WHAT IS COMPETITION?
William J. Kolasky(1)
Seminar on Convergence
October 28, 2002
Good morning. I am delighted to be here in Den Haag, addressing such a large and distinguished audience. I particularly want to thank Gertjan Lankhorst and his colleagues in the Netherlands Competition Authority for inviting me here to speak.
During the little more than one year I have served as the Deputy Assistant Attorney General for International Antitrust Enforcement, we have undertaken two major international initiatives -- the formation of a new International Competition Network and the strengthening of our bilateral relationship with the European Union. I am pleased to report that we have made more progress in both areas than any of us thought possible just one year ago. The International Competition Network ("ICN") is not only up and running, but is already serving as an important force for international cooperation and convergence and our relationship with the EU, despite our occasional disagreements, is stronger than ever. The European Commission and we worked closely together in launching ICN and making it a success. We have also stepped up the activities of our bilateral US/EU joint merger working group that now meets regularly to discuss important issues of common concern. Based on this group's work, we have developed a new set of best practices for coordinating our merger reviews, which we hope to release shortly.
Over the past year our efforts in the international area have been largely devoted to what I would call institution and relationship building and to improving our processes for merger review. Over the next year, we hope to turn our attention more to the substance of the antitrust laws and to strengthening the analytical framework we use to enforce those laws. And that is what I want to talk about today.
I will begin by discussing the purposes of competition policy. In the United States, we like to say that the purpose of the antitrust laws is to protect competition, not competitors. This principle has now become such a central part of our antitrust jurisprudence that we take it for granted. In discussions with colleagues in Europe, I have been surprised to find that this principle does not resonate over here quite the way it does for us. When I suggested in London last May that we seek to use it as a guiding principle for sound competition policy,(2) one of my fellow panelists, a European competition official, responded that it seemed an empty slogan, devoid of content. And in our debate over GE/Honeywell, Commission officials treated this principle as more of a paradox, arguing that without competitors, there can be no competition.(3) These reactions taught me that if we hope to persuade others to apply this principle in designing and enforcing their competition laws, we need to do a better job explaining what we mean by it. And we cannot do that without first defining what we mean by "competition."
I will follow by discussing how we apply the principle in practice, focusing on three areas, cartels, mergers, and abuse of dominance. On cartels, I will urge your support for strong anti-cartel enforcement. On mergers, I will argue that efficiencies should be a central part of our competitive effects analysis. And on abuse of dominance, I will discuss the need for administrable standards that permit even dominant firms to compete aggressively, lest we restrain competition in the name of protecting it.
Protect Competition, Not Competitors
When a lay person thinks of competition, he or she probably has one of two images in mind. The first is a sporting event, in which two evenly matched opponents, play a spirited, but closely contested, match like the recent World Cup final between Brazil and Germany. The second is a market that resembles a scrum in a rugby match with numerous firms scrambling for every scrap of business -- the more numerous, the more competitive.
Economists see competition differently. An economist sees competition not in terms of rivalry per se, but in terms of market performance. An economist would say that a market is perfectly competitive when firms price their output at marginal cost and costs are minimized by internal efficiency. This does not necessarily require a large number of rivals. Where entry and exit are costless, markets can be perfectly competitive even with only one firm serving the entire market. Similarly, some models of oligopoly show that in some markets prices may be driven to marginal cost, even where there are entry barriers, with as few as two competitors. In both types of markets, allocative and productive efficiency may be perfectly aligned even at relatively high levels of concentration, so that no rearrangement of productive assets could enhance total economic welfare. In these markets, antitrust intervention to preserve or create a larger number of rivals would harm consumer welfare and worsen economic performance.
Joseph Schumpeter was the first to teach us that in other markets, especially those driven by innovation, there may be a tension between allocative efficiency on the one hand and productive or dynamic efficiency on the other.(4) For example, where firms need to invest in order to innovate, prices will need to be above short-term marginal cost to provide an incentive to make the needed investments. As Schumpeter observed, these markets are often marked by "gales of creative destruction," in which one firm may serve the entire market or at least a large portion of it for a period of time, only to be displaced by another firm with a leapfrogging technological innovation that delivers dramatically improved performance or dramatically lower cost. Think of Visicalc being displaced by Lotus, which in turn was displaced by Excel. In these markets an efficient monopolist, constrained by overall market demand and the threat of entry, will often charge quality-adjusted prices that, while above marginal cost, are still below the prices that would be charged by a group of less efficient competitors. To an economist, the competitive process is working in these markets, even if it results in only one firm serving the entire market for some period of time. In such markets, government intervention to preserve rivals or create new ones will again worsen overall economic performance.(5)
To formulate a sound competition policy we need a definition of competition that takes these lessons into account. The one I would propose, and I believe it is one consistent with our case law,(6) is that competition is the process by which market forces operate freely to assure that society's scarce resources are employed as efficiently as possible to maximize total economic welfare.(7)
This formulation allows us to solve the paradox of how we can protect competition without protecting competitors. Competition is fiercest when competitors have no protection from their government; when, to paraphrase Tennyson, competition is "red of tooth and claw." It is in those circumstances that firms will strive the hardest, motivated both by the hope of success and the fear of failure. Nor should we fear that this may result in one firm serving the entire market. If a market is most efficiently served by a single firm, using the antitrust laws to prevent that outcome would not only burden society with the additional costs of having two or more firms serve that market, but would require "an effort worthy of King Canute."(8) As Judge Bailey Aldrich who died earlier this month and for whom I clerked put it, "society has an interest in competition even though that competition be an elimination bout."(9)
I want to be clear, however, as Judge Aldrich was, that antitrust still has an important role to play, even in markets in which the competitive outcome may be a monopoly. The purpose of the antitrust laws is to assure that the war is fought and the outcome determined on the basis of efficiency. The antitrust laws should intervene only when one combatant employs methods that would deny victory to the most efficient firm or create barricades to entry by equally or more efficient new entrants.
Implementing Our Objective of Protecting Competition, Not Competitors
Having tried to explain what we mean when we say that we strive to protect competition, not competitors, let me turn now to how we should implement this philosophy in three key areas: cartels, mergers, and abuse of dominance.
Our number one priority should be vigorous anti-cartel enforcement. Cartels are the very antithesis of competition. They allow small and inefficient competitors to join together to enjoy the easy life at the expense of their customers. Whereas monopolies can sometimes produce greater efficiency, cartels, by definition, have no efficiency-enhancing potential whatsoever.
There is solid empirical evidence that hard core cartels inflict serious efficiency losses on the economy. Estimates of the cartel overcharges by national enforcement agencies for 14 large multinational cartel cases in the period 1996 to 2000 range from 5 to 65 percent, with the median being around 15 to 20 percent.(10) One study found that before the UK adopted anti-cartel legislation, price fixing affected three-quarters of the British industry and reduced average labor productivity by nearly one percentage point.(11) This may sound small but is actually very large given that labor productivity growth in developed countries rarely exceeds 3-4 percent.
As most of you know, cartels are treated as crimes in the United States and have been punishable as felonies, with jail time for individuals, since 1975. We have a very active enforcement program, with nearly one-third of our lawyers working full-time in this area. We are very pleased that under Mario Monti's leadership, the European Commission has now developed an equally vigorous program of anti-cartel enforcement. Although cartels have been unlawful in most major European jurisdictions for decades, we all know that enforcement was lax at best in most jurisdictions until very recently and that cartel behavior was deeply ingrained in the culture of many European industries. It was no accident that many of the major multinational cartels we prosecuted in the 1990s were organized and led by European companies.
Over the last three years, this picture has changed dramatically. Both the European Commission and the member states are stepping up anti-cartel enforcement, increasing the size of their staffs, increasing the penalties for cartel activity, and adopting effective leniency programs. Last year alone, the Commission completed investigations into ten cartels involving a total of 61 companies and imposed fines of more than 1.8 billion euros.
While there is no way to document it, it is likely that lax anti-cartel enforcement was one of the reasons economic growth in Europe has been slower in the past than in the United States.(12) If so, you may reap a "growth premium" from the stepped-up anti-cartel enforcement we are now seeing. I hope those in the audience from the business community will support the efforts of Commissioner Monti and the national competition authorities in this area. One of the most important things you can do (other than not form cartels yourselves) is to implement effective internal antitrust compliance programs. An effective compliance program is critical, not only in preventing cartel activity by your employees, but also in detecting it early so that you can take full advantage of our leniency programs and save your company from potentially disastrous fines and your employees from prison.(13)
Let me turn next to mergers. All of us know that the rationale for most mergers is procompetitive and that most mergers have no adverse effects on competition. As competition officials, we also know, however, that some mergers do create or increase market power, and thereby reduce competition and total economic welfare. Our task as competition authorities is to screen out the few bad mergers from the many good ones. It is critical that we perform this task in a way that does not interfere unduly with the free market for corporate control, because that market plays as important a role as any product market in pressuring managers to perform efficiently.
We have devoted a great deal of effort in the United States over the last quarter century to improving our processes for reviewing proposed mergers. We now are able to clear 97 percent of all mergers in the first thirty days and we now challenge nearly two-thirds of all mergers in which we open a full-blown second request investigation. We are pleased to see that here in Europe you are now devoting an equal or greater amount of attention to strengthening your own review processes.
I have only three observations to offer, based on our own experience in this area.
The first is that efficiencies should be a central part of any competitive effects analysis. There is simply no way to evaluate whether a merger will give the merged firm the ability and incentive to raise prices either unilaterally or in coordination with other firms without examining the efficiencies the merger may produce, whether they be in the form of production efficiencies, transactional efficiencies, allocative efficiencies, or dynamic efficiencies. In evaluating efficiencies, we need to be careful to impose no heavier burden on the parties to prove those efficiencies than we impose on complainants to prove that the merger will harm competition.
My second observation is to emphasize the importance of involving professional economists in merger reviews, both at the staff level during the investigation and at the senior decision-making levels. We now have over 50 Ph.D. industrial organization economists and econometricians working at the Antitrust Division in a separate and independent unit headed by a leading academic economist. We simply could not do our job without them.
My third observation relates to predictability. One of the principal reasons for having merger guidelines is to increase transparency and to make the outcomes of merger investigations more predictable, so as to facilitate efficient transactions. To achieve that objective, it is absolutely necessary that the guidelines be applied consistently. This makes it important, among other things, that jurisdictions follow their own precedents and base their decisions on sound economics supported by strong empirical evidence, not on novel and untested theories.
The final area I want to address is abuse of dominance, or monopolization as we call it in the United States. As I said in London in May, I believe this is now the area of greatest divergence between competition policy in the United States and Europe.
I will focus today on three key topics: first, the need for administrable standards, second, the thresholds for finding dominance, and, third, our framework for separating exclusionary and predatory conduct from competition on the merits. I will close by examining our respective policies toward price cutting by dominant firms, which is probably the area of greatest divergence.
Need for Administrable Standards
Competition law has always treated concerted action more harshly than single firm conduct. The reason is simple. Concerted activity is inherently "fraught with anticompetitive risk" because "it deprives the marketplace of the independent centers of decision-making that competition assumes and demands."(14) But competition also assumes and demands that individual firms be allowed to compete freely. Efforts to regulate the conduct of individual firms, whether through the antitrust laws or otherwise, run the risk of destroying competition in the name of saving it.
For this reason, in both the United States and Europe, the competition laws regulate single firm conduct only when the firm possesses a substantial degree of market power -- what we call monopoly in the United States and what you call dominance here in Europe. In the United States, we also believe it is important that the antitrust laws allow even dominant firms to compete aggressively. Positions of dominance are generally the natural result of market dynamics due to innovation, superior management, technological characteristics, or product differentiation. Punishing dominant firms for their success, and handicapping them to protect their rivals, may have some appeal and may even produce short-term gains, but all too often the only longer-term winners are inefficient rivals protected from the rigors of competition. Therefore, we believe, as Judge Learned Hand put it in one of his most famous aphorisms, that "[t]he successful competitor, having been urged to compete, must not be turned upon when he wins."(15)
It is not clear to what extent European competition policy shares our philosophy. Under European law an undertaking with a dominant position is often said to have "a special responsibility not to allow its conduct to impair genuine undistorted competition in the common market."(16) To the extent this reflects a true difference in attitude, it may derive from our different histories. In Europe, your national markets were formerly protected by internal trade barriers, resulting in more local monopolies, and you had many more government-owned monopolies. As a result, you may have more undertakings with dominant positions that are not the natural result of market dynamics than we do in the United States. If so, greater vigilance by competition authorities may well be warranted to assure that the competitive process can operate freely to restore or create competitive conditions.
Whether or not you share our belief that dominant firms should be allowed to compete as aggressively as smaller firms, I hope you will agree that, whichever philosophy prevails, we need administrable standards to assure that our laws do not unduly interfere with the competition we are trying to protect. We all know that legal institutions are not omniscient and that some error is inevitable. We also all know that fact finding in competition cases is costly and that those costs can deter efficient conduct. Finally, we all know that there is a trade-off between cost and the risk of error -- a system that strives to eliminate all error would almost certainly be too costly to administer. In designing decision rules, it is important therefore that we take into account the relative costs of type I (false positive) and type II (false negative) errors. Because markets tend to be self correcting, we in the United States tend to put more emphasis on reducing false positives in order not to chill competition.
As we design the analytical frameworks and decision rules we will use to apply our competition laws, and especially as we try to incorporate the best economic learning into our decision-making, it is also important to remember that while technical economic discussion helps inform antitrust laws, those laws cannot precisely replicate the economists' views. Unlike economics, law is an administrative system. Rules that embody every economic complexity and qualification may well prove counter-productive by, for example, discouraging legitimate price competition.
Applying these principles, in the United States, we have developed a two-part test for monopolization. To be guilty of monopolization, a firm must both (1) have monopoly power in a well-defined market, and (2) have acquired or maintained that monopoly power through means of exclusionary or predatory conduct rather than what we call "competition on the merits."(17) I understand your abuse of dominance law applies a similar two-part test. Let me turn then to the two legs of this test.
Thresholds for Finding Dominance
Our definitions of dominance or monopoly are very similar. We define monopoly as the power to exclude competitors;(18) in Europe, you define dominance as "a position of economic strength" that gives the undertaking "the power to behave to an appreciable extent independently of its competitors, customers and ultimately of consumers."(19) In applying these tests, we have both relied principally on a firm's market share to determine whether it is dominant. We both, however, recognize the need to look beyond market shares to examine other factors that may constrain a firm's ability to raise price. One of the most important questions is how durable the firm's current market position is likely to be. As we all know, entry and even the threat of entry is a major constraint on behavior, especially in innovation-driven markets.
In identifying and examining these and other non-market share factors, I believe we need a better definition of monopoly power and dominance. We have long known that market shares are at best only a rough proxy for market power, the value of which is increasingly in doubt. The legal definition of dominance and monopoly power -- that is the ability to exclude competitors or to act independently of competitors and customers -- is also of limited use. All firms, even monopolists, are price takers in the sense that market forces determine the price that will maximize profits. We also know that in markets in which there are significant fixed costs, firms will almost invariably have some degree of market power, in the sense of being able to charge prices above marginal cost. Without some market power, there would be no investment or innovation. The difficult question is how much market power is too much. We frankly do not have a good answer to that question. I would urge, therefore, that competition authorities everywhere begin focusing on the question of how to define monopoly power or dominance in ways other than by market share.
Identifying Exclusionary and Predatory Conduct
This brings me to our next difficult question: how do we separate the sheep from the goats?(20) In other words, how do we separate competition on the merits from exclusionary and predatory conduct?
One way to think of these twin concepts is in terms of another sports metaphor. Predation is like punching below the belt and exclusionary conduct is like putting opponents in a stranglehold. The fundamental objective of both rules in boxing is to assure that the fight is decided on the merits -- may the best man win. Translated into economic terms, our competition rules are designed to assure that market outcomes are determined by the relative efficiency of the rivals. This allows us to define exclusionary and predatory conduct as conduct that is likely to exclude equally (or more) efficient rivals from the market.(21) But in order to assure that the bout is determined by the market and not by the referee, it is also important that we not intervene too often or too soon. Otherwise, in boxing all matches would be decided on points, rather than by knock-outs; and in economics, firms would have less incentive to compete hard because the prize for winning and the cost of failure would be smaller. It is for this reason that in the United States we will not intervene unless the conduct is likely to cause serious harm to consumers, not just to rivals.
Let me talk briefly about each of these two types of conduct.
In the United states, in evaluating whether conduct is exclusionary, we use a framework similar to one we use to evaluate alleged restraints of trade under the rule of reason.(22) This framework relies on shifting burdens of proof to structure the inquiry so that judges and juries cannot second-guess business conduct unless it is likely to harm competition and consumers.
First, we require the complaining party to show that the conduct is likely to "harm the competitive process and thereby harm consumers."(23) For this purpose, harm to one or more rivals will not suffice. The plaintiff must show, rather, that the conduct will prevent market forces from operating "to bring to consumers the benefits of lower prices, better products, and more efficient production methods."(24)
Second, if the complainant meets this initial burden, the alleged monopolist is required to demonstrate a "procompetitive justification" for its conduct -- that is, "a nonpretextual claim that its conduct is indeed a form of competition on the merits because it involves, for example, a greater efficiency or enhanced consumer appeal."(25)
Third, if a monopolist shows that its conduct benefits consumers, its conduct will be found lawful unless the complainant can prove that the anticompetitive harm substantially outweighs the potential efficiency-enhancing benefits. This burden might be met by showing, for example, that the same consumer benefits could have been achieved in a less exclusionary manner.
The classic example in our case law of exclusionary conduct is Aspen Highlands where a monopolist, by discontinuing a four-mountain pass it had been offering jointly with its only competitor, lowered the quality of its own product and thereby reduced demand for it in order to exclude that rival.(26) Our case against Microsoft, decided last year, provides almost a laundry list of exclusionary conduct.(27) These included: license restrictions on OEMs that thwarted distribution of competing browsers; integrating IE into Windows in a manner that deliberately made it more difficult for consumers to use another browser; agreements with IAPs, ISVs, and Apple that closed off enough significant channels of distribution to keep usage of Navigator below the critical level necessary for it to become a rival software development platform; deceiving developers into believing that MS Java would operate cross-platform; and pressuring Intel not to support cross-platform Java. In each case, court found that the conduct served to maintain Microsoft's market power in operating systems by preventing Netscape from gaining sufficient sales to become a competing platform and that Microsoft failed to show any legitimate business justifications for its actions.
No area has more bedeviled U.S. antitrust courts for the last quarter century than developing a sound test for predation, and no area has more divided lawyers and economists. What makes this area so difficult is that, as Ken Elzinger reminds us, "if you are hunting for a predator and mistakenly shoot a competitor, you injure consumers."(28)
The U.S. courts now use a two-part test for predation that we believe works quite well.(29) The first part of this test requires that the resulting prices be below "an appropriate measure of cost." The second requires that the monopolist be able to recoup its losses from the period of predation once its rivals have been excluded from the market. Let's examine each test.
The appropriate measure of cost
The reason U.S. courts will not condemn prices that are above cost as predatory is twofold. First, the exclusionary effect of prices above a relevant measure of cost simply reflects the lower cost structure of the alleged predator, and so represents competition on the merits. Second, price cutting that is above cost is beyond the practical ability of a judicial tribunal to control without courting intolerable risks of chilling legitimate price-cutting.(30)
The test that is generally used in the United States for determining whether prices are below cost is the test originally developed by Don Turner and Phil Areeda in 1975.(31) Under the Areeda-Turner test, prices can be found to be predatory only if they are below marginal cost or, if that cannot be determined, below average variable cost. Many economists have objected to the Areeda-Turner test as not economically pure.(32) Whatever merits the alternative tests they propose may have as a matter of economics, no one has yet been able to suggest a test that is as administrable as the Areeda-Turner test or that does any better job in striking a balance between cost of administration and Type I and Type II error, thereby minimizing the risk of chilling aggressive, non-predatory price cutting. For those reasons, our courts generally continue to use the Areeda-Turner test.
The courts do not use this cost-based test to establish a conclusive presumption of predatory pricing. Rather, below-cost prices create a presumption of predation. Many forms of below-cost pricing are motivated by market-expanding efficiencies. These include promotional pricing, pricing to accelerate learning-by-doing, and investment in building a large network where low prices are a way of paying the customer for the incremental value she adds to the network.(33) The courts will, therefore, allow a defendant to rebut the presumption that below-cost prices are predatory by showing that its pricing served a procompetitive, efficiency-enhancing objective.
The reason for the second requirement, recoupment, relates to our definition of competition -- we do not condemn competitive behavior unless it is likely to harm consumer welfare. Without recoupment, predatory pricing produces lower aggregate prices in the market, and consumer welfare is enhanced.(34) A recoupment requirement also serves as a useful screening device. It is often "much easier to determine from the structure of the market that recoupment is improbable than it is to find the cost a particular producer experiences."(35)
Role of Intent
Before I leave the general subject of exclusionary and predatory conduct, let me say a word about the role of intent. In the United States, we believe that intent is an unreliable guide for deciding the lawfulness of single firm conduct, especially in the heads of a jury. As Judge Frank Easterbrook has written, "Firms intend to do all the business they can, to crush their rivals if they can"; "[t]o penalize this intent is to penalize competition itself."(36)
Under our law, if intent is relevant at all, it is to "help us understand the likely effect of the monopolist's conduct."(37) Even here, we are cautious in how we use it because we know that intent evidence, especially in the hands of juries, is generally more likely to mislead than to illuminate.
Price-cutting by Dominant Firms
I'd like to close by focusing on one of the areas of single firm conduct in which I believe European abuse of dominance law is farthest apart from the U.S. law on monopolization. This is the area of price-cutting by dominant firms.
European competition law appears much more hostile toward price cutting by leading firms than U.S. law is. European law appears to take a broader view of predatory pricing than we do, and it also condemns at least two types of non-predatory price cuts -- selective above-cost price cuts in response to new entry, and fidelity or loyalty discounts. Let's look at each briefly.
The leading European Court of Justice decisions on predatory pricing are Tetra Pak(38) and Akzo.(39) In Tetra Pak, the court expressly rejected a recoupment test and instead adopted a cost-based test that also looks at the firm's intent in cutting prices. Under the court's test, when a dominant undertaking prices below average variable cost, those prices are presumed abusive; when it prices above average variable cost, but below average total cost, its prices are abusive if they are intended to eliminate a competitor.
This test for predation is substantially broader than ours. For the reasons I outlined above, we believe a recoupment test is essential because price cuts that do not create any danger of recoupment unambiguously benefit consumers, even if they are below cost. In addition, we are reluctant to base our treatment of price cuts that leave prices above average variable cost on the dominant firm's subjective intent. Two of the main reasons for cutting price are to discourage entry and to take sales away from rivals; if such price cuts are disallowed, competition would necessarily be less intense.
Whatever the relative merits of our two approaches, the European Commission at least seems to be moving away from Tetra Pak's intent-based test toward a test based, as ours is, on whether the costs are below incremental cost. The Commission's Decision of 20 March 2001, for example, established incremental cost as the appropriate measure of cost that an incumbent beneficiary of a statutory monopoly must cover in providing postal services open to competition.(40)
Above-cost selective price cuts in response to entry
Those cases that hold that price cuts can be deemed an abuse of dominance even if they do not meet the requirements for predatory pricing raise additional issues. The leading ECJ case adopting this approach is Compagnie Maritime, where the court held that it was an abuse for a dominant firm to adopt a "fighting ships" strategy of responding to entry by cutting prices even though the resulting prices were above cost where (1) the price cuts were reactive and selective; (2) the reduced prices met and beat the entrant, (3) the price cuts reduced the defendant's profits compared to what they would have been at the previously prevailing prices, and (4) the avowed purpose was to eliminate the entrant.(41)
A forthcoming article in the Yale Law Journal by Professor Einer Elhauge of Harvard outlines four concerns with this approach.(42)
First, restricting selective above-cost price cuts will often penalize efficient pricing behavior. In many markets, incumbent firms can maximize profits and output only by charging more to customers that value the product more highly, thus making them bear a greater proportion of common costs. In such markets, selective discounts to customers on the margin will be output and welfare enhancing.
Second, restricting above-cost price cuts is undesirable because it exchanges certain short-term loss for an uncertain long-term gain. Restricting such price cuts is likely to increase price and harm consumer welfare in the lion's share of cases.
Third, restricting above-cost price cuts will lessen the pressure on rivals and potential entrants to become more efficient, which will mean higher costs and lower quality for society as a whole.
Fourth, these adverse effects are worsened by implementation difficulties that are not avoidable but rather are an inherent consequence of trying to regulate firm pricing, output and responsiveness to entry.
The third area in which the European courts frequently find above-cost price cuts abusive relates to fidelity or loyalty discounts. The European courts treat such discounts as an abuse of dominance on the theory that they raise switching costs and barriers to entry, unless they are cost-justified in which case they can be viewed as "normal" competition.(43) Under U.S. law, similar arrangements have only rarely been challenged and have generally been found to be lawful unless the resulting prices are predatory.(44)
These disparate approaches point up a fundamental difference between European and U.S. law with respect to pricing by dominant firms. In the U.S., we view price cuts as inherently efficiency enhancing; as now-Justice Breyer explained in Barry Wright, "price cuts that leave prices above incremental costs are probably moving prices in the right direction -- toward the competitive ideal."(45) In Europe, by contrast, the courts see price cuts as efficiency-enhancing only if they reflect lower costs, a much narrower definition of efficiency.
We would agree that there may be some narrow circumstances in which fidelity rebates may be anticompetitive, at least when they involve bundling products as to which a firm has a monopoly with other products for which it faces competition. By foreclosing the market share rivals need to reach minimum efficient scale, bundled discounts may, in some narrow circumstances, serve to exclude equally efficient rivals from the market. We believe, however, that we need to weigh these potential anticompetitive effects against the many potential procompetitive justifications for such rebates. For example, when a manufacturer has significant fixed costs, average costs of production will exceed marginal cost, at least up to full capacity utilization. In these circumstances, fidelity rebates may be an efficient way to lower price to sell more output to customers on the margin without having to lower price on all units of output, which a firm would be unlikely to do because it would severely erode its profitability. In such circumstances, fidelity rebates will be efficiency enhancing and will benefit consumers even if rivals exit.
In summary, I would urge that we all take a hard look at this entire area of pricing by dominant firms. In the United States, we currently have two cases pending in our courts of appeals raising these issues. One is our American Airlines case, in which we accuse American on engaging in predation by adding money-losing flights on routes served by new, low cost carriers.(46) The other is LePage's, now before the Third Circuit en banc.(47) That case involves bundled discounts by the monopoly seller of Scotch-brand tape in order to exclude smaller private-label manufacturers from the market. The pendency of these two cases has given rise to a large and growing number of articles in the legal and economic journals grappling with these difficult questions.(48) I suspect you have a similar level of interest in these issues on this side of the Atlantic, and I believe we would benefit from sharing our perspectives with one another.
In closing, I again want to thank my friends at the Netherlands Competition Authority for inviting me here today. I very much value the opportunity to initiate a serious and substantive transatlantic dialogue on these important issues, which I believe are critical to the future performance of our economies. I would be happy to take your questions.
1. Deputy Assistant Attorney General for International Enforcement, Antitrust Division, U.S. Department of Justice. These remarks reflect my personal views and not necessarily those of the Department. I want to thank Christina Akers and Gloria Jenkins for their contributions. Any mistakes are, of course, my own.
2. William J. Kolasky, "North Atlantic Compeition Policy: Converging Toward What?," Address at the BIICL Second Annual International and Comparative Law Conference (May 17, 2002) (transcript available on U.S. Department of Justice Antitrust Division web site, www.atrnet.gov).
3. Götz Drauz, Unbundling GE/Honeywell: The Assessment of Conglomerate Mergers Under EC Competition Law, 2001 Fordham Corp. L. Inst. 183, 197-198 (B. Hawk, ed 2002).
4. See J.A. Schumpeter, Capitalism, Socialism and Democracy (1942).
5. A number of empirical studies have confirmed Schumpeter's basic insight. They show that the social costs associated with the static resource misallocation caused by market power are generally quite small while the gains from increases in productive and dynamic efficiency can be very large. See A.C. Harberger, Monopoly and Resource Allocation, 44 American Economic Review 2 (1954); F.M. Scherer & D. Ross, Industrial Market Structure and Economic Performance (1990). These studies have important implications for competition policy. They show, among other things, a "hump-shaped" relationship between concentration and innovation, with innovation occuring at the fastest pace in industries falling in the mid-range of concentration. See P. Aghion, N. Bloom, R. Griffith & P. Howitt, Competition and Innovation: An Inverted U Relationship, The Institute for Fiscal Studies (February 2002).
6. See e.g., Rebel Oil Co. V. Atlantic Richfield Co., 51 F.3d 1421, 1433 (9th Cir. 1995), ("Competition consists of rivalry among competitors. Of course, conduct that eliminates rivals reduces competition. But reduction of competition does not invoke the Sherman Act until it harms consumer welfare. Accordingly, an act is deemed anticompetitive under the Sherman Act only when it harms both allocative efficiency and raises the prices of goods above competitive levels or diminishes their quality.")
7. This definition is similar to that proposed by Areeda and Hovenkamp. See Philip E. Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application: Vol. 1, 4 (2nd ed. 2000), ("Today it seems clear that the general goal of the antitrust laws is to promote "competition" as the economist understands that term. Thus we say that the principal objective of antitrust policy is to maximize consumer welfare by encouraging firms to behave competitively, while yet permitting them to take advantage of every available economy that comes from internal or jointly created production efficiencies, or from innovation producing new processes or new or improved products.").
8. Omega Satellite Prods. Co. v. City of Indianapolis, 694 F.2d 119, 126 (7th Cir. 1982) (Posner, J.).
9. Union Leader Corp. v. Newspapers of New England, 284 F.2d 582, 590 (1st Cir. 1960).
10. See OECD, Report on the Nature and Impact of Hard-Core Cartels and Sanctions Against Cartels Under National Competition Laws, DAFFE/COMP (2002)/7.
11. See S. Broadberry & N. Crafts, British Economic Policy and Industrial Performance in Early Post-War Period, 38 British History (1996).
12. See The Need for Shock Treatment- the EU Barcelona Summit, Economist (Mar. 9, 2002) (reporting that the gap in GDP per capita between Europe and the U.S. widened from 1/2% in 1991 to 54% in 2001).
13. For advice on how to design an effective compliance program, see William J. Kolasky, "Antitrust Compliance Programs: The Government Perspective," address at the Corporate Compliance 2002 Conference, Practising Law Institute (July 12, 2002) (transcript available on U.S. Department of Justice Antitrust Division web site, www.atrnet.gov).
14. Cooperweld Corp. v. Independence Tube Corp., 467 U.S. 752, 769 (1984).
15. United States v. Aluminum Co. Of America (ALCOA), 148 F.2d 416, 430 (2nd Cir. 1945).
16. Nederlandsche Baden-Industrie Michelin NV v. Comm'n, Case No. 322/81,  ECR 3461.
17. See United States v. Grinnell Corp., 384 U.S. 563, 570-571, 576-580 (1966).
18. See id. at 576.
19. Hoffman-LaRoche & Co. AG v. Comm'n, Case No. 85/76,  ECR 461.
20. See Barry Wright Corp. V. ITT Grinnell Corp., 724 F.2d 227, 232 (1st Cir. 1983).
21. See Richard Posner, Antitrust Law 195 (2nd ed. 2001).
22. See United States v. Microsoft Co., 253 F.3d 34, 59 (D.C. Cir. 2001).
23. Id. at 58.
24. Interface, Inc. v. Mass. Port Authority, 816 F.2d 9, 10 (1st Cir. Mass. 1987).
25. Microsoft, supra note 22, at 59.
26. Aspen Skiiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985).
27. Microsoft, supra note 22, at 59-78.
28. Kenneth G. Elzinga & David E. Mills, Predatory Pricing and Strategic Theory, 89 Geo. L.J. 2475, 2479 (2001).
29. See Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209 210 (1993).
30. See id. at 223.
31. Philip Areeda & Donald F. Turner, Predatory Pricing and Related Practices Under Section 2 of the Sherman Act, 88 Harv. L. Rev. 697 (1975).
32. See, e.g., Oliver E. Williamson, Predatory Pricing: A Strategic and Welfare Analysis, 87 Yale L.J. 284 (1977); Janusz A. Ordover & Robert D. Willig, An Economic Definition of Predation: Pricing and Product Innovation, 91 Yale L.J. 8 (1981).
33. See Elzinga, supra note 27, at 2485. See also William J. Kolasky, Network Effects: A Contrarian View, 7 Geo. Mason L. Rev. 577, 605 (1999).
34. See Brook Group, supra note 29, at 210.
35. A.A. Poultry Farms, Inc. v. Rose Acre Farms, Inc., 881 F.2d 1396, 1401 (7th Cir. 1989) (Easterbrook, J.).
36. Id. at 1401.
37. Microsoft, supra note 22, at 59.
38. Tetra Pak International v. Commission (Tetra Pak II) (Case C-333/948)  ECR I-5951.
39. AKZO Chemie BV v. Comm'n, Case No. C-62/86,  ECR I-3359.
40. United Parcel Service/Deutsche Post AG, Case No. COMP/35.141, Comm'n decision of March 20, 2001 (O.J. L125, pp. 27-44).
41. Compagnie Maritime Belge Transports v. Comm'n, Opinion of Advocate General ¶¶ No. 111-139, C-395-96O & C-396/96P (Oct. 29, 1998).
42. Einer Elhauge, Why Above-Cost Price Cuts to Drive Out Entrants Do Not Signal Predation or Even Market Power-- And The Implications for Defining Costs, Yale L.J. (forthcoming) (manuscript on file with author).
43. OECD, Roundtable on Loyalty or Fidelity Discounts and Rebates, (May 23, 2002).
44. See, e.g., Concord Boat Corp. v. Brunswick Corp., 207 F.3d 1039, (8th Cir. 2000); Virgin Atlantic Airways LTD v. British Airways PLC, 257 F.3d 256 (2d Cir. 2001).
45. Barry Wright, supra note 20, at 232.
46. U.S. v. AMR, 140 F.Supp.2d 1141 (D.Kan. 2001), appeal pending.
47. LePage's, Inc. v. 3M, 277 F.3d 365 (3rd Cir. 2002), rehearing en banc pending.
48. See, e.g.,Elhauge, supra note 41; Aaron S. Edlin, Stopping Above-Cost Predatory Pricing, 111 Yale L.J. 941 (2002); Willard K. Tom, David A. Balto, & Neil W. Averitt, Anticompetitive Aspects of Market-Share Discounts and Other Incentives to Exclusive Dealing, 67 Antitrust L. J. 615 (2000); Dennis W. Carlton, A General Analysis of Exclusionary Conduct and Refusal to Deal -- Why Aspen and Kodak Are Misguided, 68 Antitrust L. J. 659 (2001).
What Is Competition?
Monday, October 28, 2002
Updated June 25, 2015