Home / UCLA Housing Voice Podcast / Episode 30: Skyscrapers with Gabriel Ahlfeldt

Episode 30: Skyscrapers with Gabriel Ahlfeldt

Ahlfeldt and Barr, "The economics of skyscrapers: A synthesis"

More

Episode Summary: Skyscrapers! We can’t help but find them fascinating. Some cities are full of skyscrapers, and others have none. Developers built a 70-story tower on that parcel, but the proposed building just down the street is only 30 stories. How do developers decide where to build skyscrapers and how tall they should be? And are they really a profitable investment, or simply a monument to individual power and ego? Gabriel Ahlfeldt joins us from the London School of Economics to talk about his research on skyscrapers, a comprehensive analysis that catalogs nearly every 150-meter-plus building in the world. We discuss how skyscrapers influence the built form of cities, far beyond their typical boundaries within the central business district, and what the data can tell us about their profitability, their appeal to residents and workers, and the role that planners play in shaping where they’re found and how tall they go. Skyscrapers!

  • “We bring together recent research on costs and benefits of building height, which we integrate into a general equilibrium model of the vertical and horizontal structure of an open city. We use the model to engage with the following questions concerning the positive and normative economics of skyscrapers: How do vertical costs and benefits shape the internal structure of cities? How does the distribution of economic activity between cities and rural hinterlands affect the vertical size of cities and vice versa? What are the consequences of distortionary policies that constrain the vertical size of cities on the spatial distribution of economic activity and welfare? … From our theoretical and empirical analyses, as well as our reading of the literature, we conclude that vertical costs and benefits, horizontal land use patterns, urban growth, and welfare are all mutually dependent in ways that leave plenty of room for future research.”

 

  • “[W]e have developed a new general equilibrium model of vertical and horizontal city structure to study the causes and effects of vertical growth. To engage with our main research questions, we design the model to serve three purposes. First, the model should rationalize the remarkable differences in building heights within cities that are characteristic of metropolitan skylines and account for how use-specific vertical costs and benefits affect the horizontal land use pattern. Second, the model should address the mechanisms that drive the close relationship between urbanization, vertical growth, and rising incomes between cities, which we document in our data. Third, the model should be suitable for the evaluation of a common form of land use regulation that constrains vertical space: building height limits. Here, we are interested in the effects that this distortionary policy has on the spatial allocation of economic activity within and between cities.”

 

  • “The degree of urban bias in the distribution of skyscrapers is striking. Small states in which the largest city dominates the city system, such as Panama or the United Arab Emirates, reach the highest penetration in per capita terms. 50% of the world’s skyscrapers are located in just 17 cities, though a total of 315 cities worldwide have at least one. In terms of 150-meter or taller skyscrapers, Hong Kong leads the list with 353, with New York City coming in second at 281. Of the top twenty cities around the world, nine are in China (including Hong Kong). 16 out of 20 are in east Asia. Only three cities in North America (New York, Chicago, and Toronto) make the list.”

 

  • “While the spatiotemporal diffusion of skyscrapers appears to reflect the interplay of demand and supply conditions, there is nevertheless a common belief that the height of the tallest buildings is driven by non-pecuniary motivations. As an example, the “height race” in New York City during the 1920s led to the perception that developers of tall buildings, in aiming to dominate the skyline, seek to satisfy their egos rather than maximize profits (Barr, 2016). Indeed, Helsley and Strange (2008) offer a game-theoretic model that rationalizes how developers vying to claim the prize of the “tallest building” build “too tall” to preempt rivals … Hence, it seems likely that economic fundamentals drive the heights of nearly all but a handful of the tallest buildings in a limited number of cities. Consistent with this interpretation, we find that only few buildings managed to hold the crown of being the tallest for a very long time. While the Empire State Building ruled height rankings for 40 years, the median duration is six years at the global and four years at the national scale (see Online Appendix Section B.4). This relatively rapid succession is consistent with a steady rise of the profit-maximizing building height, due to increasing demand and improving construction technology.”

 

  • “Closely related to the presumption that tall buildings serve non-pecuniary motives is the notion that they are not particularly profitable … To move beyond anecdotes that mainly refer to high vacancy in the ESB during 1930s, we collected archival records and recent annual reports up until 2009, when a half billion dollar renovation marked the beginning of a new life for the ESB. Fig. 5 presents the key results of an ex-post financial case study whose details we report in Online Appendix Section B.4. Indeed, the ESB took a greater hit than the stock market during the Great Depression. Yet, the recovery was also stronger. During the 1950s, 60s, and 70s, the ESB’s net operating income exceeds the dividend of a stock market portfolio with the same initial asset value as the ESB by a large margin.”

 

  • “Because land is allocated to the highest bidder, we can derive the horizontal land use pattern from the right panel of Fig. 10. The dark-shaded area where the commercial bid-rent for land exceeds the residential bid-rent marks the central business district (CBD). Similarly, the light-shaded area marks the residential zone where returns to residential development exceed the opportunity return in commercial and agricultural use … developers face greater net costs of height when developing residential rather than commercial buildings. Therefore, developers of commercial buildings just inside the CBD can afford to build taller at lower floor space rents than developers of residential buildings just outside the CBD.”

 

  • “In Fig. 11, we pool the urban height profile of 55 North American cities sampled by Ahlfeldt et al. (2020). Consistent with the predictions of our model, building heights decrease relatively steeply in distance from the centre of the CBD. While, of course, absolute heights increased significantly, our estimates of the elasticity of height with respect to distance from the CBD, at -0.29, are almost identical for buildings that existed in 2015 and in 1900. In Table 2, we further zoom out to cover all 125 global cities for which Ahlfeldt et al. (2020) identify “global” prime locations (our CBD proxy) … A comparison of columns 1 and 2, again, reveals a stability of the slope of the height gradient that is remarkable given the significant reductions in the cost of height suggested by the results in Section 4.1. One interpretation of the evidence is that, over the 20th century, reductions in the cost of height and the cost of CBD access have just about offset each other in their impact on the (relative) height gradient.”

 

  • “In columns (3) and (4) of Table 2, we find that the commercial height gradient is steeper than the residential height gradient. This finding is consistent with the model predictions under the chosen parametrization and can be rationalized by a relatively steep decay in the production amenity, a low input share of floor space in production, or a relatively low net cost of height in commercial development … Using a boundary discontinuity design, we estimate that building heights, on average, increase by 0.2 log points as one enters the CBD (see Online Appendix Section F.2 for details). This result provides indirect evidence for a relatively lower commercial height elasticity of construction cost, adding to the direct evidence in Section 4.1. The remaining columns of Table 2 illustrate how the height gradient is steeper in North American cities than in European or Asian cities. This suggests a role for height regulation, which we discuss in Section 7.”

 

  • “In the bottom-left panel, we illustrate the distribution of the fundamental amenity values that convert the smooth height gradients in Fig. 10 into the real-world fuzzy height gradient in the top panel of Fig. 12. The residential fundamental amenity scales utility by factors that range between about 1 and 1.3, with the great majority of locations falling into a  window centered on 1.1. The commercial fundamental amenity scales productivity by factors that fall into an even narrower range of about 0.85 to 1. Hence, relatively little variation in fundamental amenity is required to introduce sizable fuzziness into the height gradient. Intuitively, users substitute away from floor space as its price increases, leading to a more-than-proportionate increase in floor space rents as the amenity value increases … The main conclusion from Fig. 12 is that relatively small differences, with respect to the micro-geographic amenity, rationalize large differences in heights of adjacent buildings, especially if there is mixed land use.”

 

  • “There is ample evidence that real-world height limits tend to be binding (see Online Appendix Section H.2). We know from Fig. 10 that the tallest building in our unregulated stylized city is slightly smaller than 70 floors. Hence, consistent with reality, only very generous height caps do not affect the spatial structure of the city … Height limits also have an effect on the horizontal land use pattern. A height limit reduces the amount of commercial floor space at every location where it binds. The limit to vertical expansion results in the CBD growing horizontally, pushing the land use boundary and residential use outwards. The effect on the horizontal size of the residential zone is the opposite. This is because the reduction in the productivity of the commercial sector due to displacement to less productive locations leads to lower wages and a loss of population. The resulting loss of agglomeration economies reinforces this trend. Eventually, there is less demand for residential floor space so that a smaller horizontal zone is required to accommodate the population even conditional on a reduction in vertical space.”

Shane Phillips 0:04
Hello, this is the UCLA Housing Voice Podcast, I'm Shane Phillips. This episode we're joined by Gabriel Ahlfeldt at the London School of Economics to talk about skyscrapers, yes, skyscrapers. When urban planners are taught about how cities grow, they're really just given a two-dimensional model, usually one with a central business district where demand and rents, and density are highest, and a smooth decline as you move further out from the urban core. The third dimension height is implied in that model, but never really scrutinized. Skyscrapers are a sort of platonic ideal of height within cities, and so they make a great starting point in that analysis. Despite their prominence, researchers still know very little about skyscrapers, including whether they actually make for good investments on the part of developers, how they shaped the cities in which they're built, or the ways they're shaped by those cities. Gabriel, along with his co-author, Jason Barr, who I also want to give a lot of praise for the work that's gone into this research are some of the first to tackle these questions in a really comprehensive and evidence-based way. I think it serves as a great introduction to how cities evolve, both in space and in time. The Housing Voice Podcast is a production of the UCLA Lewis Center for Regional Policy Studies, and we receive production support from Claudia Bustamante and Olivia Urena. Send me your feedback or show ideas at Shanephillips@ucla.edu, give us a five star rating and a review, and don't forget to share us with your friends and colleagues on social media. We do read your feedback and are always looking for ways to improve the show. With that, let's get to our interview with Gabriel Ahlfeldt.

Gabriel Ahlfeldt is Professor of Urban Economics and Land Development at the London School of Economics coming to us via Berlin, and he's here today to talk with us about skyscrapers. As I'm sure our listeners will pick up in this conversation, there probably aren't that many people who have given more thought to the role of skyscrapers in our cities. And so we're excited to have him on to look at them through a few different lenses, Professor Ahlfelts. Welcome to the housing boys podcast.

Gabriel Ahlfeldt 2:21
Hi, nice to see you guys.

Shane Phillips 2:23
And Paavo is my co-host today. Hey, Paavo.

Paavo Monkkonen 2:25
Hey, how's it going?

Shane Phillips 2:27
So we're going to talk about Berlin first, actually, and we're going to start with our tour, what are some of the places or things you'd want to show visitors to the city, if we were in town?

Gabriel Ahlfeldt 2:36
Well, we don't have many skyscrapers to show so we will probably do something slightly different. And I will probably just give it to here just outside my place where I live, where you have a pretty nice mix of amenities. You have the little canal and all the folks sitting actually alongside the banks and having a beer, and nice conversation chats. I mean, it's the kind of vibrant urban environment that planners always strive to create, but usually never succeed. So this is where I would start. I mean, just this is I think Berlin at its best.

Paavo Monkkonen 3:11
Yeah. And before we get into the third dimension of cities, Gabriel, I was hoping you could give our listeners an overview a bit of this, this awesome Berlin paper, you have, I think, kind of one of the most epic natural experiments. I mean, unfortunate, of course, in many ways, but I think, since you've done a lot of thinking about the horizontal dimension of cities to some extent, or the kind of view from space, I think that would be super interesting for our listeners.

Gabriel Ahlfeldt 3:37
So the evidence from the Good Lawyer yeah, that's maybe the best-known paper that I've ever written. So what's so special about the paper? I think, I mean, it all starts with the question that is maybe the most popular question urban economics, which is why do cities exist. I mean, that's, that's maybe the first question I always asked my students in an urban economics class. I mean, it seems so tempting to just live kind of spread out somewhere in the countryside, where you can have a lot of space, but instead, it's all crowded into cities, and there must be a reason for that right. The century old answer, I think, is Admiration Economies, the idea that if you should kind of live together, right, you work together, you're going to be more productive. Maybe it's also more fun, just like I said, right? I mean, if you go outside the house, having a lot of people high density also is a lot of fun.

Shane Phillips 4:23
People like people,

Gabriel Ahlfeldt 4:24
But the answer... exactly! But the classic answer is that we can be more productive right? Now, that makes total sense, right? I mean, you have the flow of information, knowledge, and you have the demand in the labor market to find the job that really gives you your best productivity. That sounds so cool, but empirically kind of showing that right, I mean, this causal effect from density, on productivity - something that's not so easily done, because there's always this reverse causality in the air. I mean, you could always argue, think about Manhattan, you have a wonderful natural harbor, right? Maybe Manhattan is just fundamentally very productive, right, because it's very productive, you have high wages, and because of that people come to Manhattan. So it's not about density that makes Manhattan very productive, it could also be just the fundamentals. So how can you prove this kind of story in urban economics that density really makes places more productive? Well, you need some exogenous shock and natural experiment, as you said Paavo right, something that kind of exorbitantly changes exposure to density, and you don't find that that often. And clearly, this is where the Berlin wall comes to play a role, right in the city. The idea in the paper is that you look at West Berlin, and then suddenly you have the wall, the wall comes up, and that wall kind of isolates some areas in West Berlin, close to the historic central business district, from the historic city center. And if you believe in the idea that I mean, there are all these spillovers in the area, and they can spread from the CBD on to nearby areas, you would expect that actually, those areas which are now kind of insulated from the spillovers, they would lose productivity. And that's the paper in a nutshell, we showed that I mean. So these places, I mean, closer to CBD in West Berlin, there was a productive, extremely high rents, wall goes up, I mean, all of that is lost, and when the wall comes down, I mean, the reverse happens. And this is really the story that we tell on the paper that shows quite convincingly, that density has a causal effect on on productivity, okay. But I mean, the water side, I should also say that, I mean, in order to kind of show this, this formula, and also to kind of quantify the effects, we build this quantity of spatial model, which is kind of a contribution in its own right, and I think probably the paper by now has been more impactful actually, for the model than just for the empirical results that density matters, I would say,

Shane Phillips 6:42
The article we're discussing in more depth today is in the Journal of Urban Economics, it's titled 'The Economics of Skyscrapers: A Synthesis', and your co author on it is Jason Barr. In this article, you've developed a model here as well, that if I can simplify things a bit, is exploring what conditions lead to the development of skyscrapers and how they influence urbanization in cities, and also how restrictions on skyscrapers, especially height limits, can alter or redirect the processes of urbanization. My observation kind of coming into this as someone who follows housing discourse is that skyscrapers can be surprisingly polarizing, especially for how little of the built environment they actually account for. For some people, they represent a sort of pinnacle of urbanization, and efficiency and even beauty. And for others, they're sort of monuments to individual ego and urban inequality. And, you know, my intuition is that there's some truth to both of these views. At least that's how I, you know, kind of came into this paper feeling. But before we get into the actual content of the article, I'd be curious to hear just what drew you to this topic, why should we care about skyscrapers as a distinct building type in our cities?

Gabriel Ahlfeldt 8:00
I think it takes a more positive view in the sense that I care about the wise, rather not so much about whether skyscrapers are good or bad. I mean, I think to me, the fascinating thing is really that I mean, if you look into the history of cities, right, I mean, you have this kind of evident trend into the third dimension, I mean, the vertical growth, I mean, it's just kind of happening at a totally unprecedented rate. And at the same time, I can't even really say why, I mean, urban economics research has long been sticking as Paavo says, right, to horizontal land use pattern, right? We see cities are growing vertically, but we still kind of take the look from outer space on cities, and we also do that in the Berlin Wall paper, right? And the Berlin Wall paper, as nice as the model is, right, but there's no explicit treatment of height actually, in that model, right? And I thought I mean, okay, I mean, this is kind of a striking disconnect, and, obviously, you want to make sure that the history of Ford and Erban Economics somehow keeps up with the history of cities, and me, and also a couple of other people, have recently started to look into the vertical dimension. We now know a little bit about how rents change within buildings, as you move to higher floors. We start to have some notion of how construction costs increase as you build taller. So I thought that maybe now's a good time to try to bring together actually what we know about economics of skyscrapers in one paper, and also kind of give a little bit of research directions, so that the field has some suggestions on where we could be going.

Shane Phillips 9:31
And your model in this article is kind of an elaboration of previous models of spatial equilibrium within cities, which might be a phrase that's familiar to some people who have gone to urban planning school recently, but not many other people. So as simply and jargon free as possible, could you just give us an explanation of what a spatial equilibrium model is and what these kinds of models predict about the form that cities take, and what your model is adding with an emphasis on skyscrapers.

Gabriel Ahlfeldt 10:03
So the spatial equilibrium is a helpful concept that is based on the idea that firms and workers are indifferent between locations. So basically, you want to call a situation and equilibrium, if nobody has an incentive to move from one place to another. So how does that work? I mean, how is that possible? Well, the idea is that, of course, you have some places in the city which are more attractive than others. Okay, but what's going to happen in this equilibrium in this situation, is that the rent at these different places is going to adjust to offset for the benefits that you get from your location - of course, you want to be in a city center because you have a shorter commute, but at the same time, you're going to pay a higher rent. And if you take into account both factors, you're kind of indifferent. Okay, this is what we call a spatial equilibrium. And then we start from that, and we go modeling the city, we think about different economic agents. I mean, most typically, we think about firms and residents, right. And then you're going to have different factors that for which different agents like firms and residents have a different willingness to pay. I mean, typically, we assume that firms, they derive a real big benefit from being in a city center. Okay, so they're going to kind of outcompete residents from the city center so you get the kind of horizontal land use pattern with firms at the city center, and residents in the suburbs okay. You can become more sophisticated, like what we do in this model, we do a general equilibrium model. So we also take into account the labor market so it's not just about are you indifferent between locations within the city, it's also the question, when are you indifferent between being in the city or living outside the city, that's going to depend on the real wage. And so you can also learn something about city size. And then you can incorporate the third dimension, which is what we really do here as a novel contribution to this literature, we start thinking very explicitly about, okay, what are the costs and returns to height, okay? And we make that use specific, and, and this is, this is important, as it turns out. So basically, you have a benefit from being high up in the building, okay, you get a nicer view, pay a higher rent, okay? At the same time, you build a taller building, it's going to be more expensive, okay? Now, the rate at which these costs and benefits change, they can be use-specific okay - it can be more expensive to build a tall residential building tham a tall commercial tower, and that is something that literature has completely ignored. This then has also some implications for the horizontal land use pattern. Okay, so if it's very difficult to build tall residential buildings, then that's an additional force why we see commercial land use in city centers, where you have extremely high land rents. That's something that is completely absent in the standard revenue model but we kind of fit that in into this otherwise fairly economical general equilibrium framework. So that's, that's basically the contribution I would say

Shane Phillips 12:52
Yeah, and we'll come back to that difference between the commercial and the residential and how this influences the horizontal as well as the vertical. But we should, I think, start by being clear on how we're defining skyscrapers because what people consider a skyscraper certainly varies a lot from place to place. In parts of New York or Hong Kong, someone might not consider a 40-story tower, a skyscraper. But there are people here in LA, who might refer to a five-story building as a skyscraper, if not referring to it as something much worse. For this paper, though you were using 150 meters or taller as your threshold, which is about 500 feet. What were your reasons for picking that height? And is there anything we should know upfront about how that might bias some of your findings and your insights in this paper?

Gabriel Ahlfeldt 13:42
It's a good question, right? I mean, I think there is a theoretical and empirical dimension here. So for the model, I mean, that that high threshold really doesn't matter. So the model generates a full distribution of building heights. So you many locations in the city, they have different rents, and some places have very high rents, and you get a lot of tall buildings. And then you have some places where rents are very low, and you get short buildings, and in between, you get a full distribution, okay? And the model can tell you anything about any high threshold, or any kind of measure of vertical size. Empirically, you want to make your life easier and just kind of focus on some measure of vertical size. We pick 150 meters, because my co-author, Jason tells me that this is fairly canonical in the literature. But to be honest, it doesn't matter so much, right? I mean, you can now go and say, okay, look, let's put the cutoff at 100 meters, let's do it at 150 meters. You can actually count the total height of all the tall buildings in the city. You can also do something like just measure the height of the tallest building, right? There are many ways of how you can measure vertical size, but it turns out that they are all super correlated. So frankly, I have a hard time thinking about how really, this definition could be particularly influential for me off the conclusions that we draw on this paper, I think it's just kind of a one way of many ways of measuring vertical size.

Paavo Monkkonen 15:07
Yeah, it's interesting that that kind of incrementalism that you emphasize throughout the paper, in terms of height, I think probably would apply to kind of people's perceptions as well, right? Because once you have, you know, a city that's mostly four-storey buildings, and then you get a 10 storey building, then people think of a skyscraper, but, you know, it's all just kind of conditional on people's perception of normal, I guess.

Gabriel Ahlfeldt 15:27
Yeah, and this also happens. I mean, it's also generational specific, right? I mean, what people perceived as being a tall building, right? I mean that's very different actually if you zoom back in history,

Shane Phillips 15:38
Yeahnd, a your article actually has a bunch of cool history and stats on the evolution of skyscrapers over time, really all across the world. Could you sketch out some of that history for us especially since the turn of the 20th century. What were some of the eras of skyscraper construction, and what technological innovations made each of those eras possible?

Gabriel Ahlfeldt 15:59
Well, I think it's the necessary condition for something like a skyscraper, to emerge, basically, I mean, that was the elevator and the steel frame, right? I mean, that actually allows you for the first time to really move beyond 10 floors or something, and start having really tall buildings. So I think it's actually no surprise that somewhere around that time, right, we usually kind of record the birth of the skyscraper, although it's actually surprisingly controversial, actually, which was the first skyscraper building - Jason could elaborate on that for ages. But roughly around the turn of the century, I mean, we had the first skyscrapers because of these technological innovations, then people pushed the margins as much as they could, during the kind of famous skyscraper race during the 1920s, then you had the Great Depression, and then for some time, not so much happened. And then as far as I understand, you had the 1960s, when structural engineers started to use mainframe computing, I think that was a big game changer. Now, suddenly, you could kind of start building really tall structures in a much more efficient way - by efficient were mean, you'll use a lot less material, therefore, the building gets lighter, that allows you to go much taller, you can withstand these kind of collateral wind loads, which seems to be the big challenge in building tall buildings in an efficient fashion. And since then, I mean, the technological process hasn't come to an end, I mean, maybe it's a little bit more incremental, but it's absolutely steady. So we see more and more skyscrapers. ever greater heights popping up all over the world. And I don't think this is a development that is coming to an end As Jason tells me that at some point, we're going to have elevators that are magnetic, that don't need cables anymore, and that can actually allow skyscrapers to go to 1000 meters or even a mile or something right. So that's one part of the story, I think it's really important to think about the supply side here but it's, of course, just one side of the story, right? I mean to build a skyscraper, it's not just a question about the cost, it's also a question about the revenues, you need high rents, right? That's, that's the key region here, you need really high and the rents to finance these tall buildings. And you're going to get these high rents and environments where you have relatively high incomes, and growing urbanization, okay. And I think this explains why you first had the skyscrapers in the US right around the turn of the century. I mean, the US was a relatively rich country, cities were urbanizing fastest creating demand pressure for skyscrapers to emerge. But, of course, since then, the gravity of development has shifted a little bit, right? I mean, we have now a lot more development in the developing world, in particular, in Asia. So you have these megacities, in particular in China, where you get this combination of relatively high incomes and fast urbanization, which are basically this kind of second ingredient besides technology that facilitates vertical growth. So I think that's probably the history of skyscrapers in a nutshell.

Paavo Monkkonen 19:01
Yeah, I thought it was very interesting, kind of if you look at that graph, you have in the paper about kind of the appearance of, you know, completions by year over the last 100 years or so. Up till about 2000, it seems like urban population growth is exceeding skyscraper production but then after 2000, skyscraper production really takes off. And I guess that is probably the importance of GDP in producing these things, rather than just large cities.

Gabriel Ahlfeldt 19:28
Yeah, I agree, it's GDP that is important. Interestingly, it's also GDP growth so if you run a simple horse race, and you want to just kind of check, what is the strongest predictor of vertical growth? I mean, it used to be GDP levels used to be very important, but over time, GDP growth actually became more important.

Paavo Monkkonen 19:52
Forward looking speculative developments and optimism.

Shane Phillips 19:56
Yeah, yeah.

Gabriel Ahlfeldt 19:57
Well it make sense right? I mean, if you think about the durability of this capital stock, you're building a skyscraper, that is going to last, how long 100 years easily, right? I mean, you got to think about the future.

Shane Phillips 20:10
And I think we can just summarize quickly, but I think the concentration of skyscrapers is really interesting, too, I was pretty surprised to learn that 50% of the world skyscrapers, at least by this definition, are in just 17 cities. I know, there are a bunch of reasons for this, and we'll probably get into some of them. But could you give us a sense for why so many of the world's skyscrapers are found in just a few places? One stat just to pull out here, you know, just in terms of the locations, is it 16 of the 20 cities with the most skyscrapers are East East Asia, kind of as you hinted at, and just three now are in North America, in New York, Chicago, and Toronto; Toronto is kind of a surprise to me as well.

Gabriel Ahlfeldt 20:51
So I think the fundamental reason why you see skyscrapers and relatively few cities is that they are very expensive. And again, what you really need as a necessary condition to build them is you need high rents, and high rents, we know that I mean, they basically emerge in cities that are large, and economically prosperous, and that kind of shrinks the set of candidates, by actually quite a lot. You then have a bunch of cities, particularly in Europe, where maybe skyscrapers would be economically viable, but the planners don't like them. So you get these high constraints, and basically, you've banned skyscrapers. I mean, that plays a big role so you shrink the set even further. And I think what's going on in China is there, you you have the opposite - you actually have planners who like skyscrapers. And what they do, I mean, from my limited understanding of the institutional context is that they view the skyscraper as a potential catalyst for local economic development, which is not crazy if you think about the idea of collaboration spillovers, right? And say you put a bunch of productive people into an area, right, if you believe in the idea that there are spill overs, so maybe they can have a local impact. So what I ended up doing eventually is to heavily subsidize these skyscrapers through cheap land leases, okay, so you see, probably the opposite of Europe; in Europe, you see fewer skyscrapers that you would normally get, and in China, you see more skyscrapers that you would obtain under free-market provision because the land for skyscrapers is subsidized. And I think to some extent, this explains this Asian, Chinese bias in the skyscraper development these days, although of course, we also need to acknowledge that Chinese big cities, they just tend to be very large, right? So to some extent, it's also natural that we see many skyscrapers there.

Paavo Monkkonen 22:40
Yeah, I mean, I think it's kind of the regulatory manifestation of what some people might think about as culture, right? And talk about like, you know, it's just not we don't like living in high buildings. I mean, hear that a lot in Latin America. You know, kind of we prefer living in this manner but then, you know, maybe the regulations follow that. And that's actually what's what's shaping things. I wonder, one thing that I didn't see come up was, or maybe not, in these words, but it's kind of the coordination costs and kind of risk implied for kind of this projects of this size, right? So you might imagine countries that kind of weaker institutions or property rights might be harder to get one of these things or kind of you'd be less interested in investing in one of these things, you know, then in another context, even if the GDP was growing rapidly.

Gabriel Ahlfeldt 23:24
Yeah, I think it's an interesting thought. I mean, it makes theoretical sense but I can't remember having come across any evidence that would substantiate or reject that. So I think that's probably a good area for future research.

Paavo Monkkonen 23:37
Because also, if you think about like a lot of the financing of real estate projects of this size, have pre-sales as like an important part of revenue streams. And if you think about suburban developments with presales, well, you develop half of it, and then people are living in that. And then the purchase of those homes finances the second half, it's harder to see under construction, but occupied skyscrapers right. So that kind of model, I don't think would work as well.

Gabriel Ahlfeldt 24:01
Yeah, I agree, and and then there are also limits. I mean, to take the economical model, to developing world cities in which you have informal land markets, I mean, an informal land market and skyscraper development that doesn't go very well precisely for the reasons that you just outlined,

Paavo Monkkonen 24:14
Although there's this famous famous one in Venezuela, right, it got half built and then occupied by informally, it's pretty fascinating.

Shane Phillips 24:24
We do always like to recognize a paper that put a lot of work into the data collection or analysis here, and this one certainly qualifies. There was a ton of data you collected on skyscrapers all over the world. Could you just say a little bit about what that process looked like? What sources of data you're drawing from for this paper? And maybe anything interesting that came out of that process?

Gabriel Ahlfeldt 24:48
Yeah, I mean, that's true. And we use a lot of data in this paper, although to be fair, actually, quite a bit of the data we have used, Jason and I have independent previous research we already have had conducted. I mean, for instance, I had used this enormous database. Before the impoverished database, I think it's considered to be the most comprehensive database on on skyscrapers. It tells you exactly the location of skyscrapers, the height, I mean, the date of construction, sometimes even the construction cost I had used that in the previous project. Also the land values that we collected for Chicago. I mean, that was a major thing, right? I mean, for 150 years, digitizing historic maps that gives you exactly the block level, the land values, I mean, that took us ages and armies of arrays to help us and

Paavo Monkkonen 25:32
So this was the one with the with Dan McMillen, is that that paper?

Gabriel Ahlfeldt 25:35
Yes, absolutely. Exactly. So we used that for the for the tall buildings. The other tall buildings paper, Jason had done similar research for New York, that came in very handy. So we could do these kinds of comparisons of Chicago and New York. We also collected new data for this paper. And I think the coolest data we collected I think is the one on the Empire State Building. So at some point, Jason went to the archives and Delaware, and I think he spent a couple of days really taking hundreds of photos of all these historic files when you could determine these records on the historic balance sheets, and the conversations, the the owners of the buildings had with the authorities, I mean, to get planning permission. And then later on during the Great Depression. I mean, they were complaining a lot because the building was fairly empty, and they were at the risk of bankruptcy, at least that's what they claimed, and they wanted to lobby for tax discounts, which was really nice to read into it. And then also something which was fascinating is to see actually in these balance sheets, how then quickly, right. I mean, even I think even while they were still arguing for tax discount already, I mean, the rents kind of skyrocketed. The vacancy rates dropped, and the building became super profitabl, off course, the owners wouldn't talk about that, right. And most of the economic historians that are working in that area, they actually missed that right. And this why this building still has this this nickname of being the empty state building. Whereas actually, this was just an issue for a couple of years, and for most of its lifetime, the building was profitable, easily but the stock market in terms of net returns. So we'd be brought to light, I think, kind of nice piece of urban history, which is, which felt exciting. That doesn't usually happen in my life as an urban economists were mostly kind of sit down with a piece of paper and pencil to do some boring math or some programming right so yeah, that was cool.

Shane Phillips 27:35
Yeah, since you brought that up, let's let's jump ahead to that conversation about, you know, the profits of these buildings. I mentioned already, there's this idea that skyscrapers have more to do with the egos of their owners or developers sometimes than actual profit motives, or profit maximization anyway. I'm thinking of how people have looked to surges in skyscraper construction as a sign of irrational exuberance in the market, and maybe a signal that bubble is about to pop, sort of like record breaking skyscrapers are almost a canary in the coal mine for an overheated or somehow at risk economy. What this is all really getting at is the question of whether it makes financial sense to build buildings at these heights. And I read the article as basically saying that, yes, the these projects are pretty profitable, at least the ones you looked at in detail, it's not necessarily all about ego, whatever they cost. Could you explain how you tried to answer that question, and what you found, like a little more detail, especially on the Empire State Building but also your analysis of second and third tallest buildings and these other approaches that you took?

Gabriel Ahlfeldt 28:45
Yeah, I mean, let me just kind of quickly mentioned them, you brought up this point about skyscrapers predicting models. I think, the work that Jason suggested, actually, this doesn't work, right. I think that you can actually economy predict skyscrapers, but skyscrapers actually...

Paavo Monkkonen 29:02
But there are a few really notable anecdotes suggesting it does.

Gabriel Ahlfeldt 29:06
Well, there's always, always a difference between an anecdote and systematic evidence. I'm not going to contest your anecdote, but I think Jason would claim and I'm just passing this on, right. I mean, this is not my research, but he would claim that actually this idea that you can predict the bubble through skyscrapers. That doesn't work. I mean, but this is what's gonna be a great

Shane Phillips 29:27
Great investment strategy. If it if it did, though.

Gabriel Ahlfeldt 29:30
Yeah. I mean, we all want to be able to predict bubbles, right. I mean, I agree. More generally, I think I mean, I came to convince myself through the work in this project, that probably this idea of insane heights being driven by egomaniacs that does probably not have a lot of support, right? I mean, I wouldn't claim that there is no building in the world which seems economically unfeasible. I mean, there are a couple of cities where you find that the tallest building is a lot taller than the second tallest building. And you kind of wonder why right? So maybe you have reason to be a little bit suspicious there although you need to be careful, because especially if you're in a touristy city, and you have an observation deck, I mean, the revenues that you get from those observations can be massive. I mean, we saw that actually, in the Empire State Building, it can completely skew the, the math, I think, I mean, the at some point it really makes a difference. So I would even be careful with challenging ....

Shane Phillips 30:35
There's probably I'm just thinking of this now but there's probably also, you know, we've got in many respects, the sort of winner take all economy, especially you see this with like social media, where it's like, whoever wins, takes it all, and is worth many billions of dollars, and everyone else is just kind of nothing. In the case of like observation decks. Also, it seems like if you're the tallest or the most attractive in some way, you're gonna draw everyone and anyone who's second, third, fourth is just kind of you're maybe getting some people but just nowhere near the same level as the (tallest)

Gabriel Ahlfeldt 31:07
Yeah, absolutely. So even if you see something as suspicious as the Burj Khalifa, in Dubai, right, you could argue that maybe even that makes sense. But this situation, right, I mean, that you've seen Dubai is the huge exception, right? I mean, if you if you look at as we did, right, I mean, systematically across all cities that adopted the skyscraper technology, what you find is that usually, the second tallest building is pretty similar in height, to the tallest building. And the third tallest building basically is actually identical in terms of height to the second tallest building. And I just don't think this is a coincidence, right? I mean, this is, of course, I mean, the result in in general, that people do quite careful calculations about profit-maximizing building height, because they have an interest in maximizing their profits, and they come to similar conclusions because they are driven by fundamentals. And coming back to the Empire State Building. I mean, this is, I think the literature treats it as maybe the most overly ambitious building in history. I mean, it has this unbeaten track record of being the tallest building in the world for 40 years, which just really shows I mean, how out of proportion, it seemingly was, when it was constructed, and therefore all these nicknames about the empty state buildings. But really, I mean, as I told you, right, I mean, if you go through the balance sheets, I mean, you see, this is all just based on on a mood that emerged during the Great Depression, I mean, you build this tall building, right, now comes the Great Depression, of course, I mean, you're gonna have a problem filling that building. I mean, that's not a surprise. But soon after that, actually, you see that the stats look super healthy, right, and, and over the lifetime, I mean, the net return on investment was great. And also you have some people from the construction industry that did independent calculations on the optimal building height in Manhattan, and they come up actually with a size that is not so different from the Empire State Building. So I guess the point I'm trying to make, if even this kind of akin right of ego-driven skyscraper development that emerged from this guy from the competition with the Chrysler Building, right, I mean, if even that building was profitable, right, I mean, have a little bit of a hard time thinking about many buildings in the world that were deliberately constructed as, as loss-makers, right, just to satisfy some ego. Also because these things are just so expensive, right? I mean, if you talk about really tall buildings today, I mean, you're talking about billion dollar amounts right? I mean, you need to be really, really rich, right, to afford huge losses on such investments. So I think the story is a little bit overstated, right, I would say.

Paavo Monkkonen 33:37
One thing you hear on this a lot in Latin, well, in Mexico at least probably in other parts of Latin America, is that they're their money laundering operations, right. So you do see a few in in northern Mexico, there's a bunch of state I have some some former students that are working on research on verticalization, as they call it, in northern Mexico, where their cities that you know, they haven't had a building over 15 storeys ever, and now they have a handful of 40 storey residential towers and stuff. And they they connect it to money laundering, but I guess, in the spirit of average effects, or average relationships, this is probably not a global global phenomenon, and maybe not even true.

Gabriel Ahlfeldt 34:10
I mean, Tomas, I mean, you can easily burn a lot of money right? So I can see how money laundering would actually work very efficiently. I cannot comment actually on how widespread the phenomenon is. I mean, it's cool, cool, cool. Cool. Inside. I didn't I wasn't aware of that. Actually, it makes sense but I really don't how generalizable that phenomenon is.

Shane Phillips 34:30
Right, I mean, and just thinking about the Empire State Building in particular, and how it was the tallest building in the world for 40 years. That kind of strikes me as maybe just like a historical quirk to where you know, it was finished at the beginning of the Great Depression, so no one's going to be coming after and then you got a war, and then you've got suburbanization in the US where it's just like dispersing demand constantly away from city centers and sort of, you know, it was our policies that had nothing to do with skyscrapers that sort of drew people out of cities and presumably, you know, brought down the land rent so that they didn't really make sense to build like that again for a long time,

Paavo Monkkonen 35:11
especially for the wealthy.

Shane Phillips 35:12
Yeah, yeah, so let's get into this economic model. This is central to your paper, this model of a city and the urban economics that help shape it, and you make the point that barring restrictions, like height limits, developers will keep increasing the height of their buildings, so long as the increased rents they get from adding floors, is larger than the increase to construction costs. This seems like a good time to talk about construction costs. And I think there are two important points here, at least two. The first is that as you build taller, the cost of building increases, and it's not just that it costs money to add each additional floor of a building, which is obvious, but that adding another floor often makes every floor below it a little more expensive to build. I'll get to the second point in a moment here, but let's just hold on that observation. First, could you explain why that's the case and what your model and other data tells us about how big of an effect that is?

Gabriel Ahlfeldt 36:11
We establish in this paper, but also in the previous work with them like Milan, I mean, that this relationship exists empirically, for some reason. Actually engineers, they don't work empirically. So they don't establish these relationships into data. But they do just kind of based on engineering kind of construction, mathematics. I feel quite confident to talk about the magnitude of the effects. You want to take findings on the origins of this effect with a pinch of salt, right, because I'm not a structural engineer. But what I understand is that, of course, there are a couple of reasons that account for that. I mean, first of all, you need a lot more sophisticated structural engineering, right? When you build taller, so you need to make sure that you ensure the structural integrity of the building, that means you need better materials, right? I mean, you need a lot more sophisticated engineering in general. You also have to issue of foundations, right? I mean, the taller you build, I mean, the deeper you want to go with your foundations, and you have to the complicated issue with the bedrock, right? I mean, you want to have the bedrock at some depth, if you're gonna go deep but then at the same time, you also don't want to have to bedrock too into the surface, because then you have to blast away. Right, that all gets more complicated if the building gets taller. And then the third factor for which we also have support in our data is that you lose a lot of net floor space as you built tall right? The taller you go, the more space you're gonna dedicate to elevators and other facilities. That means that the ratio of net floor space across floor space falls significantly. And I think all these three things together, they explained why it gets more expensive in per unit terms when you built all them. Now to the magnitude, I think it's fair to say that the cost of height increases at an increasing rate, right? I mean, there's there's not really a constant elasticity here, right? which complicates things a little bit. Say you go from 10 to 20 floors, I'd say I mean, you have to live with an increase in construction cost per unit, maybe off 25% or something, you go from 250 meters to 500 meters, right, I mean, that can easily double the cost in terms of per unit construction cost, right. So this, this effect really kind of builds up exponentially. I suspect there are a lot of more kind of complex issues here around thresholds and so on, so forth. And I think we need a lot more research actually, to understand completely the shape of this, this construction cost function. That said, I find it still quite astonishing that the good old rule of thumb according to which a building cost in terms of per units increased by about 2% for each additional floor, that would have done that actually has surprisingly strong support and data. It's an imperfect approximation, of course, right. But actually, it's quite a decent approximation. So that I found interesting.

Shane Phillips 39:09
And the second point here, that I have in mind is the discontinuous relationship between construction costs and height. And this isn't really explored in your article, and I think for good reason, because it's mostly as I understand it applicable to buildings more in the 10 storey range and below

LA skyscrapers

Exactly, yeah, I think it's pretty intuitive that as you build taller the construction cost per square foot or per square meter does go up. You know, as you said, you have to invest more in structural materials and reinforcement foundation. Maybe you have to dig deeper, deeper hole. There's parking, your elevators and other circulation get more complex take up more space. All those costs get spread across the whole building, but the increase isn't necessarily linear for every single floor. So you may not have have a consistent 2% increase per square foot in construction costs as you step up from five to six to seven to eight to nine storeys instead, and I'm just making up numbers here to illustrate so don't take these too seriously, costs may go up 2%, when you add the sixth floor, and then 2% more, when you add the seventh floor, this is per square foot again but when you get to the eighth floor, they might go up 30%, or 50%. And the reason is building code requirements. Buildings between three and five storeys might have very similar construction costs per square foot, because they're structurally all wood. And I guess I should clarify, I'm thinking in the US context, here, it's going to be different in other countries, but then the costs go up for the sixth and seventh floors, because those additional levels have to be built out of concrete, or at least the main structure does. And then everything gets way more expensive when you hit for eight or nine, because now you're building the whole structural core out of concrete or steel. One important consequence of that big jump is that once you've crossed the threshold into steel or concrete construction, you often you might as well go for 20 or 50, or 100 stories, rather than just sticking around nine or 10. So again, 10 - 20 storey buildings, these aren't the focus of your article so it's not an oversight that is not discussed there. But it does seem like useful context, and I'm curious to hear your thoughts on how this relates, if at all, to the really big 150 meter and taller skyscrapers? Are there other kinds of thresholds you've come across maybe that tall buildings can hit that end up making the whole thing substantially more expensive or is this more of a kind of mid-rise phenomenon.

Gabriel Ahlfeldt 41:46
I'm not so sure. I mean the problem with the data is that it's relatively sparse. So it's not like in the data that we have, every skyscraper actually has the construction costs recorded. So we work with limited data, which unfortunately, isn't really suitable to really get into the nonlinearities, in particular discontinuities of the construction cost function. So I can't tell you too much about that. I mean, talking to architects and engineers, I mean, I hear that even for especially residential tall buildings, you could have these discontinuities, because of certain heights, I mean, then then you have to change the way you kind of do the pipes and other plumbing and stuff, right? So it seems like you're gonna kind of have this combination, but what you're describing is a phenomenon where you go beyond a certain height, and then suddenly, you incur fixed cost. And then for some floors, you can spread that fixed costs across the floor, so that the average cost goes down before it goes up again, right? This could happen multiple times. And somehow in my mind, I see a smart kid coming up with a job market paper exploiting these discontinuities in the supply of floor space. And I'll estimate the slope of the local demand, curve rights and preference, heterogeneity and all these parameters we need and quantity spatial models these days. The necessary condition to do that will be to find better data sets then that we have, right.

Shane Phillips 43:12
One other thing to note is that your model finds that commercial skyscrapers make the most sense financially within the innermost core of the city where land prices are highest. But then once you get a little ways out from the core, residential towers can actually earn a better return. At first, this kind of surprised me, because the data you have say that rents generally go up faster as you move up a residential building compared to a commercial building. But I'm guessing I think it's my understanding that the reason commercial skyscrapers are still preferable, especially in the core, is because it's cheaper to keep going taller. So to say that another way, yes, rents go up faster as you move up each floor of a residential skyscraper. But construction costs also go up faster for residential buildings as they add height. What really matters is the gap between how quickly rents go up, and how quickly costs go up as you go taller. And that gap is often bigger for commercial buildings, so it makes sense to build them taller. So first, just tell me if that's a fair interpretation of the data that you that you've pulled from and then tell me what I'm missing from that explanation because I know it's glossing over a lot of nuance, and this interplay between construction costs and rents and also land prices, which I haven't really talked about here.

Gabriel Ahlfeldt 44:34
Yeah, I think you're pretty spot on. So the way to think about that, I mean, from the perspective of a developer, is you need to figure out a net cost of height right? I mean, you've got to take into account the returns to hide the cost of height. And then the net cost of height really is what you should be concerned about and I mean, if you if you trust them in the preliminary results that we have, I mean, they're the first of its kind, we still need to build an evidence base but so far It seems quite unambiguously, I would say that the net cost of height is larger for residential building. And this is exactly as you said, right, because just kind of the engineering of a tall residential tower is more complicated. It's the thing about the ground floor area, you can have these huge floor plate areas for commercial towers and put everybody on the huge trading floor. I mean, that doesn't work for residential towers, I mean, this makes it a lot more complicated a couple of other things. So this makes always building a tall residential tower a little bit harder. I think you'll also need to keep in mind that you could totally abstract from that, and you would still get the analysis pattern with commercial land use in the center, simply because firms pay a greater premium for being in the city center. Also, workers consume less space at work than at home. So overall, the expenditure share on floor space is lower for a firm that for a household, that makes it easier for firm to afford high rent. So all of that pushes offices already in the city center. But then your logic exactly kicks in, right, and on top of that, you have the lower net cost of height for commercial offices. And that creates an additional competitive advantage for offices in the city center. And the way to think about it is that this makes the central business district a little bit bigger than it would otherwise be. And it also gives kind of rise to these height discontinuities that we see actually, for the first time in our model. So the model would predict that actually, as you go from the central business district to the residential area, you're going to see a sudden jump in height, which I think you do see in real world cities, right, you often see the CBDs standing out in terms of tall buildings, right, and being surrounded by relatively flat residential buildings.

Shane Phillips 46:53
It's not just a smooth gradient up to the tall building. Yeah, you mentioned how office workers tend to use less space. And I think that's sort of an interesting observation, in part, because it's kind of an extension of how we think about like multifamily housing and how you're sharing land between a larger number of people. And so office is kind of taking that another level where instead of sharing, you know, having a bunch of housing units, and maybe one person per 400 square feet, 300 square feet, and an office is maybe one person have per 100 square feet, I think that at least used to be the standard. You kind of talked about the difference in building heights, and your data does show really large differences in building heights, even among skyscrapers within a given city, even for buildings, just a few blocks away from each other. I know there's not a single answer to this. But why don't we see the same homogeneity in design and height for buildings, say over 20 storeys or over 150 meters tall, as we do for like modern sixth and seventh storey mid rise apartment buildings, for example. Shouldn't there be some profit maximizing design or height that's proliferated, and at least somewhat taken over the market by now? Why haven't we seen anything like that?

Gabriel Ahlfeldt 48:12
Yeah, I mean, this is an interesting question. I mean, in the paper, we talked about what we call the fuzzy height gradients. And the fact that you see these, these buildings of very, very different heights side by side within the city, and you kind of wonder how this comes about. And I think there are a couple of explanations, which are not mutually exclusive. And one of the things that I found interesting when I played around with the model, was that I need relatively little variation in economic fundamentals, and also rents to rationalize large differences in building heights. So basically, what I did is I said, okay, look, I'm gonna take Chicago, right, I have the real world kind of high profile of Chicago. And I try to explain that purely through differences in terms of local amenities. So you would have something like a subway, for instance, in one location, or a view of the Chicago River and because of that, you would get, let's say, 5%, higher floor space rents. And that would incentivize the developer to develop taller. And one of the interesting things that comes out of the model on fairly economical parameterization is that you need relatively small differences in floor space rents to get relatively large differences in heights. So if you get a 10% difference in rent, you're going to get much more than 10% difference in building heights, because a lot of substitution is going on here right? I mean, firms, when rents go up, they use less space per worker, right? I mean, when you build taller, you use less land per unit of floor space right. And that explains actually more than I had thought, right. I think differences in building heights through plausible variations and amenities, but of course, that's just one part of the story. I mean, the other part of the story clearly is that you have vintagey facts I mean, buildings are durable. They are built at different points in time and at different points in time, you have different construction technologies and you also have different rent levels therefore you also have different heights and it's very difficult to change the height of a building exposed, right. So, once you have the building at a certain height, it stands there and this creates these discontinuities, but I would also not discard the planner, right? I mean, so planners, they have very particular views about buildings and, and and heights. And usually, there is this this this trade off going on, and many, many cities where you say, okay, look, I mean, I want to build tall and the planner tells you "Yes, you can, if you give me something in return", that can be social housing in many cities, right? I mean, you will be allowed to build taller ...

Shane Phillips 50:41
Usually it's parking.

Gabriel Ahlfeldt 50:43
Good. I didn't know that right. Parking is a good example.

Shane Phillips 50:47
it's a bad example, actually, but yeah

Gabriel Ahlfeldt 50:52
I mean so if you really want to create density and get good parking in return, I mean, you kind of wonder about the objective function here is of the planners

Paavo Monkkonen 51:04
A bad one, it's a bad one.

Gabriel Ahlfeldt 51:07
That you want to maximize the congestion on the roads, right? I mean, if you want to do a skyscraper, parking lots are it,

Paavo Monkkonen 51:14
I mean, to me the durability and like what you're going to call the vintage effects seems like, you know, could account for a lot. I mean, what you're talking about the amenity effects is very interesting. I hadn't thought that it would be that predictive of kind of such a variation in heights but yeah, just with, you know, building happening at different moments in time at different rent levels, and then you know, you don't want to tear down the 10-storey building to build a 15 story building, that's usually not profitable.

Gabriel Ahlfeldt 51:44
Yeah, that must play a big role right? Although you'd be surprised, I mean, the about the planner that can be quantitatively really, really meaningful. So my colleague, Paul Cheshire, he did this study on London, proving that when you come to the Planner, and you promise to employ a trophy architect, so a guy who won the Pritzker Prize or something like that, right, then you get some additional 17 floors, right, in a city where the standard height is eight, right? And this can actually also that make quantitatively a very big difference, right?

Shane Phillips 52:17
So the huge fees that they're charging can actually be worth it, it's allowing you to build a much larger building.

Gabriel Ahlfeldt 52:24
That's what Paul would say. The whole benefit is being appropriated by the architect who walks away

Shane Phillips 52:25
That's what Frank Gehry would say too

Gabriel Ahlfeldt 52:38
Although I mean, I suspect that the planner would rather argue that they're creating a very nice visual.

Paavo Monkkonen 52:44
But Evan, don't you have a paper showing that beautiful architecture is valuable to people?

Gabriel Ahlfeldt 52:48
No, it is. I think the planner here really has probably the right intentions, right. I mean, they're trying to do something good. I don't think we have understood the net effects of this policy. So I wouldn't be in a position to really tell you whether on that it's good or bad. Right. But I would argue that probably it's not that simple that they're simply creating a means to make architects rich, right. I mean, there's also another objective that is going on here, right which one then actually dominates, right? I mean, the deadweight loss or the positive externality, I mean, I don't know, I haven't done the math,

Shane Phillips 53:23
Sticking with the subject of height, and the role of planners in this, your paper concludes with some discussion of height limits, which obviously have a big impact on the places that skyscrapers can be built, and what form they take. If you can't build a skyscraper as tall as you'd like, because of height limits, it follows that the floor space will probably be built somewhere else, maybe further from the city center and in smaller buildings. But your model has quite a bit more to say than just that. I think that's a basic observation. And I feel like the different impacts on commercial versus residential buildings were particularly interesting. Could you just say a bit about how your model treats height limits, and what restrictions on vertical growth mean for how city sprawl out horizontally or what you find in your model on that question?

Gabriel Ahlfeldt 54:11
Yeah, this is an exciting question. I could talk about that for hours right. And I suspect I mean, judging from what Paavo had as a discussion with his colleague, Michael, he would like to join in the discussion I mean, let me see how to keep this reasonably brief. So the way I think about that is that most people in the literature, they have used the closed city model, to think about the problem, okay. You take a closed city model, that means basically the population of the city is fixed, no in and out of migration, okay, now you introduce the building height. Okay, what happens of course, you reduce the supply of space in the city center, okay, that crowds out demand to the suburbs okay. You're probably going to convert I mean, some agriculture land into urban land. As a result of that, you have a city that is sprawling still, because you've constrained supply, you're going to end up with higher rents so you have a city, which is horizontally scrolling and is less affordable. And as a result of that most economists I guess, including Paavo would argue, it's actually good idea to relax height regulations, build taller because we get more compact, and more affordable cities. Now, that's all fine as long as you believe that actually, the population is fixed, right? The predictions change, they become a little bit more nuance once you assume the Open City model. Let's just think it through I mean, in this paper, we have an open city model. So in the Open City model, if I now introduce a height cap, okay, I'm going to make I'm going to restrict supply, I'm going to make the city less affordable or expensive. As a result of that I'm going to have people leaving the city, or you can also think about fewer people moving into the city, if you think about California, okay. Yeah, we're both right. So the city is now small in terms of the population relative to the counterfactual, and now it is entirely possible that you actually end up with a city that is more compact, less horizontally sprawling, because you have fewer people in the city, right so you get a very different prediction. I mean, the other way around to think about that is if I take the city and I relax the constraints, right, set rarely and make the city cheaper. But now the problem could be that a lot of people are just moving to California refilling the city, right, and then rent rate goes up again, and the cities as unaffordable as it was before, right. Potentially, if you have some people in the city who do not gain from the collaboration economies, they could even be worse off, right, because rents go up and wages don't go up right. But don't get me wrong, I'm not taking any stance here at this point I mean, which of the two models is correct, I'm just saying that it's something we need to think about carefully right? I mean, the standard predictions that we get, right, they they make sense, if you believe that the population is very mobile. If you suspect, however, that people in the US are very mobile, especially when it comes to California, right. I mean, you could be in a world where the only reason why they don't move to California is because California is just very expensive. And once you make it a little bit cheaper, more people come right, then the predictions get a bit more ambiguous. So what we really want to have, right, and that's, that's what I'm working on right now is you want to have, of course, a model with imperfect mobility, right? I mean, you want to want to think about okay, what is the true motivation and necessity, right? People are going to respond to immigration incentives but of course, they're not perfectly mobile, right? So you're probably not going to get utility equalization, you're going to have some adjustment, but this adjustment is imperfect, and then you end up with something in between the close city model and the open city model, and then he essentially depends on the migration and the society, whether you're going to see that as a result of high constraint the city's horizontally sprawling or not - I think. It's no longer clear.

Shane Phillips 58:04
The first city I thought of kind of in this model was San Jose, actually, just because it has, I looked this up beforehand, its tallest building is 298 feet tall. And it is a city of more than a million people and one of the two metro areas of the richest metro areas in the entire world

Paavo Monkkonen 58:22
with increasing GDP, I'd gather.

Shane Phillips 58:24
Yeah, I think the GDP is going up. You also took a historical look at how much rents go up as you move to higher floors at different points in time using a very detailed dataset that you had for New York City. I guess maybe it sounds like Jason collected for New York City, and I thought there were a few things worth mentioning here, too. One was that there was a sharp drop in the premium paid for higher floors after the September 11 attacks, which I don't think is surprising, but seems like good evidence that your data is reflecting real-world trends. The other thing that stood out to me is that the premium never really recovered from what it was before the 2000s. And that might be due to September 11. It's probably impossible to be sure. But you'll note that there was a lot more skyscraper construction starting around the turn of the century. And so this may just be reflecting a greater supply of skyscraper units or, you know, units and offices on higher floors. So they're not quite as scarce as they used to be. Especially given the super tall buildings, these sort of penciled in towers going up in Manhattan, which are well above the heights of the former Twin Towers, I think, and other famous skyscrapers going up all over that city and around the world. The supply and demand explanation seems more likely to me than the concern about future terrorist attacks or other kinds of fears. Tell me if any of that sounds wrong to you, though, and if you have any other ideas, what might be behind the change, but beyond that, what else should we know about how the appeal or value of upper floors, higher floors has changed over time.

Gabriel Ahlfeldt 1:00:03
I think you're spot on, not just with the fact that Jason collected the data yeah, so that's true but also with your interpretation of the results. I thought it would be a cool exercise to do this. Nobody really had done that before so it's the first time that had a time series of hate premium, we need to be a little bit careful with interpretation because it's just the first piece of evidence, and we'll see if that holds up for the people, and investigate that in broader samples. But I think it's fair to say that when I first saw the results, I was a little bit surprised. I know urban economics, we do have this idea that we are all getting smarter and richer and more creative. And as a result of that our amenity preference increases. And I thought I mean, "nice view, I mean, that's killing the amenity, right". I mean, if we have an increasing amenity preference, which makes sense, right, because of a view is purposely a luxury good. I mean, if you're very poor, you don't care about a view. But I mean, if you're sufficiently rich, you start spending money on that. So I thought that maybe over time, the high premium would actually increase. And then we found the opposite, which was surprising to me, and I think you're right, 911, that's one explanation. But there's kind of a sense that maybe that effect would be more temporary, right? I mean, if you have something that has more persistent effect, you want to have a slightly different explanation, then yes, I think if you believe that you have a distribution of preferences, for this height amenity some people like it a lot, some people like it, okay, but not so much right. Then if you have very few flats that offer that amenity, you pick the one with the highest willingness to pay, and you get a huge high premium, okay. And then as you increase supply, you go to moving down a distribution of the marginal willingness to pay eventually, you end up with a lower high premium. So that you should not do, I think, is to conclude the preference for the height amenity on average decreased over time, I don't think we can say that, right? I can only say that the marginal buyer now has a lower high preference. But this can be because the marginal buyer is now a very different person than it used to be 25 years ago, right.

Shane Phillips 1:02:18
I also kind of wonder there's, you know, the amenity of the view. But there's also the disamenity of just getting up to a home on an 80th floor or something like, that's got to be just kind of annoying, I don't know what to take a two-minute elevator ride or whatever. In the same way that like, you know, living in the Hollywood Hills is very glamorous, but you've also got to drive up the steep like, you know, one lane, windy roads that aren't all that pleasant so there is some cost there. And people are generally willing to pay that cost, but it's not nothing so I think that that's an important point as well,

Gabriel Ahlfeldt 1:02:53
Just adding to that, the cost of going up in the building that increases linearly but it's not so clear that the view amenity increases linearly. In fact, my co author Dan McMillan, who lived in Tbilisi, Chicago, he tells me from his own experience, he knows that there's a turning point around the 50th floor. After that, actually, the view gets worse because you can't really see in detail anymore so many what's happening is that the view that kind of average floor use with diminishing (returns)

Shane Phillips 1:03:25
Well, Gabriel Ahlfeldt, thank you so much for joining our show and talking with us about skyscrapers today.

Gabriel Ahlfeldt 1:03:31
It was a pleasure, thank you guys.

Paavo Monkkonen 1:03:32
Thanks.

Shane Phillips 1:03:38
You can read more of Gabriel and Jason Barr's research at our website. lewis.ucla.edu. Show Notes and a transcript of the interview are there too. UCLA Lewis Center is on Facebook and Twitter. I'm on Twitter @shanedphillips. And Paavo is there @Elpavo thanks for listening. We'll see you next time.

About the Guest Speaker(s)

Gabriel Ahlfeldt

Gabriel Ahlfeldt is Professor of Urban Economics and Land Development at the London School of Economics and Political Science.
Learn More

Suggested Episodes