What September’s weaker homebuilding estimates mean for buyers

Buyers waiting for more newly built homes received a weaker signal in September: U.S. single-family construction starts were estimated to fall 4.6% from August. If sustained, a slower pace could narrow future choice, even while builders finish homes already underway.

The Census Bureau and HUD release, published October 20, put single-family starts at an annualized pace of 859,000, against a revised 900,000 in August. The uncertainty matters: applying the published margin of 10 percentage points gives a range from a 14.6% decline to a 5.4% increase. Because that includes zero, the survey cannot confidently establish a monthly fall.

Think of three moments on a building site. A permit is local approval to build. A start generally means excavation has begun for the footings or foundation. A completion usually comes when a house’s finished flooring is installed; occupancy can qualify it sooner. These different stages tell different stories about supply.

The headline figures are seasonally adjusted annual rates: monthly activity is adjusted for recurring seasonal patterns, then multiplied by 12. They describe a pace, rather than a forecast for the year or the number of homes currently under construction. Census definitions make that distinction explicit.

September’s single-family completion pace was 840,000. These totals do not follow the same group of homes through a single month. Subtracting completions from starts cannot establish a housing shortage, and neither number predicts the price of a home in a particular town.

Permits for privately owned homes offer another useful distinction. Single-family approvals fell 4.0%, yet overall approvals slipped just 0.2%. Authorizations for homes in buildings with at least five units rose 5.9%, Reuters reported. Here is how the broader offset works:

These are annualized rates, not September counts. Other home types are calculated by subtracting single-family approvals from the Census totals. Their increase offset most of the single-family decline, concealing a shift in the homes being authorized.

The pullback can make sense for builders even when would-be buyers want more choice. Reuters linked September’s weakness to expensive mortgages and an overhang of unsold new houses. Santander economist Stephen Stanley told the news agency that slowing construction could help prevent inventories from building further. That is an assessment of builders’ incentives, rather than proof that housing needs are met everywhere. Reuters

Borrowing costs can also squeeze both sides of the transaction. Higher mortgage payments can reduce what buyers can afford; more expensive construction finance can discourage builders from beginning projects. The Federal Reserve described both mechanisms in its March 2024 housing assessment, including tighter bank lending standards for builders. That historical explanation helps show how weaker demand and less building can coexist; it does not measure the causes of September 2026’s change.

For subsequent context, Freddie Mac’s October 15 survey put the average 30-year fixed mortgage rate at 7.40%, unchanged from the previous week. It is a national benchmark drawn from mortgage applications, not an individual buyer’s offer. Its October timing also matters: it describes borrowing conditions after the September construction activity.

Even a completion is not necessarily another home available to buy. A new house can be counted as sold when a contract is signed or a deposit accepted, before construction begins. For a buyer looking now, the homes still available locally matter more than the national count of newly finished buildings.