Bill Kerr: What is a thriving city and what have we learned from Covid-19 about building resilient, healthier urban centers? These questions are especially pressing amid rising volatility, living costs, and inequality.

Welcome to the Managing the Future of Work podcast from Harvard Business School. I’m your host, Bill Kerr. I’m joined by Harvard’s Ed Glaeser, leading urban economist and a researcher I’ve had the pleasure with of collaborating over the years. Ed’s pioneering work has redefined how we understand cities as engines of innovation, growth, and upward mobility. His influential book, Triumph of the City, explains the dynamism of urban areas. Survival of the City, co-authored with David Cutler, shows how crises can spur transformation. We’ll explore the after-effects of Covid-19, including the rise of remote work. We’ll also look at the policy picture from land use, climate, and transportation to workforce development, and we’ll consider the role of AI in the urban context. Ed, welcome back to the podcast.

Edward Glaeser: Thank you for having me back on.

Kerr: Ed, you have long studied cities and their importance and functioning. Can you give me a two-minute snapshot of the intellectual journey you’ve been on?

Glaeser: Sure. So I think when you think about cities, it’s useful to think about a three-legged stool, the first of which is the things that make cities magical, the interactions that occur in cities, the things that Alfred Marshall, the great English economist, was writing about in the 19th century when he wrote that “in dense clusters the mysteries of the trade become no mystery but are as it were in the air.” And a lot of my first 10 years of working on cities was about exploring this, was about exploring agglomeration economies, the fact that we become more productive when we’re enmeshed in a maelstrom of economic activity. When do they show up? What makes them more powerful? How does learning in cities work? Do we learn from our peers? What are the downsides of not being integrated with the heart of the city’s economy? The second leg of the stool is the downsides of city life, the demons that come with density, the fact that when two people are close enough to share an idea face to face, they’re also close enough to share a virus or bacteria—the fact that when someone is close enough to sell you a newspaper, they’re close enough to rob you. And dealing with understanding how to mitigate those downsides is a huge challenge. And particularly when we think about the cities in the developing world, taming those demons is really the first-order concern. And then the third part of the city is the physical city, the infrastructure. Now as social scientists, I think we should never tire of reminding people that the real city is made of flesh and blood, not concrete and glass, and that it’s really important to make sure that, whatever you do in the city, it’s built for humanity’s needs. But let’s face it, the buildings are important, the roads are important, managing that space is important. And if there is one theme of my work on that, it is the downsides of over-regulating the creation of new urban space. And so those three legs have been what I’ve been obsessed with for the last 35 years, and I think I’ll spend the rest of my life being obsessed with them as well.

Kerr: That’s great. Maps well into the things we want to talk about. I’m going to start maybe with the second leg there, which is that cities can transfer things that are less desirable. And when you joined us in 2020, we were in the midst of the pandemic, and you were talking about some of the early handling of the Covid-19 virus and what cities were doing. What’s your assessment of how cities, in particular, performed during the pandemic?

Glaeser: Well, I’ll give you two facts. The first of which is the incredibly strong correlation between metropolitan area education and Covid death rates. So when you compare the least educated America’s metropolitan areas with the most, there’s a fourfold difference in death rates from Covid-19, which is really extraordinary. As opposed to at the most educated areas—the Seattles, the San Franciscos—one in 1,000; in the less educated areas, four in 1,000. Really huge differences. Some combination of that is both the bodily weakness of less educated people in many cases going into this, and also the fact that less educated places had less stringent lockdowns, had less of a switch to work from home. But I did not feel then—nor do I feel now—confident that what the right answer was how much lockdown there should be or how much work from home there should be. These are difficult questions. But the downside of the lockdown really shows up in which cities have thrived over the five years —from 2019 to 2024. So for the paperback version of our book—and I’ve been updating it ever since we made a “50 Largest Metro Areas” list and then we created a ranking of which of these cities did well or poorly. W—we did this on the basis of housing price growth, permit growth, employment growth, and wage growth. And it’s clear: The Sun Belt rocks this list. Now, as I have often argued, there’s no variable that better predicts metropolitan area growth in the 20th century than January temperature for many reasons. They tend to be pro-business, they tend to be pro-housing, and they tend to be anti-ice in January. All of these things happen to come together, but over the last five years it’s on steroids. And the sort of Austin, I mean, the educated Sun Belt places are particularly strong. I think if you have a two-factor model of economic growth, which is we like human capital and we like relatively pro-business policies. Those are things the American Sun Belt does well, and they seem to have particularly matter during this time period. Interestingly in the coastal areas, areas that we would’ve thought of as regional champions like San Francisco, have done particularly poorly, in part because their housing prices just have not gone up during this time period. Their permitting is, they didn’t start with much permitting, they have zero now. And so the fact that education was particularly related to work from home, I think, has been really hard for them to get through.

Kerr: The other thing that was prominent during the pandemic was essential workers, frontline workers, and how they were performing. Five years on, is there any evidence of differences across cities or permanently changing attitudes toward that workforce?

Glaeser: Well, the wages are way up. I mean, leisure and hospitality wages look great relative to five years ago. I don’t think that’s an easy thing to do with the mortality statistics, but I would be shocked if you did not see more deaths in the frontline workers. But they seem to have been compensated by higher wages, and I think probably those wages are going to be sticky for the foreseeable future and they’re going to be higher than they have been.

Kerr: Well certainly the demographic changes that we’re experiencing as a country—and looking ahead with the current administration, lower rates of immigration—there should be a lot of labor supply pressure that will keep those wages up.

Glaeser: Absolutely.

Kerr: So let’s just continue on the pandemic. A lot of us, and you mentioned this, started working from home more. In early 2025, we’re seeing a concerted pushback against that. It’s obviously happening live right now at the federal level with the federal government being ordered to be back in the office. But many private employers are also pushing hard toward return-to-office policies. How’s that affecting cities and suburbs? And maybe one way of thinking is how far do we go in one direction as a consequence of the pandemic? And is this reversion going to help Manhattan have its vibrant future?

Glaeser: So nationally we’ve been about—from the current population survey—we’ve been about 10 percent full-time work from home for the past two or three years. I don’t think that’s likely to impact the Manhattan towers very much. I think that the larger issue is the people who are coming in one or two days a week. Those numbers are about 12 percent—and far from declining, they’ve actually been increasing slightly over the past two to three years. Nationwide, about 16 percent of hours are remote. And everything I’ve seen, I mean, the Kastle technology stuff—which views the people entering into the fancy office buildings in downtown areas by looking at their various cards or other things that they use—that’s stayed at 50 percent down, relative to the pre-pandemic level. And it’s hard not to see how that’s an enduring challenge for these big cities. In part, we’re reminded the cost of the single-type zoning, creating a very large office park in the center of your city, it makes it particularly hard to commute to it. A three-block commute is very hard in Manhattan. I think it’s much less severe in the Sun Belt, just much less of a deal. And obviously, in suburban areas, it’s much less of a deal. And I think we now have a lot of evidence that there are costs from working from home, which need to be offset against the benefits. I like a lot the work of my colleague Mandy Pallais—I like her work with Natalia Emanuel and Emma Harrington, I like their work on their own. I think one thing that we learned from this is, while there can be static advantages from working from home, there are dynamic costs. And a headline fact of Emanuel-Harrington, which is also in the great Nick Bloom’s early work on this, is that even though call-center workers, when you send them home, they can make as many calls, their probability of being promoted drops by 50 percent, presumably because they’re not learning as much or the boss isn’t learning about them.

Kerr: So there’s a lot in there, and I think first I’d highlight and agree with the differences across locations and cities. Whether your headquarters was in Dallas versus Manhattan is going to play a lot toward the future from there. But let’s continue on where you ended. So maybe elaborate a little bit more on the role of the workplace and face-to-face interactions for someone being able to learn the job, learn the skillsets that are necessary for the future.

Glaeser: Well, I see this as having a couple of different strands. So the work of Harrington, Emanuel, Pallais, they’re able to look at very particular forms of skills and sharing, and they’re able to look at how this functions. So the Pallais work looks at computer programmers who help each other, and they’re able to look at the exact help that they give, and they’re able to show that people who aren’t in the same building don’t help each other, and they’re able to show that women particularly lose. And this is amazing work. Now, of course, there’s also a bunch of work which just looks at who do you work with in the same office and see whether or not that has an impact. But in some sense, the background for that work is that we have studied the segregation of the neighborhood for 60 years, and that segregation may be particularly important for kids whose lives are determined by that neighborhood, but it’s much less so for adults. It’s about the segregation of skills, rather than the segregation of race. The early work on neighborhood was all about race, partially because that’s the data that was available. But our literature is all about, are unskilled workers working by themselves or working for people that they can interact with? This is part of I think understanding a deep puzzle of cities, which is, we still think cities are pretty good for lower-income adults who come and find their way there, that they provide pathways out of poverty, they provide lots of opportunity. But we have now a large body of evidence suggesting that it’s not really great for kids. And I think of, particularly, our colleague, Raj Chetty’s work where he really finds that dense urban areas are really bad for upward mobility for kids, particularly going to a big city school district that’s bad, living in a denser part of the city that’s bad. I think one potential explanation for that is just the lives of kids are just segregated in a way that’s very different. So if you’re an adult who wakes up, sure you’re living in a segregated neighborhood, but you’re getting out of the subway, you’re going into a midtown tower, you’re experiencing more integrated life. And that appears to be true with my work with Cody Cook and Lindsay Currier, we’ve tried to identify kids, versus adults from their cell phone records. And the kids do seem to live slightly more segregated lives. Our work, yours and mine, and my earlier work with a series of co-authors looking at Brazil, looks at what happens to people who come to Brazilian cities, particularly people who start in poverty. And we found that there was plenty of upward mobility in the south and very little in the north. And in a purely statistical sense, that gap can be almost entirely explained by the level of segregation at the firm or establishment level.

Kerr: And that’s a big question going forward, because if we think about firms trying to move to some of their comparative advantages or what makes them very competitive, they’re increasingly outsourcing various functions to other vendors. And that often comes with some skill segregation that’s happening in the workplace.

Glaeser: And I think one thing we’re going to get into AI later, but there was a point that was raised to me by a friend two weeks ago, which is, what if AI kills the apprenticeship models, that’s so common in many of our skilled jobs? So think about a young attorney at a fancy law firm who spends the first 10 years of her life redlining and proofreading and doing a whole bunch of crappy stuff, but she picks up by osmosis what it’s like to be a great lawyer, and she becomes a great lawyer. What if AI gets rid of those? Now my own guess is the firm will internalize that, because her wages will go down, she’ll pay for the privilege of being around, and they’ll figure out something that human being can do to make that happen. But making sure we keep the skill growth pathways, they’re such a critical part of being a knowledge economy going, seems like a really important part of the future of work.

Kerr: Do you have a view on that? Well, not going all the way to AI yet, but just the capacity of using digital labor platforms? So I may not get the job at the prestigious law firm, but perhaps there is a job on Upwork that I can work on that is connected to a law firm, and I start to maybe interact with the company that way. Is that at all a substitute for that mentorship or that proximity?

Glaeser: I think it is, but between nothing and the full mentorship, I think it’s closer to nothing than it is to the full mentorship—being able to sit next to the partner every day for eight hours a day.

Kerr: Yeah, I’m sure that if we’ve rank ordered occupations, there’s some spectrum at which that physical close proximity—I think we would believe it’s true for academics—that it is really important, and other places—maybe digital marketing—it’s something that could be more arm’s length from each other, definitely. So let’s go back to New York City. And you’ve mentioned, I think, everyone’s—beyond the pandemic—No. 1 bogey about cities, which is the commuting times and costs. New York has just implemented congestion pricing, which I believe you’ve advocated for in the past. And so is there a sense of how this particular implementation is going or what we’re learning about congestion pricing from New York but also other cities? I believe London’s done it and elsewhere. Is this the part of the answer toward the future?

Glaeser: So, Emily Oster has two amazing students who put together a web page that actually tracked live traffic, relative to a previous time period, and just showed the increase in speeds, the decline in traffic times, particularly—say, for example, on the Holland Tunnel getting in the morning. So it does really appear to be having an impact. We saw that earlier in London 20 years ago. It’s great that they’ve done it. As economists, it’s hard not to cheer for pricing and for charging for the social cost of one’s actions. And I think it’s sensible. The thing is that you do have to weigh this against the operating costs. You need to have a system that’s actually having a large enough impact to justify that kind of a fixed cost. My primary worry is that the administrative costs are not completely out of control.

Kerr: What other policies—if you had the ear of a mayor of a, call it, 2 million population kind of level city—would you be talking about for workforce development and needs, particularly in the post pandemic era? What are the issues that they need to keep their eye on or places where somebody’s doing something innovative that they should mimic?

Glaeser: Look, I’m all about vocational training in lots of ways. I don’t see it as a substitute for traditional K–12. I think it should wrap around. I think businesses should be at the table. I think we should be experimenting with it. Part of the reason I love vocational training is you can pay for performance, because you actually know if someone’s learned to be a plumber or an electrician or a coder by the time they’re done. And so you can have free competition without that much worry. You can compete, you can provide the thing, but you’re not going to get paid if you don’t teach the kids. What exactly are the skills that should be taught? Again, the local business community should be involved. I’m not ruling out older people, but I think certainly the high school population is critical. In some places, we can look to existing community college programs or programs that have been put together with different businesses. Your colleague, Jan Rivkin, for example, has these great examples of Siemens working with places in the south to train workers.

Kerr: All right. So now I’m going to start a stopwatch here, because I know you could go for the better part of three days when we start opening up the housing, land use conversation. But maybe begin with a little bit with something that, again, you and I have been working on: why the construction sector, when you look at any measure of the output per employee, has not kept pace with automobile manufacturing or other sectors. What’s the hypothesis set that you’re working with?

Glaeser: So our hypothesis is that what land use regulation has done is, it has artificially shrunk the size of projects. As the size of projects has shrunken, the firm size has also shrunken. And you can certainly see that in the data—that manufacturing firms are huge; construction firms, particularly in the area of residential housing, particularly in single family residential, it’s tiny. Tiny firms don’t have R&D labs. They don’t have massive investments in patents and new technology. And so you have this very simple way of making housing that’s continued along for decades and where the rest of the American economy is... I mean, look at Nvidia. There’s nothing in 1980 that looks like Nvidia, but the guys who made houses in 1980 kind of looked like the guys who make houses in 2022. And it’s this very same static thing that’s a very sad thing to see.

Kerr: And in the automobile sector, roughly 2 percent of the employment is in companies that have less than five employees, but in construction, it’s 40 percent. And do we have a sense of what is the economic cost that that is bringing about? Or how closely can we pin it down on… Could we look like the automobile sector or Nvidia if we made some changes here?

Glaeser: So we went through an exercise where we asked what happens if you just grow the firms and you just take some fraction of the relationship between productivity and size as being causal. And we found that there were really quite sizable increases in productivity that would come just from having larger firms. And that was a static calculation. That wasn’t even asking, what would happen over the next 30 years if you had large firms that were actually paying for the cost of an R&D lab by expecting to use that R&D lab on thousands and thousands of units.

Kerr: And I think something on the order of $1 trillion every five years of output is being lost because of the challenges we have in building things in the country.

Glaeser: Absolutely. Government regulation influences the cost of building as well.

Kerr: So let’s go in the paper and in the work there’s this remarkable graph that shows that the productivity in the manufacturing sector and in housing actually tracked very closely from 1900 up until about 1970, and then it starts to diverge significantly. And you can look at it in patents and innovation and the like. But I think sometimes it might be helpful, concretely, what would it look like if it was different? And there’s Levittown and some of these examples of prefabricated elements, how would this innovation take root, and what could a policymaker try to do locally to maybe help a little bit in their space?

Glaeser: The Levitts were essentially mass-producing housing in post-war America, using Henry Ford’s techniques to housing. Levitt himself became disillusioned with that. By the 1960s, he thought the only solution was actually mass production in a factory. And I think that’s probably the more feasible answer. You can even do that for high-rise. Marc Vlessing at Pocket Living, they’ve done prefab high-rises that are really pretty neat. I think we’d be thinking about modular. You produce it in a factory, you produce a lot of them in a factory, and then you have a certain amount of customization. And then we just let large companies figure out how to make this stuff cheaply—with very little labor, a lot of capital. I mean, we could just build a ton of housing for very little. The question is how do you get to it from policy? This is different, it’s a regulation of process. And it means that it’s not just that you have to satisfy some broad regulatory entry cost, but every project you do, there are a whole bunch of rules that you have to pay on that particular project. And that just is a totally different ballgame.

Kerr: And I was thinking, the person who built our home in Lexington only operates in Lexington. Won’t even go one town over that’s two miles away because all those processes and the boards and everything become different.

Glaeser: Absolutely, absolutely.

Kerr: In the U.S., do we have any places where the state has just set it up in a consistent way or anything that we can appeal to that would be just, “Let’s at least allow someone to scale to the level of Massachusetts, if not all the way across the country.”

Glaeser: So Texas is easier. Texas, the firms are bigger, the projects are bigger, it’s clearly easier to get permitting. Houston famously has no Euclidean zoning, and it’s probably the best you’re going to get within the U.S. Overall, in the Sun Belt, I think we’ve seen many Sun Belt areas, Miami, Phoenix, become increasingly like California and Massachusetts over the last 40 years. And you’ve seen a slowdown particularly in the most desirable areas. So our co-author Joe Gyourko and I have been working on a Brookings paper on this, and the density effects really seem kind of small. We used to be demand-driven and now we’re supply-driven in terms of this. And I think the right answer is something like the state level. I had a great conversation with the chairman of the Illinois Housing Authority, who was proposing doing mass-produced housing for public housing in Chicago, which I think was a fantastic idea. But I threw out the idea of, what about having statewide zoning for certain prefab houses? It doesn’t require any local zoning. That would be an amazing thing in terms of just supercharging the prefab housing business and just making it just much easier for everyone to build.

Kerr: That’s interesting. Very interesting. So we evoke Texas here, and I think the bogeyman that many would have about reducing land use regulations is that we’re going to have sprawl, unmitigated sprawl. And sometimes you challenge my view of words and issues. First is “sprawl” a bad thing? But second off, how much do you think this would check or be an issue that we would find if we had more state-level work?

Glaeser: So, is sprawl a bad thing? So I tend to think carbon emissions are a bad thing, and there are aspects of sprawl which are related to carbon emissions. I also think protecting certain green spaces is a good thing. And sometimes sprawl can eliminate that as well. But the right thing to do is to not artificially handicap in really crucial industry like housing in order to reduce carbon emissions and protect green spaces. Reduce carbon emissions, ideally, with a carbon tax or if not with some other second-best policy that’s directly aimed at it, and by, say, for example, not artificially subsidizing highways. So it subsets the fact that we’ve been using general tax revenues to pay for America’s highways in a big way since 2005 instead of using the gas tax, that’s a big problem, and we really shouldn’t be doing it. And there’s lots of things that are good about sprawl—the fact that ordinary Americans find houses at a reasonable price, there’s a lot to love about that and I think we shouldn’t just think that fighting sprawl is worth any price. So I think these things can coexist.

Kerr: Yeah, I remember once you told me something along the lines of, “If you care about the nature and the environment and green spaces, move to the city.”

Glaeser: That’s right. And I believe very strongly there are huge environmental advantages of allowing building in core urban areas, precisely because you don’t do that pressure.

Kerr: So as we’re recording this, Los Angeles, has been just ravaged by fires. And, of course, there’s many cities that are exposed to the implications of climate change and so forth. Maybe with the particular of L.A., are there any thoughts as to how to approach that rebuilding process? And just more broadly, how much has the climate mitigation impacts been elevated to the forefront of the mayor’s office?

Glaeser: I think it’s got to be pretty central. I mean, climate is clearly a thing in which cities are going to have to deal with, whether or not it’s flooding or fires, in a big way. And that’s primarily about adaptation rather than about mitigation. I mean, we’ve been thinking about mitigation for a while, but how is it that you actually protect these cities just physically from this happening? I feel fairly confident that L.A. is going to have that conversation and do sensible things. What I’m not so sure about is that they’re going to take this as an opportunity to rethink their density levels, because L.A. really is way too low density relative to the level of demand for that place. And a denser and more compact city, quite honestly, will be easier to protect. So you build up a little bit more in the center, you create more areas for barriers, be easier to make safe.

Kerr: Yeah. Demographically, we’re going to continue to grow our rural population for the next 50 years, but virtually all of that growth is going to happen in developing countries. And so the question of the cities that have to be available to absorb that population—I don’t know that anyone has fully grappled with how to approach that challenge—but where are we in that frontline of the cities for the developing emerging world and going forward?

Glaeser: So I think as in everything, we’re thinking about this combination of infrastructure, institutions, and incentives. So, for example, in dealing with disease, some part of this involves getting clean water in and getting sewage out. That requires both functioning institutions to do that. You need to have an incentive to pay for a connection to the water system, because if the actual physical cost of connecting to the water main, even if a philanthropic organization has built the water main, but if the physical connection costs $1,000 in a country with per capita GDP is below $2,000, they’re going to walk away from that. Just as New Yorkers did in the 1840s and 1850s after the Croton Aqueduct was built, and they weren’t willing to pay for the connection to that and they continue to die of cholera. And then we need institutions to properly manage this, to actually protect it. I mean, maintenance is one of the great challenges of the world and it’s not like we’re great on infrastructure maintenance in the U.S. either.

Kerr: Yeah. We talked obviously about the climate related challenges to big cities like Miami or L.A. You’ve also thought deeply about rural poverty. How do we approach struggling people, versus struggling places, and separating those two pieces? Where are we today in terms of place-based policies toward rural development and what would be the prescription that we should be thinking about right now?

Glaeser: So one of the downsides of helping poor places is you encourage people to stay in poor places. Another downside is maybe you’re going to actually raise the price of living in those poor places, and that’s going to be a downside for them. And there are lots of upsides of people moving or finding other ways to make things work. I have morphed on this partially because of the rise of prime age joblessness, right? We’ve gone from 5 percent of prime age males being jobless to 15 percent over the last 60 years. That’s led me with a view that we need to do something about the joblessness problem. And I don’t know a better way other than subsidizing employment—so earned income tax credit or maybe something that is better for targeting the particular impacted population seems to me to be the right structure. I think we need to be open to place-specific kinds of policies on this, and maybe place-specific policies around entitlement reform as well. You could think about, for example, reforming disability, let’s say, in West Virginia—we’re going to lower the base pay, but we’re going to say you can work more. The cutoff for how much you can work given that you’re receiving disability insurance goes up because we want to do more to encourage work in a weak labor market like West Virginia than we do in Seattle. Similarly, if you gave me a fixed pot of money of EITC money, I’d probably throw a bit more of it to West Virginia than I would to Seattle. And I’ll make the third point on minimum wages—and we talked about the really interesting work of our former student, Jeff Clemens, on minimum wages—is the impact of minimum wage is not some universal constant. If you put a minimum wage in a place like Seattle over the last, from 2010 to 2015, it’s not going to make much difference. This labor market is thriving, they’ll figure it out. If you do the equivalent high minimum wage in West Virginia, it has all sorts of potential for a catastrophe. And it’s just really important to think about your labor market interventions and how they’re going to relate to the local labor.

Kerr: And even at the state level, there can be urban-rural differences in the binding effect of minimum wages.

Glaeser: Absolutely.

Kerr: And it moves over time. Boom years is going to be less, once the recession hits, it could be more in certain places.

Glaeser: Absolutely, absolutely.

Kerr: And to your points also about, as you think about these employment policies, let’s just loop it back and try to think about the education effects and the spillovers that we could have for the next generation of workers and what might come about there. So we could not have this podcast without having some section or conversation about artificial intelligence, and that can go in many different directions. But let me just start with, what is the cutting edge of using AI—and just advanced technology more broadly —with respect to cities?

Glaeser: Well, so our friend and colleague Mitch Weiss is convinced that we can use AI in an amazing way to make city government much more functional. In general, city governments tend to be relatively reticent about new technologies. So I believe both that he’s right, that one could. I’m also not at all sure that city governments will be, given that it’s very hard to get their workers to change. And it would be really neat to use AI to think about making it easier to get permits or dealing with things like that. In the world of urban research, there you are seeing AI and you will see AI. So, for example, the work of Arpit Gupta and Alex Bartik, they are using AI to make sense of local zoning codes throughout the U.S. and to figure out how they work. With a whole set of co-authors, including Carlo Ratti, we’ve been using AI to see how people move around in 1980 and 2010 using movies that the great [William] “Holly” Whyte, a great urbanist from the ’60s and ’70s—actually the man who discovered Jane Jacobs—he made these movies of how people walk around in urban areas, hundreds of them.

Kerr: No way.

Glaeser: Yeah. And we have ones from the 2010s. And so we can show you the extent to which we now use AI to categorize how people are moving. And so we can see that there’s much less pairing up in these urban spaces than there was 30 years ago. There’s much less lingering than there was 30 years ago, especially Downtown Crossing. Downtown Crossing in Boston is where you see the largest change. The simplest model is just a busier time.

Kerr: It’s just social interaction.

Glaeser: Yeah. Totally, totally. We’ve talked about this last time, but I’ve been involved in this agenda of using smartphone data from Uber to measure all the road roughness everywhere in America. And it’s both tragic the way that most of the cities that we have, there’s no correlation between road roughness and subsequent repaving, but there could be. And that’s another way in which we could be targeting city services better using data that we get from lots of smart areas.

Kerr: And maybe just continue on the AI and technology theme. Many will have this question going forward: Are cities going to be as important for the recombination of ideas that happens in cities? Will the technology be so advanced and sophisticated towards that particular purpose that cities will be less important for new ideas' development?

Glaeser: I’m, as usual, inclined to not think so in part because all of these information technology innovations that we’ve had over the last 550 years, they’ve all tended to make face-to-face contact more, not less, valuable. AI will do amazing things. No one’s taking anything away from that. But the comparative advantage people will have, [they] will have from the extra thing they bring to AI. And that’s likely to come from human contact or some hybrid thing between humans and AI and so forth.

Kerr: So I guess we’ll end on the final question. It’s the trillion-dollar question. Cities in the future. Optimistic? We had the triumph of the city. Is it going to go from great to greater? Or what’s on your vision here?

Glaeser: So cities as a whole, you bet. So I continue to be as optimistic as I always have been and I’m not going to move on that. And I particularly think the things that we’re going to continue to see in the developing world despite climate change are just going to be amazing. There’s just so much as you’ve written about the gift of global talent, the global talent reservoir there is enormous, and it’s a gift to America when they come to America as immigrants, but they’re also a gift to their cities. Particular cities, it’s harder to know, right? There are cities that will screw up, there are cities that will be harmed. And it’s going to be a hard slog for the big urban office markets for the next five to 10 years; that they’re the wrong physical shape for what we need at this point in time. And it can take cities 50 years to work the way out of economic shock, a preponderance of the wrong type of structures. And so it will be tough for some cities, it will continue to be difficult. But, overall, it’s going to be an urban century and I’m looking forward to see the amazing things that will come out of that.

Kerr: Great. Ed, thanks so much for joining us today.

Glaeser: Thank you so much for having me on.

Kerr: We hope you enjoy the Managing the Future of Work podcast. If you haven’t already, please subscribe and rate the show wherever you get your podcasts. You can find out more about the Managing the Future of Work Project at our website hbs.edu/managingthefutureofwork. While you’re there, sign up for our newsletter.