Click to visit City Journal California

By Edward L. Glaeser

Triumph of the City: How Our Greatest Invention Makes Us Richer, Smarter, Greener, Healthier, and Happier

City Journal

Edward L. Glaeser
Why Economists Still Like Cities
As the Information Age dawns, the dismal science has cheery news for urban America.
Spring 1996

Are cities doomed? Economists have amassed a wealth of data that allow us to answer: emphatically no. Cities are essential to the Information Age economy, and those with a diverse mix of industries and a highly skilled workforce—preeminently New York—have a bright future. Indeed, the communications revolution gives cities a whole new vitality. In study after study, the evidence of cities’ continuing vibrancy is unambiguous. Here's what the data show:

A Hoover Institution study I conducted with economist David Maré found that, on average, workers in big cities earn considerably more than workers in small cities, who in turn earn far more than their rural counterparts. According to 1990 census data, workers who live in metropolitan areas surrounding cities of more than 500,000 people earn 34.4 percent more than workers who live outside metropolitan areas, and 10.8 percent more than those who live in metropolitan areas with smaller central cities. And earnings are higher in cities than in suburbs: within metropolitan areas, on average, individuals who work within the central city earn 17.3 percent more than those who work beyond the city limits. These results stand even if we make them go through a large number of statistical hoops and further tests. For example, workers with the same level of education and those in the same demographic group get substantially higher wages in cities.

Standard labor economics tells us that high wages reflect higher productivity: a company won’t pay a premium for workers unless it gets value for its money. And indeed, economists Antonio Ciccone and Robert Hall have established a direct connection between urban density and productivity. Looking across the U.S., they found that industries in denser areas are much more productive than elsewhere.

Why? A look at earning patterns over time suggests that it is because the dynamic, economically diverse urban environment allows workers to develop skills superior to those they would be able to develop outside a city. For it is longevity in the urban workforce that enhances an employee's value. New migrants to cities experience less than a 1 percent wage gain from their move. But after five years, those workers are earning over 10 percent more than similar workers who stayed in rural areas. These wage gains represent permanent skill accumulation: when urban workers leave cities, they experience no significant wage declines. And there’s no evidence that technological or social developments are diminishing cities' role as the training ground for America’s workers: I found little change over the past 15 years in the urban wage premium.

Although workers of all educational levels earn more in cities, the urban wage premium is greatest for the most educated. Little wonder, then, that cities continue to attract the brightest and best-educated young workers. The National Longitudinal Survey of Youth tells us that young urban dwellers have significantly higher IQ scores on average than young people living in rural areas. And the average resident of a big-city metropolitan area has about one more year of education than his rural counterpart.

Today, of course, the ingredients for urban success are widely different than they were during the first great age of American urbanization a century ago. During the Industrial Age, urbanization and the rise of manufacturing were so closely linked as to be almost one and the same thing. In 1870, economist Alberto Ades and I have found, the statistical correlation between a state's concentration of manufacturing and its level of urbanization was 86.6 percent—demonstrating what to statisticians is an extraordinarily powerful connection between the two variables. Heavy industry and urban prosperity went together for the simple reason that cities made mass production possible. A big city meant a ready-made market for manufactured goods, a market comparatively inexpensive to serve because no long-distance transportation costs were involved. So factories located in large cities could achieve real economies of scale: they could spread out fixed costs of production over massive quantities of products.

As the twentieth century progressed, it became far less advantageous for factories to be located in large cities. Transportation costs declined dramatically, making a factory’s proximity to a large urban market much less crucial. The advantages of large-scale production diminished as technological advances made it economical to produce higher-quality, semi-customized goods in smaller batches and as consumers demanded those specialized products. As a result, manufacturers were able to locate plants in suburban and rural areas that offered less expensive real estate, a more favorable tax and regulatory climate, and a better quality of life for executives and workers.

If the U.S. economy had continued to center on heavy manufacturing, these developments may well have spelled doom for the cities. But traditional manufacturing has declined greatly relative to high-wage, high-skill industries, from telecommunications to finance to advertising. In 1950, 33.7 percent of American workers had manufacturing jobs; by 1990 this proportion had declined by almost half, to 17.4 percent.

The shift away from heavy industry has been the salvation of urban America. For while the decline in urban manufacturing has been unavoidable, the growth of high-skill industries has enabled many cities to prosper.

Unsurprisingly, cities that depend on manufacturing have tended to decline. Economists José Scheinkman, Andrei Shleifer, and I, in a study for the Journal of Monteary Economics, examined how cities’ dependence on manufacturing in 1960 affected their growth during the subsequent three decades. We found that every 20 percent increment in a city’s 1960 concentration of employment in manufacturing corresponded with a 16 percent decrease in population between 1960 and 1990. That same 20 percent increase in dependence on manufacturing in 1960 led to a 50 percent decline in manufacturing employment and a 13.4 percent decline in non-manufacturing employment between 1960 and 1990.

Declining manufacturing centers, especially smaller cities, dot the American landscape. East Chicago had 58.9 percent of its labor force in the manufacturing sector in 1960. Its population declined from 57,669 to 33,892, or 41 percent, between 1960 and 1990. During the same period Flint, Michigan, and Youngstown, Ohio, lost 28.5 percent and 44.2 percent of their populations, respectively. By contrast, cities in the Midwest and Northeast that don’t depend on smokestack industries showed substantial growth. These include Springfield, Illinois, the state capital, which grew by 20 percent; Stamford, Connecticut, which has attracted financial firms and corporate headquarters and which grew 17 percent; and Champaign, Illinois, a college town, which grew 28 percent. In the West and the South, the nonmanufacturing winners are even more striking. Phoenix, a young metropolis with a diverse economy and a population that recently surpassed 1 million, more than doubled in population between 1960 and 1990; Durham, North Carolina, with strong research, health care, and service sectors, grew by more than 50 percent in the same period.

The key to urban success or failure in today’s economy is simple: high-skill cities prosper; low-skill ones stagnate or decline. The National Bureau of Economic Research study found that cities that had a high proportion of college-educated workers in 1960 grew in population over the subsequent three decades, while those that didn’t, shrank. Compare the striking success of the university town of Madison, Wisconsin, where 20.9 percent of residents were college-educated in 1960, with the relative failure of nearby blue-collar Milwaukee, with only 5.9 percent college-educated in the same year. In the next three decades, Madison’s population grew by half, while Milwaukee’s declined 33 percent during the same period. Or compare two southern Virginia cities: Portsmouth, once a shipbuilding center (4.1 percent college-educated in 1960), shrank by 28 percent between 1960 and 1990, while Hampton, home of NASA’s Langley Research Center (10.4 percent college-educated in 1960), grew by 53 percent.

Education's importance for urban success is growing rapidly. Scheinkman, Shleifer, and I found that the connection between initial levels of education and population growth was much greater from 1970 to 1990 than it had been in the previous 20 years. Between 1950 and 1970, one additional year of schooling on average in a city increased its growth rate by 3.8 percent. For the 1970 to 1990 period, that figure soared to 8.1 percent.

Even the less well schooled residents of well educated cities prosper. James Rauch and David Maré have both marshaled overwhelming evidence that higher average education levels in a city translate into higher wages for everyone—not just the well-educated themselves. That is, if two workers have identical levels of schooling, the one who lives in the better-educated city is likely to make more money—perhaps because the sharing of knowledge in a highly educated workforce makes each individual worker more productive.

It’s not just the promise of financial success that makes cities magnets for talented workers. Young workers flock to cities in part because they will be exposed to a far wider range of accomplished elders and to a wider range of experiences than their rural peers. Economist Alfred Marshall's century-old observation is as true today as ever: “Great are the advantages which people following the same skilled trade get from near neighborhood to one another. The mysteries of the trade become no mystery but are, as it were, in the air.”

Young medical residents in New York, for example, encounter a much broader range of medical problems than their rural counterparts. Artists and actors benefit from the city’s rich cultural environment. Young lawyers, developing skills at breakneck pace over grueling hours, learn in New York not only from the partners at their own firms but also from their opponents and from lawyers they meet socially.

Large urban labor markets also encourage what economists call the coordination of labor: a big city makes it much easier for a worker or entrepreneur to find a niche. As Adam Smith noted, the bigger the market, the finer a division of labor is possible. By contrast, Smith wrote, in small villages “every farmer must be butcher, baker, and brewer for his own family.” Hard data confirm Smith’s observations: in a study for the National Bureau of Economic Research, Alberto Ades and I found a correlation of over 87 percent between the division of labor and the degree of urbanization across states in the nineteenth century. We lack comparable data for modern periods, but a casual look at the yellow pages shows just how specialized big-city businesses can be. The Manhattan yellow pages list such esoteric specialties as necktie renovating (one entry), nautical instruments (three entries), and papaya products (one entry). Not one of these entries appears in the San Francisco yellow pages, let alone in those of the typical small town.

Economists Gary Becker and Kevin Murphy have argued that cities support the division of labor because coordination between buyers and sellers is easier in cities. Since it's easier to shop around for suppliers in cities, it is also easier to hire outsiders to perform tasks that in an unspecialized economy would have to be performed in-house.

The bigger and more diverse the labor market, the easier it is for workers to find the right career and the right job. This is crucial for productivity: economists Robert Topel and Michael Ward have found that more than one-third of wage growth for young men comes from finding the right employer. In a city, a young worker can move across firms to find the match that is best for him. If he turns out to have chosen a field that is not ideal, he can switch careers. journalists become financiers; lawyers transmogrify into real estate investors.

Economists have pointed to other advantages of the coordination of labor. Julio Rotemberg and Garth Saloner suggest that workers in large cities are more willing to invest in improving their own skills because such investments are more likely to pay off, since they have a superior bargaining position. In a one-company town, by contrast, the company has all the bargaining power. This would explain why greater competition among companies in a city corresponds with higher wages and productivity. Economist Alfred Marshall suggested in his 1890 masterwork, Principles of Economics, that a major function of cities was to insure workers against firm-specific risk. Since Wall Street has so many trading firms, workers know that even if their firm gets into trouble they will normally be able to look for work down the street. Sherwin Rosen emphasized that cities would be especially good for the most able workers, who have a wider market for their talents.

The same features that give cities an economic edge on rural areas also give the most successful cities an edge on others. Competition, for example, is a key to economic growth. My own research has shown that the higher the degree of competition within industries—measured as the number of firms in an industry relative to the total number of workers in that industry—the more the industry tends to grow. This explains why successful local industries, like electronics in Silicon Valley or finance in Manhattan, often have a myriad of small, competitive firms.

A diverse mix of industries is also a major ingredient of urban success. A study that José Scheinkman, Andrei Shleifer and Hédi Kallal, and I published in the Journal of Political Economy found that as the concentration of employment in a given city's largest five industries fell by 10 percent—that is, as the city’s economy became more diverse—that city’s rate of growth in employment rose by 9 percent between 1960 and 1990.

Individual industries, too, grew faster in diversified cities than elsewhere. In a sample of city industries from 1956 to 1987, two of the fastest growing were business services in Albuquerque and durable wholesale trade in San Jose—cities where the five biggest industries employed, respectively, just 21.7 percent and 29 percent of all workers. In contrast, the slowest-growing city industry was textiles in Gadsden, Alabama, where the five largest industries employed 40.6 percent of workers.

It’s not just that more diverse cities are safer from downturns in a particular industry. Standard financial theory tells us that if you invest in many things, your portfolio will be less volatile—safer from sudden declines—than if you invest in only one thing. But if such a portfolio theory of cities were correct, diversified cities would grow relatively predictably, while those with fewer industries would be subject to boom-and-bust swings. In fact, we found that the reverse is true: more diversified cities have higher levels of variance.

Why? Whereas the portfolio effect depends on one investment not affecting another—if my Mobil stock goes down, that doesn’t mean my Disney stock will—industries within a city feed off one another in myriad complex ways. A booming retail sector gives business to the advertising industry; tourism boosts retailing; growth in the financial industry creates opportunities for industries that provide financial institutions with services, from telecommunications companies to law firms. Thus, lean times in me industry may bode ill for others, even in economically diversified cities—but in the long run, diverse cities will thrive.

Economic evidence also explodes the myth that big is necessarily bad. It’s a common misconception that there are limits to growth, that cities can grow only so much before beginning an inexorable decline. As the great, but often wildly wrong, critic Lewis Mumford put it: “The forces that have formed our cities in the past are now almost automatically, by their insensate dynamism, wrecking them and threatening to destroy whole countries and continents.”

Quite the contrary: there is no statistically significant connection between a city’s size and its subsequent growth. Cities that grew between 1950 and 1960, for example, also grew between 1960 and 1990, regardless of how large they were to begin with. Of the six cities with more than 1 million inhabitants in 1980, three—Chicago, Detroit, and Philadelphia—lost population in the ensuing decade, while the other three—New York, Los Angeles, and Houston—gained population.

Big cities—especially those with a highly educated populace—have enormous advantages as centers of knowledge, information, and skills. Spatial proximity may no longer be important for the delivery of widgets, but it is vital for the transfer of ideas: this is the principle on which universities, think tanks, and industrial parks are built. It’s also the reason why firms in highly skilled industries tend to congregate near one another: high-tech concerns in California's Silicon Valley, financial firms in Manhattan, cable TV companies in Denver. In Silicon Valley, AnnaLee Saxenian writes in Regional Advantage, “entrepreneurs came to see social relationships and even gossip as a crucial aspect of their businesses. In an industry characterized by rapid technological change and intense competition, such informal conversation was often of more value than more conventional but less timely forums such as industry journals.”

Economists Adam Jaffe, Michael Trajtenberg, and Rebecca Henderson have illustrated this point on a broader scale by looking at patent records. Big cities, they found, are centers for innovation: in a sample of patents issued in 1975, 29.4 percent of all domestic university patents and 30.7 percent of all top corporate patents came from inventors living in the New York, Los Angeles, Philadelphia, and Boston areas, which together had only 18 percent of the U.S. population. They also found that new patents are much likelier to cite previous patents of geographically proximate inventors—even if those inventors worked for different firms. So a cross-fertilization effect seems to be at work.

That’s not surprising. Cities spur innovation not only by concentrating expertise within particular industries but also by bringing together different industries. As Jane Jacobs argued in her magnificent The Economy of Cities, the most important ideas come from unplanned combinations of existing ideas, and cities where such spontaneous innovations flourish will themselves thrive. This role of unplanned creations explains why economic diversity is so important for urban growth. Cities that do fewer things have less opportunity for unplanned combinations of ideas.

What, then, of the popular notion that fax machines, e-mail, and other high-tech means of communication have made geographic proximity obsolete? The conventional thinking is that electronics are a substitute for cities: why put up with traffic, noise, and other urban blights when you can live on a ranch in Montana and conduct business by e-mail?

It’s too early, of course, for economists to refute this notion conclusively—though the evidence we have argues against it. Despite the proliferation of information technology in the 1980s, the share of the U.S. population living in the four largest metropolitan areas rose slightly during that decade, from 19.47 percent to 19.64 percent. The total population of the central cities in those metropolitan areas rose 4.7 percent over the same period. The share of the U.S. population living in metropolitan areas with more than 3 million inhabitants has stayed constant over the past 20 years, and the share living in all metropolitan areas has risen steadily. As we’ve seen, the productivity of urban workers is still much higher than that of rural workers. And where are all those wired workers conducting business at home? As the newsletter Edge City News reports, the nation’s ten biggest concentrations of home-based workers are all in major metropolitan regions, including three each in the New York and Los Angeles areas.

The notion that telecommunications can replace face-to-face meetings defies common sense. Phones have been ubiquitous for decades, but that hasn't stopped businessmen from consummating deals and brainstorming ideas around the conference table. While electronic communication can help us communicate simple ideas, and can put masses of raw information at our fingertips, its ability to transfer more complicated ideas—and, crucially, skills—is sharply constrained. I can tell my wife that I will be late for dinner over the phone, but it would be almost impossible for me to teach her how to use a new software package on the phone. Nor would the fax machine and email make it much easier: we actually need to sit down and use the software together. Likewise, experienced doctors teach young interns face-to-face while treating patients; lawyers forge their skills through actual courtroom experience; 26-year-old Wall Street traders learn by watching their 36-year-old elders on the trading floor. This is true of any high-skill industry.

Moreover, unplanned combinations of ideas are critical for innovation and economic growth. Telephone calls, faxes, and e-mail messages are, by their nature, planned, reaching only the people they’re intended to reach. Cities are unparalleled in their ability to provide the constant, spontaneous interchange of knowledge and insight in which new ideas thrive.

Increasingly, knowledge is economic power: according to the U.S. Bureau of Labor Statistics, male college graduates’ earnings advantage over male high school graduates doubled from 37 percent in 1979 to 74 percent in 1992; for women the increase in the wage premium was similar, though slightly less pronounced. The differences between the highest- and lowest-paid workers in general continue to grow—and standard economic theory tells us that means that skills command more and more of a premium in the market. As the overall mass of knowledge rises, the variety of specialized skills grows. Since cities play such a crucial role in developing skills, an increase in the demand for skills is good news for cities.

The basic forces shaping the Information Age economy bode very well for cities—and best of all for dense, economically diverse cities like New York. But it’s vital that urban leaders recognize where the future of cities lies: not in seeking handouts from Washington or trying to resuscitate unskilled industry, but in creating a hospitable climate for high-skill industries and smart, well-educated workers. That means minimizing the tax and regulatory burdens that workers and entrepreneurs must bear, building and maintaining the basic infrastructure that undergirds a vibrant economy, and providing excellent municipal services—from schools to police to sanitation—that make urban life attractive for today’s skilled workers. The cities that succeed in doing these things will be the economic centers of the twenty-first century.

respondrespondTEXT SIZE
More by Edward L. Glaeser:
If you liked this story, you may also be interested in:
If you enjoyed
this article,
why not subscribe
to City Journal? subscribe Get the Free App on iTunes Or sign up for free online updates: