The Lincoln Institute provides a variety of early- and mid-career fellowship opportunities for researchers. In this series, we follow up with our fellows to learn more about their work.
What can an 18th-century orphanage tell us about housing affordability today? Quite a lot, says Thies Lindenthal, professor of real estate finance at the University of Cambridge.
In Europe, many institutions, like orphanages and hospitals, historically held large portfolios of market-rate rental properties to help fund their primary operations, Lindenthal says, and some kept detailed financial records that span several centuries. That’s allowed Lindenthal and his colleagues to trace the long-term evolution of urban housing affordability by analyzing nearly half a million rent observations and other data from seven major European cities during a period of more than 500 years.
Their findings, detailed in a recent working paper, may sound surprising: In inflation-adjusted terms, real urban rents have increased only about 0.17 percent a year, on average, over the last half a millennium. “If you compare what you’re paying for a 50-square-meter place in London now to what you’d have paid, I don’t know, 100 years ago, it is not more expensive,” Lindenthal says. However, modern Europeans do pay higher rents in total—consuming more and better living space than people in centuries past.
In 2019, Lindenthal was awarded a David C. Lincoln Fellowship, which supports scholars and practitioners conducting new research on land value taxation and its applications.
In this interview, which has been edited for length and clarity, Lindenthal elaborates on his centuries-spanning real estate research, reflects on what artificial intelligence might mean for academic researchers, and reveals the actual reasons so many people tend to like historical housing (hint: it’s not really about the architecture).
JON GOREY: What is the general focus of your work?
THIES LINDENTHAL: I’m working on real estate finance topics, and mostly I’m looking into the risk-return profile of real estate as an investment. So at the asset level, what are the returns that people have achieved, what are the returns they believe they achieve, what are the returns they should achieve? And then trying to measure properly what kind of risks they are exposed to and what kind of risk-adjusted returns that they actually got in the end.
JG: What are you working on now, or hoping to work on next?
TL: I’m now in week four of my experiment to do my job with [Anthropic’s AI tool] Claude. It is quite a ride. I’ve always been interested in using big data, or biggish data, and machine learning to answer the types of questions that we work on in our field. But I think we’re now in the next step, where the machines can actually help us not just do individual tasks, but also tie them together—anything from data management to building up empirical pipelines to running tasks to interpreting tasks to doing quality control to a degree, and then, in the end, even producing reports or presentations based on that.
And that is quite a change. I mean, for academics, the stuff that we’re good at—not everything, but a lot of these things—Claude is extremely good at. So we have to really think about what it is that we can add to the mix.
JG: You’ve tracked certain real estate data across hundreds of years. Can you talk about how you’re able to do that—what kind of records exist from 400 years ago?—and what these very long-term historical property trends can tell us?
TL: We go back to the archives and find rents, and find prices, and sometimes you can find costs for portfolios from institutional investors across Western Europe. So we have good data from the Netherlands, from Belgium, from France, and somewhat okay data from England. These are records from what you’d call institutional investors—hospitals, orphanages—people who use real estate to generate income to then spend that money on running a hospital or something like that. They’re reliant on these very large portfolios of properties across a number of cities—Amsterdam, Brussels, Paris, London. And from these records, we can see how real estate, as an investment, is linked to other investment classes—for instance, how it’s linked to government bonds—and we can see how rents and prices are in equilibrium over the long run.
One thing that I find so important about these long-run data is that we can observe cities that are relatively free, so there’s very little interference from any type of regulation. You can see that in the long run, if rents and prices can adjust in a free, supply-and-demand kind of way, they don’t go through the roof, they are just nicely coming back to fundamentals.
And the other thing that’s interesting is that, if you actually start to try out policies that aim to improve affordability, for instance, or try to increase the quality of the housing stock, then that can actually work. So we’ve seen free markets, where stuff is not going through the roof. But then also we see the 1930s, 1940s, ’50s and ’60s in Europe, where you see a lot of different things being tried out in these cities, and houses get cheaper, houses get better, people live in better places.
Affordability, depending on how you look at it, might get worse, because people pay more—but also they live in much better quarters.… If you’re tracking affordability across the centuries, you’ll see that the price per square foot of space in a city has not become more expensive; on the contrary, it has become a lot cheaper.
But we don’t live in the same way as people lived 100 years ago—and luckily so. We have running water, we have more space, it’s better, it’s bigger, it’s healthier. A lot of progress has been made there. We are demanding a lot more housing services. Stuff got cheaper, but we consume a lot more of it. So overall, we pay more, but the fact that stuff has become expensive is also just reflecting the fact that stuff has gotten better.
JG: What’s one thing you wish more people understood about real estate finance?
TL: The number one insight is that the returns you get for real estate investments are realistic returns that represent the risk. You’re not getting something that is exceptionally good, that is better than other investments, no—but also nothing that is worse. This idea that you have to get into real estate as an investment, because it will make you rich? No. That’s simply not true. Some people are lucky, and some people are not lucky, but on average, you get a fair return.
JG: Have you encountered anything surprising or counterintuitive in your research?
TL: Maybe eight years back, there was this concept in the UK that the best way to create more supply of housing, to overcome the NIMBYism and the deadlock of supply not coming through, would be to just build in historical styles—to just say, you know what, we won’t build this modern stuff anymore, no more glass, no more steel, we’re going to build cozy, Victorian-style terrace houses. And if we do that, we build more beautifully, and everything will fall into place. And that was a bit of wishful thinking.
We did a study here, looking at the transaction prices that we see in the market, trying to account for the fact that the [older] buildings are in nicer locations and have bigger gardens and more greenery around them, and are better quality in terms of materials and so on. And that’s why they achieve higher prices. And the interesting thing was that, in a solid and data-driven way, accounting for these quality differences, we saw no premium at all for old architecture—not for your own property, but also no premium because your neighbors have a certain architecture.
So yes, green space matters. Yes, the quality of the materials matters. But if a house looks ’60s or ’80s or 1860s, that was not a big driver. I’m a bit of an architecture snob, I think good architecture matters. But it is more complicated than just saying, it has to look Victorian style or something. There’s still value in architecture, but I think that a good chunk of the home buyers don’t care.
What really surprised me is a different experiment that we ran, where we showed 2,000 pictures of houses from around the world to people in an app, and they could like or dislike each house. It was a bit like Tinder for houses: They could swipe left, or swipe right. We took the responses to these images and trained machine learning models to capture the aesthetic preferences of people. I was expecting to see some form of cluster [based on demographics or geography], buildings that everybody dislikes, or buildings that everybody likes, and we found surprisingly few of those. That was something I found really, really surprising. People like greenery, so that was across the board. People like space, people dislike density, and so on. So there’s stuff that showed up across the board, but it was not so much about the properties themselves, it was more about the setting in which these properties were located.
JG: What’s the best book you’ve read lately?
TL: One of the books that I really enjoyed was Careless People [A Cautionary Tale of Power, Greed, and Lost Idealism by Sarah Wynn-Williams], that is a really good book.
JG: When it comes to your work, what keeps you up at night? And what gives you hope?
TL: Well, it’s not very creative, but the AI revolution keeps me awake, and gives me hope. I think it’ll fundamentally change what we do, how we live; it will change society.
The problem is, you have the same people who are described in Careless People at Facebook, I presume that the same kind of people are running Open AI, and that is not good news. It’s the same kind of unchecked power. I mean, if you realize the power that is coming out of the big AI companies, and then read Careless People, that will keep you awake at night.
Jon Gorey is a staff writer at the Lincoln Institute of Land Policy.
Lead image: Thies Lindenthal, professor of real estate finance at the University of Cambridge. Credit: Courtesy photo.
