Housing voucher waiting lists — a 25-month average wait, and HUD has told agencies to stop issuing new vouchers

What is happening?

In the United States the queue for a housing voucher is often not a queue you can join. One national tracker counted 213 Housing Choice Voucher waiting lists open until further notice as of 2026-06-01, against 893 lists that were open for at least one day at some point in the preceding three years. People who do reach a list wait years — HUD data compiled by USAFacts puts the average wait for subsidized housing at 25 months in 2025, down from 27 months in 2024, with state averages running from 10 months in West Virginia to 57 months in Maryland. Behind the queue sits a supply gap the queue cannot clear: NLIHC reported in March 2026 that there are 11 million extremely low-income renter households against 35 affordable and available rental homes for every 100 of them, a shortage of 7.2 million homes.

The 2026 funding year narrowed the entrance rather than widening it. HUD wrote to housing agency directors on 2025-12-22, urging them to stop issuing new vouchers apart from named exceptions, to pause new project-based voucher agreements, and to lower payment standards. A second letter on 2026-02-18 repeated the instruction, naming HUD-VASH and Foster Youth to Independence as the exceptions, and said HUD expected a 99 percent proration of housing assistance payments while warning that some agencies would still not have enough funding to support every current participant.

Whose problem is this?

The body that decides how many vouchers exist is Congress, the body that rations them under a fixed amount is HUD, and the body that has to tell an applicant the list is closed is the local housing agency. None of the three carries the wait.

Where does this problem end?

The boundary matters because a closed waiting list produces no record of the household that was turned away, so the size of the queue is not the size of the need.

What is the state now, and what should it be?

Now

Should be

Needs a new measurementno source we opened states an official federal target for how many eligible households should be served, how long a wait should run, or what share of waiting lists should be open. The strongest statement of intent we found is that the FY2026 amount should be sufficient to renew existing contracts, which is a target for holding the current caseload rather than for shortening the queue.

How big is it?

The affected population is between 7.2 million and 8.25 million households as of 2026, and the two ends are different constructions of need rather than two estimates of one number.

The lower end is a direct measurement. NLIHC counts 11 million extremely low-income renter households and a shortage of 7.2 million rental homes affordable and available to them.

The upper end is an inference. NLIHC also states that only one in four households that qualify for housing assistance receive it. Applying that share to the same 11 million gives 8.25 million households that qualify and do not receive assistance. NLIHC does not itself apply the one in four share to the extremely low-income renter count, so the two populations may not be identical.

No source we opened gives the national count of households actually enrolled on Housing Choice Voucher waiting lists. Because many lists are closed to new applicants, an enumerated queue would understate this range rather than bracket it.

Under what conditions does it arise?

Three conditions keep this state in place, and none of them is a failure of any single office.

1. A fixed appropriation meets moving rents. The FY2026 bill provides 34.9 billion dollars to renew tenant-based rental assistance contracts, an amount described as sufficient to renew what already exists. When renewal costs rise faster than the appropriation, HUD prorates housing assistance payments — 99 percent expected for 2026 — and the difference has to come out of the number of vouchers actually in use. 2. Rationing happens at the door and leaves no record. An agency that cannot fund new admissions closes its waiting list. A closed list generates no count of the households it turned away, so the visible statistic is the number of open lists rather than the number of people who need one. 3. The unit shortage is larger than the program. At 35 affordable and available homes per 100 extremely low-income renter households, a voucher issued today is a claim on a home that may not exist, which keeps the queue long even in years when funding holds.

What has been tried?

Both directions were tried in the same year — money was appropriated to hold the caseload, and guidance was issued to shrink it where the money fell short.

What was found?

Why is it still unsolved?

Supply shortage — there are not enough funded voucher slots and not enough affordable homes, and the appropriation is the channel through which both shortages reach a household standing at the door.

The shortage is double, and that is why one fix does not close it. Even if every household on every list were handed a voucher tomorrow, NLIHC counts only 35 affordable and available rental homes for every 100 extremely low-income renter households, so the voucher would be a claim on a home that does not exist. Building units without funding vouchers leaves the same households facing rents they cannot pay. The two shortages have to move together, and they are decided in different rooms.

The proximate channel in the 2026 funding year is the appropriation. Renewal costs track market rents, the appropriation is set once a year, and when the first outruns the second HUD prorates housing assistance payments and instructs agencies to stop admitting new households. The reduction lands entirely on people who have not been served yet, because the program protects the caseload it already carries. That is the mechanism by which a budget decision becomes a closed waiting list.

This reads as a shortage rather than as a queue that is merely slow. A queue clears when the service rate exceeds the arrival rate. Here the service rate is set by an amount sized to renew existing contracts, so admissions are roughly bounded by exits and the queue does not clear at all.

What observation would mean it is solved?

Candidates — (a) the number of Housing Choice Voucher waiting lists open to new applicants rises well above 213 and stays there (b) the average wait falls durably below 25 months (c) the shortage of affordable and available homes for extremely low-income renters falls from 7.2 million.

(a) can be satisfied by opening lists that no funding stands behind. An open list with no voucher behind it moves the rationing from the door onto the list, and the household experiences the same outcome later.

(b) is the most dangerous of the three. The average wait already fell 7.4 percent between 2024 and 2025. A wait measured only over households that were eventually housed can get shorter precisely because fewer households are admitted, and no source we opened says which of the two produced that decline.

(c) can improve while the program shrinks, since the unit count and the voucher appropriation are set separately. None of the three is sufficient alone, and (a) and (b) are the pair that can move in opposite directions without either being wrong.

What is it connected to?

Fills with researchhomelessness counts, landlord refusal to accept vouchers, source-of-income discrimination law, public housing and project-based rental assistance as parallel queues, and the roughly 60,000 households NLIHC reported as at risk of losing Emergency Housing Vouchers. Relation type and evidence grade were not confirmed in this round.

What these sources do not say

• The national count of households enrolled on Housing Choice Voucher waiting lists. The trackers count lists, not people. Every national figure here describes need or supply, never the queue itself.
• The share of housing agencies whose lists are closed. No source we opened gives the number of agencies administering the program, so the open-list counts of 213 and 398 have no denominator and cannot be turned into a percentage.
• Why the trackers disagree. One counted 213 lists open on 2026-06-01 and another counted 398 on 2026-08-07. Neither page documents its scope, its definition of an open list, or how often it refreshes.
• Why the two 2026 funding accounts differ. A HUD letter dated 2026-02-18 expected a 99 percent proration and warned of large shortfalls, while a 2026-05-06 note on the allocation notice anticipated full funding. Nothing we read reconciles them.
• How many agencies actually entered a shortfall in 2026, and how many households lost assistance as a result. HUD stated that it could not guarantee funding would be available to resolve those shortfalls, but published no count.
• Any 2025 or 2026 voucher waiting-list opening with a published application volume. The Los Angeles lottery figure is from 2022, and the 2026 Houston figure covers a public housing list rather than vouchers.

See the evidence

No primary HUD document was read directly. Every federal figure above reaches us through NLIHC or NAHRO summarizing a HUD notice or letter, and the HUD budget management letter, the HUD voucher dashboard and the CBPP fact sheets were all unreachable. Two independent readings agree on the core supply figures — the NLIHC release and the NLIHC landing page both give a 7.2 million shortage and 35 affordable and available homes per 100 extremely low-income renter households. The as-of dates given for the NLIHC figures are the publication date of The Gap 2026 rather than the survey year behind it, which was not confirmed, and the landing page states the shortage and the ratio but no national count of extremely low-income renter households, so the 11 million figure rests on the release alone. Three divergences are left open rather than resolved. First, the renewal amount appears as 34.9 billion dollars in the NLIHC summary of the FY2026 bill and as 34,557 million dollars in the NAHRO summary of the allocation notice, and the scope difference was not confirmed. Second, the 2026-02-18 letter expects a 99 percent proration while the 2026-05-06 allocation note anticipates full funding. Third, the two open-list trackers report 213 and 398 for dates about two months apart. This is a Path A output built from research rather than from an observation record, so observation_refs is empty and provenance_mode is press-derived.

People affected

Estimated range 7,200,000–8,250,000 As of 2026

Sensitivity The range is not a confidence interval. The two ends are different constructions of need rather than two estimates of one quantity: the lower end is a deficit of housing units affordable and available to extremely low-income renters, and the upper end is a count of households that qualify for assistance and do not receive it. The more important limit is that no source we opened gives the national count of households actually enrolled on Housing Choice Voucher waiting lists. Trackers count lists, not people, and one of them counted only 213 lists open until further notice as of 2026-06-01. Because many lists are closed to new applicants, an enumerated queue would understate this range rather than bracket it. In the other direction the range is too wide, since not every extremely low-income renter household would apply for or accept a voucher if every list were open.

Regional breakdown State values were confirmed only for the average wait for subsidized housing, which ran from 10 months in West Virginia to 57 months in Maryland in 2025. That is a duration, not a population, and durations do not sum to a national total. NLIHC publishes state tables for the affordable housing shortage, but no state-level breakdown of the population counted here was produced in this round, and splitting a national count by state population would be proportional allocation, which the distribution of extremely low-income renters and of open waiting lists does not follow.