The Mission of the Manhattan Institute is
to develop and disseminate new ideas that
foster greater economic choice and
individual responsibility.

Civic Bulletin
No. 40 March 2006

Making Cities Skilled

Edward L. Glaeser

Director, Taubman Center for State and Local Government; and Director, Rappaport Institute for Greater Boston John F. Kennedy School of Government, Harvard University

I’m a Senior Fellow at the Manhattan Institute and a contributing editor of City Journal.

Throughout history, cities that have been centers of great learning have usually also taken their place as economic powerhouses. Here in America, economists have long noticed that educated cities—that is, cities with a greater percentage of knowledgeable and skilled residents—have fared better economically than their unskilled counterparts, but the exact relationship between these forces hasn’t always been well understood. If anything, the debate over the role that skilled workers play in a city’s success has become more heated and a bit more confusing in recent years. That’s largely because a professor at Carnegie Mellon in Pittsburgh by the name of Richard Florida has dazzled the public policy world with his theory that skilled workers want to live in places that are hip, cool, bohemian, and gay-friendly, among other things.

To make cities an economic success—you’re laughing, but this is accepted wisdom right now—the good professor has told mayors and economic-development officials that they need to attract the creative class with fancy lifestyle amenities: for instance, by building bike paths, because, of course, we all know that skilled workers all bicycle; by hosting rock festivals, because the new generation of skilled workers see themselves as the rock stars of the business community; and by redeveloping their downtowns into trendy food and entertainment districts dotted by artist studios, independent bookstores, and live music venues, or what we used to call when I was in college, “bars with bands.”

Part of what has made these ideas such a success is that they have partially absolved public officials of the tough choices of governing. Professor Florida, for instance, tells mayors that they often don’t understand what the creative class really wants. For instance, mayors have spent so much time obsessing about the quality of their school systems when, it turns out, much of the creative class is single and really cares more about bars with bands. Mayors and governors also ought not to worry so much about things like high tax rates that are driving business away, since apparently, the creative class really doesn’t care that much about taxes.

Into this debate strides Professor Glaeser to bring some common sense to the discussion, though it is common sense backed up by the weight of impressive research. The director of the Taubman Center for State and Local Government at Harvard’s John F. Kennedy School, Professor Glaeser has, through a series of studies, traced the importance that skilled workers play in the economies of cities and what cities should do to attract and retain them. Drawing on his own work and that of others, Professor Glaeser warns policymakers that these workers still care deeply about what we all care about, such as the quality of a school system. He tells us that while they don’t migrate exclusively to the lowest-tax communities, they still seek fair and balanced taxes, and, above all, they expect safe streets. After the frills and embellishments of the Florida agenda, Professor Glaeser urges a return to basics. It’s my pleasure to present to you one of the most distinguished observers of urban economy, Professor Ed Glaeser.


Thank you very much. I’ve been a big fan of the Manhattan Institute for almost fifteen years, and it’s a special pleasure to be able to try to discuss my ideas in this august gathering.

Let me start with the economic approach to cities. Economists define cities as “the absence of physical space between people and firms.” As such, the attraction that cities have is their power to reduce transportation costs. It’s important that we interpret the concept of transportation costs broadly. In the old days, it was transportation costs for goods; today, it’s transportation costs for people and, even more important, transportation costs for ideas. After all, what is the magic of New York if not the ability of people’s ideas to hop from one person to another to fuel productivity, invention, and reinvention? Cities grow when people want to be near other people in that city or to something else that’s near that city. This simple framework helps us make sense of a great deal of America’s urban history.

Almost all our cities in the seventeenth and eighteenth centuries formed to eliminate the transportation costs for moving goods. There were ports, places where the river met the sea, places that had a real comparative advantage in moving goods. There were seaports and, later, railroad hubs; then there was the self-reinforcing growth of industry—the cycle where manufacturing firms located around an initial port in an initial downtown. Certainly, this is the case of New York, starting as early as the 1820s, when industries such as sugar refining, textiles, and publishing came and were the dominant industries in the city in the first half of the nineteenth century to take advantage of New York’s role as a commercial hub. Indeed, as late as 1950, seven out of the eight largest cities in the U.S. were fundamentally manufacturing cities, including New York. They had more manufacturing than the U.S. did as a whole.

By 1990, that would change: six out of the eight largest American cities would be under-concentrated in manufacturing. In an earlier era, water transport was much cheaper than moving goods by land. I’m always struck that in the classical era, it was cheaper to move goods from one end of the Mediterranean to the other than to move goods seventy-five miles over land. In an era in which water transportation is so much cheaper than its land equivalent, water was everything. Of the top twenty cities in 1900, eight were on the Atlantic (generally, where rivers meet the sea), three on the Mississippi River, three on the Ohio River, three on Lake Erie, two on Lake Michigan, and one was on the Pacific. There’s not a single city that was not tied to a waterway.

Over the twentieth century, this urban landscape built around manufacturing, transportation costs, and waterways received massive shocks. There was a massive move to sun and sprawl, to decentralization of population within metropolitan areas, a transition across metropolitan areas to places that were car-based, rather than public transportation–based. These phenomena should be seen as part of the same basic pattern—the revolution in transportation costs. In 1900, it was important to be near the coal mine, near the Great Lakes; in 2000, it was irrelevant.

In 2000, increasingly, cities are located around places where smart people want to live rather than around places where businesses have some inborn transportation cost advantage. Over the twentieth-century consumer cities have risen as the primacy of consumer tastes have dominated, rather than producer cities, and cities are increasingly built around cars, rather than walking or public transportation, with the prominent exception of this great metropolitan area. Figure 1 is the real cost of dollars per ton-mile by rail over the twentieth century. It’s more than a 90% reduction in the real costs over the past hundred years. This decrease by rail understates the real amount that transportation costs have declined, because it doesn’t even begin to factor in the incredible advantages created by highways, which allow the disbursement of production within metropolitan areas away from the old hubs that were the core of traditional monocentric cities and the rise of other forms of transport as well.

FIGURE 1. Declining Transport Costs: Rail

Naturally, this decline has been associated with the decline of manufacturing.

Manufacturing cities declined, unsurprisingly. Manufacturing suburbanized, as their firms are big users of space, and manufacturing left the U.S. as a whole. After all, if it’s so cheap to move goods, why not locate them all in China or wherever it’s cheapest to produce goods? Even within the U.S., manufacturing firms no longer locate near their suppliers or customers. There is moderate statistical correlation between the co-location of manufacturing firms that supply to one another, but it’s a tiny fraction and is unimportant relative to location based on where relevant workers are. That is not true of business services, because delivering their product still involves face-to-face contact, which involves people’s time. Business firms are still driven by suppliers and customers, and that is part of New York’s magic.

There is no reason to think that the decline of manufacturing firms or the exodus of manufacturing from cities is inefficient or bad. It is a big mistake to think that we’re going to reinvent cities around nineteenth-century solutions. Accompanying the decline in manufacturing cities has been the rise of the Sunbelt, which is part of a more general phenomenon of locating in places where consumers want to live. Many things are bound up in this: the ending of disease in the South, the ending of the pre–civil rights days of the South, and the rise of air conditioning. But one interpretation is that if the city of 1900 was located in a place where businesses have a comparative advantage because of proximity to the coal mines or the river, in 2000 businesses are in a place where consumers want to live, which is warm and dry and looks like Las Vegas or Phoenix. These places are cities, but they’re not necessarily the cities that we’re used to looking at. Density, by and large, is associated with decline. New York is an outlier in this. You’re seeing a general pattern that people have moved from walking cities and high-density cities to car-based cities.

This phenomenon has gone on unabated in the 1990s. There is a relationship between cars per capita across cities in 1990 and the growth of a city between 1990 and 2000. It’s not hard to understand why cars are so popular. The average commute by public transportation in this country is 47 minutes, the average commute by car is 23 minutes, and most forms of public transportation involve time costs from 15 to 20 minutes. Think of going to the number four, five, or six subway line in Manhattan, waiting for it to pick you up, and then walking from wherever it drops you off to your final destination. That’s the fixed time cost that’s involved, and it’s hard to fight against. It doesn’t mean, obviously, that you should be driving around in a car in Manhattan. That’s certainly not what we’re suggesting, but we can understand why car cities have been so popular. It’s foolish to think that that’s necessarily going away, although it is also clear that New York, and perhaps Boston and downtown San Francisco, can survive despite the disadvantage of being focused around a slower transportation technology.

Let me speak a bit about New York, since I’m here. The era 1800 to 1900 was spectacular. Between 1900 and 1930, a lot of the growth was in the outer boroughs and particularly in the consolidation of Manhattan; since 1940 or so, there is a decline relative to the U.S. as a whole. Manhattan’s period of greatest growth was one in which its waterborne advantage as the world’s port was most important. Despite the relative decline of New York, it is important to remember that New York is one of only two of the ten largest cities in 1930 that did not lose population between 1930 and 2000. (The other city that did not lose population in that period is Los Angeles.) The other eight cities all lost population, often in big numbers: Philadelphia, down from 2 million to 1.5 million; Detroit, down from 1.6 million to 950,000; Cleveland, from 900,000 to 478,000; St. Louis, from 821,000 to around 348,000.

I often sense when reading City Journal that there is an attempt to argue that all cities have in common revitalization and a future; while I think that New York, Boston, and San Francisco clearly have a future, it is less clear to me that they have all that much in common with Cleveland or St. Louis. It’s less clear that we can predict that those cities will come back. Of course, the continued strength of New York—its unique power—has increasingly been in the area of finance. New York looks more and more like a one-industry town. The top three-digit industry—that’s three-digit, which is a very small industry—has 28% of the payrolls in the latest economic census. The top four industries combined—four three-digit industries—have 56% percent of the payroll (a comparable number for Boston would be 37%). This shows an incredible focus on finance as the engine of the city’s economy, which is worrisome, especially for people who believe that industrial diversity is a key element in continuing urban reinvention and innovation.

The primacy of New York and finance is another legacy. The port comes out of financing the early investment in shipping and its continued dominance in New York. It’s because there is no industry where up-to-date information is more valuable. There is no other industry where you can make more money, more quickly, by knowing something quicker than anyone else does. In that world, the high-density world of Wall Street is perfect, because ideas cross hallways and streets more easily than oceans and continents; being in this information hive of downtown Wall Street is the comparative advantage of the city. It’s less clear that that will continue. Wall Street should be seen as the metaphor for what has driven cities beyond the world of sun and sprawl. What has driven the cities that have survived, even in the declining region? Those cities that have survived have reinvented themselves. They’ve become centers for idea creation, which is the main focus of my conversation here.

Moving ideas and skilled people has been what made cities work. It’s no longer about the port; it’s about the people. Entrepreneurship is part of the equation. I’m going to tend to use the share of people who have college degrees as my measure of skill. Understand that this is a metaphor at best, an imperfect measure of the general level of skills and entrepreneurship that we’re talking about. The policy vision that this will tend to push is that if we have skilled workers, the employers will follow. We should have an employee-based view of public policy. This is more generally the relationship between 1980 and 2000, between the share of the population with college degrees and the growth of the metropolitan area over the next twenty years. It’s a fairly pervasive hundred-year pattern, since the occupational patterns of 1880: the more skilled the occupations in 1880, the faster the growth of the city in every subsequent decade. The only time that the statistical significance gets weak is during the 1930s, which was, in many ways, an unusual decade.

The density of colleges has predicted growth since then. Schools predict population growth, employment growth, income growth, and housing growth. Schools are a reliable predictor of which cities do well and which cities do poorly. You might think that this is a phenomenon of our most recent information age, but, surprisingly, it appears to be true in the prewar era as well. One way to understand this fact is that the major element of modern urban success is speeding the flow of ideas. High density has its comparative advantage in facilitating face-to-face transmission of new knowledge; new innovations are created in urban areas. There’s great evidence on patent citations that shows the importance of physical proximity for speeding the flow of ideas, so patents are much more likely to cite other patents that are physically proximate, made by inventors who are physically proximate to one another.

To further confirm the role of cities as forges of human capital, places where new ideas are created, look at the massive wage difference between people who live in big cities and people who don’t, and then look at migrants who come to those big cities and migrants who leave. There’s no instance in which a migrant who comes to New York immediately gets a 30% wage hike, or a migrant who leaves New York receives a 30% wage decline. Over time, the rate of wages increases faster in big urban areas, and you accumulate human capital. When people leave, they appear to take that human capital with them. Increasingly, idea-oriented industries are over-concentrated in city centers, so if you look within cities, the industries that are nearest to the city center are finance, insurance, and real estate; manufacturing is sprawled farther out. Boston would be just another declining, cold, manufacturing city if it weren’t for its preponderance of human capital. One of the remarkable things in terms of the connection between human capital and ideas is how different this correlation is between regions.

If you’re in a cold region of this country, the Northeast or the Midwest, skills are almost everything. Champaign-Urbana, Illinois, is doing a little worse than would be expected, perhaps. Barnstable is doing a little better, but basically, things are on this regression line. One way to view this is Las Vegas, with its warm weather and its sprawl, and its low regulations, without any need for anyone to be smart and figure out new ways to make the place do well. But look at it. It’s down there—fantastic growth, no college graduates.

An alternative interpretation of what’s going on here is that skills in cities are particularly important for reinvention. The economic history of Boston shows that there have been at least five periods of reinvention, when this city has undergone a major shock and looked as though it was finished. But time and time again, smart people figured out some new thing to do. The crucial thing, obviously, is that you have smart people and that they want to stay there—and that they don’t immediately respond to a negative shock by moving on to the next city. Boston has an advantage because it’s the first consumer city in America, as opposed to New York City, which was settled by Dutch tradesmen. Boston was settled by people who wanted to be there fundamentally for consumption reasons; as such, they didn’t leave immediately.

The key to reinvention is to keep skilled people from leaving. That brings us to the actual policy issue: How do you make cities skilled? If skilled people are so important, how do you keep skilled people in your city? How do you create this engine for urban reinvention? The biggest problem with the skilled base is how permanent this is across space. There’s an 89% correlation between the share of college graduates in 1960 and 1990. That’s almost a perfect fit. That tells you that very few cities have been able to break their historical destiny in this regard. It’s awfully hard to change what you have. The biggest trend over the past twenty to thirty years has been the tendency of skilled cities to become more skilled.

Historically, a big problem for many cities has been that being good at manufacturing meant that they were less-skilled places, so cities such as Philadelphia and Detroit tended to attract huge numbers of unskilled people, which can be a real difficulty for the city later on. The fact that a city was good at doing something for less-skilled people—though it was great in 1950—was terrible in 2000, at least for urban success.

Figure 2 shows the relationship between the percentage of adults with college degrees in 1980 and the growth in percentage of people with college degrees from 1980 to 1990. To the cities that have, more has been given. This same effect appears in the 1990s in Figure 3; if anything, it’s gotten even stronger. The more skilled the place was initially, the more skilled it has become over time. This makes it hard for the Clevelands and the St. Louises of the world to break what they have. One reason for this is that skilled people within the firm have gotten more important. Skilled entrepreneurs used to hire unskilled workers. Think Henry Ford: maybe not skilled in terms of a B.A., but by any reasonable definition, a very skilled guy. He provided tens of thousands of jobs for relatively unskilled people. Bill Gates—again, not a skilled person by this measure; clearly, the measure is problematic—provides thousands of jobs, but jobs that are focused on high-skill areas. This is one reason to think that skilled people are increasingly sorting with other skilled people and decreasingly providing jobs for unskilled people. The evidence seems clear.

FIGURE 2. The Divergence of Human Capital

If you look at the relationship between the average share of workers in an industry who have college degrees and the average share of managers in the industry with college degrees, or the average share of professional workers in that industry with college degrees, that number was about 35% in 1970. It has risen to over 50% in 2000, so there’s been an increased tendency across industries and firms for skilled people to work in the same industries and firms. Figure 4 shows the relationship between the share of managers with college degrees in 1970 and the share of workers with college degrees in 1970, and Figure 5 shows that same relationship in 2000: 1970, less; 2000, more.


We also see this effect in wages. In general, the wage effect of living around skilled workers has gotten larger. A substantial increase occurs between 1970 and 2000, but the effect has been almost double for college graduates than it has been for high school dropouts. The wage increase for living around other college graduates is much stronger now for college graduates than it is for high school dropouts. That was not true in 1970. If you want to understand why initially skilled cities have become more skilled, it’s right there in the wage data in the labor market: if you are a college graduate, the premium from living around other skilled people has gone up over time, and, as a result, college graduates or other skilled people have moved into skilled cities. Figure 6 shows the stunning correlation between the share of the population with college degrees and wages across cities. This graph looked much flatter in 1970. Today, if you want to understand how rich a city is, you need only look at how many college graduates it has.

Before we get into the Florida bashing, let me say that Richard is sometimes right on: he generously cites my own research for providing the basics for arguing that skills are important. But there are two basic questions. Before we even get to the debates with Richard, the first question is: Should you be industry-based, or should you be people-based? Should you be providing a Silicon Valley sort of production-based strategy, or should you be getting smart people and letting them innovate? The good news is that certain types of industry-based strategies might help. After all, industry and skills go together.

The bad news is that in the best-case scenarios, the top bureaucrats of governments still choose losers. I go back to David Weinstein’s super work in the 1970s and 1980s on MITI, which was the hero of the Western world. Yet when you looked at the numbers, MITI chose losers. Over and over again, MITI chose bad companies—not good, comparable companies with Japan. If you compare the human capital that was involved in MITI with the average economic-development policy in a small city in the U.S., it’s incomparable. City governments did not have the resources that are capable of picking which industries are going to win.

Cities often focus on growing industries in which they have no conceivable comparative advantage. How are cities with incredibly low skill levels and no major universities going to be serious centers of biotech? Yet you hear over and over again from small towns throughout the U.S. that they imagine themselves as the next biotech center. This is an important cautionary tale for why choosing industries is not great. In many cases, they don’t even focus on the new industries. They focus on the old ones: they try to get smokestack industries or something like shipbuilding. Why would you think that by focusing on low-skilled, nineteenth-century industries that you are going to reinvent your city for the twenty-first century? The hallmark of the modern economy is unpredictability and innovation, and it argues for letting businesses and entrepreneurs make these decisions rather than trying to micromanage which industries you’re going to be good at. This is not to say that some business-related policies don’t make sense, such as low taxes and low regulation and things that will attract entrepreneurs, but micromanaging the industrial side doesn’t make sense.

From the producer approach, we come to the two visions of the consumer city, and here is where Richard and I finally part ways. Both of us tend to think that the important thing is attracting skilled workers. Let them innovate, let them figure out how to make a dollar or how to employ people in the city. The question is, which vision do you like? The Florida vision is dense, bohemian downtowns, tolerance, arts. It seems to offer a magic bullet, as was suggested in his speech for declining cities. It’s popular, partly because it seems cheap.

A more traditional vision, which I’m fonder of, is safe streets, quick commutes, and good schools. It favors medium density over high density. It recognizes that the car isn’t going away and that most middle-income people are attracted to the car. This advice is more relevant outside of Manhattan. It suggests hard work and real government spending, but low taxes relative to services. Without controlling for percentage of college education, variables such as index of creative occupations, the gay index, and patents per capita generally predict growth, so Richard is right about that. Once you control for percentage of college graduates, though, all this disappears. None of these measures has any ability to predict the future other than that. There’s no way with the data to reject the view that the number of college graduates, or some reasonable proxy for skills, is everything. Nothing else appears to predict anything about urban success.

I think that the key is to focus on people with skills. What do people with skills like? By plotting the relationship between density and skill levels we find that the places with the most skills are middle densities. If Florida were right, the most skilled people would be living in the densest areas, which is certainly not what we are seeing.

As for cars and skills, is it true that skilled people are so enthusiastic about bicycles? New York is an outlier, but on average, people with skills drive their cars. It’s one thing for the smart growth people to be enthusiastic that Boston’s going to reinvent itself as a public transportation–friendly place. Try convincing a pair of thirty-five-year-old biotech people from UNC to leave their $350,000 McMansions and their twenty-minute commutes to come and take the Red Line for fifty minutes to get downtown. It’s very hard to do that. Skilled people generally have shorter commutes, but not always.

What else do skilled people like? They live in places with lower murder rates, and they don’t like crime. The higher the murder rate, the less the growth in skills between 1970 and 2000. Figure 7 looks at the murder rate in 1975. On the skill rate today and on the rate of change, there is lots of room around the line, but in general, crime rates do appear to matter; running a decent police force has been very important for revitalizing Manhattan. You don’t get away with a few coffeehouses if you’re going to have places that are totally crime-ridden.

FIGURE 7. The Murder Rate and Growth in Skills 1970-1990

There are very few controlled experiments, but the weight of the evidence suggests that spending on education, if it’s effective, is good. Of course, most spending on education isn’t all that effective, but good schools at least offer the promise of growing your own skilled workers and attracting parents who care about good schools for their kids. So it’s hard not to think that this is significant. You do see this in terms of location within metropolitan areas. Skilled people scoot to the sides of boundaries, where they are able to send their kids to good public schools rather than to bad public schools. Redistribution and high taxes on the rich clearly did terrible things to cities like New York during the Lindsay era, or Boston under James Michael Curley. They tried to run local safety nets that, though they were going to right the wrongs of the world on the local level, generally created disasters, as rich people fled and as the general tax base declined precipitously. As much as we may want to take care of our poorest residents locally, it’s a very hard thing to do because of the mobility of skilled people. We often make things worse because the skilled people emigrate.

Finally, let’s discuss housing policy. Zoning and land-use regulations currently keep housing prices artificially high and restrict urban growth. The past work that I’ve done with the Manhattan Institute has argued that the rise in housing prices in Manhattan has been associated with a very different change from the Manhattan of the 1920s, which built without limit. That is a problem for the local economy. It doesn’t necessarily hurt the average skill level, though. This is a pet policy agenda of mine, to deregulate certain types of housing construction, restricting housing and actually pushing prices up, in some cases. The down side is that you don’t get skilled young people, and if you have the view that having people in their thirties who are skilled is particularly important, then artificially pushing up prices is a great problem.

Skills are particularly critical in cold areas of the country. The future for dense cities in the twenty-first century is primarily associated with speeding the flow of ideas and idea-centered industries. I’m skeptical about the ability to create various government top-down industry policies that are going to create that. The answer is to attract smart people and then leave them be. I am at one with Richard, if what he’s about is eliminating barriers that stop coffeehouses. That would be a perfectly reasonable thing to deregulate. But if you think that you’re going to save a city by quick fixes such as creating a funky downtown, it’s hard to imagine it working. All the available evidence suggests that most skilled people—for example, a thirty-eight-year-old married couple or a twenty-seven-year-old single person—want good, cheap schools, fast commutes, and safe streets. These things do not come cheap or easy.

MR. BOB WEISSBERG: I lived twenty-eight years in Champaign, Illinois. You’re taking what is essentially a nineteenth-century concept and moving it forward. My impression is that cities today, in many cases, have no relationship to what cities were even twenty-five years ago. I’m thinking about white-collar counties, for example. I know a bit about St. Louis. If you go out into the suburbs, what you find is amazing: enormous amounts of industry and amenities. The same is true in Detroit. There are also what are now called “rim cities.” Outside of Chicago, for example, there is an area around the airport with no population, but with endless office buildings, hotels, and so forth. But it doesn’t appear on the map.

How do you account for major differences in annexation laws that cut across the country? In the South, places like Houston expand crazily because they can annex anything around themselves without the acquiescence of the surrounding counties. In the East and the North, you have to get the cooperation of the people you’re going to annex, and it never happens. There is all kinds of growth in the South, but it’s not true growth. It’s growth because places like Charlotte can annex hundreds of thousands of people and therefore grow.

PROF. GLAESER: The starting point is that the growth to warmth that I showed you is at the state level, which, at the very least, should avoid any of this issue in terms of city definitions. The general rise of the Sunbelt is not the result of an artificial statistic, such as not counting the changing urban borders in the right way. All the facts that I’ve told you are true, at the metropolitan-area level and at the city level. I’ve been a bit sloppy going back and forth to some of the graphs that are metropolitan areas, which are multicounty units that would include the places that you’re talking about that are on the edges of the city. Some of them I’ve shown you have been cities, but none of the results are an artifact of looking at cities relative to metropolitan areas.

Embedded in your question is the rising importance of the automobile. Think about Detroit: there are two types of areas in Detroit. There are places in Detroit that were built pre-automobile, or built when not everyone had automobiles, and those places have been hit by two negative things: the rise of the car and the decline of cold places. Detroit has other places that feature car-based living outside the area, which have done well from the automobile but badly on the basis of being. Those places will do worse than comparable areas in Phoenix or Las Vegas, but certainly better than central-city Detroit. We can also do this at the county level, and county boundaries are also fixed. But it’s a mistake to think that any of this is a function of changing city boundaries, although those topics are endlessly fascinating to me as well.

MR. MICHAEL MEYERS: I’m executive director of the New York Civil Rights Coalition. I wonder if you would comment on the—and I’m surprised you did not—racial and economic divide in the cities, particularly with respect to the 1970s: the mass exodus of the white middle class to the suburbs, leaving behind a poor, black and Hispanic population with terrible schools and high taxation. I’d also like to hear your comments with respect to the new gentry that are moving in, who are bringing skills with them.

PROF. GLAESER: The New York of the 1970s, which is the New York in which I grew up, obviously was hit hard. The decline of manufacturing in the city and the rise of the automobile resulted in a situation in which the people who couldn’t afford two-car lifestyles on the urban frontier got left behind and very badly hurt. It is evidence of why we can’t as a society count on localities to handle redistribution. They don’t have the resources, and particularly with sufficiently mobile populations, they can never really engage in local redistribution. It has to be a responsibility that we take to the states or to the federal level. The tragedy is not that there was a movement to cars but that we didn’t respond to that movement by recognizing the need for higher levels of government to step in and take on the responsibilities that the cities were no longer able to administer.

Bound up in the question was the suggestion that many whites left because of racial issues. I have looked at the decentralization of population and employment, and the relationship between the share of minorities in the center-city population. The same sprawl occurs everywhere over time. It’s not as if there was only sprawl in places that had substantial minority populations downtown. The car was a powerful wrecking ball coming down on all traditional downtowns, and it—not flight from racial issues—is the dominant factor.

The second question is the effect of skilled people coming back in to the cities. There are obviously many problems related to that, which creates a yawning gap in equality in the region, but it certainly beats the alternative. The cities that have not had skilled people come back in are in far worse shape than the cities like New York, which now have resources to take care of their less advantaged citizens. While there clearly are tensions associated with inequality in New York, it is important that we do not respond in the way that occasional inequality was responded to in the past—by deciding to declare class or race warfare. Down that path lies urban ruin.

MR. RICHARD VALCOURT: Before retiring, I taught political science and urban affairs at Hunter College here in New York. In theory, I agree with much of what you’re saying, but I have some problems with the practical implementation. Let me give a couple of examples closer to what is now home for you in Massachusetts, because I came from that area. Regarding innovation: Newport, Rhode Island, has done very well in innovating after the navy took its big ships out of Newport and moved them south, after which some entrepreneurs turned the city around and made it a good place for tourists.

But then we have Fall River and what is now known as the South Coast. Thirty years ago, I was a member of the board of directors at the chamber of commerce there, and no matter what we did, it was all to no avail. The city of Fall River, as well as New Bedford and Taunton, and all those cities never went anywhere, despite their location, their trained workforce, and their institutions for learning. But in recent years—and this is where I might differ with you—you talk about Boston as a city that has come back alive. In that way, Fall River and New Bedford have come alive, too, because the suburbs have expanded, which has changed much of that area. You can’t buy houses any more in Taunton, which is about thirty or thirty-five miles from Boston, and the same thing is happening in Fall River.

But this morning, the Wall Street Journal editorialized against the city of Fall River and a couple of congressmen from there, saying that they are shortsighted by not allowing an LNG plant to be located on the shores of the Taunton River. That relates to what you were saying about industries moving into places where they shouldn’t because it’s going to devastate the community. Sure, we have a need for LNG, but the problem is being perpetuated there, because Fall River, after seventy years and the departure of the textile industry, is finally beginning to grow, thanks to Boston’s expansion southeastward and southwestward. Now there are other competing forces that want to locate some sort of industry there. Everything that you were saying might be right, but when you consider the implementation, there are many variables that you haven’t factored in.

PROF. GLAESER: At the local level, there are many variables that differ from city to city. It’s also true that urban success is never entirely predictable; it’s hard to tell when and if a particular city is going to take off.

Obviously, there are many different wrinkles involved in different localities. I’m a good friend of the mayor of Haverhill, Massachusetts, who deals with these issues on a day-to-day basis and has focused on attracting skilled people from Boston. Some places have such a poor endowment in terms of historical artifacts that they’re never going to revive. But focusing on skilled workers and on how to make the place friendly for entrepreneurship is a reasonable starting point. Obviously, everything at a city level has to be on a case-by-case basis beyond that.

MALE VOICE 1: You mentioned three cities—San Francisco, Boston, and New York—as places that are not going to decline, almost no matter what happens. You left out places that you think are not going to revive. I generally agree with you. I’m interested to hear in more detail why you named San Francisco, which has been dominated by tourism for many decades and has a notoriously low-wage, low-skilled economy. It had a renaissance in the 1990s with the fabled “reverse commute,” for which Silicon Valley millionaires would buy homes in San Francisco and then drive south to work every day, but that petered out in 2000. Yet housing prices are as high as ever. San Francisco is actually gaining population for the first time in many years, and it seems as though it’s economically bulletproof, no matter how bad the city government is, how high the taxes are, or how bad the housing market is. It’s still alive, and it looks as though it’s going to remain alive for many decades. What do you think accounts for that?

PROF. GLAESER: First, we can ask about the region and then we can think about the city within the region. With the combination of being built during the car era and having the best climate outside of Tuscany, the fundamentals for San Francisco are mind-bogglingly good, so we don’t need to look very far for why this place with a fantastic climate and enough initial investment of colleges has been able to do so well in the modern era. The interesting twist has been downtown San Francisco’s revitalization. It was not obvious that that was going to happen. I take this as being associated with a combination of two things, one of which is the consumer city aspects of San Francisco, and the other, its connection with the region. San Francisco is doing well, but it’s not doing all that well relative to the rest of its region.

Reverse commuting is increasing, but the share of population or the share of employment that’s in the city has actually been decreasing steadily over the past three decades. What’s interesting is that the city itself has managed to revitalize itself, which owes much to the fact that it is an appealing place for high human capital people to continue to live. It’s a beautiful city, and it has great natural amenities. In a world in which increasing wealth means that there are people who are willing to pay for those amenities and that lifestyle, the city has done extremely well. It’s a very skilled city.

You mentioned tourism. Tourism is a great predictor of how well cities have done over the past twenty or thirty years, in the U.S. and elsewhere, but not because tourism itself is a great powerhouse industry, at least outside of Orlando. In most of these places, tourism is important because it reflects being a place that people want to come to. There aren’t many tourists in Cleveland. There aren’t many tourists in many colder places in the country. There are many tourists in San Francisco because it’s a beautiful place, and its revitalization has much to do with its consumer city revitalization. It’s a luxury good, and as the society gets richer, people are increasingly willing to pay for luxury goods.

New York has some of that, so some of New York’s revitalization over the past twenty years—and some of Boston’s, as well—has to do with the charm of the city, with its amenities. But New York is much more driven by the economic advantages of its downtown, of the financial engine that is at the center of the region’s economy, whereas San Francisco is a bauble held aloft by the economic vitality of Silicon Valley. New York is much more of an older city in the sense that it combines economic health and being a consumer city in one. But the lesson is, if you don’t continue to induce rich people to live in Manhattan, or in Williamsburg, the economic vitality of the city is in a lot of trouble.

MALE VOICE 2: There are many doom-and-gloom stories about megacities: Mexico City, Calcutta, Beijing. What can you say about how your framework applies to the gloomy megacities of the future?

PROF. GLAESER: The big cities of Latin America, which are the ones that I’m most familiar with, are enormously fascinating. These cities have always been in poor countries that were doing a bad job of delivering welfare to their citizens, which is more observable in the cities. Observers from the U.S. wander around Rio de Janeiro but not around the traditional poverty-stricken rural areas of Brazil. People who move to these cities are not irrational; generally, their lives are made better by moving to these cities.

Many of these cities are artificially large because they are the product of government policies that strongly favor the capital cities relative to the non-capital cities. In many cases, it’s impossible for businesses to get anything done unless they are, for example, in Mexico City or Buenos Aires. Brazil, because of its strong federal structure, would be an exception. The largest cities in places that are unstable democracies are 50% larger than the largest cities in places that are stable democracies that don’t have policies that favor being close to the center of power.

These places are, in the long run, suffering from exactly the same transportation cost problems that we are. In all these places, the car is moving on unabated. Increasingly, sprawl is going on with much less management than it’s even getting here. There are many reasons for the public policy concern that goes along with that. I tend to focus on why the medium-size cities haven’t been growing more, and the answer is generally politics and political favoritism toward the primary city. I would also focus on getting some management of the inexorable move toward car-based living in these places.

MALE VOICE 3: A writer for Forbes magazine, Richard Karlgaard, advocates the arbitraging of high-skill, high incomes with smaller, lower-cost cities, telecommuting, making a big income, and living in a place where that big income goes a longer way. What would your prescription be for cities that don’t have the natural advantages, other than price, to attract people to take their big-city incomes to where they have short or no commutes and other amenities that are advantages for smaller places?

PROF. GLAESER: I’m assuming that the places you’re talking about already have relatively safe streets and good commutes. Beyond that, I would probably focus most on education for children. I would have an education-based policy for many of these places. I would also imitate Las Vegas, in terms of deregulation and having young entrepreneurs come to start up their macadamia-roasting plant or whatever crazy but brilliant idea comes along.

You brought up telecommuting. It remains an open question as to whether information technology is going to be critically negative for central cities. The right way to view this is that cities facilitate face-to-face interactions, and electronic technologies facilitate electronic interactions. Those things are not always substitutes for each other. In many cases, they’re strong complements; we’re moving toward a more interactive society if it’s going to see both these things. First, people who interact with one another electronically are also people who interact face-to-face with one another. Second, over time and space, the rise of the telephone has been strongly associated with more, not less, urbanization. Indeed, the pundits 120 years ago were arguing that the telephone was going to spell the demise of urban areas—before the greatest run-up in urban areas in world history. Third, the most famous example of geographic clustering in the modern age is the industry that has the best access to all forms of information technology. No place more than Silicon Valley would have this, but even though they live in a place that’s lower-density than Manhattan, it is a very dense place and a car-based city. So there’s no sense in which information technology is a danger to Manhattan or to dense places as long as it’s accompanied by good policies that ensure the continuing desire of skilled people to live in this region and in this city.


Center for Civic Innovation.



   What makes cities grow?
   Dr. Edward Glaeser, Director of the Taubman Center for State and Local Government at Harvard University’s John F. Kennedy School, has extensively researched the determinants of city growth. He finds that urban areas thrive when they maintain safe streets, reduce barriers to building, and keep taxes low. Above all, however, Dr. Glaeser has discovered that residents of faster-growing cities are better educated than their counterparts in struggling cities. His research suggests that, of all the forces shaped by policy decision making, education is the most important predictor of urban growth.


Home | About MI | Scholars | Publications | Books | Links | Contact MI
City Journal | CAU | CCI | CEPE | CLP | CMP | CRD | ECNY
Thank you for visiting us.
To receive a General Information Packet, please email
and include your name and address in your e-mail message.
Copyright © 2009 Manhattan Institute for Policy Research, Inc. All rights reserved.
52 Vanderbilt Avenue, New York, N.Y. 10017
phone (212) 599-7000 / fax (212) 599-3494