System Is Shutting the Door on Condo Ownership

The condo you can almost afford sits inside a building nobody will lend against. That’s the knot at the centre of American housing in 2026, and three finance veterans at the Urban Institute have spent months working out how it got pulled so tight. Their answer isn’t a villain. It’s something more ordinary, and far harder to fix: nobody, anywhere, is in charge of making the system work.

Edward Golding has seen this market from nearly every seat that matters, having run the Federal Housing Administration and spent more than two decades inside Freddie Mac building the models that decide which mortgages are safe to buy. His coauthor Laurie Goodman has published over 200 papers and sits in the Fixed Income Analysts Hall of Fame, which is, improbably, a real thing.

What they describe is close to a paradox. On paper the condo is the affordable answer. Across the 18 largest metro areas the median condo runs about two-thirds the price of a single-family house ($414,000 against $460,000), and in coastal Florida it drops below half. Last year 65 percent of condo mortgages backed by Fannie Mae and Freddie Mac went to first-time buyers, compared to 57 percent for other homes. If you’re young, urban and trying to get a foot on the ladder, the condo is often the only rung you can reach.

And yet almost nobody is building them, and a growing share can’no’t be sold. New condo starts have cratered from 100,000 to 150,000 a year before 2008 to a sliver of that now, and the condo share of multifamily completions has fallen from roughly 35-40 percent to about 5 percent.

A Building with a Mind of Its Own

To see why, you have to grasp what makes a condo legally strange. Buy a house and the bank lends against a thing you control. Buy a condo and you control your unit, but the roof above it, the walls around it, the elevator, the reserves, the insurance, all of it belongs to a homeowner’s association, a collective entity that isn’t even party to your mortgage. The lender underwrites a single unit while sitting exposed to the decisions of a group it never meets. The collateral, in effect, has a mind of its own.

For decades that wrinkle stayed manageable. Then, in the small hours of 24 June 2021, Champlain Towers South came down in Surfside, Florida, killing 98 people. Investigators later found the building had never met code and that its association had deferred critical repairs flagged in its own recertification, with settlements running past a billion dollars.

The shock did what shocks do to a financial system. It made everyone cautious all at once. Fannie Mae and Freddie Mac tightened their rules, demanding assurances that a building carried no critical deferred maintenance and no looming special assessment. Reasonable instincts. But here the authors locate the real failure: the agencies reached for blunt instruments, hard yes-or-no eligibility cutoffs, and applied them across a varied universe of buildings. The condo share of agency purchase mortgages, which peaked at 13.7 percent in early 2021, had slid to 8.7 percent by the end of last year.

The Trouble with a Hard Line

The problem with a categorical cutoff is that it punishes prudence. Well-run associations levy the occasional special assessment because that’s how a sensible building handles a roof or a boiler that no reasonable reserve could fully prefund. Treating any disclosed assessment as a red flag, Golding and Goodman argue, confuses a normal tail event with a distress signal.

The better tool, they keep circling back to, is pricing: charge for the incremental risk rather than slamming the door on it. A building at 49 percent owner-occupancy is not doomed next to one at 51, yet the old rules behaved as if it had stepped off a cliff. And underneath all of it sits a mundane problem, which is that nobody ever agreed on how to write any of this down. Fannie Mae, Freddie Mac and the FHA each run their own approval system, none recognises the others, the building-health questionnaires arrive as narrative PDFs that can’no’t be aggregated, and there isn’t even a condo flag in the main federal mortgage database, so the government can’t easily count the thing it’s failing to finance.

That fragmentation hands the market to large banks with proprietary databases and shuts out the community lenders best placed to serve a neighbourhood. Every lender pays the same fixed cost to investigate the same building, over and over, for no gain.

Then there’s insurance, which may prove the hardest knot of all. In coastal Florida, in California, in any place the climate is rewriting the actuarial tables, master policies for condo buildings have grown ruinously expensive or simply vanished, and the moment a project loses its master policy it loses agency eligibility and financing evaporates for nearly every would-be buyer.

What the authors propose is less a moonshot than an overdue cleanup. Get the agencies onto one shared, cross-recognized approval system with a public registry. Give owners a sanctioned way to finance a special assessment through a second mortgage or cash-out refinance, instead of letting the bill push a building toward the spiral they grimly call “stranded.” And let the agencies use their market weight to standardize the data, the way a twenty-year push once dragged property appraisal out of bespoke narrative and into something a computer could read.

None of it’s glamorous. There’s no breakthrough here, no switch to flip. Just the patient institutional work of deciding, at last, that someone should be responsible.

The condo market is broken, the authors write, with the flatness of people who have stopped being surprised by it. The cost of leaving it that way lands on exactly the households the market was built to serve. In a country arguing endlessly about how to make housing cheaper, one answer may already be sitting in plain sight, locked inside a building nobody will lend against.

Frequently Asked Questions