Society · Housing affordability · September 10, 2026

Affordable apartments sit empty while America’s poorest renters remain shut out

A growing mismatch is exposing a flaw hidden inside the word “affordable”: homes can satisfy a government income formula, compete poorly with new market-rate apartments, and still cost far too much for people living at the bottom of the income ladder.

7.2Mnational shortage of affordable and available homes for extremely low-income renters

35affordable and available homes for every 100 extremely low-income renter households

74%of extremely low-income renters spending more than half their income on housing

3 in 4eligible renter households estimated to receive no federal rental assistance

The strangest image in America’s housing crisis is not an overcrowded shelter or a block of luxury towers. It is an empty apartment carrying an “affordable” designation while a renter with almost no money sleeps in a car, a shelter or a temporary room a few miles away.

That contradiction has become visible in fast-building cities where apartment supply has expanded, market rents have softened from their peaks and some income-restricted properties are having trouble filling units aimed at moderate-income tenants. An Associated Press review published this week found more than 4,500 apartments classified as affordable sitting vacant in Austin, Texas, a vacancy rate near 16%. Denver and Portland, Oregon, also reported elevated vacancies in some higher income bands of subsidized housing. At the same time, the country remains desperately short of homes that the poorest renters can actually pay for.

The two facts do not cancel each other out. They describe different slices of the rental market. The National Low Income Housing Coalition estimates that extremely low-income renters — households generally at or below the federal poverty guideline or 30% of area median income, whichever is higher — face a shortage of 7.2 million affordable and available homes. Only 35 such homes exist for every 100 extremely low-income renter households nationwide, and no state has enough.

The problem is therefore not simply “too few affordable units.” It is a targeting problem, a financing problem and, increasingly, an administrative problem. America has built millions of apartments under programs meant to hold rents below specified ceilings. But the formulas that make a project financially feasible often place its rents above what a household at 20% or 30% of local median income can bear without an additional subsidy. When ordinary market rents fall close to those restricted rents, a second problem appears: the regulated unit may demand far more paperwork while saving the tenant very little money.

An empty unit can be “affordable” under a program formula and still be out of reach for the lowest-income renter. The label describes a rent or income restriction, not a universal promise that every low-income household can pay the rent.

01

Two different shortages are hiding inside one word

Affordable is not the same as deeply affordable

Housing policy uses income bands because costs and wages vary dramatically from one metropolitan area to another. HUD calculates local income limits from median family income and other adjustments. A household at 80% of area median income may qualify for one program; a household at 50% may qualify for another; and “extremely low income” is generally tied to 30% of the local benchmark or the federal poverty guideline, subject to federal rules. These are not interchangeable groups.

That distinction matters because rent ceilings in many subsidized developments are designed around an assumed income level rather than around each tenant’s actual paycheck. In the Low-Income Housing Tax Credit program, the nation’s largest production engine for income-restricted rental housing, traditional project tests generally require a share of units to be occupied by households at or below 50% or 60% of area median gross income. The newer average-income option permits a mix of designated income levels from 20% through 80%, but the project average must meet the federal limit. State housing finance agencies then use allocation plans, subsidies and scoring rules that can push projects toward deeper affordability — or leave most units clustered around levels that make the development easier to finance.

Extremely low incomeRoughly the bottom of the ladder

Federal housing rules generally anchor this group around 30% of local median income, with a poverty-guideline floor and other adjustments. These renters face the largest absolute shortage.

50%–60% AMIThe tax-credit center of gravity

Many LIHTC units are financially structured around these bands. Their rents can be reasonable for a moderate low-income worker yet impossible for someone living on a much smaller fixed income.

Market rateA moving competitor

Market rents rise and fall with supply and demand. When a construction boom creates concessions and lower asking rents, restricted units can lose their price advantage.

The National Low Income Housing Coalition’s 2026 analysis makes the imbalance stark. Nearly one-quarter of U.S. renter households are extremely low income. Eighty-seven percent of those households spend more than 30% of their income on housing, and 74% spend more than half. That level of cost burden leaves little room for food, transportation, medicine, utilities, child care or an emergency expense. It also means that a rent which looks discounted on a property’s leasing sheet can remain economically impossible.

“Affordable” can describe a ceiling

A restricted rent may be below a program maximum but not calibrated to the actual income of the renter occupying the unit. Deep affordability usually requires another layer of subsidy.

“Available” is a second test

A home does not solve the shortage for the poorest renter if it is occupied by a higher-income household, located too far away, inaccessible, unavailable, or still priced beyond that renter’s budget.

More apartment supply can reduce market rents, an important win for renters overall. But new supply does not automatically create homes affordable to households with the very lowest incomes.

02

Why the tax-credit system can stop short of the poorest renter

The financing gap is real, not just semantic

The Low-Income Housing Tax Credit, created in 1986, does not operate like a monthly rent coupon. Federal tax credits are allocated through state agencies to help finance construction or rehabilitation. In return, qualifying units must follow income and rent restrictions for long periods. HUD’s database shows how central the program has become to U.S. affordable housing production, while IRS rules define the income tests that projects must satisfy.

The system is effective at attracting private capital into below-market housing, but it does not erase the operating costs of a building. Property taxes, insurance, maintenance, payroll, utilities, debt service, security and repairs still have to be paid. Those expenses do not fall by half simply because a tenant earns half as much. This is why the deepest income bands often require extra resources such as project-based rental assistance, local operating support, additional capital subsidy or cross-subsidy from higher-rent units.

A Washington-area nonprofit developer described the arithmetic to the AP using a unit targeted at 60% of area median income. The apartment brought in about $1,715 a month while mortgage and operating costs were about $1,575, leaving only a small cushion. Cutting the rent roughly in half to reach a tenant near 30% of area median income would not cut those expenses in half. Without another subsidy, the project would run a substantial operating deficit.

Tax credits lower the cost of developing income-restricted housing, but rent revenue still has to cover ongoing operations. Reaching very low income bands often takes an additional subsidy layered on top.

The central mismatch is simple: the renter’s income can fall much faster than the building’s expenses. A production subsidy can make construction possible without making every unit affordable to someone living on the lowest incomes.

Critics of the tax-credit system argue that its financial and compliance structure is too complex and that direct rental assistance can target need more efficiently. Supporters answer that vouchers cannot rent apartments that do not exist, that private-market landlords do not always participate, and that LIHTC properties provide durable income restrictions in places where land and construction costs would otherwise make below-market housing difficult to build. Both points can be true. Supply subsidies and tenant subsidies solve different parts of the same equation.

That is why comparing the programs as if policymakers must pick only one can be misleading. Tax credits primarily help create and preserve physical units. Housing Choice Vouchers primarily help qualifying households bridge the gap between income and rent. For the poorest renter, the strongest arrangement is often a deeply targeted unit with rental assistance attached — but those combinations are scarce because each additional layer costs money.

Housing affordability is ultimately a cash-flow question. A household paying more than half its income for rent has little buffer for utilities, food, transportation, health care or an unexpected bill.

03

Austin shows what happens when the income bands drift apart

Strong overall construction, weak deepest affordability

Austin is an unusually clear case because two housing stories are happening at once. Years of rapid growth and a wave of apartment construction expanded the city’s rental inventory. That new supply has put pressure on asking rents and forced many market-rate properties to offer concessions. For renters able to qualify for those apartments, the competition is welcome.

But the city’s own Strategic Housing Blueprint scorecard shows how unevenly new housing has served different income levels. The 2024 scorecard tracks a ten-year goal of 20,000 homes affordable to households at 30% of median family income or below. Through 2024, only 543 had been built. By contrast, the city reported 15,819 homes in the 61%–80% MFI band, surpassing that band’s ten-year target of 15,000, and 16,067 homes in the 31%–60% band toward a 25,000-unit target.

Those numbers do not mean Austin built “too much” affordable housing. They mean the supply is concentrated higher on the income ladder than the unmet need. The city notes that households at 30% MFI or below represent 17% of Austin households, yet only a tiny share of new production has been affordable to them. The AP reported that the Austin Housing Department has responded by giving preference to funding proposals that include units at 30% AMI.

Austin’s building boom has improved competition in parts of the rental market. The city’s own scorecard nevertheless shows that production for households at the bottom of the income distribution remains far behind its goal.

The vacancy pattern adds another layer. According to CoStar data cited by the AP, more than 4,500 Austin units classified as affordable were vacant, nearly 16% of that stock. One affordable-housing developer reported a 12% vacancy rate in its units aimed at households near 60% AMI. Those are not the same apartments that a household at 30% AMI can necessarily afford. In other words, the city can have vacancies in one regulated income band and a severe shortage in another.

This is an important correction to a common reading of vacancy statistics. An empty restricted unit does not prove that the broader affordable-housing shortage has disappeared. It may indicate that the unit’s rent is too close to market alternatives, that screening takes too long, that the property is poorly located for the target renter, or that the income restriction is misaligned with where the deepest demand sits.

Supply can work — and still miss a group

Austin’s market-rate construction has helped create renter competition. The lesson is not that new supply failed; it is that broad supply and deeply subsidized housing are complementary rather than identical tools.

Vacancy is a diagnostic signal

When income-restricted units remain empty while lower-income households are unhoused, agencies should examine rent bands, leasing friction, location, unit size and whether rental assistance is attached.

04

Denver and Portland reveal the same pressure from different angles

When restricted rents approach the open market

The mismatch is not unique to Texas. In Denver, the Colorado Housing and Finance Authority reported vacancy rates cited by the AP of about 13% among tax-credit units targeted at 60% AMI and 21% among units targeted at 80% AMI. Yet the shortage for the city’s poorest households remains severe. A vacancy in a unit aimed at a moderate-income renter does little for someone whose income falls far below the eligibility band unless another subsidy can close the rent gap.

In Denver, elevated vacancy rates have appeared in some higher AMI tax-credit bands even as very low-income renters still face a shortage of housing they can afford.

Portland presents a similar comparison. The Portland Housing Bureau reported more than 1,700 vacant affordable units and an overall vacancy rate of roughly 7.5%, according to the AP. Many of the vacancies were in housing targeted around 60% AMI. For a one-person household, that band was around $54,000, with rent for a one-bedroom capped around $1,444. CoStar’s average market rent for a one-bedroom was about $1,581.

A difference of roughly a hundred dollars a month matters. But so does the leasing experience. An unrestricted property may offer a promotion, approve an applicant quickly and require a familiar set of documents. An income-restricted property may need to verify assets, pay stubs, bank accounts, recurring transfers and household composition to satisfy compliance rules. When the savings are large, applicants tolerate that process. When the savings shrink, some households decide the convenience and flexibility of a market-rate unit are worth the extra rent.

When the restricted rent sits close to a nearby market rent, paperwork and waiting time become part of the effective price a tenant considers.

City	What the vacancy data show	What they do not show
Austin	More than 4,500 affordable units vacant; nearly 16% overall vacancy in the AP’s cited market data.	That the 30% AMI shortage is solved. Austin’s official production for that band remains far below target.
Denver	High vacancy reported in some 60% and 80% AMI tax-credit units.	That households far below those income bands can pay the same rents without deeper assistance.
Portland	More than 1,700 vacant affordable units; many around 60% AMI as restricted rents move closer to market rents.	That application friction, location and unit mix are irrelevant. Each can affect whether a qualified renter chooses the unit.

05

For the poorest renters, the missing ingredient is often rental assistance

Vouchers bridge income to rent — when families can get them

Housing Choice Vouchers are designed around the household rather than the financing of a building. Eligible renters generally contribute a share of their income toward rent, while the subsidy covers an approved amount up to program standards. That structure makes vouchers particularly valuable for extremely low-income households because the assistance adjusts more directly to what the household can pay.

The weakness is scale. The 2026 Gap report says three out of four eligible renter households receive no assistance because federal housing aid is not funded as an entitlement. Waiting lists can remain closed for years. Even a household that receives a voucher can face difficulty finding a landlord, passing screening, locating an accessible unit or completing a lease before the search period expires.

Income verification protects program integrity, but complicated documentation can also slow leasing and discourage eligible renters when comparable market-rate apartments are faster to obtain.

Tax-credit properties are important in this equation because federal law generally prevents owners from refusing to lease a qualifying unit solely because the prospective tenant holds Section 8 assistance. That makes LIHTC and vouchers potentially complementary: the tax credit helps create the apartment, and the voucher can make its rent affordable to a household whose income is far below the unit’s nominal AMI band.

But attaching assistance to enough units requires money, and so does expanding tenant-based vouchers. That is the unavoidable budget reality behind “deep affordability.” A home affordable at 60% AMI may need one public subsidy. The same physical home affordable to a household at 30% AMI may need an additional subsidy for operations or rent. The need is greatest where the tenant has the least ability to cover that gap.

A vacant 60% AMI apartment and a homeless 20% AMI renter are not evidence of excess supply. They are evidence that the subsidy and the need are sitting on different rungs of the ladder.

Extremely low-income renters are not one type of household. They include low-wage workers, older adults, people with disabilities, students and caregivers whose incomes leave almost no margin for rising housing costs.

06

Paperwork has become part of the affordability equation

Compliance protects public money but can raise vacancy

Affordable-housing compliance exists for a reason. Income-restricted units are a public benefit, and agencies need to confirm that households meet program rules. Without verification, scarce subsidies could be diverted from eligible tenants. The challenge is that the process can be designed in ways that impose a much higher burden than ordinary apartment leasing.

Property managers may need to confirm income from multiple jobs, bank balances, irregular gig payments, child support, benefits or transfers between accounts. A worker whose earnings fluctuate can be harder to certify than a salaried employee. A household may submit documents, wait for review, discover that another statement is needed, then watch a market-rate unit down the street approve an application the same day.

That administrative gap is most damaging when restricted rents are close to market rents. The prospective tenant is not comparing rent alone; the tenant is comparing time, uncertainty, documentation, move-in dates, security deposits, application fees, location and the risk of losing another housing option while waiting. A property can therefore be compliant, well maintained and nominally affordable yet still struggle to lease.

Streamlining verification does not require abandoning eligibility rules. Agencies can reduce duplicate documentation, standardize forms and improve data sharing while preserving compliance.

That points to a policy lever that costs less than constructing a new building: make the leasing process easier. State housing agencies and local governments can review which documents are required, how often they must be resubmitted, whether electronic verification can substitute for paper, how quickly files are processed and whether applicants receive clear status updates. Owners can coordinate waitlists and pre-certification so an eligible household does not restart the same paperwork at every property.

Administrative reform will not close a 7.2 million-home national shortage. But it can reduce the absurd outcome of a unit sitting empty because qualified renters cannot navigate the pathway into it. In cities with elevated vacancies, that is a practical first question to investigate before concluding that affordability targets themselves are unnecessary.

07

What a better-targeted housing system would measure

Not just units produced, but who can actually live in them

Most housing scorecards are built around production: how many units received financing, how many opened, how many are income restricted. Those are essential numbers, but the vacancy paradox shows why they are incomplete. A city can hit a production target in one income band while missing another by a wide margin. A state can allocate credits efficiently while a property takes months to fill. A federally assisted unit can exist without a voucher holder being able to find it.

A more useful dashboard would track affordability by income band, unit size and neighborhood; vacancy and time-to-lease by the same categories; the share of units paired with rental assistance; application processing times; reasons applicants are rejected or withdraw; and how often restricted rents sit within a narrow range of comparable market rents. That information would tell policymakers whether the problem is insufficient supply, weak targeting, excessive paperwork, a location mismatch or some combination.

Housing need is local. A unit’s income restriction matters alongside transit, schools, accessibility, bedrooms, neighborhood costs and whether the household can realistically reach the leasing office and complete the process.

Target more units at 30% AMI and below

State allocation plans and local gap financing can reward projects that serve the income band with the largest shortage rather than concentrating new units where financing is easiest.

Pair units with rental assistance

Project-based or tenant-based subsidies can bridge the difference between a tax-credit rent and what an extremely low-income household can pay.

Keep building ordinary housing too

Market-rate supply can reduce rent pressure and concessions can help moderate-income tenants. Deep subsidy should supplement broad supply, not substitute for it.

Measure vacancy by income band

A single “affordable vacancy rate” can hide a glut at 80% AMI and a shortage at 30% AMI. Programs should publish more granular data.

Simplify verification

Faster, standardized and reusable income certification can lower leasing friction without weakening eligibility rules or financial oversight.

Preserve what is already deeply affordable

Older subsidized properties, public housing and units with attached assistance are difficult to replace once lost. Preservation can be cheaper than recreating the same affordability.

There is no single program that can solve all of this. Zoning reform, faster permitting and more construction can relieve broad scarcity. Tax credits can finance long-term restricted housing. Vouchers can make rents workable for the poorest households. Public housing and supportive housing can reach people with needs the private market will not reliably serve. The policy failure comes when one tool is asked to do the job of all the others.

What to watch next

• Whether cities with rising restricted-unit vacancies change their AMI mix or application rules.
• Whether state housing finance agencies direct more 2027 tax-credit awards toward 30% AMI units and projects with rental assistance.
• Whether market rents continue softening in high-construction metros, narrowing the discount on 60% and 80% AMI apartments.
• Whether Congress expands rental assistance enough to reach more than the minority of eligible households currently served.

Older adults and people with disabilities make up a large share of extremely low-income renters. For households on fixed incomes, even a modest rent gap can be impossible to absorb.

08

The takeaway: vacancy and scarcity can exist at the same address

The contradiction disappears when income is put back into the picture

America’s affordable-housing debate often swings between two oversimplified claims. One says the only problem is not enough construction. The other says market supply does not matter because the poorest households cannot afford new apartments. The vacancy evidence from Austin, Denver and Portland shows why reality is less convenient. More construction can lower market rents and create valuable competition, while the bottom of the income distribution can remain stranded because even discounted rents exceed what those households can pay.

The strongest measure of success is therefore not the number of doors labeled affordable. It is whether the housing system connects the right renter to the right rent at the right time. A unit at 60% AMI should be judged by whether households in that band can and want to lease it. A household at 20% or 30% AMI should be judged against the supply actually affordable to that income, including any voucher or project-based assistance needed to make the math work.

That framing also explains why empty units deserve attention without becoming an argument to abandon subsidized housing. Persistent vacancy wastes scarce public resources and should trigger changes in rents, targeting, marketing or paperwork where rules allow. But moving those units down the income ladder usually requires real money. The cost of deeper affordability cannot be solved by relabeling the apartment.

For the renter sleeping in a shelter while a regulated apartment sits empty, those distinctions may sound bureaucratic. Yet they determine who gets a key. The next phase of U.S. housing policy will be judged not only by how many apartments it produces, but by whether it can close that last, stubborn gap between a unit that is affordable on paper and a home that is affordable in real life.

The policy goal is not a full spreadsheet or a low vacancy rate by itself. It is a stable home that a household can actually afford and keep.

Housing affordability, explained

What does AMI mean?

AMI stands for area median income, a commonly used shorthand in housing policy. HUD calculates local income limits using median family income and program-specific adjustments. Percentages such as 30%, 50%, 60% and 80% are used to define eligibility or rent limits for different programs.

How can an “affordable” apartment still be too expensive?

Because many programs cap rent using a designated income band rather than the tenant’s actual income. A rent designed around 60% AMI may be manageable for a household near that level but unaffordable for someone at 20% or 30% AMI without rental assistance.

Why not simply lower the rent on every vacant restricted unit?

Owners can sometimes lower asking rents, but a large reduction may leave too little revenue to cover debt and operating costs. Reaching the deepest income levels often requires an additional subsidy rather than a pricing decision alone.

Do housing vouchers solve the problem?

Vouchers can make existing apartments affordable to very low-income tenants, but the program serves only a fraction of eligible households. Voucher holders can also face long searches and landlord participation barriers.

Does a high vacancy rate mean a city built too many apartments?

Not necessarily. Vacancy can be concentrated in a specific income band, neighborhood or unit type. In a fast-building market, higher vacancies can also reflect healthy competition and falling rents. The important question is whether the vacancies line up with the households experiencing the greatest shortage.

Reporting sources

• Associated Press, Sept. 7, 2026 — city vacancy figures, developer examples and interviews from Austin, Denver and Portland.
• National Low Income Housing Coalition, The Gap 2026 — national shortage, cost-burden and rental-assistance estimates.
• HUD USER, FY 2026 Income Limits — federal income-limit definitions and relationship between local median income and assisted-housing eligibility.
• HUD USER, Low-Income Housing Tax Credit program background — traditional 20-50 and 40-60 set-aside structure and rent restrictions.
• Internal Revenue Service, Section 42 guidance — current explanation of LIHTC minimum set-aside and average-income rules.
• City of Austin, 2024 Strategic Housing Blueprint Scorecard — production by median-family-income band through 2024.

The housing paradox is not that America has too many affordable homes. It is that too many of the homes it calls affordable are financed for one rung of the income ladder while the deepest unmet need sits several rungs below.