Episode 76: How Housing Supply Responds to Rising Demand with Nathaniel Baum-Snow

Shane Phillips 0:05
Hello! This is the UCLA Housing Coice podcast, and I'm your host, Shane Phillips. Our guest this time is Nathaniel Baum-Snow, and we're gonna get a little wonky. Like, "microgeography of housing supply elasticity" kind of wonky. But, actually, as many times as I have read about and had supply elasticities explained to me, this might be the first time it ever felt at least somewhat intuitive. And despite the econ textbook terminology, there are plenty of more straightforward yet still very interesting ideas in this conversation.

Shane Phillips 0:41
One is that normally, when we talk about housing supply elasticity, or the responsiveness of housing supply to increasing prices, we're just talking about building new homes. If prices go up 1%, does the housing supply also go up by 1%? Or by half a percent or 2%? That actually misses, or groups together, a bunch of other important sources of housing, like the distinction between redevelopment and housing built on undeveloped land, or between housing units and floor space. Then there's the role of additions, renovations, and even just reduced demolition, which together account for something like a quarter or a third of the floor space added across neighborhoods. I think Nate's work also strengthens the case, maybe somewhat indirectly, for opening up exclusionary, affluent neighborhoods to more diverse and affordable kinds of housing, and we could always use more evidence for that.

Shane Phillips 1:37
The Housing Voice podcast is a production of the UCLA Lewis Center for Regional Policy Studies, with production support from Claudia Bustamante and Gavin Carlson. As always, you can email me with show ideas, comments, or questions at shanephillips@ucla.edu.

Shane Phillips 1:53
Also, an announcement! UCLA has a new, full-time Master's in Real Estate Development program kicking off in Fall 2025, and applications will be open until January 15th. I'll share more about the program in future episodes, but if this sounds like it might be up your alley, you can learn more at luskin.ucla.edu/mred. That's M R E D. With that, let's get to our conversation with Nathaniel Baum-Snow.

Shane Phillips 2:31
Nathaniel Baum-Snow is a professor of economic analysis and policy at the University of Toronto, and he's with us to talk about his recent research on the microgeography of housing supply: the degree to which different neighborhoods respond to increasing demand with more housing, and also the different forms that housing takes -- whether new, renovated, or otherwise. Nathaniel, thanks for joining us, and welcome to the Housing Voice podcast.

Nathaniel Baum-Snow 2:56
Very happy to be here.

Shane Phillips 2:58
And my co-host today is Mike Manville. Hey Mike.

Mike Manville 3:00
Hey guys, great to be here.

Shane Phillips 3:02
So Nate, we start every episode by asking our guests for a tour of a city that they know well. Where do you want to show us around?

Nathaniel Baum-Snow 3:09
So I thought I'd talk a bit about the city of Toronto, where I live. This is a city that, obviously is in Canada, so it's not one of the cities that's included in the study that we're going to talk about later, about housing supply, but it's the city I live in, and it has a lot of similarities to a lot of US cities. But I figure, you know, a lot of the audience might not be as familiar with the history of Canadian cities and their structure, so I think it's interesting to think about. Like many US cities, kind of in the Midwest, Toronto was settled about between 100 and 150 years ago, with huge increase in population because of immigration, mostly from Europe. But at that time, unlike many of the US cities receiving similar amounts of immigration, the powers that be in Toronto were very much against multifamily housing, and thought of it as dirty, unsafe, and so they really limited -- they put a lot of effort to limit the construction of multifamily housing. And so despite all of this immigration to Toronto, you see, today, most of the neighborhoods near downtown, that was settled that time, are single family homes -- even in these very high demand locations. And you think about a city like Chicago, you know has a lot of multifamily housing -- maybe not high rises, but multi family housing -- in these urban neighborhoods. Whereas in Toronto, most of these neighborhoods are full of single family homes -- small single family homes, but single family homes nonetheless.

Shane Phillips 4:44
Sort of rowhouse-style, right in many places?

Nathaniel Baum-Snow 4:46
Yes, rowhouses and small, semi-detached houses, yeah. And in some sense that makes it a very livable city and a nice human scale city, but it's also makes it a very expensive city, because unlike, you know, Detroit and Cleveland and Chicago, to some extent even, which were settled with these large immigration waves at similar points in time, Toronto is still a city that has huge amounts of growing demand, a lot of it's still driven by immigration. Canada's immigration policy is very generous, and Canada's population growth rate is much higher than the US population growth rate, and Toronto is one of the main destinations for immigrants to Canada.

Shane Phillips 5:32
I think I read, not that long ago, something like 50% of immigrants end up in that metro area. Some very large share, anyway.

Nathaniel Baum-Snow 5:40
A very large fraction, yeah. And you actually have net out-migration of native born Canadians from Toronto because the housing costs are becoming so high.

Shane Phillips 5:49
I think the same is true of California, actually.

Nathaniel Baum-Snow 5:53
Yeah, so, there are a lot of similarities there with US cities. But one difference I've been looking at some of the data on this recently is that actually Toronto is kind of in transition into a multi family high rise city. The demand is so strong for living in Toronto that it makes sense for condo developers to be building these sort of high rise condos. And the planning regime is such that the city, even though it's resistant to this, gets overridden by the province giving the right to condo developers to tear down single family homes and build condos. At least in many it's difficult, but in many neighborhoods, they've been able to do this.

Shane Phillips 6:38
It's been restrictive for so long too, that it's sort of jumping from these single family or attached houses to 50 story condos and kind of nothing in between, exactly.

Nathaniel Baum-Snow 6:48
And so this is very disruptive. Like some neighborhoods are really feeling the disruption of this transition. And I think it's kind of interesting you don't see that happening. I think it really in any US cities that I can think of right now, even though maybe in a few of them, like San Francisco or Boston, for example, think the demand conditions would justify that sort of transition in some neighborhoods, it's just that the municipalities have so much control over redevelopment rights and zoning restrictions that They just don't allow it to happen. And in Toronto that, just because of the way the planning regime is set up, it is allowed to happen, or has been allowed to happen with this current provincial government, which is sort of conservative and more market oriented, and that means that Toronto is going through kind of a unique transition that we haven't seen very much of in many US cities, and maybe it's sort of a nice, potentially positive lesson for the possibilities. On the other hand, housing costs have been rising very quickly in Toronto, just like in some of these high demand US cities too, so we won't know for another decade or two kind of how successful allowing construction of these condos has been for kind of alleviating the supply crunch and alleviating the growth in housing costs. But it's an interesting case study, at least. So

Shane Phillips 8:17
Mike and I were in Toronto, I guess a year and a half ago for the ACSP conference, it was my first time there. I really liked it, and I brought this up before, but it's very interesting how Canada sort of under supplies apartments, whereas the US under supplies condos. The Turner center actually just published something on construction defect liability laws that was really interesting. And it compared the share of new multi family that is owner occupied in I think it might have been all of Canada versus California. And in Canada, it was about 38% was owner occupied. So basically condos. And in California over a 10 year period, I think it was like 2011 2021 it was 3% so we are just not capable really of building that kind of housing in our cities, here in this state right now.

Nathaniel Baum-Snow 9:09
Yeah. I mean, Toronto does have this. I view it as a problem. And some people would disagree with me that there are a lot of very strong tenant protections, which disincentivizes the construction of purpose built rental housing, right? And I think that's resulted in a lot of these multi family housing construction projects being condos rather than rentals. Interesting, and this is, I mean, this is a real policy debate in the policy question that maybe topic of another day,

Shane Phillips 9:41
several podcasts, yes,

Nathaniel Baum-Snow 9:43
but that is a challenge here for sure. All

Shane Phillips 9:47
right. Well, we will stick to today's topic today and talk about an article by Nate and his co author, Lu Han, published this year in the Journal of Political Economy, titled The micro. Geography of housing supply. Again, this is about how responsive the supply of housing is to increasing demand for housing, or the supply elasticity. So just for example, if we were measuring housing demand by prices, if prices go up by 1% does the supply also increase by 1% or half a percent or less, studies of supply elasticity are not uncommon, but a few things set this article apart from most others. One is that it estimates supply elasticities at the neighborhood level across 50,000 census tracts in 300 metropolitan areas across the country. This is the US. The other is its detailed examination of the different contributors to housing supply, which include new production on undeveloped land or previously developed land, increased renovations or additions and fewer tear downs. In economics parlance, they've decomposed the supply response into its constituent parts. The paper also looks at changes in floor space, which might not and frequently does not, increase at the same rate as the number of units. Looking at supply responses at the neighborhood level allows them to estimate the contribution of things like the amount of undeveloped and flat versus hilly land and the restrictiveness of local land use regulations, and to better understand how these various geographic and regulatory factors constrain city's ability to meet residents' housing needs. We're also going to talk about some potential applications of this research, including an illustrative example that looks at the impact of the opportunity Zone program here in the US so Nate, we have talked about supply elasticities on this show before, but I don't think our listeners, and perhaps not I either would mind a refresher on them, what they are, give us a bit of the context and motivation here. How should we conceptualize housing supply elasticities? Why should we care about them, and why in particular, should we care about them at the neighborhood level, rather than only at the city or metro area level, where they are more often measured?

Nathaniel Baum-Snow 12:08
Great. So at a high level, what we want to think about when we think about supply elasticities, is how costly it is to build additional units of housing. And so if there is increasing demand for housing, how that gets manifested in prices versus quantities? So if supply elasticities are high, that means that the cost of building additional units of housing, quantities of housing is very similar to the cost of building prior units of housing. And so that means that any demand increase gets manifested mostly as quantities and prices don't go up as much, very much, whereas, if supply is very inelastic, when demand increases, we would tend to see not many more units of housing built instead, prices would be bid up a lot for the existing units of housing. And so supply elasticities matter because it tells us how cities or neighborhoods accommodate demand growth. Do they accommodate demand growth by growing themselves in terms of the quantities of housing units and population, or do they accommodate demand growth just by getting more expensive,

Shane Phillips 13:23
so that people cannot afford to keep moving there essentially, exactly

Nathaniel Baum-Snow 13:26
and so are we stuck in a situation where, when there's been a lot of discussion about this for cities like San Francisco and Boston, where The number of housing units is very hard to increase. And so given that the economies of these cities, well, especially Boston these days, San Francisco has had a bit of a decline recently with work from home, but I think Boston is still there. The demand for living in Boston is growing because of the industrial mix in the city. If it's really, really costly to build additional units of housing, mostly, that's going to result in just existing housing becoming more expensive, and that we can kind of summarize that phenomenon by saying the supply elasticity is very low.

Shane Phillips 14:10
Can I kind of tease that out a little bit? Because I think that's I'm not sure I've actually heard it defined that way before, but it does make a lot of sense. It's sort of at some level, this gap between how much it costs to build a new home and how much an older home costs is going to decide whether it makes sense to invest in a new home to build more homes, versus just for people to bid up the older stuff. Because if a new home costs $500,000 to build, but you can buy an old one for $300,000 then no developer is going to build a new home for $500,000 because people are not willing to pay that. Absolutely, homes are $300,000 and even if you can get a premium on that new home, it's probably not going to be $200,000 extra. And so prices are going to have to come up to some extent before it will make sense to build the home. At that very high price,

Nathaniel Baum-Snow 15:01
absolutely so actually, in your example, you think about a city like Cleveland or Detroit, where in many neighborhoods in these cities, the typical home price is very, very low and actually well below the cost of constructing a new home. In those neighborhoods, supply is going to be very inelastic for exactly the reason you say, right, yeah, because even if demand grows a little bit to live in that neighborhood, the price of a home is still below what it would cost any developer to build a new home. And so all of that demand growth is going to be manifested as price increase off of this very low base. I think that's sort of natural and less of a policy concern than sort of the Boston or San Francisco case, where prices are already quite high, and there's also strong demand growth for these cities that ends up being manifested just as prices getting even higher. Yeah,

Shane Phillips 15:56
yeah. That distinction is definitely important. Yeah. I

Mike Manville 15:59
think you know, one way you could flesh that out, maybe, is to add like a third city into the mix, like Las Vegas. You have a Detroit or a Cleveland, where demand is just extremely low, and so even if the price goes up, its probably not going to push the city to the point where it profitable to build a lot of new housing. But then you have places like Las Vegas and Boston, which both have very strong demand, where it probably would be sort of above construction costs to the motivation to build would be there. And yet, those two cities take very divergent courses, absolutely, whereas Las Vegas has spent the last 30 years just really adding a lot of housing, Boston has spent 30 years essentially letting its housing become more expensive,

Nathaniel Baum-Snow 16:38
exactly. And a big part of that has to do with land availability. Sure, like in Las Vegas, you have large tracts of land that can be built on in the suburbs and the exurbs Boston, partly because of planning restrictions, but also some because of topological restrictions, makes it much harder to build in the places where people would want to live. I mean, obviously you can't build on the ocean, and that's restrict supply in the Boston area. I

Shane Phillips 17:08
think Boston from 150 years ago might disagree with that proposition.

Nathaniel Baum-Snow 17:12
Well, they did fill in the Back Bay.

Mike Manville 17:14
The Back Bay is the great

Nathaniel Baum-Snow 17:16
exception, so the marshland maybe you can build on, but once you get

Mike Manville 17:20
out. But it's, it's nothing like just endless desert stretching perfectly flat. Lot harder to create new

Shane Phillips 17:27
land out of the out of the bay than to just use the stuff that's already there. All right? Well, I didn't give you the chance to answer the last of those questions. Which is, which is why we should care about the neighborhood level, supply elasticity in particular,

Nathaniel Baum-Snow 17:41
right? So many economic development policies are neighborhood targeted. And from a policy perspective, it's really important to know whether, say, government investment in a neighborhood is going to result in more housing units at the same price, or no more housing units but higher rents. Because in the first case, residents and potential new residents who move there and fill those new housing units are going to benefit. In the second case, landlords are going to benefit from this policy. And you know, I think it's not for me to say whether we should care about residents or landlords, sometimes with the same people, but probably most voters would care more about the residents than the landlords, and so sort of the incidence of local economic development policies are going to depend crucially on the local supply elasticities.

Mike Manville 18:31
Just to add to that, I think most, most economic development proposals, when they are floated, are not usually framed as, let's help the landlords.

Shane Phillips 18:39
That's a good sign of where people's interests lie, that's for sure. So at a high level, then how do you measure supply elasticity? As you point out things like local construction costs and land use regulations, parcel size, land availability, this is all discussed in the paper. Can all affect the type of housing that gets built and how much of it gets built. So the same level of demand can produce different amounts of supply depending on those other factors. So given that all of these things play a role, what are we really measuring when we measure supply elasticity? Is it the combined influence of all of these things together, or are we trying to isolate something more specific.

Nathaniel Baum-Snow 19:21
So first, I want to be very clear there are different margins of supply elasticity. So we want to be clear about which supply elasticity we're talking about. You can talk about a units supply elasticity, a housing unit supply elasticity, which is if developers see the house price go up by 1% how many more housing units do they build? In percentage terms, you can talk about a floor space supply elasticity, which says, if developers see the price per square foot go up by 1% how many more square feet do they build? Old. Those are related, but not the same thing. And actually, from a policy perspective, I think a lot of policymakers probably care more about the housing unit supply elasticity more than the forest based supply elasticity, and probably we'll get into this later, but they actually do look quite different, and maybe we don't care as much about developers building bigger McMansions when prices go up, we probably care more about them building more housing units that can actually house more families. So either way, it's a difficult empirical challenge to measure these supply elasticities, because what you need is different settings in which demand conditions vary, but in a way that's not related to cost of construction,

Shane Phillips 20:47
and that's just as a means of isolating specifically the role of demand

Nathaniel Baum-Snow 20:52
exactly so what we want is two neighborhoods, say, as a thought experiment, two neighborhoods that are identical ex ante, but experience different demand shocks, and that generates different price increases for these two neighborhoods that then we can compare and see how much extra quantity is built in these two neighborhoods that's associated with these different price increases and

Shane Phillips 21:15
just so people, so everyone's on the same page. An example of a demand shock might be something like Amazon announces it's going to open a headquarters in one of these cities and not the other. They're gonna have a lot of new workers, potentially residents, higher incomes, all that kind of stuff,

Nathaniel Baum-Snow 21:30
exactly, so something that's not related to the cost of construction. So in this study, the way we do this is we compare neighborhoods in the same metro area, but one of which is exposed to improving labor market conditions in commuting destinations relative to another. So you think about two neighborhoods that are at the same distance from, say, the Center of San Francisco in 2000 and then you had the tech boom that primarily happened in Silicon Valley. The one neighborhood that's closer to Silicon Valley is going to have a bigger demand increase than the neighborhood that's farther from Silicon Valley, and we see how many then we can trace out how much the price increased in that neighborhood as a result of this demand shock and how many more units of housing were built. And so in a city like San Francisco, what you see is that, essentially, the price increased a lot more in the neighborhoods that are closer to Silicon Valley, but not many more units of housing were built, and that kind of gets mapped into a low supply elasticity.

Shane Phillips 22:40
And can you talk about how you actually measured the housing unit growth, renovations, floor space? This was mostly Zillow data, it seems like, but the kind of Zillow data we are not allowed to access without special permission. Okay,

Nathaniel Baum-Snow 22:53
so I want to be very clear what Zillow has done. They have collated as much of the assessment, the housing assessment data that they could get their hands on from all the local jurisdictions, all the the assessment offices across the country. And these are property tax assessments. They're public data. Okay, all that Zillow has done is it's taken this public data from all these just different jurisdictions and harmonized it so that it can be put into one database, sort of together. And so the underlying data has millions and millions of housing units from across the US, starting in 2000 running up to 2010. Is what we used for our study. But it goes up into the present day. And you know, if you go and look at your own House's assessment information, the assessor knows records the square footage, the number of rooms, the number of bedrooms, the lot size, all sorts of attributes like this that they use to determine basically, how much your property is worth, and how much you owe in property tax. So that's the underlying information Zillow put together into their Z tracks data set, which we used for this study. And they were very generous in making it available to researchers for a while, and now they have decided that they don't want to do that anymore, so you can now go buy it from,

Shane Phillips 24:23
yeah, they probably just realized how much it was worth and how much people like us really value it. I

Nathaniel Baum-Snow 24:28
really, I really do appreciate that Zillow did this for a while, at least. Now you can buy it from another company called CoreLogic, which is kind of the most common source for this information. But I want to be clear that this is the underlying data. Is fully public. It's just what Zillow did was they harmonized it,

Shane Phillips 24:47
yeah, yeah, which is a very challenging thing to do, even across, let's say, all the counties in California, which we're doing some work in that general area right now. But to do it for the whole country is a whole other thing that's for sure.

Nathaniel Baum-Snow 24:59
Sure that's for sure.

Shane Phillips 25:00
Well, let's drill down on these different measures, or these supply elasticities. So we've talked just generally about housing units and floor space. But can you say a little more about how those are broken out? I guess, specifically the new units, because you're looking at undeveloped land versus undeveloped land, I'm curious what exactly that means. But then also renovations and additions and basically not tearing down housing is sort of a supply response tearing down older buildings. So can you just give us the full picture on all that?

Nathaniel Baum-Snow 25:32
Yeah. So one way to think about this is, if you look at a given neighborhood over time, you can measure the change and the total amount of floor space in that neighborhood. Now, that change can come for a number of reasons. One reason is new housing units are built and they have some floor space with them. Another reason is that existing housing units are expanded. So maybe you add an addition onto the back of a house or that sort of thing. Another reason is you can have some tear downs. So existing homes are torn down. Maybe some of those tear downs are replaced with new homes that are bigger. But you can, you can separate out that tear down margin. So for from the perspective of floor space, we can look at all those things and for new the new units part of it, they can either be redevelopment, or they can be on land that was not developed before. So this is sort of urban expansion. And so in this study, we look at all of those margins of response, both for floor space and for housing units, just to see when prices go up, where the housing supply response comes from. And as you can imagine, most of it is indeed the construction of new units. But importantly, something like a third is from the fact that existing units are either subdivided or changed, reconfigured in some way, and that as prices go up, you have fewer tear downs, that when prices go up, maybe you have some really old house rather than letting it decay and so it's not livable anymore, you refurbish it and keep it in the housing stock. Okay, that's an important margin of response that I think had not been widely recognized before. The other thing that we find, that I think is pretty interesting and important, is that we find very little evidence that redevelopment responds very much to prices. I think a lot of this comes from the fact that it's just like the zoning regulations and planning regimes are so strict in most jurisdictions in the US that redevelopment is just incredibly difficult and sort of idiosyncratic, right? If you want to tear down a single family house and build multi family housing there, the number of steps you need to go through in most jurisdictions to get approval for that makes it really hard, and sometimes it happens, but it's kind of idiosyncratic. You got to have the right connections on the planning board, and you have to know, have the right connections to the city council member, and you have to have the right support from the neighborhood organization, and all these things just makes it not very closely related to crisis, whether this can happen. And so most of the supply response that we see to price increases in terms of new units, comes on prior undeveloped land. That

Mike Manville 28:31
point you make about, you know, the teardowns becoming less common as prices rise. That seems like one of those things where it might have a slightly different flavor once you get into the Palo altos and San Francisco's of the world, right? Absolutely. And then if the price goes up in Detroit, yeah, you were, you were going to just let this house crumble into a vacant lot. And now you don't, whereas in Palo Alto because you can't really redevelop you've got this little cottage, but what you can do is sell it to someone who knocks it over and builds a huge house. Exactly, yeah, yeah. So the number

Nathaniel Baum-Snow 29:03
of units doesn't change, but the floor space changes.

Mike Manville 29:05
Floor space changes, and the the average price of a unit goes up because now you've got this gigantic unit, as opposed to a smaller one, a starter,

Shane Phillips 29:13
$2 million home, yes, 70 years old.

Mike Manville 29:17
I mentioned Palo Alto, but the entire area north of me, in the mid City, Los Angeles. So that's the story there the Melrose district. It's single family home zoning, and what you have is a series of bungalows that if you talk to people in that neighborhood, they're like, We must protect these bungalows from evil apartments. But what people do when they buy those bungalows is they knock them over and just build these gigantic boxes that fill up the whole lot that are, in essence the size of apartment buildings, but legally reserved for one family.

Nathaniel Baum-Snow 29:45
Yeah, yeah. And there's a huge potential there for densification that has not been realized.

Mike Manville 29:51
Yeah. I mean, you really could with the stroke of a pen. I mean, it is the size of an apartment building right now. I mean, you could just say three families can live here. They'd more than fit. I

Shane Phillips 30:01
did want to come back and just sort of flag, because this has been interesting to me for a while, this difference between floor space and housing units, I think part of it is just that it's a lot harder for average people to measure floor space. It's not something that's reported in the, you know, in the census, in the American Community Survey, that kind of thing, you do have to usually go to the assessor data to get that information, whereas it's very easy to track through a bunch of different channels, how many homes are permitted, how many housing starts there are, how many are added to the housing stock, and so forth. But really, you know, like if Los Angeles is primarily building 700 foot one bedroom and two bedroom apartments, and Houston is primarily building 2500 square foot town homes. The same number of units in one city versus the other is not probably housing the same number of people. It's not actually accomplishing exactly the same thing. And that's not a knock on the 700 square foot units. We need those. But it's just we don't really have that detail, that resolution that would give us a better picture of how many people can this housing stock really accommodate in one city versus the other. So I do appreciate having those two things broken out here, and I did just want to point out that so for the changes from 2000 to 2010 you found that the housing stock grew by about 7% overall nationwide, in the census tracts that you're looking at over that 10 year period, but floor space increased by 14% so about double, right? And so yeah, is that mostly McMansions in the suburbs? Is it tear downs of smaller, single family homes replaced with bigger ones? Is it those and a whole bunch of other things? What's going on there?

Nathaniel Baum-Snow 31:40
It's mostly the fact that the new homes that were built during this period were, on average, much larger than the existing homes got it. And you think this period most of the construction happened between 2000 2007 before the housing crisis, right? Kind of stopped things. And that was a period when you had, you know, you think about like the Inland Empire, right? A lot of these big homes built in the suburbs, in the sort of exurbs that ended up being bought by people who couldn't really afford it because of the cheap credit. And there is going to be this natural composition effect that, since new homes tend to be built on undeveloped land, and undeveloped land tends to be at the fringes of metro areas. It's going to be the case just naturally that new homes are bigger than existing homes, right? It makes sense for homes at the edges of metro areas to be big relative to those in the center, because that's where land is cheap. Where land is cheap, you're gonna use more land and less vertical structure in the homes you build. So that's sort of natural, but I think there was this additional kind of cheap credit environment during this period that accentuated that natural force.

Shane Phillips 32:53
So we've been talking about measuring housing supply elasticities, but we have not actually given any housing supply elasticities. So let's start off with the results just for all of the census tracts grouped together across all these different metro areas. And then maybe we can talk about the kinds of variation we see between neighborhoods within metro areas after that. But this big picture national what is the breakdown for these various elasticities?

Nathaniel Baum-Snow 33:21
So it depends, sort of the details of how you do it. But to give kind of a few numbers, if you look nationwide at all the neighborhoods, on average, you get a housing units supply elasticity of about point three, which is to say that if prices were to increase by 10% you would have about 3% more units built, and floor space supply elasticities, which are about, on average, about point five, which is to say that if the price per square foot increases by 10% you get about 5% more floor space built. Okay, so this is once again, goes back to the point that you brought up earlier Shane about the fact that new housing tends to be bigger than existing housing, and the fact that you know the floor space responses and the size responses tend to be bigger than the units responses in terms of supply elasticity. Some of this is natural, but some of this comes from planning regulations that have sort of minimum lot size. Zoning is really pervasive across the country. So even in undeveloped areas, you know, maybe developers would like to build more density, but they're not allowed to because of the way the land is planted out. And so instead, they build bigger houses, yeah, and these this higher floor space supply elasticity than unit supply elasticity that we find reflects that. So these numbers that I just gave you are averages across all neighborhoods nationwide, and there is a lot of variation, both within and between metro areas. So as a general rule, these numbers tend to be very close to zero right near the centers of metro areas. These are in the central business districts. And around the central business districts, the land is already very heavily developed. And as I said before, it's very we find like it's very difficult to take land that's already developed and redevelop it to something denser. And so then it's natural that kind of the overall supply elastices will be very close to zero in these types of neighborhoods that are already fully filled in. And then as you move out to the urban fringes, you get to supply elasticity estimates that are more like so for units more like point 5.6 and for floor space, more like point 7.8 relative to the averages of point three for units and point five for floor space. So those are kind of, kind of some of the patterns. And the other pattern we find looking across metro areas is that, as you might imagine, the places where there's more regulation. Where it's more difficult to build, you tend to find lower supply elasticities, but interestingly, you also tend to find for the units, the new units that are built, they tend to be bigger, which goes back to our conversation before you know about these single family home neighborhoods that are replacing bungalows with large houses that are still single units but have a lot more floor space. And then the metro areas that have more land available for development tend to have higher supply elasticities as well, which is a result that's both of these results have been sort of in the literature with Albert sizes work from 2010

Shane Phillips 36:37
this is going to dumb things down more than is probably appropriate. But holding incomes constant, is it fair to say that if your supply elasticity is below a one, housing is probably getting more expensive relative to income over time, whether at the square foot level or at the unit level. Right?

Nathaniel Baum-Snow 36:58
So if you're holding incomes constant, then it's got to be that demand growth comes from population growth. And so then, yes, absolutely. So if you're getting population growth, but it's becoming more and more expensive to build these new housing units, and so prices get bid up at a high rate because of the inelastic supply, then indeed, it be the case that the share of expenditure on housing in household budgets has got to be going up. Yeah. So yes, that's exactly right.

Shane Phillips 37:28
But because incomes are not constant, it is possible that if you had a supply elasticity under one, housing could still be getting more affordable relative to incomes.

Nathaniel Baum-Snow 37:39
It could be but could happen. It could happen. So So, I mean, I think it does fit into this whole phenomenon of increasing income inequality. So for high income people, I think this has been happening because their income has been increasing relatively quickly, but below the median incomes have been falling, actually, over time. And so with any house price increase, housing has become an increasingly high share of the budget. And actually, you look at renters, I've been looking at this data recently. In this time period, 2000 to 2010, the fraction of income spent on housing by renters in particular, went up a lot. It's increased a little bit recent in the past, like few years as well, but it increased a lot between 2020 10, from like 25% to 30% something like that, on average certain age.

Shane Phillips 38:30
I wonder if that maybe has something to do with the fact that renters are more concentrated in more urban areas, and those are the areas that did not build much housing, even during the boom in the 2000s

Nathaniel Baum-Snow 38:41
Yeah, absolutely. And there is some composition effect that during this period, you know, you had some movement into homeownership as sort of the higher income renters were able to get access to credit that they didn't have access to before and became owners.

Shane Phillips 38:55
Yeah, kind of skimming off the top of the renter group by income, exactly. Yeah. Well, let's move on to the neighborhood level. Obviously, we cannot talk about all of the 10s of 1000s of neighborhoods that you estimated elasticities for individually, but in the article, you spend a good amount of time discussing how the elasticities change as you move away from the central business district of metro areas. And I thought that was really interesting. So maybe we can trace out that same path. Could you walk us through how supply elasticities change as you move from the center of a metro area to its edges? Sure,

Nathaniel Baum-Snow 39:30
absolutely so. As I said, in the downtown areas, there isn't a lot of land available for development, and that tends to mean that you have very, very low supply elasticities. They're near zero in most metro areas, right in downtown regions, and this

Shane Phillips 39:47
is despite them generally having less strict land use regulations. It's just the lack yes of available undeveloped land sort of overwhelms the influence of laxer regulation. Exactly.

Nathaniel Baum-Snow 39:59
Exactly. And so the places where you tend to have more strict land use regulations is kind of the inner ring suburbs. These are the neighborhoods and the areas and a lot of metro areas you can kind of think about. They tend to have a lot of single family homes and very strict minimum lot size zoning regulations. I think a lot of the motivation historically has been exclusionary, honestly, to keep lower income people out of these inner ring suburban neighborhoods by making it too expensive for them to afford homes there. Yeah, so that regulation force pushes in one direction and makes housing supply elasticities lower in these sort of inner ring suburbs and kind of the outer neighborhoods of central cities as well. But on the other hand, you tend to have more land available for development. So those two forces over some range of CBD distance, kind of offset each other to give you unit supply elasticities around point 3.4, and then once you get out into the exurbs, there's fewer planning regulations and a lot of open land. And then you see supply of elasticities start to go up again, into kind of the fringes of metro areas. And so then it's not surprising that most of the new construction we see tends to be on undeveloped land, you know, in these fringe areas, I

Mike Manville 41:19
mean, and I think that's an area where, I think maybe for, again, for different reasons, but that's just gonna be constant across lots of central cities that are varying economic fortunes. You know, it's just extraordinarily difficult to build in downtown San Francisco, both because there's so little vacant land and there's, it's a very stringently regulated area. But then you think about Detroit, where there is actually a lot of vacant land. But you know, to our earlier point, the price goes up, and that's just gonna it's gonna prevent some tear downs, maybe. And then to Nate's point, right next to Detroit, you've got, like, gross point, the Uber exclusionary suburb who's like, whole existence is designed to keep Detroit from coming in, and they're not gonna build. But so what you do see happen in Michigan is the development just jumps over that first ring, goes into that farmland, and then you add a lot of housing there, when the Detroit region's economy improves and it ends up looking a lot like what happens in the Bay Area, as development jumps all the way out into the interior, even though those two central cities couldn't be more different In many ways, absolutely.

Nathaniel Baum-Snow 42:20
And so you have this story of many central cities in the US that you know, like Detroit, the demand to live there has just created so much that it doesn't make sense to build anything new, right? And I mean, I think actually, Detroit's been relatively successful in the past few years of sort of starting to come back from that, but they're starting at a very low base, and I think they're still at prices that are well below construction costs. Yeah.

Shane Phillips 42:47
Well, it's interesting seeing I was there a year or two ago just for like half a day passing through, and walked around, mostly in the downtown, and kind of immediately around there. And they're actually building quite a bit in the downtown, but I think it's really just there, and maybe a few kind of close in neighborhoods, but the city sprawls out much further than that, and there's just, you know, 1000s, 10s of 1000s of vacant parcels of homes that used to exist and have since been torn down and are just, you know, waiting for something to happen. That's right, I did want to follow up on this, though. So I think one thing that stood out to me about this trend that you saw as you go from the CBD Central Business District out to the edges of the metro area is that all of the supply elasticities sort of showed the same trend of increasing as you went out. And so is there anything to say about why that would be the case. I mean, even, like with tear downs and stuff, it was true even there, that as you go out, yeah, that that response is more likely.

Nathaniel Baum-Snow 43:48
So the main force for units and floor space is going to be the construction of new units, right? So I think, I think the logic there is natural for tear downs, that was not actually something that we thought about in a lot of detail. And I'm trying to remember, I think, for the suburbs, it's going to be very hard to find estimates there, just because the houses are new enough that there aren't many units that are subject to tear down. So most of the estimation of tear down elasticity is going to come kind of from the central cities where the housing units are at least, like 50 years old or something, so they'd be subject to tear down risk. Got it. So I think most of this is really just driven by just where you can build new units, and that's generating these patterns. But the other thing, which I think is important is looking between across metro areas, you have the planning regulation, the zoning regulations really mattering to affect the composition of the housing stock. And that's something that would that we saw, you know, we see kind of for every type of metro area, you get sort of bigger houses built near the. And smaller houses built near the center, and they tend to be built on undeveloped land. And that's all very kind of common across all types of places. But what the planning environment does is it accentuates the size of these common forces, right? So in very restrictive metros, basically, you still see that, but everything has just dampened a lot. And so those places like San Francisco, even in the suburbs, it's just very hard to build stuff, just because zoning regulations even in the suburbs are very strict, whereas in Houston and Dallas, it's very easy to build stuff. So you see the same pattern, but it's just with higher supply elasticities throughout. And that's kind of, I think, where this comes up, and it's true with tear downs too, because in San Francisco it's hard to build new stuff. You tend to see a little bit more of this tear down elasticity, because getting the right to tear down something and build a bigger house there is also hard. Yeah, I don't know. Maybe its maybe its less true in Los Angeles, I'm not sure. But I think in San Francisco you do see a lot of these old houses that have been maintained kind of for this reason, because the redevelopment cost is so high, just because of the conforming to the building codes and getting the building permits and all these things.

Mike Manville 46:12
Yeah, I think in Los Angeles, and we've actually looked at this a little bit, the demolitions that happen, but they really tend to happen to the old single family homes. And certainly there's plenty of people who say, Oh, I'm going to restore my old bungalow and maybe add an addition to it. It's very hard to tear down multi family housing in Los Angeles, both because of tenant protection laws, rent control laws and things like that. And just because it's like, you know, it's complicated to do in the middle of the city, take down an apartment building. You know, if you have another alternative you might want to take it. So I think what the difference between LA and San Francisco in that regard might just be the sheer amount of land that's that's detached single family homes.

Shane Phillips 46:49
There's a part of the paper Nate that I was hoping you could elaborate on a bit. I'm just going to read the quote here. You say quote, we find that each component of housing supply becomes more elastic, moving out from urban centers and that there is more variation within than between metro areas. Just want to underline that in housing supply elasticities, this pattern is in part, but not entirely, due to the increasing fraction of land available for development with CBD, distance, initial development intensity, availability of flat land and zoning regimes are all important determinants of local housing supply. That's sort of an addendum or an addition to the fraction of land available for development. The quote is long since passed. I'm just talking now. The part about the increasing fraction of land available for development is what we were just talking about. But how about those other things that you listed? I emphasize there being more variation within than between metro areas, because it does seem relevant here. My reading is that, broadly speaking, all metro areas have an increasing fraction of land available for development as you move away, just more undeveloped land as you get away from the center. And that increasing fraction is very important for predicting housing supply elasticity, maybe, in some ways the most important thing, absolutely, yep. But despite that shared characteristic across these metro areas, each Metro can still vary a lot from the others in terms of its topography, the presence of lakes and rivers and oceans, how restrictive the land use regulations are in its various jurisdictions, and where those places are distributed, with more and less restrictive regulations, how intensively developed different areas already are, and so on. So is that a reasonable takeaway and interpretation of this? And if so, or if not, what are the implications of that greater variation within metro areas compared to between them, whether we're talking about just for research purposes or for policy or whatever.

Nathaniel Baum-Snow 48:44
As you say, Yeah, you're absolutely right. There's kind of a common set of features about neighborhoods near central business districts relative to neighborhoods far from central business districts. Those near to central business districts tend to have very little land available for development, and then they tend to allow for tall buildings to be built, but already so heavily developed that its unlikely they end up to getting built, at least for residential purposes. And then you hit the inner suburbs, where you have a lot of regulation, and that's very common across metro areas. A little bit more land available for development. Then you get to the edges, where there's lots of land available. Of land available for development and low regulation. So that's why it's common to see more building there. So everything you say is totally fair. It's totally right. And so what's the implication? I think, coming back to what we talked about the very beginning, when we have policies that are targeted towards neighborhoods near to central business districts, which tend to be the poorer neighborhoods. It's important to recognize they tend to have low supply elasticities these neighborhoods, and so policies that increase the demand for living in these neighborhoods, given the low supply elasticities are. Gonna result in higher rents and housing prices in these neighborhoods, to a much greater extent than the same policies if they were implemented in the suburbs.

Shane Phillips 50:08
What kind of policies are we thinking of that would increase demand in a specific neighborhood or set of neighborhoods,

Nathaniel Baum-Snow 50:14
you could have local economic development policies like investing in improving the quality of the streetscape, the parks, the business environment, giving businesses tax credits for locating there, which is part of what the opportunity Zone program we talked about in the paper does anything that is going to make these neighborhoods more desirable.

Shane Phillips 50:39
And in theory, if you were to make it easier to build on developed land, the supply elasticity could go up, and so you might absolutely solve this problem. It's not inherent to these neighborhoods, although the fact that they are generally lower income does mean that probably the rents and prices of new homes are going to be a little lower than they might be elsewhere. All else equal. And so maybe even if you've kind of solved the regulation problem, it still might not be the top place that developers want to build. It might have a lower supply elasticity than if you made the same reforms or changed the regulations in the same way in an inner ring suburb, for example. Yeah,

Nathaniel Baum-Snow 51:20
there's kind of a threshold that you got to get over first to get up to, kind of like, forget the land cost, just the cost of building materials and labor for building new stuff. And if you can get demand high enough over that threshold, then we can start talking about making the zoning and planning regulations loose enough so that densification can occur, and then, you know, rents won't go up as much as they otherwise would have, even though they will still go up. And you know, you're going to have a lot of incumbent residents. Maybe, if you're able to do a big push like that, that will not be happy they chose to live there because rents were low, even if amenities were not that high. You know, it tends to be low income people. They care. Money is really, really valuable to them, right? And so they're willing to give up more on amenities to achieve low rents than higher income people are. And so in these sorts of local economic development policies, we've got to be careful. We want to be careful about sort of doing things that actually maybe make neighborhoods nicer, but also make it so that they're not appealing to the incumbent residents who are living there anymore. They're appealing to outsiders, and that actually could make the incumbent residents worse off. It

Mike Manville 52:30
seems like this part of the paper, and you know, when you talk about the opportunity zones, its putting some newly precise numbers on the classic sort of case against place based economic development policy, which is this, this fear that any improvement you make on a place based level is capitalized into the land value. And that's not ostensibly the target of the policy, you know, and it really one of the things that was funny to me about the opportunity zone when it was introduced, is that it seemed to harken back to an older era of economic development that we associated more with, I think, like the 1990s and 1980s of like, we're going to lure businesses into these neighborhoods, we're going to pump investment into them. And it just, it raised the same set of concerns, which is just, well, you know, if you've got a bunch of incumbent residents who aren't necessarily going to get these jobs, what are you doing besides raising their rents? And I think that's always been a more of a rhetorical point, right? Just sort of like, well, this is what's going to happen. The rents are going to go up. And I think what Nate's paper has done is showed that there's actually real reason, empirical reason, to think that is what happens when you, you know, open up the doors to an inner city, Chicago neighborhood or something, with a bunch of tax credits, you know, you in the event that it increases investment at all, which it may not, its going to just make rents go up,

Nathaniel Baum-Snow 53:45
right? And the other thing I think is important to recognize, and this is not, this is kind of beyond the scope of this paper, but I think is important is that new housing tends to be built for high income people, sure, because new housing is relatively high quality housing. So if you have a policy that incentivizes new housing construction and it has no restrictions associated with it, it's going to bring in high income people. That's not necessarily a bad thing. It could be. It's probably even a good thing.

Shane Phillips 54:13
But there are consequences. There are consequences, good and bad as always, yeah,

Nathaniel Baum-Snow 54:17
and but we also got to recognize that if you put restrictions on it, and you make it more expensive to do this because you require, you know, some income mixing and some inclusionary element to the new housing less is going to be built, and the total housing stop is going to increase by less, and somewhere down the line, that's going to mean that someone's housing costs are higher as a result, relative to a world in which you didn't constrain now, it still might make sense, but it's only going to make sense if we have put a very high social weight on mixed income communities. And you know that that's maybe a reasonable trade off, but we got to recognize that its trade off.

Shane Phillips 54:57
Yeah, yeah. Well, I think this is a good time to get. Into the opportunity zone illustration or case study here, could you tell us about what you looked at there, just as a sure kind of demonstration of the potential applications of this research?

Nathaniel Baum-Snow 55:10
Yeah, absolutely. So we looked at the census tracts that were identified as opportunity zones. This is part of the 2017 so called tax cuts and Jobs Act, the Trump tax cuts, which gave tax breaks to investors in these low income neighborhoods. And

Shane Phillips 55:34
this is specifically in real estate investors, correct? Or is it other kinds of investments as well? Yeah,

Nathaniel Baum-Snow 55:39
so I don't remember all of the details. There is definitely real estate was the most important element. I think there were some ways of benefiting from these tax breaks where you didn't have to there was going to be a real estate component. In any case, anytime you invest in something that's place based, part of the returns are going to have to do with real estate in one way or another. But I don't remember the exact details of got it, of what you needed to get these tax breaks. But certainly if you invested in real estate and building new real estate in these neighborhoods, you were going to get these tax breaks. So what we looked at is just trying to do some calculations of, well, let's look at the supply elasticities in these neighborhoods that are opportunity zone neighborhoods, they're low income neighborhoods in urban areas around the country, as compared to other low income neighborhoods that were not identified as opportunity zone neighborhoods. And what you tend to see is that the supply elasticities are lower in these opportunity zone neighborhoods. And so what that means is, if you give a tax break for new real estate development in these neighborhoods, mostly what it does is it doesn't change how much real estate development there is. It's just kind of a giveaway to the developers for whatever they would have done anyways. And so that, because of that, the consumer surplus, or the sort of value to residents who are not landlords in these neighborhoods is very low because it doesn't generate a lot of new housing, which means that rents are not going to be affected very much. And it's just really, as Shane said, sort of an illustration of the fact that if you have a targeted policy towards low supply elasticity neighborhoods, mostly that's going to get reflected in benefits for kind of the supply side, the landowners and the developers, rather than the demand side, the renters, given that these are low income mostly rental neighborhoods, Yeah, I

Shane Phillips 57:41
feel like we're going to have to do an episode on the opportunity Zone program itself at some point. It's very interesting because, as Mike said, it's, it's sort of hearkening back to a previous era of urban planning and housing policy. You could say, yeah, the 80s, 90s. You could even potentially say, like, urban renewal kind of stuff, when in the planning profession, in academia, I think the mainstream view is shifting, or already has shifted more toward this Affirmatively Furthering Fair Housing. We need to put more housing in higher resource neighborhoods, because we've been concentrating it in lower income and often communities of color for decades, and not really allowing it elsewhere. And so we've kind of been swinging back and forth as the presidential administrations change, where Obama had the 2015 AFFH rule, which was intended to kind of institutionalize more so that policy from the fair housing act of 1968 and then the Trump administration walked that back and also did this opportunity zone stuff. And then the Biden administration now is pushing AFH again, and we'll see what happens in November. So because we could move on from there, I just had one more question, hoping that there's a bright spot here. You saw this consistent pattern where declining land availability was associated with lower supply elasticities, which I interpret, and I kind of have heard you say, as us just not being very good at infill development, right? In theory, there is no reason we couldn't rebuild our urban cores bit by bit over time, taller and denser, but we don't do much of that in most places, even where the demand is very high. I don't think any of us finds that particularly surprising, but it is still a bummer. So I'm curious if you saw any exceptions in the data. Were there places that were able to build a substantial amount of housing in response to demand shocks, despite having relatively little undeveloped land, whether that's in or around a CBD or inner ring suburbs, places that were able to redevelop land efficiently? Are there any bright spots, or is it all all sadness for us?

Nathaniel Baum-Snow 59:44
So this actually comes back to the very beginning of a conversation, because I actually see Toronto as an example of a city that has been able to do this where. So you have this neighborhood in Toronto called Young and Eglinton, which was a neighborhood of single family. Homes that is now a neighborhood of condo towers, and the local residents have been up in arms forever, since I can remember about all the condo development in this neighborhood, but there's been a lot of new housing units built. It's also been very disruptive there. Thus they've had to carve out the school districts and kind of this swiss cheese sort of way to accommodate the new students in the schools. They've had to rebuild the water mains and the sewer lines. They're building a new subway line to serve this neighborhood that has been very disruptive as well. The traffic is has gotten really bad. You can see why municipalities don't want this to happen, because it is really disruptive, and its much easier to build dense neighborhoods from scratch than to densify existing residential neighborhoods.

Shane Phillips 1:00:58
This is sort of also the difference between saying we're going to pick a specific neighborhood where we're going to concentrate all of our growth, versus allowing it to be instead of 40 story condo towers just taking over a single neighborhood, maybe its five story or seven story apartment buildings allowed in many places and no one places, is sort of responsible for accommodating all that growth. I can also see the benefits of concentrating it, though, because, yeah, if you do have to upgrade your sewer lines and you want to build new transportation infrastructure, better to do it all in one place, right? And maybe you couldn't justify expanding the subway line if everywhere is just getting a little more housing rather than one place getting a whole lot

Nathaniel Baum-Snow 1:01:39
another example, which I think is partly a missed opportunity, is the seaport district in Boston, which, you know, was kind of a clean slate industrial district, and they built a kind of second tier underground bus line through it, called the Silver Line, which is not really a it's sort of a subway. It's better than a regular bus, but not as good as a subway.

Shane Phillips 1:02:05
It's got the sub part, yes.

Nathaniel Baum-Snow 1:02:08
And they, you know, allowed housing to be built there, but it ended up being Convention Center, a lot of hotel rooms, and housing that was not really high rise. Sort of mid rise housing. And so I'd say it was a moderate success story, but didn't reach the potential it could have, because there were still quite a few planning restrictions. I do think some neighborhoods in Miami have, you know, built a lot of housing relatively successfully, mostly for foreign investors, honestly. But you know, as long as they're being rented out to somebody, then I don't think there's anything wrong with that. I don't know the data on how much that's actually happening, but it's pretty rare. And, you know, I've looked at data on the so-called superstar cities in the US. These are cities that have really high prices and look like they've had a lot of demand growth over the past 20 years, and they have been building housing units at much lower rates than any other type of location. And their rate of housing unit growth is just about the same rate as population growth, whereas in every other type of location, the rate of housing unit growth is much higher than population growth rates. Which makes sense because household sizes are getting smaller over time, and so part of the housing affordability challenge--

Shane Phillips 1:03:27
But they're not really getting smaller in places like Los Angeles, from what I've seen.

Nathaniel Baum-Snow 1:03:31
Exactly.

Shane Phillips 1:03:32
Which is probably evidence that people would like to have smaller household sizes if they could, because they are everywhere else, but it's just not affordable for too many people here.

Nathaniel Baum-Snow 1:03:41
Yeah, so I think one of the sources of the affordability challenge is just that you have a lot more single people than you used to, and they each want to have their own housing unit, right? So you have this kind of mismatch between the attributes of the housing stock, which is a lot of single-family homes, and the attributes of kind of the demand conditions, which are a lot of single people who probably should be happy to live in multi-unit housing, but there just isn't very much of it available, and developers find it very hard to build because of zoning regulations.

Shane Phillips 1:04:11
True also for -- I mean, this includes single people, but -- older adults who are alone or living with their partner, who are still in a three or four bedroom home and maybe don't have anywhere that they could move to downsize and stay in the same community

Nathaniel Baum-Snow 1:04:27
Right, absolutely.

Shane Phillips 1:04:28
All right, Nathaniel Baum-Snow, thank you so much for joining us on the Housing Voice podcast. It was great to have you.

Nathaniel Baum-Snow 1:04:34
It's been my pleasure. Thanks a lot.

Shane Phillips 1:04:40
You can read more about Nate's work on our website, lewis.ucla.edu. Show notes and a transcript of the interview are there too. The UCLA Lewis Center is on various social media, I'm on Twitter at @shanedphillips, and Mike is there at @MichaelManvill6. Thanks for listening, we'll see you next time.

Transcribed by https://otter.ai