JOIN THE DISCUSSION
Subscribe to Our Emails
Receive weekly newsletters, reading lists, event announcements, promotions, and BR news.
Startups aren't the magic bullet for economic growth.
Photograph by Gabriel Jorby
In June 2014, a sleek tour bus parked beside a Family Dollar store in Pittsburgh’s gentrifying East Liberty neighborhood. It looked like a rock band was arriving at its concert venue. The bus had come to visit AlphaLab Gear, a startup hub located just blocks away from Google’s colorful Pittsburgh offices in a converted Nabisco factory.
The bus and its passengers were part of an initiative called Rise of the Rest led by former AOL CEO Steve Case. It’s a Silicon Valley investor’s tour of startups in cities not typically considered tech hubs.
Consistent with the startup ethos, AlphaLab Gear’s floorplan is open, its ceiling ducts exposed. Case and a panel of judges sat at an elevated table as companies pitched their ideas from below. New firms with only a handful of employees made wild claims about their future growth. The best startup pitch—a shoe sole that generates enough energy to charge a cell phone—received a prize before Case and his team hopped back on the bus and drove on. First to Cincinnati, then Nashville, then to the seventeen other cities that the tour has visited to date.
Case’s bus tour is one example of a broader effort to identify and develop the “Next Silicon Valley.” Slate identifies thirty-one cities around the world that have received the moniker “Next Silicon Valley” from news outlets including The Wall Street Journal, The Economist, NPR, and the BBC. New York City has claimed the title Silicon Alley; Kansas City considers itself Silicon Prairie; and Dublin, Ireland calls its tech scene Silicon Docks. In many of these places, city governments have subscribed to a kind of Silicon Valley mimicry. They have launched incubators, supported investor incentives, and sponsored innovation challenges like hackathons with big prizes—all with the hope that high-growth startups will follow.
The Silicon Valley Consensus is that innovative cities grow faster than non-innovative ones—but that’s not always the case.
Michael Piore, an MIT economist, calls governments’ enthusiasm for supporting high-tech entrepreneurship the “Silicon Valley Consensus.” The Washington Consensus of the 1980s claimed that free trade and deregulation were the most promising growth policies for developing countries. The Silicon Valley Consensus suggests that innovative cities grow faster, that startups are the only real hope for job creation, and that high-tech growth helps rich and poor alike.
But, like the Washington Consensus, the Silicon Valley Consensus offers a false promise.
Why Pittsburgh and not Cleveland?
While some American cities that have seen rapid growth since 1980 have been tech hubs, such as Boston and Seattle, other high-growth cities, such as Nashville and Madison, have excelled in diverse industries. Less than half of the twenty-five metros with the fastest income and employment growth since 1980 have a concentrated high-tech sector.
The case of Pittsburgh is telling. Its economy is among the fastest-growing of U.S. metro areas, but its growth did not originate with high-tech startups. It began its rise long before Case or Google showed up. The steel industry along Pittsburgh’s three rivers collapsed dramatically in the early 1980s. Between 1979 and 1983, the metro area lost over 100,000 jobs, nearly 9 percent of its total workforce. Imagine the riverfront—formerly crowded with busy steel mills—beset with an empty, eerie calm.
Soon after the city’s economic collapse, Pittsburgh’s mayor and the presidents of its two leading universities—with support from the business community—crafted a recovery plan they called Strategy 21. It focused on expanding public infrastructure like highways and the airport while also developing new research infrastructure at Pittsburgh’s universities: the Pittsburgh Supercomputing Center, the Software Engineering Institute at Carnegie Mellon, and the Robotics Institute, also at CMU. Strategy 21 wasn’t a reflexive response to crisis. It built on decades of foundation grants and university initiatives that had sought to help Pittsburgh diversify its economy in anticipation of steel’s decline.
The recovery “wish list” totaled $550 million. Since the coalition of Pittsburgh’s leaders couldn’t fund these projects on their own, they reached out for federal and state support. Congressional lobbying helped the coalition win funding worth over $100 million from the Department of Defense for the Software Engineering Institute. Backing from Westinghouse helped Pittsburgh’s universities win National Science Foundation funding for the Supercomputing Center. And Pennsylvania’s state government proposed grants and loans that would help make Strategy 21 a reality. Pittsburgh’s strategy was to listen to what businesses say they care about most in survey after survey: skilled workers and reliable infrastructure. Although many of Pittsburgh’s neighbors have tried to attract businesses with tax breaks and cheap capital, businesses consistently rank these factors as less important.
By the mid-1990s, employment and income in the Pittsburgh metro area started to take off faster than that of its neighbors. Over a decade later, Pittsburgh weathered the recession far better than most of the country. Today the riverfront hums with commuting cyclists. Sunlight reflects off glassy buildings affiliated with local universities and corporations. Since 1990, the metro area’s income per capita has grown 153 percent. That’s 22 percent faster than the average for large American cities during the same period. And Pittsburgh’s employment rate grew 5 times as fast as the national average—10 percent compared to 2 percent.
Pittsburgh’s investment in transportation and R&D infrastructure was not focused on incubating startups as much as it was designed to attract top technical talent to the city and keep it there. The same technical talent would prompt companies like Mellon Bank—after it merged with the Bank of New York—to keep thousands of employees in Pittsburgh instead of moving them to New York City.
Today, although investors like Case focus on Pittsburgh’s startups, young tech companies comprise only a tiny segment of the Pittsburgh economy. The growth of financial institutions, hospitals, universities, and advanced manufacturing has been far more significant. This is not to say that startups have not been growing. However, startup growth is a consequence—not a cause—of Pittsburgh’s economic success.
Skeptics might claim that Pittsburgh’s growth was lucky. It is possible that infrastructure had little to do with it: perhaps other cities made similar investments in infrastructure and experienced no growth at all. Some might point to Cleveland—only a two-hour drive northwest of Pittsburgh—as an example of big plans with few results.
As Cleveland’s industrial economy began declining in the early 1980s—albeit less sharply than Pittsburgh’s—a group of Cleveland business leaders assembled a committee called Cleveland Tomorrow. Although the local business community had helped elect the mayor, the initiative was a decidedly private enterprise. Neither government nor university officials played a leading role in the effort. The core membership came from Cleveland’s manufacturing industries.
Cleveland Tomorrow hired local consultants from McKinsey and Co. to diagnose the city’s economic problems and propose fixes. In a way the report paralleled Pittsburgh’s Strategy 21: there were recommendations to re-invest in manufacturing, and there were ideas for diversifying the local economy. However, the proposals to revitalize manufacturing were much grander than the ideas to diversify.
Cleveland Tomorrow ended up investing $105 million in the recommendations. They funded a Productivity Center aimed at improving labor productivity as well as the Advanced Manufacturing Program and the Cleveland Research Institute to generate new technologies and provide resources for companies focused on advanced manufacturing. The business leaders also invested in a venture capital fund to support Cleveland entrepreneurs and additional business resources to shepherd their startups to growth.
Today the Advanced Manufacturing Program survives as an organization called MAGNET that operates as a consultant for manufacturing businesses. The Cleveland Research Institute and other attempted partnerships with local universities from the 1980s have mostly dissolved. The venture capital fund was profitable, but not because Cleveland-based firms took off. When the fund could not find enough startups in Cleveland to fund, it started backing companies outside the region that turned out to be successful.
The Cleveland Tomorrow experience shows that not just any infusion of cash generates growth or transformation. Investments in transit infrastructure and R&D infrastructure—supported by a lasting coalition of government, industry, and universities—are more likely to pay dividends than tax relief for plants or easy capital for startups.
Decades later, in 2004, the state of Ohio passed the Third Frontier initiative, a series of private grants and financial incentives for high-tech enterprises to form in or move to Ohio. Around the same time, a gaggle of new economic development initiatives were formed in Ohio with similar ambitions: encourage new startups and give them the funding and mentorship to succeed. The program is primarily focused on supporting private industry, not public infrastructure.
Although it is too early to guess whether the program has been transformative for Ohio regions like Cleveland, there’s a larger question at stake. Does the basic premise behind programs like the Third Frontier make sense? Are cities with more innovation more likely to grow and thrive? Are startup hubs better equipped for job creation than cities with older firms? Are Silicon Cities more likely to help the middle class? In other words, even if the Third Frontier and programs like it succeed at turning their regions into innovative startup hubs, where would that leave these cities and their residents?
Innovative Cities Don’t Grow Much Faster
City governments have a basic goal: more local jobs with higher pay. Cities are searching for ways to support income and employment growth. Innovation has been sold as a pathway to both. The OECD, for example, says innovation “provides the foundation for new businesses, new jobs, and productivity growth.” The Silicon Valley Consensus suggests that innovative cities tend to grow faster than non-innovative ones, but that’s not always the case.
Since 2000, large U.S. cities with more innovation—quantified by the number of patents for inventions—have not, on average, experienced more growth in income per capita or employment. If we include the dotcom bubble of the 1990s, more innovative cities showed slightly more income growth but no more employment growth than less innovative cities. For example, if a city with low innovation like Chattanooga suddenly achieved San Francisco’s level of innovation, it would see only an additional 0.5 percent income growth per capita per year. By contrast, Chattanooga’s income per capita already grew at over 5 percent per year during this period.
For another example, the San Jose metro area—the heart of Silicon Valley—stands out as the most innovative in the U.S. It has more than double the patents for inventions of the next most innovative metro. Yet San Jose is tenth in a composite of income and employment growth since 1980. Cities like Nashville and Portland have experienced more growth during this period, but lag far behind on prominent measures of innovation.
More than anything else, these data suggest caution for the Silicon Valley evangelists. While some of the most innovative metro areas like Silicon Valley and Austin have achieved high growth, the trend is not universal. Many of the fastest-growing cities in the past decades were able to grow without incubators or venture investment.
Startups are not the only source of growth and should not be the main focus of economic policy.
Startups Don’t Necessarily Lead to Job Growth
Silicon Valley boosters might argue that even if patents aren’t associated with job growth, it’s still startups and small businesses that create jobs. The U.S. Small Business Administration says that small businesses represent “66 percent of all net new jobs since the 1970s.” The Kauffman Foundation, which advocates for and performs research on entrepreneurship, reports that “new businesses account for nearly all net new job creation.” Claims like these have generated a chorus of praise for small business and startups among politicians. “Small businesses are the backbone of our economy,” President Obama says. “Startups are engines of job creation,” says a White House statement.
Economists confirm part of the political rhetoric: a disproportionate chunk of net new jobs (jobs created minus jobs lost) come from young firms (firms zero to five years old are considered “startups”). However, the size of a business isn’t closely correlated with its likelihood of creating jobs. John Haltiwanger and colleagues show that small firms are just as likely to create jobs as large firms of the same age. Yet young firms are more likely to create jobs than old firms of the same size. On the surface, this affirms that startups are an important source of job growth at the national level.
However, there are several reasons to question whether our economic policy should fixate on startups.
First, startup jobs represent only a small chunk of overall employment. For most metropolitan areas, young firms provide less than 20 percent of private jobs. And many of these firms are not startups in the way that Silicon Valley might imagine them. These young firms include retail or service businesses that operate strictly in the local economy. They are not necessarily exporting a product or service that would bring new economic activity to a region.
Second, although startups account for the bulk of net new jobs, it’s less clear whether cities with more startup are more likely to experience overall employment or income growth. A thriving city economy needs healthy firms of all ages. Not only can mature firms benefit from active young companies that compete with them. Young companies often draw on seasoned talent from mature firms to feed their initial growth. Many of the most successful entrepreneurs in Silicon Valley and elsewhere start their careers at mature companies.
Third, even if we believe that startup jobs are great for the economy, we still can’t be sure that government support for startups is leading to more startups jobs. Of course governments make bold claims about how many jobs their tax or incentive programs have created. For example, Ohio boasts that $154 million in startup investments (half of which was public money) from its Third Frontier program have created 2,584 jobs since 2006.
By claiming credit for these jobs, Ohio is making the dubious assumption that these startups would not have hired people if the Third Frontier Program didn’t exist. The startups that succeed with support from Third Frontier could have succeeded with support from another investor. If Ohio really wants to measure the effect of Third Frontier, they would compare companies that just barely qualified for Third Frontier funding and the companies that barely missed qualifying. Ten years after they applied to Third Frontier, what’s the difference between Third Frontier startups and the others?
Startups are important to a local economy. New businesses generate excitement, support experimentation, and challenge their more mature competitors to improve. However, startups are not the only nor the most promising source of growth, and they should not be the overwhelming focus of economic policy.
Tech Jobs Are Not Always Good for Working People
Over the past several years, grassroots activists in San Francisco have protested what they see as the tech community’s contributions to exploding rents and widening inequality. The symbol of tech-inspired injustice became the Google Bus: a massive unmarked shuttle lumbering through city neighborhoods, stopping periodically to drop tech workers at stops near their homes. The perception is often that Silicon Valley’s dominance benefits only those who work in the industry and disrupts the lives of those who do not.
Silicon Valley enthusiasts disagree. They argue that growth in the technology industry provides spillover benefits to the economy as a whole. For every tech job created in a city, the Bay Area Council argues, four other jobs are created in the local economy. Tech workers increase demand for local services like restaurants, dry cleaners, and salons—hiring tech workers has a multiplier effect on local economies, the argument goes.
Both are partially correct. While it’s true that low-income households in innovative cities have not experienced more wage growth than low-income households in other cities, it’s not clear that higher-income households have been that much better off in innovative cities than elsewhere. Growing inequality is a problem everywhere—not just in places with a thriving tech industry. And while tech hiring does generate some positive spillovers, it’s important to remember that innovative cities are no more likely to experience employment growth than non-innovative cities.
High-tech cities are an example of both success and failure when it comes to generating income growth at all levels. In San Jose, income growth for the top 10 percent is far higher than median income growth, and median income growth is far higher than income growth for the bottom 10 percent. Income growth in Austin, by contrast, has been among the most even of all U.S. cities.
• • •
The flash and celebrity of Silicon Valley’s rise has overshadowed success stories in the rest of the United States that possess equally powerful lessons for economic growth. When Steve Case arrived in Pittsburgh and Nashville and Charleston, the symbols of these cities’ successes were incubator spaces and chic startup offices.
However, these cities began their rise decades ago without a high-tech “scene.” Even cities that we consider tech hubs, such as Austin and Durham, began their rapid growth not with startups but with big investments in knowledge infrastructure from the 1970s onward. For Durham it was the establishment of Research Triangle Park as a non-profit supported by local industry, universities and the governor. In Austin it was a joint effort between the University of Texas, Ross Perot, and the state government to attract MCC, a research association focused on semiconductors.
The stories behind these cities’ successes are more complex than one event or company or investment. They are the function of long-term coalitions and advances across many industries—some of which have grown more than expected, and others of which have shrunk less.
The Silicon Valley consensus has formed (and billions of public dollars have been invested) despite scant evidence that cities can become the next Silicon Valley—or that becoming the next Silicon Valley would even be desirable. The consensus has suggested that struggling cities should look west if they wish to rise. It would be more promising for cities to look in all directions—and perhaps in the mirror—for lessons on how to grow.
...we need your help. While reading "Unraveling the Silicon Valley Consensus" by Ben Armstrong, you might have noticed the absence of paywalls at Boston Review. We are committed to staying free for all our readers. We've also gone one step further and become completely ad-free. This means you will always be able to read us without roadblocks or barriers to entry. It also means that we rely on you, our readers, for support. If you like what you read here, pledge your support to keep it free for everyone by making a tax-deductible donation.
Vital reading on politics, literature, and more in your inbox
Readers Also Liked
Printing Note: For best printing results try turning on any options your web browser's print dialog makes available for printing backgrounds and background graphics.