Defining Legitimate Competition: How to Clarify Pricing Abuses Under Article 82 EC

Fordham International Law Journal, Dec 2002

This Article discusses the principles under Article 82 of the Treaty Establishing the European Community concerning anticompetitive or exclusionary abuses involving pricing issues. This Article is structured as follows. Part I outlines the basic economic thinking behind price discrimination and identifies the principal legal situations under Article 82 in which it arises. Discriminatory pricing should only be prohibited (and therefore needs to be justified) in a small number of situations. Parts II-III discuss the specific situations under Community competition law in which price discrimination and the legality of pricing practices may be relevant. Part II discusses rebate and discounting practices, including target (or sales growth) rebates, fidelity or loyalty rebates, and rebates in return for exclusivity. Part III discusses price discrimination that gives rise to distortions of competition between customers. This concerns Article 82(c), a provision that has some parallels with the Robinson-Patman Act under U.S. law. In practice, it will be rare that a profit-maximizing firm will have the ability or incentive to charge different prices to comparable customers to such an extent that competition between those customers will be significantly distorted. Part IV discusses predatory pricing. It is important that prices that remain above average variable cost should nearly always be treated as legal, since rivals will usually be able, and should be encouraged, to compete in that scenario. Community competition law should only treat pricing above average variable cost as unlawful where there is evidence of other abusive behavior linked to that low pricing, in other words a clear plan to eliminate a rival by using a range of illicit practices. Part V discusses a specific instance of predatory pricing--cross-subsidization. Cross-subsidy cases are in essence cases in which the abuse, if there is one, is predatory pricing. Finally, the Conclusion summarizes the author's comprehension of what the correct principles under Community competition law concerning pricing practices should be.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://ir.lawnet.fordham.edu/cgi/viewcontent.cgi?article=1866&context=ilj

Defining Legitimate Competition: How to Clarify Pricing Abuses Under Article 82 EC

FORDHAMINTERNATIONAL LAWJOURNAL Fordham International Law Journal John Temple Lang Robert O'Donoghuey Copyright c 2002 by the authors. Fordham International Law Journal is produced by The Berkeley Electronic Press (bepress). http://ir.lawnet.fordham.edu/ilj - 2002 Article 5 y This Article discusses the principles under Article 82 of the Treaty Establishing the European Community concerning anticompetitive or exclusionary abuses involving pricing issues. This Article is structured as follows. Part I outlines the basic economic thinking behind price discrimination and identifies the principal legal situations under Article 82 in which it arises. Discriminatory pricing should only be prohibited (and therefore needs to be justified) in a small number of situations. Parts II-III discuss the specific situations under Community competition law in which price discrimination and the legality of pricing practices may be relevant. Part II discusses rebate and discounting practices, including target (or sales growth) rebates, fidelity or loyalty rebates, and rebates in return for exclusivity. Part III discusses price discrimination that gives rise to distortions of competition between customers. This concerns Article 82(c), a provision that has some parallels with the Robinson-Patman Act under U.S. law. In practice, it will be rare that a profit-maximizing firm will have the ability or incentive to charge different prices to comparable customers to such an extent that competition between those customers will be significantly distorted. Part IV discusses predatory pricing. It is important that prices that remain above average variable cost should nearly always be treated as legal, since rivals will usually be able, and should be encouraged, to compete in that scenario. Community competition law should only treat pricing above average variable cost as unlawful where there is evidence of other abusive behavior linked to that low pricing, in other words a clear plan to eliminate a rival by using a range of illicit practices. Part V discusses a specific instance of predatory pricing–cross-subsidization. Cross-subsidy cases are in essence cases in which the abuse, if there is one, is predatory pricing. Finally, the Conclusion summarizes the author’s comprehension of what the correct principles under Community competition law concerning pricing practices should be. John Temple Lang Robert O'Donoghue* "Whether any particularact of a monopolist is exclusionary, rather than merely a form of vigorous competition, can be difficult to discern: the means of illicit exclusion, like the means of legitimate competition, are myriad. The challengefor an antitrustcourt lies in stating a generalrulefor distinguishingbetween exclusionary acts, which reduce social welfare, and competitive acts, which increaseit."' INTRODUCTION The above statement neatly encapsulates one of the most difficult issues in antitrust law: how do you have clear rules that distinguish between legitimate and unlawful conduct? Nowhere are these problems more acute than in the area of pricing behavior. The basic question is an apparently simple one: how should antitrust law draw the line between legitimate and desirable price competition, on the one hand, and undesirable and unlawful price competition, on the other? It is important that this balance is correctly struck: the welfare cost and chilling effect of discouraging legitimate price competition is considerable; the cost of allowing unlawful pricing to go unchecked may be no less serious. Although a now universally-accepted distinction is drawn in the European Community ("Community") competition law between exploitative and exclusionary (or anticompetitive) * Mr. Temple Lang is a counsel and Mr. O'Donoghue is an associate at Cleary Gottlieb Steen & Hamilton (Brussels). Mr. Temple Lang is also a Professor at Trinity College, Dublin and a Senior Visiting Research Fellow at the University of Oxford. Nicholas Levy, Christopher Cook, and Cynthia Ngwe (Cleary Gottlieb), Professor George Hay (Cornell University), and Derek Ridyard (RBB Economics) made very useful comments on this Article, but they are not responsible for any errors or omissions on the authors' part. 1. United States v. Microsoft, 253 F.3d 34, 58 (D.C. Cir. 2001). abuses, 2 very little effort has been made to clarify the general principles about the kinds of behavior that are contrary to Article 82 of the Treaty Establishing the European Community ("Article 82") which prohibits abuse of a dominant position. The case law and practice has arisen pragmatically, and largely in response to complaints to the European Commission and appeals to the Community Courts against Commission decisions adopted on the basis of such complaints. With the exception of specialized Notices and guidance in the telecommunications and postal sectors, the Commission has not attempted to develop any kind of general or comprehensive statement on abusive behavior. There have been several consequences of this unplanned growth. First, the Commission and the Community Courts have dealt with individual cases that were said to raise questions of abuse by reference to the facts of the individual case, seemingly without having any clear general analytical or intellectual framework for doing so. Second, a number of basic questions have not been answered or even discussed, because due to the accidents of litigation or otherwise, they did not arise in any of the cases that have been decided. Finally, the influence that economic thinking has had on the Community rules on distribution, horizontal agreements, and mergers has not been felt, to the same extent or at all, in the interpretation and application of Article 82. This Article discusses the principles under Article 82 concerning anticompetitive or exclusionary abuses involving pricing issues. Pricing cases have been chosen for several reasons. First, a fundamental goal of Community competition law is to encourage price competition, including price competition from dominant firms.' Second, pricing practices are relevant to every 2. See BELLAMY & CHILD, EUROPEAN COMMUNITY LAW OF COMPETITION 5TH ED. sec. 9-072 (P.M. Roth QC ed., 2001); see alsoJohn Temple Lang, Abuse of DominantPositions in European Community Law, Presentand Future: Some Aspects, in FirtH ANNUAL FORDHAM CORP. L. INST. 41-55 (Barry Hawk ed., 1978). 3. In Compagnie Maritime Belge Transports SA and others v. Comm'n and others, Joined Cases C-395/96 P and C-396/96 P, [2000] E.C.R. 1-1365, Advocate General Fennelly explained the role of price competition in Community law as follows: Price competition is the essence of the free and open competition which it is the objective of Community policy to establish on the internal market. It favors more efficient firms and it is for the benefit of consumers both in the short and the long run. Dominant firms not only have the right but should be encouraged to compete on price. Id. at 1-1411, para. 117. "Community competition law should.., not offer less efficient 2002] company that is, or may be, dominant: every company has to have a pricing policy, and needs to know what the constraints on its policy may be. Third, because low prices nearly always benefit consumers, any antitrust objections to them should be looked at critically to ensure that the rules are clear and no more than necessary in the circumstances. If the rules are not clear or they are too restrictive, there is a significant risk that legal advisers will be tempted to give overcautious advice. This in turn could lead to a chilling of desirable price competition, with potentially significant welfare implications. Fourth, it is on pricing issues that the Commission seems most clearly to have gotten away from both sound economics and good law. A number of Commission statements on pricing practices come perilously close to stating per se rules against certain forms of pricing behavior. Other statements are liable to be taken out of context and give rise to confusion. It is on pricing issues that the Commission is most obviously running the risk of discouraging legitimate and desirable competition. It is on pricing issues that a clear and comprehensive statement of the legal and economic principles is most urgently needed, not only to guide the thinking of the Commission, companies, and their lawyers, but also for the guidance of national competition authorities which are intended, under the Commission's proposals for decentralization of Community competition law, to apply Article 82 more than they have in the past. Lastly, pricing issues involve some 'of the more significant differences between U.S. and European Union ("EU") antitrust law, and these should be minimized where possible. This Article is structured as follows. Part I outlines the basic economic thinking behind price discrimination and identifies the principal legal situations under Article 82 in which it arises. The freedom of a dominant company under Community competition law to charge different prices for the same product or service has given rise to much discussion, and in our view to unnecessary confusion. There is no general principle that a dominant company must not charge different prices for the same product or service. We argue that. discriminatory pricing should only be prohibited (and therefore needs to be justified) in a small number of situations. First, where a dominant company's offering of undertakings a safe haven against vigorous competition even from dominant undertakings." Id. at 1-1418, para. 132. different prices to customers distorts competition in a meaningful way between those customers. This is a "secondary-line" abuse contrary to Article 82(c). Second, price discrimination may require justification where it leads to the unlawful exclusion of rivals, contrary to Article 82(b). This broad category of "primary line" abuses covers predatory pricing and a range of other pricing practices that may give rise to exclusion concerns. Price discrimination may also be relevant in another situation, that is where a vertically-integrated dominant firm applies less favorable terms to companies that it supplies but that also compete with its downstream business. This mainly concerns situations resembling essential facilities and is not discussed here. In each of these situations of price discrimination, we argue that the principles of law and economics that apply are different, the anti-competitive effects which the law is intended to prevent are different, and the defenses which may be relevant are different. But outside of these specific situations, Community competition law should make clear that different prices may be freely charged, and no defense is needed. This is an important conclusion, because it frees dominant companies from unnecessary and anticompetitive constraints on legitimate competition, and clarifies the legal analysis. Parts 11-111 discuss the specific situations under Community competition law in which price discrimination and the legality of pricing practices may be relevant. Part II discusses rebate and discounting practices, including target (or sales growth) rebates, fidelity or loyalty rebates, and rebates in return for exclusivity. It is perhaps in regard to these practices that certain Commission statements appear to deviate most from established economic thinking. In a number of cases, the Commission has made statements that seem to establish per se rules against the use by dominant firms of these types of normal commercial practices. We argue that Community competition law on discounting proceeds from the wrong premise. If the price is not predatory, it should benefit from a very strong presumption of legality, since it will generally be pro-competitive and based on efficiencies. The situation should only be different where the conditions attached to obtaining the favorable price are anticompetitive. The basic antitrust question is whether the customer would have to agree to an anticompetitive condition, that is to buy exclusively or almost exclusively from the dominant firm, to obtain the most favorable discount. These cases raise difficult issues and involve consideration of a wide range of factual elements concerning the market context and foreclosure effect of the rebate in question. However, we argue that some basic distinctions are clear. If the rebate is not individually negotiated with the customer, it will usually be simply a list price or a generally applicable volume discount that should be regarded as unobjectionable. If the rebate is individually negotiated, we argue that real issues only arise if it leads to exclusive or near-exclusive purchasing in circumstances liable to have a material adverse effect. The availability of a number of defenses should also be considered. The important question we address is how Community competition law should distinguish between price reductions constructed to oblige buyers to buy exclusively from a dominant supplier, on the one hand, and price reductions legitimately constructed to enable buyers to get the best price for the maximum quantity they wish to purchase, on the other. Part III discusses price discrimination that gives rise to distortions of competition between customers. This concerns Article 82 (c), a provision that has some parallels with the RobinsonPatman Act under U.S. law. We argue that, in practice, it will be rare that a profit-maximizing firm will have the ability or incentive to charge different prices to comparable customers to such an extent that competition between those customers will be significantly distorted. Moreover, in many cases, valid defenses will be available to justify those price differences. This means that the only clearly identifiable situation under Community competition law in which the non-discrimination principle would apply and none of the valid defenses would be available is likely to be where a State-owned or controlled company charges different prices to domestic and foreign buyers for protectionist reasons. In most other situations, a dominant firm will have no incentive to treat similarly-situated buyers so differently. Part IV discusses predatory pricing. As in the United States, this topic has given rise to much discussion in the EU, but very few cases in which a successful claim has prevailed. In over forty years of Community competition law, there have only been three instances in which a dominant firm's prices have been found to be predatory. In general, Community competition law's approach to predatory pricing seems reasonable. However, there are several cases under Community competition law that have FORDHAMINTERNA TIONAL LAWJOURNAL treated pricing above average variable cost (and even pricing above average total cost) as exclusionary. These cases are problematic because neither the Commission nor Community Courts have developed a clear analytical framework to explain how price-cutting that remains above cost harms consumers. We argue that it is important that prices that remain above average variable cost should nearly always be treated as legal, since rivals will usually be able, and should be encouraged, to compete in that scenario. Our conclusion is that Community competition law should only treat pricing above average variable cost as unlawful where there is evidence of other abusive behavior linked to that low pricing, in other words a clear plan to eliminate a rival by using a range of illicit practices. Part V discusses a specific instance of predatory pricing cross-subsidization. We argue that cross-subsidy cases are in essence cases in which the abuse, if there is one, is predatory pricing. There is no abuse of cross-subsidizing in the absence of predatory prices, since a price that is above cost does not, by definition, need a subsidy. We also consider what costs must be taken into account in that situation to determine whether the lower price is predatory, and in particular, the issue of allocation of common costs between a competitive and a monopoly market in the light of the Commission's recent Deutsche Post decision. Finally, the Conclusion summarizes what we believe should be the correct principles under Community competition law concerning pricing practices. I. PRICE DISCRMINATION UNDER COMMUNITY COMPETITIONLAW A. Some Basic Economic Concepts Discriminatory pricing is a broad term that covers a range of situations in which a company charges different prices for the same product to similarly-situated customers. Distinguishing between desirable and undesirable price discrimination lies at the heart of the underlying welfare objectives of antitrust law. The economics of price discrimination are complicated, but may be briefly summarized as follows.4 As a basic premise, economists 4. Derek Ridyard, ExclusionaryPricingandPriceDiscriminationAbuses UnderArticle 82 - an Economic Analysis, EUR. COMPETITION L. REV. 2002, 23(6), 286-303, has produced a 2002] consider that marginal pricing - pricing at the level of the extra cost of producing the last unit of production - maximizes consumer welfare. This creates problems for industries (of which there are many) that have high fixed costs and need to recover as much of those costs as possible in order to survive in the longterm. There is also an additional problem in many "new economy" industries where marginal costs are very low, but research and development and innovation costs are high. In these situations, it makes sense that a company may wish to price above marginal cost in order to recover some fixed costs for those who are willing to pay more and at or near marginal cost for those who can only afford to pay less but might not otherwise be able to afford the product or service in question. So if marginal costs are, say, 10% of the list price and there is a customer who is unwilling or unable to pay more than 50% of the list price for the product, it is in the interests of both, the customer and the dominant company to grant the 50% discount. The dominant company gets a significant positive contribution to its revenues from the sale. The customer gets a product, which it could not otherwise afford. The transaction is economically rational and pro-competitive, and it would harm both parties to prohibit it. In other words, economists do not tend to view price discrimination with any particular suspicion, but think that its effects may be benign or at least not obviously anticompetitive. B. Price DiscriminationUnderArticle 82 Despite much discussion of price discrimination in the decisional practice of the Commission and the case law of the Community Courts, it is worth recalling that the only provision of the Treaty Establishing the European Community (the "EC Treaty") that prohibits discrimination is Article 82(c), which prevents dominant companies from "applying dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage." 5 However, the list of good summary of the basic economic principles, and a useful criticism of the lack of economic rigor in some recent Commission decisions on pricing under Article 82. 5. Consolidated version of the Treaty establishing the European Community, art. 82(c), O.J. C 340/3, at 209 (1997), 37 I.L.M. 79 at 71 (ex art. 86) [hereinafter Consolidated EC Treaty], incorporatingchanges made by Treaty of Amsterdam amending the Treaty on European Union, the Treaties establishing the European Communities and certain related acts, Oct. 2, 1997, O.J. C 340/1 (1997) (amending Treaty on European abuses in Article 82 is not exhaustive,6 and at least four clear situations of discrimination can be distinguished: " Where the dominant enterprise is selling to companies not otherwise associated with it, and the companies are in competition with one another. In this situation, it can be an abuse if the difference in treatment is big enough to create a competitive disadvantage for the companies subject to the higher price or the less favorable treatment. This is the situation envisaged by Article 82(c). It is a "secondary line" abuse and is discussed in Part III; • Where the dominant enterprise is treating customers differently on the basis of their nationality. Discrimination on the grounds of nationality is outlawed generally by the EC Treaty and such behavior is unlawful under Article 82(c). This is also a secondary line abuse and is discussed in Part III; " Where the dominant enterprise discriminates in the terms and conditions and prices that it offers in such a way that leads to the unlawful exclusion of rivals. In this situation, the clause in Article 82 that is primarily applicable is Article 82(b) which prohibits "limiting production, markets, or technical development to the prejudice of consumers."7 A dominant enterprise is prohibited by this clause from limiting the production, marketing or technical development of its competitors, as well as its own.' This is a "primary line" abuse. In practice, discrimination in this sense is most likely to occur in two situations. First, where a dominant firm makes a price reduction conditional on the buyer's making all or nearly all of its purchases of the product in question from the dominant Union ("TEU"), Treaty establishing the European Community ("EC Treaty"), Treaty establishing the European Coal and Steel Community, and Treaty establishing the European Atomic Energy Community and renumbering articles of TEU and EC Treaty). 6. See CompagnieMaritime Belge, [2000] E.C.R. at 1-1475, para. 112. 7. Consolidated EC Treaty, supra n.5, art. 82(b), O.J. C 340/3, at 209 (1997), 37 I.L.M. at 71 (ex art. 86). 8. See C66perative vereniging 'Suiker Unie' UA and others v. Comm'n, Joined Cases 40-48, 50, 54-56, 111, 113-14/73, [1975] E.C.R. 1663; Radio Telefis Eireann (RTE) and Independent Television Publication Ltd. (ITP) v. Comm'n, Joined Cases C241/91 P & C-242/91 P, [1995] E.C.R. 1-743; Klaus H6fner and Fritz Elser v. Macrotron GmbH, Case C-41/90, [19911 E.C.R. 1-1979; Non-contentious proceedings brought by Job Centre Coop. arl, Case C-55/96, [1997] E.C.R. 1-7119. 2002] enterprise. If this kind of abuse is in question, it is the condition on which the price reduction is, given which is exclusionary, not the price reduction itself. This situation is discussed in Part II. Second, where there is predatory pricing in the broad sense. This situation is discussed in Part IV; and Where a vertically integrated dominant enterprise's "upstream" activities (e.g., selling an important raw material or other input) discriminate in favor of the dominant enterprise's own downstream activities, and against its competitors in the downstream market, in the terms on which it supplies the raw material or input. The rule is strict, so that if the difference in treatment has any economic significance, there is no need to prove that the downstream competitors are suffering any special competitive disadvantage. Since this situation mainly deals with situations resembling essential facilities rather than pricing issues, it will not be discussed in this Article. There are several reasons for distinguishing between these categories. First, there are differences in the strictness of the applicable rules in each case. Second, a case can only be analyzed correctly when it is clear into which of these categories it may fall. Finally, different kinds of defenses apply for each of the different categories of abuse. II. DISCOUNTING AND REBATE PRACTICES Companies rely on a range of different pricing and discounting practices to capture market share and retain the business of existing customers. These include "loyalty" or "fidelity" rebates, target (or market share growth) rebates, and similar incentives. Such practices are common in many industries and are an essential competitive tool in many others. There is no obvious reason why antitrust law should view such practices with any particular suspicion, since they typically lead to lower prices, which should in nearly all cases benefit from strong presumptions of legality. Leaving aside issues of bundling and tying, the situation might be different in only two instances: first, where the lower price is predatory; and second, where the lower price is conditional on the customer purchasing all or nearly all of its requirements from the dominant firm. These objections are different; in one case it is the price that is the problem; in the other it is the exclusionary conditions attached to the favorable price. This section discusses discounts that exclude competitors by making discounts dependent on the customer's buying exclusively or almost exclusively from the dominant firm. (Predatory pricing is discussed in Part IV.) Community competition law takes the view that rebates that are conditional on a customer's purchasing all or a large part of its requirements from a dominant firm may be abusive. In other words, it recognizes that the conditions attached to the price may be unlawful. While this is not universally accepted as correct, it has some doctrinal basis' and is not obviously wrong.'0 However, the difficulty is that a number of recent Commission statements seem to suggest a general rule that any discount that is conditional on, or simply creates incentives for, a customer's buying some of its requirements from a dominant company is abusive. Other Commission statements suggest that no discount can be offered by a dominant firm unless justified by cost savings 9. See 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. 615-39 (2000). While there is some agreement under U.S. law that rebates conditional on exclusive purchasing may, absent clear efficiencies, be unlawful, there has been little if any judicial endorsement of a similar analysis for rebates conditional on a customer's purchasing a large part of its requirements from a dominant firm. Instead, the U.S. courts have tended to examine whether the discounted price is predatory; if it is not, it is simply a lower price that competitors should be free to match in the absence of an exclusive dealing requirement or a discount that is conditional on exclusive dealing. Thus, in Concord Boat Corp. v. Brunswick Corp., 207 F.3d 1039 (8th Cir. 2000), the 8th Circuit sanctioned a discount plan in which discount levels increased along with the market-share percentage purchased from the seller. The Court found that the discounts were above cost; only a single product was involved; the rival sellers offered similar discounts; and the ease of market entry precluded anticompetitive results from the discounting, as a new firm could enter the market and challenge the prices offered by the dominant firm. In general, U.S. plaintiffs have not been successful in challenging incentive schemes. See FLM Collision Parts, Inc. v. Ford Motor Co., 543 F.2d 1019 (2d Cir. 1976); Western Parcel Express v. United Parcel Service of America, Inc., 190 F.3d 974 (9th Cir. 1999); Virgin Atlantic AirwaysLimited v. British Airways plc, 257 F.3d 256 (2d Cir. 2001), and LePage's Incorporated v. 3M (Minnesota Mining and Manufacturing Company), 2002 WL 46961 (3d Cir. 2002). In SmithKline Corp. v. Eli Lilly and Co., 575 F.2d 1056 (3d Cir. 1978), the plaintiff was successful, but that was on the basis of discounts that were conditional on tied sales. 10. There is no obvious analytical reason why a discount conditional on a customer's purchasing, say, 95% of its requirements (assuming they are known to the dominant company) should be treated differently from a requirement to buy exclusively to obtain the discount. This means that there is a legitimate basis for analyzing such cases under exclusive dealing principles. 2002] or some other objective reason. The Commission has also said that any practice aimed at increasing a dominant firm's market share may be unlawful. These statements cannot be right. Set forth below is a synopsis of the existing law (Section A) and available defenses (Section B), followed by certain comments that seek to clarify that law (Section C). A. Community Law on Discounts and Rebates Apart from predatory pricing, the Commission and Courts have identified three main categories of discounting practices that may be found abusive under Article 82: (1) rebates conditional on exclusive purchasing; (2) loyalty rebates; and (3) target rebates. 1. Rebates In Return For Exclusive Purchasing Rebates granted by a dominant supplier on the condition that a customer purchase its requirements for the relevant products exclusively from that supplier have generally been found abusive. In Suiker Unie, for example, the Court of Justice found that the potential loss of the challenged rebate created an overwhelming incentive for customers who would otherwise have considered purchasing some of their needs from other suppliers to deal with the dominant company. If a customer made one purchase from a competitor of the dominant supplier, the customer lost the entire rebate on all its purchases from the dominant supplier over an entire year. The Court of Justice found that this system placed customers who also bought sugar from other sources at an unjustifiable disadvantage, enabling the dominant supplier to "control" the amount of sugar that its customers bought from foreign producers. Since its customers all depended at least in part on Suiker Unie's deliveries (as customers' storage facilities were inadequate and they needed regular supplies), the disadvantage of losing a rebate applicable to an entire year's purchases would have outweighed any advantage gained by buying some sugar from third parties even if such purchases could be made at more favorable prices. No competitor could sell one consignment at a price that gave the buyer a cost saving equal to the lost rebate on a year's purchases from the dominant supplier. As a result, the rebate gave other producers no chance of competing with the association, foreclosing them from the market.1 2. Loyalty Rebates In Hoffmann-La Roche, the Court of Justice condemned a "loyalty" or "fidelity" rebate, that is to say "discounts conditional on the customer's obtaining all or most of its requirements... from the undertaking in a dominant position.' 2 Roche had offered different customers different prices for identical quantities of the same product, depending on whether or not they agreed to limit purchases from its competitors. The Court examined whether the system had in fact foreclosed rivals from the market and found that it had done so: The fact of agreeing with purchasers that they will buy all or a very large proportion of their requirements from only one source ... removes all freedom of choice from purchasers in their selection of sources of supply, and ties them to one supplier. The special price offered by Roche is the consideration for the abandonment by its purchasers of their opportunities to obtain substantial proportions of their requirements from competitors. Should a purchaser not observe his obligation of exclusivity - by purchasing some of his requirements from another vitamin manufacturer - the fidelity rebate is forfeited not only in respect of the amount of such purchase, but in respect of all his purchases from Roche.' 3 The Court found further that this rebate system had not evolved out of legitimate efforts by the dominant firm to increase its sales, but were rather motivated by specific exclusionary intent. The rebates were: "designed to deprive the purchaser of or restrict his possible choices of sources of supply and to deny other producers access to the market," "intended to give the purchaser an incentive to obtain his supplies exclusively from the undertaking in a dominant position," and "designed through the grant of a financial advantage to prevent customers from obtaining their supplies from competing producers." 4 2002] The Court followed the same line of reasoning in Irish Sugar, where the dominant supplier's fidelity rebate was conditional on the customer's purchasing "all or a large proportion" of its retail sugar requirements from it.15 The Court found that the rebate had foreclosed competitors from the market: The fact that ADM [a customer] previously obtained its supplies from SDL [Irish Sugar's distributor] before being canvassed by ASI [Irish Sugar's competitor] confirms that the granting of that rebate . . . had the effect of tying the customer to the supplier in a dominant position or, in other words, of recovering a customer who was inclined to switch to the competition.' 6 Moreover, the Court found that the rebates formed part of a plan designed specifically to exclude Irish Sugar's competitor: [The] approach to ADM took place in the context of a strategy devised jointly by [Irish Sugar and its distributor] to prevent the expansion of the market by ensuring the fideEliutyroloufxitsbrcaunsdtoomnertsh.'e7 Irish retail The Commission's decision in Soda-ash/Solvay was based on similar reasoning. Solvay had adopted a pricing structure under which it offered customers a standard list price on a basic contractual tonnage amount (usually calculated to represent around 80% of the customer's total annual requirements), and payments and discounts for marginal purchases above the basic amount (the "top slice," for which the customer had potential alternative sources). The top-slice rebate was only given where Solvay was the customer's sole or principal supplier.18 The Commission found that this rebate scheme made it: difficult or impossible for an existing or potential supplier to enter as second supplier for the marginal tonnage, since in order to match the substantial pecuniary advantages offered by Solvay and obtain the order for the top 'tranche' of business, they would have to sell at unprofitable or at 'dumping' prices. 19 Solvay argued that the system was justified, since it simply represented volume discounts that depended on customers reaching objective and pre-determined purchase thresholds. However, in view of Solvay's position as defacto exclusive supplier to most customers and the substantial documentary evidence that: (1) the system was "specifically intended to ensure the loyalty of the customer and exclude or limit competition,"2 0 and (2) the whole purpose of the rebates was "to remove or restrict the opportunity of other producers or suppliers of Soda-ash to compete effectively with Solvay," the Commission rejected this defense. 21 3. Target Rebates The Commission and the Community Courts have treated certain target rebates in a similar fashion to loyalty rebates. The leading case is Michelin, where the Court of Justice found Michelin's rebates linked to annual sales targets abusive. 2 Michelin granted rebates to tire dealers based on an annual sales target that was established individually for each dealer on the basis of several criteria, including the dealer's estimated sales potential and Michelin's share of the dealer's total tire sales. The dealer did not know the criteria that Michelin used in calculating the target, which was not confirmed in writing but only orally by Michelin's representative. Moreover, it was very difficult for the dealer to ascertain how much it was earning on sales of Michelin tires, since dealers would often not discover what their final rebates were until they opened the envelopes that Michelin's representative gave them at the end of each year.23 The Court found that this system had the effect of binding tire dealers to Michelin, restricting their effective choice of supplier. Crucial to this judgment was the fact that the reference period for the rebates was one year (i.e., if the customer achieved the sales target, it received a retroactive discount on its entire year's purchases from Michelin), which meant that even a small percentage reduction in the discount rate could signifi19. Id. at 33, para. 52. 20. Id. 21. Id. at 36, para. 61. 22. See NV Nederlandsche Banden Industrie Michelin v. Comm'n, Case 322/81, [1983] E.C.R. 3461 [hereinafter Michelin]. 23. Id. at 3461, para. 28. cantly affect the dealer's profit margin for the whole year. This pressured the dealer into .buying from Michelin: [A] ny system under which. discounts are granted according to the quantities sold during a relatively long reference period has the inherent effect, at the end of that period, of increasing pressure on the buyer to reach the purchase figure needed to obtain the discount or to avoid suffering the expected loss for the entire period.24 Another element that the Court cited was the fact that Michelin was much -larger than its main competitors (around 65% market share, compared to 8% for the next-largest supplier). In the Court's view, Michelin's sheer size in the relevant market made it effectively an essential trading partner for tire dealers, who were forced to do business with Michelin. Moreover, the level of the targets on the basis of which Michelin granted the rebates represented a significant proportion of each dealer's total annual requirements for tires. Given this, Michelin's competitors could not, by offering .discounts on their comparatively small sales volumes to the customer, equal the amount of the conditional rebate from Michelin that could be lost if the customer dealt with the competitor, to such an extent that it would lose the Michelin rebate. Thus, dealers were reluctant to deal with Michelin's competitors because Michelin's target rebates represented an important proportion of their total annual income, they were uncertain (dealers could not be sure of meeting them, even toward the end of the year),25 and the risk of not achieving a Michelin target outweighed any possible benefit that a smaller supplier might have offered by selling a comparatively small amount of product even at lower prices than Michelin offered. The system thus significantly restricted dealers' ability to choose among suppliers, particularly near the end of each annual reference period. Furthermore, the Court found that Michelin's target rebates were "calculated to prevent dealers from being able to select freely at any time in the light of the market situation the most favorable of the offers made by the various competitors."26 This confirmed the Commission's view that the system was 24. Id. at 3517, para. 81. 25. Id. at 3517, para. 82. 26. Id. at 3518, para. 85. "clearly aimed at tying the dealers closely to [Michelin] and thus making it difficult for other producers to gain a foothold in the market. '27 In the face of this evidence of exclusionary intent, the Court rejected Michelin's proffered justifications for the system. In 2001, a Michelin scheme with similar effects was condemned by the Commission in Michelin 11.28 Similarly, in Irish Sugar, the Court found the dominant supplier's target rebate system on the retail sugar market abusive because it foreclosed competitors. Irish Sugar's market share was above 85% and had been for almost a decade, and there was only one domestic competitor on the market.29 Irish Sugar had set the reference period of the challenged target rebate (six months) specifically to coincide with competitors' launch of new brands on the relevant market, evidencing its exclusionary intent. ° The rebates were also offered only to certain customers of these new-entrants and target purchase levels were fixed at a figure near to the customer's total requirements for the relevant product. 1 The Court found that these measures effectively foreclosed competitors from the market; in fact, one such competitor went out of business only months after Irish Sugar started implementing its rebates.12 Most recently, in Virgin/BritishAirways, the Commission likened the travel agent commission system employed by British Airways ("BA") to the target rebate system condemned by the Court of Justice in Michelin, as both were created with the intention of, and had the effect of, preventing firms from selling their products in competition with the dominant supplier. BA's travel agent commission system involved successively higher annual sales commission rates for travel agents selling BA tickets as they met various sales targets, with the targets defined as a percentage 2002] of that agent's sales of BA tickets in the previous year." The Commission focused on BA's significant market share (42% of the market at the date of introduction of the rebates, compared to 5.8% for the next-largest supplier and less than 4% each for all others). As a result of its much larger sales base, BA could offer travel agents large monetary rebates by giving relatively small percentage discounts on their total annual purchases from BA. By contrast, BA's competitors would have had to offer very large percentage rebates on their lower sales volumes in order to equal the rebate payments from BA.34 Because BA's target rebates represented an important proportion of their total annual income, the Commission said that travel agents were reluctant to deal with BA's competitors, since the risk of not achieving a BA target outweighed any possible incentive that a smaller airline might have created by making an attractive offer to sell additional flights. As in Michelin, therefore, travel agents were, according to the Commission, left with no realistic option as to the airline with which they dealt. The Commission also focused on BA's intent in implementing the system, concluding that BA had designed the rebates with the aim of foreclosing competitors: "[BA's rebates were] intended to eliminate or at least prevent the growth of competition to BA in the UK markets for air transport."35 The decision is currently on appeal before the Court of First Instance. B. Defenses A number of possible defenses or justifications for different prices or other conditions, which might be contrary to Article 82 are available. The Court of Justice in United Brands held that a dominant company is entitled to charge differential prices based on a number of factors. The factors that the Court expressly mentioned were differences in costs (e.g., transport costs, taxation, customs duties, the wages of the labor force, the differences in the parity of currencies), and competitive conditions (market33. See Commission Decision No. 2000/74/EC, O.J. L 30/1, at 2, para. 3 (2000) [hereinafter Virgin/British Airways Decision]. 34. Id. at 8, para. 30. The Commission calculated that a competitor would have to offer a discount of 17.4% on a travel agent's entire purchases from that airline in order to compete with a BA rebate of 0.5%. Id. 35. Id. at 23, para. 118. FORDHAM INTERNATIONAL LAW JOURNAL ing conditions, and the density of competition).36 The Commission has also accepted price reductions for "special customer status (e.g., [company] employees, affiliated companies, global and multinational accounts, educational and non-profit institutions, or government institutions").3 7 The list is not exhaustive but the principal defenses are set out below. 1. Volume-Based Discounts/Economies Of Scale The Community Courts and Commission have invariably found standard volume rebates (e.g., offering a 10% discount to all customers whose purchases exceed a certain threshold level) unobjectionable. This probably reflects a number of considerations. First, such a system is non-discriminatory in the sense of Article 82 (c), since it does not result in the application of dissimilar conditions to equivalent transactions. Second, in most cases, some cost savings probably result from serving larger customers.38 Finally, the commercial reality in most industries is that large customers expect to receive better supply terms than smaller customers. In Hoffmann-La Roche, the Court ofJustice held that quantity discounts linked to customers' purchasing volume would be permissible.39 It found, however, that, on the facts, the price advantages granted were not based on the differences in volumes bought from Roche, but were expressly conditioned on the supply of all or a very large proportion of a customer's total requirements by Roche.4" Similarly, in Irish Sugar, the Court accepted that Irish Sugar's border rebates in the retail sugar market would have been justified if they had been related to the purchasing volume of Irish Sugar's customers.4 1 In that case, however, the 2002] The Commission also seems to recognize these dangers, even if in practice they do not seem to have deterred it from treating prices above average variable cost as unlawful on the basis of circumstantial evidence of intent: The Commission emphasizes that it does not consider an intention even by a dominant firm to prevail over its rivals as unlawful. A dominant firm is entitled to compete on the merits. Nor does the Commission suggest that larger producers should be under an obligation to refrain from competing vigorously with smaller competitors or new entrants. 122 Finally, antitrust enforcement that depends on circumstantial evidence of intent as a necessary legal condition is haphazard and risks using hyperbole as a basis for intervention. Inflammatory documents are easily concealed and more easily taken out of context. c. Our Suggested Approach Although there appears to be some view that prices above average variable cost exceptionally may be anticompetitive, it is hard to see how a clear legal rule could capture the relatively small number of situations in which that occurs. Condemning above-cost pricing should be approached with considerable reserve, since price competition is almost always desirable and it is very difficult, if not impossible, to formulate a legal rule to distinguish between an above--cost low price that will eliminate a competitor and one which will not. It is notable that the proponents of the theory that above-cost pricing can be exclusionary have not been able to formulate a satisfactory legal rule to that effect. Baumol's idea that the dominant firm should be required to maintain the low price for a period of eighteen months or so seems unworkable and an overreaction to the actual scope of the underlying problem. Edlin's idea that the dominant firm should be prevented from responding with substantial price cuts or significant product enhancements until the entrant becomes "viable" suffers from a similar problem and from definitional issues. Neither thesis could form the basis of any sensible legal rule. In our view, the only exception to the rule that pricing above average variable cost is always legal concerns situations where there is clear evidence of a cumulative pattern of other exclusionary 122. ECS/AKZO Decision, O.J. L 374/1, at 21, para. 81 (1985). abuses in addition to low pricing. This exception would provide legal advisers with a workable framework. It is also consistent with the AKZO case, which states that prices above average variable cost but below average total cost may only be unlawful if they form "part of a plan to eliminate a competitor." 2. The Legal Rules: The Need For Recoupment Under EU Law One difference between EU and U.S. law on predatory pricing is that recoupment is not an element of the test in EU law. Some U.S. courts have held that, in addition to proving that the pricing complained of is below an appreciable measure of cost, there must be a reasonable prospect of recovering the losses incurred by the below-cost price through the dominant firm's subsequently raising prices. In Brooke Group, the Supreme Court laid out a two-prong test for recoupment. First, "below-cost pricing must be capable.., of producing the intended effects on the firm's rivals, whether driving them from the market, or ... causing them to raise their prices to supra-competitive levels." 123 Such a determination is to be based on factors such as "the extent and duration of the alleged predations, the relative financial strength of the predator and its intended victim, and their respective incentives and will. The inquiry is whether, given the aggregate losses caused by the below-cost pricing, the intended target would likely succumb."' 2 4 Second, it must be shown that "the predatory scheme alleged would cause a rise in prices above a competitive level that would be sufficient to compensate for the amounts expended on the predations, including the time value of the money invested in it.' 25 This determination will depend on "an estimate of the cost of the alleged predation and the close analysis of both the scheme alleged by the plaintiff and the structure and conditions of the relevant market."126 If either element of recoupment is not possible, the claim fails. Factors such as new entry, the ability to expand capacity, and whether the market is diffuse will influence to what extent recoupment is possible.' 27 123. Brooke Group Ltd., 509 U.S. at 225. 124. Id. 125. Id. 126. Id. at 226. 127. Id. 2002] Although not universally applied by the U.S. courts, the theoretical underpinnings of the recoupment test have much to commend them, even if, in practice, they constitute a considerable barrier to plaintiffs trying to establish a predatory pricing claim. The basic objection to predatory pricing implicitly assumes that the ability to recover losses will make short-term pricing cutting both anti-competitive and profitable because higher prices can be charged once market exit has been caused and rivals are deterred from entering again. If the price-cutting does not lead to some recovery by the dominant firm, this will in nearly all cases mean that rivals have been able to offset the effects of those price cuts or re-enter at a later stage, or that new entry is possible and will keep prices at competitive levels. Reaction in the EU to the need to have recoupment as a legal condition has been mixed, but generally hostile. As explained above, in Tetra Pak II, the Court of Justice rejected the notion of recoupment as necessary, at least in so far as the egregious circumstances of that case were concerned. Likewise, the UK Competition Authority, the OFT, takes the position that recoupment is relevant only if a dominant firm uses revenues from a dominated market in order to engage in predatory conduct in a non-dominated market, i.e., in cases of cross-subsidization.' 28 According to the OFT, if pricing below cost occurs in a dominated market, it can be assumed that the dominant firm can recoup losses afterwards. More recent judicial statements have been supportive of the need for recoupment as a necessary legal condition for predatory pricing. In CEWAL, Advocate General Fennelly stated that recoupment was implicit in the Court of Justice's statements in AKZO and Hoffman-La Roche and that some form of recoupment "should be part of the test for abusively low pricing by dominant undertakings.1 29 Indeed, in that case, the Advocate General stated that the sharing of losses resulting from the price-cutting among the CEWAL members was in essence a form of recoupment. This seems only partly correct. While loss-sharing may make it easier to share the cost of eliminating a competitor (the first limb of the recoupment test under U.S. law), it does not directly affect the possibility for supra-competitive prices to be 128. See OFT 414, supra n.47, at 16. 129. CompagnieMaritime Beige, E.C.R. 1-1419, para. 136. maintained thereafter (the second limb under U.S. law). That said, it seems reasonably clear from the CEWAL case that CEWAL's insistence on exclusive contracts, its 100% loyalty rebates, its reputation for taking significant retaliatory measures against its only competitor, and the high fixed costs that would have be incurred by a new entrant would have been sufficient to ensure that recoupment was feasible. Likewise, in AKZO and Tetra Pak II, the companies concerned were dominant in a wide range of products but only engaged in selective price-cutting for one product that a competitor offered. Given their dominance in a wide portfolio of products, it was probable that the threat of retaliation against a rival supplying only one of those products was a credible deterrent that would have allowed recoupment because other rivals would have been dissuaded from entering. In other words, all cases in which predatory pricing was found under Community competition law seemed primafacie capable of supporting probable recoupment. There are, however, reasons to treat a strict recoupment requirement with caution. First, it is often difficult to prove what the dominant company could do successfully at an unspecified time in the future. It would be necessary to show that there would be no entry by more competitive or more determined rivals, and that when the dominant company increased its price, it would not attract new entry. It would also be necessary to show that the price elasticity of the product was such that, although buyers were accustomed to low prices, they would be willing to pay significantly higher ones in the future. All of this suggests that the burden of proof is crucial. If the burden of proof was on the party alleging illegal low prices, it would make it difficult to bring a successful case. If the burden of proof was on the dominant company, it would be obliged to prove a negative, that is, to prove that it would be unable to recoup its losses if it tried to do so. Second, predatory pricing by a dominant company may have anticompetitive effects even if the dominant company does not or could not recoup its losses. The most effective form of predatory pricing is one where a company discourages market entry, or causes exit, by signaling to actual or potential competitors that their profitability in the market in question will be low as long as the dominant company is price leader in that market. This signaling would be more effective, and the effects of it would last longer, if the dominant company did not have to recover its losses, but held its prices only a little above competitive levels. This discouraging or signaling effect is particularly likely to be important if the dominant company is active on several markets, because predatory pricing on one market may discourage market entry on the others. This is particularly important in air transport, where predatory pricing, if it occurred on one route, would discourage entry on the other routes on which the dominant airline was operating. Finally, predatory pricing may have anticompetitive effects even if the rival is not forced out of the market, but instead decides to raise its prices to approximately the prices of the dominant company. In particular, in a concentrated market predatory pricing may demonstrate the dominant company's ability and willingness to retaliate against aggressive pricing by a competitor, and so may give rise to oligopolistic pricing. In such circumstances it would be extremely difficult to prove that recoupment had occurred, even if it had. In the circumstances, an express recoupment requirement might complicate the law further and impose an unfair burden on the party that bore that burden. Further, the absence of an express recoupment requirement does not, in our view, risk serious divergence between EU and U.S. law. In EU law, the prohibition on predatory pricing only applies to companies that are dominant. If a company is found dominant, that already suggests a high market share, a certain degree of immunity from normal competitive forces, and high barriers to entry in the relevant market concerned. An incumbent with a large share in such a market may be able to exclude new entrants effectively through below-cost price-cutting, particularly where there are large sunk costs or structural features of the market that require a minimum efficient entry or network effects. Thus, in many cases, the prevailing market conditions that contribute to dominance may also offer a good indication that rivals' exclusion will lead to anticompetitively high prices in the future, and that recoupment is therefore probable. 3. Defenses For Below-Cost Pricing The Commission's decisional practice and the Community 146 Courts' case law have often assumed that the only explanation for below-cost pricing is predation. This seems overly restrictive, since there are a number of legitimate pro-competitive reasons for short-term below-cost selling. Different defenses apply depending on whether the price charged is above or below average variable cost. a. Prices Below Average Variable Cost Where the price charged is below average variable cost, it usually will be safe to assume that it is predatory, since a dominant firm will normally have no interest in charging such a price in the medium to long-term. However, a few defenses might still be applicable. The first is the principle established in the General Motors judgment, which suggests that it may be a defense if the company genuinely did not know the facts that showed that its price was unlawful and corrected the price as soon as it found out 30° The second possible defense is that the dominant company is launching a product or service in a new market, and so inevitably the first sales, whatever the price charged, will not cover the average variable costs that are being incurred or have already been incurred. A defense on these lines must be permissible; otherwise dominant companies would be unable to enter new markets. The third possible defense is that the low price or free gift is a short-term promotion or trial offer made as a means of getting attention for the product or service in question. There is nevertheless an element of reasonableness for belowcost selling in the case of new markets or promotional offers: long-term or repeat promotional offers may be tantamount to predatory price-cutting. As the OFT states: A dominant undertaking which adopts a one-off short-term promotion [below average variable costs for a limited period] is unlikely to be found in contravention of the Chapter II prohibition [abuse of a dominant position]. However, a series of short term promotions could, taken together, amount to a predatory strategy. 131 In addition, the strategic objectives behind loss-leading are also important: price-cutting designed to exclude rivals' access to 130. See General Motors Continental NV v Conm'n, Case 26/75, [1975] E.C.R. 1367, 1380, paras. 20-21. 131. See OFT 414, supra n.47, at 12. 2002] DEFININGLEGITIMATE COMPETITION market "gateways" should be treated more harshly than loss-leading that is genuinely intended to allow consumers to test new products. A final defense might be that the product is being phased out. Some revenues are better than no revenues where fixed-cost recovery is concerned and the scope for exclusion in such circumstance is limited. Loss-leading raises more difficult issues. Loss-leading is practiced by companies selling a number of products, and is designed to attract buyers to the seller in the expectation that, once the buyer is on the seller's premises or committed to certain purchases anyway, the buyer will buy enough of other products to provide a profit greater than the loss on the product used as the loss leader. The most common example is probably food sold in a supermarket. Loss-leaders in these circumstances tend to involve different products on each occasion. But the same kind of issues are raised by a company which consistently sells capital equipment at a loss with a view to recovering the loss on subsequent sales of spare parts, consumables, or maintenance or repair services. In all these situations it seems that there is a valid defense if it is reasonable for the dominant company to expect that, as a result of the sale below cost, revenue will be obtained from other sales, which would not otherwise have been made, and that the expected or average additional revenue will exceed the amount of the loss. In other words, it is lawful to sell a system or combination of products in this way even if the initial sale is made at a loss (except, perhaps, where the same products are always sold at a loss and there is a rival which only produces that product). It would be useful to be able to show that competitors were able to use the same strategy if they wished, even if the competitors were clearly not subject to the obligations of dominant companies. However, if the competitors were not in a position to use the same strategy successfully, selling below cost in a market "gateway" might have serious exclusionary effects. This was the situation in the Napp Pharmaceuticalsjudgment of the UK Competition Appeal Tribunal discussed. above. Finally, the defense of meeting competition explained in detail in Part III above will also be applicable. Even dominant firms should be allowed to compete where there is a genuine price war with rivals. However, in the case of pricing below average variable cost, the dominant firm should be allowed to meet, but not undercut, the rival's price. Otherwise, the dominant firm could always put rivals out of business through predatory pricing by arguing that it was responding to a competitive offer. Moreover, unlike in the case of pricing above average variable cost, such a rule should not lead to price collusion between the dominant firm and its rivals: the dominant firm will not have any interest in agreeing on prices below average variable cost. b. Price Above Average Variable Cost But Below Average Total Cost If the price in question is above average variable costs but below average total costs, different questions arise and a wider range of defenses will be available. Under the second AKZO rule, selling at such a price is lawful unless there is evidence of exclusionary intent. There may be several explanations other than predation for pricing above variable costs but below average total cost: * As in the case of pricing below average variable cost, the dominant firm should be allowed to respond to competitive offers. However, unlike in the case of pricing below average variable cost, the dominant firm should be allowed not only to meet the rival's price but also to undercut it. Otherwise, there is a risk, for the reasons explained in Part III above, that the rival would benefit from a price floor and that the firms would be tempted to collude on pricing. " The sale is being made during a period of reduced demand in which no supplier of the product or service is able to sell at a price sufficient to cover its average total costs. A dominant company in such circumstances must be free to sell what it can at whatever prices it can obtain, for cash-flow reasons. Loss-minimizing is also the basis for the defense that goods are being sold below average total costs because they are obsolete, deteriorating, or would cost so much to store until they could be sold at a higher price that the loss would be minimized by immediate sale. The principle of loss-minimizing or profit-maximizing is probably the principle applicable in one difficult and controversial type of situation, in which the marginal or average variable cost of each additional sale is near to zero, but the effect is that competitors can sell their products only with difficulty or not at all. If, however, this is the 2002] effect of legitimate economies of scale or scope, a price at the average variable cost is not predatory, once the capital costs have been covered. Before they have been covered, the start-up defense may be applicable. Below cost pricing in these circumstances should not be regarded as predatory in the absence of some exclusionary intent. Even where there is exclusionary intent, it will be necessary to evaluate whether loss-minimizing or predation is the real reason for the low price. * It should be a defense to show that the price in question, although below average total costs, was a loss-minimizing price at the time it was charged. A dominant company with high capital costs may be obliged to sell at prices well below its total costs for a substantial initial period until it reaches a certain scale of operations, or a minimum number of customers in a network industry. • If product storage costs over the long-term would result in some selling below average total cost, it may be more rational to sell below average total cost immediately to make some saving. All these defenses have to be assessed in light of the facts as they were known to the dominant company at the time it made the below-cost sale. If the price chosen was a reasonable and rational price to minimize losses at that time, the defense is valid even if, with hindsight, it appears that another strategy would have been more profitable or would have had less exclusionary effect. Similarly, a loss-leader sale is lawful if the information available to the company showed that on average it was probable that the loss incurred would be recovered from sales of other products or services, which would not have been made without the below-cost sale. The same principle applies to yield management by airlines, which leads them to sell the last seat on a plane just before it takes off for a minimal price because if they did not sell it at that price, the seat would be empty. In the case of yield management, as in the case of other sales of systems or loss-leaders leading to sales of other products, the price is not predatory unless the company is likely to make a loss overall. Whether the above defenses should be available where the pricing is accompanied by other exclusionary practices will depend on the nature of those exclusionary practices and the defense claimed. If the other exclusionary practice is an exclusive purchasing commitment, it will not be relevant to say that the dominant firm was responding to a competitive offer. Similarly, if there is evidence of tying practices in addition to price-cutting, it will not usually be a justification to argue that the dominant firm was responding to competitive threats in the market for the tied product. In contrast, where goods are obsolete, deteriorating, or would incur substantial storage costs if unsold, the fact that the discounted price is coupled with an exclusive purchasing commitment for the quantities in question should not invalidate the defense of loss-minimizing. V. CROSS-SUBSIDIES - A SPECIFIC CASE OFPREDATORY PRICING A. Introductory Remarks Cross-subsidies are not mentioned in Article 82, and brief references in the case law of the Community Courts do not make clear when a cross-subsidy might be an abuse. 32 In general terms, cross-subsidization occurs where a company uses funds generated from one area of activity to fund activities in another area of its activity. A cross-subsidy may give rise to an antitrust problem if the dominant company has a monopoly or near-monopoly position in one market, and also has activities in another related market where it is in competition with competitors who sell only in the second market. Competitors in the second market have to meet all the costs necessary for their production for that market ("stand-alone costs"). The horizontally integrated dominant company however has several kinds of costs. It has "incremental costs" which arise only because of its operations in the competitive market, and which would cease if its operations in that market ceased. It also has or is likely to have costs which are common to its operations in both markets, but which would be unaffected by cessation of its activities in the competitive market. It also has costs which arise only because of its operations in the market in which it has a monopoly. The problem for the competitor is that the dominant company is able to spread its common costs over two sets of operations instead of only one in other words, it has economies of scale or scope. 2002] Various approaches have been used to detect cross-subsidies, but there are essentially two approaches that should be mentioned. First, if the monopolist's prices in the second market cover the incremental cost of producing products in that market, there is no cross-subsidy. A second approach is to calculate the stand-alone cost of producing each output separately from other outputs: if the price for each output covers those costs, there is no cross-subsidy. A number of regulatory issues are raised by cross-subsidies, particularly in the context of utilities and regulated markets, including the need for structural and accounting separation between reserved monopoly and competitive businesses. In the context of Article 82, it is assumed that cross-subsidy cases are in essence cases in which the abuse, if there is one, is predatory pricing.' Multi-product companies cross-subsidize all the time, whether or not they realize that they are doing so. There is no abuse of cross-subsidizing in the absence of predatory prices, since an above-cost price does not by definition need any subsidy. The important issue therefore under Article 82 is which costs must be taken into account in determining whether the lower price in the competitive market is predatory. Two different approaches are possible. The approach which has always been understood to be the correct one is to require the dominant enterprise to allocate its common costs, on some appropriate basis, between the monopoly market and the competitive market. This would not deprive it of the benefit of economies of scope or scale, but would ensure that in other respects it would be competing on an equal footing in the competitive market. Its economies of scope would give it a legitimate 133. In UPS Europe SA v. Comm'n (Deutsche Post AG, intervening), Case T-175/ 99, [2002] 5 C.M.L.R. 2, the Court of First Instance said that "the mere fact that an exclusive right is granted to an undertaking in order to guarantee that it provides a service of general economic interest does not preclude that undertaking from earning profits from the activities reserved to it or from extending its activities into non-reserved areas." Id. at para. 51. The Court went on (para. 55) to say that the use of profits from reserved activities to buy another company could raise problems where the funds used derived from excessive or discriminatory prices or from other unfair practices (that is, presumably, practices contrary to Article 82). In such a situation, it is necessary to look at the source of the funds to see if the purchase of the other company resulted from abuse of a dominant position. Cfi Criminal proceedings against Paul Corbeau, Case C320/92, [1993] E.C.R. 1-2533. competitive advantage, but it could not use its internal cost allocation to create a barrier to entry into the competitive market. The second approach is to essentially apply the AKZO rules to the competitive market only. If there is no evidence of a plan to eliminate a competitor, only average variable costs would be relevant; average total costs need not be calculated (and so there is little or no need to allocate common costs, because most common costs are fixed). In fact, if there are no incremental fixed costs, the AKZO average variable costs test and the pure incremental-costs-only test come to the same thing. However, the AKZO case was treated as involving only one product market, and the "cross-subsidy" which predatory prices always imply was within a single product market for a relatively short period. The AKZO case, in other words, is analogous to the case of the last seat on the airplane on a given flight, not to a situation in which a dominant company is operating in separate markets with some common costs. It would not therefore seem right to apply the AKZO test to a two-market situation and to apply it over a period of several years in circumstances where: (i) a large proportion of the dominant company's costs in the competitive market were common costs; and (ii) the dominant company's revenues in the competitive market were above its costs there only if that proportion was allocated exclusively to the uncompetitive market. It is obviously a weakness of the first AKZO test that it ignores all nonvariable costs altogether, in spite of the fact that both the dominant company and its competitor will normally have some costs of this kind. In such circumstances the internal cost allocation by the dominant company could create a barrier to entry into the competitive market. B. The Case Law The issue of cross-subsidization has not been raised directly in any case before the Community Courts. The issue was raised indirectly in Tetra Pak I.L 4 Tetra Pak was found to have committed a range of pricing and other abuses in two different but related markets; aseptic and non-aseptic machinery and cartons. Tetra Pak's market shares in the aseptic and non-aseptic markets were approximately 90% and 50%, respectively. There were also 134. See Tetra Pak II, O.J. L 72/1 (1992), on appeal Tetra Pak v. Comm'n, Case T83/91, [1994] E.C.R. 11-755. important associative links between these two markets. The Commission's case was that Tetra Pak had engaged in predatory pricing in relation to its Tetra Rex non-aseptic carton by pricing below average total cost. This finding assumed that Tetra Pak was able to incur losses in the non-aseptic sector by substantial profits made in the monopoly aseptic sector. Tetra Pak argued before the Community Courts that it had not engaged in crossfinancing from the aseptic to the non-aseptic sector. The Court of First Instance did not rule on this point, but simply noted that the "application of Article 86 [now Article' 82] of the Treaty... does not... depend on proof that there was cross-financing between the two sectors. " 135 The issue has now been considered directly in the Commission's decision in Deutsche Post.13 6 The case concerned a complaint brought by the international express parcel delivery company, UPS, against the incumbent German postal operator. The allegation was that Deutsche Post was using profits from its reserved monopoly in the reserved postal sector to cross-subsidize a loss-making business in the competitive parcel sector and to engage in predatory pricing. That complaint was upheld by the Commission, which found that the parcel service was operated at a substantial loss for several years. Without the cross-subsidies from the reserved area, Deutsche Post would not have been able to finance below-cost selling in the competitive parcel area for any length of time. The Commission prohibited Deutsche Post's sales below cost in the parcel area and ordered the structural separation of that business from the reserved area. The case had several unusual and significant features. First, Deutsche Post had a statutory monopoly, and also a legal duty to provide a universal postal service throughout Germany at standard postal rates. For this purpose it was obliged to maintain an infrastructure, which it was able to use both for its monopoly and its competitive services, but no part of which involved an incremental cost of providing the competitive service. Second, some of the incremental costs' of its competitive service were fixed costs. In other words, some of the infrastructure used in this 135. Tetra Pak, [1994] E.C.R. 11-389, para. 186. 136. See Deutsche Post, O.J. L 125/27 (2001). See also European Commission, Guidelines on the Application of the EEC Competition Rules in the Telecommunications Sector, O.J. C-233/2, at 17 (1991) (predatory behavior as a result of cross-subsidisation). service was distinct from the infrastructure which Deutsche Post needed and used for its universal service. Third, the Commission decided to use the concept of incremental costs in the competitive market. Fourth, Deutsche Post's incremental costs of providing the competitive service in 1990-1995 were above its revenue from that service: in other words, by this test its prices were predatory. In condemning Deutsche Post's predatory pricing on the basis of cross-subsidization, the Commission made a number of important points: * Cross-subsidization occurs where the earnings from a given service do not cover the incremental costs of providing that service and where there is another service or bundle of services, the earnings from which exceed the standalone costs. " The service for which revenue exceeds stand-alone cost is the source of the cross-subsidy and the service in which revenue does not cover the incremental costs is its destination. A profitable reserved monopoly is likely to be a permanent source of funding, i.e., overall revenues in the reserved area exceed its stand-alone costs. This means that, when establishing whether the incremental costs incurred in providing a service in the competitive sector are covered, the additional costs of producing that service, incurred solely as a result of providing the service, must be distinguished from the common fixed costs, which are not incurred solely as a result of this service. * The Commission decided that infrastructure that had to be maintained in order to provide the universal service could not be incremental (which was clearly correct). The Commission therefore decided that the cost of providing and maintaining the infrastructure needed for the universal service need not be divided between the monopoly and the competitive service, but could be attributed solely to the monopoly in calculating Deutsche Post's costs in the competitive market. The Commission did not take its analysis further because Deutsche Post's activities in the competitive market were still predatory. C. Comments 1. Allocating Costs Between Two Operations After much controversy in various industries, the Deutsche Post decision shows that in these rather unusual circumstances the dominant company is not required to allocate its common costs, on any basis, between its two kinds of operations. In other words, it may legally allocate all its common costs to its monopoly operations, even if they also benefit its competitive activities. If it does this, its costs in the competitive market will be only incremental costs, and these will be less than the stand-alone costs of its competitor (all other things being equal). How much less will depend on the extent of the economies of scope or synergies between the dominant company's two sets of operations. Presumably the incremental-costs-only approach would have been seen to be wrong if Deutsche Post had used every post office for rent-free banking, insurance, and travel agency activities. Merely concluding that even by this standard there was predatory pricing may not have been enough, since if common costs had to be allocated the losses would have been much greater, and competitors' claims for compensation correspondingly larger; the quantum of predation is often important. The Commission's conclusion that all the common costs could be attributed to Deutsche Post's monopoly activities was due to its universal service obligation. In other words, the same conclusion would arguably not be reached in the case of a dominant enterprise with no universal service obligation, whether or not it has a statutory monopoly. In the unusual case in which it has been used, this approach means that the problems of correctly allocating common costs between the two sets of activities does not arise, and the question of whether the dominant company's prices in the competitive market are predatory depends only on whether they are below its incremental costs in that market. The fact that the competitor in that market has not got the advantage of economies of scope may be a difficulty, a handicap or a barrier to entry, but it is not one created or aggravated by the dominant company (except in price squeeze cases). Indeed, even if the dominant company was obliged to allocate its common costs in some appropriate way between its two kinds of operations, the total costs attributable to its operations in the competitive market might still be below the stand-alone costs of its competitor, since the dominant company would have economies of scope and the competitor would not. Whether any such approach discourages a dominant company from entering a competitive market does not depend on the merits of the Deutsche Post decision. It depends, in essence, on whether the principle is subject to an exception of the kind always assumed to apply in predation cases, which accepts that a company's initial activities in a new market are likely to be unprofitable. (This question did not arise in Deutsche Post, as the company had lost money for five years, and had been in the competitive market before that.) Provided that an exception of this kind exists, the Commission's approach does not seem to discourage legitimate competition by dominant companies. 137 In the absence of a universal service obligation, however, the incremental-costs-only approach would be too favorable to the dominant company because it would create or legitimize a barrier to entry into the competitive market. In the absence of any objective criterion such as the universal service obligation, the dominant company would have too much freedom to decide which of its costs in the competitive market were incremental and which were not. A rule should not be adopted if its application would lie essentially within the discretion of the company to be bound by the rule. The dominant company's incremental costs (because they can include some fixed costs) are likely to be higher than the average variable costs of its operations in the competitive market. If it is accepted that the incremental-costsonly approach is not appropriate in the absence of a universal service obligation (or the equivalent for some other reason), then it seems that the right approach would be to require allocation of common costs on some appropriate basis. What the best basis should be will depend on the circumstances. 2. Which Costs Do You Allocate? It is said that neither economics nor cost-accounting impose 137. See Mats A. Bergman, A Prohibition Against Losses? The Commission's Deutsche Post Decision, 2001 EUR. COMPETITION. L. Rv. 351, 351-54 (2001). Bergman suggests that the decision "will curb competition on the merits. Taken to the extreme, this standard can be interpreted as a prohibition for a firm that holds a monopoly in one market to show red figures in competitive markets." Id. at 351. This seems incorrect, for the reason stated in the text one single "correct" method of cost allocation. This is no doubt true, but it does not follow that allocation, on some consistent and reasonable basis, is unnecessary or impossible (although it may be difficult), or that no common costs need to be attributed to the competitive activities. Unless there is a universal service obligation, the dominant company should allocate or apportion costs and can still get the benefit of the economies of scale or scope to which it is entitled. If, on any reasonable cost allocation method, its prices are above its costs, then the details of the cost allocation method are unimportant. This means that it is much easier for a complainant or regulator to argue a cross-subsidy case than a price squeeze case. It has been pointed out that the most obvious basis for cost allocation, in proportion to turnover in the two sectors involved, is defective if the lower price is predatory because it results from an unlawful cross-subsidy. If this was the position, the cost allocation might have to be recalculated using a corrected turnover in the lower price sector. It might be important e.g., for the rights of injured competitors to claim compensation, to calculate the extent of predation accurately. However, cost allocation between malleable and related interoperable products in the software industry, where the marginal cost of selling another copy of an existing product is in any case almost zero, cannot solve the antitrust issues that arise. There may be a real problem if the competitive market is not really an independent market, but is always and necessarily merely a by-product of the market in which the dominant company is dominant (that is, the competitive market is uneconomic except in combination with the other market). If this is the situation as a result of the inherent economics of the two markets, then the incremental-costs-only approach might be right, because the barrier to entry into the competitive market is due to the competitor's underlying need to enter the main market and not to the dominant company's cost allocation. But in this situation it would be necessary to prove objectively that independent activities in the competitive market were inherently uneconomic, and were not uneconomic for competitors only because of predatory pricing by the dominant company. The fact that the dominant company has a statutory monopoly does not seem relevant to the question of abuse. It is the statutory duties, if any, of a dominant company which justify the incremental-costs-only approach, not its statutory privileges. Whether it is legally impossible for a competitor to enter the monopoly market, and not merely very difficult or impossible for other reasons, should not alter the rule on the definition of abuse. If there is no statutory monopoly in a given market, entry by new competitors is always possible in theory, however unlikely, difficult or uneconomic it may be in fact. From an economic standpoint, an absolute legal impossibility is not very different from a near-impossibility for other reasons. Finally, it is worth mentioning another, different, principle. The Community Courts have accepted that a statutory postal monopoly may be necessary because some services are inevitably loss-making and therefore require cross-subsidization so that the postal service can break even overall. However, they have made it clear that the monopoly is only justified if all the services in question are necessarily part of the core activities and no less restrictive solution can be found."' 8 If they are distinct services, there is no justification for the monopoly applying to them, and the normal rules about cross-subsidies, whatever they are, would apply. This principle also confirms that a postal monopoly and related competitive services should be considered separately, and that neither should be allowed or required to cross-subsidize the other. CONCLUSION From the foregoing, it is apparent that much of Community competition law on pricing practices is concerned with the form of the pricing measure rather than its economic effect and implications for consumer welfare. Certain Commission statements also seem to create per se rules, or at least strong presumptions of illegality, against certain pricing practices. Much of Community competition law in this area proceeds from the wrong premise. Lower prices should in nearly all cases benefit from a strong presumption of legality. The situation should only be different 138. See Corbeau, [1993] E.C.R. 1-2533; see also European Commission, Notice from the Commission on the application of the competition rules to the postal sector and on the assessment of certain state measures relating to postal services, O.J. C-39/2, at 10 (1998). Deutsche Post's overall profits were substantial, and this calls into question either the justification for its reserved monopoly or the effectiveness of the regulatory authority. See also Ufex v. Comm'n, Case T-613/97, [2000], on appealCase C-94/01, La Poste. where it is clear that, in the specific context of the market under consideration, they distort competition in some material way between customers or create a handicap for competitors that is not merely the result of the dominant company's offering a lower price. In our view, the Commission should clarify its position on pricing practices in a way that properly reflects current economic thinking and its recently-stated objectives of safeguarding consumer welfare rather than the position of competitors. In light of the wording of Article 82, there are strong arguments for saying that only the following pricing practices by dominant companies are unlawful under Article 82: PRINCIPLE 1. Charging a lower price on condition that the customer buys exclusively from the dominant company will generally be contrary to Article 82(b). Despite some statements to the contrary, there is no rule of Community competition law which prohibits a dominant company from agreeing to give a price reduction on condition that the customer buys at least a specified quantity during a specific period, or gives a reduced price for a quantity which both parties believe is likely to be the buyer's total requirements during the period. A basic distinction should be made between: (i) discounts that are generally available; and (ii) discounts that are individually negotiated with the customer in question. If the quantity or target is not individually negotiated with each buyer, it should normally be unobjectionable, since it cannot (except by coincidence) correspond to the buyer's total requirements. Even where the discount is individually negotiated with the customer, it is still necessary to further distinguish several situations: 139 * A dominant company should be free to grant a quantity rebate for all of a customer's purchases if that rebate is not conditional on that customer's not buying from third parties. This is simply an unconditional price reduction. " A dominant company can grant a discount on additional purchases made by a customer above a certain 139. These situations assume that none of the discounts would lead to predatory pricing. quantity; that is simply a pro-competitive price reduction, and the buyer would be free to buy the additional quantity from any other supplier which offered a more favorable price for that quantity. A price reduction given on a buyer's purchases from a dominant supplier in a given period, which is granted only on the condition that the total quantity exceeds a target figure should usually be regarded as pro-competitive. However, issues may arise in the context of target or loyalty rebates if a price reduction that is applicable for all or most of a customer's purchases from the dominant company for a certain period is likely to be lost if the customer buys from another source. The antitrust objection is that rebates of this kind can create strong marginal incentives for the customer not to make the additional purchases from a company other than the dominant firm so that, in practice, they amount to exclusive or near-exclusive purchasing commitments. Whether the target rebates gives rise to appreciable foreclosure is highly fact-specific and requires consideration of a series of factors, including: (i) the size of the discount; (ii) whether the discount increases in a linear manner, or in steps, and if so what the quantity is for each step; (iii) the structure of market demand and whether the rebate is share-stealing or market-growing in effect (i.e., barriers to entry to the emergence of new buyers); (iv) the length of the reference period; (v) whether the dominant firm is an unavoidable trading partner; and (vi) the proportion of the total market subject to the price reduction. AVAILABLE DEFENSES. In the case of non-predatory discounting practices, the unlawful practice is not the lower price but the conditions on which it is available. This means that it usually will not be a defense to show that the lower price was matching or undercutting a competitor's price. However, several other defenses will be available. In cases of exclusive buying contracts contrary to Article 82(b), or where the conditions of sale impose a loss or penalty if the buyer buys from another supplier, it should be a defense that the seller was making a substantial investment in order to supply the buyer, and this investment would be economic only if the buyer bought exclusively from the dominant company. For other rebates, the following defenses will usually be available: (1) volume discounts/economies of scale; (2) reduced costs; (3) payment for services provided by the buyer; and (4) the lower price is a pro-competitive measure to help a class of companies, e.g., those starting up or making substantial investments. There is no requirement that the rebate in question should be justified by precise cost savings. PRINCIPLE 2. Charging different prices in similar transactions to customers that compete with each other is only contrary to Article 82(c) if the price difference is so large that it creates a significant competitive disadvantage for the customers paying the higher price. There is otherwise no general prohibition on a dominant company's charging different prices for the same product. (Discrimination by a vertically-integrated dominant company against competitors in a downstream market that rely on the dominant company for some input is subject to different and stricter obligations than those applicable under Article 82(c).) However, the situations',in which a dominant company would have the ability or incentive to charge different prices for the same product to such an extent that they give rise to distortions of competition between end-users will in practice be rare. This means that Article 82(c) will mainly be relevant where State-owned or controlled enterprises discriminate against companies from other Member States. Article 82(c) is therefore much narrower in scope than the (muchcriticized) price-discrimination provisions of the U.S. Robinson-Patman Act; in addition, there are defenses under Article 82(c) that are excluded by that Act. AVALABLE DEFENSES. As in the case of rebate practices, the following defenses are generally valid: (1) quantity rebates; (2) meeting or undercutting a competitor's price; (3) reduced costs; (4) payment for services provided by the buyer; and (5) start-up prices. PRINCIPLE 3. Charging a price (whether selective or not) which is below the dominant company's average variable cost of selling the kind of product or service in question will usually be contrary to Article 82(b) unless a defense is available. AVAILABLE DEFENSES. In cases of prices below average total cost but above average variable cost, it should be a defense that the price was loss-minimizing or profit-maximizing in the circumstances at the time. The defense of meeting competition will also be available and the dominant firm should be allowed to meet and undercut a rival's offer, as long as the price remains above average variable cost. 11. See Suker Unie, [1975] E.C.R. 2001 - 04 , paras. 502 - 26 . 12. See Hoffman-La Roche & Co. AG v. Comm'n, Case 85 /76, [1979] E.C.R. 461 , at 540, para. 89 . 13. Id . at 467, para. 24. The Commission reached a similar conclusion in its decision . See Hoffman-La Roche Decision , O.J. L 258, at 514, paras. 22 and 24 ( 1997 ). 14. Id . at 540, para. 90 . 15. Irish Sugar plc v . Comm'n, Case T-228/97 , [1999] E.C.R. 11 - 2969 , at 11-3049, para. 194. See also Commission Decision No. 97 /624/EC, O.J. L 258/1 ( 1997 ) [hereinafter Irish Sugar Decision]. 16. Irish Sugar , [1999] E.C.R. 11 - 3051 , para. 198 . 17. Id . 18. See Commission Decision No. 91 /299/EEC, O.J. L 152/21, at 24, para. 16 ( 1991 ) ("Soda-ash/Solvay Decision"). 27. Commission Decision No. 81 /969/EEC, O.J. L 353/33, para. 38 ( 1981 ) [hereinafter Michelin I Decision] . 28. See Commission Decision No. 2002 /405/EEC, O.J. L 143/1 ( 2002 ) [hereinafter Michelin II Decision] . 29. Irish Sugar , [1999] E.C.R. 11 - 2984 , para. 23 . 30. Id . at 11-3052-53, para. 203. See also Irish Sugar Decision, OJ . L 258/1, at 29, para. 152 ( 1997 ). 31. Id . at 11-3052-53, para. 203. See also Irish Sugar Decision, OJ . L 258/1, at 29, para. 154 ( 1997 ). 32. Irish Sugar , [1999] E.C.R. 11 - 3059 , para. 222 . 36. See United Brands Co. and United Brands Continental B.V. v . Comm'n, Case 27 /76, [1978] E.C.R. 207 . 37. Maurits Dolmans & Vincent Picketing , The 1997 DigitalUndertaking, EUR . COMPETITION L. REv . 1998 , 19 ( 2 ), 108 - 15 , at 113. See also id. at 114 (regarding customized or non-standard product offerings). 38. This consideration cannot, however, be the sole justification for the positive treatment of volume discounts. As Ridyard explains, "there is almost no plausible cost function that would make such a discount scheme 'cost-related' in the sense that the differences in price were explained by differences in the costs of supply." Ridyard, supra n.4 , at 289. 39. See Hoffmann-La Roche , [1979] E.C.R. 461 . 40. Id . at 521, para. 22 . 41. Irish Sugar , [1999] E.C.R. 11 - 3042 , para. 173 . 132. See e.g. Merci Convenzionali Porto di Genova SpA v . Siderugica Gabrielli SpA, Case C- 179 /90, [1991] E.C. R. 1- 5889 , para. 19 .


This is a preview of a remote PDF: https://ir.lawnet.fordham.edu/cgi/viewcontent.cgi?article=1866&context=ilj

John Temple Lang, Robert O'Donoghue. Defining Legitimate Competition: How to Clarify Pricing Abuses Under Article 82 EC, Fordham International Law Journal, 2002,