Before the Amazon HQ2 hysteria started, Bob Duffy twice asked Jeff Bezos to make an investment in Rochester. The upstate New York locale once employed about 60,000 tech workers, when the film empire of Eastman Kodak was one of the most valuable enterprises in the world. Duffy worked at Kodak as a messenger and dark room production tech for a year after high school. His wife, Barbara, worked in the company’s HR department for 20 years. Their house is mortgaged through ESL Federal Credit Union, the bank that founder George Eastman established in 1920 to help Kodak employees finance their homes. “Kodak had such an impact in almost every aspect of civic life,” says Duffy, who served as Rochester’s mayor from 2006 to 2010. “If you needed something done, Kodak was there to help you.”
In September 2017, Duffy read about an exciting opportunity. Amazon planned to open a second headquarters that would house 50,000 employees. In the words of CEO Bezos, this would be a “full equal” to the corporation’s Seattle home base. The company promised billions of dollars in direct investment, and further investment from commercial and real estate developers was sure to follow. Metropolitan areas across the continent with at least 1 million residents were invited to apply.
Why not Rochester? thought Duffy. Amazon could be the city’s Kodak for the modern era. In its 208-page bid, submitted in tandem with nearby Buffalo, the city proposed housing Amazon on the old Kodak campus, the symbol of a faded tech giant. But Rochester was knocked out of the running for HQ2 in January. “We always held out hope that we would make the top 20. I was disappointed when we didn’t,” Duffy says. “Many midsized cities do struggle and they do work very hard for growth. … [Amazon] could come in and actually transform the environment with their investment.”
With the benefit of hindsight, it’s now clear that few of the 238 communities that applied for HQ2—including many of the 20 finalists—ever really stood a chance. On November 13, the online retailer announced that HQ2 will not be an HQ2 at all; instead, the company will open two smaller sites in Long Island City, a Queens neighborhood in New York, and Crystal City, a Virginia suburb of Washington, D.C. Those cities already house Amazon’s two biggest offices away from the West Coast. They’re nexuses of financial and governmental power. And they’re just a few miles from two of Bezos’s lavish homes. Amazon broke the rules of its own game, then picked the most obvious candidates.
Increasingly the U.S. economy is centered on so-called “superstar cities” like New York, D.C., and San Francisco. These places, like the giant corporations that call them home, benefit from scale, deep talent pools, and network effects that make their power more entrenched as time goes on. Other locales struggle to keep pace, just like the brick-and-mortar retailers trying to compete with Amazon’s might. The HQ2 frenzy resonated so broadly because many places knew a chance like this would never come around for them again.
Now that it’s over, the cities that lost are left to make sense of a process that was always stacked against them. Meanwhile, Amazon is set to reinforce an economic system that is increasing inequality, monopoly power, and political polarization in one fell swoop. “We really do have two classes of places,” says Mark Muro, a senior fellow of the Metropolitan Policy Program at the Brookings Institution. “We have the superstar metros that are highly technology based … and then we have hundreds of places that are sort of left behind.”
Fifty years ago, when Kodak was flourishing, big ideas were often born in small places. Comcast was founded in Tupelo, Mississippi. Walmart emerged in Rogers, Arkansas. Intel started making computer chips in a patch of Northern California that would soon be christened Silicon Valley. Back then, entrepreneurs sought low-cost areas in which to launch businesses and industry titans required lots of space in which to situate their factory plants and warehouses. Wages and employment in the nation’s biggest cities grew at the same rate as in smaller cities and rural areas. Economists believed that this trend would continue indefinitely, and economic prosperity would be evenly distributed across the United States in a phenomenon known as convergence.
But something changed in the 1980s. The advent of the personal computer upended the nature of work, transforming some jobs and eliminating others. Kodak once needed factory workers to manufacture cameras and film, chemists to research film development techniques, engineers to design camera models, logistics specialists to distribute their physical goods to stores, and marketers to sell people on the idea of a “Kodak moment.” It employed tens of thousands of people in Rochester across these fields. But the PC, the digital camera, and eventually the smartphone rendered many of these jobs obsolete. Today, Apple has about 800 people dedicated to designing the iPhone camera, clustered in Silicon Valley alongside hundreds of thousands of other engineers. The manufacturing duties for gadgets like the iPhone have been shuttled to parts suppliers in China.
The American economy has switched from a cycle of convergence to one of divergence, with more and more economic growth consolidating in the most prosperous cities. According to a recent report by the Brookings Institution, metro areas of more than 1 million people have accounted for 72 percent of the nation’s employment growth since the 2008 financial crisis, even though they comprise only 56 percent of the country’s population. Among the most prosperous 2 percent of metros, the growth rate is even faster. This is one of the digital age’s many grand ironies, considering technology makes it easier for people to collaborate whether they’re in the same room or thousands of miles apart. “At the dawning of the internet, there were books written with titles like ‘The Death of Distance,’” Muro says. “It turned out that, yes, more transactions could happen remotely, but the value and importance of clustering increased. … Digital skills were rewarded. And the ability to manage projects, solve technical issues, and so on.”
Today, the five tech giants that lord over the U.S. economy—Amazon, Apple, Facebook, Microsoft, and Alphabet, Google’s parent company—all are based in either the Bay Area or Seattle. The next crop of mega-corps, such as Uber, Airbnb, and Netflix, are headquartered there as well. With fewer places earning the spoils of the digital economy, cities have taken to competing aggressively for whatever scraps these companies might offer: a warehouse here, a data center there. Government officials increasingly resort to offering tax breaks to lure firms that promise to bring jobs. The number of megadeals per year valued at $50 million or more in incentives has doubled since the 2008 recession, according to Good Jobs First, a Washington, D.C.-based organization that tracks government subsidies.
It was in this environment that Amazon launched its HQ2 sweepstakes. The company laid out broad criteria for its second headquarters: a population of at least 1 million, proximity to an international airport, a functional mass transit system. But it remained tight-lipped as cities that clearly did not meet these requirements offered to change their names to “Amazon,” hand-delivered cacti, and performed other shameless stunts to attract attention. These antics were doubly beneficial to Amazon, since they generated free promotion and contributed to a treasure trove of data about cities where the company might later want to plant warehouses, Whole Foods stores, or surveillance technology. “A less charitable view is that this was a mass data vacuuming effort to scarf up a lot of data about dozens of other places that they will likely want to locate smaller units,” Muro says.
Some cities appealed to Amazon’s pragmatism, citing their low cost of living as a lure away from the coasts. Others clung to the romantic notion that Amazon would view its coming headquarters in civic terms, using it to revitalize a struggling city or bridge the gap between the caricatures of “coastal elites” and “true heartland Americans.” In a letter pitching Detroit, local business magnate Dan Gilbert invoked the image of the hard-working Midwesterner and offered the company a chance at “reinventing” a major metropolis.
Kansas City was among the heartland locations that thought its offer of a bucolic life away from the housing and overcrowding challenges facing coastal cities would appeal to Bezos, who lived in Albuquerque, Houston, and Miami as a child. But the city discovered its distance from the coasts was a problem, not a perk. “We were told when we didn’t make the final 20 that there was some concern about the size of Kansas City and the location in the middle of the country,” says Tim Cowden, CEO of the Kansas City Area Development Council. “We think that’s a real positive.”
In fact, tech companies like Amazon prize regions dense with highly skilled workers, whether they’re employees or not. Unlike the era in which Kodak excelled, both employers and employees now expect jobs to be temporary stepping stones, not lifetime commitments. Therefore workers are attracted to places with ample opportunities in their field. Research shows that innovation (as measured by patents filed) happens overwhelmingly in large metro economies. Going it alone in a city with a nonexistent or developing tech ecosystem is a risk, especially for a company premised on always being a Day 1 startup. “These companies thrive on brain cells,” says Greg LeRoy, executive director of Good Jobs First. “They’ve got to be in labor markets where there’s lots of people moving around, among tech companies that are plugged into what’s going on. You don’t get that in Cleveland or Calgary or Albuquerque.”
Amazon was plainly aware of these dynamics when it launched the HQ2 selection process. But even as recently as a week before the announcement, Bezos continued to cast the search as a Bachelor-esque quest for a soulmate, saying the final decision had to be made “with your heart.” It seems much more likely that Amazon had a more cynical strategy all along.
When the Amazon folks came to visit Indianapolis in March, Maureen Donohue Krauss didn’t want them to see a single piece of trash on the 12-mile stretch of highway between the airport and downtown. The city collected 560 55-gallon bags of garbage along the route and ensured that every pothole was filled before hosting Amazon officials for a quick 24-hour tour of the region. “We spiffed up for company,” says Krauss, the chief economic development officer for the city’s chamber of commerce. “This was a direct quote from them: ‘We know what Amazon’s culture is. We want to understand your culture to see how it would work together.’”
The Indy cleanup was just a small example of the lengths some of the 20 finalists went to in order to land HQ2. Pittsburgh offered Amazon nearly $10 billion in financial incentives. Maryland pitched $8.5 billion and Newark pledged $7 billion. Dallas said it would build an Amazon University, a joint project between local colleges to train engineers in a facility next to city hall, and Atlanta promised Amazon a private car on its MARTA rail system and an executive lounge at Hartsfield-Jackson airport. While New York and Virginia offered Amazon $2.8 billion and $573 million in tax breaks, respectively, the fact they didn’t offer the highest amounts—and that New York is paying so much more than Virginia—is proof that Amazon was eager to play one city against another. “The whole thing was a show with the ultimate purpose of getting the best possible benefits from one or more cities on the short list they already had in mind,” says Chris Tilly, a professor of urban planning at UCLA. “I don’t think there was a genuine process of scoring the map of the United States.”
Though Amazon chose the most obvious places, economic developers in the smaller competing cities insist the HQ2 process was constructive for them. In Indianapolis, where Salesforce has its largest hub outside San Francisco, Krauss believes the competition helped the city shed the “flyover country” label. “We really solidified what our regional story is,” she says. “Being on that list for us was really important because it gave us recognition of things that we knew and others did not know.”
Rich Fitzgerald, county executive of Pittsburgh’s Allegheny County, echoed those sentiments. Though his region is widely associated with its past as an industrial town, it’s also home to Carnegie Mellon University and corporate AI research labs. “It did bring to the attention of a lot individuals and a lot of companies what a great place Pittsburgh is,” he says. “It gave us a look from companies that might not have had us on their list.”
City officials in these places say they’re not trying to compete with New York, and that seeing HQ2 land in another Midwestern city would have been more disappointing than seeing it go to the biggest hubs on the East Coast. But Amazon’s choice may accelerate an economic imbalance that ultimately makes it tough for smaller cities to recruit top companies. New York’s tech sector grew by 57 percent between 2010 and 2016, according to a report by the state comptroller’s office. Google is buying up enough real estate there to house 20,000 employees. (That’s only 5,000 fewer than Amazon’s planned headcount, with none of the self-aggrandizing fanfare.) And D.C. is likely to become a greater priority for many major tech firms as the industry comes under increased political scrutiny.
Other cities could try to incubate homegrown alternatives to the current megacorporations, much as Intel once relocated tech’s center of gravity from New York to California. But startup creation is mired in a yearslong slump, in part because of the market power of giants like Amazon. And of the new businesses being created, half are launching in New York and Washington, D.C.
The migration of tech from the West Coast to the East Coast will put even greater strain on public infrastructure and housing costs in cities already bursting at the seams. But tech companies are no strangers to causing cities to creak under the weight of their success. It’s up to everyone else to accommodate them or compete.
On Monday, General Motors delivered the second-biggest jobs news of the month: The company is planning to lay off 15 percent of its salaried workers and is shutting down five North American factories in Ohio, Michigan, Maryland, and Ontario. In total, the automaker will cut 14,000 jobs.
In the same release announcing the layoffs, GM said it plans to double its investment in electric and autonomous vehicles over the next two years. Much like the transition from the old Kodak technology to digital cameras, these new cars require workers with different skills and more education than their analog predecessors. GM’s self-driving division, based in San Francisco, will soon open a satellite office in Seattle, mostly employing engineers. GM is now competing for workers with Google, Uber, and, yes, Amazon.
After the 2016 election, I wrote about how self-driving trucks had the potential to disrupt our labor markets and government as workers rendered obsolete by artificial intelligence drifted toward more extreme political candidates. I wondered whether we might feel the ramifications of tech’s influence on politics in a decade. In fact, it’s already happening. According to the Brookings report, Hillary Clinton won just 15 percent of U.S. counties in the 2016 presidential election, but those counties accounted for about two-thirds of the national GDP. “At this point, more places feel on the outs [of the economy] than in the center,” Muro says. “That is very unhealthy and a profound negative externality that is disrupting the stability of the country.”
This imbalance does not only create cynicism about politics among conservatives. In this year’s midterms, 59 percent of voters in Senate elections backed Democratic candidates, yet the party lost two seats in the chamber. This is a particularly pronounced version of a trend that’s been escalating for decades, as more Democrats crowd into deeply blue states like California and New York. The last time there was some level of parity between the number of states that trended Democrat or Republican in congressional races was 1980, just before the PC sparked the Great Divergence and the rise of superstar cities.
One solution Muro offers is for the federal government to host its own civically minded version of the HQ2 search, asking cities far from the coasts to compete for funding that would designate them as “rising tech hubs” and spur economic growth. An approach managed nationally could lead to a wider distribution of prosperity than a sweepstakes that encourages cities desperate for growth to go at each other’s throats.
We’ve created a system in which people in rural areas feel left behind by the economy, and people in urban areas feel frustrated that our governance does not accurately reflect the will of the people. By planting a flag in the middle of America, Amazon could have helped reverse these intertwined trends. But that wasn’t a factor in the HQ2 search, however much some people wanted it to be.
Amazon captured the country’s imagination by allowing people to believe that their communities could defy the economic odds. But this is a company guided by data, not “heart,” no matter how much Bezos tries to spin the decision. Splitting up HQ2 makes sense. Putting it in New York and D.C. makes sense. Dumping factory workers in favor of software engineers makes sense. But these are also signs that the ruthlessly efficient AIs movies have taught us to fear have already arrived in corporate form. The people who don’t live in the superstar cities—or can no longer afford to—have to pay the price. “It’s another variation on the theme of the 1 percent getting richer at the expense of the 99 percent,” LeRoy says. “It’s sort of the geographical equivalent of our inequality today.”