Collusion to Control a Powerful Customer: Amazon, E-Books, and Antitrust Policy

BY JOHN B. KIRKWOOD, 69 U. Miami L. Rev. 1 (2014).

Introduction: Few recent antitrust cases have generated as much interest—and controversy—as “the e-books case.” It involved five leading publishers, two of America’s best known high-tech firms (Apple and Amazon), and a secret conspiracy, fueled by Steve Jobs himself, to force Amazon to charge higher prices. According to the Justice Department, the publishers—encouraged and assisted by Apple—agreed to impose a new pricing model on Amazon, which sharply increased the prices of new and bestselling e-books. In the Justice Department’s telling, it was a routine case: “Stripped of the glitz surrounding e-books and Apple, this is an unremarkable and obvious price-fixing case appropriate for per se condemnation.”

But it was hardly routine in the eyes of many inside and outside the industry. To the contrary, the response to the Justice Department’s first round of proposed settlements, which allowed Amazon to resume its price cutting on e-books, was intense and highly critical. The settlements generated 868 public comments, and they were “overwhelmingly negative. ”To be sure, many of them were submitted by Amazon’s competitors, who may have been cloaking self-serving objections in public interest terms. Other comments, however, could not be so easily swept aside. For example, the Authors Guild, which represented individuals who were not participants in the conspiracy and who would normally prefer competition in the distribution of their ideas, opposed entry of the consent orders. Numerous commentators also criticized the government’s action, including Professor Richard Epstein, who concluded that “the betting here is that this law suit is a mistake.”

What united the critics was concern with Amazon’s dominant position as an e-book retailer and its aggressive tactics, particularly its below-cost pricing as a seller and its hard bargaining as a buyer. Its low prices could drive out other booksellers and ultimately increase retail prices, while its buyer power could reduce gains to publishers and authors, stunting the development of new titles. One common refrain was that the government was focused on collusion by the publishers and Apple when the real problem was the threat of an Amazon monopoly. Scott Turow, President of the Authors Guild, declared that “Amazon was using e-book discounting to destroy bookselling, making it uneconomic for physical bookstores to keep their doors open. Michael Shermer stated: “What this lawsuit probably will do . . . is return to Amazon the power to monopolize the e-book market through predatory pricing to the detriment of publishers, authors and, ultimately, readers.” David Carr asserted that Amazon was not only using below-cost pricing to block entry into e-book retailing, but wielding its buyer power to coerce better terms from publishers. Steven Pearlstein asked the ultimate question: “[A]re monopolies so bad that we might want to tolerate a little price-fixing by customers or suppliers in order to break them? Could a little anti-competitive behavior actually be procompetitive?”

This article addresses that question. It asks whether suppliers ought to be allowed to collude to counteract the power of a large customer, either in the circumstances of the e-books case or in other settings. The issue is important. If the e-books case was a mistake, it would cast doubt on the value of antitrust enforcement, suggesting that it seeks to preserve “competition” at the expense of consumers. Likewise, if collusion to control a powerful customer can never be permitted, no matter how beneficial the consequences, it would raise questions about the very purpose of antitrust law. Is the goal to preserve rivalry for its own sake, or is the aim to make consumers, small suppliers, and the economy better off? Finally, if, as a normative matter, collusion to control a powerful customer should be allowed in some circumstances, antitrust law must address the pragmatic issue: Can those circumstances be identified with sufficient clarity to permit a workable and limited defense?

Part II begins by summarizing the e-books case. This review leaves little doubt that the defendant publishers conspired, with the enthusiastic assistance of Apple, to impose a new pricing model on Amazon and other retailers, causing the prices of popular e-books to rise substantially. The question is whether that conspiracy was nevertheless desirable because it reduced Amazon’s power and aggressive behavior, benefitting consumers in the long run.

To lay the groundwork for that inquiry, Part III reviews the state of the law and the scholarship on whether collusion is permissible if it offsets the power of a large customer. The case law is clear. Virtually all cases apply a rule of per se illegality to hard-core collusion—agreements among rivals that set prices or output but do not involve any productive integration—even if the colluders are selling to or buying from a firm with substantial power. Only one case has recognized the benefits of offsetting power. Similarly, most scholars would not let firms collude to exert countervailing power, concluding that a defense for such behavior would do more harm than good. Several economists have shown, however, that when small suppliers face a customer with monopsony power, they can protect themselves and benefit consumers by colluding to force the monopsonist to raise its input price and increase output.

Part IV evaluates whether this—or any other rationale—justified the e-books conspiracy. The economists’ rationale did not because Amazon did not possess monopsony power. While it did have considerable countervailing power, there was no evidence that the exercise of this power had harmed consumers or that it was likely to pose a much more severe threat in the future. Amazon’s below-cost prices on many titles were a concern, but these prices were probably not predatory, and even if they had been, the proper response was not collusion but an antitrust lawsuit. Neither Amazon’s power as a buyer nor its behavior as a seller justified a conspiracy that imposed substantial price increases on consumers.

The larger issue, however, remains: Should suppliers be allowed to collude in some circumstances to offset the power of a large customer? Part V concludes that the answer is yes and proposes a limited defense to the per se rule to protect the most important instances of procompetitive collusion. That defense would have three elements. First, colluding suppliers would have to prove that their customer had buying power. In addition, the suppliers would have to demonstrate that this power, whether monopsony power or countervailing power, was legally acquired, substantial, persistent, and durable.Second, the suppliers would have to prove that their action was desirable. If the customer’s power was monopsony power, no further proof would be required, for counteracting monopsony power through the creation of supplier power is ordinarily procompetitive. If the power was countervailing power, however, the suppliers would have to furnish additional proof, because the exercise of such power is often procompetitive and, when that is so, it should not be offset. To show that the customer’s exercise of countervailing power had been anticompetitive, the suppliers would have to demonstrate that it had reduced their ability to innovate and that collusion would lead to more innovation. Third, the suppliers would have to prove that their collusion did not create downstream market power, for if it did, it would normally harm consumers.

These requirements are demanding and would seldom be satisfied. True monopsony power is unusual in the American economy, and substantial, persistent, and durable monopsony power is even rarer. While countervailing power is more common, suppliers could not often prove that the exercise of such power had an adverse effect on innovation. Finally, upstream collusion frequently produces downstream market power and, when that is so, suppliers could not collude. In short, Parts IV and V affirm the traditional antitrust hostility to collusion, both by rejecting the asserted justifications for it in the e-books case and by demonstrating that the circumstances required to justify it in other cases are highly restrictive. Nevertheless, there are some situations in which collusion would be justified, and it appears possible to identify and protect those cases through a limited and practical defense. When the necessary facts are established, collusion to control a powerful customer ought to be permitted. . . . Full Article.

Recommended Citation: John B. Kirkwood, Collusion to Control a Powerful Customer: Amazon, E-books, and Antitrust Policy, 69 U. Miami L. Rev. 1 (2014).

Leave a Reply

Your email address will not be published. Required fields are marked *