Printer Friendly

Reforming the Robinson-Patman Act to serve consumers and control powerful buyers.

I. Introduction

Measured by the standards of contemporary antitrust policy, the Robinson-Patman Act has two major flaws. First, it is a protectionist statute. Its fundamental aim is not to promote competition in order to benefit consumers, but to protect small business from competition. As the Supreme Court has explained, (1) Congress was concerned that a large firm could use its buying power to obtain discriminatory prices and promotional benefits from its suppliers and then wield the resulting competitive advantage to take business from--or even eliminate--its smaller competitors. Prohibiting such behavior, however, is frequently at odds with the proconsumer focus of the other antitrust laws. When a large buyer uses the concessions it has induced to take business from small rivals, it normally does so by lowering prices, increasing services, or otherwise benefitting its customers. (2)

The second major problem with the Robinson-Patman Act is that it fails in many cases to control the very buyer-driven discrimination it was passed to prevent. While several provisions in the Act limit its effectiveness, (3) the most important is the meeting competition defense, which allows a seller to grant a discriminatory price or promotional benefit to a large buyer if the seller believes in good faith that a competing seller is making a comparable offer. This provision, designed to protect innocent sellers, frequently protects large, aggressive buyers as well, since they typically induce concessions by playing sellers off against each other. This gives each seller a meeting competition defense but insulates the resulting discrimination from attack.

As a result, the Robinson-Patman Act, the antitrust law most directly targeted at buyer power, is usually unable to reach discrimination induced by a powerful buyer. It is highly likely, for example, that Wal-Mart, the world's largest retailer, and Amazon, the country's largest book seller, (4) obtain preferential prices and promotional benefits from their suppliers. Yet no one has brought a successful Robinson-Patman action against these buyers. To be sure, neither retailer has generally harmed consumers through the exercise of its buyer power. But there are many ways in which the exertion of buyer power could harm consumers, and as these buyers grow, those consequences become more probable. (5)

There are two major reasons, then, to revise the Robinson-Patman Act. The first is to reorient the statute, changing its fundamental goal from protecting small business to protecting consumers. This would bring the Robinson-Patman Act into the antitrust mainstream, confining R-P enforcement, like enforcement of the other antitrust laws, to practices that pose risks to competition and consumers. (6) This change would also increase the statute's appeal, making government enforcers more willing to bring R-P actions and courts more willing to sustain them.

The second reason to reform the Act is to convert it into an effective tool for combatting buyer-driven discrimination that threatens competition and consumers. If a powerful buyer like Amazon does extract discriminatory concessions that are likely to harm consumers, the antitrust laws ought to have a remedy. But the Shennan Act and the FTC Act may not provide it. (7) While these statutes are written broadly enough to reach anticompetitive buyer-driven price discrimination, they have never been used to stop it. To the contrary, several courts have declared that price discrimination furthers the purpose of the Shennan Act. If the Robinson-Patman Act were repealed, therefore, neither the Shennan Act nor the FTC Act may plug the gap. (8) Of course, Congress could close the gap by amending one of these laws, but the simplest and most straightforward legislative fix would be to revise the existing antidiscrimination provisions in the Robinson-Patman Act.

In this article I propose two sets of changes. First, Congress should delete the Act's competitive injury language, forcing plaintiffs alleging discrimination to show harm to competition, not just to individual competitors. Second, the Act's two principal defenses--meeting competition and cost justification--should be curtailed. One option is to eliminate the defenses in purely equitable actions. In other words, if a plaintiff seeks only equitable relief, the defenses would be unavailable. (9) Alternatively, the defenses could be eliminated in treble damage actions as well as equitable actions if the plaintiff proves by clear and convincing evidence that the challenged discrimination is like to harm competition. (10) But whatever option is chosen, neither defense should stand in the way of relief when the plaintiff establishes that a powerful buyer has induced an anticompetitive concession. (11)

These reforms are important. Critics have complained about the Robinson-Patman Act for years, arguing that it distorts competition, reduces efficiency, and harms consumers. (12) The first change would put an end to these criticisms, refocusing the Act on safeguarding competition and consumers, not on shielding small rivals for their own sake. The second change would ensure an effective remedy if a powerful buyer like Amazon or Wal-Mart does extract concessions that reduce rather than promote competition. Concern with Amazon's power remains especially high. It is the dominant retailer of books in the United States, with a share of e-book sales approaching monopoly levels, (13) and many commentators believe that the company already has too much power. (14) In addition, numerous authors fear that its demands will deprive publishers and authors of the revenues they need to maintain a rich supply of new books. (15) There is ample reason, in short, to ensure that anticompetitive buyer-driven discrimination can be stopped.

Part II of this article briefly reviews the goals of the existing Robinson-Patman Act, showing that in secondary line cases, (16) the Act's ultimate aim is indeed protectionist. Part III examines the need for a reoriented statute. It identifies the principal settings in which buyer-driven discrimination would harm consumers (or small suppliers) and presents evidence that powerful buyers like Amazon and Wal-Mart have triggered, or threatened to trigger, a number of these scenarios. Part IV explains why enforcement of the other antitrust laws may not reach many of these harmful cases. Part V describes the changes in the Robinson-Patman Act that would convert it into a statute that serves consumers and effectively controls powerful buyers. Finally, Part VI briefly compares other methods of controlling buyer power (structural relief and common carrier regulation) and concludes that, as a general rule, a Robinson-Patman Act remedy would be superior.

To simplify the analysis, the remainder of the discussion focuses exclusively on price discrimination. If its conclusions are correct, however, they should apply to promotional discrimination as well. A reformed Robinson-Patman Act should treat promotional discrimination essentially the same way it treats price discrimination. (17) Further, the article concentrates on suppliers with market power, since they are the most likely to engage in secondary line discrimination. Suppliers without market power cannot discriminate among their customers except in very limited circumstances. (18) In those circumstances, however, when a buyer with monopsony power forces those suppliers to disfavor the buyer's rivals, injuring the suppliers as well as the rivals, those sellers would deserve protection from the exercise of buyer power. (19) In the more typical case, in which a powerful buyer extracts concession from suppliers with market power, the suppliers deserve protection only to the extent that protecting them benefits consumers. (20) In other words, when a buyer exerts countervailing power against a supplier with market power, the overarching goal should be consumer welfare, not supplier welfare. (21) This is, of course, the preeminent aim of the original antitrust laws; it is not the fundamental purpose of the Robinson-Patman Act.

II. Purpose of Robinson-Patman Act

It is widely understood today that the fundamental goal of the original antitrust laws--the Sherman Act, the Clayton Act, and the FTC Act--is to preserve competition in order to protect consumers. (22) Firms that gain market power without justification can raise the prices they charge, reduce the services they provide, or exploit consumers in other ways, taking their wealth without providing compensating benefits. Congress passed the original antitrust laws primarily to prevent such exploitation and most cases now see the overarching aim of antitrust enforcement as protecting consumers from anticompetitive conduct. (23)

The Robinson-Patman Act, however, was passed for a different purpose. Enacted during the Great Depression, it was designed to ease the plight of small business, not help consumers. Congress amended Section 2 of the Clayton Act to make it more difficult for large firms to obtain discriminatory advantages from their suppliers and use them to take business from smaller competitors. Today, that protectionist purpose is manifest only in what are called "secondary line" cases.

In "primary line" cases, the Supreme Court declared two decades ago that the goal of Robinson-Patman enforcement is the same as the goal of Shennan Act enforcement. In a primary line case, the plaintiff is a competitor of the discriminating seller. Typically, the plaintiff claims that the seller is using geographic price discrimination to carry out a campaign of predatory pricing. Suppose, for example, that the seller has a monopoly in Pittsburgh but faces a new entrant, the plaintiff, in Cleveland. By charging the monopoly price in Pittsburgh, the seller can amass funds to finance below-cost pricing in Cleveland and attempt to drive out the plaintiff. If its predatory campaign succeeds, the seller can restore monopoly pricing in Cleveland, harming competition and consumers there.

In Brooke Group Ltd. v. Brown & Williamson Tobacco Corp,, (24) the Supreme Court observed that "primary-line competitive injury under the Robinson-Patman Act is of the same general character as the injury inflected by predatory pricing schemes actionable under [section] 2 of the Sherman Act." (25) As a result, the Court brushed aside contrary precedent and ruled that the standards for liability in a primary line Robinson-Patman case are essentially the same as those in a Sherman Act case. (26) The Court even stated that the Robinson-Patman Act should be construed "consistently with broader policies of the antitrust laws." (27) Today, in short, the aim of primary line Robinson-Patman enforcement is to protect consumers from anticompetitive conduct.

In a secondary line case, however, the overarching goal is to protect small business, not consumers. Speaking of secondary line cases, Calvani and Breidenbach declared, "It is quite clear that the underlying predicate of the Robinson-Patman Act was not consumer welfare. Rather, the Act is protectionist legislation." (28) In a secondary line case, the plaintiff is a customer of the discriminating seller and typically alleges that it was injured because a larger rival obtained a lower price from the seller and used the resulting competitive advantage to undercut the plaintiff, depriving it of sales or profits. The purpose of secondary line enforcement is to prevent this harm to a competitor, even when consumers benefit from the lower prices that the large buyer charges. In FTC v. Morton Salt Co., (29) the leading case on secondary line injury, the Supreme Court stated, "The legislative history of the Robinson-Patman Act makes it abundantly clear that Congress considered it to be an evil that a large buyer could secure a competitive advantage over a small buyer solely because of the large buyer's quantity purchasing ability." (30) This focus on competitive advantage, rather than consumer harm, led the Court to create the "Morton Salt inference," which allows a fact finder to infer competitive injury simply from a substantial and sustained price differential. (31) In its most recent secondary line case, the Court reaffirmed the Morton Salt inference. (32)

The Robinson-Patman Act's protectionist purpose in secondary line cases has deeply troubled many observers, for it can stop a large buyer from inducing a concession that benefits consumers. As noted earlier, when a powerful buyer obtains a concession and uses it to take business from smaller rivals, it does that by lowering prices, increasing services, or otherwise making its product more attractive to customers. This plainly benefits consumers, as long as it does not lead to some form of long-term harm. (33) Moreover, the demands of a powerful buyer can ultimately cause the collapse of an oligopolistic price structure, which not only enhances consumer welfare but also benefits smaller buyers. (34) Because Robinson-Patman Act enforcement can prevent or deter these desirable effects, (35) the Act has been subject to relentless criticism, (36) and many have called for its repeal, including the Department of Justice in 1977 and the Antitrust Modernization Commission thirty years later. (37) Repealing the Robinson-Patman Act rather than reforming it, however, would make it more difficult to challenge major categories of buyer-driven discrimination that harm consumers.

III. Consumer Harm from Buyer-Driven Discrimination

Buyer-driven discrimination can reduce competition and harm consumers or powerless suppliers in at least ten different settings. In prior articles, I have described these scenarios in detail. (38) Here, I list them and use several of the most important to explain why Amazon's behavior and the behavior of other powerful buyers has generated such concern. The first group describes the ways in which discrimination induced by a powerful buyer can reduce competition upstream, in the market or markets in which the buyer purchases inputs from suppliers. In the second group, buyer-driven discrimination diminishes competition downstream, in the market or markets in which the buyer and its rivals sell their products to consumers.

A. Harm to Upstream Competition

When a buyer exercises countervailing power to obtain discriminatory price cuts from a supplier with market power, (39) competition among suppliers could be reduced in at least five ways: (1) the buyer's exercise of countervailing power could diminish the returns that suppliers earn from research and development, curbing their incentive to innovate; (2) by depressing suppliers' profits, the exertion of countervailing power may cause suppliers to curtail the variety of products they offer, reducing consumer choice; (3) if the powerful buyer concentrates all its purchases in a single supplier, it could give that supplier monopsony power over small, powerless suppliers further upstream; (4) the exercise of countervailing power may turn into monopsony power if the structure of the supplying tier changes and suppliers lose their market power; and (5) the powerful buyer's exercise of its power may cause suppliers to collude in response, which may lead to higher consumer prices.

Amazon's power and behavior already have triggered the fifth scenario. Reacting to its below-cost pricing of many e-books, and fearing the exercise of its buyer power, five of the nation's leading book publishers, with Apple's enthusiastic assistance, colluded to adopt a new pricing model for e-books.

They imposed the new model oil Amazon and other retailers and raised retail prices sharply. (40) Prompted by a complaint from Amazon, the Justice Department challenged the conspiracy, obtaining settlements from all five publishers. (41) Apple insisted on litigating but lost in the district court. In an elaborate and sometimes scathing opinion, Judge Cote held that the high-tech giant had collaborated with the publishers to elevate e-book prices. (42) Because Apple has appealed, its ultimate liability is unknown, but there is little doubt that in the e-books market a powerful buyer provoked a backlash among its suppliers and the result was substantial consumer harm. (43)

The e-books conspiracy is over, however, and is unlikely to be repeated. At this point, Amazon's buyer power is most likely to hurt consumers in the ways described in the first two scenarios. In other words, Amazon's "take-no-prisoners attitude toward its suppliers" (44) may eventually reduce publishers' profits--and authors' incomes--to such a degree that it diminishes the number or variety of new books published. Amazon's public dispute with Hachette greatly heightened these concerns.

Hachette was the first major publisher to try to reach an agreement with Amazon after the e-books conspiracy ended. Because Hachette would not grant the concessions Amazon demanded, Amazon retaliated, not by cutting off Hachette, as it had done to many smaller publishers, (45) but by discouraging the sale of Hachette books. Amazon withdrew its usual discounts, eliminated quick shipment of Hachette titles, and even suggested that customers consider other options. (46) These actions, which not only hurt Hachette but also deprived its authors of income and exposure, (47) provoked intense public criticism. One of the country's leading literary agents stated, "It's very clear to me, and to those I represent, that what Amazon is doing is very detrimental to the publishing industry and the interests of authors." (48) Ursula Le Guin equated Amazon's tactics to government control of publishing: "Amazon is using censorship to gain total market control so they can dictate to publishers what they can publish, to authors what they can write, to readers what they can buy." (49) Franklin Foer wrote that in its drive for more revenue, Amazon might reduce the price of a novel to the "price of a Diet Coke" or might "prod the publishing houses" until they give up more profits. (50) "Either way, the culture will suffer the inevitable consequences of monopoly--less variety of products and lower quality of the remaining ones." (51) George Packer summed up the issue for many: "[T]he big question is not just whether Amazon is bad for the book industry; it's whether Amazon is bad for books." (52)

Although numerous authors remain dissatisfied with it, (53) Amazon settled its dispute with Hachette after it reached a similar agreement with Simon & Schuster. (54) Both agreements generally allow the publisher to set the retail prices of e-books but give it an incentive (a greater share of retail sales revenue) if the price is set lower. (55) The agreements also maintain the author's share of revenue. (56) They do, however, require the publisher to increase its promotional payments to Amazon (57) and thus may enlarge Amazon's share of revenue. It is possible, then, that these agreements may reduce the publishers' profits. But both Hachette and Simon & Schuster pronounced themselves pleased with the terms, (58) ending, for the moment, the public clamor over Amazon's buyer power.

While its dispute with Amazon imposed temporary losses on Hachette and its authors, there is no evidence that it led to a market wide decrease in innovation or variety. Indeed, it does not appear that Amazon's aggressive exercise of countervailing power has ever resulted in an overall decline in the number, variety, or quality of books published. Despite their intense criticism of Amazon, neither The Authors Guild nor any other group has shown that Amazon's power as a buyer has produced lasting harm to consumers. (59) In short, apart from short-term disruptions caused by its disputes with particular publishers, Amazon has not triggered the anticompetitive effects described in the first or second scenarios.

Instead, to date Amazon has used its buyer power to produce lower prices and better services for consumers. On this point, there is wide consensus. Surveying the history of the company, Brad Stone stated, "The publishers had seen over many years what Amazon did with additional leverage. It exacted more concessions and passed the savings on to customers in the form of lower prices and shipping discounts." (60) Franklin Foer, one of Amazon's strongest critics, noted that the retailer does not use its "power to hike up prices." (61) To the contrary, Amazon, Wal-Mart, and Google "have ushered in an era of low prices." (62) Joe Nocera rejected the view that Amazon is a monopoly, pointing out that "most people turn to Amazon for e-books not because there are no alternatives but because its service is superior." (63) Matthew Yglesias attributed Amazon's success not to its power but to what it provides consumers: "Everyone gets their e-books from Amazon because they're just as cheap as Apple's e-books, but they work on a much broader range of devices." (64)

Despite these benefits, there is reason to fear that Amazon's buyer power may cause consumer harm in the future. Its market shares, already large, (65) are likely to grow as Barnes & Noble struggles and Amazon relentlessly pursues lower costs and greater volume. As Amazon grows, moreover, its demands may escalate, reducing the profits of publishers and the advances of authors. Foer warns that author advances are "the economic pillar on which quality books rest, the great bulwark against dilettantism. " (66) Krugman agrees: "By putting the squeeze on publishers, Amazon is ultimately hurting authors and readers." (67) Amazon already has the power to discriminate among titles. It exempted The Way Forward by Paul Ryan, the former Republican vice presidential candidate, from the penalties it was imposing on other Hachette books, while treating Daniel Schulman's Sons of Wichita, a profile of the Koch brothers, like other Hachette titles, offering it without a discount and without prompt delivery. (68) Although Amazon did not say why it favored Ryan's book over Schulman's, Krugman suggested the discrimination was politically motivated, (69) a possibility that heightens the concern with Amazon's power.

In sum, buyer power can reduce competition upstream in at least five different ways. In Amazon's case, it has provoked a conspiracy among publishers and led to abiding fear that the number or diversity of new books published will be curtailed. (70) Moreover, Amazon's buyer power, like the power of Wal-Mart or Toys "R" Us, could also harm competition downstream.

B. Harm to Downstream Competition

By inducing suppliers to discriminate in its favor, a large buyer can also reduce rivalry in the downstream sale of its products or services, harming consumer welfare and economic efficiency. Like upstream harm, these adverse effects can occur in at least five different settings: (1) the large buyer may induce its suppliers to raise its rivals' costs, enabling it to increase prices or otherwise exploit consumers in downstream markets; (2) the buyer may extract price cuts or other concessions from its suppliers, who may react to the decline in their profits by increasing prices to the large buyer's rivals, allowing it to raise its prices or otherwise harm consumers; (3) the buyer may obtain discriminatory concessions that are so large and long-lasting that they enable it to drive out or greatly diminish the market share of smaller rivals, increasing downstream concentration and making tacit or explicit collusion more likely; (4) even if downstream prices never rise as a result of the elimination of the buyer's rivals, their destruction may deprive consumers of choices they preferred and depress overall consumer welfare; and (5) the discriminatory concessions induced by the buyer may permit it to become less efficient, less dynamic, and less responsive to changing consumer preferences. (71)

The first scenario occurred in Toys "R" Us, (72) where the nation's largest toy retailer induced leading toy manufacturers to refuse to sell popular models to Costco and Sam's Club, then a new and rapidly growing threat to established retailers. By raising the costs of these emerging rivals, Toys "R" Us curtailed their growth and deprived consumers of the option to purchase desirable toys at warehouse club prices.

The second scenario--what Dobson and Inderst have called the "waterbed effect" (73)--may have taken place during the bookstore wars of the (1990) s and early (2000) s, when Barnes & Noble and Borders, then the nation's leading bookstore chains, obtained substantial price and promotional concessions from book publishers. (74) In response, the publishers reportedly increased the average list price of their books, (75) causing prices at independent bookstores (who typically sell books at list) to rise. This also made it easier for Borders and Barnes & Noble to raise their own retail prices.

The bookstore wars also exemplified the third scenario. Armed with substantial discriminatory advantages, the national chains gained market share at the expense of independent bookstores, driving literally thousands of them out of business. (76) With independents weakened, retail markets concentrating, and publishers raising list prices, the national chains reduced the discounts they offered consumers. (77) The American Booksellers Association summarized the effect: "Rising list prices combined with disappearing discounts to consumers has meant that the chains have actually ultimately driven higher prices to consumers." (78)

The destruction of thousands of independent bookstores may also have led to the fourth scenario, in which consumer welfare falls, even though prices remain low, because consumers lose the ability to shop at outlets they value highly. An even better illustration of this scenario is the devastating impact that Wal-Mart's entry has had on many local retailers, which has hollowed out the centers of numerous small towns and left them blighted and vacant. (79) Concern over this effect has caused almost three dozen local jurisdictions to vote to prohibit the entry of Wal-Mart (or other big box stores) into their communities. (80) These votes suggest that consumers in these areas valued the preservation of a vibrant community center more than they valued Wal-Mart's low prices.

Professor Adelman described the fifth scenario in his classic study of The Great Atlantic and Pacific Tea Company. (81) He discovered that during the Great Depression, as suppliers found it more difficult to resist price cutting, A & P obtained greater discriminatory concessions. (82) Adelman concluded that this discrimination was anticompetitive, since it diminished pressure on the large chain to become more efficient and serve consumers better:
   [Pressure on the pocket-book nerve, which should have shaken A & P
   out of its lethargy, was dulled; the company was anesthetized by
   the concessions made to it.

   Hence, if our criterion for public policy is the promotion of
   competition to increase output and lower prices, then our public
   policy should largely condemn the discriminations in favor of A &
   P. (83)

Although Amazon's power could trigger any of these five scenarios, Amazon has avoided the first three by relentlessly pursuing a proconsumer image, marked by low prices, prompt shipping, and attractive services. (84) Given this track record, it seems unlikely, at least in the near term, that Amazon would attempt to raise book prices to supracompetitive levels and thus cause the principal type of consumer injury--high prices--described in the first three scenarios. (85) Likewise, there appears to be little immediate risk of the fifth scenario. Amazon's heavy emphasis on operational efficiency and cost cutting suggests it would not soon use the concessions it obtains to let itself become bloated and unresponsive to consumer preferences. At present, then, the principal risk is the fourth scenario, in which a powerful buyer reduces consumer welfare by eliminating smaller rivals that consumers value.

This risk cannot be dismissed. Over the last two decades, aggressive competition from the national chains drove thousands of independent bookstores out of business. (86) In addition, one of the two national chains, Borders, recently went bankrupt, no doubt due in part to Amazon's low prices, wide selection, and rapid delivery. (87) Thus, the ranks of brick-and-mortar booksellers have been severely depleted, and Amazon's continuing ability to undercut them threatens the viability of those that remain. Indeed, because of competition from e-books and online bookselling--competition spearheaded by Amazon--Richard Posner predicted that physical bookstores will fail in large numbers and may vanish altogether. (88)

While Posner characterized this as economic progress, it could actually reduce consumer satisfaction. Plainly, large numbers of consumers benefit from Amazon's pricing, selection, and service. But many consumers also value the warmth, convenience, and personal service of their neighborhood bookstore, plus the ability to pick up and browse through a book before making a purchase. (89) One might think that the market would sort out this conflict, letting competition between Amazon and independent bookstores determine which type of retailer benefits consumer most. There is a collective action problem, however, that may prevent the market from reaching the optimal outcome. (90) It is possible, therefore, that if Amazon continues to compete in the way it has--using its buyer power to obtain substantial concessions and using those concessions to grow at the expense of brick-and-mortar booksellers--it could reduce overall consumer welfare. If that occurs, however, and the Robinson-Patman Act has been repealed, the other antitrust laws may not provide a remedy.

IV. Limitations of Other Antitrust Laws

The fundamental goal of the other antitrust laws--the Sherman Act, the Clayton Act, and the FTC Act--is to protect consumers and small suppliers from anticompetitive conduct. (91) But these laws do not protect consumers and small suppliers from all anticompetitive conduct. Each of these antitrust statutes has textual, precedential, or policy limits that prevent it, or may prevent it, from reaching every form of rivalry reducing behavior that harms consumers or powerless suppliers. As a result, despite their breadth, the other antitrust laws may not halt every category of anticompetitive discrimination induced by a powerful buyer like Amazon or Wal-Mart. Indeed, they may not halt most harmful buyer-driven discrimination.

A. The Sherman Act

Section 1 of the Sherman Act prohibits agreements in restraint of trade. In theory, that prohibition would reach a discriminatory price cut induced by a powerful buyer that is likely to harm competition upstream or downstream. The discriminatory price would be embodied in an agreement--the contract of sale between the supplier and the buyer--and if the discrimination were likely to harm competition, the agreement would restrain trade. Thus, Professor Hovenkamp states that if the Robinson-Patman Act were repealed, "[section]1 of the Sherman Act would be sufficient to reach the relatively few injuries to competition that result from a supplier's pricing to its own dealers." (92) There are many reasons, however, why section 1 would not be a reliable bulwark against anticompetitive discrimination. First, the courts have never upheld its use for this purpose. Second, they frequently say that price discrimination promotes the policies of section 1. And finally, they may be unwilling to stop it when the supplier was meeting competition or the discrimination did not enhance the buyer's market power. (93)

To be sure, section 1 clearly reaches a conspiracy of competing buyers to force a supplier to engage in price discrimination. When a plaintiff alleges a horizontal combination, the courts are willing to entertain a section 1 claim. (94) But if the discriminatory price is contained in a purely vertical contract between the supplier and the buyer, and that contract does not require the supplier to discriminate against other buyers--the price in the contract is simply lower than the price charged competing buyers--the courts have been reluctant to find a violation. Indeed, no case has ever held that a vertical, nonexclusionary contract containing a discriminatory price offends section 1.

Some decisions have implied that such a contract could never offend section 1. A Tenth Circuit opinion declared, "We do not think section one of the Sherman Act requires the manufacturer to offer the same price to all its customers." (95) A Ninth Circuit decision stated, "[T]he courts have held that such an agreement, without proof of an arrangement to exclude others from the buyer's market does not give rise to a section 1 claim.", (96) A Third Circuit panel, after noting that "price discrimination simpliciter ... is usually not a Shennan Act violation," (97) suggested that a violation could be found only if the discrimination had been induced by a group of conspiring dealers. (98)

Other courts have indicated, however, that price discrimination may violate section 1, even if it is embodied in a vertical, nonexclusionary contract, if the discrimination is anticompetitive. In the best-known opinion, authored by then-Judge Breyer, the First Circuit ruled against the plaintiff because it had not shown anticompetitive effects. (99) But the court left open the door to section 1 liability, noting that it had simply held that "evidence of a violation of the Robinson-Patman Act, showing injury only to competitors, does not automatically show a violation of the Sherman Act as well." (100) Other cases have also said that price discrimination may violate section (1) if it produces an unreasonable restraint. (101)

At the same time, however, the courts have also set forth multiple reasons why price discrimination should not ordinarily--if ever--be challenged. First, when a supplier with market power grants a price cut to a large buyer, the supplier normally reduces its price closer to marginal cost in order to gain or retain the buyer's business. But the supplier does not usually cut its price below marginal cost, since that would cause it to lose money on every unit sold to the buyer. As a result, a discriminatory price cut is not typically a predatory price cut, and courts are understandably hesitant to condemn nonpredatory price cuts. As Breyer put it, "judicial efforts to prevent firms from charging low, nonpredatory prices, despite an occasional beneficial result, would more often significantly interfere with the achievement of the Sherman Act's basic and important low price objective." (102)

Second, a rule against discriminatory price cuts will tend to keep prices high, hurting consumers. Judge Breyer stated,
   If suppliers cannot charge low, nonpredatory prices without the
   threat of antitrust actions, they will hesitate to cut their
   prices. If suppliers must cut prices to all competing dealers or to
   none, if they cannot decide to favor a single dealer, ... they may
   well decide not to cut prices at all, perhaps to the benefit of the
   dealers, but certainly to the detriment of the Sherman Act's
   ultimate beneficiary, the consumer. (103)

For this reason, several other courts have declared that price discrimination furthers the aims of the Sherman Act. (104) Likewise, the Third Circuit declared that a powerful buyer is "generally free to bargain aggressively when negotiating the prices it will pay," since that tends to promote vigorous competition. (105) To be sure, if a buyer-induced concession actually reduced competition, courts might recognize an exception. But it would be difficult to persuade a court that price discrimination is truly anticompetitive when it believes the practice is generally beneficial. Moreover, even harmful discrimination would be outside the bounds of section (1) when either of the next two considerations apply.

According to decisions in the Ninth and Tenth Circuits, section (1) does not bar discriminatory price cuts offered to meet competition. In Zoslaw, a Ninth Circuit panel stated that "the Sherman Act is intended to encourage ... competition between sellers," and thus, if sellers are competing for the business of a large buyer, section 1 does not apply. (106) In AAA Liquors, the Tenth Circuit noted that Congress created defenses for meeting competition, cost justification, and changing conditions in the Robinson-Patman Act. This shows, the court declared, that "Congress considered price discrimination to be reasonable in at least these circumstances." (107) If this means that a discriminatory price cannot be an unreasonable restraint if it was granted to meet competition, it would decimate Sherman Act enforcement, since large buyers commonly induce preferential benefits by forcing suppliers to compete for the buyer's business.

Finally, the Sherman Act, like the other major antitrust laws, is directed at practices that increase a firm's market power, allowing it to acquire power it did not have or maintain power it would otherwise lose. (108) But several forms of anticompetitive buyer-driven discrimination described above would not enlarge the buyer's market power. They would reduce the vigor of upstream competition but would not give the buyer more power than it otherwise had. For example, if a buyer demanded concessions that reduced the profits of its suppliers and curtailed upstream innovation or variety, there would be harm--to upstream competition and consumers--but it would flow from the exercise of the buyer's countervailing power, not from an increase in this power. (109) As a result, section 1 would not apply.

Section 2 would be a more useful tool in one--but only one--circumstance: where a powerful buyer obtains such substantial concessions from its suppliers that it can drive out most smaller rivals and acquire, or threaten to acquire, monopoly power. In that case, the buyer's conduct would constitute monopolization or an attempt to monopolize. But that circumstance is narrow. Courts rarely have found monopoly power when the defendant's share of the relevant market was less than 70 percent, (110) and have apparently never found it when the defendant's share was below 50 percent." (111) To establish an attempt to monopolize, moreover, a plaintiff would have to show that the defendant's behavior created a dangerous probability of achieving such a large share. (112) There are many situations, however, in which buyer-driven discrimination would reduce competition without creating either monopoly power or a dangerous probability of such power. In fact, virtually every one of the scenarios identified in Part II is possible without a high degree of single-firm power. (113) All of them require a substantial degree of countervailing power, but a buyer can exercise substantial countervailing power with a share significantly below 70 percent. (114)

Moreover, section 2 would rarely apply, even in the scenario in which it otherwise fits, if courts read in a defense for meeting competition. (115) To be sure, in a section 2 case, the defendant would not be a supplier meeting competition, but the powerful buyer. Under the Robinson-Patman Act, however, a buyer is not liable for inducing a discriminatory price unless the discriminating supplier is also liable, (116) and the discriminating supplier would not be liable if it had a meeting competition defense. As a result, there is no buyer liability under the Robinson-Patman Act where the buyer plays suppliers off against each other, since each supplier would have a meeting competition defense. (117) If courts were to apply this logic to a monopolization case, most section 2 challenges to price discrimination would fail.

B. The Clayton Act

The only provision of the Clayton Act that addresses price discrimination is section 2, the section amended by the Robinson-Patman Act. If Congress repeals rather than reforms the Act, it would remove section 2, leaving nothing in the Clayton Act to address price discrimination. (118)

C The FTC Act

Like section 1 of the Sherman Act, section 5 of the FTC Act is broad enough to reach many categories of buyer-induced price discrimination. If a large buyer extracts a discriminatory concession from its suppliers and uses the resulting competitive advantage to destroy smaller competitors and harm consumers, that behavior would constitute an "unfair method of competition." It would reduce competition downstream through methods that are unfair to both competitors and consumers. Thus, over forty years ago, the FTC successfully established that buyers who induce discriminatory promotional benefits engage in unfair methods of competition. (119)

But like section 1, section 5 is unlikely to serve as a comprehensive or reliable safeguard against anticompetitive buyer-driven discrimination. In the first place, the FTC may refuse to pursue any concessions given to meet competition, which would remove most preferential pricing from challenge. Second, the Commission may decline to apply section 5 unless the discrimination at issue enabled the buyer to increase or preserve its market power, rather than simply exploit that power. This position, common in antitrust law, would exempt several categories of buyer-driven discrimination from attack. (120) Indeed, it is difficult to call the pure exploitation of buyer power an "unfair method of competition." (121) Third, the Commission might be loath to pursue price discrimination on the ground that it is generally procompetitive. This attitude has increasingly typified the agency's approach to Robinson-Patman enforcement, which has now withered to the point of non-existence. Its last Robinson-Patman case was filed fifteen years ago, and it was targeted at exclusionary behavior by a seller, not discrimination induced by a powerful buyer. (122) The last time the Commission attacked buyer-driven price discrimination was in (1988), when it sued six publishers for favoring Barnes & Noble and Borders over independent bookstores. It eventually dismissed all six cases, however, without obtaining relief. (123) To be sure, the FTC might be more aggressive if it thought that a particular instance of discrimination was actually anticompetitive, but its track record is not encouraging. Finally, private parties could not take up the slack, since they cannot enforce the FTC Act.

In short, if the Robinson-Patman Act were repealed, it would be risky to rely on the other antitrust laws to halt anticompetitive buyer-driven discrimination. While some cases might proceed, many would not because of concerns about (1) discouraging procompetitive discrimination; (2) inhibiting meeting competition; or (3) penalizing firms for exploiting--rather than enlarging--their market power. Thus, if Congress wants to stop protectionist Robinson-Patman enforcement, but preserve a bulwark against harmful buyer-driven discrimination, it should reform, not repeal, the statute.

V. Reforming the Robinson-Patman Act

Congress should make three principal changes to the Clayton Act: the competitive injury language added by the Robinson-Patman Act should be removed, the meeting competition defense should be eliminated in certain actions, and the cost justification defense should be similarly restricted. (124)

A. Competitive Injury

Section 2 of the Clayton Act has three effects tests. Price discrimination may be illegal if the "effect of such discrimination may be" (1) "substantially to lessen competition"; or (2) to "tend to create a monopoly"; or (3) "to injure, destroy, or prevent competition with any person who either grants or knowingly receives the benefit of such discrimination, or with customers of either of them." (125) The third test was added by the Robinson-Patman Act, and it is this test that allows section 2 to be enforced in a protectionist manner. Unlike the first two tests, which require an impact on market wide competition, the third is satisfied simply by showing that the challenged discrimination made it harder for the disfavored customer to compete with the favored customer. As Judge Breyer observed in Monahan's Marine, the "single word 'with' makes a considerable practical difference. It means that the Act protects those who compete with a favored [customer], not just the overall competitive process." (126) Eliminating the third test would harmonize the Robinson-Patman Act with the other antitrust laws by making harm to competition, rather than injury to a competitor, the ultimate test of liability. It would mean, in the ordinary Robinson-Patman case, that a violation could be established only by showing that the challenged discrimination threatened to harm consumers. (127)

Professor Hovenkamp proposes that "the secondary-line provisions of the statute should be amended to bring them back to the problem that was the focus of Congress's original concern: the powerful buyer who induces a lower price that is contrary to the interest of the seller as well as its other buyers." (128) In particular, "a reasonable accommodation is to protect the manufacturer who uses rebates or other price concessions as a device to reward more effective dealers, but not to protect powerful dealers who use their power as large buyers to force suppliers to sell to them at a lower price than rival dealers must pay." (129) This proposal would make sense as a "reasonable accommodation"--a compromise that would reduce, but not eliminate, the Act's protectionist reach. (130) But if Congress wants to change the fundamental objective of section 2, it should remove the third effects test--the "competitive injury" provision--altogether. (131)

B. Meeting Competition Defense

The rationale of the meeting competition defense is understandable. When a large buyer tells a supplier that it will lose the buyer's business unless it meets the price a competing supplier has offered, the supplier often has no choice but to comply. Assuming the buyer's threat is credible, the supplier's options are clear: if it meets the competitor's price, its margin will fall on its sales to that customer; but if it refuses, it will lose all the profit it made on those sales. For this reason, the Robinson-Patman Act protects a supplier that reasonably fears such an outcome. If the supplier has a good faith belief that a competitor is in fact offering a lower price, the supplier can meet that price even if the result is discrimination in favor of the large buyer. (132)

The meeting competition defense is another example of the Act's protectionism. In this instance, the protected party is a supplier, not a small competitor of a large buyer, but it is protectionist nonetheless. It insulates discrimination from Robinson-Patman attack even if the discrimination reduces market wide competition and consumer welfare. When a discriminatory concession is granted to meet competition, neither the supplier that grants the concession nor the buyer that induces it can be held liable, even if it triggers one or more of the anticompetitive consequences described in Part III. (133)

To control harmful buyer power more effectively, therefore, the meeting competition defense needs to be cut back. It may not make sense to eliminate it entirely. There are legitimate reasons to protect a supplier from treble damages and attorneys' fees when a large buyer demands that it meet a competitor's offer. If the demand is credible, the supplier faces a significant loss of business if it does not match the competitor's offer. And it cannot realistically predict, during a time-sensitive negotiation, whether the concession is more likely to reduce rather than promote competition. It would be unfair, therefore, at least in most cases, to penalize a supplier with treble damages and attorneys' fees if it decides it has to meet competition and the concession turns out to reduce competition.

To avoid this unfairness, the meeting competition defense could be eliminated only in purely equitable actions. If a supplier establishes the defense, it would not be exposed to treble damages or attorneys' fees. The problem with an equitable remedy, however, is that it sacrifices the deterrent benefits of treble damages and attorneys' fees. Suppliers will have less fear of caving into buyer pressure and meeting a competitor's offer if they know they will not be punished for doing so. They will simply be ordered not to do it again. (134)

The optimal approach, then, requires balancing fairness and deterrence. One possible resolution is to expose suppliers to attorneys' fees but not damages if they meet a competing offer and the resulting discrimination poses a significant threat to competition. Another approach is to allow plaintiffs to recover both treble damages and attorneys' fees from suppliers that meet competition, provided the plaintiff shows by clear and convincing evidence that the resulting discrimination was likely to harm consumers or small suppliers. But whatever approach is chosen--eliminating the defense in purely equitable actions, eliminating it in equitable actions in which the plaintiff can recover attorneys' fees, or eliminating it in treble damage actions as well if the plaintiff makes an especially strong showing of probable harm to competition--it is essential to curtail the meeting competition defense. Otherwise, the Robinson-Patman Act will remain a largely ineffective instrument for combatting buyer-induced discrimination.

C. Cost Justification Defense

Whatever option is chosen for the meeting competition defense, the cost justification defense should be treated similarly. To be sure, the cost justification defense is less of a problem since it is frequently rejected. (135) Yet, as Professor Hovenkamp has documented, it "has been accepted numerous times." (136) And when it is accepted, it could insulate anticompetitive buyer-driven discrimination from attack. Although the dangers of upstream injury are reduced when the discrimination is cost justified (since suppliers may not lose profits when they grant such a concession), the risks of downstream injury remain. Even if a discriminatory price cut is cost justified, a powerful buyer can use the competitive advantage to drive out or weaken its smaller rivals, enabling it to raise prices downstream or deprive consumers of choices they value. When a plaintiff can establish actual or probable harm to consumer welfare, a court ought to stop the discrimination--and perhaps impose treble damages--even if the discrimination was cost justified. (137)

The changes discussed in this Part would reorient the Robinson-Patman Act, shifting its overarching aim from protecting small business to protecting competition. They would also enable the Act to control powerful buyers more effectively, eliminating (or at least substantially reducing) their ability to harm consumers or small suppliers through discriminatory concessions. There are other ways of controlling powerful buyers, however, such as breaking them up (to destroy their power) or subjecting them to common carrier regulation (to prevent them from exercising it). Part VI briefly compares the three methods.

VI. Other Methods of Controlling Buyer Power

Congress could jettison the Robinson-Patman Act altogether and attempt to control buyer power through structural relief or common carrier regulation. Structural relief would be limited, however, by the reach of section 2 and the need to demonstrate that it would not cause a substantial loss in scale economies. Common carrier regulation would require a new regulatory agency and necessitate the kind of detailed governmental oversight that is widely disfavored. As a general solution, therefore, neither method is likely to be superior to reforming the Robinson-Patman Act.

A. Structural Relief

If a big buyer like Wal-Mart or Amazon were broken into smaller pieces, the resulting firms--the "baby buyers"--would have significantly less ability to extract concessions from suppliers. Each baby buyer would purchase in smaller quantities, the suppliers could now play the baby buyers off against each other, and each baby buyer would be less able to integrate backward into the supplier's business. Each new buyer's negotiating leverage, in short, would be materially reduced.

The magnitude of the benefit, however, depends on the extent of structural relief. Breaking a firm into two pieces will have less impact on its buyer power than breaking it into four pieces. But breaking it into four pieces may raise its operating costs significantly. Consider a large retailer like Wal-Mart or Amazon. These firms enjoy economies of scale in multiple dimensions of their business, as Judge Hogan explained in Staples, (138) One of the most important is distribution. A large retailer typically employs a set of warehouses (or "distribution centers") located strategically across the country to minimize its shipping costs and maximize its shipping speed. If those distribution centers were divided among four successor firms, none of the firms could serve the national market as efficiently as the original chain. As a result, shipping costs and shipping times would probably rise, forcing consumers to pay more and wait longer. (139) The principal drawback of structural relief, in short, is that it may cause a significant reduction in operating efficiency. (140)

In addition, structural relief may not be effective in achieving its ultimate aim: eliminating the defendant's monopoly power and restoring competition. In a well-known study, Robert Crandall found that structural remedies typically fail to enhance competition in the relevant market. (141) In a more recent review, Professor Carstensen came to the opposite conclusion. He determined that when firms were broken up, whether through a litigated order or a consent decree, the results were generally beneficial. (142) But whether structural relief is typically beneficial or usually ineffective, this dispute underlines the importance of a careful inquiry into its impact in each particular case. Structural relief should not be ordered unless it would produce successor firms that are (1) likely to be viable, (2) likely to compete with each other, and (3) unlikely to have significantly higher costs.

Given these requirements, it is not surprising that courts are hesitant to order structural relief. Indeed, in the last thirty years, they have never imposed it. (143) Professor Carstensen thinks the primary reason is the shrinking scope of section 2 of the Sherman Act. As the courts have tightened the standards for section 2 liability, making it more difficult to show illegal exclusionary conduct, they have shifted the decisive issue from market structure to conduct, making it much more probable, if they do find a violation, that the remedy will be behavioral rather than structural. (144) Finally, even if a court were willing to impose a structural rather than a behavioral remedy, it normally cannot do so absent a violation of section 2. Yet, as explained earlier, most forms of anticompetitive buyer-induced discrimination do not violate section 2 of the Sherman Act.

In short, while structural relief might be optimal in some cases, it is unlikely to constitute the ordinary remedy for buyer-induced price discrimination.

B. Common Carrier Regulation

To achieve a broader remedy, Congress could subject powerful buyers to common carrier regulation. To do this Congress would have to establish a new regulatory agency, charge it with overseeing powerful buyers like Amazon and Wal-Mart, and empower it to impose on them the duties of a common carrier, including the obligation to carry the products of its suppliers. So obligated, no powerful buyer could refuse to carry a product because the supplier had not make a concession the buyer demanded. This would substantially limit the buyer's power, since a major source of a buyer's leverage (as Amazon has repeatedly demonstrated) is the willingness to cut off a supplier that fails to bend to its demands. (145)

Common carrier regulation, however, is likely to entail extensive oversight of the buyer's prices, terms, and services. A big retailer like Amazon or Wal-Mart cannot exist unless it earns a margin sufficient to cover its costs. In order to survive, it must be allowed to purchase products at a price low enough to enable it to resell those products in competition with other retailers and still earn a margin sufficient to cover all its legitimate costs, including its costs of capital. It must, in short, be allowed to make a competitive rate of return. On the other hand, the firm will exercise its buyer power to the maximum extent it can, paying its suppliers less and raising its margins to supracompetitive levels. Unless its power is limited, therefore, the buyer is likely to earn excess returns. The basic task of the new agency would be to moderate this conflict, reducing or eliminating the exercise of buyer power while allowing the firm to make a competitive profit.

The agency could pursue this goal by regulating the prices the buyer pays to each of its suppliers. Given the number of products carried by retailers like Amazon and Wal-Mart, however, this would be an immense task. The more manageable approach would be to regulate margins directly. (146) But that would not be easy. As noted, the buyer's margin must be sufficient to cover its legitimate costs and provide a competitive return (given the risks incurred), and the agency must determine both. Because costs and risks may differ by product category, the agency may have to repeat this analysis across multiple product classes. Moreover, because costs and risks are likely to change over time, continuing oversight would be required.

But this would not be all. In addition to margins, the agency would have to regulate the services the buyer provides to its suppliers and the prices it charges for those services. Recall that Amazon attempted to extract concessions from Hachette by selectively depriving it of services like rapid shipping and preordering, services it provided to other publishers. To forestall that tactic, a powerful buyer would have to offer all its services to all its suppliers. To prevent the buyer from charging some suppliers a higher price than it charges others, or charging all suppliers a supracompetitive price, the regulatory agency would have to regulate the price for each service. Otherwise, the firm could exercise its buyer power through the prices it levies for its services.

One important type of service--so important that it is addressed in section 2(d) of the Robinson-Patman Act--is promotional service. Buyers commonly furnish a variety of promotional services to their suppliers, such as placing a supplier's product where it is likely to be noticed, selling the product at a special discount, or featuring the product in the buyer's own advertising. In order to prevent a buyer from selling such services to some or all suppliers at a premium price, the agency would have to regulate the prices the buyer charges. Finally, the agency might have to police the quantity and quality of each service, or the buyer might obtain a supracompetitive margin by surreptitiously reducing the volume or caliber of the services it furnishes.

In brief, common carrier regulation would require oversight of the margins earned, prices charged, and services provided by powerful buyers. Such comprehensive regulation would be difficult to implement, particular in the new agency's early years, when it would lack experience and a repository of data. Over time, its skills would improve, but this may not be enough to avoid significant adverse effects on efficiency and growth. (147)

C. Reforming the Robinson-Patman Act

Reforming the Robinson-Patman Act is likely be the more cost-effective approach. A court would not have to determine the feasibility and utility of a structural remedy. Nor would a new agency have to regulate the margins, services, and terms of the country's most powerful buyers. Instead, upon proof that discrimination in favor a large buyer had led to a reduction in upstream or downstream competition, the court would enjoin that discrimination. Such an injunction--prohibiting the pertinent suppliers from offering prices to the powerful buyer that are more favorable than the prices they offer to competing buyers--would be easy to understand and write. (148)

This simple device, moreover, would substantially limit the buyer's power. Under the injunction, a supplier could not offer the buyer a lower price than it offers competing buyers. It would have to treat all rivals the same. Further, if the buyer wanted a concession from the supplier, it would have to persuade the supplier to offer the concession to all competing buyers, greatly increasing the costs of the concession. As a result of both constraints, an injunction would sharply reduce, if not remove, the buyer's ability to exercise its power in anticompetitive ways. It could not gain an advantage over competing buyers and it would be much less able to obtain a concession from suppliers. In short, a simple injunction against discrimination would substantially diminish a supplier's ability to give--and therefore a powerful buyer's ability to extract--preferential prices that are likely to harm competition. (149)

A simple injunction, however, may not always be the ideal remedy. It would prohibit the buyer from obtaining any concessions, even those that would increase competition and benefit consumers. Suppose, for example, that Amazon had induced book publishers to give it a preferential discount of 10%, allowing it to buy best sellers and other titles for 10% less than other retailers paid. But suppose that only 6 percentage points of that discriminatory discount were anticompetitive. The other 4 points, though discriminatory, were procompetitive because (1) they reduced excess returns among publishers or reflected cost savings generated by Amazon, (2) Amazon was likely to pass them on to consumers in the form of lower prices, and (3) those lower prices would not drive out so many competing retailers that Amazon would gain downstream market power and raise prices. In this situation, 4 percentage points of the discriminatory discount would benefit rather than harm consumers. Thus, the ideal injunction would prohibit only 6 percentage points, allowing publishers to charge Amazon 4 percent less than they charge other retailers.

But if courts were to parse a discriminatory discount in this way, determining what portion of it was beneficial to consumers and what portion was anticompetitive, it would raise the costs of Robinson-Patman Act enforcement considerably. Courts would have to decide not only whether the challenged discrimination was anticompetitive on balance--a difficult enough task--but also whether the order should ban all of the discrimination or just part of it. This would greatly complicate the remedial portion of the case and, because market structure, costs, and other relevant factors are likely to change over time, require either continuing oversight by the court or a time-limited injunction. As a result, unless a defendant provided strong evidence that a portion of the discriminatory was procompetitive and ought to continue, a court would be justified in refusing to engage in this inquiry.

Even if judges undertook it in some cases, though, it would be less expensive than common carrier regulation, which would entail the creation of a new regulatory agency with the authority and responsibility to control the margins received and the services and terms offered by every powerful buyer in the country. In contrast, Robinson-Patman action would be triggered only by proof that a particular buyer had obtained discriminatory prices that were likely to reduce, or had reduced, competition. Such case-by-case enforcement is likely to impose fewer costs on the government and the economy than a new regulatory regime. (150)

In some instances, of course, structural relief may be the optimal remedy. It may be possible to formulate a breakup plan that reduces buyer power substantially without raising costs significantly or requiring continuing judicial oversight. Where those conditions hold, a structural remedy may be preferable to an injunction under a reformed Robinson-Patman Act. But those conditions may not often hold. Instead, structural relief may not be feasible, may not increase competition in the relevant market, or may harm consumers by depressing productive efficiency. In most cases, a Robinson-Patman remedy is likely to be superior.

VII. Conclusion

Critics have condemned the Robinson-Patman Act for decades because its goal is to protect small business rather than consumers. At the same time, authors, publishers, and commentators have expressed alarm at the growing power of Amazon, worried that its aggressive demands could curtail the development of new titles, eliminate independent bookstores, and reduce consumer choice. Yet they have wondered whether there is any effective remedy for this behavior, under the antitrust laws or otherwise. This article suggests that both problems could be addressed by reforming the Robinson-Patman Act.

First, the Act's goal should be altered. Plaintiffs should be allowed to enforce the Act only when they can show a threat to market wide competition, not just individual competitors. Second, the Act's meeting competition and cost justification defenses should be curtailed, since those defenses often block enforcement action against powerful buyers, even when their behavior endangers consumers. Finally, though not the focus of this article, conforming changes should be made in the Act's technical and jurisdictional requirements and its treatment of promotional discrimination.

These revisions would reorient the statute, making the test of liability impact on competition and consumers, not small business. They would rehabilitate the Robinson-Patman Act and increase the likelihood that the government would bring Robinson-Patman actions and the courts would uphold them. More important, they would create an effective tool for controlling powerful buyers like Amazon and Wal-Mart. They would enable a court to enjoin discriminatory concessions induced by such buyers, even if granted to meet competition or reflect cost savings, if those concessions were likely to harm competition and consumers. The Robinson-Patman Act should be reformed, not repealed.

DOI: 10.1177/0003603x15602393


This article has benefitted from comments by Warren Grimes, Mark Glick, Einer Elhauge, Peter Carstensen, Geoffrey Kirkwood, and two anonymous referees.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.


The author(s) received no financial support for the research, authorship, and/or publication of this article.

(1.) See infra note 30 and accompanying text.

(2.) The Robinson-Patman Act's protectionist thrust is now confined to "secondary line" cases--cases in which the plaintiff is a customer of the discriminating seller. Two decades ago the Supreme Court decided that in "primary line" cases--cases in which the plaintiff is a competitor of the discriminating seller--the plaintiff must demonstrate a threat to market-wide competition and consumers. See infra Part II.

(3.) For example, the Act only applies to discrimination in the sale of commodities; it does not prohibit discrimination in the furnishing of services. Moreover, it ordinarily requires "two sales" involving "competing purchasers" and one of those sales must be "in commerce." For a full discussion of the Act's technical and jurisdictional requirements, see 14 Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application ch. 23B (3d ed. 2012).

(4.) See Jeffrey A. Trachtenberg, Amazon Defends Stance in Dispute, Wall St. J., July 2, 2014, at B1 ("Amazon's overall share of new books sold increased to 40% from 12% over the last five years, measured by units, according to the Codex Group LLC, a book audience research firm. Its share of the e-book market grew to 64% from 58%, Codex said. 'They're the most powerful book retailer today by far,' said Peter Hildick-Smith, chief executive of Codex.").

(5.) For example, by exacting better terms from publishers than its smaller rivals receive, Amazon could weaken or destroy them, and that reduction in choice, if it becomes severe enough, could outweigh the lower prices Amazon offers. Moreover, by extracting greater concessions from publishers, Amazon could reduce their profits--and the earnings of authors--to such a degree that the output of new titles is diminished. For a fuller discussion of the anticompetitive consequences of buyer power and a brief review of the evidence, see infra Part III.

(6.) The other antitrust laws--the Sherman Act, the Clayton Act, and the FTC Act--also have a second fundamental goal. In cases challenging the unjustified acquisition or maintenance of monopsony power, the aim is to protect small, powerless suppliers from exploitation. See infra note 19 and accompanying text.

(7.) If section 2 of the Clayton Act (the provision containing the Robinson-Patman Act) were repealed, the Clayton Act would not address discrimination.

(8.) Only the Federal Trade Commission can enforce the FTC Act and, for more than a generation, the Commission has shown no interest in challenging buyer-induced discrimination--under the FTC Act or the Robinson-Patman Act. See infra Part IV.C.

(9.) To maintain the incentive to bring a legitimate lawsuit, a successful plaintiff could still recover attorney's fees.

(10.) The second option provides greater deterrence, while the first affords greater protection to a supplier that cannot realistically be expected to know, in the heat of negotiations, whether a concession is likely to reduce or promote competition.

(11.) Other changes in the Act should also be made: adjustments to the Act's technical and jurisdictional requirements and conforming changes to the Act's treatment of promotional discrimination. These changes, however, are beyond the scope of this article.

(12.) See infra notes 33-37 and accompanying text.

(13.) See Trachtenberg, supra note 4.

(14.) See, e.g., Franklin Foer, Amazon Must Be Stopped, New Republic, Oct. 10, 2014, 119769/amazons-monopoly-must-be-broken-radical-plan-tech-giant; Paul Krugman, Amazon's Monopsony is Not O.K., N.Y. Times, Oct. 20, 2014, A25 (", the giant online retailer, has too much power, and it uses that power in ways that hurt America.").

(15.) See infra Part 111.A.

(16.) See supra note 2.

(17.) There are some differences in the two forms of discrimination. For example, in the case of price discrimination, a lower price given to a large customer may be justified because it is cheaper for the seller to serve that customer. In the case of promotional discrimination, the most important form of justification is not lower costs but higher value. Certain promotional activities (e.g., television advertising) are likely to have higher value (i.e., generate more incremental sales) than other activities (e.g., in-store displays). When that is true, a seller would have a legitimate reason to reimburse the higher value activities more generously than lower value activities. But if the resulting discrimination leads to a reduction in consumer welfare, then the value justification, like the cost justification, should be curtailed.

(18.) For a description of these circumstances, see John L. Peterman, The Morton and International Salt Cases: Discounts on Sales of Table Salt, 21 Res. L. & Econ. 127 (2004).

(19.) The legislative history of the Sherman Act is clear that Congress intended to protect suppliers without market power--small, competitive suppliers--from monopsony power that was acquired or maintained without justification. See Gregory J. Werden, Monopsony and the Sherman Act: Consumer Welfare in a New Light, 74 Antitrust L. J. 707 (2007); John B. Kirkwood & Robert H. Lande, The Fundamental Goal of Antitrust: Protecting Consumers, Not Increasing Efficiency, 84 Notre Dame L. Rev. 191, 207-9, 233-36 (2008); John B. Kirkwood, The Essence of Antitrust: Protecting Consumers and Small Suppliers from Anticompetitive Conduct, 81 Fordham L. Rev. 2425, 2434-35, 2442 (2013). (20.) See John B. Kirkwood, Collusion to Control a Powerful Customer: Amazon, E-Books, and Antitrust Policy, 69 U. Miami L. Rev. I, 102, 162 (2014).

(21.) Countervailing power is the type of buyer power that is exercised against suppliers with market power. The other type, monopsony power, the classic form of buyer power, is exercised against powerless suppliers, typically small, competitive sellers. For a discussion of these two types of buyer power, describing their nature, prerequisites, and effects, see John B. Kirkwood, Powerful Buyers and Merger Enforcement, 92 B.U. L. Rev. 1485, 1493-1512 (2012).

(22.) As 1 noted earlier, in cases challenging the exercise of monopsony power, the fundamental goal is symmetric: preserving competition in order to protect small, powerless suppliers. See supra note 19.

(23.) See Kirkwood & Lande, supra note 19; Kirkwood, supra note 19. Most commentators agree. See. e.g., Einer Elhauge, Tying, Bundled Discounts, and the Death of the Single Monopoly Profit Theory, 123 Harv. L. Rev. 397,435 (2009) ("The Antitrust Standard is Consumer Welfare, Not Total Welfare"); Steven C. Salop, Question: What Is the Rea! and Proper Antitrust Welfare Standard? Answer: The True Consumer Welfare Standard, 22 Loy. Consumer L. Rev. 336 (2010); Herbert Hovenkamp, Implementing Antitrust's Welfare Goats, 81 Fordham L. Rev. 2471, 2477 (2013) ("In sum, antitrust policy in the United States follows a consumer welfare approach in that it condemns restraints that actually result in monopoly output reductions, whether or not there are offsetting efficiencies and regardless of their size."). Some commentators assert that the preeminent purpose of antitrust law is protecting "competition" or "the competitive process." If that means preserving market rivalry when it benefits consumers in sell-side cases and competitive suppliers in buy-side cases, it is unobjectionable. Indeed, as Professor Grimes has pointed out to me, both terms enable one to state the ultimate goal quite compactly, without having to refer to consumers in sell-side cases and suppliers in buy-side cases. Neither term is self-defining, however, and if they are interpreted to mean preserving all the rivals in a market, or existing rivalry between every one of them, regardless of the impact on consumers or small suppliers, they would be highly problematic. For a fuller discussion, see Kirkwood, supra note 20, at 131- 32.

(24.) 509 U.S. 209 (1993).

(25.) Id. at 221-22 ("the essence of the claim under either statute is the same").

(26.) Specifically, a plaintiff must prove that (I) the seller's prices were "below an appropriate measure of [its] ... costs," id.; and (2) the seller "had a reasonable prospect [under the Robinson-Patman Act] or, under [section] 2 of the Sherman Act, a dangerous probability, of recouping its investment in below-cost prices." Id. at 224.

(27.) Id. at 220 (quoting Great Atl. & Pac. Tea Co. v. F.T.C., 440 U.S. 69, 80 n. 13 (1979)).

(28.) Terry Calvani & Gilde Breidenbach, An Introduction to the Robinson-Patman Act and Its Enforcement by the Government, 59 Antitrust L.J. 765, 770 (1991); see also Hovenkamp, supra note 3, at 12 ("Clearly, the class targeted for protection was not consumers, who benefitted from the chains' success; rather, the class comprised the various small businesses and intermediaries who lost market share, profits, or in some case their entire businesses as a result of more efficient distribution methods."); Mark A. Glick, David G. Mangum, & Lara A. Swensen, Towards a More Reasoned Application of the Robinson-Patman Act: A Holistic View Incorporating Principles of Law and Economics in Light of Congressional Intent, 60 Antitrust Bull. 279-317 (2015) ("The explicit goal of the RPA is to protect small businesses").

(29.) 334 U.S. 43 (1948).

(30.) Id. at 43. The Court also stated that the Act's competitive injury language was "intended to justify a finding of injury to competition by a showing of 'injury to the competitor victimized by the discrimination.'" Id. at 49. See also Glick, Mangum, & Swensen, supra note 28 ("The primary congressional goal in enacting the Robinson-Patman Act was to protect small businesses from being underpriced by larger competitors who possessed buyer power that allowed them to extract concessions from manufacturers."). As Glick, Mangum, & Swensen note, the Act's protectionist thrust is limited by its cost justification defense. A supplier may discriminate between a large buyer and its smaller rival so long as the difference reflects the savings the supplier realizes in serving the large buyer. The Act's goal, then, is not to protect small firms from all price differentials but to protect them from non-cost-justified differentials. See id. ("the cost justification defense ... was designed to filter out cases where differential pricing was a result of efficiencies rather than buyer power").

(31.) See John B, Kirkwood, Buyer Power and Exclusionary Conduct: Should Brooke Group Set the Standards for Buyer-Induced Price Discrimination and Predatory Bidding?, 72 Antitrust L. J. 625, 633-34 (2005).

(32.) Volvo Trucks N. Am., Inc. v. Reeder-Simco GMC Inc., 546 U.S. 164, 177 (2006). Some statements in Volvo suggest that the Court intended to end the protectionist thrust of secondary line enforcement. The Court repeated its declaration from Brooke Group that Robinson-Patman Act proscribes "price discrimination only to the extent that it threatens to injure competition," id. at 176 (quoting 509 U.S. at 220). It also indicated it "would resist interpretation [of the Act] geared more to the protection of existing competitors than to the stimulation of competition." Id. at 181 (emphases in original). But these statements do not mean that the Court has repudiated the Act's traditional goal. On the contrary, the Court expressly endorsed the Morton Salt inference and stated that a "hallmark" of the competitive injury a plaintiff must show in a secondary line case is "the diversion of sales or profits from a disfavored purchaser to a favored purchaser." Id. at 177. The Court would have said neither of these things if it had wanted to impose a new obligation on a secondary line plaintiff--an obligation to show an adverse impact on competition and consumers. Likewise, the Court never stated that a secondary line plaintiff had to establish market power, entry barriers, higher prices, or any other indicia of consumer harm. What seems to explain the Court's divergent statements is that the Court was distinguishing traditional Robinson-Patman cases from attempts by a plaintiff to make new law. In a traditional case, a plaintiff may still prevail by showing competitive injury in the customary way. But in a novel case, a case like Volvo, in which the plaintiff can win only if a new, more expansive interpretation of the Act is adopted, the Court signaled that the new interpretation should not be accepted unless it would promote competition. See John B. Kirkwood, The Robinson-Patman Act and Consumer Welfare: Has Volvo Reconciled Them?, 30 Seattle U. L. Rev. 349 (2007)

(33.) Long-term harm to consumers can occur in various settings, as Part III explains.

(34.) See Kirkwood, supra note 31, at 646. This need not happen; a powerful buyer's demands may lead to a persistent two-tiered pricing structure, in which the oligopolists maintain supracompetitive prices to the buyer's smaller rivals.

(35.) See, e.g., Hovenkamp, supra note 3, at 150-51;
   [A] a rule that penalizes firms for beating rather than meeting the
   bids of other firms, combined with the basic RobinsonPatman Act
   rule forbidding firms from bidding lower than the price it has
   charged in reasonably contemporaneous sales to others, serves to
   blunt the hard competitive bidding that antitrust policy generally
   wishes to encourage.

See also id. at 151 ("[W]here collusion or oligopoly already exists, the Robinson-Patman Act often serves to preserve rather than undermine it.").

(36.) See id. at 146 ("Very few statutes have survived such long-lived and unrelenting criticism as has been directed against the Robinson-Patman Act.").

(37.) See U.S. Dept, of Justice, Report on the Robinson-Patman Act (1977); Antitrust Modernization Commission, Report and Recommendations (2007); Hovenkamp, supra note 3, at 13 ("there have been many calls for its modification or repeal.").

(38.) See Kirkwood, supra note 21, at 1536-58; Kirkwood, supra note 31, at 647-51.

(39.) As the introduction noted, this article focuses on the exercise of countervailing power. The exercise of monopsony power can also result in discrimination and suppliers can be harmed as a result, but the incidents appear to be less frequent.

(40.) For a detailed description of the conspiracy, and an analysis of its possible justifications, see Kirkwood, supra note 20.

(41.) See United States v. Apple, Inc., No. 12-CV-2826 (S.D.N.Y. Sept. 6, 2012) (accepting the first three proposed settlements, with Hachette, HarperCollins, and Simon &Schuster); James B. Stewart, Booksellers Score Some Points in Amazon's Spat with Hachette, N.Y. Times, June 21, 2014, at B1 ("The five publishers--Hachette, HarperCollins, Macmillan, Penguin and Simon & Schuster (all the so-called majors except Random House)--settled the charges with the department and agreed not to collude.").

(42.) See United States v. Apple, Inc., 952 F. Supp. 2d 638 (S.D.N.Y. 2013).

(43.) The key to the consumer harm was the creation of downstream market power. When suppliers collude in response to buyer power and that collusion creates downstream market power, then suppliers acquire both the ability and the incentive to exploit consumers. See Kirkwood, supra note 20, at 160-63.

(44.) David Streitfeld, Investments by Amazon Are Piling Up, as Big Losses, N.Y. Times, Oct. 24, 2014, B1, http://

(45.) See Kirkwood, supra note 20, at 146.

(46.) See Krugman, supra note 14 ("Amazon began delaying their delivery, raising their prices, and/or steering customers to other publishers."). In addition, Amazon removed preorder buttons on Hachette books. See Jeffrey A. Trachtenberg & Greg Bensinger, Amazon, Hachette Reach a Truce, Wall St. J., Nov. 14, 2014, at B1, (noting that "Preorders are often key to catapulting books onto the national bestseller list.").

(47.) See id. at B2 (quoting Simon Lipskar, president of the literary agency Writers House: "Our writers have been suffering terribly because their sales have been significantly diminished as a result of this dispute.").

(48.) David Streitfeld, Literary Lions Unite in Protest over Amazon's E-Book Tactics, N.Y. Times, Sept. 29, 2014, at B1, http://

(49.) Id. at B7; see also David Streitfeld, Writers Feel an Amazon-Hachette Spat, N.Y. Times, May 10, 2014, at B1, B7, http:// www.nytimescom/2014/05/10/technology/writers-feel-an-amazon-hachette-spat.html ("'Like all repressive regimes, Amazon wants to completely control your access to books,' Sherman Alexie said in a Twitter post.").

(50.) Foer, supra note 14.

(51.) Id.

(52.) George Packer, Cheap Words: Amazon Is Good for Customers. But Is It Good for Books? New Yorker, Feb. 17,2014, http://

(53.) See, e.g., David Streitfeld, Publishing War Is Over, but Fear of Amazon Isn't, N.Y. Times, Nov. 14, 2014, at Al, http:// (noting that "a large group of authors" continues to call for Amazon to be investigated for antitrust violations).

(54.) Id. at B6 ("The multiyear agreement [with Hachette], which includes both e-books and print books, broadly follows a deal Amazon recently worked out with Simon & Schuster.").

(55.) Id. (quoting a Forrester analyst's statement that "Hachette got Amazon to allow them to control pricing while also cutting the amount of money Amazon takes if the publisher does engage in discounts").

(56.) Id. ("Michael Pietsch, Hachette's chief executive, said the percentage of revenue on which Hachette authors' e-book royalties are based 'will not decrease under this agreement.'").

(57.) Id. ("Amazon got increased co-op funds").

(58.) See Jay Greene, Hachette, Amazon Finally Write Ending to Book Feud, Seattle Times, Nov. 14, 2014, at Al ("Executives from both companies [Hachette and Amazon] expressed pleasure in getting it done."); Jay Greene, Foes Want Probe, but Antitrust Law Favors Amazon, Seattle Times, Oct. 26,2014, at A1, A13 (noting that Simon & Schuster called its agreement with Amazon "economically advantageous").

(59.) See Kirkwood, supra note 20, at 149; see also Albert A. Foer & Tyler Patterson. E-Books and Amazon: The Need to Hear Two Hands Clapping, Comp. L. Insight, July 31, 2012, at 10 (while expressing deep concern with Amazon's monopsony power, acknowledging that "we do not at this point in time have evidence of monopsonistic abuse or even a dangerous probability of it").

(60.) Brad Stone, The Everything Store: Jeff Bezos and the Age of Amazon 256 (2013); see also id. at 278 ("As suppliers had learned over the past decade, no matter the category, Amazon wielded its market power neither lightly nor gracefully, employing every bit of leverage to improve its own margins and pass along savings to its customers."); see also Streitfeld, supra note 48, at B7 (in explaining its dispute with Hachette, Amazon has "said that it was trying to make e-books cheaper and thus more affordable for all"); Trachtenberg, supra note 4 (noting that Russ Grandinetti, Amazon's senior vice president of Kindle content, said that Amazon was simply doing what is "in the long-term interest of our customers.... The terms under which we trade will determine how good the prices are that we can offer consumers.").

(61.) Foer, supra note 14.

(62.) Id.

(63.) Joe Nocera, Amazon Plays Rough. So What?, N.Y. Times, Oct. 14, 2014, at A23, opinion/joe-nocera-amazon-plays-rough-so-what.html.

(64.) Matthew Yglesias, There's One Huge Problem with Calls for Anti-Trust Action Against Amazon, Vox, Oct 10, 2104, http://

(65.) See supra note 4 and accompanying text.

(66.) See Foer, supra note 14.

(67.) Krugman, supra note 14.

(68.) See Streitfeld, supra note 48, at B7; Krugman, supra note 14.

(69.) See Krugman, supra note 14.

(70.) One might dismiss this concern on the ground that Amazon would not want to limit the number or variety of new titles published, which would diminish its own sales. There are two reasons, however, why the retailer might tolerate some decline in innovation or variety. First, like many other large, publicly traded corporations, Amazon cares about its short term results. It might therefore use its buyer power to increase its short run profits, even if that results in some reduction in the number of new books in the future. Second, price is probably more salient for consumers than the extent of variety, at least where substantial variety remains. In other words, consumers are more likely to notice whether the price of a bestseller has been lowered than whether the total number or array of new books has been limited. As a result, Amazon is likely to be rewarded for a small but significant price reduction on popular new titles, even if it is accompanied by a small but significant reduction in variety. Neither reason indicates that Amazon would be willing to tolerate a substantial reduction in the number or diversity of new titles published. But both explain why Amazon might use its buyer power to extract short run price concessions, even if the result is some decline in innovation or variety.

(71.) For fuller discussions of these scenarios, see Kirkwood, supra note 21, at 1537-50. In addition, two papers in this volume emphasize the anticompetitive consequences of the fourth scenario, in which the elimination of small competitors reduces consumer choice. See Keith Leffler & Ted Tatos, Competitive Injury and Damages Under the Robinson-Patman Act: Morton Salt and Statistical Analysis, 60 Antitrust Bull. 318-344 (2015) (non-cost-justified discrimination distorts the competitive process and may drive out the most efficient or innovative firms); Glick, Magnum, & Swensen, supra note 28 ("A significant and growing body of literature shows that small businesses are a significant source of innovation ... and market efficiencies. Moreover ... in some instances, so-called 'Big Box' development has harmed local economies by ... reducing consumer access to the variety and valuable services offered by small retailers.").

(72.) See Toys "R" Us, Inc. v. F.T.C., 221 F.3d 928 (7th Cir. 2000).

(73.) See Paul W. Dobson & Roman Inderst, The Waterbed Effect: Where Buying and Selling Power Come Together, 2008 Wis. L. Rev. 331, 333 ("waterbed effect" arises when "better supply terms for powerful buyers [leads] to a worsening of the terms of supply for smaller or otherwise-less-powerful buyers").

(74.) See Bruce V. Spiva, Comments of the American Booksellers Association to the Antitrust Modernization Commission Robinson-Patman Act Panel 4-10 (2005).

(75.) See id. at 15 ("[I]n response to the chains' demands for ever larger discounts, publishers have gradually raised the average list prices of new books, particularly hardcovers, in order to maintain their own profitability.").

(76.) See id. at 3-4 (reporting that membership in the American Booksellers Association fell from a "high of 5,200 in 1991 to 1,791 members today, a 65% decline in less than fifteen years").

(77.) See id. at 14-15 (citing David D. Kirkpatrick, Quietly, Booksellers Are Putting an End to the Discount Era, N.Y. Times, Oct. 9, 2000, pagewanted=all&src=pm).

(78.) Id. at 15.

(79.) See, e.g., Robert B. Reich, Don't Blame Wal-Mart, N.Y. Times, Feb. 28,2005, at A1, opinion/28reich.html (declaring that Wal-Mart turns "main streets into ghost towns by sucking business away from small retailers."); Peter Applebome, Who Killed This Little Bookstore? There Are Enough Suspects to Go Around, N.Y. Times, Mar. 22, 2009, at A24 ("A store shutting down these days isn't exactly startling news. You drive around any suburban downtown these days, and you see the yawning, empty storefronts; ghosts of better days.").

(80.) See Store Size Caps, Inst. For Local Self-Reliance (Mar. 15, 2012), (identifying twenty-eight cities and five counties that have "enacted zoning rules that prohibit stores over a certain size"). In addition, community protests have stopped Wal-Mart from opening hundreds of stores. See Paul Ingram, Lori Qingyuan Yue, & Hayagreeva Rao, Trouble in Store: Probes, Protests, and Store Openings by Wal-Mart, 1998-2007, 116 Am. J. Soc. 53, 53 (2010) ("[T]he principal obstacle to the expansion of Wal-Mart has been protests by local activists. During the period starting from 1998 and ending in 2005, Wal-Mart floated 1,599 proposals to open new stores. Wal-Mart successfully opened 1,040 stores. Protests arose on 563 occasions, and in 65% of the cases in which protests arose, WalMart did not open a store.").

(81.) M.A. Adelman, A&P: A Study in Price-Cost Behavior and Public Policy (1959).

(82.) Id. at 242 ("As price structures gradually buckled in the depression, ... [preferences increased in absolute amount, and ... [by] 1935, ... gross discrimination [accounted for] well over a third of net profit.").

(83.) Id. at 242-43.

(84.) See supra notes 60-64 and accompanying text.

(85.) See Joe Nocera, Amazon's "Bullying" Tactics, N.Y. Times, May 31, 2014, at A19, opinion/nocera-amazons-bullying-tactics.html (Amazon's "goal is not to raise prices, which is contrary to Amazon's pro-consumer culture."). In the longer term, however, Amazon's goals may change. If at some point its growth ceases, it may decide to exploit its existing position and raise prices to the extent it can.

(86.) See supra note 76.

(87.) See David Streitfeld, Feed the Beast (or Else), N.Y. Times, Jul. 13, 2014, at BUI (Amazon grew from a "cute toy" and a "counterweight to Barnes & Noble and Borders" to the dominant bookseller, leaving "Borders ... dead" and "Barnes & Noble ... weak").

(88.) See Richard A. Posner, Can Bookstores Survive? Prospects and Consequences--Posner, Becker-Posner Blog, Jan. 9, 2011, /can-bookstores-survive-prospects-and-consequencesposner.html.

(89.) The ability to browse not only pleases consumers directly, it may also help publishers sell new titles, benefiting consumers in the longer term. See Letter from Scott Turow: Grim News, Authors Guild, Mar. 9, 2012, advocacy/letter-from-scott-turow-grim-news ("Marketing studies consistently show that readers are far more adventurous in their choice of books when in a bookstore than when shopping online. In bookstores, readers are open to trying new genres and new authors: it's by far the best way for new works to be discovered. ").

(90.) Suppose that many consumers are willing to pay a significant amount to keep their local independent bookstore open. Whether or not they shop at it regularly, they want to preserve the option of shopping at it. Thus, they are willing to buy a significant number of books at full price at the independent rather than purchase them at a discount from Barnes & Noble or Amazon in order to preserve the possibility of shopping at the independent. But unless other consumers are following the same strategy, many individual consumers will not pursue it, for no consumer can preserve an independent bookstore on her own. In order to ensure the survival of independents, consumers have to act collectively, and that is difficult to accomplish. For this reason, the market may not protect consumers from the elimination of independent bookstores, even though, if consumers could act collectively, they would preserve the independents. See Kirkwood, supra note 21, at 1548-49 (discussing this market failure and describing evidence that would help determine whether consumers were hurt on balance).

(91.) See supra note 23 and accompanying text.

(92.) Hovenkamp, supra note 3, at 8.

(93.) The mainstream antitrust laws generally reach conduct only if it enables a defendant to acquire or maintain market power. See infra note 108 and accompanying text. A concession induced by a powerful buyer may reduce a supplier's profit--and thus may inhibit innovation--but it would not add to the buyer's market power unless the buyer used the concession to weaken or destroy smaller rivals. If it does not, the discrimination is unlikely to violate section 1.

(94.) See, e.g., Callahan v. A.E.V., Inc., 182 F.3d 237, 248-49 (3d Cir. 1999refusing to dismiss the plaintiffs' Sherman Act claim that competing retailers had conspired to induce preferential prices from their suppliers); Mack Sales & Service, Inc., v. Mack Trucks, Inc., 2005 WL 724117, at *5 & *6 (E.D. Pa. 2005rejecting summary judgment for defendants on plaintiffs Sherman Act claim that a group of dealers conspired among themselves and with the manufacturer to obtain more favorable pricing).

(95.) AAA Liquors, Inc. v. Joseph E. Seagram & Sons, Inc., 705 F.2d 1203, 1207 (10th Cir. 1982). (96.) Zoslaw v. MCA Distributing Corp., 693 F.2d 870, 886 (9th Cir. 1982).

(97.) Callahan, 182 F.3d at 248.

(98.) Id. at 249.

(99.) Monahan's Marine, Inc. v. Boston Whaler, Inc., 866 F.2d 525, 529 (1st Cir. 1989stating that this is not "a case of serious, demonstrated injury to the market itself").

(100.) Id.

(101.) See, e.g., Zoslaw, 693 F.2d at 886 (indicating that vertical contracts containing discriminatory prices may violate section 1 if "they are found to be unreasonable."); Alliance Shippers v. Southern Pac. Transp. Co., 858 F.2d567,570 (9th Cir. 1988) ("vertical arrangements resulting in price discrimination ... may violate Section 1 under a rule of reason analysis").

(102.) Monahan's Marine, 866 F.2d at 527.

(103.) Id. at 527-28.

(104.) See, e.g., Zoslaw, 693 F.2d at 886 ("the Supreme Court has recognized that the price discrimination which results where buyers seek competitive advantage from sellers encourages the aims of the Sherman Act"); AAA Liquors, 705 F.2d at 1207 (a requirement to charge uniform prices would inhibit suppliers from "starting or responding to a local price war. Such a holding would discourage competition rather than encourage it, contrary to the policy of the Sherman Act.").

(105.) West Penn Allegheny Health Sys., Inc. v. UPMC, 627 F.3d 85, 103 (3d Cir. 2010).

(106.) Zoslaw, 693 F.2d at 885.

(107.) AAA Liquors, 705 F.2d at 2107 n.-5.

(108.) See Letter from Professors of Law, Economics, Business, Communication, and Political Science to Chairwoman and Commissioners of the Federal Trade Commission (Jan. 29, 2015), at 4 ("antitrust liability requires identifying anticompetitive conduct that creates or maintains market power.") (on file with author); Kirkwood & Lande, supra note 19, at 201 (Congress' central purpose in passing antitrust laws was "preventing firms that have unfairly acquired or maintained market power from charging consumers supracompetitive prices"); Kirkwood, supra note 19, at 2429 (fundamental goal of antitrust law is "protection of consumers from anticompetitive conduct--conduct that creates market power, transfers wealth from consumers to producers, and fails to provide consumers with compensating benefits."); Kirkwood, supra note 20, at 30 n. 163 (Sherman Act and Clayton Act are "targeted largely, if not exclusively, at practices that create market power.").

(109.) If the buyer simply pocketed the concessions its suppliers made, there would be no harm to the buyer's competitors and thus no increase in the buyer's market power. Put differently, when a buyer demands a concession front a supplier, it exploits its existing market power but it does not--without engaging in additional, exclusionary conduct--enlarge that market power, either by acquiring more than it had or by protecting what it had from erosion.

(110.) Exxon Corp. v. Berwick Bay Real Estate Partners, 748 F.2d 937, 940 (5th Cir. 1984) ("[Monopolization is rarely found when the defendant's share of the relevant market is below 70%."); see also Colo. Interstate Gas Co. v. Natural Gas Pipeline Co. of Am., 885 F.2d 683, 694 n.18 (10th Cir. 1989) (observing that to establish "monopoly power, lower courts generally require a minimum market share of between 70% and 80%"); United States v. Dentsply Int'l, Inc., 399 F.3d 181, 187 (3d Cir. 2005) ("[A] share significantly larger than 55% has been required to establish prima facie market power.").

(111.) See U.S. Dep't of Justice, Competition and Monopoly: Single-Firm Conduct Under Section 2 of the Sherman Act 22 (2008), ("The Department is not aware ... of any court that has found that a defendant possessed monopoly power when its market share was less than fifty percent."). The Obama administration withdrew this report because of disagreements with its enforcement approach. See Press Release, U.S. Dep't of Justice, Justice Department Withdraws Report on Antitrust Monopoly Law (May 11, 2009), http:// ("[T]he Section 2 report will no longer be Department of Justice policy [because the report] raised too many hurdles to government antitrust enforcement and favored extreme caution and the development of safe harbors for certain conduct within reach of Section 2.").

(112.) Colo. Interstate Gas Co., 885 F.2d at 694 (stating that section 2 requires a "dangerous probability that the defendant's conduct would propel it from a non-monopolistic share of the market to a share that would be large enough to constitute a monopoly for purposes of the monopolization offense").

(113.) The only exceptions are the third and fourth scenarios relating to upstream competition, both of which result in monopsony power. Neither scenario, however, would trigger a violation of section 2. In the third scenario, the powerful buyer does not itself acquire monopsony power (an upstream supplier does). And in neither scenario does the buyer engage in conduct that increases its own market power (in the fourth scenario the monopsony power arises because of a change in the upstream market structure that is not caused by the buyer).

(114.) See, e g., Misha Petrovic & Gary G. Hamilton, Making Global Markets: Wal-Mart and Its Suppliers, in Wal-Mart: The Face of Twenty-First Century Capitalism 131 (Nelson Lichtenstein ed., 2006) ("Even for the biggest manufacturers of packaged consumer goods, from Procter and Gamble to Clorox and Revlon and from Del Monte to Nabisco and Sara Lee, the amount of business with Wal-Mart--typically ranging between 15 percent and 30 percent of total shipments--creates a significant dependency on the retailer's demands."); Paul W. Dobson & Roman Inderst, The Waterbed Effect: Where Buying and Selling Power Come Together, 2008 Wis. L. Rev. 331, 356 ("[A] buyer (or a group of buyers) could wield substantial buyer power ... at levels of size and market share considerably below those that are needed to establish seller power in the final market."); Benjamin Klein & Kevin M. Murphy, Exclusive Dealing Intensifies Competition for Distribution, 75 Antitrust L.J. 433, 449 (2008) ("[Significantly lower wholesale prices can be achieved by retailers with relatively small market shares as long as the retailer has the ability to influence the share of its customers' purchases ... obtained by a chosen manufacturer."); see also Kirkwood, supra note 20, at 1503-4 (citing additional support).

(115.) See supra notes 106-7 and accompanying text.

(116.) See Automatic Canteen Co. of Am. v. F.T.C., 346 U.S. 61 (1953); Great Atl. & Pac. Tea Co. v. F.T.C., 440 U.S. 69 (1979).

(117.) See Great Atl. & Pac. Tea Co., 440 U.S. at 78; see also infra note 133.

(118.) The Robinson-Patman Act also addresses promotional discrimination and it, too, would no longer be covered by the Clayton Act.

(119.) See, e.g., Grand Union Co. v. F.T.C., 300 F.2d 92 (2d Cir. 1962); Giant Food Inc. v. F.T.C., 307 F.2d 184 (D.C. Cir. 1962); Alterman Foods, Inc. v. F.T.C., 497 F.2d 993 (5th Cir. 1974).

(120.) See supra note 108 and accompanying text.

(121.) In the pure case--when a buyer extracts a concession from a supplier but does not use the concession to take business from a rival--there is no "competition" involved. The supplier does not compete with the buyer and the buyer's competitors are not injured.

(122.) See McCormick & Co., FTC Dkt. No. C-3939 (Apr. 27,2000) (statement of Chairman Pitofsky & Commissioners Anthony & Thompson) (indicating that McCormick was using discriminatory price cuts not to appease powerful buyers but to take customers from a competing spice producer).

(123.) See, e.g., Harper & Row Publishers, Inc., 122 F.T.C. 113 (1996).

(124.) As noted earlier, other changes would also be desirable: adjustments to the Act's technical and jurisdictional requirements and conforming the Act's treatment of nonprice discrimination to its treatment of price discrimination. But these are beyond the scope of this article.

(125.) 15 U.S.C. [section] 13(a).

(126.) Monahan's Marine, 866 F.2d at 529.

(127.) In two kinds of cases, described in the third and fourth scenarios in Part III.A, it would be possible to show a violation by demonstrating harm to upstream suppliers. In both scenarios, the suppliers are injured by the exercise of monopsony power.

(128.) Hovenkamp, supra note 3, at 156.

(129.) Id. at 169.

(130.) It would not eliminate the Act's protectionist reach because it would allow lawsuits in the numerous cases in which discrimination induced by a powerful buyer is procompetitive.

(131.) It might also be desirable to eliminate section 2's incipiency language. At present, a plaintiff does not have to show that a discrimination is likely to reduce competition or create a monopoly. It need only show that the effect of the discrimination "may be" to lessen competition or "tend to" create a monopoly. In other words, a plaintiff need only establish a "reasonable prospect" of competitive harm. See Brooke Group, 509 U.S. at 224. That standard seems too lax if it allows plaintiffs to challenge instances of discrimination that create a risk of competitive harm but are likely, on balance, to promote competition.

(132.) See 15 U.S.C. [section] 13(b).

(133.) The supplier is not liable because it has a meeting competition defense. The buyer is not liable because, to establish buyer liability, a plaintiff must prove both that the supplier violated the Act and that the buyer knew, or should have known, of the violation. See Great Atl. & Pac. Tea Co. v. F.T.C., 440 U.S. 69, 78 (1979); Automatic Canteen Co. of Am. v. F.T.C., 346 U.S. 61 (1953). If the supplier has a meeting competition defense, the plaintiff cannot establish buyer liability.

(134.) Moreover, without the spur of treble damages and attorneys' fees, private plaintiffs are less likely to file actions. Thus, if the only relief available is equitable relief, buyer-induced discrimination is less likely to be challenged.

(135.) See Texaco, Inc. v. Hasbrouck, 496 U.S. 543, 561 nn.-18, 21 (1990) (noting the exacting standards of the defense and the views of some authorities that it is "difficult, expensive, and often unsuccessful," even "largely illusory in practice.").

(136.) See Hovenkamp, supra note 3, at 174, n. 8.

(137.) Moreover, whenever the cost justification defense is available to a supplier, it should not be subject to hypertechnical requirements. To the contrary, suppliers ought to be able to cost justify a preferential discount if they can show that the discount was "reasonably related" to the lower costs of serving the favored buyer. See American antitrust Institute, The Next Antitrust Agenda: The American Antitrust Institute's Transition Report on Competition Policy to the 44th President 135 (Albert A. Foer ed., 2008) ("A reasonable relationship test would preclude the courts from denying the defense simply because of minor defects in a defendant's cost study ").

(138.) See F.T.C. v. Staples, Inc., 970 F. Supp. 1066, 1087 (D.C.D.C. 1997) (new entry unlikely because a "new office superstore would need to open a large number of stores nationally in order to achieve the purchasing and distribution economies of scale enjoyed by the three existing firms.... Economies of scale at the local level, such as in the costs of advertising and distribution, would also be difficult for a new superstore entrant to achieve").

(139.) These effects are likely to occur whether the retailer is a brick-and-mortar retailer like Wal-Mart or an online retailer like Amazon. In the case of a brick-and-mortar retailer, the relevant distribution costs and delivery times are incurred in moving goods from its warehouses to its retail stores. In the case of an online retailer, distribution costs and delivery times are incurred in shipping goods from warehouses to consumers.

(140.) The D.C. Circuit emphasized this drawback in Microsoft. It overturned an order breaking up the company because the trial judge had not investigated the impact of structural relief on the company's efficiency. See United States v. Microsoft Corp., 253 F.3d 34, 105-06 (D.C. Cir. 2001). To highlight the risks involved, the Court quoted Learned Hand's statement in ALCOA: "[A] corporation, designed to operate effectively as a single entity, cannot readily be dismembered of parts of its various operations without a marked loss of efficiency." See 253 F.3d at 106 (quoting United States v. ALCOA, 91 F. Supp. 333, 415 (S.D.N.Y. 1950)). The Court also repeated Microsoft's claim that it had "only one sales and marketing organization, ... one basic research organization, one product support organization, one operations department, one information technology department [etc.]." Id. (quoting Defendant's Offer of Proof). The Court noted: "If indeed Microsoft is a unitary company, division might very well require Microsoft to reproduce each of these departments in each new entity rather than simply allocate the differing departments among them." Id.

(141.) See Robert W. Crandall, The Failure of Structural Remedies in Sherman Act Monopolization Cases, 80 Or. L. Rev. 109 (2001).

(142.) See Peter C. Carstensen, Remedies for Monopolization from Standard Oil to Microsoft and Intel: The Changing Nature of Monopoly Law From Elimination of Market Power to Regulation of Its Use, 85 S. Cal. L. Rev. 815, 825 (2012):
   These examples support the proposition that such interventions in
   markets that had longstanding monopolistic or oligopolistic
   structures and conduct can on balance have positive effects on
   those markets, including stimulating technological innovation, and
   that there is little hard evidence of any significant adverse
   effects on efficiency resulting from these corporate

(143.) Id. at 817.

(144.) See id. ("[T]he courts responded [to the open-ended standards of an earlier era] by developing stricter standards for the proof of a violation," which, in mm, "shifted the focus of monopoly law to the merits of specific conduct rather than the long-run implications of continued dominance of a market by a firm that could be reconfigured to restore workable competition").

(145.) Franklin Foer pointed to this form of regulation as a possible solution to Amazon's buyer power. See Foer, supra note 14 (one of the "ideas percolating" for controlling Amazon is to "deprive it of the ability to use its site to punish recalcitrant suppliers.").

(146.) Specifically, the buyer's gross margin--the difference between its sales revenues and its cost of goods sold. In the case of a retailer, this essentially means the difference between its retail prices and its purchase prices.

(147.) See Carstensen, supra note 142, at 821 ("The experience of Australia, New Zealand, the United Kingdom, and much of the rest of Europe was that direct economic regulation did in fact inhibit economic growth and innovation.").

(148.) The injunction would also cover non-price discrimination, though that is not the focus of this article.

(149.) The injunction would not only prohibit the supplier from extending discriminatory price cuts to the powerful buyer. It would also bar the supplier from agreeing with the powerful buyer to raise the prices it charges competing buyers. To be sure, the order might not bar an agreement by the supplier to refuse to deal with rival buyers, since the Robinson-Patman Act does not apply to refusals to deal. But such an exclusionary agreement--a form of raising rivals' costs--is clearly covered by section 1 of the Sherman Act.

(150.) To be sure, a reformed Robinson-Patman Act is likely to be more difficult and costly to enforce than the existing Act. But the fundamental goal of the existing Act is problematic. It seems better to have the right target--protecting consumers and powerless suppliers--than having the wrong target--protecting small competitors--even if the wrong target is easier to hit. In fact, though, it is very difficult to hit the existing target. See Leffier & Tatos, supra note 71 (plaintiffs lost 99 percent of the reported secondary line cases during 1996-2006 and 96 percent of the cases during 2006-10). These dismal results are due, at least in part, to the widespread perception that the Act is anticompetitive. Thus, if the Act were reformed, the number of successful Robinson-Patman cases might actually rise.

John B. Kirkwood, Seattle University School of Law, American Antitrust Institute, Seattle, WA, USA

Corresponding Author:

John B. Kirkwood, Seattle University School of Law, American Antitrust Institute, 901 12th Avenue, Seattle, WA 98122-1090, USA.

COPYRIGHT 2015 Sage Publications, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:To Defend or Reform? The Law and Economics of the Robinson-Patman Act
Author:Kirkwood, John B.
Publication:Antitrust Bulletin
Date:Dec 22, 2015
Previous Article:Using the economics of the pass-through in proving antitrust injury in Robinson-Patman cases.
Next Article:Private damage actions under the Robinson-Patman Act.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters