Memo: Streamlined Rental Assistance Payment Demonstration

To:       Interested parties
From:  Will Fischer and Barbara Sard
Re:       Streamlined rental assistance payment demonstration

Policymakers are considering measures to sharply expand the housing voucher program, including President Biden’s proposal to make vouchers available to every eligible family.^[1] Vouchers are a highly effective form of rental assistance, but 3 in 4 eligible households receive no rental assistance due to funding limitations. Making vouchers available to more households would do more than any other single policy change to reduce homelessness, overcrowding, and housing instability and would reduce poverty and racial inequity substantially. At the same time, improving the system for delivering voucher subsidies — which has significant shortcomings — could make an expanded voucher program more efficient and effective at improving the well-being of low-income people.

This memo proposes a demonstration testing streamlined delivery of rental assistance through electronic payments administered by state agencies that administer housing vouchers or Supplemental Nutrition Assistance Program (SNAP) or other benefits for people with low incomes. (State agencies that don’t administer housing vouchers would in some cases partner with local public housing agencies that administer vouchers today.) The streamlined payments would be subject to fewer federal administrative requirements; for example, there would be no requirement that units be subject to third-party housing quality inspections or that landlords enter into contracts with housing agencies. Also, the new system would permit families that keep their rent and utility costs relatively low to retain the savings, subject to the restriction that they use subsidy funds only for housing costs.

The streamlined payments would maintain many core features that make the existing program effective at assisting the lowest-income families, including the policy of setting rent payments at 30 percent of a family’s adjusted income. Importantly, we are not proposing to test conversion of rental assistance to a cash benefit that recipients could use for non-housing purposes. That approach could have significant unintended downsides, including the risk that the benefits would count as income in SNAP or other programs, which would reduce their value considerably.

The policy changes we propose to test could have several significant benefits over the existing delivery system, although the demonstration would be crucial to determining their extent and identifying any tradeoffs. In brief, those potential benefits include:

• Making it easier for families to find a unit they can rent. Landlords may be more willing to rent to a tenant if they will receive a single monthly payment rather than having separate payment arrangements with public housing agencies (PHAs) and tenants (as occurs today) and if inspection and other federal legal requirements are eased. Greater landlord acceptance will be crucial to successfully implementing a major voucher expansion and would likely give households access to a broader range of neighborhoods.
• Reducing administrative costs. The changes proposed here could substantially reduce voucher administrative costs, which are several times higher than comparable costs in SNAP, such as by using income determinations from SNAP or other programs instead of costly and duplicative determinations by housing agencies, eliminating housing inspection costs, and achieving economies of scale by having large state agencies (rather than the current patchwork of over 2,000 state and local agencies) administer the program.
• Eliminating jurisdictional barriers to housing choice and facilitating partnerships with state-administered benefits. Giving states a role in administering benefits would enable families to move seamlessly across local jurisdictional boundaries without the often-cumbersome “portability” process used to shift assistance from one housing agency to another. It also could help states pair rental assistance with Medicaid and other services to address the needs of people who are homeless or have severe disabilities that make it difficult to live independently.
• Providing families with more simplicity and autonomy. Families could choose a unit to rent without needing agency approval of the unit or the rental charge. For example, a family could rent a unit for up 20 percent less than the maximum subsidy amount and use the remainder to lower their own payments for rent and utilities, which would free up resources for other needs and could reduce hardship. Or they could use the remainder to save for future housing costs, such as a security deposit on a new unit or a downpayment.

A five-year demonstration could be conducted at the same time the existing voucher program is scaled up. Its findings could inform adoption of some or all of the proposed policy changes for existing vouchers (including those that would still be administered by local public housing agencies) as well as for further expansions toward the goal of providing vouchers to all eligible families that want them. The more families that receive voucher assistance, the more important it will be to identify improved administrative mechanisms and policies.

That is partly because sharply expanding the voucher program would raise the stakes and make it more crucial to examine how best to deliver assistance. In addition, a larger voucher program would make state-administered alternative housing subsidies more feasible and desirable by expanding the group of households that receive both rental assistance and state-administered benefits like SNAP. A full entitlement would add to the potential advantages of alternative subsidies, since it would eliminate the need for waiting lists and local admissions preferences, which are one rationale for local voucher administration. A full entitlement also would make it easier to push to upgrade all substandard rental housing rather than focusing on inspecting individual voucher units, since all low-income people would be able to afford standard-quality units.

Potential Benefits, Risks of Distributing Voucher Assistance by State-Run Electronic Transfer to Families

The Housing Choice Voucher (HCV) program serves families through more than 2,000 state and local PHAs, often with overlapping jurisdiction. A streamlined system could rely on state agencies to distribute rental assistance directly to families via electronic benefit transfers (EBT), similar to other federally supported benefits to low-income families. About 30 states already have state-level agencies that administer housing voucher programs for part or all of the state, and every state has a state agency that administers SNAP. To realize the greatest efficiencies from a state-administered housing voucher program using EBT — and to adhere most closely to regular housing market practices — landlords would have a legal relationship only with their tenants, unlike the current HCV program where landlords also have a separate contract with a public housing agency.

Similar to SNAP, in which state agencies load benefits monthly on families’ EBT cards for food purchases at grocery stores or other locations, administering agencies could draw on federal funds to deposit the rental benefit amount to a family’s EBT card or account.^[2] Families would then use the card (or funds in their electronic account) to pay the full rent — the subsidy plus the tenant share — to their landlord rather than agencies making subsidy payments to landlords. Such payments could be made online to landlords’ accounts. (Small landlords may have to establish accounts for this purpose, similar to how small retailers have adapted to use of electronic payment systems.)

Though benefits would be paid directly to families under the demonstration, they could only be used for monthly rent and utilities or other housing-related expenses, just as SNAP benefits are restricted to use for food. Families would generally be required to spend at least 80 percent of the maximum subsidy level on regular rent and utility costs, to discourage families from renting units that are in poor condition or too small for their health and safety. Families could use any remaining funds for other housing-related costs (such as future security deposits to facilitate a move), for rent or utility payments if they have an unanticipated increase in utility costs, or for a downpayment on a home purchase. They could also use the remaining funds to reduce their own payment toward rent and utility costs (normally about 30 percent of their income) and use the savings for other, non-housing purposes.

The demonstration would not test conversion of benefits to unrestricted cash or allow any rental assistance funds to be used for non-housing purposes. While unrestricted cash would give families even more flexibility, the risks would outweigh that benefit. A cash benefit would very likely count as income for SNAP and other programs (which exclude housing vouchers from income calculations today because they are only used for housing), reducing a family’s benefits in those programs to a degree that would offset much of the value of the rental assistance. SNAP benefits alone typically fall by about 30 cents for every dollar of income once a family’s income is above a very low level. Even if cash benefits under a demonstration project did not count toward other programs like SNAP, if unrestricted payments to families were ultimately adopted it would be very hard to maintain this protection in other means-tested programs, with the real risk that the effective value of the subsidy would be reduced substantially. In addition, conversion to cash could potentially reduce landlord participation, since a major benefit of renting to voucher holders is that they receive a reliable subsidy that must be used to pay the rent, and undermine political support for the program.

The EBT system we propose would use policies similar to the current housing voucher program to determine eligibility and benefit amount, but streamlined to facilitate state agency administration and reduce administrative costs. Electronic transfers would be made monthly based on the difference between HUD’s standard for the maximum rental subsidy for the family’s size in the zip code where the family’s rental unit is located^[3] and 30 percent of the family’s income after deductions, as determined periodically by the administering agency. (This is the same required contribution to rent and tenant-paid utilities as in the HCV program.) Administrators could rely on income determinations by other means-tested federal benefits, such as SNAP, TANF or SSI, to further simplify program administration.

This streamlined system would substantially reduce administrative costs. For example, the federal government paid state and local housing agencies an average of $854 in 2020 for each voucher in use to cover their administrative costs,^[4] more than double the annual administrative cost per SNAP case.^[5] SNAP administrative costs largely reflect states’ procedures for determining families’ eligibility and countable income and transferring the resulting benefit amount to EBT cards; similar steps — and the economies of scale realized through administration by state-level agencies — would apply in the proposed streamlined housing voucher program. The streamlined approach also would eliminate many costly, additional components of current housing voucher administration, including the time and paperwork required to: issue vouchers to families before they search for housing, enter into a contract with a private landlord on behalf of each individual family, inspect families’ rental units periodically to determine compliance with quality standards, and implement cross-agency procedures that apply when families move to a different agency’s jurisdiction. States could achieve additional savings by making the many families that also receive SNAP or other federal low-income benefits eligible for housing voucher benefits automatically, and by relying on income determinations from SNAP or other programs instead of having housing agencies conduct costly and duplicative determinations.^[6]

Such an EBT system could also make housing vouchers more effective for families. In many parts of the country, particularly in lower-poverty areas, it is very challenging for families to find landlords that accept housing vouchers;^[7] nearly a third of families issued vouchers don’t succeed in using them, on average.^[8] By making the rental process nearly identical to that in the unsubsidized rental market, taking a government agency out of the landlord/tenant relationship, and eliminating local jurisdictional barriers to families’ ability to choose where to live, the proposed streamlining could help families use their subsidies to rent units of their choice.^[9] It also could give more families the option of living in a wider range of neighborhoods, including neighborhoods with better schools and other features that provide greater opportunities, and thereby reduce racial and economic segregation.^[10] In addition, families may prefer the relative simplicity and autonomy of making their own decisions about what unit to rent without the requirement of a PHA’s approval of the home’s condition and rent amount.

To streamline the delivery of housing voucher benefits in this way, however, would modify or eliminate some features of the current HCV program that are intended to help families. Most significantly, PHAs would no longer verify that families’ units comply with federal minimum standards concerning housing conditions and adequate size for the number of family members, or play a role in enforcing these requirements by withholding subsidy payments. (See box, “Rental Subsidy Expansion Could Reframe Approach to Improving Housing Conditions.”) Nor would families get the benefit of federally required legal protections in the landlord-tenant relationship, such as the requirement of a written lease (usually for 12 months initially). In addition, some families would receive smaller subsidies than under current program rules (though some would receive larger subsidies) due to the simplified determination of the maximum housing benefit amount, which would require the use of zip-code-level rent maximums and eliminate local flexibility to adjust subsidy caps.^[11] And while eliminating bureaucratic procedures that discourage some landlords from accepting housing vouchers may increase some families’ housing opportunities, other landlords may be newly deterred by loss of the security that PHAs’ direct payment of rent subsidies provides.^[12]

A demonstration could help determine whether adopting a streamlined EBT system would realize the anticipated benefits, before making such significant changes in the administration of expanded housing voucher benefits program-wide, and identify any negative impacts on housing safety or stability. A demonstration also could explore if there are ways to minimize the loss of tenant protections resulting from the administrative redesign. And it could function as a pilot to work out various technical issues that would need addressing before launching such a program on a large scale.

Research Demonstration Design

We recommend that a research demonstration be conducted in at least eight sites, selected to represent the likely major drivers of potentially varying impacts in areas with different housing market conditions, as discussed below. In each site, 2,000 households, drawn from public housing agency HCV waiting lists, would be enrolled: 1,000 for the “treatment” group, which would receive streamlined EBT benefits, and 1,000 in the control group, which would receive regular HCV assistance. The demonstration would last for five years: one year for design and piloting of the interventions,^[13] three years for random assignment and implementation, and a fifth year for preparation and publication of the final evaluation report. We estimate the demonstration would cost $598 million: $578 million for three years of housing benefits and related administrative costs and $20 million for design, piloting, and implementation of the new treatment at eight sites plus the evaluation.^[14]

Vacancy Rates, Housing Stock Condition and Safety Standards, and Prevalence of Racial and Ethnic Segregation Likely Affect Intervention’s Impact

Decades of experience with housing vouchers have shown that families have more difficulty using vouchers to rent units in low-vacancy (“tight”) housing markets; research also shows that Black and Latino families are less likely than white families to be able to use vouchers to rent homes in areas where they constitute a small racial or ethnic minority.^[15] As explained above, there are many reasons to anticipate that state-level administration of simplified housing benefits would encourage more landlords to rent to recipients of federal housing assistance and would reduce the barriers families often face to moving to new communities with greater opportunities. But to test whether these changes in administration and program design can reduce or eliminate those challenges, the demonstration should select sites that vary on vacancy rates, housing conditions and standards, and prevalence of segregation.^[16]

It is also vital to determine whether such program simplification — particularly the elimination of agency-supervised federal requirements concerning housing conditions — would leave families living in worse quality or more crowded housing than under current program rules. Areas of the country differ in the extent of defective conditions in the lower-cost rental housing stock, and in the strength of rental housing safety standards and compliance mechanisms. To help determine whether these aspects of housing market conditions can effectively substitute for HCV program requirements, the demonstration should also select sites that vary significantly by the quality of the lower-rent housing stock and the legal environment regarding enforcement of housing quality standards and tenant rights.^[17]

Therefore, to yield the most policy-relevant results, the demonstration should be designed to produce findings from areas with substantial differences in the three sets of housing market-related features, as the table below shows. These key variables would require implementing the demonstration in at least eight different types of locations.

We recommend enrolling 2,000 households at each site in order to yield statistically significant findings when impacts differ by 5 percent or more. (Significant differences among sub-groups would have to be somewhat larger to be detected.) Funding the full estimated cost of the 16,000 rental subsidies for the three-year implementation period from random assignment would assure that the demonstration could be completed within five years from selection of the participating sites and evaluation team.^[18]

To administer EBT-delivered streamlined rental subsidies as well as the regular housing vouchers for the control group most efficiently, the demonstration should require collaboration between the state agency administering SNAP and a state-level public housing agency that administers HCVs.^[19] Involving SNAP administrators in the demonstration would bring their expertise in using EBT to deliver funds for time-sensitive expenses for low-income families; it also would facilitate using SNAP income determinations for participants in the demonstration, most of whom likely receive SNAP. In states without a state-level HCV program, one or more local public housing agencies could apply jointly with the SNAP administrator. In either case, the EBT rental subsidies would be made available to families initially residing in the same geographic areas as the control group.