Building Costs vs. Housing Prices: Why Construction Isn't Driving the Crisis

[MUSIC PLAYING] Historically, one major reason has consistently been cited for the growth and the cost of housing in this country. And that's the growth in what it costs to build homes. But history is changing. If you look at the growth rate of prices and costs, they used to move together on average, but not super closely. What's happened lately is they basically have become completely decoupled. Welcome to The Pie. I'm your host, Tess Vigland. Economists are always talking about The Pie, how it grows and shrinks, how it's sliced, who gets the biggest share. In this show, we're talking about the most pressing matters of the day, seen through the lens of economics. The pie is a production of the University of Chicago's Becker Freeman Institute for Economics. And in this episode, yep, I have a cold. Actually, I'm recovering from about with our old and horrible friend, COVID. I mention it because we pre-taped today's interview a couple of weeks ago. So I'll sound just a little different for that part of the show. Today, we're talking about housing and the cost thereof. The median price of an existing home in this country in June was $435,300, according to the National Association of Realtors. That was a record, the latest in the slew of them, and up 2% from just a year ago. It's taken as an article of faith that a good portion of that kind of increase is because of ongoing increases in what it costs to build a house. But as we're about to hear, that's not necessarily the case. There's a lot more to home in on than just building materials. I'm Chad Sieverson. I'm the George C. Chao Distinguished Service Professor of Economics at the University of Chicago Booth School of Business. Well, Chad, let's start with some assumptions or ideas that are, I think, taken as sort of gospel when it comes to housing prices, which hit a record high in June of this year. And the one that gets probably the most traction is that we have a supply problem, right? Not enough homes. And that that is in part because of the sky high cost of building them. Can you walk us through that historically? Has that been the case, at least, in theory? Yeah, the basic economic theory would say the price of homes is a reflection of the combination of supply and demand factors. And one big supply factor is the cost of production. And if you look over the sweep of history at a very broad level, yes, it's got more expensive to build houses. And the price of houses has gone up. And this is all sort of relative to inflation. So it's not just everything gets more expensive because inflation, but even relative to everything else, both the price of houses and the cost of building them has gone up over time. However, if you start breaking it down over the last, say, 75 years where we've got better data on it, for a while, the cost of building actually grew faster than the prices of houses did. But more recently in the last half century or so, and especially in the past, let's say 20 to 30 years or so, house price growth has considerably exceeded the growth of the cost of building houses, at least in national averages. So while they are moving in the same direction in terms of the quantitative connection, the relationship between prices of houses and the cost of building them has gotten looser and looser over time. You know, there's this saying that there is no national housing market, right? Because everything varies so widely, not just among regions, but even interest state, interest city. Does that ring true for building costs? That's true to some extent for costs as well. Although if you look at the variation across cities in building costs, it's not nearly as large as the variation in house prices is. So while there's definitely more than zero variation, it's not as big as price variation. And if you break down those building costs into two components, one is sort of the materials you need to put house together, the lumber, the drywall, the concrete, et cetera. There's less variation in those things than the rest of the building costs, like how much you have to pay workers to come assemble the house and contracting fees and stuff like that. And that kind of makes sense if you think, well, those materials, they can be traded and moved across space to some degree, where it's much more difficult to do that with workers. And so that's why you get less variation in building costs coming from materials and more coming from the labor component. But all that said, even both of those have less variation than the price variation of houses itself. - And I should mention just briefly that we're really not including land costs here, right? This is like where the land has been bought, so we're looking at the building costs. - Yeah, that's a very important point. Yeah, so the, of course, whatever the land is worth is in the price of the house. So when I talk about house prices, that's in there. But when I talk about building costs, it very specifically excludes the price of land. It's really about, hey, I got this box of wood and concrete and granite, et cetera, et cetera. I got to put it together. How much is it going to cost me to assemble this box on a chunk of land that's already been bought? - Okay. So how do you go about measuring those construction costs? What kind of data are you looking at? Most of the paper uses RS means data to measure construction costs. Now, this is a data set that's used widely in the construction industry. And it's a company that has been operating for a long time, basically has folks out on the ground serving construction companies as well as the companies that sell construction materials. And at a very detailed level saying, okay, here's what it costs to get a square foot of, granite countertops these days. Here's what it's going to cost for so many board feet of two by fours. And here's what you have to pay to get a plumber to come work for a day. And they add all that stuff together for a typical house and you can actually vary what you define as typical if you've got this data. How much it's going to cost to build that house or in some cases, how much it will cost to build, a house of a certain type per square foot of that house. And so it really is a kind of ground up aggregation attempt to take prices in construction and construction material markets in pretty close to real time and then put this cost index together every quarter and every year. - You chose to focus on certain cities and you went back to 1950. Oh, the good old days that I don't remember. So which cities did you choose and why and why 1950? - So we went back to, as far as we could and still get pretty decent data basically. And the 1950, of course, the post-World War II sort of was the start of the real big, the massive housing construction boom, of course. And so that seemed like a logical place to start and the data was available back to then. The RS means data starts then. And then in terms of the cities we looked at, we looked at whatever RS means has have broken out by city. And then we combined it with house price indexes that were also available by city. Those didn't really come along the city level stuff didn't come along until the late 70s. And so all the sort of across city comparisons we make in the paper are sort of post 1975. The longer sweep versions were just about national averages. - Okay. So let's talk about some of the patterns that you did start to see. One of which was comparing the rise in construction costs to the rise in the consumer price index, the CPI, which measures overall inflation. Tell us a little bit about that. - Yeah, so construction costs have generally exceeded inflation, which is actually in some ways it's a little surprising and it's also kind of related to work that I had done earlier with Austin Gools B. And the surprising bit is if you look at in general the price of physical stuff, so what an economist would call goods as opposed to services, you know. That stuff is actually tended to see price increases that are slower than inflation, at least over long run periods. And that's because we tend to have pretty fast productivity growth in goods production or we have had historically that's been the case. And so relative prices of goods compared to services and other things in the economy have tended to fall. So goods prices aren't rising as fast. Now you might look at a house and say, well that's a physical thing, that's a good. And you know, it's assembled for use the way a car is. at least they might think so. And it's true that the. prices of cars have fallen relative to inflation over the logs we've of history. But you don't see that with the costs of putting houses together. So there's something about we're not seeing efficiency gains in putting together house goods in the same way we are with with other goods. And this is tied to this work I did with Austin Goolsby, which looked at productivity in the construction sector over the log run. Also going back to 1950. And there we found that since 1970 productivity growth in the construction sector has actually been negative, which is really unusual, and indicates that cost reductions in construction just have not been forthcoming. And you compare that to say manufacturing in the US manufacturing scene really big productivity gains over the past half century construction has not. And so that might be in tied part to this fact that we are seeing construction costs of houses go up faster than inflation while other costs of putting together other physical stuff like cars and airplanes, etc, etc have been growing more slowly than inflation. So basically what you're saying is the robots are not taking the construction jobs. That's one way, you know, that's one possibility. It turns out, yeah, capital is one thing that actually leads to productivity growth. Turns out Austin, I didn't find a lot of evidence that the construction sector was under investing in capital. There are other problems going on, but you certainly can say that they didn't have so much automation that it drove down their costs because that certainly has not happened. All right, so then you also compared that growth in construction costs to the overall rise in housing prices. And you've mentioned this already, but can you walk us through that finding a bit more? Sure. So, you know, 1950 to the mid-70s or so construction costs actually grew faster than house prices, but then it turned around and in the 50 years sense, house price growth has considerably exceeded construction cost growth. Now, how fast that differential has opened up has varied over those last 50 years. There's cross-sectional variation across cities and the relative rates and so on and so forth, but just talking about long-run national averages, you could just summarize it as saying over the last half century house prices have gone up faster than the cost of building houses and by quite a fair amount. What were some of the cities that saw the biggest gaps between the hike and the cost of building a home and the growth in housing prices and did those trends stay true over time? Sure. So, in general, there is some evidence that cities that see faster house price than construction cost growth in one five-year period are less likely to see it in the next five-year period. Nevertheless, there are some cities that over the decades have seen quite a few more periods of faster price growth and cost growth. So, maybe these won't surprise any of your listeners. You're talking about, say, California coastal cities, San Jose, San Francisco, Los Angeles. They've tended to have over the long run a much higher growth in house prices than in the cost of building those houses, but it's not just out in California. Miami has seen that as well. New York City, Seattle, Washington, mostly coastal cities. I think that is fair to say, at least over long periods of time. It's mostly the coastal cities, although within any given, say, half decade period, you can have other cities see quite a bit faster house price growth and construction cost growth. And I'm curious, how do you account for elements like the size of homes? I mean, building costs in a place like Los Angeles with multi-gazillion dollar mansions, that's got to be a different beast from mid-town America with ranch homes or, I don't know, even a craftsman. Yeah, that's fair. So, the city-specific construction cost indexes, they really are city-specific. So, if there are systematically different things about building houses in, say, Los Angeles, compared to Houston or Portland, Maine, whatever you have, those are captured in the index because those things are put together. We also do a comparison in the paper in terms of price versus cost levels where we literally compare not just the cost of and price of building a home, but the price per square foot to the cost per square foot, which RS means actually makes available those per square foot costs as well. So, cost per square foot and what was the other one? So, one is to just at the city level, you've got the average prices of houses and the average cost of building houses. The way that the cost index is put together by RS means is city-specific in the fact that if you use different things to put together a house or more things to put together a house in Los Angeles, then in, say, Houston, the cost index for Houston will reflect those different things than the cost index for Los Angeles. So, as we're comparing prices to cost within a city, in some sense, we're already getting those differences in terms of what goes into the house and how big the houses are. Separately, we do another analysis where we compare not just the price of a house to the cost of building the house, but the price of houses per square foot in the city to the cost of building a square foot of a house in the city. So, we're actually literally making a comparison on the square foot level. So, we're sort of taking the house size differences out altogether. Okay. You also made note in the paper of some differences before and after COVID, presumably the supply chain issues affected construction costs, like they did pretty much everything else that we buy. And I'm curious if you saw spikes one way or the other, not just with COVID, but I don't know, going back to the 2008 financial crisis. Were there spikes of times when it certainly did make sense that construction costs would be a major part of why home prices would go up or was that really even over time? That's a great question and spikes is a good word to use. Basically, there were two periods where there was the biggest difference between house price growth and cost growth and in the positive direction that is prices were going up fast and the cost. And that was just prior to the global financial crisis, the sub, if you want, to call it the more industry appropriate, in this case, the subprime mortgage crisis of the mid-2000s and then the during and post-COVID period. Okay. So those were the two periods in our data where again, house price growth was faster than cost growth by the most amount. What's different about those two is that that gap in the subprime mortgage crisis period was all coming from the fact the prices were going up. It wasn't because costs were going up and prices were going up but prices were going up. More, it was just all prices. Cost growth was not excessive at all in that period. On the other hand, during the post-COVID period, both of them were going up. So costs were going up and prices were going up. It just happened to be that prices were going up so much faster. You still had this really big gap between price and cost growth. But you definitely did see the sort of residue of the supply chain problems that we saw coming out of COVID show up in the data. So costs were going up a lot and a lot of those costs were coming from materials which I mentioned before. So you can break down costs into materials components and sort of a more labor intensive component. Labor was going up too, but you could really see the spike in materials costs in the COVID period compared to say the decade before. So that's an interesting contrast between two periods that both a divergence between prices and costs. It was sort of despite the increasing costs that was happening in the COVID period. So interesting. So Chad, this is a working paper, but can you give us some takeaways here after looking at 75 years of housing costs and related building costs? How much of a relationship is there? And what does that tell us about why home prices are so sky high right now? Or maybe it doesn't tell us anything is your point? No, I think the takeaway is there is a connection between cost and prices and I guess you'd expect that. But it's never been super strong and it's just gotten at least well, I should be a little careful. If you look across cities, you'll see that cities where it's more expensive to build houses have higher house prices. Now, said is a ton of variation in house prices even for cities that have pretty similar construction costs. So house prices move around for all sorts of other reasons besides construction costs. So if you look at any say five year period, it's extremely hard to predict which cities see the highest house price growth just by looking at their cost growth. There's basically no real connection anymore between those two growth components. So whatever connection there was in growth rates and it wasn't ever super strong seems to have disappeared. There's still a little bit left over in price and cost levels as you look across cities, but in terms of changes, that's gone away. And I think that's one of the big conclusions of the paper. And if you want to sort of take that and run, you can say, well, okay, if it's not costs, changes that are driving house price, changes, what else is it? Yeah, well land prices are probably a big part of that. Changes in land prices, and then that leads to the obvious next question is, okay, what explains changes in land prices? And I think there we go to a big public policy debate today, which is about land use rules and permitting and things like that. How easy is it to actually find a plot of land? You can build something on. And all of that sort of gets capitalized into the price of the land itself. It's not going into our construction cost index because that again is you've already got the plot of land. I was at cost, you put the box on it. This is about what is it cost you to get the land that basically gives you the right to build a house. And there, many other things seem to be driving the variation in that. All right. Well, Chad, you are actually the first guest over the last year who has now been on twice and you've already been through our lightning round. But I'm going to ask you again because it's been a few months. I'm going to ask you a couple of the questions, two out of three. So what books are you reading right now or would you recommend? I'm going to assume it's a little different from last fall. Yes, actually, I've got one I just finished, which is Hate the Game by Darryl Fairweather. Darryl is a university Chicago PhD alum who is now the chief economist at Redfin. And she wrote a book about economics as it applies to, well, just all sorts of things. And it's really, you know, there are a lot of like, hey, here's economics in real life kind of books. This is different from any other one like that I've read. Really unique, very entertaining. But also, you want to see how an economist thinks inside their head when they're walking through the world. She basically writes it down. I really, as I read this, it was just like, yep, that's kind of how we think about these things. And so it's a way to sort of see into the mind of an economist. And the way we use our models to think about every day kind of stuff. I think you might find it fascinating. So that's Hate the Game, not the player and the player would be that's a good question. She explained it for God. Okay, I'll read it. I'll read the game. The subtitle is economic cheat codes for love, life, and work. Oh, wow. Yeah, and it's Darryl Fairweather. All right. And finally, Chad, if you had to make a pie with the current contents of your kitchen, what would it be? Now this might have been the answer. We've got raspberries, blackberries, and cherries from our tree in the backyard, actually. Well, that's cool. Yeah. So how about that fresh? All right. Mixed berry pie. Yeah, mixed berry pie. I think we talked to you in September, October, last year. So it was probably different berries. Maybe, but we're pretty regular blackberry raspberries. Okay, fair enough. Well, Chad, thank you so much for talking with us today. You're welcome, Chad. I enjoyed it. The pie is a production of the Becker Freeman Institute for Economics and part of the University of Chicago podcast network. If you'd like to keep in touch with the latest economic research from the University of Chicago, you can visit bf i dot u Chicago dot edu slash subscribe. Carry the two is the show that pulls back the curtain to reveal the mathematical and the statistical gears that turn the world. Co-host Katie Wood Koski and Ian Martin bring unique perspectives from the fields of mathematics and statistics to convey how mathematical research drives the world around us. With each episode tackling, with each episode tackling a different topic, subscribe to Carry the Two, part of the award-winning University of Chicago podcast network. Our theme music was composed by story mechanics production assistants from the BFI communications team. I'm Tess Vigland your host and executive producer. Thanks for listening and we'll see you next time.