Where we build homes helps explain America’s political divide
But for some reason, the law of supply and demand appears to have broken down in the U.S. housing market. Though prices are hitting record highs, available homes remain at record lows.
What’s going on here? We began our inquiry in those few corners of the nation where builders do appear to be heeding the clarion call of fast-rising prices: Florida, Idaho and the Carolinas, the top housing permitters in the country relative to their population, according to the Census Bureau’s Building Permits Survey.
Not coincidentally, all four of those states have seen some of the fastest home-price growth in the nation since the pandemic began, according to Zillow data. But their markets have cooled as the Fed has raised rates and home builders have caught up to demand. Idaho, in particular, has seen prices fall 9 percent since their 2022 peak — though they remain 52 percent above pre-pandemic levels.
The Census Bureau survey reaches out to about 8,400 local jurisdictions every month to ask how many housing units they’ve authorized, how much they’re worth and whether they’re for single or multifamily units. A recent expansion of the monthly permit data includes almost every county in the country, allowing us to paint a beautifully complete picture of where new homes soon will be built.
Housing permits are remarkably good predictors of future home-building. Almost every U.S. home (about 99 percent) is permitted, and almost all single-family permits lead to new housing. In addition, about 4 in 5 multifamily permits end in completed apartments or condos, and many of those that aren’t completed are reclassified to become single-family developments.
It’s true, however, that homes are taking longer and longer to complete lately. Supply-chain issues, labor shortages, volatile-and-rising prices and a shift toward sprawling multifamily complexes have made for a record number of homes under construction this year, even as the number of homes permitted and completed remains far from record highs.
The new housing mini-boom in places such as Florida and Idaho looks very different from the tract-house tsunami that swamped the global economy in 2007. Many of the new units are in apartment buildings and condo complexes. In fact, South Dakota, Colorado, Florida and Arizona now trail only D.C. in the production of multifamily housing units. In some recent months, even states as rural as South Dakota and Montana have seen the majority of their new housing units coming in multifamily complexes.
Meanwhile, many of the states producing the least housing are the traditional multifamily powerhouses, states larded with dense, high-rise-heavy megalopoli such as Pennsylvania, Illinois and New York.
You may see a political tilt to this data. Seven of the 10 top home-building states voted for Donald Trump in 2020, while seven of the bottom 10 went for Joe Biden. But seven out of 10 isn’t the sort of clean sweep that sets off our data detectors. Overall, we saw little relationship between a state’s politics and its housing production.
U.S. politics do polarize along rural-urban lines, of course. But while you might think the least-dense places are the ones building the most homes, we found little relationship between how rural a county is and how many homes it has permitted. The variables that relate to home-building — job growth, newer housing stock, higher migrant population and so on — tend really to be reflections of population growth.
Saying that housing and population growth correlate is like saying chickens and eggs correlate: Of course they do. The real prize is teasing out which came first. Does building homes keep prices low and attract newcomers, or do developers merely respond to demand from folks who were coming anyway?
It’s probably both, said Adam Ozimek, a Lancaster, Pa., bowling-alley tycoon and chief economist at the bipartisan Economic Innovation Group. “Some places build more when they get population inflows,” Ozimek told us, “but population inflows are also driven by which places are building more and thus more affordable.”
Because housing regulation is local, we recalibrated our political analysis to focus on counties instead of states. And there, clear political divides do emerge. Red counties permit more housing than blue ones. Even in states that voted for Biden in 2020, the typical Trump-voting county permitted twice as much housing per person as its Biden-voting neighbors and rivals.
That could indicate that these localities can build more cheaply and be more responsive to housing needs, as Ozimek says. Consider one of the few clear political trends we saw in the data: The more remote workers a red county had from 2017 to 2021 — during the pandemic remote-work population shuffle — the more housing it’s currently building. Meanwhile, even blue Zoomtowns beloved by remote workers aren’t building much more housing than less-desirable blue locales.
But why? When we cornered Chris Herbert, director of Harvard’s Joint Center for Housing Studies, he humored our endless speculation about more restrictive zoning, NIMBYism and environmental regulation in blue counties. And then he gently explained the more mundane reality: It all boils down to land availability.
Zoning, NIMBYism and regulations — “all those things matter” when you’re trying to build housing, Herbert said. But land scarcity is the most important.
Think about the housing boom that preceded the Great Recession. Many of those new houses were erected on marginal land. After lenders were devastated in the housing bust of 2006 and 2007, the funding and motivation to build sprawling subdivisions on suburban outskirts dried up.
So what’s happening now “is a lot more infill of single-family housing in closer-in communities, where you’re not going to have room for large-scale developments and where the land is going to be worth a lot more,” Herbert said. Single-family land scarcity, he said, “has been a big factor keeping the supply down.”
Inspired by Herbert, we dug deeper and realized that this one variable — how much land is available for development — could help explain some fundamental facts of American geography.
In a blockbuster 2010 paper, Albert Saiz, an economist at the Massachusetts Institute of Technology, analyzed satellite data to estimate how much land was actually available for development within a 50-kilometer (31-mile) radius of each major U.S. city. He found that available land, when combined with measures of land-use regulation, could go “very far to explain the evolution of prices” from 1970 to 2000.
Saiz even took it a step further, showing that a lack of land can cause stricter regulation. If a place has less room to build — due to mountains, wetlands or oceans, for example — each square foot of dirt costs more. Homeowners also may push local officials to regulate the land more aggressively in an effort to protect their investment and safeguard a scarce resource.
From 2013 to 2018, zoning and related restrictions added about $410,000 to the cost of a quarter-acre lot in the San Francisco metro area, $199,000 in Los Angeles, $175,000 in Seattle and $152,000 in greater New York, according to an analysis by economists Joseph Gyourko at the University of Pennsylvania and Jacob Krimmel, now with the Federal Reserve.
The comparable figure for Phoenix sat at $22,000. Atlanta was $15,000. Dallas was a mere $2,000. Not coincidentally, perhaps, many such Sun Belt metros have produced floods of new housing.
But why do blue cities tend to have less land available for development? Perhaps it works the other way: Perhaps land-restricted places tend to evolve into Democratic strongholds.
We don’t have data for this, but logically higher home prices and regulation in land-light cities should make much of their housing accessible only to educated, well-compensated professionals, right? In this simple mental model, coastal cities have less room and thus, by definition, attract the elite. And in American politics right now, Democrats dominate the professional classes.
We’ve long heard Democrats derided as the “coastal elite,” but we never stopped to wonder why all those blue counties hugged the coasts in the first place. Exceptions are easy to find, but the subtle effects of coastal land shortages, over time, could help explain that most prominent feature of America’s political geography.
That effect could be compounded, Saiz told us, by the simple truth that coastlines, lakes and other natural obstacles to construction make cities more beautiful, and thus more desirable to those who can afford such amenities, as his research with Gerald Carlino of the Federal Reserve Bank of Philadelphia shows. And the presence of an educated workforce will cause the city’s economy to grow faster, further expanding economic divides.
“High-amenity areas are more desirable and tend to attract the highly skilled,” Saiz said. “These metros tend to have harder land constraints to start with, which begets more expensive housing prices which, in turn, activate more NIMBY activism to protect that wealth.”
The pandemic may have disrupted that loop. Paraphrasing a recent newsletter from onetime Bureau of Labor Statistics economist Joseph Politano, it seems that if we can’t build more housing in these big, in-demand cities, then perhaps the current migration across our geographic divide, to (often beautiful!) red places building more housing, could be the next best thing.
Hi, friends! The Department of Data covets queries. Where are measles on the rise? Which animals are most common in U.S. zoos? What are the most popular baby names in Japan? Just ask!
If your question inspires a column, we’ll send an official Department of Data button and ID card. This week we’re going to break slightly with precedent and mail a handful of buttons to the Building Permit Survey staff at the Census Bureau. Their expansion of the survey’s county data inspired this article, and they’ve been super responsive to our endless questions about the data.
Posts
