Overview 

The United States Department of Housing and Urban Development (HUD) posted a notice of proposed rulemaking on March 2, 2026 (Docket No. FR-6520-P-01), entitled, “Establishing Flexibility for Implementation of Work Requirements and Term Limits,” hereafter the “Proposed Rule.” The Proposed Rule would allow, but not require, eligible Public Housing Agencies (PHAs) to implement work requirements and time limits for non-disabled working age recipients of Housing Choice Vouchers, Project-Based Vouchers and public housing; and allow the same flexibility for certain Owners of Project-Based Rental Assistance developments. Only PHAs that are not in receivership and are not designated as troubled performers, and only Owners that are not in default and meet other criteria, would be eligible. 

In this comment, I first describe the two biggest problems with federal rental housing assistance—inequitable treatment of low-income families and discouragement of upward mobility. Deep subsidies and lengthy spells make federal rental assistance highly inequitable, providing a typical recipient household close to a quarter million dollars of subsidies over their total benefit spell while similar households receive nothing from this program. Federal rental assistance discourages upward mobility by imposing an implicit 30% tax on earnings that exacerbates already high implicit taxes from other benefit programs and the tax code, which weakens the financial benefits of work and marriage. I then argue that the Proposed Rule would substantially mitigate both of these problems, making federal rental housing assistance more equitable and encouraging upward mobility. I conclude by offering specific recommendations to further improve the Proposed Rule.

Federal rental assistance is deeply inequitable 

The federal government spends over $60 billion annually on rental assistance.¹ Among all non medical in-kind transfer programs, federal rental assistance is the second most costly following  the Supplemental Nutrition Assistance Program (SNAP), which allocated $95 billion in benefits  in fiscal year 2025.²It also substantially exceeds spending on the Temporary Assistance to Needy Families program, which totaled $35 billion in aid from federal and state governments in  fiscal year 2024.³ 

Despite the large amount of spending on federal rental assistance programs, only about one in four eligible households actually receives the benefit.⁴ That is because, unlike SNAP, rental housing assistance is not an entitlement, and so when available funds are exhausted, many eligible households are excluded from assistance. Many PHAs operate waiting lists that require  eligible households to wait multiple years to begin receiving benefits, and in other areas waitlists  are closed altogether to new households.⁵ Proposals to turn rental housing assistance into an  entitlement, granting aid to all eligible households, would be prohibitively expensive given the  already high cost of serving only one in four eligible households. As of 2024, 4.4 million households received rental assistance from mainstream federal programs compared to 22.1 million households receiving SNAP.⁶ Among the federal rental assistance programs, the largest is Housing Choice Vouchers (2.3 million households), followed by Project-Based Rental Assistance (1.3 million households), and public housing (0.8 million households).⁷ 

A direct implication of high spending on and relatively low enrollment in rental housing assistance is that subsidies are deep. In 2025, the average federal subsidy for Housing Choice Voucher recipient households was $1,255 per month, 3.6 times larger than the $353 average monthly SNAP subsidy to assisted households.⁸ And unlike SNAP which provides consistent maximum benefit allotments across all states (except for Alaska, Hawaii, and federal territories), federal housing subsidies vary widely. As an extreme example, a family who lives in a three bedroom apartment in Washington, DC is eligible for a Housing Choice Voucher of up to $5,301  per month, or $63,612 per year.⁹ Meanwhile, in Jefferson County, Alabama, the maximum  federal subsidy for a three-bedroom apartment tops out at $2,090 per month, or $25,080 per year. Although housing costs are higher in Washington than Jefferson County, the higher cost reflects other characteristics of the area (e.g., public transportation, school quality, job availability, etc.) that are effectively subsidized by the larger benefit amounts in Washington. Given research showing that low-income families with the same income levels are no better off in low-cost areas than high-cost areas, it is not clear why families in high-cost areas should receive more generous subsidies.¹⁰ 

These characteristics make federal rental housing assistance deeply inequitable in a given year. Most eligible households receive no subsidy whatsoever. Those who receive rental assistance obtain an average federal subsidy worth over $15,000 per year, which in extreme cases can reach over $60,000 in the highest cost areas. 

Rental assistance is even more inequitable once accounting for the typical length of receipt. In  recent research, Howard Husock and Bruce Meyer use administrative data on the universe of  federal rental housing assistance recipients (about 4.5 million households) to estimate typical  spell lengths.¹¹ Figure 1 reports their key result: Among all non-elderly, non-disabled tenants  receiving rental assistance, 63 percent of spells were expected to last over 10 years, and 47% were expected to last over 15 years. 

A household receiving the average monthly benefit of $1,255 for the median duration of about 15 years would receive a total subsidy of $225,900. For context, a family receiving no subsidy would need to work at a $15/hour job full time for more than 7 years to earn that same amount.  A household in Washington DC renting a three-bedroom apartment for 15 years could receive a total subsidy of up to $954,180, the equivalent of working full time for 27 years at DC’s $17.95/hour minimum wage.¹²

Figure 1. Share of ongoing spells in federal rental assistance programs expected to last for various amounts of time

Source: Husock, Howard, and Bruce D. Meyer. 2025. “Subsidized Housing and Upward Mobility.” National Affairs 67. Fall. Available at https://www.nationalaffairs.com/publications/detail/subsidized-housing-and-upward-mobility; Author’s calculations 

Notes: Programs include public housing, tenant-based rental vouchers, and project-based rental vouchers.

Federal rental assistance discourages self-sufficiency 

Maximum benefit amounts for federal rental assistance programs vary by location, as PHAs set payment standard tied to the Fair Market Rent in an area, typically the 40^th percentile of market rents. After applying a limited set of deductions, a household must contribute 30% of its monthly income toward the rent, and the federal government pays the remainder.¹³ This effectively means that for every additional dollar earned by a household, it must pay an additional 30 cents toward rent and so the federal subsidy falls by 30 cents. This is equivalent to a 30% tax on earnings.  

As a means-tested program, it is necessary to phase out benefits as a household’s income rises, reflecting their greater ability to pay and reduced need for government assistance. However, a 30% implicit tax on earnings is a substantial disincentive to earn more (or potentially to earn more but shield it from reporting to program administrators), especially when considered on top of the implicit and explicit taxes the same family already faces. SNAP imposes an implicit tax of around 24% or more for families after deductions have been applied. The Earned Income Tax  Credit (EITC) phases out at a rate of 21% for a single parent with two children after reaching  earnings of $23,890.¹⁴ The same family sees their Additional Child Tax Credit (ACTC) phase  out at a rate of at least 10% once attaining earnings of around $42,000, depending on their tax  situation.¹⁵ They also must pay a federal payroll tax of 7.65%. Thus, a family receiving housing assistance and SNAP can face an implicit tax rate exceeding 90% when all benefits are phasing out simultaneously. It is entirely possible for families to face implicit tax rates exceeding 100%, especially if they receive benefits from additional programs such as Supplemental Security Income, heating and energy assistance, childcare assistance, and other state benefit programs, not to mention state income taxes and tax credits. They could also risk losing Medicaid coverage.  

Figure 2 reports the take-home pay for a single parent with two children who in addition to refundable tax credits receives only SNAP and a Housing Choice Voucher (red line). A shallow slope indicates that take-home pay increases only by a small amount when earnings rise. For earnings between $26,000 and $66,000, the family’s implicit tax rate exceeds 50% for any $1,000 increase in earnings. For earnings between $43,000 and $49,000, the implicit tax rate exceeds 100%, meaning they are made financially worse off by increasing their earnings within this range. Because this figure excludes state income taxes and other benefit programs, it understates implicit taxes. Without housing assistance, implicit tax rates would be less extreme (black line).

Figure 2. Post-tax post-transfer income by earnings with and without housing assistance, single parent with two children, 2026

Source: U.S. Department of Agriculture; U.S. Department of Housing and Urban Development; Internal Revenue Service; Author’s calculations 

Notes: Post-tax post-transfer income subtracts federal income taxes and the employee portion of the payroll tax, adds Supplemental Nutrition Assistance Program (SNAP) benefits, and in the case of the red series adds benefits from the Section 8 Housing Choice Voucher program. Taxes and benefits are simulated for a single parent with two dependent children, with a monthly rent of $2,000, assumed less than the payment standard in the area. All income is accrued from earnings. State income taxes are not modeled. Taxes are simulated based on tax year 2026. SNAP and Section 8 Housing Choice Voucher are simulated based on rules in effect in April 2026. For SNAP, family takes the standard deduction and excess shelter cost deduction. For housing assistance, family takes the annual dependent deduction of $500 for each child.

Figure 3 indicates that among all benefit programs shown modeled in Figure 2, housing assistance is the largest driver of high implicit tax rates. It provides the deepest subsidy, and after applying dependent deductions phases out at a 30% rate until it completely phases out. As a result, housing assistance increases the implicit tax on a family by 30 percentage points until earnings exceed $80,000. Housing assistance arguably poses the worst poverty trap among all means-tested programs targeted to non-disabled working-age adults with children. Empirically, housing assistance recipients do not show signs of upward mobility. As previously noted, the average nondisabled working-age recipient will remain dependent on the program for almost 15 years. In addition, just 8.6% of assisted households were married couples as of 2021.¹⁶

Half of non-disabled working-age recipients do not work at all, and half of those who work earn less than $20,000.¹⁷

Figure 3. Post-tax post-transfer income and benefit payments by earnings, single parent with two children, 2026

Source: U.S. Department of Agriculture; U.S. Department of Housing and Urban Development; Internal Revenue  Service; Author’s calculations 

Notes: Benefits are simulated for a single parent with two dependent children, with a monthly rent of $2,000, assumed less than the payment standard in the area. All income is accrued from earnings. Earned Income Tax Credit (EITC) and Additional Child Tax Credit (ACTC) are simulated based on tax year 2026. SNAP and Section 8 Housing Choice Voucher are simulated based on rules in effect in April 2026. For SNAP, family takes the standard deduction and excess shelter cost deduction. For housing assistance, family takes the annual dependent deduction of $500 for each child.

Causal evidence indicates that receipt of housing assistance reduces work and increases single parenthood. Brian A. Jacob and Jens Ludwig find that randomly assigned Housing Choice  Vouchers from a waiting list in Chicago decreased employment by 4 percentage points and decreased annualized earnings by $1,316 (obtained by multiplying the reported quarterly effect  by four), among non-disabled working-age adults.¹⁸ A HUD-sponsored randomized controlled trial found that vouchers decreased employment among welfare-recipient families in the short  run, an effect which faded over time, and increased single parenthood in the long run.¹⁹ Another HUD-sponsored randomized controlled trial offering vouchers to homeless families significantly decreased employment after 20 months while the difference was not statistically different after 37 months. Separations between spouses and partners increased after 37 months.²⁰

Time limits and work requirements would make federal rental assistance more equitable and promote self-sufficiency 

Given a fixed amount of Congressionally appropriated funding for federal rental assistance, making it more equitable requires reducing the amount of benefits to those already receiving assistance and using the savings to serve additional households. One way to do so is to reduce the annual benefit level paid to each existing recipient household.²¹ 

Another way is to limit the duration of housing assistance receipt so that the program becomes more equitable over time. For example, rather than subsidize one household for 15 years, three households could be subsidized for five years each, or five households could be subsidized for three years each. Time limits could dramatically expand the reach of federal rental assistance programs over time and greatly reduce the amount of time families must remain on the waiting list, potentially better targeting assistance to households when they most need it. 

Work requirements would also contribute to making federal rental assistance more equitable. To the extent that work requirements increase earnings, the housing subsidy will automatically shrink for existing recipients, freeing up funds for new recipients. To the extent that work requirements do not lead to increased employment, sanctions on existing recipients would also reduce their benefit and free up funds for new recipients. Thus, regardless of the effect of work requirements on employment, they are likely to make housing assistance more equitable by expanding the pool of recipients. 

Time limits and work requirements would also lead federal rental assistance to better promote upward mobility. Time limits would reduce the number of years during which families face an especially high implicit tax on earnings, reducing their implicit tax rate by 30 percentage points in years after the time limit expires when they otherwise would have been enrolled in housing assistance. Even while receiving housing assistance, families may plan ahead by seeking out employment before the time limit hits, minimizing the risk of financial disruption when transitioning off benefits and preserving the possibility of accessing housing assistance in the future. 

Work requirements would incentivize the half of non-disabled working-age recipients of housing assistance who do not currently work to enter the workforce. Those who work fewer hours than the number ultimately established as the minimum would be incentivized to work more. A work requirement substantially improves work incentives because it changes a work penalty to a work reward. Rather than lose 30 cents per dollar earned, a recipient would gain some or all of the housing subsidy as a result of working a sufficient number of hours.

Time limits and work requirements were successful during the 1990s, in combination with expansions of the Earned Income Tax Credit, in encouraging upward mobility. The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 reformed the Aid to Families with Dependent Children (AFDC) cash welfare program. Like housing assistance, AFDC subjected low-income families to high implicit taxes on earnings, though typically even higher.  AFDC was renamed Temporary Assistance for Needy Families (TANF) and the new program imposed time limits and strengthened work requirements. Cash welfare caseloads fell, employment among single mothers increased dramatically, and single parenthood stopped growing. While there is disagreement regarding whether these successes were more a result of time limits and work requirements for cash welfare or expansions of the Earned Income Tax Credit (in combination with the strong late-1990s economy), it is worth noting that both reforms ultimately increased work incentives and reduced rewards for non-work, as establishing time  limits and work requirements for housing assistance would do.²²

Comments on specific aspects of the Proposed Rule 

The specific provisions of the Proposed Rule as drafted are generally well-crafted and appropriate. The Proposed Rule would allow PHAs and Owners of project-based assistance to implement time limits of two years or more and work requirements of up to 40 hours per week for nondisabled working-age (18-61) adults. PHAs and Owners could choose to implement both time limits and work requirements, just one or the other, or neither. Individuals in college or caring for a disabled individual or for a child age 6 or under must be exempt from work requirements, though PHAs and owners could exempt other populations as well. Hardship exemptions would be required. 

Throughout, the Proposed Rule strikes a reasonable balance between promoting maximum flexibility for PHAs and Owners and still providing guardrails to mitigate risks from major policy reforms. And because all changes are completely discretionary, any PHA or Owner not willing or prepared to enact the reforms can maintain the status quo. While extending time limits and work requirements throughout the nation would improve housing assistance programs by making them more equitable and encouraging of upward mobility, it is useful to initially roll out changes more gradually to assess their effectiveness. Ideally the evidence from PHAs and Owners who decide to implement time limits and work requirements would be used in consideration of future regulatory or legislative changes to federal rental housing assistance programs.

The Proposed Rule, as is, would be a major improvement to federal rental housing assistance programs. Nonetheless, I provide three minor recommendations below for consideration as HUD drafts its final rule:

Recommendation 1: Reference state/local minimum wages rather than the federal minimum wage for purposes of determining compliance with work requirements. 

When characterizing the allowable work activities for a work requirement, the Proposed Rule states: 

A PHA or Owner would be able to determine whether the individual has met the work  requirement, for example, if they have worked the minimum number of hours in self employment or if they earn weekly wages equal to the Federal minimum wage multiplied  by the minimum number of hours in the PHA or Owner’s work-requirement policy. 

It may be more appropriate to reference the state or local minimum wage instead of the Federal minimum wage, to better reflect the labor market conditions and expected pay in the recipient’s local market. According to the National Conference of State Legislatures, 34 states, territories and the District of Columbia had a higher minimum wage than the federal minimum of $7.25 per hour, as of January 5, 2026.²³ 

Recommendation 2: Clarify that PHAs and Owners may set different work hour  requirements based on the number and roles of work-eligible adults in the family. 

When specifying the conditions for determining the work hours under the work requirement, the Proposed Rule states:

PHAs and Owners would be able to specify the number of hours per week a work-eligible adult must engage in work activities, except that a PHA or Owner could not require any individual to participate in work activities for more than 40 hours per week. 

From the Proposed Rule, it is clear that the PHA and Owners can determine which work-eligible adults are subject to the work requirement, but it is not clear whether PHAs and owners can select different hour requirements for different adults. For example, it may be worth explicitly allowing for say a 20-hour requirement for the head of household and a shorter 10-hour work requirement for a second adult in the same household. 

Relatedly, the Proposed Rule already allows for PHAs and Owners to calculate work hours at the  family rather than individual level, which is sensible. It may be worth explicitly allowing PHAs  and Owners to set different family hour requirements depending on the number of work-eligible  adults in the family. For example, a PHA or Owner may wish to set a minimum of 20 hours for a  family with one work-eligible adult, while setting a minimum of 30 hours for a family with two  work-eligible adults. 

Recommendation 3: Allow PHAs and Owners to set lifetime term limits, subject to statutory authority. 

In specifying the flexibility allowed under term limits, the Proposed Rule states: 

HUD notes that the proposed rules do not propose any additional eligibility criteria for admission to the program and consequently families or households which exit housing after hitting a term limit may then reapply for housing assistance. However, the household would have to go through the PHA or Owner’s waiting list and selection process in order to be readmitted. 

This makes it clear that the term limit must be applied per spell, ruling out term limits that apply over the lifetime. HUD should reconsider this aspect of the Proposed Rule and consider allowing for lifetime limits, subject to statutory authority to make this change. A lifetime limit would allow tenants to bank some of their assistance for later when facing future negative financial or familial shocks, strengthening the incentive to move out of housing assistance before the time limit expires. In addition, a lifetime limit would better reduce inequities in the program by ensuring more families can receive assistance at some point in their lifetime. HUD could allow PHAs and Owners to establish a lifetime limit (e.g., 5 years) that exceeds the length of a spell limit (e.g., 2 years), as well as allow them to establish only a lifetime limit without specifying an individual spell limit. This local flexibility would improve the Proposed Rule by catering these decisions to local context, such as existing unmet demand for rental housing assistance.