This week we’re joined by Chris Redfearn of USC and Anthony Orlando of Cal Poly Pomona to discuss their paper "Houston, you have a problem: How large cities accommodate more housing."
We talk about why "pro-business" Texas housing markets are catching up to "pro-regulation" California and what it might mean for future city growth.
Scroll down below the audio player for an edited excerpt of our conversation, or click here for an unedited, AI-generated transcript of the entire conversation.
Jeff Wood: I want to kind of crystallize what you all are saying in the paper, which is basically, cities accommodate growth by building more housing units on the periphery, then when that land "runs out," eventually, the cheaper land is consumed and developers begin to focus on infill, which is what Anthony was talking about. And then the number of tracks participating in growth is reduced over time suggesting the ossification of neighborhoods. So those are the three things that I picked up from the paper is that, like, we grow until we can’t, we start to do infill — which is more expensive — and then neighborhoods are also ossifying at the same time.
I want to talk about that kind of change from the growth on the outside to changing to infill and some of the kind of construction things. Because part of the paper that was really interesting to me as well is thinking about construction costs or land costs or the things that make it harder to build after you’ve run out of land on the periphery to what I would call sprawl.
Chris Redfearn: Yeah. So the reason I’m smiling here is just because this is a big can of worms, right? The paper I think in some ways is very descriptive and very aggregate in some ways. We dive in and do this mapping locally, which is I think, you know, a step in the right direction, but there’s still lots of details left to be sorted out here. I think, like a lot of urban housing economists, about development as a black box. When prices rise there’s a force at work, and some atomistic small industry that doesn’t have much industrial structural power goes ahead and builds things and we get outcomes, but I think Anthony and I have both spent enough time with enough actual practitioners to see that the difference between single-family home construction and multifamily construction and type one, you know, tall construction. They are really different and they have different sets of risks and different sets of capital providers.
They’re just different and they work in these different orbits. That’s the starting point there, we need to think about differentiating these things. What you get with the infield development is that the scale is just gone, right? So the master plan developers in the periphery — like so much of LA was built by master developers who would tie up 10,000 acres and they would subdivide and phase and they would put in the schools and the infrastructure and they had a plan in mind. But, you know, we talked before about how hard it is to change cities once you build. Going back into in infill means finding a 50-by-150 site and then squeezing on it as much as you can. And that’s hard. I mean it, it’s just super hard from a design perspective. You want do subterranean parking on a site like that. You lose a bunch of land just to getting an underground or not. That’s really expensive and that makes it more expensive.
One of my students is an infill developer and he tells me these horror stories about when you try to do a small site, you can’t get the plumber to park on the project. Like, you’ve got this site and you want your plumber to drive right up with all his tools and all the piping and all that kind of stuff and they may not be able to park on the site, he has to park down the street. And so just little things like that make development more complicated. It’s an entire headache getting his crew on the site and the materials parked and in a way that greenfield [development] never has to do. With greenfield, you’ve got acres of space and you can park anywhere where you want and just start tearing up dirt and putting infrastructure in. It’s really easy and when you inherit some other site you have to go underground and upgrade all the facilities, which is not trivial.
And there’s a bunch of infrastructure you have to sort of think about working around, which you don’t have to do with greenfield. And so we just like inherently the entire business model changes in ways. So when we start rotating from exurban greenfield development into infill development, just the level of complexity, cost, time, horizon risk, all those things get higher. I think part of what we’re seeing in the falling elasticities of housing supply is exactly this — it’s just harder to do.
Anthony Orlando: And before all of that you’ve got to get a parcel that actually works for an infill development like that. And so there’s a lot more land assembly that’s required for infill development and for multifamily than there is for greenfield and single family. There aren’t large parcels just freely available in the city center anymore. You usually have to piece together multiple parcels and sometimes they are odd shapes and odd sizes and, you know, there’s good work showing that the more land assembly you have to do, the more expensive the development. In a place like Los Angeles County, almost half of large developments these days require some form of land assembly, according to the latest studies that we’ve seen.
Chris Redfearn: We actually look at, sort of, where are there big sites that are heavily amenitized? There are these post-industrial sites, right? So the thing about the Arts District in Los Angeles, a place that was largely abandoned as manufacturing left California, that’s a big site and it’s relatively near lots of amenities, right? There’s downtown, there’s transit, there’s lots of things there. So you know, one of the areas that we’ve seen a lot of development is in these post-industrial sites. And that’s true in Houston, it’s true in lots of places is where manufacturing’s rotated. But that’s a finite number if you look at the map and sort of how much area is there, you know, there’s not much of it. And so we’re filling like this, the latest splurge in where we’re, we’re getting units built is we’re not getting it built in the much in the infill. We see like we have lots of fights over infill, it’s not a big number. Most of the numbers are still exurban, and then we see pos-industrials being a significant channel also.
But my big frustration there with the infill stuff is that they build these stumpy buildings. This is your chance to build towers, and if you build four-over-twos there, you’re gonna get stuck with four-over-two density and it’s gonna take another generation or two to or through it into a higher density. So there’s this kind of interesting story about we’re building inelasticity in those places, because we’re making it lower density than we ought to.
Jeff Wood: And related to that though, you’re talking about how the changes happen in these different states. So California versus Texas. You have California which hit that edge and started to do more info early on, and so you have the higher prices. We keep on talking about running out of land, but nobody ever runs out of land. It’s just like running out of huge tracks of property for single family home development. But they started that earlier and then Texas might be hitting this wall soon too. So that is another part of your paper that I think is really interesting, that you’re basically saying the regulatory environments maybe don’t matter as much as we say they do and Texas is just behind California and maybe a place like Las Vegas, which I sent you an article saying Las Vegas is saying that it’s gonna run out of land in eight years. Maybe Las Vegas is just behind Texas.
Chris Redfearn: Phoenix is in the same boat too. They’re running out of land also. So you know, I think we’re — I don’t know Anthony if you wanna say we’re willing to make a a strong statement about this — but I think we’re suggestive but we’re still working on it. We hate to sound like an economist here, but I think what we’re really pushing here is — we know that regulation matters at the margin. And we know that topography matters at the margin. The literature has really demonstrated that those, they’re abundantly true. And so the question is, is that explaining everything? And that’s where we are — we’re looking at other margins here. In this case the land assembly story is one of them. And the other thing we haven’t really talked about is filtering, right? How did we provide more affordable housing in the past? That’s when we were going out to the greenfield. So the lack of the greenfield isn’t necessarily a bad thing by itself for other reasons, but that just means that the chain of houses that went from a upper middle class house that was built in the seventies and then someone left to go to a bigger house further out, then a bigger house further out, left behind a depreciated house.
That’s been the primary mechanism for delivering middle class housing, that now we have people moving infill. And that’s tying this back to the whole story about amenities; at some point people have preferences about what they like. And so at one point people just wanted a big house away from everybody like getting away from noxious downtowns. There’s the primary move after World War II. A lot of downtowns were really industrial and polluted and crowded and people wanted to get away from that. So we’ve been suburbanized, you know, for 150 years. But you know, once you get out far enough out, you got a big enough house and you got those things, you also want some other amenities and so you wanna move back in. I think what we’re gonna see is this dynamic about where people are in their cycle — we’re gonna have land development that is running out at the same time people want higher amenities and both those things are gonna work towards more infill.
Anthony Orlando: One way that we visualize that in the paper is we map out what the distribution of neighborhoods looks like over time in these different cities. And it’s not surprising that in Los Angeles you find that the number of neighborhoods that have just a ton of vacant land goes down over time. But the number of neighborhoods in Houston goes down over time, too, that have a lot of vacant land. It’s not like we’re filling up all the land. You never fully run out of land in the United States. You can always go further away. But within what we consider to be the metropolitan area where people are willing to commute, the number of neighborhoods where there’s a lot of land available has gone down dramatically.
And it’s really actually quite stunning to see in the Texas cities — how the changes in those graphs look a lot like what Los Angeles’s graphs looked like just a few decades earlier. Now I don’t wanna say, and Chris was trying to make this point too, we don’t wanna say that regulation doesn’t matter, it does matter. There’s a lot of literature suggesting that it does. It’s not surprising. You make it harder to build, you make it more expensive to build, it’s gonna be more expensive to get housing. But when we first started studying this and we start first started presenting some of our findings to other urban economists and planners and policy analysts, maybe about five or six years ago, one of the first presentations we gave, we went through a list of studies by other economists that looked at the actual impact of land use regulations in different cities.
So, for example, if we were to change the zoning in one city to a different type of zoning, how many new units would we actually be adding and how would that actually impact prices? Economists have been good at studying this and they’ve found statistically significant effects but what’s the magnitude that we’re talking about? We started adding up the magnitudes from their own studies and what we found is if you change some of these regulations, you don’t come anywhere close to adding enough units to actually close the gap in what anybody measures to be the housing shortage that we have in the United States.
So it matters but even if we changed a lot of the regulations that we’re worried about, we still don’t think that would completely fix the problem. It might not even come close.
