Who Can Actually Rent It

1. Who this housing is now for

The system was largely designed around working households whose incomes were expected to rise. It increasingly serves households whose incomes will not.

• Households with an older adult, meaning aged 62 or over, are 42 percent of all HUD-assisted households in 2024, up from 33 percent in 2014, on the same definition in both years. About 1.9 million such households receive assistance, against roughly 6 million eligible.¹
• A third of the poorest renter households are headed by someone 62 or older. There are about 11 million of these households nationally.²
• On a single night in January 2023, more than one in five people experiencing homelessness were aged 55 or over: about 98,400 aged 55 to 64 and 39,700 aged 65 and over, roughly 138,000 in total, measured against all people counted rather than adults only.³
• Assisted older households stay about nine years on average, against about four for families with children.⁸

When HUD proposed work requirements and time limits on assistance in February 2026, older adults and people with disabilities were exempt.⁹ That exemption is the clearest official acknowledgment available that for these households, assistance is not a stage on the way somewhere. It is where they live.

2. The homes exist. Someone else is in them.

There are 7.2 million rental homes in the country affordable to the poorest renters, meaning those at or below 30 percent of their area’s median income or the poverty line, whichever is higher, against 11 million such households. That is 65 affordable homes per 100 households, which is a shortage but not a catastrophic one.

Then look at how many are actually obtainable:

The difference is 3.4 million homes: affordable units occupied by households with higher incomes. Nearly half the homes whose rents are affordable to the poorest renters are lived in by households earning more than that.

That is not a story about waste or bad management. Those units are full and rent-paying. It is a description of who ends up inside them.

And it is invisible in every operating statistic a housing agency looks at. CohnReznick’s surveyed national portfolio reported median physical occupancy of 97 percent at the end of 2024.⁷ Nothing in that number tells you who is not there.

3. Two different gates, often mistaken for one

An applicant has to get past two things, and only one of them is written down.

The income ceiling is regulatory. It is the most a household may earn and still qualify, set as a share of the local median income and published by HUD every spring.⁴ It is a maximum. There is no matching statutory minimum: federal rules do not require a household to earn any particular amount to live in an income-restricted home.

The income test belongs to the landlord. Because rent in these homes is fixed by a published table rather than by what the resident earns,⁵ the property has to satisfy itself the resident can pay a fixed amount. It does that the way landlords generally do, by requiring documented income at a multiple of the rent. Income screens commonly use a multiple of the monthly rent. This brief models three scenarios, at three times, two and a half times and two times rent, rather than asserting a national norm, because no comprehensive source for prevailing multiples exists.¹⁰

This is the distinction that matters for policy. Being over the ceiling is an eligibility problem, created by regulation, and fixable by regulation. Failing the landlord’s income test is a private practice layered on top of a rent the rules had already fixed. They exclude the same household at different levels of the same system, and because no single body owns both, no one publishes the space between them.

Throughout this brief, figures described as a screen are calculated at a stated multiple. They are not a federal requirement and are not presented as one.

4. What this means for an older adult on Social Security

The 2026 average Social Security retired-worker benefit is $2,071 a month, $24,852 a year.⁶ At a three times rent screen, that income qualifies a rent of at most $690.33 a month. The cheapest 60 percent one-bedroom rent in the country is $315; the national median is $1,012.¹¹

How to read the table. An area counts as passing at a given level if the same household both sits at or below that level’s published one-person income ceiling and could rent that level’s one-bedroom under the stated screening multiple. Both conditions have to hold in the same area at the same level.

One person at that income, one-bedroom, across all 2,644 HUD income limit areas:

Read the first column and the second column as different problems.

The 30 percent row is an eligibility exclusion. In 92.7 percent of areas her income is above the published limit for the deepest homes. No screening assumption is involved. The average retirement benefit has risen past the deepest affordability tier in most of the country.

The 60 and 80 percent rows are a screening exclusion. She is comfortably under those ceilings almost everywhere. She cannot document three times the rent. In the 0.6 percent of areas where rents are low enough to clear a three times screen, the income limits are correspondingly low and she is over the ceiling instead, which is why the two columns leave nothing behind.

The screen is the variable, not the market. At two and a half times, some level opens in 51 percent of areas. At two times, 96 percent. Same person, same buildings, same published limits.

Rental assistance removes the problem this screen is measuring. With a Housing Choice Voucher the tenant contribution is generally income-based rather than a fixed amount, so there is no fixed rent for an income multiple to be applied against, and she is under the 60 percent income ceiling in 99.4 percent of areas.¹² Assistance does not exempt anyone from a property’s other admission criteria, such as credit, rental or criminal history screening.

One local comparison. In Multnomah County, Oregon, a single older renter in good health needs $38,100 a year to meet every basic need at market rent, according to the Elder Index.¹³ A property there applying Portland’s capped two and a half times screen would require $43,320 to admit her to the subsidized one-bedroom. The screened requirement is $5,220 more than the income she would need to live without the subsidy at all.

5. Why building more, by itself, does not fix this

The screen is not a policy anyone wrote. It is what ordinary underwriting does when it meets a rent that regulation has fixed in advance.

In programs where rent moves with income, there is nothing for a screen to test. Public housing from 1937, Section 202 housing for older adults from 1959, and project-based rental assistance all set the resident’s share at roughly 30 percent of income, with an operating subsidy covering the difference. You cannot be too poor for that structure. It was built for people with very little.

Then the engine changed. Congress repealed the construction part of Section 8 in 1983, and in 1986 created the Low-Income Housing Tax Credit, which has been the country’s main way of producing affordable homes ever since. The credit pays for construction, not for operations. With no operating subsidy, rent cannot follow each resident’s income. It has to be fixed by a table. And a fixed rent has to be tested against something.

The financing shift is documented; what we claim here is about the conditions it created, not about where tenant screening originated.

It also explains why new homes are built at 60 percent rather than 30. A rent set at 30 percent of area median income usually will not cover the cost of operating the building without an ongoing subsidy, and that subsidy is scarce. So the deepest affordability is the thinnest supply, and the level that does get built is the level whose rent produces the highest screening requirement.

The evidence that this is the binding mechanism is in HUD’s own tenant data:¹⁴ in HUD’s 2023 tenant data, 57.2 percent of reported tax credit households had incomes at or below the deepest tier, and 48.3 percent received monthly rental assistance. The poorest residents of these buildings are largely there because a subsidy came in with them. The building alone did not reach them.

Legislation signed in July 2025 made a 12 percent increase in the credit allocation permanent from 2026 and lowered the bond financing threshold from 50 percent to 25 percent, changes projected to finance roughly 1.22 million additional affordable homes over 2026 to 2035.¹⁵ Built the same way, at the same levels, those homes will be reachable by an older adult on a fixed income only where assistance is attached. Rental assistance currently reaches about one in four eligible households.

6. What does change it

Three things, and only three.

Attach assistance to the home. A voucher, or assistance tied to the building itself, makes the tenant contribution income-based, so a rent-to-income multiple has no fixed rent to bite on. Other admission criteria still apply. This works, it is well understood, and it is rationed.

Set some homes at a lower level. Since 2018 a building may put individual homes anywhere from 20 to 80 percent of the local median, as long as the building’s average stays at or below 60. Lower-designated homes carry lower rents and therefore lower screening requirements. This is a decision made building by building, and it is available now.

Cap the multiple. The multiple is a property decision almost everywhere. We identified caps in Portland, Oregon, at two and a half times rent, and statewide in Colorado at two times. We did not survey every state and local jurisdiction, so that is what we found rather than a claim about what exists. Moving a property from three times to two times rent opens a level to this household in 96 percent of areas instead of 8.

None of these require new construction. All of them decide whether construction reaches anyone.

7. What we already have, and what it would take

America built serious financial machinery to create housing. Tax credits, investor equity, permanent debt, insurance, guarantees, and regulatory credit for the banks that take part. Roughly $28.9 billion in tax credit equity closed in 2024 alone, about 80 percent of it from banks.¹⁶ That works because the federal government did not simply declare affordable housing valuable. It manufactured an asset, defined what counted as a unit of production, set the standard, and gave several parties a reason to show up.

There is almost nothing equivalent for using housing that already exists.

And unlike most claims about untapped capacity, this one has been measured. Our own research, The Rooms Already Exist, counted it from Census microdata:¹⁷

Every state and the District of Columbia holds more spare bedrooms in older adults’ homes than it has older renters in housing trouble. Not one state is short.

And the same finding names both households at once. Alongside those 4.5 million older renters are 7.7 million older homeowners, about one in four, paying more than they can afford to stay in homes they already own. That is 12.2 million older households under housing pressure, on both sides of the same front door. One arrangement addresses both: the renter gets a home she can afford, and the homeowner gets income that may be what keeps her in hers.

The economics, plainly

Building a new subsidized apartment is expensive, and it should be. The most recent published figures put total development cost in California at roughly $515,000 to $726,000 per unit for 2020 to 2023, depending on wage requirements.¹⁸ Costs vary widely by state and that is a high-cost example, but the order of magnitude is not in dispute: producing a new deeply affordable home is a several-hundred-thousand-dollar act of capital formation, and the resulting home, as sections 3 and 4 show, is one an older adult on Social Security usually still cannot rent without assistance attached.

Activating a room that already exists costs a fraction of that, and there is a public price on it. Portland’s home sharing pilot, launched in February 2026 with $500,000 in city funding, pays a home provider $1,500 for the first qualifying room and $750 for each additional room, after thirty days of successful rental through a qualified provider, with rent capped at $250 a week including utilities.¹⁹ The room needs no land, no construction financing, no thirty-year compliance period and no new heating system. It is already built, already heated and already standing.

That is the case for Housing Capacity Parity. Not that building is wrong: many households need independent, deeply subsidized, accessible or family-sized homes, and America needs substantially more construction than it is doing. The position is build, preserve, convert and activate. The argument is that a country which manufactured an entire asset class to finance the first three has left the fourth with no unit of measurement, no incentive, no risk infrastructure and no recognized event at which anything happens.

An arrangement between two people has neither of the gates in section 3. There is no published income limit to exceed and no landlord multiple to document, so nobody is turned away here for earning too much or for earning too little. It is the one form of supply this population is not screened out of on income, and it is the one the system does not count.

Verification still happens, and it is a different kind. Identity checks confirm that the person on the other side is a real person, and a comprehensive background check is made available to the party someone is considering sharing with rather than used by us to disqualify anyone. Those safeguards exist because a platform connecting older adults and people with low incomes would otherwise be attractive to fraud and predation. What is absent is the financial gate, not the protection.

What a shared arrangement does not carry is the rest of what regulation provides. There is no rent held down by a published table, no thirty-year compliance period keeping it there, and no guarantee of the independence, accessibility, or security of tenure a subsidized apartment is built to offer. It is a different form of supply rather than a cheaper version of the same one. That is precisely why the position is parity and not substitution: the argument is that this capacity should be measured, incentivized and financed on equal terms with new construction, not that it should stand in for it.

What parity would require is set out in the framework itself: a defined unit of production, an incentive for placing a room into affordable service, financing mechanisms that already exist being pointed at it, standardized risk and verification, and clarity on how room rental income interacts with means-tested benefits.²⁰ None of that is exotic. All of it was done once already, for construction.

8. What this means, depending on where you sit

If you set housing policy. All three levers in section 6 are already in your hands and none requires new construction money. The scoring rules your state uses when it awards the housing tax credit can favor buildings that designate homes below the 60 percent level. Waiting-list preferences for older applicants are already permitted, and a preference for local residents must treat someone who works in the area the same as someone who lives there.²¹ And a cap on the rent multiple a property may require is the single change that moves the household in section 4 from 8 percent of the country to 96 percent.

If you fund or finance housing. The landlord’s income test is why capital put into construction does not convert into access for this population at the rate the unit count suggests. A project that adds fifty homes at the 60 percent level adds fifty homes an older adult on Social Security usually cannot rent unless something else comes with them. Two better questions than the unit count: what income multiple will the manager apply, and how many of these homes carry rental assistance.

If you build or operate housing. You know this from lease-up already: applicants who want the home, need the home, and cannot document the income. That pressure rarely shows up as an empty building. It usually resolves the other way, as someone better off taking the home rather than the home sitting empty, which is why surveyed occupancy sits at a median of 97 percent while the households these homes were intended for are outside them.⁷ A full building is not evidence that the intended household was reached, and occupancy is the wrong measure to run a mission against.

If you fund organizations rather than buildings. This is the part with no financing machinery behind it, and it is where HomeShare America needs support.

The rooms exist. The households on both sides exist. What does not exist, and what no capital stack currently pays for, is the connective work between them: identity verification, background checks for both people, a matching process built on how two households actually live, agreement templates written for shared homes rather than borrowed from leases, and the support that carries a match past the first difficult month. That work is efficient precisely because most of the money in a home sharing arrangement reaches a household directly rather than a transaction chain, which is also why it generates no fee pool to fund itself the way development finance does.²⁰

We operate that platform through partners already on the front lines: Area Agencies on Aging, community action agencies, housing authorities and community development corporations who meet older adults at the moment housing is coming apart. Those partners do not need a new program. They need a functioning platform to refer into, and the platform needs operating support to be there when they do.

The proportions are the argument. Housing every cost-burdened older renter in America would take about 13 percent of the spare bedrooms older adults already own. Physical bedroom capacity, on paper, is abundant. How much of it can realistically be activated is the open question, and answering it is part of the work rather than something we claim to have settled. What is already clear is that the constraint is not a shortage of rooms. It is the absence of infrastructure to connect them, which is currently the only part of America’s affordable housing system asked to run on goodwill.

How these figures were checked

Every area-level figure comes from HUD’s published FY2026 tables, processed by script rather than transcribed. The build reads the HUD workbooks directly and emits one record per income limit area, failing rather than filling any gap. Maximum rents are derived from the statute for every one of the 2,644 income limit areas, and the national percentages in section 4 are sums over those same area-level records rather than separate estimates.

Those results are then checked against a second implementation that shares no code with the first. It re-reads the same HUD workbooks with its own parser, re-derives every rent from the statute and recomputes both gates independently. It reproduces all four levels to the tenth of a percentage point, with zero mismatches across all 2,644 areas, and the build stops if that ever stops being true.

The underlying income limit tables are published by HUD and are public. The method, the derivation and the full national results are set out in the companion technical brief, Between the Ceiling and the Floor.

This product uses HUD published data but is not endorsed or certified by HUD.

Companion work. Between the Ceiling and the Floor is the technical brief behind these figures, with an interactive tool covering every county in the country. Housing Capacity Parity sets out the policy framework this evidence points toward. The Value of a Match prices what home sharing is worth to both households.

Statement of interest. HomeShare America operates a home sharing platform, and sections 7 and 8 make a case for funding that work. We publish this research for one reason: home sharing is the one form of affordable housing capacity the country does not measure, incentivize or finance, and our intention is that it be recognized and treated on equal terms with new construction. We are not arguing that it should replace construction. The analysis in sections 1 through 6 describes the regulated system on its own terms and stands independently of anything we do.

Suggested citation. HomeShare America. (2026). Who Can Actually Rent It: income eligibility, tenant screening, and who America’s affordable housing reaches.

Companion work

Between the Ceiling and the Floor is the interactive tool behind these figures: set a household and an income and see both limits drawn for any county in the country, with a checklist of what to ask before applying. Housing Capacity Parity is the policy framework this evidence points toward, The Rooms Already Exist counts the spare bedrooms this brief cites, and The Value of a Match prices what one arrangement is worth to both households.

Notes and sources