Skip to content

Housing, on the record

Housing · Oct 29, 2026 · 7 minute read

When public housing becomes Section 8, who gets to come home?

Across the country, public housing is being rebuilt and moved onto Section 8 contracts. The repairs are real. So is the question every family asks when the moving trucks come: will we be let back in?

865,449public housing homes left, December 2025, down from about 1.4 million in the mid 1990s
555,000homes Congress now lets housing authorities move to Section 8, with no end date (2026 law)
41%of public housing households were headed by a Black resident, 2025
27 monthsaverage wait for families who got a voucher, 2025

National Low Income Housing Coalition from HUD data (December 2025); Congressional Research Service on P.L. 119-101 (2026); HUD, Picture of Subsidized Households (2025). Checked October 29, 2026.

Our mission is to make sure Black communities are not simply informed about the decisions shaping our future, but equipped to help make those decisions. An informed community is a powerful community.

On this page: What the law is, and what Congress meant it to do and Why this is different from what happened in Atlanta

Is this paid with your tax dollars? Yes, in part, and the rest is borrowed or invested money. Here is how it moves, so you can weigh it for yourself. A conversion takes the federal tax dollars that already ran a public housing property and turns them into a long term Section 8 contract, usually 15 or 20 years, that has to be renewed. Because the rent is now guaranteed by that contract, the housing authority can borrow from banks and sell federal low income housing tax credits, which are a tax break the public gives up, to pay for repairs. HUD says conversions bring in more than $18 for every $1 of public housing money. Government auditors counted it differently and came to about $7.44, and HUD's own 2019 evaluation shows the answer depends on what you count as public money. Some residents see a smart way to fix homes Congress never paid to fix. Others see public land and public dollars passing into private hands. SourceLine takes no side. These are separate stages, never one total, and the decision stays with you and the people you elect.

Start with what is working. In HUD's own evaluation, the repair bill at converted properties fell by about two thirds, from $12,981 a home to $4,608, while it more than doubled at public housing that did not convert. HUD reports more than $27 billion in construction at 1,950 properties. Most families never had to leave: 82 percent of residents surveyed stayed in the same property through the conversion. That matters, because Black families are at the center of this. In 2025, 41 percent of public housing households and 48 percent of voucher households were headed by a Black resident (record, HUD, Picture of Subsidized Households). Before most conversions began, the properties that later converted were even more Black than public housing as a whole: 58 percent against 45 percent.

What the law is, and what Congress meant it to do

Public housing peaked at just over 1.4 million homes in the mid 1990s. By December 2025, 865,449 were left. Much of what remained was aging, and in 2010 HUD counted a repair backlog of $25.6 billion that yearly budgets were not covering.

Congress answered in 2011 with the Rental Assistance Demonstration, called RAD, inside a spending law signed November 18, 2011. Its stated purpose was "to conduct a demonstration designed to preserve and improve public housing". It began as a test capped at 60,000 homes. Congress kept raising the cap, to 455,000 by 2018. This year it went further. The 21st Century ROAD to Housing Act became law on July 11, 2026, without the President's signature. It raises the cap to 555,000 homes, removes the end date so RAD becomes permanent, and requires HUD to report each year on what conversions do to residents.

The law itself says a conversion "shall not be the basis for re-screening or termination of assistance or eviction of any tenant family". HUD's rules add the promise families care about most. Any resident moved out for construction "has a right to return," and "permanent involuntary displacement of residents may not occur". Residents cannot be screened again for income, credit or criminal history because of the conversion. Big rent increases must be phased in over three or five years. Since 2025, residents behind on rent get 30 days' notice instead of 14. After one or two years, a resident can ask for a voucher and move elsewhere, though agencies may limit how many such vouchers they hand out each year.

Why this is different from what happened in Atlanta

Many Black families remember an earlier program, HOPE VI, which began in 1992. It tore down about 93,000 public housing homes by 2009 and planned to replace only about half of them as public housing. Housing authorities expected 61 percent of residents to come back; by 2006 that estimate had fallen to 38 percent. In an Urban Institute study of five cities, only 5 percent of families had moved into a new HOPE VI home by 2005.

Atlanta became the first city in the country to end all traditional public housing, by 2011. Georgia State University researchers counted about 50,000 households relocated since 1994. In the last round, families were offered vouchers with no option to return.

HOPE VI made no promise to bring families home. RAD does, on paper. That is why the question for RAD is not whether the promise exists, but whether it is kept.

Who heads the households, by program, 2025

Vouchers

48%

Public housing

41%

Section 8 buildings

34%

Share of households headed by a Black resident. HUD, Picture of Subsidized Households, 2025.

On this page: What the record shows, and where it stops; Pros and cons, side by side; What supporters point to; What critics point to; A view, labelled; Rural America is losing homes too; Who decides, from your street to Washington; and What you can do

What the record shows, and where it stops

HUD says more than 90 percent of residents choose to stay or return, but it does not say how it counted. A study published by HUD in 2021 found that about a third of residents were relocated during conversion, and that most working age residents who moved did not go back to their original apartment, which is not the same as leaving the property. A peer reviewed study in 2026, covering 4,410 developments in 43 states and the District of Columbia, found no rise in eviction filings after conversion, while warning that it could not see families pushed out in quieter ways.

Oversight has been thin. In 2018 the Government Accountability Office found HUD was not tracking what conversions did to families, and that 27 percent of finished conversions involved no construction at all. In 2025 HUD's Inspector General inspected 190 converted homes at three housing agencies, one of them in Georgia. More than 74 percent failed HUD's own quality standards, and more than 60 percent of those had a defect that threatened life or safety. Advocates say HUD will not release its summary of resident complaints.

SourceLine has not found a published national count of families who were displaced or never came back. Until one is published, every claim about RAD, for or against, is resting on partial numbers.

Pros and cons, side by side

What supporters point to

Supporters say RAD saved homes that were falling apart, using borrowed and invested money instead of waiting on Congress for repairs it was not paying for. They say the new developments mix families, seniors and working people, sometimes with shopping nearby, instead of concentrating families in survival mode in one place. There is research behind part of that idea. In the Moving to Opportunity study, children who moved to lower poverty neighborhoods before age 13 earned 31 percent more as adults. Adults in that study did not earn more, but they had less diabetes, less extreme obesity and less distress.

What critics point to

Critics say the same moves can break up the neighbors, churches and family ties that help people survive, and that most displaced residents, when asked, would have preferred to stay. Moving to Opportunity was a voluntary move with a voucher, which is not the same as being moved for construction, so its findings do not settle what RAD does to families who had no choice. Teenagers who moved in that study saw slightly negative effects. Critics also point to what happens on the waiting list. Families who got a voucher in recent years waited about 27 months on average, and in some large cities years longer (record, HUD, Picture of Subsidized Households, 2025). Only about one in four eligible households gets help at all.

A view, labelled

Kazemde Ajamu, who founded SourceLine and serves as Vice Chairman of the Lithonia Housing Authority, sees RAD as a good thing as long as residents are protected and get the first right to come back. That is his view, not SourceLine's reading of the record, and SourceLine reports every housing authority, including Lithonia's, from the same sources and by the same methods.

Rural America is losing homes too

In the countryside the program is different but the stakes are the same. The Agriculture Department's rural rental housing had 406,290 apartments in October 2023, and 20 percent of the households were Black. In the rural South the share is far higher: 58 percent in Georgia, 67 percent in Louisiana, 69 percent in South Carolina and 75 percent in Mississippi (SourceLine's arithmetic from USDA's counts). Those homes are disappearing as owners pay off their mortgages and leave the program. Nearly 15,000 apartments in 621 properties left between June 2021 and March 2026, and the Housing Assistance Council projects the whole stock could be gone by 2056. The 2026 housing law made permanent a fix that lets rental help continue after a mortgage ends.

Who decides, from your street to Washington

Your housing authority board votes on whether and how to convert, and in most places the mayor or the county commission appoints it. Your state's housing finance agency awards the tax credits that pay for most rebuilding. Congress made RAD permanent in 2026: the Senate voted 85 to 5 on June 22, and the House voted 358 to 32 on June 23. HUD writes the rules on the right to return. Where you live, below, names the people who decide for your address, shows how your members of Congress voted, and gives one question to ask each of them.

What you can do

Know your numbers. Enter your address under Where you live to see Black homeownership, public housing, Section 8 and vouchers where you live, how long families waited, and who decides.

If a conversion is coming where you live, go to the resident meetings, ask for the relocation plan in writing, and keep your copy. The right to return is written in HUD's rules.

Next dates that matter. HUD's first yearly report on RAD's impacts under the new law is the one to watch. Your housing authority board meets in public, usually every month.

---

Found this useful? Pass it on.

Facts in this story were checked October 29, 2026.

Key figuresBy the numbers

Black households in rural rental housing, 2023

Mississippi

75%

South Carolina

69%

Louisiana

67%

Georgia

58%

United States

20%

Share of households in USDA rural rental housing (Sections 515 and 514) that are Black. SourceLine's arithmetic from USDA Rural Development, Multi-Family Housing Occupancy Statistics, October 2023.

Where you live

Enter your ZIP code or address to see Black homeownership, public housing, Section 8 and vouchers where you live, how long families waited, and who decides.

Weigh in

What does this story mean for you, your family or your community? Tell us in your own words.

Leave blank to stay anonymous.

Your thoughts go to SourceLine, not to your representatives. After review, they may be shown on this page word for word, with the name you give or as Anonymous. We never ask for your email or address.

Get SourceLine by email

When a county commission, city council or school board decides something that touches the Black community, we will tell you what happened, who voted and how to be heard. Pick what matters to you.

What do you want to hear about?

We send you one email first to confirm your address. SourceLine never sells or shares your email address.

How we counted, and where every number comes from

Figures carry the year and source shown. HUD's Picture of Subsidized Households (2025) is loaded in SourceLine's database and supplies the Black share of households by program and the average months new households waited. Homeownership is from the Census Bureau, American Community Survey, 2020 to 2024. Conversion counts differ across sources by date and by what they count; the story uses HUD's own August 2026 overview for construction and properties and does not state one national total of converted homes. Leverage ratios are shown as each source measured them and are not combined. HOPE VI figures are estimates from the Congressional Research Service, GAO and the Urban Institute. Rural shares are SourceLine's arithmetic from USDA's October 2023 occupancy report. Congressional votes are from SourceLine's database of House and Senate roll calls on H.R. 6644. Proposed, approved and spent money are never added together.

Sources

See docs/research/2026-10-29-rad-facts.md for every source with its link. Primary records: Public Law 112-55 (2011); HUD RAD Notice Revision 4 (2019) and Supplemental Notices 4B (2023) and 4C (2025); HUD RAD Overview (August 2026); HUD, Final Report on RAD Evaluation (2019); GAO-18-123 (2018); HUD Office of Inspector General 2025-CH-0002 (2025); CRS R49354 on P.L. 119-101 (2026); CRS R41654 and RL32236; GAO-04-109; USDA Rural Development Multi-Family Housing Occupancy Statistics (October 2023). Research: Hayes, Gerken and Popkin, Cityscape (2021); Lochhead, Hepburn and Ellen, Housing Policy Debate (2026); Oakley, Ruel and Reid, Cityscape (2013); Chetty, Hendren and Katz, American Economic Review (2016); Housing Assistance Council (2026, as summarized by NLIHC).

Checked October 29, 2026. Found an error? Write to info@mysourceline.com.

About these figures

Every figure on this page shows the year it describes. Public agencies release their numbers on their own schedules, often a year or more after the work of counting is done. The Census Bureau, for example, publishes its five year community survey each December, and the national reading test is given every two years. That is why some figures are not from this year. SourceLine updates each one as soon as the agency publishes it.

For some counties and small towns, SourceLine has not found meeting records online. Where it has not, we add them as they become available. A page that says SourceLine hasn’t updated a record yet means exactly that.