Earlier this year, the Institute on Taxation and Economic Policy, a liberal think tank, reported that Amazon, one of the most valuable corporations in the world, paid no federal taxes on a supposed $11.2 billion in profits in 2018. Many Americans felt outraged, and shortly thereafter Senator Elizabeth Warren introduced a plan to force companies like Amazon to pay their “fair share” of taxes.
But in this case, the outrage was somewhat misplaced. We should not be astonished that Amazon pays no taxes, for the simple reason that it doesn’t actually turn a profit. While the company used accounting techniques to show a positive cash flow on paper, its zero-dollar tax bill more accurately reflects the nature of the business.
Today, many firms, not just Amazon, have adopted a growth strategy based on rapid expansion and negative cash flow. They are propped up by investors and by low interest rates that provide cheap and easy access to capital. They can’t be unprofitable forever, the thinking goes, and they must have an exit strategy, even if they don’t share it publicly. Until then, they continue to hemorrhage cash in their quest for an ever greater market share. The orthodox narrative on Wall Street is that these firms are reinvesting what would otherwise be profits, instead of sharing them with investors and shareholders. This narrative suggests that we are witnessing one of the greatest wealth transfers in the history of capitalism. By investing all their profits back into the firm, these companies are essentially transferring wealth from their investors to us, the consumers.
However, it’s unclear how, or even whether, that’s actually happening. Selling below cost is a classic way for aspiring monopolists to seize market share from smaller competitors who can’t afford to consistently lose money. This technique, known as predatory pricing, is bad for consumers, and the economy as a whole, because it drives companies out of the market not because they’re less competitive or efficient, but because they don’t have enough funds to survive without turning a profit. That’s why predatory pricing is illegal under federal antitrust law.
Today the U.S. economy is rife with spectacularly valuable corporations that fail to turn a profit, relying on the continuing faith of investors. It’s not just Amazon: Uber, Netflix, and WeWork are some of the many other examples. To the average person, these companies appear to be using super-low prices to gain market share. But if predatory pricing is illegal, how can this be happening?
The answer is that what the average person thinks about Amazon’s business strategy doesn’t matter, because the Supreme Court has all but defined predatory pricing out of existence. Taking cues from the conservative law and economics movement, the Court has held that the strategy is irrational as a matter of economic theory, because for it to pay off, the monopolist will have to recoup today’s losses by raising prices dramatically in the future. But that won’t work, the logic goes, because when they do, competitors will swoop in and offer the same service or product at lower prices, frustrating the entire scheme. Under that thinking, the Court has set up rules making it nearly impossible to prove that predatory pricing is happening.
But the Court and most antitrust scholars have been making a systematic mistake. The prevailing doctrine assumes that there is only one way for a company to recoup its losses once it has cornered the market: raising prices. It ignores the other half of the profit equation: costs. This is a serious error, because giants like Amazon have tremendous power to lower costs by squeezing other parties, like employees and suppliers. When you take both costs and prices into account, predatory pricing begins to look much more rational, and therefore more common, than the courts have imagined.
To the average person, companies like Amazon, Uber, Netflix, and WeWork appear to be using super-low prices to gain market share. But if predatory pricing is illegal, how can this be happening?
In early June, the Federal Trade Commission, Department of Justice, and House Antitrust Subcommittee all announced that they would be opening investigations into tech companies. So far they haven’t said anything about whether they will look into predatory pricing. It would be a mistake not to. Amazon and the other “unicorns” on Wall Street claim to be heavily investing in innovation now to reap the benefits in the future. But there is reason to believe that in addition to making these legitimate investments, they are spending heavily to subsidize selling below cost. In other words, they may be getting away with predatory pricing in broad daylight. If federal regulators don’t start asking the right questions, however, we may not know until it’s much too late.
Predatory pricing is not a new phenomenon. It was one of the allegations brought against John D. Rockefeller’s Standard Oil in 1911, when the Supreme Court decided to break up the company. In subsequent decisions, the Court came to hold that the practice was illegal under federal antitrust laws. The rationale was simple. It’s good when a company gets so efficient that it can charge less than its rivals. But if a company reduces prices below its own costs, then that doesn’t reflect efficiency, and must instead be aimed at cornering the market. That’s harmful to consumers in the long run, because eventually the company will start charging higher prices than it would if it hadn’t crushed the competition.
Through the late 1970s, cases involving predatory pricing were common, but critics pointed out that the courts lacked a consistent framework for deciding whether it was happening in a given situation. In 1975, a groundbreaking Harvard Law Review article established a straightforward test: to prove that a company’s pricing scheme is predatory, a plaintiff must show both that the prices are below the cost of producing a single product or service and that using that strategy to eliminate competitors is economically rational. The scheme counts as rational if the firm will be able to recoup its losses in the future “through higher profits earned in the absence of competition”—the whole point of having a monopoly.
But just a few years later, this formula was subtly revised. In his 1978 magnum opus The Antitrust Paradox, which provided the blueprint for a conservative counterrevolution in the field, Robert Bork narrowed the definition of “recoupment” from turning a profit to specifically charging higher prices. And any company whose strategy depended on raising prices in the future, he reasoned, would run into the problem of new competitors emerging to undercut them. Predatory pricing, he concluded, was almost always irrational, and courts should be highly suspicious of parties bringing forward such claims.
That position, like most of Bork’s views on antitrust, made its way into the official doctrine of the Supreme Court, which assumed that investors and shareholders would refuse to allow companies to lose money on a scheme that was unlikely to work. And so, according to the judiciary and mainstream antitrust lawyers, we have little to fear from the practices of companies that sell products and services for ridiculously low prices, year after year, without making a profit.
But the rise of platforms that are both insanely valuable and persistently unprofitable has made the Court’s assumptions look increasingly shaky. As Amazon, Netflix, and Uber have shown, investors and shareholders can be more than willing to tolerate losses if they expect the firm to eventually translate these losses into a money-making scheme. The question is: How?
As I argued in a paper recently published in the Oxford Journal of Antitrust Enforcement, scholars on both sides of the antitrust debate have been overlooking the other side of the predatory pricing equation: lowering costs without passing those savings along to the consumer.
Take Amazon. While the company publicly claims to be profitable, it has reported a cash outflow in statements filed with the Securities and Exchange Commission as recently as the end of 2017. That means that—once various debt obligations were taken into account—Amazon was losing money. For example, for the 2017 calendar year, it reported a net cash outflow of $1.5 billion. This raises the question: If Amazon is indeed losing money because it is pricing below cost, what’s its strategy for recouping those losses down the line?
In my paper, I argued that Amazon, which dominates nearly 50 percent of all e-commerce, might eventually recoup its losses by growing so efficient that tomorrow’s costs drop far below today’s prices. Then, in the absence of competition, it would face little pressure to pass the savings onto customers.
However, in the weeks that followed the publication of the paper, a more ominous hypothesis occurred to me. It begins with the fact that Amazon’s e-commerce business is really composed of two main parts. In the first, Amazon operates as a retailer, buying products in bulk from vendors and then selling directly to consumers. Under this model Amazon bears all of the costs associated with storage, fulfillment, and shipping. This operation is similar to a traditional brick-and-mortar shop. The second branch is the Amazon Marketplace: a virtual mall in which sellers pay Amazon for the right to display and sell their goods on its platform. The crucial difference is that in the Marketplace, sellers shoulder the costs associated with storage and fulfillment. In recent years the share of the Marketplace has grown dramatically. In 2001, 6 percent of merchandise sales on Amazon came through the Marketplace. Today, the figure is around 58 percent.
Here we see the potential for Amazon to recoup its losses in a way that Robert Bork never imagined. Instead of raising consumer prices, Amazon can sell the same products, for the same price, but push more and more vendors to become third-party sellers on the Marketplace—offloading the costs of fulfillment and allowing Amazon to charge those same entities higher fees.
This recoupment process appears to be under way. Bloomberg recently reported that Amazon was preparing for a “supplier purge”—ceasing to buy from thousands of wholesale vendors, and pushing them instead to become sellers on the Marketplace. At the same time, there is anecdotal evidence that Amazon is making terms less generous to sellers, so that Amazon keeps a bigger cut of the money coming in. In early May, I was contacted by the CEO of a company that has generated over $65 million in revenue on the Marketplace over the last five years. He spoke on the condition of anonymity, since he relies on Amazon for his livelihood, but agreed to let me look at his company’s books. His story matched accounts I’ve heard from other sellers.
Until recently, the CEO explained, Amazon treated his company well, because sellers like him solved a key problem: big brands were refusing to sell directly to Amazon, meaning their products were unavailable on the site. Sellers on the Marketplace filled this crucial void. They would buy products from the big brands and then sell them on Amazon, thereby circumventing the embargo. Amazon treated sellers like royalty in exchange. For instance, the CEO explained, the company charged unusually low fees for storing his company’s products, and if Amazon made a mistake in the fulfillment process or lost products shipped by sellers, it would quickly reimburse them, no questions asked. And most importantly, it offered its own fulfillment services, known as “Fulfilled by Amazon,” or FBA, at extremely low prices—far lower than what it would have cost the CEO to fulfill orders on his own. Those were the good days, he said, when his business grew from just under $1.5 million to more than $6 million in annual revenue.
The Supreme Court assumed that investors and shareholders would refuse to allow companies to lose money year after year. But the rise of platforms that are both insanely valuable and persistently unprofitable has made the Court’s assumptions look increasingly shaky.
But now he fears that his company’s days are numbered, due to ever shrinking profit margins. What happened? First, FBA became much more expensive. If back in 2014 fulfillment fees stood at 17 percent of the seller’s total costs, they are currently hovering at about 27 percent. Second, Amazon now charges much more for inventory storage: over the past four years, the monthly rate per cubic foot that the CEO pays has increased by over 40 percent. (Add to this the fact that more recently, the CEO said, he has had to pay Amazon for advertising for his products to have a chance of appearing at the top of search results. In the past, a successful product could top the list without a boost from ad money.) Amazon touts that it continuously invests in research and development and improving vertical integration (such as buying its own fleet of planes), ever striving toward greater cost efficiencies. It’s hard for the CEO to square that with the fact that fulfillment services fees have grown by about 60 percent as a proportion of his costs over the past five years.
Why not leave Amazon fulfillment, then? It’s too risky, the CEO explained: even if he could raise enough capital to set up an efficient fulfillment infrastructure, his company would risk bankruptcy in the event that Amazon suspended its account for any missed or delayed deliveries. And leaving Amazon entirely would be suicide for most sellers.
All in all, it appears that if Amazon was indeed engaging in predatory pricing, it has now moved on to the recoupment phase by shifting the costs of fulfillment onto third parties and by squeezing higher commissions and fees from those sellers. By Amazon’s own account, those represent one of the company’s fastest-growing sources of revenue. In 2018, Amazon’s cut of the revenue from third-party sellers totaled $42.7 billion, which translated to nearly one out of every five dollars the company made.
The threat of predatory pricing goes far beyond Amazon. Uber, for instance, has been up front in public announcements, including the paperwork for its IPO, about the fact that it is running its Uber Pool service at a loss in order to gain market share. And it has even discussed the need to recoup its losses by lowering the cut of fares that drivers receive. It’s far from clear that Uber’s strategy will work, as its lackluster IPO suggests—the ride-share industry may be impossible to monopolize, since the barriers to entry for new rivals are relatively low. But Uber is just one high-profile example. WeWork, another cash-burning mammoth gearing up for an IPO, is another. It could be leasing and buying office space around the world in order to attract customers, only to recoup its losses by squeezing landlords once it snatches up a large enough market share of the world’s prime real estate.
Ultimately, however, we generally lack concrete evidence that these companies have been charging below cost—the key way to determine whether predatory pricing is taking place—because that information is not part of mandatory corporate disclosures. Similarly, we can’t know for sure that the squeeze Amazon sellers are feeling these days is part of a recoupment plan. But the available indicators should be triggering alarm bells in Washington. As part of their upcoming investigations of the tech industry, the Federal Trade Commission, Department of Justice, and House Antitrust Subcommittee should make companies turn over the information that would show whether they’re violating the prohibition on predatory pricing. For a more forward-looking solution, regulators should create new rules requiring companies that consistently lose money to submit a confidential report showing whether they’re pricing below cost.
The good news is that at least one powerful member of the judiciary, which wields disproportionate power over how antitrust law is applied, seems somewhat attuned to the dangers of corporations that use their position as a dominant buyer to impose unfair terms on sellers—a phenomenon known as monopsony power.
Instead of raising consumer prices, Amazon can push more and more vendors to become third-party sellers on the Marketplace—offloading the costs of fulfillment and allowing Amazon to charge those same entities higher fees. This process appears to be under way.
As the judge put it in a 2017 decision of the U.S. Court of Appeals for the D.C. Circuit, “[T]he exercise of monopsony power to temporarily reduce consumer prices does not qualify as an efficiency that can justify an otherwise anti-competitive [activity]. . . . Although both monopsony and bargaining power result in lower input prices, ordinary bargaining power usually results in lower prices for consumers, whereas monopsony power usually does not, at least over the long term.”
That was Brett Kavanaugh, now a member of the Supreme Court and, perhaps surprisingly, a potential swing justice when it comes to antitrust law. In May, Kavanaugh joined the Court’s four liberals to allow an antitrust case to proceed against Apple based on the terms it imposes on third-party developers in the App Store. “A retailer who is both a monopolist and a monopsonist may be liable to different classes of plaintiffs—both to downstream consumers and to upstream suppliers—when the retailer’s unlawful conduct affects both the downstream and upstream markets,” he wrote in the majority opinion.
There’s no doubt that negative cash-flow juggernauts like Amazon have used technological innovations to provide consumers with excellent services at a fantastic bargain. The question is whether they are also using illegal tactics to push competitors—including tomorrow’s would-be innovators—out of the market. If they are, and they aren’t stopped soon, we will likely come to find that the bargain wasn’t worth it.
