The Billionaire Who Took Over the Internet

Tax exemptions, labour practices, and explosive growth: how Jeff Bezos made Amazon one of the world’s biggest companies

The head of Jeff Bezos, a bald white man, smiles toward the screen while surrounded by shiny piles of gold.
The Walrus/Wikimedia Commons/flickr

Amazon’s founder and CEO, Jeff Bezos, became the world’s richest man in the summer of 2017, edging out former Microsoft CEO Bill Gates, who’d held the title for a decade. Bezos retained the crown even after his 2019 divorce, the world’s most expensive, and also retained control of company voting rights. His $108 billion fortune (all figures US) arose from his enormous company, which in recent years has traded the title of largest company in the world with tech colossi Microsoft and Apple.

Amazon says its huge growth and success owes to being “obsessed with the consumer.” In fact, the record shows that the company’s rise has been fuelled by the growth-boosting momentum of economic network effects, horrifying levels of labour exploitation at both the white- and blue-collar levels, and a boost from a fortunate sales-tax exemption.

In the very early startup days of the company, Bezos explained to a colleague why he was so hell-bent on fast growth: “When you are small, someone else that is bigger can always come along and take away what you have. . . . We have to level the playing field in terms of purchasing power with the established booksellers.” But, once Amazon itself reached great stature, it did not encourage a level playing field; rather, it used its own increased leverage and bargaining power to drive whole market sectors to surrender their profitability to Amazon’s growth. As a former content writer for the company’s front page wrote, “As the public saw it, we had morphed from David to Goliath.”

Bezos founded Amazon after several successful years at a Wall Street hedge fund. From the very beginning, the plan was to “exploit” the internet, as Bezos’s biographers describe the business model. The strategic value of running a retail business online was that it had almost no “transaction costs”—the term economists use to describe the costs of making a deal. In-person retail requires significant transaction costs, like driving to a store, paying for gas or transit, and conforming to retail hours. Shopping via the then brand-new World Wide Web offered major convenience, as it does to this day. These lower transaction costs have been a large part of online retail’s expansion at the expense of brick-and-mortar stores.

Bezos chose books to begin building his empire for purely utilitarian reasons—they shipped relatively well, didn’t spoil, and were available from wholesalers. Sales grew quickly in the mid-1990s, with the company expanding into new merchandise categories like DVDs and music, then toys and electronics. The growth was stupendous, with top-line revenue shooting up by over 50 percent during some quarters. This famously fast growth owed to many reasons, but the really crucial factor was the network effects. In markets run through networks, like the telephone network, services get more valuable as more people use them. The real net effects arose as Amazon gradually became not just a sprawling online retailer but a platform—a network based on allowing users to bring their own goods to sell at the company’s central, virtual location. Higher sales of all types meant higher volumes with lower costs, allowing Amazon to lower its prices, a short-term boon to the consumer but also a move bringing more business to the company, allowing it to further cut costs and, of course, to attract more third-party sellers to its platform.

By 2002, third-party sales had become one-third of cash flow. But the company kept its foot on the gas to build its platform power. That push included the creation of Prime, Amazon’s loyalty program, which offers fast shipping (and now streaming media) for an annual fee. This again helped drive up volumes and push down costs, and it allowed more aggressive negotiations with suppliers. Ten years later, Amazon’s third-party sellers sold nearly 40 percent of the items shipped, with over two million retailers on the rechristened Amazon Marketplace. By last year, 58 percent of all sales came from Marketplace, with almost 1 percent of all Americans selling goods on Amazon. Taking a flat commission, in the neighbourhood of 10 percent, off each sale, the platform was profitable and, above all, had become a magnet for thousands of small and midsize companies wanting to sell on Amazon’s incredibly visible website without having to create their own web presences.

More impressively, to protect and expand the platform, Amazon actually began lowering the prices charged by the independent sellers and paying the difference itself. This “discount provided by Amazon” measure makes no damn sense at all unless considered alongside the importance the company puts on its platform power. As Amazon and other tech platforms expand globally, Amazon has resorted to actively “recruiting” small merchants to sell on its site, to build out its platform power in other major economies like India. The business press reports that Amazon has gone to the extent of lending small firms the money for inventory in order to get it available on amazon.in, thereby gravitationally attracting more business. The company even faced a legal challenge in 2018 from eBay, which accused Amazon of poaching its higher-value sellers by infiltrating its internal member messaging system.

As Amazon grew, it continued to lose money overall—an enduring feature of most of the company’s history. How has Bezos kept his investors from getting spooked? Business reporting has regularly concluded that Amazon’s gigantic, platform-fuelled growth rate has kept them quiet. This growth into ever more industries means insane future profits and enormous market power, and investors have by and large been willing to accept this. But this all means Amazon’s bold acceptance of near-term losses or tiny profits, along with Bezos’s famous claim that “your margin is my opportunity,” can be attributed to one thing—network effects.

As of this writing, Amazon is now the second-biggest company in the world by market value (behind Microsoft and ahead of Apple, Google, and Facebook), with $177 billion in yearly revenue and a few billion in actual yearly profit, thanks mainly to its remote-computing cloud-services unit. Amazon is among the leaders of the “optimization” of business processes, using metrics to evaluate every decision and then building preferences into the algorithms that run a large part of the day-to-day activity on the platform. Today, Amazon relies on these systems to regularly “scrape” pricing data from competitors’ sites and then automatically match or slightly underprice them in order to leverage its platform power into a greater market share.

Amazon has proven itself to be a true bully of the markets, with a long history of using its huge market weight to undo client after competitor after customer. It’s a fascinating story in which the company again and again justifies its ever-growing power using the laws of the jungle whose name it took.

Amazon’s explosive growth soon proved Bezos correct that, “when you are small, someone else that is bigger can always come along and take away what you have.” But now Amazon was bigger, and it would do the taking. The company’s enormous power is now relatively well recognized in the mainstream, including what economists call “monopsony” power—the ability to exert power as a huge buyer, which is distinct from the selling power of a monopoly. This became relevant fairly early in Amazon’s negotiations with UPS, the major US commercial shipping company. Amazon had contracted with UPS for much of its shipping, but as its stature rose, Amazon approached FedEx and began quietly integrating their data systems while increasing shipments.

When UPS declined to negotiate rates, Amazon virtually cut it off as a shipping vendor, such that UPS went from receiving millions of Amazon parcels daily to only a handful. Amazon’s cultivation of FedEx, combined with its ability to rely on the US Postal Service to deliver a large number of boxes—allowing it to continue smooth service for buyers—caused UPS to cave after seventy-two hours and offer Bezos lower rates. Journalist Brad Stone, in his 2013 book The Everything Store: Jeff Bezos and the Age of Amazon, observes that the experience taught “the company an enduring lesson about the power of scale and the reality of Darwinian survival in the world of big business.”

As Amazon expanded into more and more product categories, certain holdouts came to its attention. One was diapers.com, a baby-product startup run by entrepreneurs who idolized Bezos. Their delivery service was a godsend to exhausted parents, and the small outfit attracted millions in venture-capital investments. Spotting potential prey, Bezos’s team told the diapers.com founders that Amazon was readying a large launch into the category and therefore it would be in everyone’s best interest to negotiate a buyout immediately. When this “offer” was declined, Amazon cut prices for baby products by up to 30 percent. Amazon later said its pricing had nothing to do with this upstart company, but the small businesspeople noticed that, when they tweaked their posted online prices, the Amazon website’s algorithms changed its own prices accordingly. The gigantic company’s web-crawling software was automatically tracking the prices of its small competitor.

While the startup flirted with selling itself to Walmart, Amazon ratcheted up the existential pressure by announcing a new loyalty program offering a whopping additional 30 percent off diapers in return for joining a subscription program, Amazon Mom. Staff at diapers.com estimated the program was losing $100 million a quarter, and once diapers.com caved and sold, Amazon closed its money-losing program to new members, although it later had to hastily reopen it after the Federal Trade Commission started sniffing around. These moves show that Amazon wasn’t trying in good faith to compete. Rather than being “obsessed with the customer” and seeing who could best satisfy the market demand for the lowest sustainable price, Amazon purposefully ran the category at a giant loss in order to drive its smaller, cash-poorer rival out of the market, then monopolized it.

But books were Amazon’s original market category, and in the end, they were the site of its cruellest predations. Beyond the low transaction costs of the business model, there was the fact that years of previous consolidation in the bookselling industry had left large swaths of the United States with few or no bookstores, creating a near-captive market for Amazon. By 2004, the company was selling a large portion of the books purchased in the US. So it began to display its now-famous aggression with publishers, and here I quote at length a recounting by Stone:

Amazon approached large publishers aggressively. It demanded accommodations like steeper discounts on bulk purchases, longer periods to pay its bills, and shipping arrangements that leveraged Amazon’s discounts with UPS. To publishers that didn’t comply, Amazon threatened to pull their books out of its automated personalization and recommendation systems, meaning that they would no longer be suggested to consumers. . . . Amazon had an easy way to demonstrate its market power. When a publisher did not capitulate and the company shut off the recommendation algorithms for its books, the publisher’s sales usually fell by as much as 40 percent. “Typically it was about thirty days before they’d come back and say, Ouch, how do we make this work?” says Christopher Smith, a senior book buyer at the time.

          Bezos kept pushing for more. He asked [then head of Amazon publisher relations Lyn] Blake to exact better terms from the smaller publishers, who would go out of business if it weren’t for the steady sales of their back catalogs on Amazon. Within the books group, the resulting program was dubbed the Gazelle Project because Bezos suggested to Blake in a meeting Amazon should approach these small publishers the way a cheetah would pursue a sickly gazelle.

          As part of the Gazelle Project, Blake’s group categorized publishers in terms of their dependency on Amazon and then opened negotiations with the most vulnerable companies.

Stone adds that, “soon after the Gazelle Project began, Amazon’s lawyers heard about the name and insisted it be changed to the less incendiary Small Publisher Negotiation Program.” Stone notes that these lower wholesale book prices allowed Amazon to lower its retail prices, which attracted more consumers and put more pressure on physical bookstores. The record is full of episodes where executives openly acted as predators to remove competition. “By his own admission,” Stone writes, “[one executive charged with squeezing publishing houses] took an almost sadistic delight in pressuring book publishers to give Amazon more favorable terms.” By the mid-2000s, Amazon had finally reached a scale where it could match Walmart’s price cuts, and for some years it even won contracts to run the online commerce sites of Target and Borders books after they had fumbled the development of their own web presences.

In 2004, as the company turned the screws on publishers, it began to prepare the Kindle. Amazon had an early flirtation with online music, but Bezos realized that Steve Jobs had beaten him to the punch with the 2001 iTunes release, and he “ultimately concluded that if Amazon was to continue to thrive as a bookseller in a new digital age, it must own the e-book business in the same way that Apple controlled the music business,” writes Stone. This led the company to develop hardware for reading electronic versions of books, a risky strategy given Amazon’s lack of manufacturing experience. Its eventual e-reader tablet, the Kindle, was far less versatile than a smartphone but had longer battery life and a better display for reading.

To make the new book format and device a success, Bezos apparently pulled the e-book price of $9.99 out of thin air, which, as a result of his power, had serious consequences for the publishing world. Amazon knew publishers would be horrified at this price point, which would be money-losing for them, but it needed them to commit scarce resources to digitizing their large libraries in order to establish the platform power of the new device. So the executive staff decided not to give advance notice of this money-losing price point lest the publishers balk. The publishing houses’ first knowledge of it came when Bezos announced the product publicly in 2007.

The Kindle 2, released in 2009, was the breakout version, and in the years until Apple’s release of the iPad, in 2010, and Barnes & Noble’s release of the Nook e-reader, in 2011, Amazon held a staggering 90 percent of the e-reader market. To the publishers, writes Stone, “Amazon’s dawning monopoly in e-books was terrifying.” The enforced low prices “tilted the field” further against physical bookstores and gave “Amazon even more market power. The publishers had seen over many years what Amazon did with this kind of additional leverage. It extracted more concessions and passed the savings on to customers in the form of lower prices and shipping discounts, which helped it amass even greater market shares—and more negotiating leverage.”

Similar to other great monopoly capitalists, Bezos has a long record of stopping at nothing to clear the market of competition. The lengths he was willing to go to crush even small competitors are rough to read about. He targeted the online shoe retailer Zappos by spending $30 million building a competing website, endless.com—which offered different styles and sizes of shoes with free overnight shipping—demonstrating Amazon’s usual strategy of losing money to pull market share away from vulnerable competitors. This bite apparently wasn’t enough to break Zappos, so Bezos, unbelievably, added a five-dollar bonus to its overnight shipping rates, meaning customers were paid five bucks to buy shoes. Faced with this plus several pressures arising from the Great Recession after 2007, including lower consumer spending and freezing credit lines, Zappos’s investors sold out to Bezos, after which the shipping bonus was discontinued.

Amazon’s broader retail platform has grown largely through the company’s willingness to antagonize higher-end companies by allowing independent sellers to continue aggressively discounting. Companies like Sony and Black & Decker have gone back and forth with Amazon on allowing regular sales of their products on Marketplace at rock-bottom prices, but they usually come crawling back because of Amazon’s unique market strength.

Amazon’s corporate arm wrestling extends, of course, to the public sector. After grandly announcing, in 2017, its need for a new headquarters, or “HQ2,” to complement its enormous footprint in costly Seattle, the company received an ocean of applications from 238 North American cities and regions. By design, the process was a media frenzy, with endless free publicity for the company while America fell over itself trying to offer free things to the world’s richest man. Many of the bidders were among the poorest in North America, offering money and subsidies from budgets already bleeding and cutting back public services. Detroit offered the company thirty years tax-free. New Jersey and the city of Newark put together a staggering combined package of $7 billion in tax cuts and subsidies, and Governor Chris Christie openly dared other states to top it. The Wall Street Journal noted that the company “has advised HQ2 candidates to keep this phase private.” (Amazon ultimately split its investments between New York and the Washington, DC, area.)

The company now earns half of every dollar spent online, and it continues its relentless expansion: into food with its acquisition of pricey premium grocer Whole Foods, which it’s integrating into its Prime discount and delivery service; into online advertising to compete with Google and Facebook; and into shipping by land, sea, and air with its creation of a sea-freight forwarding service, plans for an air cargo hub, and the beginning of a “last-mile” delivery business. And Bezos himself bought journalism institution the Washington Post for half a billion dollars. But it’s Amazon’s growth in utterly new industries that is now getting the most attention.

Cloud computing refers to the development of technology allowing information-processing needs, like storage, database management, or processing power, to be provided online instead of through a company’s own hardware. The data is stored and processed in computing centres built by tech companies—huge windowless warehouse-like buildings filled with tall rows of computer servers managing data or doing calculations for clients around the world. A quickly growing portion of today’s computing requirements are done through cloud computing, and Amazon is the market originator and leader (with Microsoft a strong second).

The project grew out of the company’s ambitions to be seen as a tech company, like Google, rather than a mere retailer, like Walmart. The company sat on a giant pool of data on customers’ buying and rating behaviour, the currency of the tech age, and it made some available to publishers early on, later creating APIs (application programming interfaces) that let software developers collect price and product data. Other APIs let clients use Amazon’s shopping cart and payment system. These projects were later grouped together as Amazon Web Services (AWS). These à la carte, online-enabled services at first attracted small businesses that had a hard time paying for all their own IT hardware and staff, but soon large firms took notice. The service grew quickly, in part, again, due to low prices that Amazon’s investors tolerated in exchange for breakneck growth.

The efficiency of cloud computing is that it reduces the need for expensive in-house hardware and software, which traditionally allow for high data-processing speeds and the handling of heavy data traffic but are often only needed in the busiest conditions and are otherwise unnecessary. Amazon’s own system was built to manage the traffic flood the week after Thanksgiving, and the rest of the year, much of it sat idle. “Most of the system’s capacity went unused most of the time,” observes Nicholas Carr in The Big Switch: Rewiring the World, from Edison to Google, and Bezos said in 2006 that “there are times when we’re using less than 10 percent of capacity.” Now computing power can be bought as a utility, like hydro power, and business observers hope this development will end up benefiting smaller firms and startups, which would otherwise struggle to afford the big IT assets of large corporations.

Amazon utterly dominated the cloud-computing market for some time and is now the largest player, at about 51 percent market share, followed by Microsoft’s Azure cloud service, at 13 percent, and Google, at 6 percent. And, unlike the rest of Amazon’s famous cash-burning business model, AWS is profitable—highly profitable. Indeed, only when AWS blew up did the company begin to take a few billion in profit a year; cloud computing is responsible for half of Amazon’s profits to this day.

And, crucially, big corporations are not the only ones seeing the value in Amazon’s services—other big institutions, like the Pentagon, are noticing as well. Since 2013, Amazon has hosted data for the Department of Defense, the CIA, and other security agencies. Amazon’s dominance of the security cloud is thus unlikely to wane soon. The Wall Street Journal reported that, in 2016 alone, the three biggest cloud providers together invested over $30 billion in their gigantic data centres, filled with cold-stored banks of computer servers, to run their corporate services. The business paper found “the massive investment is creating a barrier for would-be rivals that would need to spend tens of billions of dollars to match the computing capacity Amazon, Microsoft and Google already have,” and quoted an investment banker: “They’ve created a powerful moat.” High barriers to entry are a classic feature of long-lived monopolies and oligopolies.

For all this potential, the cloud has very real downsides. As with any utility, service can be disrupted, like when a snowstorm pulls down power lines and you have to sit in the cold until service is restored. In 2012, a large electrical storm, or “derecho,” on the Atlantic seaboard caused heavy rains, flooding, and downed trees, leading to the failure of local power grids. Carr relates that, as computing access for millions of people hung in the balance, “Amazon’s technicians acted quickly . . . But though the outage was brief, the recovery proceeded haltingly. It took the company many hours to reboot its servers, reconstruct its databases and its virtual machines, and get all its systems working again.” Amazon and the other cloud oligopolists are investing heavily in their centres and systems to avoid these kinds of disruption in the future, but this episode illustrates that, when a few firms run an industry, the stakes are raised if anything goes wrong.

Those stakes increasingly extend not just into big corporate databases but into the home, too, since it’s the power of the cloud that’s allowing a new wave of small hardware products that use specialized cloud connections to punch above their weight, including “smart speakers.” These small computers and stereos deployed around the house allow “voice computing,” in which simple verbal commands or queries activate the speaker to run a function like playing music or searching online.

As with the cloud itself, Amazon has led the smart speaker market with its various Echo models, followed by the Google Home, the Microsoft Invoke, the Facebook Portal, and the Apple HomePod. Market analysts find Amazon leads the market in part by offering characteristically rock-bottom prices because it sees the speakers as “an entry point for e-commerce,” according to the Wall Street Journal. Thus, having its speaker, rather than another company’s, in a home means regular chances to push its own retail products, to entice more households into Prime, to collect more data, and to charge monthly streaming fees.

These always-listening devices have already led to a number of innovatively creepy episodes, like the one involving a Seattle woman whose Echo mistook parts of a conversation as commands, recording the conversation and then sending it to a person on the user’s contact list without the user’s consent. And, when a coalition of child- and privacy-advocacy groups tested the Echo Dot Kids Edition service, they found that the information it recorded about children’s requests and searches couldn’t be deleted by employing parental controls but required contacting customer support.

Gradually, the tech majors are eroding the walls of consumer privacy, and the speaker market is a bleeding-edge example of that. Users now allow the speakers’ algorithms to help decide if one outfit looks better than another on them, and with Amazon’s purchase of a company making smartphone-connected, camera-equipped doorbells—and its new service delivering products inside your home when you’re not there, or even inside your car—the company integrates itself ever deeper into customers’ lives.

But all these cutting-edge, creepy cloud-based toys rely on human labour to be produced and to operate. And Amazon’s labour history is less than prime.

In his 2010 book Amazonia: Five Years at the Epicenter of the Dot.Com Juggernaut, James Marcus tells the story of being hired by Amazon in 1996 and signing paperwork featuring the header “Job-Related Stress”:

The Employee . . . recognizes that his position may involve a high degree of job-related stress. As a condition of employment, the Employee agrees that he will not bring an action against the Company for . . . damages alleged to have resulted from such job-related stress, and will indemnify the Company.

The clause was well founded, especially in the company’s early days, when coders and warehouse workers alike were consumed in “Death Marches”—as other tech giants playfully call them—where major software reconfigurations or warehousing clearance pushes were completed over many days, with sleep brief and on-site. In general, workers often didn’t take even weekends off, giving up a precious legacy of the labour movement. Stone relates that Bezos “was imagining a different culture for Amazon, one where employees worked tirelessly for the sake of building a lasting company and increasing the value of their own ownership stakes.” But, during hiring interviews, “if the potential employees made the mistake of talking about wanting a harmonious balance between work and home life, Bezos rejected them.”

Of course, the ramification of this work policy is that women, whom society still widely expects to take the primary caregiver role for a family’s children, will be disproportionately discouraged from working or advancing in such an environment. Silicon Valley’s well-documented problem of drastic underrepresentation of women is due in significant part to its embrace of a fanatical worth ethic that leads to sneering at any work-life balance, meaning Bezos’s rejection amounts to form of institutionalized sexism.

The white-collar staff was subjected to nightmarish conditions similar to those at other tech giants, including long double shifts in the warehouse when the company was shorthanded during the holiday season—orgies of work whimsically called “Operation Save Santa.” For years, Bezos refused to subsidize bus passes, an effort to prevent workers from leaving earlier to catch a bus instead of working late and driving home. A 2015 New York Times investigation documented how “emails arrive past midnight, followed by text messages asking why they were not answered.” Unsurprisingly, management is given a prepared set of “leadership principles” that are a heinous caricature of libertarian authoritarianism. A standout: “Leaders are right a lot. They have strong judgement and good instincts.”

And, as the company grew, “Bezos drove employees even harder, calling meetings over the weekends, starting an executive book club that gathered on Saturday mornings. . . . As a result, the company was not friendly toward families, and some executives left when they wanted to have children,” writes Stone. When a female staff member asked when policies supporting work-life balance might be put in place, Bezos responded that “we are here . . . to get stuff done, that is the top priority. . . . If you can’t excel and put everything into it, this might not be the place for you.” Amazon is very much part of the gender imbalance in the broader market economy, where women’s reproductive decisions can eject them from the workforce.

Yet, despite the staff’s extensive sacrifices, like delaying the start of a family or never seeing their existing one, the picture is different for the CEO. Although Bezos terrorized the workforce and consumed its waking hours, before their separation, his wife MacKenzie became pregnant, and the couple moved into a giant mansion in Medina, near Bill Gates’s estate. According to Stone, MacKenzie would drop off Bezos daily in a minivan after taking the kids to school, then retire to her separate apartment to write. When their first son was born, Bezos took time off. As he should. But, at Amazon, simply having a basic family-work balance requires major privilege.

The company is more notorious for its history with its blue-collar workforce. The warehouse “pickers”—who hasten from shelf to shelf with carts, gathering items, then back to the central belts that send the combinations off for sorting and packaging—are the sweating workforce behind the delightful convenience of shopping online. From its earliest days, when knives sold in the new kitchen-goods category came sliding down the conveyor system with no packaging, Amazon has worked its blue-collar employees to the bone. Endless hours for limited pay is the norm, as are escalating performance targets that have workers continuously running from wing to wing of the warehouses.

Pickers carry handheld scanners to confirm the merch and direct themselves to the next shelf, which also allows the company to track them on a productivity-per-second basis. Managers frequently chew out the workforce for not meeting goals and threaten them with firing, sometimes on a daily basis. The demands rise and rise, suggesting a purposeful strategy of running workers to exhaustion and then replacing them. One picker wrote that the harried workers still show one another desperate support, like trying to cover for a coworker who’s running late, but with little wiggle room for solidarity.

The most nationally infamous episodes involve the Allentown, Pennsylvania, fulfillment centre’s issues with heat. Amazon balked at providing AC for its midwestern and mid-Atlantic facilities, perhaps not realizing the feverishly hot and humid summers in the region. When summer temperatures reached triple digits Fahrenheit in the warehouses, rather than spring for climate control, the company decided to add a few minutes to breaks and hand out water or Gatorade to the workers running from shelf to shelf in the stifling heat.

But, in Allentown, the escalating heat index and ludicrous productivity demands meant employees worked through pain and heat stress to the point where some became faint or passed out on the warehouse floor. This situation reached such a cartoonish extreme that Amazon hired paramedics to sit in an ambulance on-site, ready to treat heat-stressed and dehydrated employees. Some were taken to local hospitals and replaced by unsuspecting new hires.

This nightmare persisted until, after admitting temp worker after temp worker for heat strain, an ER doctor called government regulators. This complaint was followed by other witnesses, including “a security guard who reported seeing pregnant employees suffering in the heat,” according to the Allentown Morning Call, whose bold reporting on a major local employer earned national prominence. Subsequent Occupational Safety and Health Administration inspections found fifteen workers fainted from heat on one June day, leading to the installation of more fans and the distribution of cooling bandanas. Workers taking time off to recover incurred disciplinary points against them, which could add up to firing. Some took to chanting, “End slavery at Amazon.”

Stone suggests the bad PR around this and other episodes led to Amazon’s installation of AC in more fulfillment centres. Thanks to government disclosure rules requiring publicly traded firms to report their median worker salaries, we know Amazon pays an average fulfillment centre worker $28,446 a year, or $13.68 an hour, typical of the US warehouse industry but not reflective of Amazon’s incredible revenues and market stature. On the other hand, Amazon’s federal filings say, “We consider employee relations to be good.” So that’s reassuring too.

These kinds of practices help explain the enduring interest in unionization. But, like the rest of Silicon Valley, the company has resisted unions for its workers, and while Stone notes that the Teamsters and the United Food and Commercial Workers have “tried to organize associates” in the fulfillment centres,

Amazon’s logistics executives quickly met these campaigns by engaging with employees and listening to complaints while making it clear that unionizing efforts would not be tolerated. The sheer size of Amazon’s workforce and the fact that turnover is so high in the fulfillment centers make it extremely difficult for anyone to organize workers. . . . The unions themselves say there’s another hurdle involved—employees’ fear of retribution.

This is understandable after episodes like Amazon’s closure of a call centre, in 2001, that the company claimed was unrelated to labour organizing there. Joe Allen wrote perceptively in Jacobin that “Walmart consciously built its distribution centres in remote, conservative regions of the country to avoid the threat of unionization,” a pattern followed by manufacturers and other giant retailers like Amazon. Operating globally does mean some traditions of worker solidarity will be unavoidable, as when thousands of European workers went on strike in late 2018, some chanting, “We are not robots.” In Spain, Amazon was reported to have even requested that police intervene in a strike and enforce worker productivity levels, which the baffled police declined to do.

In the United States, reports uncovered that Amazon even had its warehouse staff sign noncompetes in their hiring contracts, barring them from work in any field competing with Amazon in any way—an almost-impossible standard to clear. After public backlash, Amazon announced it would stop requiring this, but only of its hourly temp workers.

Many earnest book lovers now often turn to their Kindles for much of their casual reading, downloading books and reading them on the go. However, few read the tablet’s fine print, which gives the company control over access to Kindle-formatted e-books even after they’re downloaded onto users’ physical machines. And so there was surprise, in 2009, when Amazon deleted some editions of certain books—the rights to which hadn’t been properly cleared—from users’ Kindles.

Amazon controls the wireless network that allows downloading and synchronizing e-books, and the company is able to delete them without user action or permission. Among the books remotely deleted were editions of Nineteen Eighty-Four by socialist writer George Orwell. Orwell would probably nominate Bezos, the corporate tyrant, as Big Brother’s capitalist twin.

Excerpted from Bit Tyrants: The Political Economy of Silicon Valley by Rob Larson. Copyright © 2020 Fernwood Publishing. All rights reserved.

Rob Larson
Rob Larson is a professor of economics at Tacoma Community College and author of Bleakonomics and Capitalism vs. Freedom from Zero Books.