In his book Fulfillment: Winning and Losing in One-Click America (2021), Alec MacGillis uses the example of Amazon to illustrate the growing divide between rich and poor areas of the country, in addition to the widening income gap between individuals and races. “The country had always had richer and poorer places,” MacGillis says, but the gaps have been growing wider. “In 1980 almost every area of the country had mean incomes within 20% of the national average. Only metro New York and Washington, D.C., were above that band, and only parts of the rural south and southwest were below it. But by 2013 the northeast corridor from Boston to Washington and the northern California coast had incomes more than 20% above average, and a huge swath of the country’s interior had incomes more than 20% below average.To the extent that the problem was noticed, it was often described as an urban-rural divide, and rural America was in crisis,” but more than that was going on. 

“The divide was also between cities – between a handful of winner-take-all metropolises and a much larger number of left-behind rivals. In the first six years after the Great Recession of 2008, job growth was almost twice as fast in large metro areas as in small ones, and income grew 50% faster. In the 1960s, the 25 cities with the highest median income included Cleveland, Milwaukee, and Des Moines; now the richest cities were on the coasts. By 2019, more than 70% of all venture capital was flowing to just three states: California, New York, and Massachusetts. As this regional inequality grew, so did its consequences. There was, above all, the political cost: a rising resentment in the left-behind places that made voters susceptible to racist and nativist appeals. Regional inequality was making parts of the country incomprehensible to one another. 

At the same time, whole sectors of the economy were being taken over by certain companies. This trend had been underway for decades as the federal government relaxed its opposition to corporate consolidation, and it caused regional imbalance in all sorts of ways. Airline mergers led to less service in smaller cities, which made it harder for them to attract businesses. Consolidation in agriculture meant that less of the money spent on food ended up with those who’d produced it. Mergers in sectors like banking and insurance meant that many small and midsize cities lost corporate headquarters and the economic and civic benefits that came with them. Put simply, business activity that used to be dispersed across hundreds of companies large and small, whether in media, retail, or finance, was increasingly dominated by a handful of giant firms. As a result, profits and growth opportunities once spread across the country were increasingly flowing to the places where the dominant companies were based.

Having played an outsized role in this zero-sum sorting, Amazon provides an ideal frame for understanding all of this, including the extreme inequality between its founder’s outlandish personal fortune and the modest wages of the majority of its employees; the nature of the work most of them are engaged in: rudimentary and isolating, out on the edge of town, often with unreliable hours and schedules; the immense influence the company’s amassed over both the state and federal government; and the unraveling of the civic fabric the company contributes to, through its undermining of face-to-face commercial activity and the tax base of countless communities.”

MacGillis describes how Seattle, Washington became a “hyper-prosperous city” by attracting tech entreprenurs and workers “in the decade since the Great Recession, adding 220,000 jobs, and engineering or research and development branches for more than twenty Fortune 500 companies, including Facebook, Google, and Apple. By 2018, per capita income in metropolitan Seattle had grown to nearly $75,000, but this rising wealth wasn’t spread evenly. By 2016, a city once known for its lack of extreme poverty and wealth, had matched San Francisco for high levels of income inequality. By 2018, the median cost of buying a home ($754,000) was higher in Seattle than anywhere in the country except the Bay Area. Rent, which had been on par with the national average before 2010, had also increased by 57% in only five years to an amount three times higher than in the rest of the country. This hyper-prosperity had many corporate fathers, among them Starbucks, Nordstrom, and Microsoft, but one loomed far above them. There were now 53,000 people working at Amazon in Seattle and its suburbs, earning an average of $150,000 a year. 

Amazon began with no more than ten warehouses scattered around the country. By 2017, it was up to more than a hundred. It needed more dispersed capacity because it was now handling 40% of all e-commerce sales in the country, double its next nine rivals combined. In 2016, the company trained its sights on the Dayton, Ohio area, picking a site for a new giant warehouse south of the city on I-75, one of the country’s biggest freight highways. As was typical for it, it sought incentives from both the state and the local community (Monroe). It insisted on total secrecy, to keep the negotiations from attracting public attention until the deal – a ten-year state tax credit worth $3.8 million and an exemption from local property taxes for fifteen years – was done. Monroe wasn’t alone in its magnanimity – in 2017 alone, Amazon collected well over $100 million in subsidies to open fulfillment centers around the country.”

MacGillis says that “nationwide, the poorest 10% were earning only 4% more an hour, adjusted for inflation, than they’d earned 40 years earlier. A third of all U.S. jobs paid less than $15 per hour, and the bottom half in income had less wealth, adjusting for inflation, than it had 30 years earlier.” 

Turning to Washington, D.C., MacGillis describes how, after 9-11, “a metropolitan area already made prosperous by the governmental lobbying industry was lifted to a whole new level of wealth by the national security industry.” Between 2008 and 2012, while the country was struggling to recover from the Great Recession, “the area’s high-net-worth households increased by 30%, and real estate prices in metro Washington increased more than prices in any other city. As wealth flowed into Washington, so did the media. Newspapers were in rapid decline in cities large and small across the country, their business model devastated by the triple whammy of Craigslist offering classified ads for free, Amazon eviscerating the department stores that bought many newspaper print ads, and Google and Facebook siphoning off the digital ad revenue that replaced print ads. From 2005 to 2015, 12,000 reporting jobs (one out of every four) vanished across the country. City councils, school boards, and major trials went barely covered, and candidates for state and local office un-vetted. But in Washington over that same period, the number of reporters doubled. It was easier to make digital journalism work at national rather than regional scale. When the economy collapsed in 2008, this meant that the national media was concentrated in a place where things were going much better than average, and so could be mistaken about the country’s real condition.” 

MacGillis notes that a revolving door between political and economic power had turned in Washington from the start, but it had usually just spun treasury secretaries back and forth from Wall Street. Goldman Sachs CEO Henry Paulson was George W. Bush’s treasury secretary, and another Goldman Sachs exec, Robert Rubin, was Bill Clinton’s. Rubin and his successor, Lawrence Summers, thwarted an attempt to regulate financial derivatives and pushed to repeal the Glass-Steagall Act, which separated commercial and investment banking, moves later implicated in the financial collapse. The movement continued under Barack Obama, who failed to hold accountable the bankers most responsible for the collapse.

As tech companies expanded their influence operations in Washington, they started using the revolving door, too. Facebook hired Joel Kaplan, George W. Bush’s former deputy chief of staff. Google hired Susan Molinari, a former Republican congresswoman. She was one of more than 400 former members of Congress employed as lobbyists, earning as much as tenfold their $174,000 congressional salary. In 1970, only 3% of members of Congress became lobbyists on leaving office; three decades later, more than 40% did. 

In 2013 Jeff Bezos bought the Washington Post, abetting his rapidly growing interests in the capital. Between 2012 and 2017, Amazon’s spending on lobbying quintupled; by 2018, it had the largest lobbying office of any tech firm in Washington, with 28 people, in addition to more than 100 lobbyists on contract at a dozen firms around the city. The company lobbied more federal agencies than any other tech company. It lobbied on the sales tax, which it still didn’t assess on most of the third-party sales that now made up more than half of its U.S. retail business; against regulations for drones, which it wanted to use to deliver packages; to maintain the discounted delivery rates it enjoyed with the Post Office; on government procurement, seeking to become the one-stop shop for federal purchasing; and against any effort to bring antitrust scrutiny to the company.”

Throughout the book, MacGillis gives us examples of the lack of workplace safety at Amazon’s blue-collar job sites, one of the most dramatic being the death of 52-year-old Jody Rhoads in 2014 in a forklift accident at a warehouse in Carlisle, Pennsylvania. Exhausted from working 10½ hour shifts, she’d crashed into some shelving at neck-level.” Amazon employees represented Rhoads’ death to OSHA, the Occupational Health and Safety Administration, as having occurred from natural causes, because “there wasn’t any blood or anything,” but an autopsy revealed “multiple traumatic injuries due to a pallet truck accident,” including “intestinal bleeding, a lacerated liver, and a bruised heart.”

The next city that MacGillis focuses on is Baltimore, Maryland, a major port and site of steel production, and “the country’s sixth-largest city – larger than Washington, Boston, San Francisco, or Houston in 1950.” He notes, however, that even then “decline was quietly underway, in the steady leakage of white residents for the suburbs. In 1940, even before the flood of workers that arrived later in the war years, Baltimore had the highest proportion of Black people of the country’s ten biggest cities: a fifth of the city, 168,000 people. Initially constrained to slices of the city’s east and west side, in homes packed to the point of squalor, they sought more humane conditions, and the more prosperous – those with union jobs – advanced into new neighborhoods.” MacGillis adds that the steel industry in Baltimore and all over the country eventually declined due to failure to modernize and compete with Japan. “Faring no better was shipbuilding – the Arab oil embargo had devastated an over-capacity tanker business, and by 1978, employment at the shipyard had dropped below 4,000, half its peak…The white flight that had started stealthily a few decades earlier was now a rush. There had been a riot in 1968, in the days after Martin Luther King, Jr.’s assassination, which claimed six lives and resulted in more than 5,000 arrests; there was crime and drugs; and there was now deindustrialization. The city’s population fell by 119,000 over the 1970s. By 2012, Baltimore’s population had fallen to 621,000, placing it 26^th in the country. Only its harbor, which handled coal headed to Europe and cars arriving from Germany; the biomedical empire of Johns Hopkins; and its location on the northeast corridor spared it the isolation of postindustrial cousins like St. Louis, Buffalo, Cleveland, and Detroit.” 

In September 2015, the top elected officials from metropolitan Baltimore gathered at the site of the former General Motors plant for the official opening of an Amazon “fulfillment”  warehouse as big eighteen football fields. “GM jobs at Broening Highway had paid an average of $27 per hour, plus generous benefits. Now, a decade later, the jobs at the same site paid $12 or $13 per hour and came with much thinner benefits. This hadn’t kept local and state leaders from showering Amazon with incentives to open the warehouse here – $43 million in all. 

There was growing agitation nationwide about the company’s wages and working conditions. Most notorious had been the 2011 episode at a warehouse near Allentown, Pennsylvania, where the company had stationed medics outside to handle workers fainting from the heat instead of paying for air-conditioning. Median pay across the company was only about $13 an hour, $28,000 a year. As recently as 2012, the company had had only 88,000 employees worldwide. But over the rest of the decade, it would grow with astonishing speed, making it the second-largest private-sector employer in the country after Walmart. By late 2019, it had more than 750,000 employees worldwide and 400,000 employees in the United States, the overwhelming majority of them in the company’s more than 200 fulfillment centers, sortation centers, and other delivery facilities. 

Warehousing and distribution used to be considered somewhat higher-skilled jobs: one could make over $20 per hour and stay years at a time. At Amazon, it was a more fleeting existence. Workers tended to be younger, turnover was high, and the seasonal workforce was often transient in the form of retirees traveling the country in vans and RVs. So much impermanence came with a major benefit for the company: it made it easier to stamp out union efforts to organize workers. Also deterring solidarity was the atomization of the warehouses themselves, where the layout and algorithms seemed almost designed to isolate employees from one another. Where organizing efforts managed to gain traction anyway, the company deployed tried-and-true defenses, hiring law firms that specialized in blocking unions and fomenting fears about union greed and corruption – the sort of tactics that had helped drive union representation to only 6% of the private-sector workforce nationwide.” When Amazon opened a second huge warehouse in the Baltimore area in 2017, “it got $19 million in state and local tax incentives. The company put out the word that it was hiring in August 2018, at a starting rate of $13.75 an hour. Those who passed the test got an email: ‘Congratulations on your offer of employment with Amazon.com Services, Inc.’ Before they could click to accept, they had to e-sign an agreement never to disclose anything about their work.”

Next, MacGillis describes what happens to businesses that sell products on the Amazon website, using examples of office suppliers in El Paso, Texas. “One day in 2017, Sandy Grodin, owner of an office supply business, got a call from one of his largest customers, El Paso Independent School District, informing him that it might be shifting its purchasing to Amazon. Around the same time, Amazon itself reached out to him, inviting him to start selling through Amazon Marketplace, with Amazon getting a 15% commission on every sale. While on the phone with Amazon, Grodin had his purchasing director, Heidi Silva, see what a type of Avery address labels were selling for on Amazon. They were $15.25 a box, below the $18 Grodin got them for from the wholesaler. He then had Heidi call Avery to ask for a pallet of them at the Amazon price. The guy at Avery, confused, came back a few minutes later, and told Heidi that the product on Amazon was counterfeit. By this time, the Amazon pitchmen were asking Grodin if he had any questions. Grodin asked if they could recommend another independent dealer who’d joined the Marketplace, so he could get a reference. They refused, saying the identity of Marketplace members was proprietary and confidential. Grodin then asked about counterfeiting on the site. When the Amazon reps asked what he meant, he said he’d come across an item selling well below the usual cost and checked with the manufacturer, who’d said it was counterfeit. Flustered, they asked Grodin what the item was. ‘That’s proprietary and confidential,’ he said, hanging up. Grodin asked Silva to track the prices on Amazon, every hour for a week, for the top 20 office-supply items ordered by El Paso Independent and compile them in a spreadsheet. The prices were fluctuating wildly, hour by hour. He then showed school district repesentatives the spreadsheet, told them about the 15% cut, and used his 2016 profit and loss statement to demonstrate the effect that selling on Amazon would have on his business. The district’s purchasing director shook his hand, thanked him for being so transparent, and said his recommendation to the superintendent would be that the district move cautiously with Amazon and only use it for items the district couldn’t buy locally.” 

Amazon was also trying to get Teresa Gandara, co-owner of another El Paso office supply company, to sell her products on its website. At a 2018 trade show, she attended a workshop entitled ‘Amazon, the Invisible Competitor: Does Your Business Have a Plan?’ presented by Mike Tucker, onetime pen salesman to the federal government. He explained that Anne Rung, a former head of the General Services Agency, the purchaser of goods for the federal government, had taken a high-level job with Amazon and met with a top GSA official to discuss a shift toward the e-commerce portal during the one-year ‘cooling off’ period during which former government officials aren’t allowed to lobby former colleagues. This had led him to conclude that it was time to sound the alarm. Tucker sketched Amazon’s famous ‘flywheel’ of success: use customer service and Prime’s free delivery to capture online traffic, use high traffic to compel other firms to join the Marketplace, then draw on those sellers’ expertise and the stream of data their transactions generate to see which of their products sell well, and start selling near-copies of those goods under Amazon’s name. For many merchants, Amazon could collect more than 30 cents of every dollar spent, once you added up commissions, fulfillment fees, advertising on the site, and account management deals, some of which were optional for selling on the site, but hard to avoid for sellers hoping to succeed. It was making a 20% margin on third-party sales, compared with only a 5% margin on its regular retail sales. In 2013 the number of third-party vendors selling via Amazon’s warehouses jumped by two-thirds in a single year. By 2017, the company’s cut of third-party sales was up to $32 billion, then $42.7 billion in 2018, a fifth of the company’s total revenue. 

Amazon had eliminated twice as many jobs at independent retailers as it had created, Tucker said – in 2014, it had sold $2 billion worth of goods in Illinois and $1 billion in Missouri without employing a single person in either state. (All told, only a quarter of all retail shopping now took place in independent stores, down from a half in the 1980s.) By driving out so many local businesses, it was devastating local and state property tax collection, and that was before factoring in the hundreds of millions it had received in tax subsidies for warehouses. There were also intangible costs: reduced street life, civic engagement, and social capital. The problem was, these costs weren’t perceptible to the average American. ‘The damage Amazon is doing is nearly impossible for consumers to detect in real time. It appears so consumer-friendly, it’s hard to see them as a monopoly. Its placelessness contributes to its invisibility, and makes it harder to fight.’ The Independent Office Products and Furniture Dealers Association (of which he was the director) was doing its best to fight back at the national level, Tucker said, but it was up to local dealers to resist in their own towns and cities, where it mattered most. They needed to urge local elected officials to reject the federal U.S. Communities contract Anne Rung had helped negotiate, by showing them how its model cost them money, and reminding them how much the local tax base depended on prospering small businesses. Tucker then turned the stage over Sandy Grodin, who told his story of successful resistance. 

Teresa Gandara left the session on fire. When she got home and learned that Amazon would be a special guest at the annual expo for vendors with whom El Paso did business, she asked for an appointment with Bruce Collins, the city’s purchasing director, who was managing the event. She showed him the information she’d gotten at Tucker’s workshop about Amazon’s effect on local businesses and tax revenue, and asked him to cancel Amazon Day at the expo. Learning that that would be up to his boss, retired army colonel Cary Westin, Gandara met with him, too, but not long into her presentation, he cut her off and told her the city of El Paso wasn’t interested in ‘protectionism for small businesses.’ On the evening before Amazon Day, Gandara appeared on the local TV news, saying that the city was sending ‘our tax dollars’ out of the local economy.’”

At this point in the book, MacGillis says that Amazon “allows its third-party sellers, a third of whom are based in China, to sell counterfeit goods, along with clothes made in dangerous Bangladeshi factories that other retailers had stopped buying from, and toys, infant sleeping mats, and other products that had been declared unsafe by regulators.” He then tackles the subject of Amazon’s data centers. “The cloud lives in data centers, vast windowless structures that started proliferating in certain corners of the American landscape at the close of the 20^th century as communications and commercial life moved online. In theory, data centers can be anywhere with proximity to fiber-optic cable, plentiful water, and cheap electricity. In reality, they cluster together in a handful of locations. Even more so than other aspects of the digital landscape, the cloud is dominated by a few places and a few companies – those with the most capacity and connections. The biggest cluster by far is in northern Virginia, where commercial internet providers are drawn by the area’s concentration of military contractors, high-tech companies, plentiful land, and cheap, coal-fired electricity. 

Amazon created Amazon Web Services, its cloud-computing branch, in 2003, and began offering its first data storage service in 2006. By 2017, AWS was providing cloud services to, among others, GE, Capital One, News Corp, Verizon, Airbnb, Slack, Coca-Cola, and even direct rivals like Apple and Netflix, bringing in more than $17 billion in revenue, a tenth of Amazon’s total. Between its dominance of the cloud and its dominance of online sales, Amazon has positioned itself as a gatekeeper extracting fees – what economists call ‘rent’ – on two of the biggest realms of digital commercial activity: data storage and e-commerce. You could almost compare it to a tax, except this tax was being collected by a corporation, not by a duly elected government. 

Each data center consumes as much energy as 5,000 homes. 

MacGillis relates how Amazon added a second data center in northern Virginia and a second data center hub in the metropolitan area of non-industrial Columbus, Ohio. It demanded and got the usual large incentives there, “including a 15-year exemption from property taxes, which, for a standard data center, would be worth $5.4 million, and insisted on special treatment at every step: accelerated building permit approvals, waivers of routine fees, and total secrecy. Three Columbus suburbs rushed to comply, also agreeing to credit back to the company 10% of the payroll taxes it would be paying on its employees, which would number only 25 at each center. This was on top of another tax break the company negotiated with the Ohio Tax Credit Authority, a sales tax exemption for equipment worth $77 million. The town of Dublin went yet further, giving the company 69 acres of farmland appraised at $6.75 million on which to build the center.”

MacGillis describes how Elena Schlossberg and other residents of Haymarket, Virginia fought successfully against Amazon’s effort to site its new data there at the community’s expense. “The thought of a new 100-foot-tall power line line cutting through Haymarket on a 120-foot-wide right-of-way at a cost of $65 million for the sake of serving Amazon alone irked Schlossberg, who organized a meeting at the local high school and filled it to capacity with 1,200 people. In January 2015, she fired off an email to Jeff Bezos, urging him to put the data center in the area that the county had set aside for the purpose, further east, or at the very least to run the power lines along the I-66 right-of-way, a route that would leave the lines partially buried and be less disruptive, though it would cost Dominion Power more. She concluded: ‘The community would be happy to welcome you, but what we won’t do is sacrifice our homes and our unique resources for your gain alone.’ She received no response to the email, nor to subsequent hard copy letters, so she and her fellow activists took the fight to the legislature in Richmond, pushing for a bill to encourage Dominion to take the mostly buried route. FInally, Dominion removed the Rural Crescent route from consideration. In June 2017, the hearing examiner overseeing the case for the utility commission ruled that the line would instead go through an historic African American community on Carver Road, the only place Blacks had been allowed to buy land in the county after emancipation. Those who’d opposed the earlier routes, led by Schlossberg, could have sat back now that their upscale homes and subdivisions had been spared, but they didn’t. A multiracial assortment of 70 people turned out to protest at Dominion’s regional headquarters and at the entrance to the data center itself, where security guards stood watch and signs warned of attack dogs. The pressure grew so great that in late 2017 county legislators announced that they were refusing to grant Dominion an easement needed to build the Carver Road route. Finally, Dominion announced that it would adopt the more expensive, less disruptive route that Schlossberg and her allies had been urging, along the interstate and partly buried below ground. Two months after the route reversal, however, the Virginia House of Delegates approved Dominion’s proposal to pay for the line with a monthly fee on all ratepayers. Meanwhile, Amazon filed a 78-page application to state regulators seeking a special discounted rate for the power it would be using at its data centers. There was no way of knowing exactly what the terms of the discount were: one version of its application sat under seal with state regulators, and the public version was heavily redacted.

In September 2017, Amazon announced that it would be opening a second headquarters somewhere in North America. The new headquarters would house 50,000 employees making an average salary of $150,000 and involve a total investment of $5 billion. To select the lucky city, the company threw open the selection process to all comers. The second-highest-valued company in the world after Apple, Amazon had other offices scattered around the country – San Francisco, Washington, New York, Boston, Los Angeles, and Austin all had at least 1,000 Amazonians, but its growth demanded a full second headquarters, not mere satellites…

Amazon would spend $4 billion on films and TV shows in 2018 and soon control a third of the market for streaming video…

A survey done in June and July of 2018 found that Amazon was the most trusted institution in the country among Democrats – ahead of government, universities, unions, and the press, and the third most respected among Republicans, ahead of only the military and police…

In January 2018, the company announced the 20 finalists for its second headquarters, and they were skewed toward BosWash hubs and creative-class boomtowns, the sorts of places already overflowing with high-tech talent and likely to be most appealing to recruits. On the list: Boston, Washington, New York, and Austin. Not on the list: Detroit, Baltimore, Cleveland, and St. Louis. What few Middle American cities made the cut were those that had already established themselves as winner-take-all cities within their regions, such as Columbus and Nashville.

Nick Hanauer, the early investor who’d helped bring Bezos to Seattle, had since grown critical of the company and was unsurprised that Amazon wasn’t going to use the HQ2 process to lift up a struggling city. ‘Right a social wrong? Are you fucking kidding me? Jeff Bezos is a straight-up libertarian. Those people believe that the only important thing is how well Amazon does, at the exclusion of everything else. They’re going to go where it’s best for Amazon. How that impacts the place? Other than how the place impacts Amazon is not in consideration for them.’

For years, Amazon had been strikingly unengaged in politics and civic affairs in Seattle, an absence that grew all the stranger the more it expanded. This was a reflection of the libertarian politics of its founder, who saw government not only a hindrance, but as irrelevant. Mike McGinn, the mayor from 2010 to 2013, hadn’t met Bezos once while leading the city. This lack of engagement stood in contrast with Microsoft, which had long been involved in local schools, even urging its engineers to teach computer science classes. Amazon’s disassociation extended to philanthropy: Bezos had given away conspicuously little wealth, locally or otherwise. In April 2018, he made headlines by saying that he viewed his space-exploration company, Blue Origin, as more worthwhile than anything else he could do with his money. 

Historically, Amazon had kept keep its tax bill down by claiming few profits. Now that it was finally racking up profits, it was funneling them through an office in Luxembourg, avoiding paying the federal government $1.5 billion, according to the IRS. In 2018, it paid zero corporate income taxes for the second straight year, despite doubling its profits to more than $11 billion. In fact, it so successfully gamed the tax code that it received a $129 million rebate. Overall, from 2009 to 2018, the company paid an effective tax rate of 3% on profits totaling $26.5 billion. Its success in avoiding taxes made it all the more remarkable that the company was making so much money from government contracts from federal agencies and the military to school districts.

Socialist Kshama Sawant had come to Seattle from India with her husband, a Microsoft engineer. Having won election in 2013 to represent the city council’s Third District, which included the Central District, she argued for a larger tax on big business than what the council’s liberal wing was proposing, and she believed in giving the target of the tax a clear face. “TAX BEZOS,” signs said. Amazon announced that it was so opposed to the $75 million tax legislation that it was pausing construction on its next big tower, a 17-story building that was to accommodate an expansion of 7,000 to 8,000 jobs, and reconsidering its lease of 700,000 square feet in another tower, space for up to 5,000 jobs. Mayor Durkan met with the bill’s council supporters, Amazon officials, and others to produce a Mother’s Day compromise: the tax would be shrunk by more than a third, to $275 per employee, collecting only $47 million a year. Amazon’s share would be about $12 million, at a time when its revenue exceeded $230 billion. With the compromise in hand, the council voted unanimously on May 14^th in favor of the new tax. Two days later, opponents announced that they were gathering signatures to put a referendum question on that fall’s ballot to repeal it. The No Tax on Jobs referendum push would be heavily funded: by Vulcan, a development company; Starbucks; and Amazon. On June 12^th, less than a month after voting unanimously to pass the tax, the council voted 7 to 2 to repeal it. A month later, the city’s median home price would hit $805,000, and by the end of the year, Seattle would post the largest increase in homelessness of any city in the country. Amazon, meanwhile, would reach $2 billion in quarterly profit for the first time a couple weeks after the vote; that year’s Prime Day, the annual summer shopping extravaganza, would surpass even Cyber Monday and Black Friday in sales. Its market capitalization would break the $1 trillion mark a few weeks after that, on September 4^th.

The first Amazon Fulfillment warehouses in Ohio opened late in the summer of 2016. Like the data centers, they were in the Columbus area, but in a different part of it – not in the more upscale exurbs on the city’s northern edge, but in the humbler outlying areas to the south and east, close to I-270, the beltway, and I-70. Calls started coming right away for nearby EMTs. West Licking Fire Station 3, three miles from the Etna warehouse was getting one emergency call a day at first, then several a day once the holiday season ramped up, with workers pressed to work up to 60 hours a week to get deliveries out in time. The calls ran the gamut: shortness of breath, chest pains, cuts, and fractures, but some were more serious. On April 7, 2017, there was a crash involving a forklift at the Obetz warehouse. There were no serious injuries, but a worker alerted OSHA a few weeks later that the forklift was being driven by someone who was legally blind. OSHA investigated, and indeed, Amazon had assigned forklift-driving to someone who had failed the vision test for an Ohio driver’s license – and it had another forklift driver in training who was also legally blind. It was nothing the local EMS and fire departments couldn’t handle, but it rankled them was that they were providing their services for free while Amazon would pay no local property taxes – the taxes that paid for the functions of local government, from the schools to the police and fire departments – for fifteen years. The warehouses were in Licking and Franklin Counties, but they were not really of them. They drew many hundreds of cars and trucks onto local roads every day, and they put out the emergency calls, but Amazon paid for neither snowplows nor ambulances. The basic social compact applied to others, but not to them. In 2017, voters in the area served by West Licking Fire Station 3 would be asked to approve a five-year, $6.5 million property tax levy to keep the fire department going, making up what the company withheld. The taxpayer would have to make up for what the company didn’t provide in other ways, too. In 2018, it emerged that one in ten Amazon employees in Ohio was making so little on the job that they were receiving food stamps. Nationwide, the company was one of the top employers of food-stamp recipients in at least five states. 

With more than 110 fulfillment centers around the country, Amazon now had warehouses within 25 miles of about half of the U.S. population, in virtually every state, and it was delivering half its orders. It leased 60 planes, and contracted with freight airlines for use of many more. Pilots complained of having to work 18-hour days (one plane contracted by Amazon crashed near Houston in February 2019, killing all three onboard), and it was planning a 100-plane hub at the Cincinnati airport. It had started handling some of the shipping from China itself, sending more than 4.7 million cartons of consumer goods in 5,300 containers in 2018 alone. It was setting up almost 200 delivery stations around the country, in addition to the fulfillment centers, to get packages closer to buyers’ homes. It had bought a fleet of 20,000 diesel vans and 7,500 trailers, and was about to order 100,000 electric vans for its drivers. Except they weren’t Amazon’s drivers, technically – they worked for contractors who, while they might deliver exclusively for Amazon, were separate entities. This meant drivers were denied whatever benefits Amazon provided employees. It also meant Amazon could avoid liability if the drivers got in accidents, as was happening with increasing frequency as time pressures mounted with the shift to one-day delivery.

On November 13, 2018, Amazon announced that Washington, DC – or to be precise, a northern Virginia suburb across the Potomac – had won the competition to be its second headquarters. At first, it looked as if the Washington area – specifically, the inner suburb of Arlington, Virginia – would share the prize of the second headquarters with another city. After all the drama about seeking one city for 50,000 jobs and $5 billion investment, the company had decided to split the loot, concluding that it would be easier to find the requisite space and workers in two cities than one. This lent the whole sweepstakes a bait-and-switch aspect – the prize, it turned out, wasn’t as grand as advertised. More upsetting to many of the contestants, though, was the identity of the two winners. After all the hype about a nationwide search, about scouring every corner of the land for a second home and encouraging cities far and wide to spend countless hours assembling bids, the company had chosen the two most obvious candidates: the two wealthiest and most influential cities on the east coast: New York and Washington. (Nashville would get the consolation prize of a new satellite campus with 5,000 jobs.) It was hard to avoid the conclusion that everyone else had fallen victim to a giant ruse. They’d generated free publicity for the company, handed over reams of information about their city (useful for the company’s future expansion calculations), and ramped up the bidding, all so that the company could extract a bigger price from the cities it was likely headed to all along: the seat of the federal government and the center of global finance. 

In picking the east coast’s two winner-take-all cities, though, the company miscalculated. New York had reached a point of such hyper-prosperity that it had given rise to a vocal minority recognizing that there could be too much of a good thing. There were streets so congested that bicyclist deaths were rising sharply, with much of the congestion due to the growing numbers of trucks and vans delivering boxes. The city was becoming a parody of the plutocratic metropolis, so when word emerged of yet more towers to accommodate the company led by the greatest plutocrat of them all – subsidized by nearly $3 billion in taxpayer incentives, a newly energized minority, led by tribunes like newly elected U.S. Representative Alexandria Ocasio-Cortez, rose up and said ‘no.’ SCAMAZON read the graffiti sprayed on an abandoned restaurant near the chosen site in Long Island City. The company was stunned by the ingratitude and effrontery, and pleaded its case at one city council hearing after another. It didn’t help that the city was getting warnings from Seattle. Teresa Mosqueda, the council member who thought she had a compromise agreement with Amazon, came to New York and told them to get everything in writing. “That type of bad behavior when it comes to public policy making shouldn’t be replicated in another city,” she said later. A few weeks later, it was over. One day, the company was negotiating with the critics over the treatment of the 5,000 workers at its mammoth Staten Island warehouse, and there was talk of allowing them to organize freely in exchange for approval of the subsidies for the new headquarters. This was a bridge too far for Amazon. A day later, the company pulled out. Not entirely – there would still be several thousand white-collar Amazonians in the city, mostly at Hudson Yards, the giant new development on the west side of Manhattan. But there would be no HQ2 extravaganza.

So there was one lone winner after all. Washington and northern Virginia had mustered nothing close to the resistance of New York. With 25,000 employees, Amazon would be the area’s largest private-sector employer. The area offered the technical talent the company was looking for – there would be easy recruiting amid all the homeland security IT contractors and rival cloud providers. Other cities, such as Boston and Austin, could have offered that, too, but what they couldn’t offer was proximity to power. The company would soon pass the $4 million mark in quarterly spending on lobbying, nearly five times what it had spent five years earlier and second only to Facebook among tech giants. All told, Amazon had spent $80 million over the course of the decade on seeking to influence the federal government.

On July 16, 2019, the U.S. House Judiciary Committee’s subcommittee on antitrust, commercial, and administrative law convened a hearing to address “online platforms and market power.” It was a remarkable moment: after years of watching as a small handful of companies grew to dominate life and commerce online, members of Congress in both parties were all of a sudden showing concern about the extent of that dominance. Their concern had been stoked by a small but effective cadre of thinkers and activists articulating the nature and costs of the companies’ dominance. These included Stacy Mitchell with the Institute for Local Self-Reliance, who documented the companies’ effect on smaller businesses, local communities, and democracy in general: “These companies,” she wrote, “have created a form of private government – autocratic regimes that are tightening their control over our main arteries of commerce and information.” They also included Lina Khan, who while still in law school had written a groundbreaking paper showing how antitrust enforcement had failed to address the threat of the new monopolies, Amazon in particular. Regulators had erred, she argued, in focusing only on the question of whether companies were raising prices for consumers, not on whether they were causing broader distortions in the marketplace and society in general. Amazon, she argued, was stifling competition through predatory pricing and the structural advantages that came with being a dominant online platform in ways that over time would leave consumers with lower-quality products and less choice and innovation, even if they seemed to be benefiting from lower prices today. Her arguments were being wielded by a new coalition, dubbed Athena, that included dozens of unions and activist groups, such as the Warehouse Worker Resource Center, which had long fought on behalf of the thousands of Amazon workers in the San Bernardino Valley. Even some of Amazon’s founding employees had started to have grave misgivings. ‘I think the characterization of Amazon as being a ruthless competitor is true,’ said Shel Kaphan, a programmer who’d worked with Bezos at the time of the company’s founding. ‘Under the flag of customer service they can do a lot of things that might not be good for people who aren’t their customers.’ 

The Federal Trade Commission, the Department of Justice, several state governments, and the European Commission had all launched investigations into the company’s practices. Now executives from four giants – Amazon, Google, Facebook, and Apple – had been called to testify on the question of whether they’d gotten too big for the country’s good. The subcommittee’s chairman, David Cicilline, a Rhode Island Democrat, noted that Amazon controlled nearly half of all online commerce in the United States; that half of American families now had an Amazon Prime account; and that Amazon’s closest competitor, eBay, controlled less than 6% of the market for online commerce. ‘This hearing isn’t just about the companies before us today,’ he said. ‘It’s about ensuring that we have the conditions for the next Google, the next Amazon, the next Facebook, and the next Apple to grow and prosper.’ Committee members asked Amazon’s legal representative whether third-party sellers had any viable alternatives for selling their goods online other than Amazon. They asked about the rising fees that Amazon was charging them to sell on the site – an average of 27 cents on each dollar of sales, a 42% increase over five years. They asked about the company’s punishment of sellers that violate its terms by suspending their accounts and shutting down their product pages, freezing them out of the marketplace. And they zeroed in on the inherent conflict between Amazon both controlling the platform on which nearly half of all online sales were made and selling its own products on that platform. They asked whether the company privileged its own products by promoting them higher on the page or pricing them below cost to drive competitors out of business, and whether it used sales data to come up with its own versions of hot-selling items. In October 2020, the Democratic staff of the House subcommittee released a 449-page report on its investigation into the tech giants’ dominance, calling on Congress to take action to break up the companies. ‘To put it simply, companies that once were scrappy, underdog startups that challenged the status quo have become the kinds of monopolies we last saw in the era of oil barons and railroad tycoons,’ the report stated. ‘These firms have too much power, and that power must be reined in and subject to appropriate oversight and enforcement. Our economy and democracy are at stake.’

The corona virus pandemic took a series of related developments in American life and accelerated them. The news organizations that had already lost the majority of their advertising revenue to Silicon Valley were now losing what little remained as a result of the halt in commerce, and the legacy retail companies that had survived the upheaval of the prior two decades were careening toward extinction. JCPenney, Neiman Marcus, and J.Crew filed for bankruptcy. Macy’s temporarily shuttered all 775 of its stores and furloughed nearly all of its 125,000 workers; its stock fell by 75% in two months. Amazon’s Seattle neighbor, Nordstrom, announced it was laying off thousands. The toll was no less widespread among the country’s small independent businesses. All told, 25,000 retail stores were expected to go out of business by the end of 2020, a figure that nearly tripled the mass closure figures of recent years. Meanwhile, the companies that had for several decades been capitalizing on the trends reshaping the economy were growing larger and more successful. By late May 2020, the five biggest tech companies – Apple, Facebook, Microsoft, Amazon, and Google’s Alphabet – had added a stunning $1.7 trillion to their combined market cap in just two months, a rise of 43%. The combined value of these five companies was now a fifth of the S&P 500, and, collectively sitting on $557 billion in cash, they were only going to grow larger. They used it to finance new acquisitions and to raise their spending on research and development to nearly $30 billion – more than NASA’s entire budget – even as their smaller rivals retrenched.

The biggest winner was Amazon. Its first-quarter sales were up by more than a quarter over those the year prior, at a time when overall retail sales were plunging. Its stock surged so much in mid-April 2020 – up by more than 30% on the year before, as the pandemic was nearing its deadliest period – that Jeff Bezos’s net worth increased by $24 billion in the span of two months. In late July, Amazon announced that its profit had doubled in the second quarter, with sales up by a stunning 40% from those a year earlier. On the news, its share price surged even higher; by early September, it was up by 84% on the year, more than double the rise of the other tech giants. To handle the surge in business, the company had, between January and October, added more than 425,000 employees worldwide, bringing its total number of non-seasonal employees in the U.S. to 800,000 and its global total to more than 1.2 million, up by half from a year earlier and now behind only Walmart and China National Petroleum, and that tally didn’t even include the 500,000 drivers who were delivering its packages. To house these workers, the company went on a building and leasing spree, opening 100 buildings in September, on its way to occupying nearly 100 million additional square feet of warehouse space by the end of 2020, a roughly 50% increase. Its warehouses weren’t the only part of the company in high demand: its data centers were ramping up capacity for customers like Zoom, as hundreds of millions of daily human interactions shifted online. The midsummer announcement of Amazon’s massive pandemic profits had come on the same day that the Commerce Department reported that the U.S. economy had shrunk by nearly 10%, the largest quarterly drop on record. In other words, Amazon was flourishing more than ever at one of the lowest moments for the country as a whole. 

Even as the pandemic accelerated the concentration of wealth and power within some of the most dominant companies, it was possible to imagine ways that it might disperse prosperity and dynamism more broadly across the country. Some of the many New Yorkers who’d fled the city (there were upscale apartment buildings in Manhattan left almost entirely vacant) were considering permanent relocation. There was talk that the pandemic might spell the end of the office, that we would, at long last, be free to work from anywhere. While rents and condo prices started to retreat from their stratospheric highs in the wealthiest cities, the rising demand for suburban New York office parks and bedroom communities, Western ski-resort real estate, and Hamptons school districts suggested, however, that the benefits of any dispersal would remain well within the bounds of the winner-take-all metropolis and its satellites, and wouldn’t trickle out to smaller cities farther afield. Meanwhile, those same smaller cities were poised to suffer deep cuts as the federal pandemic rescue package targeted municipal aid only to cities with populations over 500,000. By August, American Airlines was announcing it was ending flights to fifteen small and midsize cities, making it more likely that their isolation and decline would accelerate. The pandemic’s toll was going to fall hardest on the places that had the least cushion to fall back on. 

The same was true of the impact on individuals. Nationwide, people who earned more than $28 per hour had seen their employment levels return to pre-pandemic levels by the fall of 2020, while those who made less than $16 per hour had seen more than a quarter of their jobs wiped out. At Amazon’s warehouses in France, union demands over safety measures had forced a weeks-long shutdown and an eventual deal that included a reduction of shifts by 15 minutes, without a reduction to pay, to allow for more social distancing at crowded shift changes. In the absence of unions at the U.S. warehouses, discontent took other forms. WELCOME TO HELL read the graffiti inside truck trailers, out of sight of warehouse cameras. FUCK BEZOS. Workers began sharing their disquiet in online back channels, and at some warehouses, they organized protests, signaling that the pandemic might set in motion a new era of workplace activism.”

MacGillis concludes by saying that “federal action is needed to provide a check on so vast and powerful a company, along with fellow industry giants.”