Did your grandparents and great-grandparents really pay less for their homes than you did, even after inflation? The Department of Data might have the answer.
The Painted Ladies at Alamo Square in San Francisco. (R. Krubner/Classic Stock/Getty Images)
This even holds true when the data in question addresses one of the biggest, most important economic objects in the known universe: the U.S. housing market.
Despite its centrality to our lives — and the global economy — we could never say with too much confidence if we’d really paid more for our homes than our grandparents had, after inflation, or which cities had the hottest housing markets in the past century.
For years, the best weapon economists had for this type of analysis was a home-price index going back to 1890 compiled by housing data hall of famer Robert Shiller, the longtime Yale economist. But many folks who used it seemed to have forgotten that even Nobel Bob himself had seemed to view the series as provisional.
“The home price index we constructed … is imperfect, and I hope to improve it someday,” Shiller wrote in “Irrational Exuberance,” “but for now it appears to be the best that can be found for this long a time period.”
He was right! The data, while super valuable, was cobbled together — its earliest observations came from a 1934 survey that essentially asked folks to remember what they’d paid for their homes years ago. That approach beats the bejabbers out of having no data at all, but its reliance on the vagaries of human memory obviously leave room for improvement.
Updating the Shiller series had the potential to unlock some fundamental mysteries about where the U.S. housing market went wrong, but it also had the potential to be a monumental, career-consuming quagmire. Ronan Lyons of Trinity College Dublin, Allison Shertzer of the Federal Reserve Bank of Philadelphia, Rowena Gray of the University of California at Merced and David Agorastos, then at the University of Pittsburgh, took it on anyway.
They spent the past decade painstakingly compiling and processing 2.7 million classified ads from 30 U.S. cities. It was a monumental effort — their acknowledgments section lists more than 100 people and grants.
In the early stages, Lyons remembers settling in on vacation in County Limerick, after his pregnant wife and his 2-year-old were asleep, to pour himself a beer and pore over old listings and figure out the best way to transmute them into usable data.
“This is not machine learning,” Lyons told us. “This is artisanal data.”
This artisanal artistry changed how we think of U.S. housing.
Their index goes through 2006, when — as we may have noticed — the bottom began to fall out of the newspaper classified-ads market. But we’ve extended it to the present day with the help of the excellent home-price indexes from our friends at the Federal Housing Finance Agency.
Their work shows that, after adjusting for inflation, housing prices were essentially flat from 1890 to 1940, with the housing boom preceding the Great Depression being canceled out by the big ol’ bust that followed. Until relatively recently, making a profit selling your home wasn’t really part of the housing equation — your home value probably kept pace with inflation, and that was that.
“Inflation-adjusted capital gains went from being unimportant (close to zero) in the prewar era to quite important in the post-1970 and especially in the last ten years of our sample,” Shertzer told us via email, referring to the 1997 to 2006 period.
We reckon that’s the single most important thing you can take from their data. For generations, homes were a commodity. It’s only recently that they’ve become an appreciating investment.
What changed?
Well, for starters, the hottest housing markets clearly had. For the earliest years in their study, we estimate that Atlanta had the highest prices in the nation, but they were soon passed by the mighty Motor City itself. But as automobiles went from high-tech oddity to high-volume commodity, Detroit slid down the list to be replaced by massive coastal commercial centers.
It probably wasn’t just America’s long-term, world-beating, Soviet-scrapping run of economic growth in the post-World War II era. After all, U.S. cities have beat the world before without blowing up their housing market.
But shifting economic fortunes don’t seem to be the key driver, since not all superstar cities have seen substantial sticker shock. Instead, the first big change may have been driven by the mortgage market. Before the Great Depression, mortgages as we know them didn’t really exist. If you couldn’t pay in cash, you dropped a huge down payment — think 50 percent — and borrowed the rest in a lump sum.
“You had to pay it all back in, say, five years’ or seven years’ time,” Lyons told us. “And if your home’s value collapsed or you lost your job just when you needed to roll over your loan, you were in big trouble.” That big trouble is, in fact, a big part of what made the Depression so Great.
But between the world wars, Lyons said, those who were spending their time reading thousands of old newspaper housing sections would “start to see the first ads where they’re like, ‘Why not pay mortgage like it’s rent? Why not pay an amount every month and then after a certain amount of time, you’ve got the whole thing paid off?’”
And that innovation, with help from the New Deal, the GI Bill and our old friend Fannie Mae, would — by the Space Race Sixties — spawn the particularly American beast called the 30-year fixed-rate mortgage.
That simple phrase probably explains a bit of the housing market’s long post-baby boom boom. As mortgages evolved, buying a home got easier and easier. That alone could push up prices. And that’s before that magical modifier, “fixed rate,” goes to work.
In many countries, your mortgage rate rises and falls with the cost of borrowing. So it’s nice enough when rates drop, but it doesn’t give you a huge incentive to leap into the housing market to lock in that low-low rate for a generation.
In America, as those of us who tried to buy a home in 2021 can tell you, home buyers seem to respond to lower rates with more urgency. In practice, this means falling rates translate to higher prices. And since 1980, when mortgage rates topped an ear-piercing 18 percent, rates have seen a long-term slide — even after adjusting for inflation.
But the alchemy of falling interest rates and fixed-rate mortgages can’t be the true culprit. Because while rates apply to us all, the home price run-up has bypassed much of the country.
While at least four cities saw prices grow fivefold from the 1890s to 2020s, even after adjusting for the cost of living, home prices in many others — Atlanta, St. Louis, Pittsburgh — have barely kept pace with inflation.
If you owned a home in one of those cities in 1890 and sold it now, the purchasing power of your proceeds wouldn’t really differ from the purchasing power of the money you spent on it 135 years earlier.
Meanwhile, doing the same thing in San Diego or Los Angeles might have given you an every-grandkid-gets-a-trust-fund increase of about 1,000 percent, adjusted for inflation.
When we stopped to think about that, we couldn’t get it out of our heads. So many of us have internalized the lesson that homes are speculative, flippable investment vehicles, yet in much of the country — Cleveland, Memphis, Detroit, we could keep going — housing has been a truly quotidian commodity. There, home prices simply keep pace with inflation over the long run, no different from spaghetti or sprockets.
Sure, some of the cities with flat home prices aren’t economic superstars, but the home-price winners didn’t just get there because everybody wanted to move to the same few spots.
Consider that Dallas, Houston, Seattle and Portland, Oregon, all had what the researchers would classify as high demand for housing. But prices in Dallas and Houston have only roughly doubled in price since 1890, compared with a more than sixfold jump in Portland, or almost fivefold in Seattle.
What’s the difference? Lyons had a simple answer.
“If prices go up,” Lyons asked us rhetorically, “does supply come on stream to follow? Do people look to build homes?”
Since 1970, the metros where housing stock grew the least relative to population growth — think Los Angeles, San Francisco, San Diego or Seattle — saw some of the fastest home price growth. While metros that built enough housing — such as Atlanta, Phoenix and Charleston, South Carolina, saw home prices rise much less rapidly, even as their populations soared.
Why aren’t those cities building as much? The economists have an answer — and it’s not just geography. They find home prices grew faster in places with more restrictive zoning, though they warn us that could be because more zoning leads to higher prices, because higher prices lead to more zoning, or — most likely — both.
Hello, there! The Department of Data continues collecting queries. Tell us what piques your curiosity: Do older homes depreciate? At what age does the human mind really peak? Do kids today get more days off from school? Just ask!
If your question appears in a column, we’ll send you an official Department of Data button and ID card.