loan modifications for individuals with a large amount of debt. Another option for states is to use internal loan data from their own housing-finance agencies as a proxy for overall homeowner needs, though it would be important to consider how these homeowners might differ from the overall pool—for example, if they are more or less likely to be at risk of delinquency.
Timing of need
Another way that data could help states design their programs is in determining the potential timing of the need for assistance. Analyzing data on delinquencies over time in combination with information on mortgage servicing can help states map how policies protecting homeowners might affect the timing of applications for HAF assistance.
For example, MBA data can provide a view of the timing of delinquencies early in the pandemic, when many of the potential homeowners still in need would likely have applied for forbearance. Given the timing of the pandemic’s onset and the surge in delinquencies, original forbearance applications may have occurred in most states at the end of the first quarter of 2020, with the second-largest tranche in the second quarter—though this may vary by state. Federal protections offer up to 18 months of forbearance to many homeowners with loans backed by federal entities, as long as the homeowners applied before a certain time.3 In states where this is the case, it is likely that many homeowners’ forbearance would be expiring around the end of the third quarter and throughout the fourth quarter of 2021. States can examine this information to determine when their populations will be most at risk, but they may also want to consider their own local protections, moratoriums, and foreclosure timelines to further specify when exactly homeowners might be at risk of instability.
In states where many homeowners applied for protection in early 2020, HAF programs may need to be ready up front for a potential influx of applications from homeowners exiting forbearance. This scenario raises several questions, including how to quickly scale up channels, what mix of channels to use, how to use triaging (based on, for example, potential foreclosure timelines) or waitlists to manage demand, and how to involve housing-counseling agencies to ensure homeowners can access additional protections offered through their servicers.
Using data to understand the magnitude and timing of need can help states understand what policy questions are most relevant to their constituents. Going further to build dynamic models of the impact of different decisions and scenarios can also help states test various potential policies and better prepare HAF plans.
Focusing on the ‘who’ and the ‘where’
Understanding who is most at risk and where homeowners are experiencing distress can help states target their outreach and support. The COVID-19 crisis has disproportionately affected people of color across every dimension, and housing is no exception.4 Data also show that FHA loans have been recovering more slowly than conventional loans and that a higher percentage of the FHA loan pool is still in forbearance.5 Our analysis of Home Mortgage Disclosure Act (HMDA) data shows that, historically, Black and Hispanic and Latino households have been disproportionately represented in the FHA origination pool compared to White households. States can plot historical FHA originations to show where concentrations of these at-risk households may occur. A geospatial lens has also been helpful for states to understand where services are most needed. States can use geographic data at whatever granularity is available to make plans for their own outreach to homeowners. For example, municipality-level data on concentrations of FHA originations or of relative levels of delinquency can help states prioritize which local governments to work with first on outreach and education for homeowners, including physical or digital advertisements, media campaigns, and partnerships with local organizations such as housing-counseling agencies and community groups.
Planning for multiple channels
States have several potential channels for distributing HAF funds. Three of these channel options are outlined below, including potential criteria to help make each approach a success, as well as important considerations for states contemplating establishing multiple channels to potentially reach a broader swath of eligible homeowners.
A retail approach
A retail approach is one in which homeowners can apply directly for funding. Some retail models reach homeowners through an application process run by a government agency, while others leverage community organizations as front doors, such as housing-counseling agencies or community-development financial institutions.
There are a few areas in which states may want to consider investing in people, processes, operations, and technologies to ensure successful delivery of a retail program. Critical considerations for this program include accessibility, user-centric design, and adequate staffing and tools. A rigorous performance-management approach—including regular data collection, common formats for applications, daily stand-ups, weekly calls and Q&A sessions, trainings, and office hours—for staff or any processing partners can greatly facilitate the operational aspects of managing a retail program. Investments in technology to process applications can also help, but states will likely want to ensure these investments do not add complexity or disrupt how any existing agencies involved in the process already work. Staff may struggle to learn and incorporate new technology quickly enough to be effective, so states may consider harnessing the systems and technology an agency already has while being mindful that a user-friendly process can ensure that homeowners in need successfully complete their applications. States may also consider speeding up the review process through flexible documentation and eligibility rules as outlined by the US Treasury.6 As an example, states could use geography or even occupation as a proxy for determining the eligibility of a borrower. States can examine the workability of different proxy approaches and determine the impact they would have on the review process.
As they consider establishing a retail HAF operation, many states may already be struggling to administer another large federal relief program: emergency rental assistance (ERA).7 Adding a HAF program can strain already-stressed capacity if it is administered by the same entities. To the extent that a HAF program might use infrastructure (such as technology or staff) that overlaps with ERA programs, states have been careful not to disrupt ERA operations, trying to find synergies with certain parts of the process and adding new net capacity where needed.
Another consideration when working to administer a retail program in a timely way is to account for operational impact when making policy decisions. For instance, HAF funds can be used for more than just missed payments, according to US Treasury policies. However, as states debate these options, they can consider lessons from the Hardest Hit Fund and other Great Recession–era interventions, which revealed how difficult it is to quickly and accurately underwrite homeowners to determine who may need assistance beyond making up for back payments, such as principal or interest deductions or other loan modifications.8
A servicer-driven approach
States could work with mortgage servicers to identify potentially eligible homeowners, conduct outreach, or cure arrearages. Servicers are the companies that handle the day-to-day tasks of managing loans, including sending statements and processing payments.9 Each state will need to decide how to manage and oversee relationships with servicer partners in distributing relief funds. The following are two potential options:
- Servicers could act as a funnel for a retail program for homeowners in need by determining who within their portfolios is behind on payments and fits the eligibility requirements of the program. Servicers could provide this information to the state or to community partners who could then reach out to homeowners, or they could send program and application information to homeowners directly.
- Servicers could apply directly to the state on behalf of homeowners. They could obtain permission from homeowners to share information with the state and solicit any documentation that they do not already have, such as attestation of hardship, that is required for an application. They could then work directly with the state to provide this information in bulk by bundling applications and documentation from multiple homeowners in a common format. Each state could determine what would be reviewed by which party. Homeowners could then have their arrearages cured directly through servicers, saving them the process of submitting applications and potentially providing them relief faster.
Arrangements with servicers could also be facilitated centrally (for example, by the US Treasury) rather than by each state arranging a separate agreement. A national approach may add scale and leverage beyond what states can do alone, as was the case with the Hardest Hit Fund.10 Some states have begun to show interest in servicer options in their draft plans; they may want to continue to explore this approach further with their fellow jurisdictions and with the US Treasury.11
An internal-portfolios approach
Finally, in cases where a state has an internal portfolio of mortgages, such as through housing-finance agencies, it may have an opportunity to cure arrearages directly for those portfolios. This approach allows states to target many of the homeowners who are most at risk as housing-finance agencies focus lending on lower-income and underserved households and communities.12 Similar to the second servicer approach described above, this channel could allow homeowners to become current on their mortgages without having to go through a retail application process.
A joint approach
While states may wish to focus on establishing a single channel to reduce operational complexity, launching multiple channels could increase the likelihood of reaching homeowners. Having a retail channel alongside a servicer channel, for example, would provide an option for each of two potential populations of homeowners: those who feel more comfortable seeking help directly from the state, and homeowners who might potentially find a retail application process overwhelming and prefer to have a servicer handle materials on the back end. Similarly, establishing an internal-portfolios approach—if a state believes it may be faster to launch than a retail or servicer channel—might allow states to get relief to a subset of homeowners quickly while they work to establish other channels.
If more than one approach is used, it will be important to ensure that states, administering agencies, and advocates clearly communicate how these programs work together so as not to confuse homeowners who could be eligible for help through multiple channels.
The US Treasury has opened its portal for HAF plan submissions. Some states may have submitted their initial plans and are now beginning the difficult work of establishing their programs and refining their analysis and approaches. Others may still be developing their HAF plans for later submission. As states either prepare, submit, or begin to execute and refine their plans, the above considerations may assist in thinking through how to use data and insights from other programs to refine their approaches for the greatest impact. As homeowners throughout the country potentially come to require support through the next phase of the crisis, states can use these insights to help those most in need.
