Policy LabPolicy analysis with the thesis up front and every source locked · built on PLACES I NEVER HEARD OF

The housing engine: bank fees into first-loss equity

2026-11-02 · universal income, housing, homelessness, national
ThesisA universal income paid into a housing shortage becomes rent. Before the money can raise living standards, there have to be homes to live in. The design makes the banks that earn the pool's fees build them: half of every fee dollar goes into new housing as 25 percent first-loss equity, which draws in four dollars of ordinary lending for every one. The money works down a ladder: supportive homes with services for the roughly 770,000 people without housing first, then the three to five million home shortage, then retrofits and lagging regions. Tenants never need to own.

Give every adult $52,000 in a city with no vacant homes and the first thing that happens is that rents rise until the money is gone. Housing supply is the condition on which everything else in this series depends, and the design does not leave it to chance. It attaches a building obligation to the one party that profits from running the system: the banks.

The mechanism

Banks earn 0.5 percent of pool inflow (piece 2). Half of that fee must be reinvested each year as 25 percent first-loss equity in new housing. First-loss equity is the money that absorbs the first 25 percent of any loss on a project, which is exactly the capital that is scarce and expensive today and the reason so many projects never start. With a bank holding the first loss, the other 75 percent can be ordinary deposit-funded construction lending, which the Community Reinvestment Act already asks banks to do4. One dollar of fee becomes four dollars of housing. In the model, half the fee times four equals twice the fee in housing capital: $111 billion in 2027, $258 billion in 2045. At $350,000 a home, falling 3 percent a year as modular and automated construction spread, that is about 320,000 extra homes a year at launch, on top of the 1.4 million starts a year the country builds now3.

The ladder

The money is not spread evenly. It works down a priority ladder, and a bank cannot move to the next rung in a region until the one above is met.

RungWhat it means
1. Supportive homes with servicesAbout one million homes for the roughly 771,000 people without housing1 and those at risk, with wraparound services ($20,000 per person a year, funded by a bank tax credit, $6 billion in 2027 rising to $20 billion). The evidence is that housing with services ends chronic homelessness for most people who get it5. The model has the million homes in place by 2030.
2. The shortageThe three to five million homes the country is short2, closed by the mid-2030s at the model's pace. Ordinary rental and for-sale housing, mixed income, wherever the shortage is.
3. Retrofits and regionsEnergy retrofits, accessibility, and building in regions that lost population, so that the places where a universal income goes furthest get homes worth living in.

Tenants never need to own. The point of the engine is homes, not mortgages. A tenant in a bank-equity building pays rent set to cover operating costs and the ordinary loan, not a return on scarcity, and the bank's equity return comes from the building being full for decades.

Housing capital this year
bank equity plus the lending it unlocks
Extra homes this year
on top of ordinary building
Extra homes, cumulative
Supportive million reached
rung one
Four-million shortage closed
rung two, at the midpoint estimate

What the model cannot fix

Money is not the bottleneck for housing in most of the country; permission is. Zoning, permitting, parking minimums and environmental review stop more homes than financing does, and a bank with equity to deploy and nowhere allowed to build will deploy it in the few places that allow building, which are not where the need is. The design's answer is outside the model: the local essentials grant (piece 6) and the land value tax give local governments a reason to permit (the tax is the same on an empty lot as a full one), and the rollout ties the second rung to state zoning reform. We flag CRA box-checking as a real risk: a bank can meet a dollar target with projects that do not house anyone who needed housing. The ladder is the guard against it, and the ladder needs an auditor.

Sources

  1. 2024 Annual Homelessness Assessment Report to Congress — U.S. Department of Housing and Urban Development · original · 771,480 people experiencing homelessness on a night in January 2024. Not yet locked.
  2. The U.S. housing shortage — Zillow Research; Freddie Mac; National Association of Realtors (estimates range from about 3 to 5 million homes) · original · the shortage the second rung closes. Not yet locked.
  3. New Residential Construction — U.S. Census Bureau · original · about 1.4 million housing starts a year today, the baseline the engine adds to. Not yet locked.
  4. Community Reinvestment Act: regulations and guidance — Federal Financial Institutions Examination Council · original · the existing obligation the housing requirement is attached to. Not yet locked.
  5. Permanent supportive housing: evidence review — National Academies of Sciences, Engineering, and Medicine, 2018 · original · housing with wraparound services ends chronic homelessness for most people who receive it. Not yet locked.

Numbered references in the text point to this list. "Locked" means we hold a dated copy of the file on the PLACES platform so the figure can be checked later even if the publisher changes the page.