Triumph of the City

by Edward Glaeser

  Cross-cultural communications are always complicated; things are always lost in translation. New ideas from different continents can be so unlike our current knowledge that we need to make huge intellectual leaps, which invariably means that we need plenty of coaching. We may understand the context of ideas in our society, but we are often adrift when confronted with thoughts that came from a totally different society, like the translators of the Sindhind, who didn’t understand the Euclidean math that lay behind it.

  Cities, and the face-to-face interactions that they engender, are tools for reducing the complex-communication curse. Long hours spent one-on-one enable listeners to make sure that they get it right. It’s easy to mistakenly offend someone from a different culture, but a warm smile can smooth conflicts that could otherwise turn into flaming e-mails. Cities like Nagasaki, Baghdad, or Bangalore, which specialize in international connection, develop communications experts who become adept at importing information. Such cities are convenient spots for foreigners to sample a host society’s science, art, and commerce, and vice versa.

  The success of places like Bangalore is not only about international intellectual connections. These cities create a virtuous cycle in which employers are attracted by the large pool of potential employees and workers are drawn by the abundance of potential employers. So firms come to Bangalore for the engineers, and engineers come for the firms. Urban scale also makes it easier for workers to move from job to job. In highly entrepreneurial industries, workers get ahead by hopping from firm to firm. Young people become more productive and better paid as they switch employers and acquire new skills. An abundance of local employers also provides implicit insurance against the failure of any particular start-up. In Bangalore, there’ll always be another software company. Moreover, dense concentrations of entrepreneurial talent encourage the growth of related industries, like the venture capitalists who work near Silicon Valley.

  The forces that compel concentration in a single city are clear, but it isn’t obvious why any particular city should emerge as a hub of information transfer. Why did Bangalore, out of all Indian cities, achieve this status? Bangalore does have a relatively benign climate—drier than Mumbai and much less oppressive than Delhi. But skills, not geography, are the source of Bangalore’s strength. An initial kernel of engineering expertise attracted companies like Infosys, and a virtuous circle was born, wherein smart firms and smart workers flock to Bangalore to be near each other.

  Few have gotten more out of Bangalorean proximity than the city’s three Infosys billionaires. Infosys was founded in 1981 and moved to Bangalore in 1983. In the summer of 2008, the company had close to a hundred thousand employees, and its market capitalization exceeded $30 billion. Today Infosys is a flat-world phenomenon, with vast operations in software, banking services, and consulting. In essence, Infosys is selling intelligence—whether provided by humans or machines—at lightning speed around the world, and it takes its employees’ skills seriously, educating thousands of people each year in its training center in Mysore. Fewer than 2 percent of Infosys’s job applicants get a place in that training center, making it far more competitive than any Ivy League school.

  Narayana Murthy, one of the Infosys founders, received engineering degrees from the University of Mysore and the Indian Institute of Technology at Kanpur. But Murthy may have picked up his most valuable skills in the 1970s at Patni Computers. Patni was a bridge company, an early connector between the United States and India, whose Indian founders had lived in America. They saw the opportunities for Indian software and set up a back office in Pune. Murthy worked there with the six other founders of Infosys, where they learned how to link Indian talent with American markets.

  In 1981, they left Patni to found their own company selling software to foreign clients. Murthy borrowed $250 from his wife to cover expenses. In 1982, they acquired their first American client, a software company. In 1983, they moved to Bangalore to provide software for a German spark plug producer that had located there back in 1954 and wanted Infosys close by so that information could flow readily between the two companies. Infosys was also attracted to Bangalore because it was near top-notch engineering schools.

  Over the past twenty-five years, Infosys has opened offices in the United States, Canada, Latin America, and Europe, but it remains rooted in Bangalore. The rise of Infosys may seem to suggest that distance is dead, but it can just as easily be interpreted as evidence that proximity matters as much as ever. By concentrating so much talent in one place, Bangalore makes it easier for outsiders, whether from St. Louis or Shanghai, to do business with Indian entrepreneurs. Bangalore may be luckier than any other Indian city, but only because it made its luck. Its current abundance of engineers reflects decisions made long ago by its leaders, the maharajas of Mysore and their ministers. Mysore had a long tradition of embracing new technologies. In the eighteenth century, its sultan dealt the British a fearsome defeat with the help of imported cannons, manned by imported sailors. Throughout the Raj, Mysore stood out among the princely states for its competence, but the savviest of its leaders was Sir Mokshagundam Visvesvaraya, or Sir MV, the state’s prime minister during the early twentieth century.

  Sir MV was born about thirty-five miles from Bangalore and came to the city for high school. After an illustrious career as a civil engineer, he returned to Bangalore and in 1908 became the prime minister of Mysore. Along with the maharaja, who was both fabulously wealthy and remarkably progressive, Sir MV pushed through a sweeping modernization program, including dams, hydroelectricity, steel mills, and, most important, schools. Sir MV’s motto was “Industrialize or perish,” but instead of just pushing big construction projects, he emphasized the education needed to build projects efficiently. Infrastructure eventually becomes obsolete, but education perpetuates itself as one smart generation teaches the next.

  In the United States and Europe, industrialization rarely encouraged education. Much of factories’ appeal for owners and workers alike was that they gave jobs to unskilled labor, not skilled artisans. But for Sir MV, industrialization meant training the engineers who could import technology from the West, just as he had done. He founded both the University of Mysore and Bangalore’s engineering college, which now bears his name. Those schools first generated a cluster of engineers that persists to this day.

  By the middle of the twentieth century, Mysore was fully industrialized. Its probusiness government brought Hindustan Aeronautics Limited, Hindustan Machine Tools, Bharat Heavy Electricals, and Indian Telephone Industries to Bangalore. It also attracted the German spark plug producer that would later bring Infosys there. Those early companies were important not because Bangalore’s future lay with heavy industry (it did not), but because they nurtured that cluster of engineers. Starting in 1976, Bangalore also paved (sometimes literally) the path to IT dominance by launching an extensive program to improve roads, electricity, and other utilities that would attract international IT firms.

  Education and Urban Success

  Human capital, far more than physical infrastructure, explains which cities succeed. Typically, in the United States, the share of the population with a college degree is used to estimate the skill level of a place. Admittedly, this yardstick is imperfect at the individual level. Using a college degree as a measure would classify Bill Gates, surely among the world’s most skilled people, as unskilled. But despite its coarseness, no other measure does better in explaining recent urban prosperity. A 10 percent increase in the percentage of an area’s adult population with a BA in 1980 predicts 6 percent more income growth between 1980 and 2000. As the share of the population with college degrees increases by 10 percent, per capita gross metropolitan product rises by 22 percent.

  People have flocked to skilled areas because of higher incomes, and education in 1970 does an impressive job of explaining which of America’s older, colder cities have managed to successfully reinvent themselves. Between 1970 and 2000, the population of counties where more than
10 percent of the adult population had college degrees grew by 72 percent while the population of those areas where fewer than 5 percent of people had college degrees grew by 37 percent.

  We live in an age of expertise, when earnings and knowledge are closely linked. For each worker, an extra year of schooling typically leads to about 8 percent higher earnings. On average, an extra year of schooling for a country’s entire population is associated with a more than 30 percent increase in per capita gross domestic product. The striking correlation between education and a country’s GDP may reflect what economists call human capital externalities, a term for the idea that people become more productive when they work around other skilled people. When a country gets more educated, people experience both the direct effect of their own extra learning plus the benefits that come from everyone around them being more skilled.

  The connection between urban skills and urban productivity has grown steadily stronger throughout the developed world since the 1970s. In those days, less-skilled places that were filled with highly paid, unionized factory workers often earned more than more-skilled areas. In 1970, per capita incomes were higher in industrial areas like Cleveland and Detroit than in better-educated metropolitan areas like Boston and Minneapolis. Over the past thirty years, however, the less-skilled manufacturing cities have faltered while the more-skilled idea-producing cities have thrived. In 1980, men with four years of college earned about 33 percent more than high school graduates, but by the mid-1990s, that earnings gap had increased to nearly 70 percent. Over the past thirty years, American society has become more unequal, partly because the marketplace increasingly rewards people with more skills.

  While no one disputes the robust increase in the value of skills, there are competing theories about why they’ve become more valuable. One school of thought emphasizes technological change. Some new technologies, like computers, have increased the returns for being better educated. Other new technologies, like robots in car factories, have decreased the need for unskilled labor. Not just the technologies themselves, but the rate of technological change also favors the skilled. Many studies have shown that skilled people are better at adapting to new circumstances, like the introduction of hybrid corn and computers. Like skilled people, skilled cities also seem to be better at reinventing themselves during volatile times.

  A second school of thought emphasizes international trade and globalization. According to this view, declining transportation costs made it possible to outsource less-skilled labor. Detroit’s carmakers once had a near-monopoly on American auto purchases, but today those companies face intense competition from Japan, Europe, and Korea, and this makes it much harder to sustain high wages for less-skilled workers.

  Of course, more skilled jobs are being outsourced as well. That’s one reason for Bangalore’s success. Yet so far at least, skilled Americans and Europeans seem to have gotten more from the ability to work the world market than they’ve lost from foreign competition. The most-skilled people in the rich countries have thrived by selling their ideas to the world and by using worldwide labor to produce their inventions more cheaply. The software producers in Bangalore haven’t made Silicon Valley obsolete. Instead, they’ve made it cheaper—and thus easier—for Silicon Valley firms to develop software.

  The Rise of Silicon Valley

  America’s greatest information technology hub is Santa Clara County, California, which most people know better as Silicon Valley. Much like Bangalore, the Valley achieved this status by making its luck with education. A century ago, when New York and Nagasaki were old, computers didn’t exist, and Santa Clara County was covered in orchards and farms. This agricultural community became a world capital of high technology because Senator Leland Stanford, a railroad magnate, decided to build a university on his eight-thousand-acre horse farm.

  Founding universities was, like breeding horses, a way for nineteenth-century millionaires to spend their surplus money. My University of Chicago diploma displays, in appropriately gilt letters, the name of the school’s Gilded Age founder, John D. Rockefeller. But while Rockefeller envisioned a Baptist college and hired a classicist as president, Leland Stanford opened Stanford declaring that “life is, above all, practical; that you are here to fit yourselves for a useful career.” He wanted leaders who were committed to the real world, to developing the American West, and to spreading useful knowledge.

  Stanford University’s first major high-tech start-up had its roots in the unschooled genius of Francis McCarty, the son of Senator Stanford’s head coachman. McCarty left school at twelve to work as an apprentice electrician. In 1904, at the age of sixteen, he had crafted a “spark telephone” that could send a voice seven miles over water. McCarty wasn’t the first to send words by wireless, but he was close, and his brilliance brought financial backing. Tragically, in 1906, McCarty died in a traffic accident, smashing his head against a telephone pole. He wasn’t even eighteen.

  But his backers hadn’t lost their appetite for radio, and they asked a Stanford engineering professor to recommend a suitable replacement for McCarty. He pointed them to Cyril Elwell, a bright Stanford student who had written his dissertation on electric smelting. Elwell proved an inspired choice. He worked for a year on McCarty’s design and concluded that it couldn’t provide reliable wireless service. But instead of giving up, Elwell opted for an even newer technology: Valdemar Poulsen’s arc transmitter. Elwell sailed to Copenhagen and came back to Palo Alto with a Poulsen transmitter. With financial backing from the president of Stanford, Elwell then set up the Poulsen Wireless Telephone and Telegraph company, soon renamed the Federal Telegraph Corporation.

  FTC was the pioneer firm of Silicon Valley’s radio industry, attracting talent and producing spin-offs. Lee De Forest, the inventor of the audion transmitter, came to FTC in 1910 when his own company went bankrupt. There he developed the first vacuum tube, a critical part of radio technology until 1947, when another product of Palo Alto, William B. Shockley, led the group that invented its replacement, the transistor. Even after De Forest left, FTC thrived with navy contracts and access to Stanford’s talented students. Stanford’s first PhD in electrical engineering was awarded on the basis of work done at FTC.

  Like later Silicon Valley firms, FTC produced distinguished progeny. Two Danes who had come to Palo Alto to help with Poulsen’s arc transmitter left to form Magnavox. Another FTC employee invented the first metal detector and started Fisher Research Laboratories. Litton Industries, which grew large by producing vacuum tubes for the military during World War II, was yet another FTC offspring.

  But no FTC employee did more to make Silicon Valley what it is today than Frederick Terman, who connected with the company as a kid and worked there during his college summers. His father was a Stanford professor who specialized in gifted children like his own son. The elder Terman became famous for developing the Stanford-Binet IQ test. The younger Terman went to Palo Alto High and Stanford, then headed east to get his doctorate in electrical engineering at MIT in 1924. He became a Stanford institution, serving for forty years as professor, dean of engineering, and provost, but his greatest gift was turning Palo Alto into the center of the computer industry.

  One advantage enjoyed by a university surrounded by orchards is an abundance of available land, and Terman got the idea to start an industrial park right next to Stanford. His vision, which would inspire technology-intensive clusters in Bangalore and throughout the world, was to create an area packed with technology businesses. His students David Packard and William Hewlett were two early tenants in Terman’s industrial park, but he couldn’t achieve critical mass relying solely on his own protégés. He sought tenants like Lockheed, General Electric, and Westinghouse. Most important, he convinced the new Shockley Semiconductor Laboratory to come to the valley.

  William Shockley was already a legend in the mid-1950s. Like Terman, his father had taught at Stanford. The young Shockley actually did poorly on an IQ test given by Terman’s father, which sa
ys something about the fallibility of IQ tests. Shockley was educated at MIT and then worked at Bell Labs in New Jersey. After earning a medal for his wartime work using technology to fight U-boats, Shockley was put in charge of Bell Labs’ new solid-state physics research group. This group collectively invented the transistor, and in 1956, Shockley and two of his co-workers shared the Nobel Prize in Physics.

  By that time, Shockley had left Bell Labs and headed out to California, where his enormous abilities—and a fatal flaw—would both assert themselves and both contribute to the success of Silicon Valley. Like Pericles and the Abbasid caliphs, he had a rare talent for attracting geniuses. In his first years, he searched America’s campuses and brought great young minds eager to come to Silicon Valley and work with the Nobel laureate. But Shockley was a capricious and dictatorial manager who couldn’t keep the talent that he had attracted. In one notorious incident, he made his workers take lie detector tests in order to establish who was responsible for a secretary’s cutting her hand on a pin. By attracting and then repelling genius, Shockley both brought talented people to Silicon Valley and ensured that they would be starting their own firms instead of just working for him.

  At one point, eight of his best young scientists collectively quit. A cameramaking magnate named Sherman Fairchild bankrolled them, and Fairchild Semiconductor was born. The firm stayed in Silicon Valley. Why would the “traitorous eight” want to leave a paradise packed with Terman-trained engineers? In 1959, Fairchild Semiconductor patented the first integrated circuit. Eventually, the talent also tired of Fairchild’s management. Two of them left Fairchild in 1968 to form Intel. Another left to form the venture capital giant Kleiner Perkins, which would bankroll many of the Valley’s next wave of innovators.