Section 2 and Article 82:
Cowboys and Gentlemen
J. Bruce McDonald
College of Europe
June 16-17, 2005
Professor Damien Geradin invited me to outline the main differences between the ways Article 82 and Sherman Act § 2 are interpreted and applied, as well as the vision U.S. enforcement authorities have of the way Article 82 is applied in the EU. This seems a timely topic, even shortly before dinner, as each of these jurisdictions is holding a debate on how it should deal with conduct by dominant firms or monopolists that may harm rivals or competition. The EC is deep into study on how Article 82 should be reformed, American enforcers and others are struggling to define a workable standard for exclusionary conduct, and the courts of both jurisdictions are bringing about changes.
In his correspondence to me, Professor Geradin noted that abuse of dominance is an area where there is "little convergence" between EU and U.S. law. Damien is a diplomat. The "little convergence" has sparked EU and U.S. commentators to compare the two jurisdictions' systems with questions like:
Over time, we have observed some divergence and at least a little convergence, but some degree of non-convergence might have been predicted from the different statutory language, with which we begin.
Article 82 of the Treaty establishing the European Community reads:
Section 2 of the Sherman Act of 1890 states:
Every person who shall monopolize, or attempt to monopolize . . . shall be deemed guilty of a felony, and . . . shall be punished by fine . . . or by imprisonment . . . .
The U.S. courts have stated the elements of a Section 2 violation. Monopolization requires (1) the willful acquisition or maintenance of monopoly power (2) by the use of exclusionary conduct.
Without elaborating yet on what may be required for an Article 82 violation, we can begin to see contrasts. One law exists to police the conduct of dominant firms, the other to prevent the creation or maintenance of monopolies. One requires only "abuse," the other exclusionary conduct.
More telling are the decisions of the two jurisdictions' prosecutors and the holdings of our courts, which show more starkly our incomplete convergence.
Before digging into the non-converging enforcement decisions and caselaw, let me invite this European audience to spend a moment with me in a United States antitrust court, to give you a flavor of the principles that animate U.S. monopoly law.
There are at least three important principles that govern the courts' application of Section 2. The most fundamental is that the purpose of the antitrust laws quoting from the Supreme Court
is not to protect businesses from the working of the market; it is to protect the public from the failure of the market. The law directs itself not against conduct which is competitive, even severely so, but against conduct which unfairly tends to destroy competition itself. It does so not out of solicitude for private concerns but out of concern for the public interest.(1)
Or, as one U.S. judge and antitrust scholar has put it: "Competition is a ruthless process. A firm that reduces cost and expands sales injures rivals sometimes fatally. . . . These injuries to rivals are byproducts of vigorous competition, and the antitrust laws are not balm for rivals' wounds. The antitrust laws are for the benefit of competition, not competitors."(2)
It is principles like these, and such a blunt articulation of principles, that gave a German journalist reason to call the U.S. system "cowboy capitalism."(3)
A second important principle governing the application of Section 2 is that "[a] monopolist, no less than any other competitor, is permitted and indeed encouraged to compete aggressively on the merits."(4) One reason for this is that aggressive competition, even though it may harm less-efficient firms, "is precisely the sort of competition that promotes the consumer interests that the Sherman Act aims to foster."(5) Another is that striving for monopoly is "an important element of the free-market system" because "it induces risk taking that produces innovation and economic growth,"(6) so, "[t]he successful competitor, having been urged to compete, must not be turned upon when he wins."(7)
That's cowboy capitalism.
The third principle is that "[s]ubjecting a single firm's every action to judicial scrutiny for reasonableness would threaten to discourage the competitive enthusiasm that the antitrust laws seek to promote."(8) A basic reality of competition policy is that it "is sometimes difficult to distinguish robust competition from conduct with long-run anticompetitive effects."(9) This is especially clear in the case of aggressive price cutting, because the mechanism through which competition may be excluded "is the same mechanism by which a firm stimulates competition."(10) Consequently, the legal system is apt to produce "mistaken inferences" of predatory pricing, which"are especially costly, because they chill the very conduct the antitrust laws are designed to protect.(11)" Much the same is true for conduct other than predatory pricing, and as a general matter, the Supreme Court has reasoned that the "cost of false positives counsels against an undue expansion of § 2 liability."(12)
In contrast, in Europe, large firms are not entitled to be so freewheelingly focused on their own success. In EU markets, dominant firms must keep an eye on maintaining competition by rivals. To quote the Court of First Instance:
[W]hilst the finding that a dominant position exists does not in itself imply any reproach to the undertaking concerned, [the dominant undertaking] has a special responsibility, irrespective of the causes of that position, not to allow its conduct to impair genuine undistorted competition on the common market.(13)
Requiring such care by dominant firms led one EC economist to comment that Europe expects its companies to "compete like gentlemen."(14)
With these two images in mind the cowboy capitalist and the gentleman competitor let's explore further the "non-convergence" of these two systems in the context of some particular categories of conduct and court cases.
One of the most controversial topics in U.S. antitrust today is under what circumstances a monopolist has a duty to deal with competitors, to assist rivals, to share essential facilities. And EU law on a dominant firm's duty to deal has changed in the last few years. I think it fair to say that, in both jurisdictions, the duty to deal has become less generous to the big guy's rivals, but the laws of the two jurisdictions still differ significantly.
To describe current EU law on duty to deal, a subject that all of you likely know better than I do, I'll mention the Bronner and IMS Health cases.
The 1998 Bronner v. Mediaprint decision by the ECJ appears to have been a watershed.(15) To remind us of the facts: Mediaprint published two daily newspapers in Austria, with almost half the market, and Bronner was an upstart new competitor. Mediaprint had developed an effective delivery scheme, a Mediaprint affiliate's delivering newspapers to subscribers' homes in the early morning. Bronner sought access to Mediaprint's delivery scheme, asking Mediaprint to also deliver Bronner's newspaper, Der Standard, for reasonable remuneration. Mediaprint refused, and Bronner asked the Austrian courts to order Mediaprint to give Bronner access. (I can picture the confused expressions of a legion of newspaper delivery boys, ordered by a court to carry a second newspaper on their bicycles each morning.) The national court asked the EU courts whether under EU law Mediaprint had abused a dominant position.
Bronner asserted that Mediaprint's delivery scheme was an "essential facility" and that Mediaprint therefore was required to share. Mediaprint argued that the fact it held a dominant position did not oblige it to subsidize competition by assisting competitors.
The ECJ laid out a three-part test for refusal to deal liability. The Court said that, if the market was nationwide home-delivery schemes, then Mediaprint had a dominant position, and its refusal to deal could be an abuse if three cumulative conditions were met:
(I confess to not appreciating the difference between elimination of competition and indispensability.) The Court held that there was no abuse. There were other distribution methods in fact others were used by other publishers and Bronner could establish its own delivery scheme.
The Bronner case brought EU law one step closer to U.S. law, as Bronner established that, to impose a duty to share, it was not enough that the defendant have a dominant position, but it must also be found that the facility is indispensable and that the plaintiff is unable to compete without it.
In the IMS Health case, the ECJ reinforced the Bronner decision's three conditions.(16) You know the facts: IMS provides pharmaceutical companies with data on wholesaler sales to pharmacies in Germany. Over several years, with input from the pharmaceutical industry, IMS developed a copyrighted "brick" structure, according to which German postal districts were broken into 1,860 areas, "bricks." IMS refused to license the brick structure to competitors, who felt disadvantaged because customers had begun to rely on the IMS brick structure to evaluate such data. On reference from a German court, the ECJ said that, while the refusal to grant a license is not an abuse except in exceptional circumstances, it was sufficient if three conditions were met:
These are similar to those in Bronner, although somewhat more demanding, requiring that the licensee intend to offer a new product or service. (Is this difference perhaps derived from the fact IMS involved IP rights?) The text of the Bronner decision left unclear, is it sufficient for an abuse for the refusal to deal to make it impossible for the complaining rival to compete, or must it eliminate all competition in the downstream market? IMS implies the latter, which seems certainly the right rule.
To contrast U.S. duty to deal law, I will start with the U.S. Supreme Court case Aspen Skiing, and then mention the Court's recent Verizon v. Trinko decision.
The general rule in the U.S. is that a firm, even a monopolist, is not obliged to cooperate with its competitors. Forced sharing is the exception, and the Supreme Court's 1985 Aspen Skiing decision is the leading U.S. case upholding liability for refusing to cooperate with a competitor.(17)
There are four good snow skiing areas near Aspen, Colorado. The defendant in the case, Aspen Skiing Company, owned three of them, its rival the plaintiff Highlands owned one. The two had voluntarily cooperated for years to sell a joint, multiple-day, all-area ski ticket. After repeatedly demanding an increased share of the joint proceeds, the defendant canceled the joint ticket. The plaintiff, concerned that skiers would bypass its one mountain if customers would not buy a joint ticket, tried a variety of increasingly desperate measures to get Aspen Skiing to re-create the joint ticket, even offering to buy Aspen Skiing tickets at retail price. The defendant refused, the plaintiff sued, and a jury found that the defendant had monopolized the Aspen skiing market by ending the joint marketing arrangement with the plaintiff.
On appeal, the Supreme Court was asked whether a monopolist ever has a duty to cooperate with its smaller rivals. Yes, said the Aspen Court, upholding the jury verdict. The Aspen Court did not articulate the specific circumstances in which a monopolist has a duty to share. But in last year's Trinko decision, the Supreme Court stated that what was important in Aspen was the fact that:
[t]he unilateral termination of a voluntary (and thus presumably profitable) course of dealing suggested a willingness to forsake short-term profits to achieve an anticompetitive end. Similarly, the defendant's unwillingness to renew the ticket even if compensated at retail price revealed a distinctly anticompetitive bent.(18)
This indicates that conduct that is unprofitable for the monopolist but for its anticompetitive effect may be exclusionary.
The Trinko Court's description of Aspen's reasoning is consistent with making a distinction between a first time refusal to supply (as in Trinko) and the termination of an existing relationship (as in Aspen), but certainly did not make this explicit.
The Trinko case itself involved the duties of a descendant of the former AT&T telephone monopoly, Verizon, which had inherited a monopoly in local telephone service in a region of the country. As part of Congress's efforts to eliminate such monopolies, the 1996 Telecommunications Act required that Verizon lease part of its network to new competitors at cost, which Verizon had failed to do properly, violating the Telecommunications Act. The plaintiff, Trinko, a telephone customer of one of the new competitors, claimed that Verizon's failure to share with the new competitor as the Telecommunications Act required was also an antitrust violation.
Trinko's case got a cool reception at the Supreme Court. First, the Court held, a violation of the Telecommunications Act is not ipso facto a violation of the Sherman Act. More importantly, the Trinko Court emphasized strongly that it is not unlawful for a firm to exploit its market power, that the general rule is that the monopolist has the right to refuse to deal with other firms, and that exceptions to that rule will be rare. Exceptions will be especially rare where there is government regulation: where there is regulation, do not expand antitrust. Along the way, the Trinko Court put some limits on refusal-to-deal claims, saying that Aspen "is at or near the outer boundary of § 2 liability."(19)
The U.S. antitrust agencies filed an amicus curiae brief in Trinko. The government proposed a standard under which a monopolist would be liable for refusing to deal only where it was anticompetitive and otherwise costly for the monopolist. Specifically, the standard we advocated was, "conduct is not exclusionary or predatory unless it would make no economic sense for the defendant but for its tendency to eliminate or lessen competition." The analysis of the Supreme Court in Trinko is consistent with that standard. (Again, the Court did not explicitly adopt the standard, much less say definitely whether it is a short or long run standard or whether it is limited to refusal to deal cases.)
The Trinko Court highlighted the policy reasons behind limiting duties to share. Compelling that firms share the resources that give them a competitive advantage is in tension with the underlying purposes of the antitrust laws. A duty to share lessens the incentives for a firm, and its rivals, to invest in developing those resources. Forced sharing also requires the courts to determine when resources are competitively advantageous and indispensable and at what price they should be made available, a set of decisions better left to the market. Compelling cooperation between competitors may facilitate collusion, "the supreme evil of antitrust."(20)
Finally, the Court made a point of not relying on the essential facilities doctrine. Whatever the doctrine's value in the U.S., it can be used only in the context of a Section 2 violation, which requires exclusionary conduct, and cannot be used by itself to impose antitrust liability in order to remedy what may be a structural problem in the market.
What contrasts are shown by this review of these refusal to deal cases in the EU and U.S.? There are differences in (first) the market power the defendant firm must have, (second) the nature of exclusionary conduct that may be unlawful, and (third) the market effect required to prove a violation.
First, a fundamental difference is that EU law seeks to control the conduct of firms that are dominant, while U.S. law addresses monopolies, the creation or maintenance of monopoly power. Take this practical comparison of market share thresholds. The dominance standard the power to behave to an appreciable extent independently of competitors, customers and ultimately consumers(21) allows a presumption of dominance where a single undertaking holds 50% or more of the market, and less may be enough. The U.S. standard the ability to raise price and exclude competition would rarely be proved where market share is less than 70%. Of course, neither jurisdiction relies solely on market share evidence. In any event, dominance is less than monopoly.
Second, another difference is in the nature of the conduct that may violate the law. Both describe the forbidden conduct as "exclusionary." However, the threshold for what sort of conduct qualifies as exclusionary is lower in Europe, where EU law requires only that the EC establish that the conduct in question prevents rivals from competing by means other than "normal competition" or "competition on the merits," or that it otherwise violates the special responsibilities of dominant firms.(22)
In contrast, U.S. law will not condemn conduct as exclusionary unless it violates the appropriate standard. What the standard should be is our big debate. The government advocated the "no economic sense" standard in Trinko, but others have been seriously proposed, such as an "equally efficient competitor" test and a "proportional balancing" test. In the refusal to deal context in the U.S., the "no economic sense" standard has distinct implications. For example, a firm should not be required to sacrifice profits to sell to competitors at a discount, even if that would on balance produce social gains; therefore, a monopolist's refusal to sell below the monopoly price should not be a violation, because that would require that the monopolist share its profits.
A third difference. In both jurisdictions, a firm dominant or monopolist will be liable for refusing to deal only if it has a negative effect in the market. But what is that negative effect? In the EU, at least in the past, the focus has been on the likely and sometimes presumed impact on the "effective competition structure" and on competitors, and only indirectly on any impact on consumers. EU policy has been based on the premise that consumers are best served by protecting the competitive process, and therefore in Europe it is not necessary to show direct consumer harm. When firms with market power engage in conduct determined (sometimes presumed) to be anticompetitive, consumer harm has been presumed to occur. In the U.S., more attention is given to evidence of actual or likely price or output effects, and consumer harm is more directly implicated because the prohibition is against monopolies, not just abuse of dominance.
Lastly, the U.S. gives special deference to IP holders that refuse to license. There is no free-standing duty to license in the U.S. In contrast, the EC's Microsoft decision which requires Microsoft to make available, on reasonable and nondiscriminatory terms, proprietary information about certain Windows communications protocols appears to be based upon a liability finding that Microsoft failed to supply these protocols to a rival.(23) If so, the U.S. disagrees with imposing liability for an unconditional, unilateral refusal to supply intellectual property, which goes beyond requiring firms merely to refrain from anticompetitive conduct that harms rivals and compels firms affirmatively to assist their rivals.
Another type of conduct that the EU and U.S. have addressed differently is granting fidelity discounts or loyalty payments.
An example that is often used to compare the jurisdictions' approaches is the Virgin Atlantic v. British Airways case, in which Virgin challenged in both EU and U.S. courts BA's granting loyalty payments or rebates to travel agents.
The facts: Dominant U.K. airline BA set up an incentives payment scheme under which travel agents that met certain sales targets were rewarded beyond the usual commission for selling more BA tickets. The rebates linked to the year-by-year growth of BA sales by each agent and applied retroactively, arguably giving agents the incentive at the margin to promote BA over other carriers.
Presumably BA sold more tickets by inducing travel agency loyalty. Despite the BA scheme, Virgin remained successful, but one might assume it was less successful or at least less profitable than it would have been absent the scheme.
Upon Virgin's complaint, the EC investigated and in 1999 condemned BA's practice under Article 82. The Commission decision was appealed to the Court of First Instance, which ruled in 2003. The Commission and Virgin argued that a dominant firm simply may not give discounts or incentives to encourage customer loyalty. BA argued that its rebates could not be held to be an abuse without evidence they had an adverse effect on competitors, customers, or consumers.(24)
Evaluating whether the scheme was exclusionary, the court started from the premise that such fidelity-building rebates, if paid by a dominant firm, will by their nature prevent customers from buying from the dominant firm's competitors and violate Article 82. The court agreed with the Commission that proof of an actual anticompetitive effect was unnecessary:
It is sufficient . . . to demonstrate that the abusive conduct of the undertaking in a dominant position tends to restrict competition, or, in other words, that the conduct is capable of having, or likely to have, such an effect.(25)
The CFI upheld the Commission's decision against BA.
In the U.S., Virgin brought a private action under the Sherman Act, sections 1 and 2. (Section 1 prohibits agreements that unreasonably restrain trade, and generally the challenged agreement is analyzed under the rule of reason, which balances the agreement's procompetitive and anticompetitive effects.) A federal court of appeals upheld summary judgment for BA in 2001. Under Section 1, the court held that Virgin had failed to prove any actual anticompetitive effect, such as higher prices or lower output. Under Section 2, Virgin put forth a complicated theory about how the travel agent rewards had resulted in predatory prices, diverting passengers from Virgin and other airlines. The court rejected Virgin's arguments, observing that Virgin had remained in business and there was no evidence BA had begun to charge monopoly prices.
Again, to compare, the EU approach focused on the competitiveness of rival firms, which could indirectly benefit consumers, but relied on presumed harm to rivals in the market and presumed indirect harm to consumers. The U.S. approach went to actual anticompetitive effects and demanded proof of harm to competition and to consumers.
Other types of conduct have received similarly divergent treatment in the two jurisdictions.
In the area of predatory pricing, in the Brooke Group case, the U.S. Supreme Court imposed a requirement that the predating monopolist be able to recoup the losses incurred while below-cost, which is not required in the EU. In Tetra Pak, the court said pricing below average variable cost could be deemed predatory if "part of a plan for eliminating a competitor."(26) U.S. courts, in contrast, would not consider such intent evidence.
To compare the jurisdictions' views on tying, look to Microsoft. Even though Microsoft has a 90% share and undisputed monopoly in operating systems, the U.S. appeals court required that the government's tying claim be evaluated under the rule of reason, necessitating proof of actual anticompetitive effect in the downstream browser market.(27) In contrast, the EC decision was more generous, finding an unlawful tie between the operating system and downstream media player without proof of effect on consumers.(28)
Finally, there are other categories of conduct that today are always lawful in the U.S., but may support a claim in the EU, such as monopoly leveraging, exploitation of a "shared monopoly" or collective dominance, and high pricing.
What are the reasons for the historical differences between the two jurisdictions?
A variety of factors may help explain. As I began, the language of the statutes differ. And the Treaty has been interpreted in ways that have served its goal of integration undercutting formerly state-owned or state-sanctioned monopolies and opening markets to competition. Relative to U.S. enforcers, historically EC officials appear to have been more concerned with false negatives and more confident in their ability to predict outcomes, and therefore more willing to intervene. Relative to EU officials U.S. enforcers exhibit greater attention to the long run and greater faith in the market's ability to police long run problems. The availability of the same U.S. statutes to private plaintiffs injured by antitrust violations suggests that it should be more difficult to prove a violation of U.S. law. As much as anything, we have historically taken different philosophical approaches to our economies: Europeans are gentlemen. Americans are cowboys.
1. Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 458 (1993).
2. Ball Mem'l Hosp., Inc. v. Mutual Hosp. Ins., Inc., 784 F.2d 1325, 1338 (7th Cir. 1986) (Easterbrook, J.).
3. Olaf Gersemann, Cowboy Capitalism: European Myths, American Reality (Cato Institute, 2004).
4. Olympia Equip. Leasing Co. v. Western Union Tel. Co., 797 F.2d 370, 375 (7th Cir. 1986) (Posner, J.) (internal quotation omitted).
5. Copperweld Corp. v. Independence Tube Corp., 467 U.S. 752, 767 (1984).
6. Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407 (2004).
7. United States v. Aluminum Co. of Am., 148 F.2d 416, 430 (2d Cir. 1945) (L. Hand, J.).
8. Copperweld, 467 U.S. at 775.
9. Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 45859 (1993) (citing Copperweld, 467 U.S. at 76769).
10. Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 226 (1993).
12. Trinko, 540 U.S. at 414.
13. Case T-219/99, British Airways plc v. Commission, 2003 E.C.R. at ¶ 242 (Ct. First Instance 2003).
14. Remarks by Claes Bengtsson, "Loyalty Discounts, the approach in EU," Developments in the Law and Economics of Exclusionary Pricing Practices (Mar. 18, 2004), Washington, D.C.
15. Case C-7/97, Oscar Bronner GmbH & Co. KG v. Mediaprint Zeitungs und Zeitschriftenverlag GmbH & Co. KG, 1998 E.C.R. 1-7791,  4 C.M.L.R. 112 (E.C.J. 1999).
16. Case C-418/01, IMS Health v. NDC Health, 2004 E.C.R., at ¶¶ 48-52 (E.C.J. 2004).
17. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985).
18. Verizon Communications, Inc. v. Law Offices of Curtis V. Trinko, 540 U.S. 398, 409 (2004) (citations omitted) (emphasis in original).
20. Id. at 408.
21. Case 27/76, United Brands v. Commission, 1978 E.C.R. 207, at ¶ 65 (E.C.J. 1978).
22. See Case 85/76, Hoffman La Roche v. Commission, 1979 E.C.R. 461, at ¶ 91,  3 C.M.L.R. 211 (E.C.J. 1979).
23. Case COMP/C-3/37.792 Microsoft, at 4.1.2, 5.3.1, and 6.1.1. (March 24, 2004).
24. Case T-219/99, British Airways plc v. Commission, 2003 E.C.R., at ¶ 250 (E.C.J. 2003).
25. Id. at ¶ 293.
26. Case T83/91, Tetra Pak Int'l SA v. Commission,  2 E.C.R. 755,  4 C.M.L.R. 726, ¶ 149 (E.C.J.), aff'd, Case C-333/94P,  1 E.C.R. 5951,  4 C.M.L.R. 662, ¶¶ 7375 (Ct. First Instance). See also Case C62/86, Akzo Chemie BV v. Commission,  1 E.C.R. 3359 (E.C.J. 1991),  3 C.M.L.R. 215.
27. United States v. Microsoft Corp., 253 F.3d 34, 84 (D.C. Cir. 2001).
28. Case COMP/C-3/37.792 Microsoft, at 126.96.36.199.4 (March 24, 2004).