|This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.|
Remarks of Deborah Platt Majoras
"The Consumer Reigns: Using Section 2 to Ensure a 'Competitive Kingdom'"
June 20, 2006
Good afternoon. Along with my good friend and colleague, Assistant Attorney General Tom Barnett, I am pleased to welcome you to the opening of the joint FTC/DOJ hearings on Section 2 of the Sherman Act. We are privileged that two of our most distinguished antitrust scholars, Professor Dennis Carlton and Professor Herbert Hovenkamp, are here today. Both have made substantial contributions to how we analyze market conduct, and we are privileged to have this opportunity to hear their views.
At the start of any new endeavor, it is important to reflect on why we are undertaking it. Beginning in 1990, the McKinsey Global Institute, lead by founding director, William W. Lewis, undertook a twelve-year study of the economic performance of thirteen nations, seeking to understand globalization and, more fundamentally, the disparities between rich and poor. The study showed that levels of productivity made the difference between rich and poor nations. What, though, made the difference in levels of productivity? The answer, they found, was undistorted competition in product markets. In his book in which he reports the results of the study, Mr. Lewis says, "Most economic analysis ends up attributing most of the differences in economic performance to differences in labor and capital markets. This conclusion is incorrect.
Differences in competition in product markets are much more important."1
McKinsey also asked why the highly productive United States has higher competitive intensity than other nations. Mr. Lewis sums up the answer by saying that, in the United States, "Consumer is king." More specifically, "[t]he United States adopted the view that the purpose of an economy was to serve consumers much earlier than any other society," and we continue to "hold this view more strongly than almost any other place."2 He concludes that, in fact, "Consumers are the only political force that can stand up to producer interest, big government, and the technocratic, political, business, and intellectual."3
This is why we are here. The FTC and the Antitrust Division have the responsibility to ensure that competition in U.S. markets is free of distortion and that consumers are protected not from markets but through markets unburdened by anticompetitive conduct and government-imposed restrictions. This work is critical, indeed central, to the well-being of the American people. Over the past few decades, the United States has substantially deregulated critical industries, including transportation, telecommunication, and energy, to the substantial benefit of the U.S. economy. As government regulators give way to free markets, much of the responsibility for protecting competition shifts to competition agencies and courts. While competition is distorted when governments regulate or intervene excessively, it also is true that private actors can and do distort competition.
Breaking up cartels, preventing mergers that will substantially reduce competition, and halting conduct that goes beyond aggressive competition to distorting competition is vital to promoting vigorous competition and maximizing consumer welfare. We have developed a great deal of consensus regarding appropriate antitrust policy as it relates to cartels, mergers, and other horizontal conduct, as a result of which our enforcement has become more transparent and predictable, making it easier for market participants to make decisions.
Unilateral or "single-firm" conduct, however, still vexes. Even though we can find some respectable measure of consensus around principles that should apply, we find a range of opinions from knowledgeable people about how to apply those principles to enforcement in the market. The question of the proper test that our agencies should apply to conduct of a single firm with market power now has dominated antitrust debate for several years.
We are not alone. Across the globe over the past quarter century, economic systems in which the state owns firms and central planners set prices and levels of output have given way to market competition, where the forces of supply and demand determine prices and allocate resources. U.S. government officials and private citizens have worked hard to promote the economic and political benefits of free markets throughout the world. With attempts to introduce market economies have come new competition authorities, today numbering 100, when only 15 year ago, there were less than two dozen. Even countries that for decades have had nearly total state control over their economies, such as China, are now dedicating substantial resources to drafting competition laws and establishing competition agencies.
Currently, the issue of how to evaluate unilateral conduct is the most heavily discussed and debated area of competition policy in the international arena. Last week, FTC and DOJ officials attended the European Commission's hearing to review their policy under Article 82 of the EC Treaty, which addresses conduct by dominant firms. Officials from the FTC and DOJ also recently held discussions on monopolization issues with our colleagues in Japan, Mexico, and Canada. Two weeks ago, in the Competition Committee of the Organization for Economic Cooperation and Development ("OECD"), we discussed dominance issues in two sessions. And since the International Competition Network ("ICN") established a Working Group on unilateral conduct in May, the FTC, which will co-chair the group, has received expressions of interest from more countries wanting to be involved than in any other working group.
Why the strong interest? First, many nations are facing the challenge of converting from state-owned or supported monopolists to markets with more than one participant -- no small challenge as we have learned in endeavoring to deregulate electricity markets. And, indeed, to enforcers in those nations, companies with market power are the primary evil to attack; conversely, they have not, to a large extent, had horizontal competitors to be as concerned about. Second, disagreement among competition authorities about how to treat unilateral conduct produces uncertainty in national and world markets, reducing market efficiency and imposing costs on consumers. And third, the analysis of unilateral conduct and identification of that which is anticompetitive presents unique challenges not present, or less present, in the core antitrust concern of conduct between competitors.
By now, these unique challenges are familiar. First and fundamentally, it is difficult to distinguish between aggressive, procompetitive unilateral conduct and anticompetitive unilateral conduct. As the U.S. Court of Appeals for the District of Columbia Circuit stated in its en banc decision in the Microsoft case: "The challenge for an antitrust court lies in stating a general rule for distinguishing between exclusionary acts, which reduce social welfare, and competitive acts, which increase it.4 That is tough, because as Judge Diane Wood wrote for the Seventh Circuit, distinguishing between legitimate and unlawful unilateral conduct requires "subtle  economic judgements about particular business practices."5 While difficult, it must be done and done well. Antitrust challenges to legitimate single-firm conduct have substantial negative consequences for the market and thus consumers, as does the failure to challenge anticompetitive unilateral conduct.
Second, the process of distinguishing between permissible and impermissible conduct must be relatively consistent and transparent, so that firms can incorporate it into their decision-making. While there are relatively few findings of Section 2 liability, there nonetheless are a relatively large number of different types of such conduct that potentially raise competition issues.6 The list includes: exclusive dealing, predatory and other forms of pricing, refusals to deal, tying, bundling, rebates, product design, misleading and deceptive conduct, and abuse of government processes, as well as variations on and among each. Developing clear and predictable antitrust standards, and determining how to harmonize them, is even tougher.
Third, while antitrust practitioners have had substantial success devising remedies for joint conduct, devising remedies for single-firm behavior presents significant difficulties. As Professors Areeda and Hovenkamp put it, "By contrast with [concerted conduct], unilateral behavior is . . . difficult to evaluate or remedy by any means short of governmental management of the enterprise."7 It is not an easy thing to do in a way that defines clear obligations and does not restrict competition going forward.
Competition issues related to single-firm behavior likely will continue to implicate a substantial volume of commerce. A thorough examination of the application of the U.S. antitrust laws and economic theory to real-world single-firm conduct not only should advance the ball in the United States, but also will aid U.S. efforts to promote cooperation and convergence among competition authorities throughout the world. We have much to work with. Already, a number of experienced experts have proposed the adoption of a single test for evaluating nearly all types of potentially exclusionary conduct. Some argue for a test that focuses on the impact of the conduct on consumer welfare.8 Others support analyzing whether the conduct involves the short-term sacrifice of profit.9 Others support a "no economic sense" test, which asks whether the cost of engaging in the exclusionary conduct makes sense only because the conduct serves to eliminate competition.10 Judge Posner has written that the inquiry should focus on whether the conduct excludes equally efficient rivals.11 Still other practitioners and scholars oppose the adoption of a single unilateral conduct test, and instead favor consideration of different tests for particular types of potentially exclusionary conduct.12
Moving beyond U.S. borders, there is an even greater diversity of approaches to defining unlawful unilateral conduct. For example, the Australian Trade Practices Act states that a firm is prohibited from taking advantage of its substantial market power for the purpose of eliminating or substantially damaging a competitor, or deterring or preventing competitive conduct or market entry.13 In Japan, a firm may not make it extremely difficult for competitors to carry on business or enter the market.14 Many jurisdictions, including several EC Member States, follow the EC model of enumerating a list of prohibited practices, including imposing unfair prices or trading conditions, limiting production, discriminating between customers, and tying.15
Proponents of various tests and approaches already have done a good job of laying out their relative merits; virtually all acknowledge that their preferred approach may not be perfect. At these hearings, I hope we can tackle the issue by starting with the conduct itself. The hearings will have panels that will focus on specific types of conduct that, at least to date, can implicate liability. We want the panels to discuss the conduct from the market perspective from the ground up, that is, examine why and when firms engage in it, how they do it, and what effects it produces for the firm, for other firms (customers and competitors), and for consumers. We should look at whether firms in competitive markets engage in the same conduct and, if so, examine why they do it.16 We want these discussions, to the extent possible, to include knowledgeable business people or at least their advisors. From these discussions, we then should endeavor to develop signposts for when the conduct may harm competition and when it typically does not. From these signposts, it would be beneficial to draw some guiding principles. Only then should we turn to examining the current state of the law as it has been applied to such conduct and then to determining what workable legal rules can be applied to the specific conduct at issue for the panel. Perhaps at that point, we may examine what we have learned about workable legal rules for individual types of conduct and determine whether we can pull those together into a broader test or set of rules. Even if these hearings do not produce consensus on a universal test or set of tests, I am optimistic that they can identify relative consensus on a number of principles and on how to approach a significant fraction of the single-firm conduct we encounter.
In our discussions, we must be careful not to permit labels or semantic differences to get in the way; in some discussions that I have heard, I have wondered whether we are talking past each other. In addition, the debate must not become so academic that even if it could be resolved, it might not have much practical application in the marketplace. Indeed, when I spoke with a long-time antitrust practitioner about the debate last week, he said that while he thought Section 2 issues were important, the search for the "holy grail" test might just be something in which only about 27 people have an interest.
There is, in fact, substantial consensus about the core underlying principles and factors that should underlie any evaluation of unilateral conduct. Indeed, the fundamental antitrust guideposts for unilateral conduct are quite clear.
First, the only type of unilateral conduct that should implicate the antitrust laws is conduct that produces durable harm to competition, leading to higher prices, reduced output, lower quality, or lower rates of innovation. As much as we may value the success of particular companies, the health of such companies by itself is not the concern of the antitrust laws. To think otherwise, and focus the inquiry on harm to competitors, is not only bad and discredited economics, but would introduce an element of subjectivity into the antitrust laws that would significantly undermine their effectiveness across the board.
Second, there is consensus that antitrust standards that govern unilateral conduct must not deter competition, efficiency, or innovation. This is why we frequently worry about "false positives." Pervasive and aggressive competition, in which firms consistently try to better each other by providing higher quality goods and services at lower costs, is crucial to maximizing consumer welfare and economic growth. Thus, the antitrust laws should never condemn market power that is obtained through the development of superior products and services, regardless of how many competitors are driven from the market or whether it produces a monopoly. As Justice Scalia explained for a unanimous Supreme Court in the Trinko case:
The mere possession of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful; it is an important element of the free-market system . . . To safeguard the incentive to innovate, the possession of monopoly power will not be found unlawful unless it is accompanied by an element of anticompetitive conduct.17
Third, there is consensus that the standards for evaluating unilateral conduct must be clear and practical to administer. While big litigated cases receive the headlines and produce bar association programs, much of the value of sound competition policies comes from the promulgation of practical and straightforward standards that enable firms to avoid engaging in unlawful conduct, with minimal transaction costs. The most analytically sound antitrust principles will provide little real world value if firms cannot interpret them as they make business decisions. Similarly, tests for analyzing whether conduct is exclusionary must be practical for courts to interpret and apply.
While I want to emphasize that I will use these hearings to continue developing my thinking on these issues, I do approach them with, in addition to the broad principles, a number of other hypotheses. First, any legal framework needs to avoid second-guessing business judgments that were objectively reasonable at the time that they were made. An ex post facto examination of the hypothetical effects of alternative courses of conduct is likely to chill legitimate business behavior. Second, to be practical, any legal framework must be able to evaluate conduct that both generates efficiencies and produces anticompetitive exclusion. If we only had to worry about conduct for which the effects are obvious, we probably would not be here today. Third, any test or tests must account for the fact that certain types of unilateral conduct are significantly more likely to cause competitive harm than others. For example, most would agree that unilateral above-cost pricing at monopoly levels should never be condemned under the antitrust laws. Similarly, behavior that some commentators have termed "cheap exclusion" - such as the use of government processes to unlawfully extend the life of a patent -is generally viewed as unlawful exclusionary conduct.18 This may mean that there can be no unitary test, or that we simply need a broad framework that can accommodate a spectrum or sliding scale for levels of likelihood of harm. Proposals have been made for how we might think about the distinctions that should be made, including Deputy Bureau of Competition Director Ken Glazer's proposal that we analyze conduct by distinguishing between conduct that is coercive or incentivizing.19
In Microsoft, the D.C. Circuit applied what I view as a sensible "weighted" balancing approach to Microsoft's conduct that is largely consistent with these three principles. Some have criticized the framework used in Microsoft as insufficiently structured or "unfocused,"20 but I think that if we look at how it was actually applied, it may be a workable framework that incorporates principles for which there is wide consensus. Perhaps the same criticism could be applied to the Section 1 rule of reason analysis, but as applied to, for example, joint ventures, the balancing has been weighted in the right direction.
First, the Microsoft court did not attempt to substitute ex post facto its judgment for that of business judgments that were made ex ante, or to determine what actions might have been better overall for consumers. For example, the court did not base its findings on an ex post analysis of the impact of Microsoft's conduct on the prices charged to consumers.
The Microsoft court also demonstrated that to evaluate whether certain types of unilateral conduct violate the antitrust laws does require an examination of both likely anticompetitive and procompetitive effects. For example, the court analyzed the legality of a Microsoft license provision that prohibited OEMs from modifying the initial boot sequence. Microsoft did not dispute that the restriction limited competition against IE by preventing OEMs from promoting rivals' browsers. The court nonetheless held that the restriction was not an antitrust violation because it concluded that preventing the Windows desktop from ever being seen was a substantial alteration of Microsoft's copyrighted work that could produce harm that "outweighs the marginal anticompetitive effect" of the prohibition.21 The Court performed this analysis across nearly two dozen types of conduct, examining both anticompetitive effects and procompetitive justifications, taking care to ensure not to chill procompetitive behavior.
Finally, the D.C. Circuit made clear that it did not consider all types of unilateral conduct to raise equal concerns under the antitrust laws. For example, the court stated that courts need to be very skeptical about claims that a dominant firm's design changes harm competition and, by implication, violate the antitrust laws.22 As some commentators have observed, when the D.C. Circuit, later in its opinion, addressed the lawfulness of Microsoft's development of a Java interface tool that was incompatible with Sun's tool, the court appeared to suggest implicitly that an extensive balancing inquiry was not required.23 The court ruled that Microsoft's development of the tool was lawful, holding that "[t]he JVM ... allow[s] applications to run more swiftly and does not itself have any anticompetitive effect.24
One final note: I hope that our latest panel(s) on remedies will generate productive discussion. It simply is not possible to implement sound competition policy for single-firm conduct without giving careful thought to remedies. Despite their importance, the issues relating to remedies have not received extensive attention. Although the Microsoft case received "a bit" of notoriety, I have been struck by how few productive discussions of the DOJ remedy and the en banc D.C. Circuit decision accepting the DOJ's approach (while explicitly rejecting other proposed remedies), have actually occurred. While that may have stemmed from market dissatisfaction over the remedy, these hearings will give the Section 2 remedy issue the prominence it requires. After all, devising and drafting remedial provisions in monopolization cases can be more difficult than determining whether a violation has been committed.25 Yet, if a workable remedy cannot be found, the case likely should not be brought in the first place.
At bottom, we must remember that antitrust is the means, not the end. Rather, the end is undistorted competition, driven by "King and Queen Consumer." The challenge is to keep the competition undistorted, without distorting it ourselves in the process. Thank you again for attending the opening of the Section 2 hearings. We look forward to your contributions.
25. See paper submitted by United States to OECD Roundtable on Remedies and Sanction in Abuse of Dominance Cases (June 6, 2006), available at http://www.ftc.gov/bc/international/docs/AbuseRemediesUnitedStates.pdf.