What is driving up housing costs across the US?

Housing prices nationally are at an all-time high, including in many metro areas that were previously considered affordable alternatives to coastal markets. While prices have been rising over recent decades, the average growth rates of housing stock have been in decline. In a new BPEA paper, Edward Glaeser and Joseph Gyourko explore the evolving dynamics of the U.S. housing market, focusing on six metropolitan areas and in particular on steep housing stock decreases in Sun Belt cities. On this episode of the Brookings Podcast on Economic Activity, Gyourko joins Brookings Vice President and Director of Economic Studies Ben Harris to discuss the paper’s findings and potential policies to boost affordable housing supply.

• Listen to the Brookings Podcast on Economic Activity on Apple, Spotify, YouTube, or wherever you like to get podcasts.
• Learn about other Brookings podcasts from the Brookings Podcast Network.
• Sign up for the podcasts newsletter for occasional updates on featured episodes and new shows.
• Send feedback email to [email protected].

[music] 

EBERLY: I’m Jan Eberly, the James R. and Helen D. Russell Professor of Finance at Northwestern University.  

STEINSSON: And I’m Jón Steinsson, Chancellor’s Professor of Economics at the University of California, Berkeley.  

EBERLY: We’re the co-editors of the Brookings Papers on Economic Activity, a semi-annual academic conference and journal that pairs rigorous research with real-time policy analysis to address the most urgent economic challenges of the day.  

STEINSSON: And this is the Brookings Podcast on Economic Activity, where we share conversations with leading experts on the research they do and how it will affect economic policy. 

Americans spend a sizable fraction of their income on housing. Over the past few decades, the price of housing has been rising and is now at an all-time high, higher even than at the peak of the housing boom prior to the 2008 recession. It’s no surprise then that high housing costs have become a concern for economists, policymakers, and the public. 

Today’s episode explores this issue, drawing on research from the new paper, “America’s Housing Affordability Crisis and the Decline in Housing Supply,” by Ed Glaeser and Joe Gyourko, which looks at the historical trends that lead to this moment.  

EBERLY: The authors point out that housing shortages have become much more widespread. We used to think of housing supply being constrained in coastal areas, for example, where building was limited by natural constraints. But now housing supply seems to be limited everywhere, including in cities where building was once extensive and affordable. 

The authors take a hard look at many possible explanations, including commuting costs and other limits to scale. They conclude that in some sense, we’ve done this by choice, in trying to protect the nature and the value of existing neighborhoods.  

STEINSSON: Today’s interview will be led by Ben Harris, Brookings Vice President and Director of the Economic Studies Program, who will be joined by one of the authors, Joe Gyourko of the University of Pennsylvania.  

HARRIS: Thank you, Jan and Jón. So Joe, I’m delighted to be speaking with you today about this fascinating new paper you’ve authored with Ed Glaeser. And so this paper describes the evolution of the U.S. Housing market over the post-World War II period. 

So perhaps you can start by briefly explaining the key trends in housing affordability and how this has played out across various metro areas nationwide? And maybe you could also explain why you focus on suburban markets in the paper? 

[2:45] 

GYOURKO: Okay, sure. One, thank you for having me. It’s great to be here. This is a labor of love. Ed and I have been working on the supply side of housing markets for a quarter century now, and it was a lot of fun and very interesting to take a new look at the situation because things have changed. 

So, Ben, if you’ll bear with me, I’ll start by talking broadly about prices. I know the title talks about supply, and we’ll definitely get there, I’m sure. But we have 50 years of really good data on what economists think of as real constant quality house prices. Think of that as the price of a given home over time. We have measures going back to 1975. 

And in those measures, it’s quite clear that the big coastal markets, think Boston, New York, and Washington, D.C. on the east coast, Seattle, San Francisco, Los Angeles and San Diego on the west coast. Those seven big coastal markets have grown the most over the last 50 years. But what I really want to do to get into the BPEA paper is to break down that 50 years into two 25-year periods: 1975 to the end of 1999, and then 2000 through 2024. 

There’s an index out there. I know you don’t want jargon, but there’s the S&P CoreLogic Case Schiller Price index and the beauty of that is they track 20 markets over 50 years. Now there are seven Sunbelt markets in that 20-city index and for the first 25 years, 1975 to 2000, none of them were in the top 10. Indeed, the highest appreciating market in that 25-year period, Sunbelt market, was Charlotte. It was number 13. And it had a whopping 11% real price growth over those 25 years.  

Phoenix, Miami, Atlanta, Tampa, Dallas, and Las Vegas were actually below them with real price-growth over that quarter century period ranging from –4% in Las Vegas to +9% in Phoenix. All those supply-constrained coastal markets we all know that it had been expensive for a while. San Francisco grew by 160 percent, San Diego was pulling up the rear at 69, everybody else was in between. Okay? So that’s ‘75 to 2000. 

What about the last quarter century? Here, the number one appreciating market of those 20 tracked by the index is Miami. Tampa’s number five, Phoenix is number nine, but all three of them, Miami, Tampa, and Phoenix, grew by more in real constant quality terms than did New York City’s housing market since 2000. That’s the big change, is that you start to see very, very large real price growth in big Sun Belt markets. 

And by the way, New York City grew by a healthy 70 percent, again, real constant quality. Denver’s at 68, Las Vegas at 61, remember Las Vegas was minus 4 the previous 25 years, Dallas 57, Charlotte 50, and Atlanta 32. So what is new, then, is starkly rising home prices the last couple of decades in the Sun Belt. 

What is not new is that the supply-constrained big coastal markets I talked about earlier, they still had pretty strong price growth. The other thing that’s not new, where I grew up in the industrial Midwest, I grew up near Dayton, Ohio, Detroit and Cleveland happened to be in the index I talked about. They don’t grow much at all either and that’s not new. They weren’t growing much in the previous 25 years. 

The new news, Ben, is really strong price growth in the Sunbelt region. That’s what I think got Jan and Jón interested in having a paper on this and it’s what led us to a really fun and interesting investigation of, how come? Why did this change happen?  

HARRIS: So let’s focus on how come. And so, let’s turn to the role of housing supply, which is the focus of your paper in driving these trends. So, Joe, can you provide an overview of the key changes in housing supply across the United States, what you and your co-author characterize as the decline in the intensity of housing production? And perhaps you can explain how this contributed to declines in affordability that you just discussed? 

[7:28] 

GYOURKO: Sure. So, two big changes in supply since 1950. In the paper, we go back to the first full decade after the Second World War. Aggregate supply intensity fell consistently since 1950, really over the last 75 years. And again, I don’t want to bore people on a podcast with a bunch of numbers, but in the 1950s, the average annual growth in the supply of housing, all of it, was 4% a year. There are 10 years in the 1950s in every decade, so 10 times four is 40%. 

In the 1950’s, we increased the housing stock of the United States by 40% between 1950 and 1960. In the 60’s, that growth rate falls a bit from 4% to 3.3%. So we grew by a third, 33%. You keep going. By the 1990s, that annual average growth rate is 1.3 percent, multiply by 10 and we increase the stock by 13. In the 2010s, we increased the stock on an average annual basis by 0.6 percent. So we only increased by 6 percent. That’s the decrease in the intensity. 

Now I want to say something about that. Housing is durable. So it lasts a long time and if population growth falls, either because fertility rates fall or immigration falls, we won’t need to build as much. So nobody should be alarmed by a falling rate of growth of the housing stock unless we’re growing people and then where it’s occurring. 

The second major change in housing markets is we see a dramatic convergence in supply growth rates across markets. Now, in the paper, we tell this story through six markets: Los Angeles, Detroit, Phoenix, Miami, Atlanta, and Dallas. Los Angeles is a representative of a supply constraint coastal market. Detroit is the industrial Midwest decline. And then Phoenix, Miami, Atlanta, Dallas are spread across the Sun Belt, right? From the very east coast all way to near the border with California. 

In the 1970s, I’ll use a couple of decades here. Actually, I’m gonna start with the 60s, if you’ll bear with me. The average annual growth rates were in Los Angeles, 4.1%, 2.5% in Detroit, a whopping 8.9% in Phoenix, 7.6% in Miami, 6.9% in Atlanta, and 6% per year in Dallas. So basically, in the 1960s, our four representative Sunbelt markets grew their housing stocks by from 60% over that decade in Dallas to basically 90% in Phoenix. Phoenix almost doubled its housing stock. 

The spread between Detroit, LA, and the top Phoenix market is six and a half percentage points a decade, which is 65%. That’s two-thirds of the stock. That’s a very big difference. Just go now 30 years later, go to the 1990s, and what you see is Detroit and Los Angeles are out of the building business by the 1990’s. Los Angeles’ average annual growth rate is 0.8%. Detroit’s is a half percent, 0.5. Now, we know they didn’t build, Los Angeles because of regulatory restrictions and Detroit because demand fell. Developers could not make money building houses in Detroit because their population was starting to decline. 

But Phoenix fell from 8.9% a year to 3.3% a year. Miami fell from 7.1% a year to 1.6% a year. Atlanta fell from 4.8 to 3% a year. And Dallas fell from 4.3 to 2%. The spread is no longer 6.5% from the lowest growing rate to the highest. The spread is 2.3%. I want to emphasize that happened in the 1990s, okay? Because part of our story is going to be this trend we see appears to be long run. It is not something just due to COVID or the GFC. It happened before. 

Then in the 2010s, the last decade we have, the spread it 0.7 percentage points. It’s nothing at all. The lowest rate of growth is 0.8% in Detroit. The highest rate of supply growth is 1.8% in Dallas. Basically, in the 2010s, everyone grew by the same amount. So, the story is we have lower production rates in general, but we have really strong convergence so that almost everybody’s doing something very similar on the supply side. That starts in the 1990s and it’s really equal or almost equal in the 2010s. And that is interesting, it’s perplexing, because we know from population growth the demand strength in the Sunbelt markets is really strong and it’s not in the other ones, yet they’re growing at the same rate in terms of the growth of their housing stock. 

So, that’s the big picture on the supply side and to economists like us, you go, wow, if demand is really growing and people we’re flowing into Phoenix but they’ve dropped their production growth rates, strong demand faced with very limited new supply of units is a recipe for high prices, which was my answer to your first question. The interesting thing on the price side is the highest price growth in the last quarter centuries in the Sunbelt, not the coastal markets.  

HARRIS: Yeah, so just to summarize, it used to be the case that in the Sun Belt, you’d see new construction happening at much greater rates. This was a long-term phenomenon that predated COVID. We’ve seen a convergence in the rate of new building across the Rust Belt, across the coastal cities, and across the Sun belt, which is really surprising given the different dynamics in these cities. You’ve got out migration in the Rust belt. You’ve got high regulation in places like Los Angeles, and in the Sun Belt, you just have growing demand. So, it’s odd that it would be all converging. Here, the fact that it looks similar is the real puzzle. Is that a fair summary?  

GYOURKO: That is an excellent summary.  

HARRIS: Okay, so now can we turn to why? Why is this happening? 

[14:14] 

GYOURKO: Why the supply change, not the demand? So, the bulk of the paper now, after we get rid of those facts, is why the supply changed. 

So what we document in the paper is a really dramatic weakening of the relationship between the intensity amount of building of new housing units in a market and prices in that market. So in a well-functioning market, for any good could be housing could be tiddlywinks, could be anything, if the price at which a builder can sell the good the home is above production costs, well they’re going to make a nice profit and you would expect builders to sell more and more, i.e., build more because they can sell more at a profit. In the 1970s, this is using census tract data, that was the decade with the strongest relationship between prices relative to production costs at the beginning of the decade and production over the subsequent 10 years of the decades. That is, in Phoenix, in Miami, in Dallas, in Atlanta, if prices were really high relative to construction costs, they built like mad. That’s how you get 9% a year or 90% increase in the total metro stock in 10 years. 

In the seventies, they built whenever prices made production profitable. You go to the 1990s and that relationship has weakened materially. And then you go to 2010s and it’s gone. It’s flat. Basically, by the 2010s – this is all census tract data now – whether the average price in a census tract is high or low, you build the same amount. It’s a fairly low amount but you basically build the same amount. The relationship between the price and how much you supply disappears over time. Okay? That’s the overall story. 

We then disaggregate. We know exactly where the tracks are, so we know which ones are really close into the urban core and likely to be in the central city and which ones are in the suburbs. So, we redo the analysis just using what we call close in tracks. That is, a census track that’s within 10 miles of the urban of a metropolitan area. So, the downtown CBD of Dallas, whatever. 

We do the same analysis for suburban tracks. Which we just defined, you’re more than 10 miles out. That relationship, I talked about the weakening between beginning of period price and subsequent building. It used to be incredibly strong in the suburbs, much stronger than in the central city, in the close-in tracks. It disappears completely in the suburbs. It weakens but does not entirely disappear in the closed-in tracts in the city. 

So, what’s going on, Ben, is we’re seeing something that’s preventing price from signaling and causing developers to build in the suburbs where we’re actually getting some urban infill in the closer-in tracks. That’s a big change. Turns out the cities are doing a better job with new supply than the suburbs now.  

HARRIS: And so is it fair to say, just as we’ve seen convergence across metro areas, we’re now starting to see convergence within metro areas to say that, whereas at the central city, you might expect the relationship between price and new construction to stay flat, we’re seeing that in the suburbs and the outlying areas? 

GYOURKO: Yeah, it’s flattened way more in the suburbs. 

HARRIS: This is why I thought your paper, well many reasons why your paper was just so incredibly fascinating in addition to being important, which is: if you’re an advocate or just a regular person who’s concerned with housing affordability, you could expect markets to do a lot of heavy lifting, right? So as there’s increased demand, you’d expect to be more construction, but we’re not seeing that.  

[18:10] 

GYOURKO: And by the way, that used to happen throughout the Sun Belt. That’s what’s changed. It hasn’t happened for 50 years in the coastal markets, including the one you live in, but the Sun Belt markets are starting to look like those coastal markets.  

HARRIS: And then so as far as you know, what’s driving this? One thing you guys bring up in your paper is real housing construction costs have gone up by 35% but that’s not the whole story. A lot of times people like to point to zoning and say it’s zoning. It sounds like maybe that’s some of the story, but maybe not the whole story in every market. So, you know, what’s driving this flattening of the supply curves? 

[18:46] 

GYOURKO: So we’ve been, to be fair, we do not answer that question in some really cool causal way in the paper. Let me go back to production costs. 

So, if you remember the numbers, I’m sure you took notes at the beginning on Atlanta’s, in the last 25 years, its real constant quality price goes with 32 percent. The rise in real construction costs can explain the Atlanta market. It cannot explain any of the others. Remember, Miami grew 137% in real constant quality prices since 2000. Construction costs can’t explain it. 

What I think is more likely is regulation, although again, the paper’s long enough already as you can tell from all the numbers, and the editors won’t let us expand it, but regulation is undoubtedly a big part of the story. And it’s not hard to understand in the American context, because we are really almost unique globally in that the locus of regulatory power of land use in the United States is very local. 

I live in Swarthmore, Pennsylvania, and we have a local zoning board that controls whether you, if you were a developer, could build in our community. We do that. It’s not Washington, D.C., it’s not the governor in Harrisburg, it’s us. And that’s true in most jurisdictions in the U.S. It’s either done at the state level and usually the state delegates to local levels. 

So think about how easy it is to stop new production. Then all new development has some negative spillovers. Some of it might just be temporary noise pollution from Mack trucks driving in and out of the construction site near me. That’ll go away over time. But what people appear to worry about is, if you build a lot of housing units with a lot of kids, we may clog the school system that I care about a lot in my community. And I may go, if we need to build a school, I’m not allowing you to develop at all. 

The right economics answer is we would make you, the builder, internalize that so-called negative externality, those are the terms you don’t want used, but it’s just a cost. It’s just the cost that somebody’s got to pay. And in our system, Ben, we don’t really have much of an ability to allocate those costs in a way that people want to approve development. 

And so, I think regulation is a big issue and we have to think that one through in a big way about how we might want to adjust that.  

HARRIS: And so, I know that your paper doesn’t necessarily take on policy prescriptions. It’s a sweeping paper as it is, as you noted. But if you were a housing czar, and you’re one of the top housing economists in the world, so I think it’s a fair question. If you were at U.S. housing czar, what policies would you put in place, or what approaches would you take to address affordability concerns and to boost the stock of housing?  

[21:57] 

GYOURKO: In one of the latter tables in the paper, we show that what’s going on is not just a weakening of the relationship between price and building. There is a really sharp drop in building in low density, high price tracts. So where’s low density? In the suburbs, right? Where we don’t have a lot of housing units per acre. Where are high price tracts? They’re in both the urban core and in the suburbs, they’re expensive neighborhoods throughout the metro area. But when you combine those two trades, low density and high price tracts, we’re talking mostly suburban neighborhoods. 

Now in Atlanta, in the 1970s of all the housing unit production in the metro area from 70 to 80, 90% of all new units were built in low-density, high-price tracks. That is, fairly expensive suburban areas, okay? In the 2010s, from 2010 to 2020, 40%. They had a 50% drop in the share of new production coming from relatively expensive, highly-valued suburban areas. That was Atlanta. In Dallas, the drop is only 12, 13 points. In Miami, it’s almost 30 points. So, it’s not just price, it is price and low density. That’s why I think regulation probably is an important area. 

Now we don’t know for sure, I think we’ll know within a decade, but if it is the closing, it’s economically and socially important. 

Why so now? When your market became supply constrained, New York, San Francisco, there were all these high growth markets around the country with great jobs that people could go to. If what we’re seeing is something like that happening in the Sun Belt. There won’t be other high-growth markets with great jobs. It will be a big change in the United States. That then gets me to, okay, what should you do with respect to policy? The honest answer has been, I don’t know, but it won’t stop me from pontificating, okay? 

So, what would I encourage in this? I would encourage experimentation on almost any ground and level because clearly what we’re doing now doesn’t work. More of the same, strikes me as not a good idea. So, what does that mean? While the locus of control is at the local and state levels, I think we should experiment at the federal level with both carrots and sticks. Why the feds? Because they have more resources to potentially utilize. 

You could think of sticks in the sense of, Ben, if you’re running the suburban jurisdiction that’s low density and high price and you’re not building anymore, a stick might be, you know, we’re not going to give you new park funds. We’re not gonna redo your roads that come through the gas tax and the like because if you’re going to allow more people to live there, why should we do that? Now, that will be politically tough. You work in DC, that using sticks is kind of hard, but you could imagine carrots too, right? Which is, we’ll give you more road funds, and we’ll get you more park development funds if you do that. So, I think at the federal level, we should start thinking along these lines in terms of all types of grants. 

Another way would be to run a tournament, which is ask the states to come up with the best plan for the biggest metro area, whatever metro areas they wanted to talk about in their states, and here’s our plan for increasing supply: one carrot might be, we’ll help fund it, we’ll help subsidize it in some way. So, I think we should think about everything there on the grounds that what we’re doing now does not work. Beyond that, at the state and local level, that’s where the locus of control is. We should start thinking about whether states should take more responsibility and not just delegate to localities. 

Remember my example from where I live in suburban Philadelphia. The local zoning authority, let’s say you would benefit from moving into a new house down the street from me, but you don’t live in Swarthmore, hence you don’t get the vote. And hence, that means we don’t need to care about the fact that you would benefit from moving in. The social planner and economics would care, but that’s just not how politics works. So maybe it would be better if we move that control up to the state, because while I may not care as the head of the zoning authority, well, Governor Shapiro, the guy named Josh Shapiro is the governor of Pennsylvania. He might care, because you might be a resident of Pennsylvania, and he might care. 

We have experimented a bit, the early evidence with the New York and Massachusetts experiments is, we haven’t made coastal California affordable, nor is Boston affordable, but that just says we’re not doing it right yet. 

So that’s a very long-winded way of saying I would hope we would be willing to try many, many things at all levels of government. Watch carefully, use our federalist system because if it starts to work in one area we should just copy like mad. That’s it. I wish that we had a direct answer that I know policy x will work. Honestly Ben, we do not know any specific policy that will work. We need to figure that out. That is the next challenge.  

HARRIS: So, experimentation from your lips to policymaker’s ears. Joe, this has been a fascinating podcast to match a fascinating paper. It really redefined my understanding of U.S. Housing markets. I recommend your BPA paper with that Ed Glazer to anyone with an interest in housing policy and housing economics. Joe, thank you so much for being on today. This is a terrific podcast and I really appreciate it. 

GYOURKO: Great talking to you. Thank you.  

[music] 

STEINSSON: Once again, I’m Jón Steinsson.  

EBERLY: And I’m Jan Eberly.  

STEINSSON: And this has been the Brookings Podcast on Economic Activity. Thanks to our guests for this great conversation and be sure to subscribe to get notifications about new releases of this podcast.  

EBERLY: The Brookings Podcast on Economic Activity is produced by the Brookings Podcast Network. Learn more about this and our other podcasts at Brookings dot edu slash podcast. Send feedback to podcasts at Brookings dot edu and find out more about the Brookings Papers on Economic Activity online at Brookings dot edu slash B-P-E-A.  

STEINSSON: Thanks to the team that makes this podcast possible: Fred Dews, supervising producer; Chris Miller, co-producer; Gastón Reboredo, co-producer and audio engineer. Show art was designed by Katie Merris and promotional support comes from our colleagues in Brookings Communications.