The Big Sort

by Bill Bishop

  City life was key for such economies of ideas to flourish, because cities sped both the creation and the spread of useful knowledge. That insight came from a woman who never graduated from college and was best known for her observations about city planning. In 1969, Jane Jacobs published the first edition of The Economy of Cities, a book that described how ideas scattered through urban areas, creating new industries and new wealth.* Jacobs wrote that new wealth and businesses didn't stem from isolated discoveries or unique inventions. Innovations sprang from older lines of work—"parent work," she called it. Consider, for example, the brassiere. Ida Rosenthal was a custom seamstress in early-twentieth-century New York, as Jacobs told the story. In trying to find ways to make her goods better fit her customers, Rosenthal began experimenting with different garments to be worn underneath her dresses. A result of her innovations was the brassiere. Her customers liked the new undergarment so much that in time, Rosenthal stopped making dresses altogether and went into the business of sewing undergarments full-time. She called her company Maidenform Brassiere.36

  The Economy of Cities is filled with examples of how a series of small innovations can spawn entirely new industries. The Minnesota Mining and Manufacturing Company (3M) was a small firm founded in 1902 to produce the sand used as an abrasive in the production of metal casings. To find other uses for its primary product, 3M tried making its own sandpaper (a product already invented). The adhesives the company was using failed, so 3M experimented with sticky substances. That's where the new work began to appear. In a series of innovations, 3M first produced a gummed paper that house painters found useful (masking tape), then electrical tape, Scotch tape, plastic tape, and sound recording tape.37

  Sure, there had been significant scientific discoveries, Romer said. Penicillin, for example, hadn't been cobbled together on the shop floor. But great advances didn't often stem from fundamental scientific research. If that were the case, Romer wrote (echoing Jacobs), the study of thermodynamics would have led to the development of the steam engine. The reverse happened instead: the people who solved the practical problems of making steam engines uncovered the basic principles that led to the science of thermodynamics.*38

  Cities were essential in the work of innovation not because they were efficient or practical, because they certainly were neither. They were essential because economic development was the social process of problem solving, discovery, and innovation. Henry Ford failed twice in his single-handed attempts to manufacture an entire automobile. He succeeded when he pieced together parts from other small businesses around Detroit. This process of making connections and sharing ideas worked best when people involved in related kinds of work could meet face-to-face and share new ideas.39 Almost twenty years after Jacobs wrote The Economy of Cities, academic economists developing new growth theory adopted her observations. If cities were simply accumulated sets of production factors (capital, people, and land), economist Robert Lucas wrote, they would "fly apart." Land was always cheaper outside cities. Cities were crowded and inefficient, crazily so. By any ration of common sense, everyone should move out. But they didn't. In fact, the opposite happened. Firms paid higher rents to be in cities than elsewhere. What could these businesses be paying for, Lucas reasoned, "if not for being near other people?"*40

  Research in the 1970s and 1980s explored how economies were social enterprises. A grad student in sociology in the early 1970s was curious about how people found out about new information. Mark Granovetter read a study of a southern textile mill swept by a "hysterical contagion." The first people to report being bitten by a mysterious insect were workers at the mill who had few close friends. (The biting critter was, in fact, imaginary.) The "disease" started at the social margins of the mill and then moved into the more conformist social center of the factory's workforce. Granovetter theorized from this case that new ideas (good and bad) didn't start where there were a lot of strong, stable social ties. New behaviors and ideas were transmitted first and fastest through weak, peripheral relationships. Granovetter wanted to discover how people learned about new job openings. Based on the example of the southern textile mill, he predicted that finding a new job wasn't a function of having strong friendships. The best bet for hearing of a new opening was to have a slew of weak relationships. Strong ties between people bottled up information, Granovetter theorized. That made sense. Your friends know what you know. That's why they're your friends. "Weak ties" tapped new pools of information—and weak ties allowed that information to flow.

  To test his hypothesis, Granovetter went to a Boston suburb and in terviewed people who had recently found new jobs through personal contacts. Granovetter asked his subjects how often they saw the person who had told them about the job. Only 17 percent of those newly employed people said they were often in contact with their source, 56 percent said they occasionally saw their source, and 28 percent said they saw their source rarely. Granovetter described this phenomenon as the "strength of weak ties." The more bridges connecting a person to others, no matter how tenuous or seldom used, the better his or her chances for employment. The same would hold true for a community, Granovetter proposed. Communities with lots of internal bridges, even if they were flimsy ones, would be better off than places with social groups that were tightly knit internally but unconnected to other groups in the community.41

  Historian Peter Hall examined the dynamic cities in history—Manchester, England, in the late eighteenth and early nineteenth centuries, San Francisco in the second half of the twentieth century, Berlin from 1840 to 1930—and found that they were constructed on the strength of weak ties. "Most of these places seem to have had egalitarian social structures: they were unstuffy, un-classbound, non-hierarchical places," Hall wrote.42 At the turn of the twentieth century, all the newly discovered technology in the entertainment industry—pictures that both moved and talked—was located in New York, along with most of the country's acting talent. If the economy of motion pictures depended on the accumulation of valuable stuff, the industry should have flourished in New York. Instead, the movie business took root in a no-place town called Hollywood. The reason wasn't just because of southern California's sunny weather, according to Hall. After all, there was sun aplenty much closer to New York in Florida or Cuba, where filmmakers first traveled to escape New York winters. Hollywood was a good place for a new industry to grow—especially a new industry headed largely by immigrant Jews—because it lacked an established hierarchy or social aristocracy. Los Angeles wasn't balled up with those strong social connections that can kill gumption, trap new ideas, and suffocate innovation.43

  Why, in the early 1990s, did Silicon Valley add more jobs and create more new businesses than the high-tech cluster along Route 128 in Boston? Thinking more like an anthropologist than an economist, AnnaLee Saxenian found the answer in the Wagon Wheel restaurant in Mountain View, California, where people from different firms would meet, mingle, and trade information. She noticed that people switched jobs in Silicon Valley often and with impunity. As they flitted from one office to another, they pollinated their new firms with ideas they'd gathered along the way. In Boston, Saxenian observed firms that were more cylindrical, hardened, and isolated. People tended not to change jobs. Firms preferred employees who were "in for the long haul." There weren't many bridges between businesses. There were few natural meeting places for people from different firms to mix, no Wagon Wheel restaurant. There were fewer weak ties.44

  Cities Come Back—with Weak Ties

  After the disastrous 1970s, when the future of cities seemed bleak, people returned. In the 1980s, housing prices began to rise in the same cities that had experienced drops in the previous decade. New York City gained population. Personal income in cities increased, too, reversing declines in the 1970s. Not all cities added people and income. Edward Glaeser discovered that there was a complete switch in the types of cities that grew. In 1950, seven of the eight largest cities had a higher percentage of people employed in manufacturing th
an the nation as a whole. By 1990, six of the eight largest cities had proportionally less manufacturing than the country as a whole. The urban revival wasn't led by places that made things, but by cities that produced ideas.

  When Glaeser examined the urban resurgence of the 1990s, he could see that falling crime rates helped spur a return to the cities. But safe streets alone didn't cause the urban renaissance. Glaeser found an increased demand for what cities created in abundance, and that was human interaction. The demand was economic. Face-to-face discussions may not have been important in a Henry Ford—style assembly line, but they were essential to an economy based on the development and rapid spread of new ideas. The tightly wound mass production economy that shaped American society through the 1950s was shrinking, replaced by firms specializing in knowledge. A certain group of cities had a culture and a way of life that supported this kind of economy.

  Cities mattered again. Glaeser found that there had been both a "remarkable increase in the importance of knowledge" in the 1990s and an increasing economic demand for cities that "facilitate social interactions."45 Other researchers had discovered that inventors tended to cite patents produced by other inventors who worked nearby. Patent citations decreased as physical distance between inventors widened. By the late 1990s, the tendency of inventors to cite others from the same city in their patent applications was increasing. Proximity was becoming more important in the creation and transmission of ideas.46 In late 2006, a Silicon Valley business professor named Randall Stross wrote about the "twenty-minute rule" adopted by several venture capital firms. "If a start-up company seeking venture capital is not within a 20-minute drive of the venture firm's offices, it will not be funded," wrote Stross. Proximity translated into convenience, of course, but the venture capital firms also knew that frequent face-to-face contact was essential. And despite world-flattening technology, there was something magical about place. Craig Johnson, managing director of a Silicon Valley venture capital firm, told Stross that having all the entrepreneurial ingredients nearby created an intensity that was missing from more spread-out geographies. Los Angeles, after all, had great universities and talent and boatloads of money. "But in Los Angeles, people are scattered across a wide area; everything is spread out," Johnson explained. "Like a gas, entrepreneurship is hotter when compressed."47

  The busting of the tech bubble in the early 2000s slowed the growth in these cities, but only temporarily. By early 2007, demographer William Frey was telling a "tale of two kinds of cities." Those producing patents and technology were expanding again; the older cities that made things—Buffalo, Cleveland, Toledo, Detroit—were reporting continuing declines.48

  The Rise of the Superstar Cities

  People migrate to maximize their economic returns. At least that's been the most common explanation for why people move. They pull up stakes to make or save money. Economic utility seemed to be motivating both employers and workers in the 1970s; then, big-city businesses were paying higher wages to keep and attract workers. But by 1990, there was no wage advantage to living in one of the twenty-six largest metro areas. By 2000, it cost more to live in those places than any premium received in wages.49 According to traditional economics, when this kind of imbalance occurs, either wages rise or people stop moving to such overpriced areas. But the opposite was happening. People were moving to places that cost more and where, in comparison to other cities, they would be earning less. Additionally, defying every economic precept, six out of ten well-educated young adults moved to places with slower job growth than the cities they left.50 Weird things were going on in urban America. Cities were no longer places whites fled. Instead of people jamming the roads from homes in the suburbs to jobs in the central city, people were reversing the direction, commuting from downtown homes to workplaces outside the city. By 2007, the number of whites living in Washington, D.C., was increasing, while the number of blacks was decreasing.51 There was a surge of people who wanted to live in cities for what could only be social—or even aesthetic—reasons.52

  Economist Joseph Gyourko compiled a list of metro areas so attractive, so in demand over the past two decades, that housing prices had far outstripped any wage advantages the cities might offer. Housing prices in much of America rose modestly during this time. In the sought-after cities, however, they shot out of sight. In 1940, for example, the average house in Cincinnati cost more than a home in San Francisco. Over the next sixty years, the price of a Cincinnati house increased from $65,000 to $145,000. The total population of San Francisco had changed little in that time, but the average house price rose by a factor of nine, from $60,162 to nearly $550,000. Gyourko dubbed these places "superstar cities," metro areas where residence had become, in essence, a luxury good.53 People paid for the privilege of being in cities such as San Francisco, Seattle, San Jose, Portland, Los Angeles, New York, Austin, and Raleigh-Durham because they wanted to live there, not because they expected an economic return.

  The function of cities had changed. Their reason for being—and their residents' reason for living within them—was no longer to produce salable goods and services. The city's new product was lifestyle. Glaeser called these metro areas "consumer cities"—metro areas that catered to the well-paid, well-educated people who moved there.54 Richard Florida found that these cities attracted "creative-class" workers.55 The University of Chicago's Terry Nichols Clark proposed that cities had become "entertainment machines."56 To attract the people who would power their economies, cities had first to produce the sounds, sights, tastes, and experiences desired by post-materialists engaged in their part of the Big Sort. These new social and economic arrangements had changed a fundamental American assumption: that citizens have the right to live in whatever city they choose. Just as not everyone could own a Mercedes-Benz, Gyourko wrote, Americans no longer had the right to live in San Francisco, central Austin, or northwest Washington, D.C. Both the car and the cities were extravagances.57

  Meanwhile, the way Americans sorted themselves created a new kind of cultural separation. People living in different cities literally had different ways of relating to family, to government, to strangers, and to religion. This gave people another choice. There were places where they could enjoy the comfort of strong families, bustling civic groups, near universal political participation, and abundant volunteering. And there were cities that offered anonymity, the opportunity for self-invention, and the economic benefits of loose ties.

  Using Putnam's data, we discovered that high-tech cities had cultures that differed from manufacturing towns. The high-tech towns had fewer volunteers, lower church attendance, and weaker family connections. We then found that these differences extended to politics. High-tech cities were slightly more liberal than the nation as a whole prior to 1990, but since then they have become more liberal.* Before 1990, people living in low-tech cities described themselves as slightly more liberal than the national average, but after 1990 an increasing number labeled their politics as conservative. Prior to 1990, high-tech cities were at the national average in terms of party identification. After 1990, these places were Democratic strongholds. Manufacturing cities and rural areas moved in the reverse direction, growing more Republican.

  These shifting political cultures emerged over time in presidential elections. Bob Cushing and I divided U.S. metro areas into five groups with descending levels of high-tech and patent production and then compared how these groups of cities voted in the six presidential elections from 1980 to 2000. In the earliest election, all the city groups voted much the same. The twenty-one high-tech areas were slightly more Democratic than the nation as a whole. Suburban cities adjacent to these high-tech areas (places such as Boulder outside Denver, Orange County outside Los Angeles, and Galveston outside Houston) were slightly more Republican than the national average. But in 1980, the vote in all these areas approximated how the nation voted as a whole.

  As time passed, voting patterns in the city groups diverged. The high-tech group tilted increasingly De
mocratic, so that by 2000, these twenty-one cities were voting Democratic at a rate 17 percent above the national average. (Take out the Texas tech cities—Austin, Houston, and Dallas—and the remaining eighteen metro areas were voting Democratic at a rate 21 percent above the national average.) The cities adjacent to the high-tech hubs flipped altogether, turning strongly Democratic as a group. (This was true even with the inclusion of still-Republican Orange County.) The low-tech cities and rural America grew increasingly Republican. The exceedingly close 2000 election ended with Democrats leading by a half million votes nationally. With the vote broken down city by city, however, the election was a series of local landslides that aligned with local economies. Al Gore was the winner in the high-tech cities. He led George W. Bush in these metro areas by more than 5.3 million votes. Bush made up this deficit in low-tech cities and rural America. This pattern reappeared in 2004, when John Kerry outpolled Bush by more than 5 million votes in the high-tech cities, in an election the Republican won by more than 3 million votes.

  The near fifty-fifty results in presidential politics belied the divisions appearing locally. Young voters in rural eastern Oregon abandoned the Democratic Party. In Pendleton, home of one of the country's best rodeos, only 21 percent of voters twenty-one to thirty-five years of age were registered as Democrats in 2004. Among Pendleton residents a generation older, from forty-five to sixty-five years of age, 35 percent were registered Democrats. In Portland, however, young voters were going in exactly the opposite direction, and fast. In two typical inner Portland precincts in 2005, the percentage of young people (ages eighteen to thirty-five) registered as Republicans dipped to 5.4 and 7 percent.58