Profile of Wharton Professor Gilles Duranton, Dean’s Chair in Real Estate
Professor Gilles Duranton is the Dean’s Chair in Real Estate at the Wharton School of the University of Pennsylvania.
- Could you briefly talk about your upbringing and educational experiences?
Like most people, I went to high school and did not know what I wanted to do when I grew up. I first went to a business school but did not like the prospect of working as a “salary-man” for the rest of my life. When I was soul searching while studying at a business school in France, I first thought I would become a historian, but I realized there was a big gap between thinking and reading about history and the actual job of being an historian and spending a big part of your day transcribing archival material. I then went to economics, and my quest stopped there because I liked it a lot. The irony is that now, 30 years later, I’m teaching in a business school despite my initial repudiation.
When I started studying economics back then, it was a very theory-based discipline in that most of the economics I learnt at the time was about writing down models. It didn’t have much empirical content, if any. I was interested in understanding the key forces that govern the working of cities and modelling them. At the time there was no such thing as urban economics as an established subfield of economics in Europe where I was. So with a few others like Diego Puga, Pierre-Philippe Combes and Yves Zenou, we got that started. Now, conferences on urban economics in Europe can attract 300 participants. I find this amazing and I am glad to see that many others are seeing what we saw then.
- What are your favorite or most intriguing courses to teach at Wharton?
Wharton has always been known for Finance, but traditional Finance is somewhat in decline. It’s cool that we’ve leveraged more broadly towards “analytics”. That’s really who we are today as an institution. That’s a great compromise between our “nerdiness” and what our students want. We’re trying to develop classes that are in-demand. For example, the big thing right now in real estate is proptech,property and technology, and you cannot go to an industry meeting without people talking about proptech. At the same time, as an academic these are all on-going developments and traditional channels for knowledge like academic conferences and publications in learned journals are lagging far behind. Then, the idea was to team up with an industry leader to develop the curriculum and co-teach a class. More generally, like many of my Wharton colleagues, I’m in touch with people who are pretty well-known in their industries and can provide a lot of insight to me and to our students about what’s happening “in the trenches”. That’s what I really enjoy in a business school. We can get input from really smart people from outside academia. If this sounds trivial, you probably have not spent enough time in academia.
- What do you enjoy most about your role as Co-Editor for the Journal of Urban Economics?
For science, academic journals are fundamental. Even if I’ve just been critical of them and said they lagged behind, they serve two fundamental functions. They vet and transfer knowledge. The fact that they sometimes lag behind the latest developments is actually a reassuring feature. Vetting knowledge is a slow and time-consuming activity. Absorbing new knowledge is even slower and more time consuming. Sometimes we need the dust to settle before things start to make sense. As coeditor of the main journal in my field of investigation, I’d say I enjoy the sense of duty the most, as it’s an important obligation in the profession. Vetting new knowledge is also about providing the individuals producing it with a quality stamp which helps them in their career. We’re talking about a community of perhaps 20,000 or 30,000 academic economists which mostly relies on this mechanism in the way it works, day to day. Editing is not something you do alone. You work with referees, other co-editors, journal board members, etc. This collective system is of course imperfect but relying on peer recognition is much better and fairer in my view than the judgement of one or two senior colleagues in the same institution or the whims of a dean…
Editing a journal is also important for me to stay informed about the newest advancements from people doing research, and that can filter into what I teach my students. I also understand the responsibility of being sometimes able to make a big difference in someone’s research and career. I wouldn’t claim complete responsibility for these researchers’ growth, but I hope my guidance has helped them in a meaningful way.
- How would you distinguish theoretical and empirical economics?
For a long time, the reasoning and logic of academic economists was not nearly as informed by data as it is today. But there is not as much opposition between theory and data as many would like. Actually, these are complementary activities. Data helps theory, at least theory of the more applied sort that tries to explain certain phenomena. In practice, when you develop a model, you make a claim about the world. This model needs to be disciplined by data to avoid speculating uselessly. This model, when it’s insightful, will also propose new conjectures to bring to the data and suggest novel connections.
There’s also pure theory, which is about developing tools to think about things. That’s the vocabulary of the discipline if you want. The prisoner’s dilemma is a good example. Here you face a simple situation where two players, both doing the best for themselves, end up in a situation that is worse for both of them when they act independently. There may be a good cooperative outcome to a situation (a “game”) but both players often have an incentive to be individually uncooperative and this leads to this suboptimal outcome for them. This is both trivial at some level and deeply puzzling. The rigorous study of these situations, a subfield of known game theory, has led to many insights and what seems obvious today escaped some of the greatest minds of the 20th century. Before John Nash, portrayed in the Oscar winning movie “A Beautiful Mind” set us straight on this, one of the intellectual giants of his times in John von Neuman was essentially talking nonsense about this. We also need that language from pure theory. A concept like “subgame perfect bayesian equilibrium” sounds awful to the non-initiated (or even to the initiated, believe me) but it’s part of our tool box.
This said, our discipline needs some balance. It was clearly theoretical when I was doing my PhD in the 1990s. The first big shock with the advent of public, administrative data. It brought about a big evolution in economics. These data were mostly about the labor market telling us about workers and their wages. Labor economists seized on that for the benefits of the entire profession. Of course, this is not only about the data. Meaningful statements in economics must often be causal statements. Establishing causality is hard, and we need to go beyond data. So the advent of administrative data led to some serious thinking about how to establish causality. For instance, we observe that more educated workers make higher wages to the tune of perhaps 10% higher wages for each extra year of education. Interesting, but the quantity we are really interested in is the returns to education, not this correlation. Maybe more educated workers are just smarter or more diligent at work. That allowed them to stay longer in school and makes them more productive workers today. To assess returns to education we may want to look at reforms that made pupils stay longer in school, like the lengthening of compulsory education, and perhaps compare students born in September who might have left earlier and had to stay one more year in school. Alternatively, some pupils may differ from others only because they live at different distances to school. Then, facing different costs of reaching school, those that live closer may be induced to stay longer. Of course, the key here is on the word “only”. If they differ in other dimensions, we can no longer make causal statements out of such comparison. The two examples I just gave are classic examples from labor literature. They’re actually not perfect because we subsequently discovered some flaws but they greatly helped us become better thinkers about causality. This is a great thing about science. Your average PhD student today is way more sophisticated on some issues than the entire profession was 30 years ago when I started learning about economics.
Then, for people like me interested in urban development, transportation, and housing, in the 2000s, a lot more data became available through satellites. In the past decade, the internet and other technologies have drastically increased the amount of available data even further. We have been pushed to update our theories and techniques to be able to analyze those data more effectively. We’ve been able to retrieve data for just about anything, such as people’s activities, whereabouts, and habits. Part of my current research is to track what’s happening to traffic conditions by collecting data from Google Maps. I sometimes joke that I’ve become the world’s main hacker of Google Maps. But this is useful stuff beyond the joke. Did you know, for instance, that the main source of differences in mobility across cities in the world is not how congested they are? Instead, most of the differences in travel speed across cities are explained by how fast you can travel in these cities at 3 am. Hence, urban mobility is mostly about the type of roads that are available, their quality, and the design of the road network. For many of us raised in the religion of “congestion is a big bad externality and we must have road pricing to combat it,” this is quite a surprise. Of course, congestion is bad – talk to drivers in Bogota or Manila – but there is a lot more to the problem of slow urban mobility. Urban travel in developing countries is two to three times slower than in the US and this seems to be mostly an infrastructure and urban design problem.
Then, the big question is how do we translate this research into our teaching? This is hard, particularly in a business school. The real challenge I have seen is that business students are not always intellectually minded; they are more interested in things for practical reasons, rather than just for the sake of knowing. That’s why they went to a business school instead of studying classics, after all. What makes an economist an economist is the ability to be able to see and predict things, to ask the tough questions, and to look at things in a systematic way. This does not always square well with the immediate demand of many students and we need to find a compromise where we are more than a newspaper or an industry journal article but still able to engage students.
- Fellow Wharton Professor Amit K. Gandhi spoke with Rebellion recently about the duality of his experiences at Microsoft and as a professor. Why have you chosen to stay in academia? How has your research intersected with the private sector? Public policy?
The relationship between the “real world” and the “ivory tower” is more complicated than what people often think. My two areas of research, cities and transportation, are closely related but have followed very different paths. Economists interested in cities chose to disconnect from the real world in the 1960s and develop their own intellectual tools independently from immediate social demands. Economists interested in transportation remained instead very much in contact with the questions coming from transportation planners and other policy makers.
The first time I went to a transportation conference, I had someone vehemently opposing what I had to say without using academic arguments, involving data and logic. I subsequently learnt that this person was a lobbyist. Given where I was coming from, this seemed unbelievable to me. So there is a tradeoff here. Economists who retreated from the world developed really sophisticated intellectual tools to think about what’s going on and analyze data but at the cost of a lesser relevance for many years. Those that remained closer to the social demand served society better in the short run but that came at the cost of intellectual progress. Traditional transportation economics is really good at answering a few immediate questions such as those about mode choice for transport or pricing issues but it ends up being narrow in what it does and lacks a deep appreciation that transportation affects urban development, people’s location choices, and more.
In my own case, nearly 15 years ago, I decided to become more involved with policy making, to spend more time with multilateral institutions and national governments in a bunch of countries. You need to keep a balance because demand for your time from the policy world might be infinite, especially if you’re not doing that to get rich. At the same time, I view this hands-on, application-based economics as part of my own learning. Policy-makers ask me interesting questions, and thinking about these issues that are most relevant to the public helps me eventually form ideas for interesting research projects. What I learned from the policy world was that there were some really big questions that were being debated, and then considering how we can be more systematic and scientific in our approach. Of course, this is not as simple as that. What policy makers often call “research” is really getting them some basic information that they lack. This is important but not always scientifically fruitful. The key here is the digestion of what the world is asking into fruitful avenues of investigation. I am certainly not claiming that I have it right or even that I have the right balance but I got a lot from interacting with people outside academia even if I’ve lost more policy fights than I’ve won, and by a large margin.
- In what ways have the COVID-19 pandemic and racial justice movements affected your research philosophies or ambitions?
In teaching, when you teach about the world like I do, COVID-19 is a big disruption so we must discuss it in the classroom. We had to reinvent classes; in some classes you may draw examples from the pandemic but it will not become the core of the class, while in others, the pandemic is such a huge event that it merits the central focus of the curriculum.
I’ve been impressed by all my colleagues stepping up to do economic research on the pandemic, and using the pandemic to improve our tools, techniques, and understandings. It’s a really honorable and important job to do. After a few months during the lockdown, economists had produced literally thousands of research papers trying understand, for instance, the role of transit in the propagation of COVID in New York City or whether COVID was more of a demand or a supply macro shock, which has big implications for the policies of the central bank and the government. This said, to me the central insight from economics during COVID, was to argue that people will react to what they see and hear, however imperfectly. Standard epidemiological models are fairly mechanical and assume a constant rate of propagation among the uninfected population. That’s the famous R0 at the origin of the early predictions that everyone will have been sick by May 2020. The disease instead propagates depending on how cautious people are as people react to what is happening. So the R0 is really an endogenous variable and disease propagation is not only caused by the characteristics of a virus. This is not about saying that economists are better than epidemiologists. Certainly not. However, taking these considerations into account seems like it has led to a much better understanding of the epidemic and perhaps better prediction models.
Profile of Wharton Professor Gilles Duranton, Dean’s Chair in Real Estate Written by Michael Ding & Edited by Benjamin Binday
Profile of Wharton Professor Gilles Duranton, Dean’s Chair in Real Estate
