Some say the world will end in fire,
Some say in ice.
From what I’ve tasted of desire
I hold with those who favor fire.
But if it had to perish twice,
I think I know enough of hate
To say that for destruction ice
Is also great
And would suffice.
-Robert Frost, “Fire and Ice”
Nothing in this world lasts forever, even Amazon. In 2018, Jeff Bezos told his employees, “Amazon will go bankrupt. If you look at large companies, their lifespans tend to be 30-plus years, not a hundred-plus years.” In four years, Amazon will hit that 30-year mark. Imagining the fall of the titan that is Amazon during the next U.S. presidential election seems far-fetched, and when Bezos made that announcement, he issued it as a challenge, not an inevitability. Much of this series has been spent discussing how Amazon has entrenched itself in the American economy, society, and identity. Yet the company must fail, one way or another, and today I examine how. From a natural death due to slowed growth, poor investments, or economic downturn; to various government-ordered dissolutions; to a zombification by government takeover, possibilities abound. Yet each death has perils not only for Amazon, but for our society, and a gradual dismantling now could help us avoid great pain down the road.
Of the natural deaths for Amazon, slowed growth and stagnation ranks high. Two years ago, Forbes’s Richard Kestenbaum warned that “if Amazon doesn’t find new sources of revenue growth in other industries, its expansion will slow. And because its stock price has been so influenced by revenue growth, it won’t continue to rise.” Without rising stock prices, Kestenbaum foresaw potential employees turning to other tech companies with better stock options and room for growth. The result would be a gradual decline in talent and innovation at the company, allowing competitors to topple its throne. Along the same lines, Bloomberg’s Shira Ovide warned last year that Amazon’s flat-lining international growth posed a long-term threat to the company’s longevity.
Although these warnings are reasonable, they seem to be quite early. Amazon has continued to grow rapidly over the past two years, boosted by the global pandemic. In the third quarter of 2020, Amazon posted its highest quarterly profits ever, over $5.1 billion, even though the company spent an additional $4 billion on expenses related to coronavirus. Internationally, sales in traditionally difficult markets for the company such as in Italy have remained at higher levels even as pandemic concerns have fluctuated. To the extent that Italy is an indication of the rest of the world, Amazon’s pandemic surge will not be ephemeral, and stagnating growth will likely remain hypothetical.
A less likely scenario for Amazon’s decline is that a downtick in sales would trigger a rapid fall in profits. For its first decades of existence, Amazon posted razor thin (or negative) profit margins as a percentage of its revenue. Those profits followed a large rise in free cash flow for the company (essentially the amount of money the company has on hand in a given quarter that, for Amazon, is larger than its profits because the company collects payments for its products an average of 18 days before it has to pay the cost of providing the product). Growing free cash flow allowed the company to profit on returns from investing income before payment was due, sustaining the tiny margins of profitability. This system of profitability, however, depends heavily on continued sales. Even a small decline in sales would shrink Amazon’s free cash flow and thus erase its potential investment returns.
Before 2016, this threat was more significant for Amazon than now because the company has expanded to other more profitable endeavors. In particular, its cloud computing arm Amazon Web Services has comprised a majority of Amazon’s net income since 2016 and continues to grow each year. With the addition of growing advertising buys and fees from third party sellers, the company has found new avenues of profit even as its retail arm continues to grow in size, but not so much in profits.
While a natural death for Amazon does not appear imminent, even if it did, it would likely not be desirable. Given the trend toward consolidation that has dominated the tech industry and the American economy at large for the second decade of the 21st century, it is unlikely that Amazon will be replaced by a more competitive set of businesses. Just as Amazon has been framed as the company that entered retail as a challenge to giants like Walmart, the company that succeeds Amazon will likely be even bigger.
This ever-increasing consolidation poses both economic and political dangers. The nonpartisan think tank Third Way linked industry consolidation with lower firm entry rates, suggesting that consolidated industries are bad for startups. Whatever company arises after Amazon will likely surpass its ability to strongarm small businesses and eliminate potential competitors while keeping antitrust regulators at bay. Columbia Law Professor Tim Wu also notes in his book, The Curse of Bigness, that consolidated industries increase corporate lobbying as firms are willing to invest more money when they have a greater market share. “The fewer members of the industry,” Wu writes, “the fewer among whom the gains are split.” The political dangers of consolidated industry power were among the chief fears of the founders of American antitrust law. Consolidated power in any form is a fear of Americans across time, space, and the political spectrum. If Amazon discomforts us when it employs its size to secure its standing, its successor could be worse.
While Amazon might prefer a natural end, if it continues to wield its power anticompetitively or if Congress or the courts seek a more active role in the structure of the American economy (perhaps to counter the consolidation addressed above), the government may intervene. As of now, two particular antitrust suits seem plausible: breaking apart Amazon Marketplace from the retail branch of the company and severing Amazon Web Services from the rest of the company. Though I have written more extensively on why these are plausible suits, suffice it to say that these parts of Amazon’s business give it the most leverage over its competitors (and the U.S. government). If such a case was brought to dissolve Amazon, or at least parts of Amazon, it would likely span years and presidential administrations. The last big successful trust-busting case (against AT&T) started under Johnson’s administration and ended under Reagan’s. That kind of shared, sustained commitment alone might render such an effort unlikely, but setting aside such doubt, it is worth it to examine what might happen if it was successful.
Contrary to the promises of the most avid antitrust proponents, breaking up Amazon would likely not deliver significant immediate economic benefits to consumers. Of the major antitrust cases in American history—among them Standard Oil, American Tobacco, Paramount Pictures, and AT&T—antitrust scholars largely agree that prices did not fall for consumers, nor did output increase as a result of dissolution. After the Reagan administration finalized the breakup of AT&T in a settlement, chairs of the Senate Commerce Committee and House Telecommunications Subcommittee (GOP Sen. Bob Packwood and Dem. Rep. Timothy Wirth) expressed worries that the breakup would lead to higher prices for consumers on local calls (although significant rises did not come as feared). Thus, under the current understanding that antitrust enforcement should advance consumer welfare as measured by price, it is hard to justify intervention. Recent free market thinkers (neoliberals) have even concluded that Standard Oil itself should never have been fragmented.
That people question the Standard Oil dissolution testifies to the dramatic narrowing of antitrust thought over the past decades, much to its detriment: even without direct consumer benefit through price reduction, antitrust enforcement can deliver structural improvements to the American economy and society, and even aid shareholders. Even Robert Bork, the most prominent architect of consumer price theory that governs antitrust law today, wrote in his seminal work, The Antitrust Paradox, that the Standard Oil and American Tobacco decisions were sound on the grounds that the monopolies had achieved their dominance illegitimately and outside the interests of consumers.
Beyond Bork’s limited acceptance of the two decisions, the dissolution had positive effects on the market for shareholders. The Economist magazine tells the story of John D. Rockefeller learning that his company, Standard Oil, was going to be dissolved: Rockefeller turned to his golf partner and said, “buy Standard Oil.” He was right. In the year after the company was broken up, the value of the 34 companies that sprang from it collectively doubled. By a few years later, it had quintupled. Even today, some of the world’s largest companies—Exxon, Mobil, and Chevron—are the product of that breakup.
Breaking up Amazon could also strengthen the American economy outside of direct price measurements. Wu notes that “today, concentrated economic power is used to avoid raising wages, to insist on intense conditions of employment, to abuse of ‘non-compete’ agreements, and to hire part-timers.” He adds that the new “focus on ‘allocative efficiency’ yielded almost no consideration of the ‘dynamic’ costs of monopoly, like stagnation or stalled innovation.”
Finally, taking action to reign in Amazon now could shield the American economy (and government) from the disaster of Amazon’s sudden death. On February 28, 2017, Amazon Web Services crashed, taking down huge portions of the internet across America, potentially including U.S. intelligence agency databases. Fortunately, the company was able to get its servers back online in four hours and blamed the outage on an employee error. The episode underscored the vulnerabilities of such a system, and that outage was internally-caused and isolated to only some S3 servers. Yet with Amazon hosting 2.2 million small businesses on its platform (700,000 of whom use it as their only selling place) in addition to the million-plus AWS clients, the company is too big to fail. Perhaps a slow death as described at the beginning would be tolerable, but the sudden collapse of Amazon due to circumstances seen or unforeseen would be unacceptable.
Such a collapse points to an alternative to Amazon’s death: zombification. Instead of breaking Amazon up or hoping that it dies peacefully, some people have suggested that the government could, in one form or another, take (partial) control of it. In her Yale Law Journal note, “Amazon’s Antitrust Paradox,” antitrust scholar Lina Khan suggested as an alternative to antitrust action that the government could think of Amazon as a “public utility” (though she ultimately prefers antitrust). Such a move would eliminate competition in some of the industries in which Amazon is involved (for Khan, Amazon Marketplace). In Khan’s words, a public utility regime “accepts the benefits of monopoly and chooses instead to limit how a monopoly may use its power.” Public utilities have traditionally been employed for industries such as water and grain transport that are considered important enough not to be left to market forces.
We see a glimmer of the same idea reflected in Theodore Roosevelt’s “New Nationalism” campaign in 1912, in which the nation’s most flashy (if not diligent) trust-buster called for embracing big companies that submitted to government regulation. Roosevelt called for much less regulation than the public utility framework, but the essential commitment to the benefits of monopoly remains. (Roosevelt was defeated in 1912 by Woodrow Wilson, whose “New American” plan envisioned a robust free market of small and medium-sized businesses.)
As I have already given away by my choice of language, the Amazon Zombie plan should be one of last resort—or better, not at all. The dynamism of the internet economy thrives on robust, healthy, protected competition. Before the consolidation seen in this decade, internet companies innovated rapidly in a competitive market. Attempting to design a framework for government control of a company like Amazon Marketplace, as Khan notes, is a daunting (and perhaps unworkable) task. The internet, not to mention internet retail, is too responsive to the market. Further, as this article begins, no systems last forever. Establishing a regulatory framework around Amazon that does not address the ability of Amazon to use its power to influence the government would quite possibly take us the way of Dr. Frankenstein. The government’s place is not to run this market but rather to protect the people in it.
There are a lot of ways for Amazon to meet its death (or, perhaps, undeath): a gradual decline due to slow growth and atrophy, increased antitrust scrutiny, and government takeover. Of the three options, antitrust action in one form or another to restructure the company appears to be the best. Of course, such action would be expensive, time-consuming, and difficult, but it is the only one able to address Amazon’s abuses of power while preserving a dynamic and competitive market. Antitrust action would not eliminate Amazon. In fact, none of the big antitrust cases in history destroyed the companies they targeted. Instead, they fought against the tyranny of big businesses over little ones and over governments, against the abuses of monopoly, and for competition and innovation. Our economy, like our society, is best when it is free. That freedom needs protection in both.