The average square footage in new single-family houses has been declining since 2015. House sizes tend to fall just during recessionary periods. It happened from 2008 to 2009, from 2001 to 2002, and from 1990 to 1991.
But even with strong job growth numbers in recent years, it looks like demand for houses of historically large size may have finally peaked.
According to Census Bureau data, the average size of new houses in 2017 was 2,631 square feet. That’s down from the 2015 peak of 2,687.
2015’s average, by the way, was an all-time high and represented decades of near-relentless growth in house sizes in the United States since the Second World War.
Indeed, in the fifty years from 1967 to 2017, the average size of new houses increased by two-thirds (67 percent) from 1,570 to 2,631 square feet. At the same time, the quality of housing also increased substantially in everything from insulation, to roofing materials, to windows, and to the size and availability of garages.
Source: Department of Labor, Census Bureau, HUD.1
Meanwhile, the size of American households during this period decreased 22 percent from 3.28 to 2.54 people. Needless to say, the amount of square footage per person has expanded greatly over the past fifty years. (Square footage in new multifamily construction has also increased.)
And yet, we continue to hear in survey data that Americans are “overworked,” “stressed out,” and pushed to the limit when it comes to paying for living space. If that’s the case, why do so many Americans continue to buy new housing that’s more than 50 percent larger than what their parents grew up in?
Part of it is a matter of demonstrated preference versus what they say in surveys. The demonstrated behavior or many people is simply that they prefer more house to less, even if it means more stress in making that mortgage payment every month. Another factor is the low-low mortgage rates that continue to be available to a great many borrowers. Sure, that extra 500 square feet above and beyond what your dad shared with 3 siblings might be a bit much, but if you can spread the payments out over 30 years, why not just get it?
How Government Policy Led to a Codification of Larger, More Expensive Houses
But there are other factors as well. In recent decades, local governments have continued to ratchet up mandates as to how many units can be built per acre, and what size those new houses can be. As The Washington Post reported last month, various government regulations and fees, such as “impact fees,” which are the same regardless of the size of the unit, “incentivize developers to build big.” The Post continues, “if zoning allows no more than two units per acre, the incentive will be to build the biggest, most expensive units possible.”
Moreover, community groups opposed to anything that sounds like “density” or “upzoning” will use the power of local governments to crush developer attempts to build more affordable housing. However, as The Post notes, at least one developer has found “where his firm has been able to encourage cities to allow smaller buildings the demand has been strong. For those building small, demand doesn’t seem to be an issue.”
Many involved in home sales likely won’t be shocked to hear this. In many markets, it’s the mid-priced homes that sell the fastest. In the Denver metro area, for example, homes priced in the $300,000-400,000 range are quickly snapped up. But luxury homes coming in around $700,000 or a million dollars can languish. Indeed, the Washington Post article features a Denver-area couple who were delighted to buy a new downsized 1,400 square foot house for $257,000.
As much as existing homeowners and city planners would love to see nothing but upper middle-class housing with three-car garages along every street, the fact is that not everyone can afford this sort of housing. But that doesn’t mean people in the middle can only afford a shack in a shanty town either — so long as governments will allow more basic housing to be built.
Local housing has become so inflexible as a combination of a variety of historical trends which later become nearly set in stone thanks to government policy. We have seen this at work as decades of federal housing policy has worked to encourage ever-larger debt loads which in turn leads to larger houses as well. Eventually, this sort of housing — and the sort of people who live in it — reach a critical mass politically. The people who live in the larger houses then want to make sure that the “character of the neighborhood” is preserved — by force of law — which ends up excluding new types of more economical housing. This doesn’t necessarily mean apartment buildings, of course. It can simply mean smaller, more simple single-family housing. But once existing homeowners begin to dominate the local political process, the deck becomes stacked against new homeowners who can only afford basic housing that the old-timers don’t want to see.
The result is an ossified housing policy designed to reinforce existing housing, while denying new types of housing that is perhaps more suitable to smaller households and a more stagnant economic environment.
Eventually, though, something has to give. Either governments persist indefinitely with restrictions on “undesirable” housing — which means housing costs skyrocket — or local governments finally start to allow builders to build housing more appropriate to the needs of the middle class.
For now, the results have been spotty. But where developers are allowed to actually build for a middle-class clientele, it looks like there’s plenty of demand.
Not Really Downsizing
The Post’s article covering this downsizing phenomena is titled “Downsizing the American Dream,” but this represents nothing that might be called a downsizing when compared to the alleged Golden Age of the American Dream in postwar America.
After all, by the standards of the 1950s and 1960s, the new “smaller” houses remain large and luxurious by comparison. According to a 1956 report by the US Department of Labor, “The 2-bedroom, 1-bathroom house, with less than a thousand square feet of floor area ... typified new houses in 1950.”
Keep in mind, moreover, that the average household size in 1950 was 3.37 (compared to 2.62 in 2000). Those two bedrooms and that one bathroom in 1950 saw a lot more traffic than would typically be the case today.
House-sizes grew considerably into the 1960s, but even those homes — which were often three-bedroom two-bathroom houses for families with children — still came in around 1,500 square feet well into the 1970s.
Today, the average new house has more than 1,000 square feet than a home of the 1960s — often housing no more than a couple and its dog.
But do new home buyers need all that house? It’s hard to know since housing production is caught up in a complex web of government financing, government regulation, and neighborhood NIMBYism.
To know the answer, we’d have to allow developers to build less-expensive housing, but that would require a great simplification of the political and regulatory processes developers must deal with. Expectations for housing have changed so much over the past fifty years, it’s hard to imagine a return to what households of the past would have considered to be normal, middle-class housing.
It would be an interesting experiment, though: would city planners and neighborhood groups welcome a developer who planned to build a neighborhood of 1950s retro housing? That is: new two-bedroom, one-bathroom houses of 1,000 square feet? (They’d have to exclude the asbestos siding typical of the time, and the terrible insulation of the time would need to be replaced with something more modern.)
It would be interesting to see someone try it.
- 1The historical data has been compiled from three srouces. Since 1973, see “Median and Average Square Feet of Floor Area in New Single-Family Houses Completed” from the Census Bureau (https://www.census.gov/construction/chars/pdf/squarefeet.pdf). 1960s data is from “Characteristics of Single-Family Homes: 1969,” a joint report of HUD and the Census Bureau (https://www.huduser.gov/Publications/pdf/HUD-11675_v2.pdf). 1950s and 1940 data is from “New Housing and its Materials, 1940-1956” from the US Dept. of Labor. (https://fraser.stlouisfed.org/files/docs/publications/bls/bls_1231_1958.pdf)