Triumph of the City

by Edward Glaeser

  The Fairchildren gave the Valley a new set of entrepreneurs, and others soon joined them. Many of the companies formed near Stanford focused on hardware, including Intel, Cisco, and Sun Microsystems. Two former Hewlett-Packard employees, both members of Silicon Valley’s Homebrew Computer Club, mixed hardware and software innovations when they started Apple Computer. A former Apple employee started eBay in the 1990s, when Silicon Valley also became the place for pioneering the Internet. Both Yahoo! and Google were formed by Stanford graduates not far from their alma mater.

  In some ways, Silicon Valley is like a well-functioning traditional city. It attracts brilliant people and then connects them. Walker’s Wagon Wheel played a legendary role as a place where smart entrepreneurs shared ideas with one another outside the confines of their various day jobs. Silicon Valley’s concentration is also a response to the curse of communicating complexity; all that cutting-edge technology can be pretty complicated, and geographic proximity helps the flow of information. Like all of today’s successful cities, its strength lies in its human capital, which is nurtured by Stanford University and attracted by economic opportunity and a pleasant climate.

  Yet in some ways Silicon Valley looks completely different from any older city. It is built almost entirely around the car. While there are some areas, particularly in downtown Palo Alto, where you can walk a few pleasant blocks to get an ice cream or buy a book, feet are generally useless for getting from one company to another. A few companies, like Google, run their own bus services, but public transportation is minimal. Only 3.7 percent of the people living in Santa Clara County take mass transit to work. Car-based living goes together with low density levels. There are only about 2.14 people per acre living in Santa Clara County. There is a lot of action in the Valley, but you have to drive a ways to find it.

  Santa Clara County’s economy makes little room for poorer, less skilled people. Even after the housing bust, the median housing price in the San Jose metropolitan areas remains over $550,000, making it very hard for someone who isn’t a successful computer person to buy a home. Some of the most attractive areas in the Valley have completely priced out less skilled people and the businesses that employ them. Only 22.2 percent of Palo Alto’s residents over the age of twenty-five lack a college degree.

  The Valley’s other major drawback is that it’s a one-industry town; over half the county’s payroll in its export-related sectors, such as manufacturing, information, and even wholesaling, appears to come from computer-related firms. Traditionally, single-industry cities, like Detroit and Manchester, haven’t done well in the long run because their industrial monocultures discourage the growth of new ideas and companies. Jane Jacobs explained this phenomenon by pointing out that new ideas are formed by combining old ideas. Even in information technology, some of the most successful entrepreneurs of the last thirty years have been hybrids, merging ideas from multiple industries. Michael Bloomberg created his enormously successful IT company by knowing exactly what Wall Street traders wanted to know and how technology could help them. Facebook started on a college campus, and its founders knew what kind of information undergraduates wanted to share. Proximity to customers or related industries provides valuable information that can be a wellspring of innovation.

  When eBay wanted to expand its customer base, it had to reach outside Silicon Valley in order to find a CEO, Meg Whitman, who had amassed experience selling to the American public at Procter & Gamble, Stride Rite, Walt Disney, and Hasbro. Can the Valley’s software experts continue to offset their isolation from the rest of American industry by occasionally importing smart, experienced outsiders? The Valley was a great place to develop faster and faster semiconductors, but it might not be the best place to connect technology with other businesses.

  But perhaps those connections aren’t necessary. The Internet revolution was about making technology accessible for ordinary Americans, who can search the Web with Google, use e-mail, or buy and sell on eBay. Software engineers are people, too, and they can look to their families and friends—as the Facebook founders did—to understand the needs and desires of ordinary mortals.

  In the long run, Silicon Valley will likely be hurt by concentrating too much on a single industry and by allowing too much space between its innovators. But despite the poor track record of single-industry cities, like Detroit, there are good reasons to be more optimistic about the Valley. Unlike Detroit, Silicon Valley is not concentrated in a few big firms, and that helps keep the area entrepreneurial. It has superb educational institutions and continues to invest in its schools and universities. It has arguably the best climate in the United States, and that will continue to attract rich, smart people, who are willing to pay some of the country’s highest housing prices to live in that climate surrounded by many of the world’s most innovative companies.

  The Cities of Tomorrow

  Silicon Valley and Bangalore remind us that electronic interactions won’t make face-to-face contact obsolete. The computer industry, more than any other sector, is the place where one might expect remote communication to replace person-to-person meetings; computer companies have the best teleconferencing tools, the best Internet applications, the best means of connecting far-flung collaborators. Yet despite their ability to work at long distances, this industry has become the world’s most famous example of the benefits of geographic concentration. Technology innovators who could easily connect electronically pay for some of America’s most expensive real estate to reap the benefits of being able to meet in person.

  A wealth of research confirms the importance of face-to-face contact. One experiment performed by two researchers at the University of Michigan challenged groups of six students to play a game in which everyone could earn money by cooperating. One set of groups met for ten minutes face-to-face to discuss strategy before playing. Another set of groups had thirty minutes for electronic interaction. The groups that met in person cooperated well and earned more money. The groups that had only connected electronically fell apart, as members put their personal gains ahead of the group’s needs. This finding resonates well with many other experiments, which have shown that face-to-face contact leads to more trust, generosity, and cooperation than any other sort of interaction.

  The very first experiment in social psychology was conducted by a University of Indiana psychologist who was also an avid bicyclist. He noted that “racing men” believe that “the value of a pace,” or competitor, shaves twenty to thirty seconds off the time of a mile. To rigorously test the value of human proximity, he got forty children to compete at spinning fishing reels to pull a cable. In all cases, the kids were supposed to go as fast as they could, but most of them, especially the slower ones, were much quicker when they were paired with another child. Modern statistical evidence finds that young professionals today work longer hours if they live in a metropolitan area with plenty of competitors in their own occupational niche.

  Supermarket checkouts provide a particularly striking example of the power of proximity. As anyone who has been to a grocery store knows, checkout clerks differ wildly in their speed and competence. In one major chain, clerks with differing abilities are more or less randomly shuffled across shifts, which enabled two economists to look at the impact of productive peers. It turns out that the productivity of average clerks rises substantially when there is a star clerk working on their shift, and those same average clerks get worse when their shift is filled with below-average clerks.

  Statistical evidence also suggests that electronic interactions and face-to-face interactions support one another; in the language of economics, they’re complements rather than substitutes. Telephone calls are disproportionately made among people who are geographically close, presumably because face-to-face relationships increase the demand for talking over the phone. And when countries become more urban, they engage in more electronic communications.

  Certainly some people still work alone, handling customer complaints or airline reservati
ons, perhaps, over the phone in some spot far from any city. However, most of those jobs require less skill and accordingly pay less. In the average U.S. county with less than one person per acre, 15.8 percent of adults have college degrees. In the average county with more than two people per acre, 30.6 percent of adults have college degrees. The Internet and long-distance calling make it possible to perform basic tasks at home, but working alone makes it hard to actually accumulate the most valuable forms of human capital.

  Innovations cluster in places like Silicon Valley because ideas cross corridors and streets more easily than continents and seas. Patent citations demonstrate the intellectual advantage of proximity. In 1993, three economists found that patents had a remarkable tendency to cite other patents that were geographically close. More than one fifth of all corporate patent citations were to older patents in the same metropolitan area, and more than one quarter of these citations were to patents in the same state. Correcting for the tendency of people to cite patents from the same firm, the propensity to cite patents from the same metropolitan area is about twice as likely as it should have been if citations were determined by luck. The geographic pattern becomes looser as patents age, because ideas do eventually spread across space, but even in our age of information technology, ideas are often geographically localized. More recent research continues to find that patent citations are geographically close to one another. Recent research also finds that productivity is significantly higher for firms that locate near the geographic center of inventive activity in their industry. Just as proximity speeds the flow of the most important inventions, it also enables the more mundane learning that turns neophytes into experts. More than a century ago, the great English economist Alfred Marshall described how in dense concentrations “the mysteries of the trade become no mystery but are, as it were, in the air.” Hanging around successful older engineers helps make younger engineers more successful themselves.

  Data backs up Marshall’s claim. Workers in big cities earn about 30 percent more than their nonurban equivalents, but people who come to urban areas don’t experience higher wage gains overnight. Year-by-year, workers in cities have higher wage growth, as they accumulate the skills that make them successful. Wage growth is particularly faster in cities with more skilled workers. Two decades of extra job-market experience is associated with 10 percent more wage growth in skilled metropolitan areas than in nonmetropolitan America, but only 3 percent more wage growth in less skilled metropolitan areas.

  For over a century, pundits have been predicting that new forms of communication would make urban life irrelevant. One hundred years ago, the telephone was supposed to make cities unnecessary. That didn’t happen. More recently, faxes, e-mail, and videoconferencing were all supposed to eliminate the need for face-to-face meetings, yet business travel has soared over the last twenty years. To defeat the human need for face-to-face contact, our technological marvels would need to defeat millions of years of human evolution that has made us into machines for learning from the people next to us.

  Better audio and higher-definition screens have enabled videoconferences to seem more like real live encounters, but will technology ever be able to simulate the full range of sensory inputs—eye contact, olfactory cues, the warmth of a handshake—that help make live meetings work? Furthermore, much of the value of a dense work environment comes from unplanned meetings and observing the random doings of the people around you. Fancy videoconferences will never give a young assistant the ability to learn by watching the day-to-day operations of a successful mentor. Facebook is another Internet technology that makes face-to-face interactions more valuable and effective. Studies find that Facebook typically connects people who have met in person at a party or in the same class, and that Facebook is disproportionately used by people who are good at real-life conversation. Moreover, the initial idea for an Internet social network seems to have come out of a series of murky meetings between members of a real live network of smart, ambitious Harvard students.

  Today, information technology is changing the world, making it more idea-intensive, better connected, and ultimately more urban. Improvements in information technology seem to have increased, rather than reduced, the value of face-to-face connections, which might be called Jevons’s Complementarity Corollary. The nineteenth-century English economist William Stanley Jevons noted that more fuel-efficient steam engines didn’t lead to less coal consumption. Better engines made energy use effectively less expensive, and helped move the world to an industrial era powered by coal. The term Jevons’s paradox has come to refer to any situation in which efficiency improvements lead to more, not less, consumption—one reason why low-calorie cookies can lead to larger waistlines and fuel-efficient cars can end up consuming more gas. Jevons’s paradox applied to information technology means that as we acquire more efficient means of transmitting information, like e-mail or Skype, we spend more, not less, time transmitting information.

  One might think that better information technology would reduce the need to learn from other sources, like face-to-face meetings in cities. But Jevons’s Complementarity Corollary, which follows naturally from Jevons’s paradox, predicts that improvements in information technology can lead to more demand for face-to-face contact, because face time complements time spent communicating electronically. All those electronic interactions are creating a more relationship-intensive world, just as improvements in steam engines led to a more coal-intensive economy, and those relationships need both e-mail and interpersonal contact. Better connections between people create far-ranging opportunities for trade and commerce. Information technology, from the book to the Internet, has enormously increased the scope of human knowledge and consequently made it more difficult to master. Better information technology has made the world more information intensive, which in turn has made knowledge more valuable than ever, and that has increased the value of learning from other people in cities.

  It takes time to see the far-reaching, systemic effects of new technologies, so it makes sense to look at the long path of history, over which increases in the ability to communicate at long distances have generally made cities more important. No modern innovation can equal the printing press in its impact on long-distance communication. The ability to put words on paper cheaply and in great quantities was a seismic shift in mankind’s ability to communicate with people who weren’t in the same room. Yet there is no reason to think that books hurt cities and every reason to believe that the printing press helped to create a more urban world.

  The most obvious reasons that books helped cities are that printing technology was developed in cities and cities are natural centers of publishing. Gutenberg, who grew up in the first years of the fifteenth century, set out to create a printing press with the secrecy of a medieval alchemist, but a piece of machinery as bold and as expensive as a printing press could never have been created by a solitary genius. Gutenberg needed financial backers and assistants, and they were found in cities. After his breakthrough, the technology of movable-type printing soon spread from town to town, carried by itinerant merchants, and by the 1480s, Venice had become the world center of printing. Cities have an edge whenever a technology, like printing, relies on expensive forms of infrastructure, like a printing press. Large urban markets make it easier to cover the fixed costs of these new technologies, which is one reason why telephones and broadband technology, as well as printed books, became available in cities first.

  The city’s rich, literate population provided plenty of local demand for books, but Venice also thrived because it had a ready supply of material worth printing. The city’s position at the crossroads of East and West gave it a ready supply of scholars, like the Byzantines who fled to Venice after Constantinople fell to the Ottomans in 1453 and started translating for its presses. In later centuries, New York came to dominate printing in the United States because it had access to pirated English novels coming into its port and because the city attracted a vast number of wr
iters and artists.

  But the book didn’t help cities just by boosting their publishing industries. The printed word also made the world more urban in subtler, deeper ways. One direct effect of the printing press was allowing far-flung farmers to read the Bible, but indirectly the printing press helped make the world more knowledge intensive, more democratic, more commercial, and ultimately more urban. Martin Luther described the printing press as “God’s highest and extremest act of grace,” because the Bible, translated into German by Luther himself, provided a source of religious authority other than Catholic tradition and played a crucial role in the Reformation: “Between 1517 and 1520, Luther’s thirty publications probably sold well over 300,000 copies.... Altogether in the spread of religious ideas, it seems impossible to exaggerate the significance of the Press.” The Reformation, then, supported economic, political, and social changes that made commerce in cities more attractive. Max Weber famously connected Protestantism with the spirit of capitalism and the ethical values of urban merchants and artisans. Personally, I don’t think that Protestantism has an inherent superiority in supporting cities, trade, or democracy, which all flourish today in many Catholic nations. Instead, I believe that the post-Reformation rise of cities, trade, and democracy reflects the value of religious competition, which meant more choice over church rules and doctrine and led to reforms, like the erosion of usury laws, that aided the rise of global commerce.

  The printing press both directly and indirectly—through the Reformation—also supported revolutions that created a more republican and more urban Europe. The great Dutch revolt started in 1566 near the Flemish clothproducing town of Steenvoorde when a Calvinist mob destroyed the statues of a local Catholic church. In 1581, using language that would be familiar to later English, American, and French revolutionaries, the Dutch declared that King Philip of Spain had acted illegally and thereby lost his right to rule Holland. This revolutionary Act of Abjuration drew on a recently published Protestant (Huguenot) tract. The Act itself was printed and widely posted throughout the Low Countries to bolster opposition to Spain. After almost seventy years of fighting, Holland became an independent republic, the most urban nation in Europe, and the center of a global trading network that would reach as far east as Nagasaki, Japan, and as far west as the island of Manhattan.