How Many Homes Does America Need? (1.5M–5M Gap)

Ask how many homes does America need and you get four answers from four institutions, none of them small. That spread frustrates people who want one clean number. It should not. The disagreement is about accounting, not about whether the hole exists. Everyone counting finds millions of missing units.

Here is what makes the number slippery. A shortage of goods is easy to measure: empty shelves. A shortage of housing hides inside other people's living rooms. The 32-year-old in a childhood bedroom is not on any vacancy report. Neither are the four roommates in a two-bedroom, or the family that took the apartment 40 miles out because nothing closer penciled. They all have addresses. They are all housed. And every one of them represents a unit the market failed to produce, which is exactly the quantity the vacancy statistics cannot see.

What are the three ways to count the gap?

Method one: households without units. Count existing households, count available units, subtract. This is the most conservative approach and produces the smallest figures. Freddie Mac's work has landed in the low millions. It undercounts by design, because a household that never formed does not appear.

Method two: the construction deficit. Take the pace of homebuilding from the decades before 2008, project it forward, and measure the shortfall against what was actually built. The 2008 crash gutted the homebuilding industry and it never fully returned. This method produces mid-range estimates, often 3 to 4 million.

Method three: suppressed household formation. Count the people who would live independently in a functioning market and do not: adults doubled up, adult children at home, roommates by necessity rather than choice. Up for Growth's work uses this framing and lands at the higher end.

The right way to read that chart is not to pick a favorite. It is to notice the floor. Even the stingiest count says America is short more than a million and a half homes.

Are we building our way out of it?

No. U.S. housing completions have run around 1.4 to 1.5 million units annually in recent years (U.S. Census Bureau). Household formation plus the replacement of homes lost to demolition, fire and obsolescence consumes most of that. What remains to chip at the backlog is thin, and in several recent years it has been close to zero.

Run the arithmetic against the middle estimate. A 3.5 million unit gap, closed by an extra 200,000 homes a year, takes roughly 17 years. Closed by an extra 500,000 a year, a construction boom the industry has not produced since the early 2000s, it still takes seven.

Why is the shortage worst at the bottom?

Because the missing homes are missing by category, not at random. The units that vanished are the small, modest, attached and starter homes, the ones a person on a normal income could buy or rent. Luxury units get built because they clear the fixed cost of land, permits and regulation. Cheap units do not, which is why builders only build luxury homes and why starter homes are disappearing.

The result shows up in the price ratio. Median household income sits near $80,000 (U.S. Census, 2023) while the median home sale price runs roughly $400,000 to $420,000 (NAR / Census). That is about five times income. In the 1980s the same ratio was two to three times. Nothing about a family's earning power changed by that magnitude. The supply of homes they could afford did.

Where did the missing homes go?

They were never built, and three forces explain most of it.

The 2008 collapse destroyed homebuilding capacity: firms, crews, lumber yards, lenders. Construction employment took more than a decade to recover, and the small builders who specialized in modest homes died at higher rates than the large firms.

Local zoning capped how many units may legally exist per acre. A city cannot build its way out of a shortage on land where multi-unit housing is illegal, which is the mechanism behind the zoning laws driving the housing shortage.

And investor purchases absorbed a share of the existing cheap stock, converting owner units into rentals. That effect is smaller than the supply story but real, as corporate landlords buying homes documents.

Sources: U.S. Census Bureau, NAR, U.S. Dept. of Labor, Harvard Joint Center for Housing Studies.

What does a multi-million unit gap do to a renter?

It removes your leverage. In a market with slack, a landlord who overprices a unit eats a vacancy. In a market short millions of homes, the overpriced unit rents anyway, to someone with fewer options than the last person who walked away.

That is the whole mechanism behind rent burden. The share of income going to housing rises not because renters got worse at budgeting but because the alternative to paying it is nothing. See what rent burden means and why rent is so high for how that plays out on a lease.

So how many homes does America need, and who decides?

Take the honest answer: somewhere between 1.5 and 5 million, concentrated at the bottom of the market, in the metros where the jobs are. The exact figure matters far less than the fact that no plausible estimate is small and no current building rate closes it this decade.

The gap is not a natural phenomenon. It is the cumulative result of local land-use codes written to protect property values, a construction industry that never rebuilt after 2008, and federal policy that stopped funding the bottom of the market decades ago. Every one of those was a decision. Housing shortages end when the decisions change, and not before. That is the load-bearing piece of why the American dream broke. For the fuller picture, start with the housing crisis explained.