New York City has become too dependent on the financial industry. In 2008, 44 percent of Manhattan wages were earned by workers in finance and insurance; the following year, even after the financial crisis and economic downturn had battered the industry, that share stood at a still-enormous 37 percent. And the track record of one-industry towns isnt good. No matter how loudly Chryslers provocative Super Bowl ad heralded Detroits comeback, the Motor Citys population dropped by a quarter over the last decade and now stands at 39 percent of its 1950 peak. In Russia, Soviet-era monocities like Norilsk, a mining hub, are emblems of urban decline. Economic data, bearing out what those examples suggest, show a positive link between industrial diversity and long-run urban success.
New York shouldnt try to hold finance back, of course, but it should try to reduce the cost and regulatory barriers that limit the growth of other sectors. If Gotham hopes to keep playing its historical role in leading the worlds economy, it needs to welcome companies in other fields—most likely, technology, business services, and a broad range of information-intensive industries.
I spent my childhood in Manhattan, from 1967 until 1984. New Yorks economy was far more diverse then than it is today. My friends parents were hardly a proper cross-section of the city, but they nevertheless represented a remarkable array of different industries: there were editors and philosophers, art dealers and jewelers, judges and doctors. Show-business parents, including Broadway composers and a character actor best known for a modest role in Shaft, added glamour. Developers and landlords added grit. I cant recall a single investment banker in the bunch.
But then, the city had been diverse for centuries. New Yorks first big industry, back in the eighteenth and nineteenth centuries, was sugar refining. But sugar was never going to dominate the city the way steel came to dominate Pittsburgh, because refined sugar couldnt be shipped very far, which limited the size of the market. Similarly, printing and publishing—New Yorks second big nineteenth-century industry—could grow only so much, since demand for the printed word in frontier America was limited. No such barriers limited the expansion of New Yorks third and greatest industry, garment production, since clothes are portable and represent a significant share of household budgets. But even then, New Yorks economy remained diverse: in 1950, at the height of its power, the garment industry represented only 8 percent of city employment.
The most important reason for New Yorks continuing economic diversity was the citys massive scale, which generated plenty of homegrown entrepreneurship and attracted entrepreneurs from elsewhere as well. New York was an early hub of automobile production and the film industry, for example, and though they eventually left town, they helped make the city diverse during their early years. Even when companies were born in the Midwest, their chieftains often moved their headquarters to New York to be part of a great agglomeration of business services and financiers, as John D. Rockefeller did with Standard Oil in 1885. The oil industrys presence in New York infused the city with a little prospecting swagger and gave the oil industry a taste for culture, explaining why Mobil financed Masterpiece Theater programs for years.
Most of Americas older ports—Boston, Philadelphia, Baltimore, San Francisco—shared New Yorks industrial variety. Americas inland cities were another story. Most of them exploded in the nineteenth century because they offered a huge natural advantage that couldnt be ignored, such as nearby coal mines or cornfields. St. Louis, Cincinnati, Chicago, and Minneapolis, all in the grain business in a big way, were dominated by agriculture-related firms.
These inland cities became even less diverse in the early twentieth century, when the country moved to manufacturing. Industry located in the inland metropolises for several reasons: sometimes to be near local entrepreneurs, like Henry Ford; sometimes because the older ports were too pricey for the acreage-intensive manufacturing of large industrial products; and sometimes because industry needed local inputs that were easily available in the Rust Belt. The steel industry in Pittsburgh, for example, relied on nearby coal mines. So did Buffalos steel industry, which also took advantage of iron ore from Minnesota shipped along the Great Lakes. Michigans abundant forests provided the wood for Billy Durants Flint Road Cart Company and later for his car company, General Motors.
The twentieth centurys heavy industries faced few natural limits to growth, and falling transportation costs meant that their products could be sold throughout the country, if not across the globe. By 1950, the midwestern cities found themselves heavily dominated by one or two large industries. Automobile production boasted 28 percent of Detroits total employment; metal and machinery manufacturing claimed 20 percent of Clevelands. Again, New Yorks garment industry, massive though it was, represented less than 10 percent of the citys total employment.
Finance has existed in New York for centuries, but its current dominance dates to the late 1970s, when it was a crucial component of the troubled citys resurgence. Over the next few decades, Manhattan financiers pioneered innovations—quantitative approaches to evaluating risk; ever-larger leveraged buyouts; the securitization revolution—that made finance considerably more lucrative. Just as Henry Fords immense success had led automobile production to dominate early-twentieth-century Detroit, Wall Street earnings meant that finance played an ever-larger role in late-twentieth-century and early-twenty-first-century New York.
And just like Detroits auto industry, New York finance became concentrated in fewer, bigger firms. In 1998, Manhattan had 7,313 establishments in the narrower financial sector that the U.S. Census calls securities, commodity contracts, investments, and each employed an average of 22.4 workers. By 2008, Manhattan was down to 4,919 firms in that sector, with an average of 40.3 workers apiece. During the same ten-year period, the number of companies in that sector with more than 1,000 workers rose from 19 to 33.
As finances success drove up rents, many businesses in other sectors had to leave Manhattan. Between 1998 and 2008, the island lost more than 75,000 jobs in manufacturing, transportation and warehousing, and wholesale trade. Offsetting that decline was a gain of more than 100,000 jobs in consumer industries— retail, food and accommodation, and arts and entertainment—catering to well-heeled residents and tourists. (While thats diversification of a sort, its hard to imagine that Manhattan can sustain itself primarily as an entertainment hub.)
By 2008, as I noted above, workers in finance and insurance were earning 44 percent of all wages in Manhattan. That startling statistic does somewhat overstate the power of finance in New York, though. For one thing, the industrys role was much smaller in the four boroughs outside Manhattan, where it accounted for only 6 percent of payroll in 2008; take those boroughs into consideration, and youll find finance representing 37 percent of the entire citys 2008 payroll. Also, because finance and insurance workers are often paid extremely well, the sectors share of Manhattans payroll was considerably higher than its share of Manhattans employment—in 2008, 16 percent, which does make the island look less like a one-industry place. Still, the sheer size of the financiers salaries means that the success of other industries, from the restaurant business to real estate, is tied closely to the success of finance. Just look at the doldrums that the whole city enters when Wall Street bonuses fall short.
As I also mentioned above, finance and insurances share of Manhattans payroll has declined since the financial crisis. In 2009, it stood at 37 percent. Yet it isnt clear which situation, the boom or the crash, is more representative of the long-term picture. The sectors share of Manhattans payroll was 40 percent in 2006 and 39 percent in 2002; its hard to imagine that it wont resume its climb soon.
The crash does, however, illustrate one of the short-term dangers of overconcentration. Standard portfolio theory holds that since every sector has its ups and downs, you should divide your eggs into many baskets, and certainly not deposit all of them into a basket as volatile as finance. New York has ignored that principle. Between 2008 and 2009, the citys payroll fell by $35 billion—and more than four-fifths of that decline was in Manhattans finance and insurance payrolls.
Is New Yorks concentration in finance dangerous over the long term as well? Many economists would answer no; industrial concentration, they would say, enables the development of highly specialized skills. A century ago, Alfred Marshall emphasized that within industrial clusters, the mysteries of the trade become no mystery but are, as it were, in the air. The Nobel laureate Kenneth Arrow created elegant theories about learning by doing, which would seem to be enhanced when an area concentrates in cars or finance. In the 1980s, Paul Romer, a pathbreaking economist now at New York University, made sense of long-term economic growth by emphasizing the increasing returns that came from the stock of knowledge—and in dense clusters, that stock becomes richer.
In the 1980s, Silicon Valley seemed to prove these economists right. With a little help from entrepreneurial Stanford engineer Frederick Terman, who had mentored Hewlett and Packard and helped establish Stanford Industrial Park in the fifties, the region attracted many of the best minds in computing. This great concentration of talent in a single industry made it easier for entrepreneurs to connect with one another and exchange ideas—for example, after hours at Walkers Wagon Wheel, a Mountain View restaurant popular among techies.
Other economists and urbanists, however, argue that a citys long-term success depends on its hosting many industries, since real breakthroughs pull ideas from more than one field. More than 40 years ago, Jane Jacobs argued in The Economy of Cities that new ideas came from combining old ideas. Nighttime baseball combines baseball with electric lighting; graphic computer interfaces merge old-fashioned pictures with basic computing functions. Michael Bloomberg became a high-tech billionaire not in Silicon Valley but in New York, thanks to his firsthand knowledge of what technology a stock trader needed at his desk. To innovate, in Jacobss view, you often need to borrow the insights of another occupation—and since diverse cities contain many occupations, they should encourage more leaps of insight.
At its extreme, this view predicts that Silicon Valley will eventually resemble Detroit. In the short run, industrial concentration can lead to rapid leaps along a technological path. But progress along that path will eventually grow slower and yield diminishing returns, since an industrial monoculture will not encourage radically new discoveries.
About 20 years ago, three coauthors and I examined industrial clusters within cities to test the Marshall-Arrow-Romer hypothesis against the rival Jacobs view. The data supported Jacobs. High levels of industrial concentration within the clusters in the mid-1950s were associated with less subsequent growth between 1956 and 1987. We also found that clusters in more diverse cities grew more quickly and that clusters with plenty of small firms grew more quickly than those with fewer, larger ones.
For this essay, I have studied 300 metropolitan areas during the period between 1977 and today. My primary measure of industrial concentration is the share of wages in each metro area that went to that areas four largest industries in 1977. I have found that the more concentrated each area was, the smaller its percentage growth in population between 1980 and 2010, and the smaller its percentage growth in real median income between 1980 and 2000. (The changes in the 2010 census make it difficult to use comparable income data for that year.) In other words, both population growth and income growth decline steadily with industrial concentration. The 60 most diverse areas enjoyed population growth of over 60 percent and income growth of over 25 percent, while the 60 least diverse areas saw population growth of under 23 percent and income growth of under 16 percent.
Of course, the relationship between industrial diversity and long-term growth becomes substantially weaker when we control for other factors that correlate with growth: each metro areas January temperature, average establishment size, share of employment in manufacturing, and share of the population with a college degree. After controlling for those factors, I have found that a 10 percent increase in the share of payroll going to a metro areas largest four industries is associated with about 3.3 percent less population growth between 1980 and 2010 and about 3 percent less income growth between 1980 and 2000.
While the effects of industrial concentration are statistically significant and economically meaningful, they are less important than some other factors. For example, a 10 percent increase in the share of the population with a college degree in 1980 is associated with 18 percent more population growth between 1980 and 2010 and 8 percent more income growth between 1980 and 2000. This means that industrial diversity, though it should be a policy concern, isnt important enough to trump other core issues, such as schooling.
New York City owes a great deal to its financial sector, and it should wish nothing but success for its traders and investment bankers. To encourage a more diverse local economy, it should seek not to hurt finance but to expand entrepreneurship in other sectors. But which sectors would those be? The continuing decline of New York manufacturing reinforces the point that the city will never again be a great place to produce goods. Industrial balance is likelier to come from the sector that the census calls professional, scientific, and technical services, in which the number of Manhattan employees rose from 257,000 in 1998 to 285,000 in 2008. Success in that sector would make sense, since New Yorks comparative advantage is density—a particularly valuable asset for information-intensive industries.
Sound urban policies should respect the limits of government, which does not have the ability to play venture capitalist (see Urban-Development Legends, Autumn 2011). But that doesnt mean that the government cant knock down its own barriers to local success, such as those that drive up property costs. The cost of Manhattan real estate is a serious problem for potential start-ups: directly, since they must pay for pricey office space; and indirectly, since they must pay higher wages to let their employees rent or buy pricey residential space.
Building on a long, narrow island will always be expensive, but it could be less expensive if there were fewer rules deterring new construction. The city is currently considering upzoning the area around Grand Central Terminal to allow bigger, denser buildings, which could slow price growth for the financial-sector firms located there. Increasing the supply of top-tier space for financiers would indirectly aid other sectors as well, by limiting the competition for less prime real estate elsewhere in New York. To allow new industries to flourish, New York should also rethink its 1961 zoning code, which severely micromanages real estate. And the steep real-estate prices in New Yorks vast historic-preservation districts remind us that preserving the past imposes costs on the future (see Preservation Follies, Spring 2010). A little more freedom would allow more start-ups to locate in New York.
The city should worry, too, about the many other rules that limit start-ups. If you want to open something as simple as a restaurant or bar, you have to apply for a bewildering variety of permissions from various agencies and organizations: health and building permits; a license from the State Liquor Authority, which promises an answer within six months; community approval, after community members have pored over your proposed menus. The mayors office has tried to help by creating a New Business Acceleration Team, but the city should go further, locating all necessary approvals within a single office and guaranteeing applicants a decision in, say, 30 days.
While a simpler regulatory process is a no-brainer, another helpful step, cutting taxes, wont be a free lunch. City services, such as policing, require funding, and reducing those services typically makes a city less attractive. Further, new enterprises—as opposed to large, ongoing concerns—arent likely to be troubled by the citys high tax rates because entrepreneurs usually dont earn profits for many years. Still, that doesnt mean that taxes dont matter. High income taxes can make it harder for entrepreneurs to attract workers. Income taxes also weigh on entrepreneurs who are financing their innovative activity with a day job. New York will only become more attractive to lower-margin businesses if it figures out how to lower its tax burden without substantially reducing the quality of life within the city.
New York has recently made a bold bet to support industrial diversity: encouraging a new applied-science campus on tiny Roosevelt Island, between Manhattan and Queens. After holding a tough international competition, the city awarded the land for the campus, plus $100 million worth of infrastructure support, to Cornell University. The connection between education and metropolitan success is empirically robust; Berkeleys Enrico Moretti has shown that an area with a land-grant college before 1940 is likelier to enjoy economic success today.
The Roosevelt Island investment seems like a wise gamble. We dont know, of course, whether the campuss benefits will exceed its costs, whether it will mesh well with other local engineering-related efforts, or whether the island will prove connected enough to the rest of the city to encourage spin-offs and fuel an industry. But New York needs industries other than finance, and applied science seems like a plausible one. For hundreds of years, the city has derived strength from its wide proliferation of different industries and sectors, and New Yorkers should hope that the science campus kicks off the latest episode in that remarkable story.