Single-Firm Conduct Hearings Comment: International Chamber Of Commerce (William MacLeod)

This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.


International Chamber of Commerce
The world business organization

Single-Firm Conduct As Related to Competition

Prepared by the Commission on Competition

Submission to the US Federal Trade Commission and Department of Justice Hearings on Section 2 (project no P062106) of the Sherman Act

ICC, the International Chamber of Commerce, and the USCIB, United States Council for International Business, welcome the opportunity to submit comments to Federal Trade Commission and Department of Justice Hearings on Section 2 of the Sherman Act regarding certain antitrust issues that have been identified for potential study.

As world business organizations with members from all sectors in over 130 countries, ICC and USCIB are able to draw on a rich range of perspectives from different sectors on these issues. With their long history of interest in the issues below, it is hoped that the following comments may assist the hearings to identify priority areas for further work. The issues identified and questions presented in the Federal Register notice have been previously addressed in recent comments submitted by ICC and the USCIB, specifically in the International Chamber of Commerce, Comments on the Reform of the Application of Article 82 of the EC Treaty1; the ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses2; the International Chamber of Commerce, Comments on Selected Issues for Study by the U.S. Antitrust Modernization Commission;3 and the United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses.4 Rather than repeat this material in its entirety, ICC and USCIB direct the staff to them generally, and here include only highlights addressing particular issues and questions presented in the Federal Register notice.

  1. Bundled Loyalty Discounts and Market Share Discounts:

  1. How should the structure of the market and the market shares of participants be taken into account in analyzing such conduct?
  2. There is no basis for the premise that all rebate systems established by a dominant undertaking are abusive unless they are cost-justified. International Chamber of Commerce, Comments on the Reform of the Application of Article 82 of the EC Treaty, Document No. 225/623 (12 December 2005) at 15.

    ICC has problems understanding why it would be abusive for dominant undertakings to try to 'maintain or strengthen' market shares through the adoption of rebates... Competition is generally about increasing market shares to the detriment of competitors... [D]ominant undertakings should be allowed to compete aggressively on rebates since this may lead to long-term aggressive price competition. Rebates would become abusive in limited defined circumstances when competitive strategies are not 'on the merits' or involve predatory or other anticompetitive behaviour resulting in likely foreclosure effects, for example, "full line forcing". International Chamber of Commerce, ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, Document No. 225/627 (7 April 2006) at 14-15.

  3. What are the likely procompetitive and anticompetitive effects of the conduct in the sort and long term?
  4. Any kind of fidelity rebate can have a pro-competitive role in the sense that it creates further dimension of competition (the non-linear price schedule) and it can represent a more aggressive pricing strategy: hence an additional minimal condition for rebates to be abusive should be that competitors are not able to propose similar rebates or different ones (with different thresholds),... International Chamber of Commerce, ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, Document No. 225/627 (7 April 2006) at 16.

  5. What types of cost savings, risk reduction or other efficiencies could be generated by such conduct?
  6. [R]ebates increase allocative efficiency and consumer welfare by increasing output and reducing prices. They are often preferred by customers to alternative arrangements and are often the result of hard bargaining by customers to get the best price from undertakings that, because they are dominant, would otherwise charge higher prices. International Chamber of Commerce, Comments on the Reform of the Application of Article 82 of the EC Treaty, Document No. 225/623 (12 December 2005) at 13.

    Loyalty rebates are particularly important to a pure-play innovation company and ultimately for end-consumers. Consumers may benefit, for example, from a manufacturer having a low marginal input cost, when that low cost is passed on to the consumers. This in turn will provide the manufacturer with an incentive to expand sales by competing on price.

    Additionally, loyalty rebates may facilitate efficient recovery of fixed costs. In general, consumers will face higher prices where an innovator needs to charge higher prices - resulting in lower volume - in order to recover fixed research and development costs. A loyalty rebate scheme allows the innovator to charge a relatively high price for the non-contestable share of the market, where demand is relatively inelastic, while charging a lower price (after loyalty rebates) for the contestable part of the market, where demand elasticity is higher. The company can simultaneously profit from a higher margin on the infra-marginal units without losing volume at the margins. International Chamber of Commerce, ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, Document No. 225/627 (7 April 2006) at 18.

  7. How might competitors respond to counteract a loss of sales to the firm engaging in such conduct, and would that result in harm to consumers?
  8. First, 'distorting' effect on competitors docs not necessarily mean 'abusive'. Findings of abuse should be based on a longer-term market assessment that should take into account competitors' likely response to the rebate system, customers' ability to switch and long term benefits for end-users. International Chamber of Commerce, ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, Document No. 225/627 {7 April 2006) at 14.

  9. What tests and standards should court and enforcement agencies use in assessing whether such conduct violates Section 2?
  10. Increases in allocative efficiency and consumer welfare ought to be regarded as objective justifications for rebates and should negate the assumption that such rebates are exclusionary by their very nature. International Chamber of Commerce, Comments on the Reform of the Application of Article 82 of the EC Treaty, Document No. 225/623 (12 December 2005) at 14.

    It is critical that the regulation of these practices focus on their effects on the welfare of customers in the market. The ability of rivals to match a dominant firm's discounts is at best ambiguous evidence on the desirability of the practice. Inadequate attention to demonstrable competitive effects could create law that preserves inefficient competitors while sacrificing competition. Remedial relief is not warranted when smaller competitors have difficulty competing against a dominant player that is more efficient. Efficiencies should always be relevant in unilateral conduct cases -- whether preferred in defense of the practice or implicated in the proposed remedy. Challengers of aggressive discounting should bear a heavy burden to show that intervention in the marketplace would produce remedies that benefit consumers without imposing costs that consumers will bear. International Chamber of Commerce, Comments on Selected Issues for Study by the U.S. Antitrust Modernization Commission, Document No. 225/621 (1 September 2005) at 17.

  1. Product Tying and Bundling

  1. How should the structure of the market and the market shares of participants be taken into account in analyzing such conduct?
  2. Consideration of the effects of a practice should focus on the maintenance or enrichment of market power, not structural dominance. When enforcement agencies or tribunals protect competition through the regulation of single-firm conduct, they must engage in the task of distinguishing between firms that achieve or maintain a dominant position through legitimate means and those that have done so through means that hinder the competitive process. Dominant firms often employ behavior that combines valuable innovation with aggressive marketing. International Chamber of Commerce, Comments on Selected issues far Study by the U.S. Antitrust Modernization Commission, Document No. 225/621 (1 September 2005) at V7.

  3. What are the likely procompetitive and anticompetitive effects of the conduct in the sort and long term?
  4. Market-leading companies should be able to continue producing innovative combinations of products benefiting consumers without running afoul of the prohibitions on tying unless the competition authority can rebut the innovating firm's prima facie case of efficiency gains. When companies combine formerly separate products, consumer welfare is usually increased as firms realize the efficiencies involved. These efficiencies may be the result of greater product functionality or the elimination of double marginalization, or simple convenience. Such tying or bundling may also lead to system-based competition, which may create an even more innovative and competitive market than component-based systems, as the markets for computer systems, home theaters, and cell phones aptly demonstrate. United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 23.

  5. What types of cost savings, risk reduction or other efficiencies could be generated by such conduct?
  6. [B]undling is a valuable strategy to gain broader distribution of the products or service that is subject to network effects. And the broader the distribution, the greater the value produced for all consumers. This is particularly true when the product or service in question has a low (or no) marginal costs, because the supplier can costlessly include the product or service in bundles with other products... Similarly, we believe that it should be acknowledged that bundling can generate efficiencies in multi-sided markets, i.e. markets where products or service must be matched with other products or service to have value... The complex business models resulting from multi-sided markets often require bundling practices because the consumption on one side of the market is being "sold" on the other side of the market, and piece-meal consumption on one side of the market breaks down the interdependent ecosystem. International Chamber of Commerce, ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, Document No. 225/627 (7 April 2006) at 24.

  7. Would a business typically analyze or estimate the likely cost savings from this type of conduct before or after engaging in it?
  8. Because the harm over-enforcement can cause to consumer welfare is significant in this area, the ideal test is one that greatly reduces the risk of enforcement by being administrable by competition authorities while being easily and predictably applied by businesses. It would create a safe harbour for which a business can qualify using its own readily available data, thus not diminishing the effects of efficient conduct as a result of compliance costs. United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 24-25.

  9. How might competitors respond to counteract a loss of sales to the firm engaging in such conduct, and would that result in harm to consumers?
  10. It is critical that the regulation of these practices focus on their effects on the welfare of customers in the market. The ability of rivals to match a dominant firm's discounts is at best ambiguous evidence on the desirability of the practice. International Chamber of Commerce, Comments on Selected Issues for Study by the U.S. Antitrust Modernization Commission, Document No. 225/621 (1 September 2005) at 17.

    [T]he fact that other undertakings in the market also offer bundles is a presumption that bundling generates efficiencies and meets consumer demand - if not, bundling by the dominant undertaking would provide competitors with a great opportunity to differentiate their offerings and make them more attractive to consumers. Additionally, the dominant undertaking ought to be able to compete with bundles offered by its competitors. International Chamber of Commerce, ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, Document No. 225/627 (7 April 2006) at 22.

    If there were sufficient customer demand to make the supply of the unbundled product profitable, competitors of the dominant undertaking would most likely avail themselves of this business opportunity. Id.

  11. What tests and standards should court and enforcement agencies use in assessing whether such conduct violates Section 2?
  12. [A] safe harbour based upon analysis of whether "the incremental price that customers pay for each of the dominant company's products in the bundle [covers] the long-run incremental costs of the dominant company of including th[e] product in the bundle." Assuming that this safe harbour is sufficient, then for mixed-bundle discounts or rebates that fall outside the safe harbour, the Commission should then continue the analysis by demonstrating (1) a likelihood of recoupment and (2) a likelihood of the creation of substantial market power in the relevant market for the "bundled" product in order to show that discounting through mixed bundling constitutes an abuse of dominance. Absent such a showing, mere exclusion of a competitor should not be found sufficient to establish a finding of anticompetitive bundling. United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 23.

    The fundamental inquiry when analyzing a tying arrangement should be whether competition is threatened by the practice in question. ICC members believe that this assessment is best accomplished by analyzing tying arrangements under the rule of reason. International Chamber of Commerce, Comments on Selected Issues for Study by the U.S. Antitrust Modernization Commission, Document No. 225/621 (1 September 2005) at 15.

    It can be difficult to distinguish between anticompetitive acts and vigorous competition. This is particularly true when the alleged anticompetitive act is the offering of lower prices to customers - practices such as aggressive discounting, attractive rebates, and various loyalty programs. The law should continue to demand that companies challenging such practices demonstrate anticompetitive consequences, because these types of cases may discourage practices that provide significant net benefits. International Chamber of Commerce, Comments on Selected Issues for Study by the U.S. Antitrust Modernization Commission, Document No. 225/621 (1 September 2005) at 16.

    Inadequate attention to demonstrable competitive effects could create law that preserves inefficient competitors while sacrificing competition. Remedial relief is not warranted when smaller competitors have difficulty competing against a dominant player that is more efficient. Efficiencies should always be relevant in unilateral conduct cases - whether proferred in defense of the practice or implicated in the proposed remedy. Challengers of aggressive discounting should bear a heavy burden to show that intervention in the marketplace would produce remedies that benefit consumers without imposing costs that consumers will bear. International Chamber of Commerce, Comments on Selected Issues for Study by the U.S. Antitrust Modernization Commission, Document No. 225/621 (1 September 2005) at 17.

    [T]he distinct products test itself may not be helpful for understanding market dynamics because, by definition, this test is backward-looking... A better approach in these cases would be simply to ask whether the undertaking integrating the previously distinct products can make a plausible showing of efficiency gains. Since technical tying is normally efficient, market-leading undertakings would be able to continue producing innovative products benefiting consumers without running afoul of the prohibitions on tying. Id.

    We believe that the long-run incremental costs standard is inconsistent with business reality because it requires companies to price bundles to cover sunk fixed costs that are unrecoverable. This approach ignores the economic reality that, when businesses decide how to price a product, they do not consider costs that are "sunk" or "unrecoverable," even if not a single product is sold... We believe that a more appropriate cost standard in this case would be marginal costs ("MC") or at least Average Avoidable Costs ("AAC"). When business people decide whether or not to make a marginal sale at a particular price, they generally consider the marginal cost of making that sale. Id. at 23.

  1. Predatory Pricing

  1. How should the structure of the market and the market shares of participants be taken into account in analyzing such conduct?
  2. Dominance itself should not be sufficient to establish the likelihood of recoupment, particularly in technology markets. For example, looking forward one or two years in the dominance inquiry is not sufficient to undertake a proper assessment of recoupment where significant uncertainty abounds regarding not only cost and demand but the existence of potential entrants. It is entirely possible that a firm may be dominant in the sale and/or distribution of a given product, yet be constrained by entrants with highly disruptive technologies which require greater than one or two years to mature and be successfully commercialized. International Chamber of Commerce, ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, Document No. 225/627 (7 April 2006) at 13.

  3. What tests and standards should court and enforcement agencies use in assessing whether such conduct violates Section 2?
  4. First, pricing at or above average total cost (ATC) should not provide a basis for a claim of predatory pricing. United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 21.

    Pricing above ATC ["average total cost"] is in general not considered predatory, but according to the virtually unanimous economic literature, it would be better to state explicitly that pricing above ATC is never predatory since it cannot lead to foreclosure of 'as efficient' competitors. International Chamber of Commerce, ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, Document No. 225/627 (7 April 2006) at 12.

    [T]here are not economic justifications for a change of standard from AVC ["average variable cost] to LAIC ["long-run average incremental cost"]. Moreover, we believe that the LAIC standard is inconsistent with business reality because it requires companies to price to cover average sunk fixed costs that are unrecoverable: this approach ignores the economic reality that, when businesses decide how to price a product, they do not consider costs that are "sunk" or "unrecoverable," even if not a single product is sold. Id. at 13.

    Recoupment should be a critical element of any predatory pricing claim, since consumers will benefit overall from lower prices unless the firm engaging in below cost pricing is able to recoup all of its losses on a net present value basis. It is therefore not sufficient to presume a "likelihood of recoupment" from the fact that a firm holds a dominant position and, consequently, that there are likely to be barriers to entry into the relevant market. The existence of barriers to entry is necessary for the dominant firm to recoup its losses but is not sufficient to establish that recoupment would occur. The recoupment assessment should take into account the magnitude of the likely losses, the level of increased prices following foreclosure and the period of time during which those prices would need to be charged, the time value of money, and the prospects for innovation affecting the ability to recoup as well as the prospects for entry prior to recoupment of the losses on a NPV basis. United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 21-22.

  1. Refusals to Deal

  1. How should the structure of the market and the market shares of participants be taken into account in analyzing such conduct?
  2. In a market system of free competition, even a dominant company must, at least in principle, be allowed to freely decide upon its sales strategy and distribution system. If it decides to change its distribution policy, e.g., to terminate existing distribution contracts and to establish a direct sales organization, it is its own choice for which it bears responsibility. Competitionlaw is not meant to guarantee an existing distributor relationship once and for all. As long as the supplier does not act in order to discipline a specific distributor and as long as the necessary termination periods are observed (depending on the given set of facts, the length may vary), there is no reason to intervene... Therefore, it should not "fall upon the dominant company to show that consumers are better off with the supply relationship terminated"... If there be a presumption at all, it should be in favour of the company's freedom to decide upon its distribution strategy. Only in the case where the terminated dealer can show that he was disciplined or discriminated, the supplier might be required to justify the termination. International Chamber of Commerce, ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses, Document No. 225/627 (7 April 2006) at 26.

    It is also widely recognized that forcing dominant firms to grant access to their inputs can deter innovation, both by discouraging dominant firms from investing in innovation in the first instance, and by encouraging smaller rivals not to innovate but instead to "free ride" on the innovations of others.5 The United States Supreme Court, echoing these principles, recently observed that compelling firms who have established an advantage "to share the source of their advantage is in some tension with the underlying purpose of competition law, since it may lessen the incentive for the monopolist, the rival, or both to invest" in ways that promote consumer surplus.6 United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 25.

    It is a well established principle that the rights of intellectual property holders are to be respected in all but most exceptional circumstances. In fact, there is no economic reason why cases involving intellectual property rights should be treated any differently than any other case involving a refusal to deal.7 The purpose of intellectual property law in the first instance is to provide businesses an incentive to invest in research and development activities aimed at generating new products and services. Thus, intellectual property rights are of vital importance to promoting consumer welfare. The adoption of rules and standards that create uncertainty as to when a company may be required to license its intellectual property will have a chilling effect on investment in research and development, to everyone's detriment. This is particularly true in markets that are already subject to governmental regulation. Such regulation tends to significantly reduce the likelihood of major antitrust harm. The additional benefit to competition of adding another layer of legal process will tend to be small, whereas the risk of false positives is high.8 United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 27.

  3. Would a business typically analyze or estimate the likely cost savings from this type of conduct before or after engaging in it?
  4. As in bundling and tying cases, reducing the occurrence of over-enforcement in cases involving refusals to deal while being efficient and administrable requires the consistent application of sound economics. In order not to suppress conduct that would be beneficial to consumers, appropriate standards must be adopted that condemn only conduct that is not "competition on the merits," while allowing firms to reap the fruits of their skill, foresight and industry by being able to predict the likely consequences of their actions. Meaningful guidance must be provided to firmsto enable them to know how to avoid liability using data that is readily available to them at the planning stage, and that the conduct, if challenged, will be evaluated under the same efficient standard that applied at the time the company decided to engage in the conduct. United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 26.

  5. What tests and standards should court and enforcement agencies use in assessing whether such conduct violates Section 2?
  6. The recent decision in Verizon Communications v. Law Offices of Curtis Trinko, LLP,9 deals explicitly with the challenge facing courts in their endeavors to avoid deterring beneficial conduct or imposing remedies that they are ill-equipped to administer. For these reasons the Court declined to use Section 2 to impose upon the defendant a duty to deal with a competitor, even if the refusal allowed the defendant to reap the benefits of its position in the market... In essence, the Court reiterated reasons why Section 2 should be applied cautiously to refusals to deal, and resisted the temptation to recognize new theories of liability or to declare the practice immune from attack. ICC members can understand why the Court (and the U.S. competition authorities, which filed briefs in the case) did not regard the situation before it as one of the rare exceptions to the right of parties to choose their customers. Accordingly, we would not characterize the decision as going too far. International Chamber of Commerce, Comments on Selected Issues for Study by the U.S. Antitrust Modernization Commission, Document No. 225/621 (1 September 2005) at 14-15.

    A firm's dealings with third parties and its prior dealings with rivals provide a baseline for evaluating its challenged conduct. Where a firm is willing to deal with its retail customers on certain terms (such as a certain price), claiming that its refusal to deal with a rival on those terms constitutes anticompetitive conduct makes no economic sense. However, absent discriminatory dealing or departures from prior profitable courses of dealing, decisions by either courts or regulatory agencies to enforce sharing distorts the incentives to innovate and should therefore be avoided. United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 26.

    ICC believes that patented and non-patented technical technology should be treated on the same footing and that the requirement that the refusal to license prevents the appearance of new goods or services be clearly set out. International Chamber of Commerce, Comments on the Reform, of the Application of Article 82 of the EC Treaty, Document No, 225/623 (12 December 2005) at 19.

    Similarly, there is no justification in law or economics for the proposition that trade secrets should be entitled to less protection under Article 82 than other forms of intellectual property. If trade secrets are provided less protection than other forms of intellectual property, the net effect will be less innovation and competition in the market, not more. This is simply because the protection of trade secrets enables firms to recover the investments they make in the research and development that are necessary for the firm to be able to meet the competitive pressures of its rivals, who are themselves investing in research and development for the same reason. Thus, as is the case with other forms of intellectual property, uncertainty as to the ability to recover the costs of the research and development necessary to create innovative trade secrets acts as a disincentive, to the detriment of consumer welfare. From the other perspective, there is little incentive for risking the loss of your own investment in research and development that may fail to yield the desired results when you have the option of free-riding off of the efforts of a rival. For these reasons, sound economics requires that trade secrets be protected the same as any other form of intellectual property, and that the rules and regulations impacting intellectual property rights not create ambiguity with regards to the extent of their protection. United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses (30 March 2006) at 28-29.

We hope that our comments will be helpful. We welcome the opportunity to expand upon our comments, if necessary.

Document No. 225/637
14 December 2006


APPENDICES

  1. Comments on the Reform of the Application of Article 82 of the EC Treaty (Document No. 225/623 (12 December 2005),available at htt://ww.iccwbo.org/uploadedFiles/ICC/policy/competition/Statements/Comments%20on%20t he%20Reform%20ofDIo20the%20Application%20of%20Article%2082%20of%20the%20EC%20 Treaty.pdf
  2. The ICC Comments on the European Commission Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses (Document No. 225/627 (7 April 2006), available at htt://ww.iccwbo.org/uploadedFiles/ICC/policy/competition/Statements/ICCComments%20EC%20Article%2082.pdf)
  3. The International Chamber of Commerce, Comments on Selected Issues for Study by the U.S. Antitrust Modernization Commission (Document No. 225/627 (September 1, 2005), available at htt://ww.iccwbo.org/uploadedFiles/Submission%20to %20the %20AMC.pdf)
  4. The United States Council for International Business, Submission to the Directorate-General for Competition on the Application of Article 82 of the Treaty to Exclusionary Abuses ((30 March 2006), available at htt://ww.uscib.org/docs/Final_USCIB_Arcle82.pdf)

FOOTNOTES

1. Document No. 225/623 (12 December 2005),available at http://www.iccwbo.org/uploadedFiles/ICC/policy/competition/Statements/Comments%20on%20the%20R eform%20of%20the%20Application%20of%20Articie%2082%20of%20the%20EC%20Treaty.pdf

2. Document No. 225/627 (7 April 2006), available at http://www.iccwho.org/uploadedFiles/ICC/policy/competition/Statements/ICC_Comments%20EC%20Art icle%2082.pdf

3. Document No. 225/621 (1 September 2005), available at http://www.iccwho.org/uploadedFiles/Submission_%20to_%20the_%20AMC.pdf

4. (30 March 2006), available at http://www.uscib.org/docs/Final_USCIB_Article82.pdf

5. Brief for the United States, et al, as Amici Curiae Supporting Petitioner, Verizon Communications Inc, v. Law Offices of Curtis V, Trinko, LLP, 540 U.S, 398 (2004) (No 02-682), 2003 WL 21269559 at *13-20,

6. Verizon Communs., Inc.,540 U.S. at 407.

7. Illinois Tool Works, 126 S.Ct. at 1293.

8. Verizon Communs., Inc., 540 U.S. at 407-408, 411-15.

9. 540 U.S. 398(2004).


Comments on the Reform of the Application of Article 82 of the EC Treaty

Prepared by the Commission on Competition

This paper is submitted to the European Commission by the International Chamber of Commerce (ICC). It has been prepared by its Commission on Competition.

This is the first ICC paper on reform of the application of Aricle 82 EC. It deals with a number of general points. We will in a subsequent paper address some specific forms of abuse of dominant position. We also intend to comment on the draft Guidelines when available.

We begin by reviewing the purpose of Article 82 and its implications for the benchmarks to be used.

We then comment on what we recommend as a modernized approach to dominance, relevant market and abuse.

We end by stating briefly how a modernized application of Article 82 would contribute to the pursuit of the objectives of the Lisbon Agenda.

1. Some general remarks
The purpose of Article 82

The EC Treaty makes clear that its competition rules are designed to establish "a system ensuring that competition in the internal market is not distorted" (Art. 3(1)(g)).

In applying Article 82, this means preventing practices of dominant firms that distort the normal functioning of the market.

"Normal functioning" of the market in the presence of dominant firms can have different meanings. It could mean a market that is efficient in the sense that it leads to optimal allocation of resources, provides to economic agents appropriate incentives to pursue innovation, efficiency and quality, and maximizes consumer welfare.l Under this interpretation, the rules on abuse of a dominant position are enforced against conduct of dominant undertakings where it is established that the conduct does not enhance efficiency and harms consumers.

"Normal functioning" of the market could also mean a market in which economic agents have access to the market and operate on the market without obstacles created by dominant undertakings. Under this interpretation the presence of a number of competitors on the market is of paramount importance and is considered, in and of itself, as necessary in order for the market to function so as to achieve the ultimate purpose of competition rules. Under this interpretation, the rules against abuse of dominant position are seen primarily as protecting competitors in order to protect competition. The drawback of this interpretation is that, by focusing on the presence of competitors, it may end up protecting less efficient competitors and prohibiting conduct of dominant undertakings that furthers Article 82's ultimate purpose of promoting an efficient market. In addition, protecting rivals against competition from the dominant undertaking may reduce their incentive to engage in robust and creative competition that can further efficiency and benefit consumers.

We consequently recommend that, when applying Article 82, the Commission move beyond findings relating to the effects on competitors and assess whether the conduct of a dominant undertaking is likely to have effects that promote or impede efficiency and benefit or harm consumers. As Commissioner Kroes has noted, "it is competition, and not competitors, that is to be protected"2. We elaborate this recommendation in the following sections.

The ultimate test

The Commission decisional practice and the case law of the EC court have not provided clear guidance as to which interpretation is correct. Various concepts have been used as benchmarks for assessing whether the conduct of a dominant undertaking is abuse, including "normal competition"3, "competition on the merits"4; "genuine undistorted competition"5. These vague concepts need to be clarified and elaborated so as to reflect the purpose of the rules against abuse of dominance. In its Guidelines on the Application of Article 81 (3), the Commission has done so with respect to the concept of "restriction of competition" and has put forward consumer welfare as the ultimate test.

We are of the view that harm to consumers, which is expressly referred to in Article 82 (b), is the ultimate test of abuse of dominance6, just as it is for Article 81. In Continental Can7 the ECJ made clear that the purpose of Article 81 and 82 should be consistent. If harm to consumers is the ultimate test for "restriction of competition" under Article 81, it should be so as well under Article 82.

It is our understanding that the Commission may wish to maintain the distinction made by the EC court between practices directly damaging consumers ("exploitative abuse") and practices that do so less directly by restricting competition by efficient undertakings ("exclusionary abuse")8.

In that event, Article 82 should be cautiously applied to conduct coming within the first category and only be applied to condemn a practice where it is established that the practice is likely to have a direct material adverse effect on consumer welfare in the form of higher prices or less output, so as to not stifle innovation and investment.

As to conduct coming within the second category, there is clearly a need for principles that distinguish legitimate competition by a dominant undertaking from exclusionary abuse. In making this important distinction, the Commission should not adopt a single test - such as the "profit sacrifice" test, and, for pricing behaviour, the "equally-efficient competitor" test or the "limiting production" test - for all forms of potentially exclusionary conduct because there is no consensus that any such test is applicable in all circumstances. The Commission should instead elaborate principles, such as commitment to promote efficiency and consumer welfare, and apply those principles to individual cases on the basis of a careful assessment of the particular facts9. These principles should be set out in guidelines, so that they can be taken into account ex-ante by undertakings when they decide on a given course of conduct.

Legal Certainty

As is clear from the above, we advocate that the Commission should move away from a legalistic "form-based" approach to a more economics-based approach in the application of Article 82. Such a move would be in conformity with recent developments in the other areas of EC competition law.

A more economics-based application of Article 82 would focus on increase in consumer welfare. It should not lead at the same time to reducing legal certainty as long as undertakings are in a position to assess whether their conduct has a legitimate efficiency-enhancing business justification. Much of the current uncertainty about the boundaries between permissible and prohibited business practices results from a form-based approach to certain pricing practices and the difficulty inherent in such an approach in determining whether new kinds of economic activity should be regarded as being of one type of form or another. Form-based approaches lack consistent and rigorous analysis of the concrete effects of a given practice and often have the effect of condemning profit-maximizing conduct that benefits consumers. The uncertainty that results from the condemnation of conduct that may not have any significant impact on competition or that may benefit consumers creates added risks for business, which itself reduces efficiency, and deters undertakings from applying business practices (e.g. certain pricing schemes) which in fact increase competition and are beneficial for consumers. The deterrence of desirable conduct is enhanced because of the lack of an official procedure for undertakings to make sure that a certain business practice is in conformity with the competition rules, the lack of coherent and clear case law and the level of fines inflicted for abuses of a dominant position.

We suggest that the Commission issue economics-based guidelines, which would guarantee an adequate level of legal certainty by making it clear that single undertaking conduct that enhances efficiency and benefits consumers is not an abuse, whatever its form and the degree of market power of the undertaking concerned. Such guidelines are all the more necessary to avoid inconsistencies, now that increased enforcement by NCAs and national court is to be expected.

Such guidelines would also be to the benefit of competition authorities and court by lowering enforcement costs.

2. Dominance

The current review of Article 82 is focused on the various types of abuse and the need to give clarity and guidelines in relation to conduct by dominant undertakings and the test to assess whether business practices may or may not be held to be abusive.

So far the Commission has not published any notice that gives guidance on the assessment of dominance or market power. There are documents where the Commission has given some indication on its thinking on dominance in other contexts:

  • The Guidelines on Market Analysis and the Calculation of Significant Market Power in Electronic Communications;10
  • References to the competitive assessment of mergers (which include the creation or strengthening of a dominant position) in the Guidelines on the Assessment of Horizontal Mergers 11; and
  • The Guidelines on the Application of Article 81 to Technology Transfer Agreementsl2 provide some indication of the factors to take into account in assessing the market power of the parties to a technology transfer agreement that falls outside the thresholds of the Technology Transfer Block Exemption.

While the above documents, together with Commission decisions and the EC courts' case-law on dominance.13 provide some insight into the Commission's thinking, the lack of a comprehensive framework for assessing dominance undermines legal predictability and business certainty. Therefore, specific guidance by the Commission on the assessment of dominance under Article 82 would also be welcomed in the context of the current review, since a finding of dominance is a basic element of the scrutiny of any business conduct under Article 82.

By contrast, the OFT in the UK has published Guidelines on the Assessment of Market Power which explain how the OFT will assess whether undertakings have market power when investigating cases both under Articles 81 and 82 and under the respective UK law provisions (Chapter 1 and Chapter 2 of the Competition Act 1998). In particular, the OFT explains that a company is not dominant unless it has substantial market power and acknowledges that market power is not an absolute term but a matter of degree, and that the degree of market power will depend on the circumstances of each case, including whether and to what extent the company concerned faces competitive constraints from existing or potential competitors, and other factors (such as strong buyer power). This is consistent with the position of Commissioner Kroes, who has noted that "market shares are not - on their own - sufficient to conclude a dominant position exists" and that to show dominance, "a full economic analysis of the overall situation is necessary".

We believe that this approach is the correct and appropriate starting point for the assessment of dominance and we encourage the Commission to put in place a framework providing predictability as to the tools that will be used in assessing, in the specific circumstances of each case, whether an undertaking may be in a dominant position under Article 82.

In this context, the Commission should distinguish between the following scenarios:

  • Cases where an undertaking is below the dominance threshold and, therefore, its conduct cannot be found to be abusive under Article 82;
  • Cases where an undertaking may be held to be dominant; and
  • Complex areas in fast-moving markets where caution is warranted in findings of dominance and the application of Article 82.

Cases where an undertaking is below the dominance threshold

The objective of Article 82 to promote efficiency will be furthered if undertakings have a clear understanding of when they will not be regarded as dominant. The Commission should thus develop some screening mechanisms to determine "safe harbors", e.g. undertakings with a low market share may presumptively be able to engage in certain unilateral conduct. Similarly there may be situations where an undertaking has large market shares for only a brief period of time, before the emergence of a new product or new competition, and therefore cannot be held to have any market power. Such mechanisms will reduce uncertainty and allow undertakings in "safe harbors" to engage in robust and creative competition for the benefit of consumers. In formulating its "safe harbors", the Commission should give special attention to the pronouncement of the ECJ in Hoffman-Larochel4 that the fact that an undertaking is compelled by competitive pressure to lower its own prices is inconsistent with the independence vis-à-vis consumers and competitors that is the hallmark of dominance.

Cases where an undertaking may be held to be dominant

It is now widely acknowledged that: (a) market shares are only the starting point that give a first indication of the market structure and of the competitive position of an undertaking; (b) market shares alone are not conclusive in determining whether an undertaking has market power; and (c) there is no specific market share threshold that reliably establishes that an undertaking has market power. In fact, what is of particular significance in assessing dominance is the exercise of market power over time, i.e. the ability to profitably raise prices above competitive levels.

The Commission is increasingly relying on economic analysis and empirical work in its investigations. That trend is taking place in the assessment of mergers and vertical agreements and Commissioner Kroes has indicated in recent statements that the Commission intends to use similar economic analysis in investigating potential abuses of dominant position under Article 82.

We support this commitment and recommend that the Commission give clear guidance about the economic framework, tools and evidence that it will use to assess the competitive constraints on undertakings (such as the structure of the market, existing and potential competitors, entry analysis and countervailing buyer power).

Complex areas in fast-moving markets

There has been some debate recently about the need to enforce competition law with caution and flexibility in fast-moving markets where the key features of certain industries (particularly high-tech and "new economy" industries, such as computer software and hardware, internet, mobile telephony and biotechnology) make it challenging to apply traditional competition law concepts and tools to analyse competitive issuesl5. Such industries are often characterised by huge investments in R&D and IPRs, network effects, high fixed sunk costs and low marginal costs. Competition in these markets is dynamic in the Schumpeterian sense that competition often takes place for the market in a "winner takes all" race. Undertakings may have high market shares, but are constantly subject to threat from innovative competitors and potential entrants. Some scholars and economists argue that the mechanical application of static models does not give true reflection of market power when applied to high-tech/new economy industries.

Therefore, the Commission should also give guidance on how it intends to assess dominance/market power in such fast-changing and complex markets.

Collective dominance

The Commission has taken the view that Article 82 also applies where two or more undertakings together hold a dominant position. The EC courts have endorsed this view and have set out conditions to be fulfilled in the cases that were brought before them. In the interest of legal certainty, some guidelines should be issued summarizing the case law. They should clarify that collective dominance does not apply to uncoordinated single undertaking conduct and that such conduct must be assessed based on the market power (or absence thereof) of the individual undertaking.

3. Relevant Market Definition The purpose of market definition

In order to establish whether an undertaking already possesses market power (or is likely to achieve such a position in the future), it is necessary to define the relevant market in which the undertaking is alleged to have such power. If the definition of market power is the ability of an undertaking "to behave independently of its competitors and customers':it is clearly necessary to identify the competitors.

The purpose of the Commission's Market Definition Noticel6 was to create a common framework for identifying and defining the boundaries of competition between undertakings (the hypothetical monopolist test) that could be applied to all competition analyses: mergers, Article 81 situations, Article 82 investigations and state aid enquiriesl7

Differences between mergers, agreements and abuse of dominance

Important differences exist, however, in respect of the role that market definition plays in the different analytic frameworks:

  • In the Article 81 arena, a precise market definition used to be relevant largely for non-full function joint ventures and vertical relationships. In the latter case, even that application disappeared once the de minimis and the block exemption market share thresholds were exceeded. A more economics-based approach to Article 81 in the post-modernisation world may re-focus the attention to an analysis of actual effects on the market which in turn will demand a more rigorous approach to market delineation18.
  • In respect of mergers the tide flowed in the other direction. The importance of a precise market delineation has somewhat declined, given the change of the substantive test from dominance to SIEC. In the new world of the SIEC test, the Commission will often directly look at the likely consequences of the merger on post-merger prices, without a detailed definition of the relevant market.
  • For Article 82 cases, market definition remains a crucial par of the Commission's analysis since the Commission will have to establish that the undertaking holds a dominant position in order to consider whether an abuse has taken place. However, unlike merger cases, where the analysis is prospective (i.e. will prices rise?) an Article 82 review will be historic (i.e. does a particular undertaking already hold market power?).

Avoid artificially narrow markets

It is now well understood that the inability to raise prices without facing significant substitution does not necessarily demonstrate wider markets (or significant competition) but could mean that the current price is already set at a supra-competitive level (the cellophane fallacy).19

The danger of the cellophane fallacy is that the relevant market in an Article 82 case may be narrower than in a merger situation, but only in very limited circumstances and not in all Article 82 cases.

A more economics-based approach to Article 82 therefore not only means avoiding the cellophane fallacy, but also, importantly, avoiding artificially narrow market definitions. There are a number of cases in which the Commission adopted a very narrow market definition, unconnected with the cellophane fallacy and contrary to the principles set out in the Market Definition Notice. The reason seems to lie mainly in the Commission's focus on demand-side considerations to the near exclusion of supply-side factors.

These examples can be grouped into two categories: (i) a too narrow focus on a particular customer segment and product characteristics; and (ii) a too rigid view of markets involving consumables (after-markets):

  • Too narrow focus on particular customers
  • In United Brands/Chiquita20 an important contributing factor in the Commission's market analysis was that bananas were a very important part of the diet of only one customer segment (i.e. the "young, sick and the very old")21 and this seems to have been accepted by the Court: "The banana has certain characteristics, appearance, taste, softness, seedlessness, easy handling, a constant level of production which enable it to satisfy the constant needs of an important section of the population consisting of the very young, the old and the sick"22 This conclusion seems to have been reached without sufficiently robust empirical evidence. Moreover, neither the Commission nor the Court considered whether the remaining customers (for whom switching to other fruit was viable and who could not be charged different prices) constituted a sufficiently large group constraining any price rises.

    Similarly, in Hilt23 the Commission decided that powder-actuated fastening systems (nail guns) form a distinct market from other fastening systems (such as welding, self tap screws or rivets, bolts and nuts). This was based purely on the fact that product characteristics differed and that there may not be full demand-side substitution.24

    The Commission did not consider whether the pricing of one product constrains the pricing of the other products. On appeal, the Court of First Instance similarly failed to consider whether the number of marginal customers who could switch and who could not be charged different prices was sufficiently large to act as a constraint.25

  • After markets (primary products and consumables)

    If the approach in Article 82 cases evolves away from per se prohibitions towards a more economics based approach, the relevance of market share and hence market definition may well decline, as the Commission focuses directly on the competitive harm. However, doing so without a robustly defined market removes a significant methodological safeguard against findings of dominance by instinct.

    In the context of complementary products, concentrating on demand-side substitutability will lead to the definition of separate markets for the main product and for the spare pars or consumables for that product, since the main product and its spare parts or consumables will not be interchangeable, be it at the level of supply or demand.

    The main product and the spare parts or consumables have been held to form separate relevant markets in cases like Hugin26, Hilti27 and TetraPak28. In these cases the Commission, upheld by the Court, defined the market for consumables or spare part by reference to the primary product (e.g. "Hilti-compatible"). In such cases, the manufacturer was found to hold a dominant position on the market for these spare pars or consumables29, although it may not have been dominant in the market for the main product.

    In the context of complementary products, substitutability should be explored through the analysis of the effects of an increase in price of either the main product or the spare part or consumables. An explicit analysis of the impact of a hypothetical price increase has not always been carried out, particularly in the cases mentioned above (albeit that an econometric study was presented in the Hilti case).

    This type of analysis could show, for example, that in the case of a price increase of the main product A, consumers switch to competing products B and C and to their consumables or spare parts, since the purchase of a competing main product without its spare part or consumables would be of no value to the customer. Similarly, in the case of an increase in the price of the consumables or spare part, consumers switch again from product A to product B and C and their consumables or spare pars. In such a hypothesis, the conclusion to be drawn is that both the main products B and C and the consumables or spare parts, ie the whole system, lie in the same relevant market as the main product A and its consumables or spare part.

    More recent Commission decisions: no change in market definition

    There has been an evolution in the Commission's analysis in recent decisions. In 1995, it announced that several factors had to be taken into account in order to assess dominance with respect to spare parts or consumables: price, life-time of the main product, transparency of prices of spare parts or consumables, prices of spare parts or consumables as a proportion of the main product value and information costs.

    Although it examined two separate markets, taking these elements into consideration led the Commission in the Pelican/Kyocera case30 to conclude that there was no dominance on the market for consumables. Particular features of both the market for consumables and the market for the primary product, such as the price of the consumables as a high proportion of the main product value, which was taken into account by customers from the outset, meant that consumers would switch to another main product if the price of consumables for the first main product increased. The Commission concluded that there was no evidence of possibilities for price discrimination between "old" or captive customers and new customers.

    A similar more economics-driven analysis was carried out in the Info-Lab/Ricoh case in 199931, where the Commission held that the market for the main product and the market for consumables were interrelated in such a way that competition in the main product market also constituted an effective competitive constraint in the consumables market. Ricoh was not dominant on the main product market and was not held to be dominant on the separate market for consumables.

    The Commission therefore appears to have reconciled (i) the definition of two separate markets for the main product and the consumables with (ii) the conclusion that the manufacturer of a main non-dominant product may not be dominant on the market for consumables for the main product, by analysing the links in terms of competitive constraints between these markets.

    Suggested improvements

    This approach remains debatable in that the Commission still typically considers the main product and the spare part or consumables to constitute two separate markets. The analysis of the competitive links between both markets, in particular with regard to switching costs, should, however, in appropriate cases, lead the Commission to conclude that the main product and its consumables or spare parts constitute a system which is in competition with other similar systems, where consumers would switch to another main product and its consumables or spare parts, if the price of the consumables or spare parts of the first main product were to increase by a small but significant amount.

    Although its more recent analysis has led the Commission to conclude that the manufacturer may not be dominant on the market for consumables or spare parts, it could also reach (as it has done in the past) the opposite conclusion, even if the manufacturer is not dominant on the market for the main product. In such cases, manufacturers would have to apply different commercial policies to two complementary products, causing genuine difficulties from a business point of view.

    It is therefore recommended that the Commission examine the competitive links between products and systems at the stage of market definition. The Commission would thus recognise, in line with economic analysis, that main products and their spare parts or consumables should, in appropriate cases, be considered as systems which, together with the other systems against which they are in competition, constitute a single relevant product market.

    4. Abuse

    On "abuse" we have three general comments that we hope the Commission will take into account in preparing its forthcoming guidelines. These are (1) forms of conduct not listed in Article 82, (2) the "legalistic" approach and (3) the treatment of "efficiencies".

    Forms of conduct not listed in Article 82

    It is established that the catalogue of the forms of abuse listed in Article 82 is an open one. However, in order to qualify conduct other than the forms listed in Article 82 as abuse, courts and regulators cannot limit themselves to finding that such conduct is capable of having or likely to have the effect of restricting competition. Even if one should deduce from the fact that by expressly listing certain forms of conduct as abuse the Treaty has introduced a presumption of abuse, courts and regulators can only treat forms of conduct other than those listed in Article 82 as abuse where, on the basis of a precise examination and convincing evidence, they find that the conduct has effects that run counter to the purpose of Article 82.

    In identifying the constituent elements of the abuse, the Commission should make it clear that, while harm to competitors is necessary for conduct to be an abuse, it is not sufficient. After all, inventing better products or more efficient methods of distribution, reducing price or offering better terms of trade for the benefit of consumers, and more quickly adapting to changes in the market can disadvantage rivals and maybe even cause them to abandon business. Yet these forms of conduct enhance efficiency and consumer welfare, and should thus not be prohibited by Article 82. The Commission should articulate standards that make clear that conduct by a dominant firm would be deemed to be an abuse only if it does not promote efficiency or consumer welfare.

    The Commission might be tempted to argue that even conduct that increases efficiency can be an abuse if it excludes competitors on the ground that, in the long run, the loss of competitors will reduce competition and, ultimately, consumer welfare. However, such an approach would require the Commission to establish that the long run harm from the reduction in competition exceeds the short run increases in consumer welfare and the long run improvement in efficiency attributable to the dominant undertaking's conduct. The Commission should be very reluctant to treat conduct as an abuse on the basis of such a tradeoff. Projecting and estimating the magnitude of long run harm to competition is almost always very difficult and uncertain. More importantly, were the Commission to attempt to meet that burden, it should explain how it addresses the risk of overstating long run competitive harm in the particular circumstances of the case under consideration. Indeed, dynamic, Schumpeterian competition inherently brings forth new innovations and new entry that were not anticipated and could not have been predicted. Efforts to assess long run harm to efficiency and to consumers as a result of exclusion of competitors clearly risks grossly overstating the harm if they fail to take account of these likely developments. Therefore, in cases where the conduct at issue generates short term efficiency and the Commission nevertheless considers prohibiting such conduct, it is the duty of the Commission - as the institution entrusted with the enforcement of competition rules aimed at promoting consumer welfare - to demonstrate that the balancing test between short- and long-term effects avoids the risk of overstatement of long term consumer harm.

    "Legalistic" approach

    Our second comment relates to the "legalistic" (form-based) approach mainly used so far when dealing with Article 82 cases. As the EAGCP Report points out "[t]he standard for assessing whether a given practice is detrimental to 'competition' or whether it is a legitimate tool of 'competition' should be derived from the effects of the practice on consumers" (at p. 8). A form-based approach is inadequate as a standard for such assessment. To illustrate this, we briefly analyse the treatment of rebate and discount structures.

    The issue

    In evaluating whether a discount or rebate ("rebate" for short) scheme constitutes an abuse under Article 82 EC, the European court determine whether the rebate scheme is objectively justified32. If it is not, there is an evident danger from certain CFI rulings that a rebate by a dominant undertaking will be treated asperse having restrictive and/or discriminatory effects. As a result of developments in the case law, there has been a narrowing of the types of rebate considered legitimate when implemented by a dominant undertaking. This narrowing has reached a point where all rebates, including those calculated on a quantitative basis, are deemed to be abusive when established by a dominant undertaking, unless they are the result of economies of scale that are passed on to the customer33.

    For example, although the judgement is ambiguous, in Michelin II, the CFI appears to have taken the view that under Article 82 it may not always be necessary to show an actual effect on competition: rather, the conduct may be abusive if it "tends to restrict competition or, in other words, that the conduct is capable of having that effect" (para 239). Thus, the CFI appears to consider volume rebates as presumptively abusive, a presumption that can be rebutted if there is an objective justification of the rebate scheme. However, Michelin II suggests that only transaction-specific cost justifications will suffice: i.e. the grant of a specific rebate must be linked to economies of scope gained through sales to that particular customer, in that particular transaction34.

    Furthermore, the European courts have shown an increasing willingness to find behaviour to be abusive where there is no anti-competitive effect, or even any strong likelihood of anticompetitive effect, but merely the potential for harm. These developments have created a legal doctrine which has serious consequences for dominant undertakings and which seriously restricts the pricing structures and arrangements that they are permitted to establish. Indeed, as things now stand, the rules relating to discounting are considerably stricter than those applicable to single dealing and refusals to deal. In the case of single dealing, the courts have considered the actual exclusionary effects35, while in the case of refusals to deal, the court have developed a rule of reason approach36. The stricter treatment of rebate schemes appears difficult to justify: discounting reduces the cost to consumers and is not on its face exclusionary.

    In examining how this position has been reached, it becomes clear that the legal and economic foundations for the doctrine are shallow and insufficient to support the edifice now constructed on them. In particular, these cases overlook the fact that rebates increase allocative efficiency and consumer welfare by increasing output and reducing prices. They are often preferred by customers to alternative arrangements and are often the result of hard bargaining by customers to get the best price from undertakings that, because they are dominant, would otherwise charge higher prices. Conduct of this nature should be subject to Article 82 only if there is a compelling economic basis for doing so. Increases in allocative efficiency and consumer welfare ought to be regarded as objective justifications for rebates and should negate the assumption that such rebates are exclusionary by their very nature.

    Summary of Existing Case Law

    1. In Hoffmann-La Roche37, the Court discussed exclusivity agreements and fidelity rebates which had a similar effect to exclusivity agreements, namely, tying customers to Roche for the supply of all, or a large proportion, of their purchasing requirements. These fidelity rebates were contrasted with quantity rebates based solely on the volume of purchases, on the basis that the fidelity rebates were:

      "... designed, through the grant of a financial advantage, to prevent customers from obtaining their supplies from competing producers"38.

      This distinction also applied to fidelity-type rebates set at progressive rates but based on the percentage of a customer's estimated annual requirements.

      No adequate explanation was provided why these rebates should be any more likely to prevent a customer obtaining supplies from competitors than a quantitative rebate, based solely on the volume of purchases, or even just a low price, when such a rebate or low price is also designed to encourage the customer to purchase more from the same supplier.

      The rebates in Roche were also distinguished from pure volume rebates on the basis that they were not dependent on quantities fixed objectively, but based on estimates of the annual requirements of each customer. They were driven by an aim to obtain the maximum volume of a customer's requirements, rather than just the maximum volume of sales possible. The system was therefore discriminatory:

      "... applying dissimilar conditions to equivalent transactions with other trading parties"39.

      The Court presumed that this discriminatory behaviour constituted an abuse without examining in detail whether, in practice, the rebates prevented customers from choosing their supplier or reduced consumer welfare.

    2. The reasoning of the Court in Michelin I40 is also based on the premise that fidelity rebates (but, apparently, not pure quantity rebates based "objectively" on volumes purchased):

      "... prevent customers obtaining their supplies from competing manufacturers"41.

      However, again, no adequate analysis or explanation is provided why rebates based on attaining a particular target percentage of the previous year's requirements are necessarily (and legally) more restrictive than quantity discounts linked solely to the volume of goods purchased. Because the percentage targets were calculated on an individual basis, the behaviour was deemed discriminatory and therefore abusive under Article 82 EC.

    3. In Michelin II42, the system for calculating rebates applied the same volume-based rules to all customers and was therefore not discriminatory. According to previous case law, as a pure quantitative rebate system, this system should have been considered legitimate. However, the Court held that even quantity rebates (applied by a dominant undertaking) are illegal unless they can be justified by economies of scale:

      "... a rebate system in which the rate of the discount increases according to the volume purchased will not infringe Article 82 EC unless the criteria and rules for granting the rebate reveal that the system is not based on an economically justified countervailing advantage but ten, following the example of a loyalty and target rebate, to prevent customers from obtaining their supplies from competitors"43.

    4. The possibility of restrictive or discriminatory behaviour being justified by an economic analysis had been raised in Roche and Michelin I, but such analysis was not considered to be a factor in determining whether the behaviour itself was restrictive. (Quantitative rebates not based on targets relating to previous purchases by a customer had always been considered not to have a restrictive effect and economic justification related only to behaviour which had already been deemed restrictive or discriminatory.)

    The lack of legal or economic basis for the legal doctrine

    There is no basis for the premise that all rebate systems established by a dominant undertaking are abusive unless they are cost-justified. In the early case law, there is no adequate legal or economic analysis to distinguish quantitative rebates based purely on volumes purchased from "fidelity-type" rebates. Furthermore, the need to demonstrate that a quantitative rebate depends on cost savings in order not to be deemed restrictive is new in Michelin II and is largely unexplained. That requirement ignores all sorts of other efficiencies that are often realized by rebates (allocative efficiencies, economics of scope in the case of multi-product undertakings and distribution efficiencies).

    The current case law is, in effect, a form-based rule: a rebate system which is transaction-specific, quantity-based and structured to reflect benefits of scale is legitimate if applied by a dominant company: all other rebates applied by a dominant company are, by default, unlawful, irrespective of effect or market analysis.

    Form-Based -v- Effects

    In Michelin II, the Court considered that the rebates were loyalty-inducing44 and that this practice was capable of restricting competition45. Paragraph 60 of that judgment identifies two areas for economic and factual inquiry: (i) whether there is an economic justification for the rebate; and (ii) if not, whether the rebate has anticompetitive effects46. While there has been a tendency on the part of the Court to curtail the second analysis once it determines that there is no economic justification for the rebate under the first test, there is no legal requirement for it to do so. Thus, the failure to conduct such an analysis is inconsistent with both the legal test set. forth in Michelin II and the Court's case law, which gives foreclosure effects a serious examination47. In particular, there should be no presumption that a particular type of discount and rebate will have an anti-competitive effect based on its form, especially as discounts and rebates produce immediate benefits in the form of lower prices in every instance whereas they may produce harm only on a limited set of circumstances. And, whether or not any inference of anti-competitive effects is based on evidence of actual market performance, it should be rebuttable with evidence of actual market performance.

    Given the possible pro-competitive benefits of some rebates which are not quantity-based and structured to reflect benefits of scale, a blanket prohibition of all such schemes by dominant undertakings under Article 82 could discourage potentially efficient behaviour.

    The need to move away from form-based rules

    From an economic perspective, it is difficult to find any form of unilateral behaviour that will always be harmful to competition, without considering the market context in which the behaviour occurs. It is therefore difficult to find an economic rationale for a regime in which certain types of behaviour are perse unlawful once an undertaking passes over the threshold into dominance, especially when they are per se lawful below this threshold.

    An effects-based framework of analysis under Article 82 would also be in line with the approach under the Article 81 (3) Notice, in particular if the Article 82 Guidelines mandate the same requirements in terms of quantification and balancing the pro- and anticompetitive effects of unilateral behaviour. ice believes that the draft Guidelines should fully acknowledge the analytical approach under Article 81(3) Notice.

    A rebate scheme may create an incentive for a customer to purchase all his requirements from the dominant undertaking but it needs to be established that the scheme has a negative impact on consumers. If the rebate can be easily matched by rivals, only applies to some but not to others or is of short duration, it may well have no negative effects. A detailed market analysis of the actual effects is, therefore, indispensable.

    On "efficiencies"

    A third comment relates to efficiencies. Efficiencies occur when mergers, agreements or unilateral conduct give rise to reduced prices, improved quality or other positive effects that benefit consumers. Efficiencies may be either of a quantitative (cost) nature or of a qualitative (dynamic) nature. When unilateral conduct enables an undertaking to realize economies of scope or scale or to utilize more efficient production or distribution methods, it can reduce costs. Unilateral conduct can also increase output when, for example, it better aligns incentives of distributors and other producers of complements with the dominant undertaking. Unilateral conduct by undertakings can also promote dynamic efficiency, where it leads to increased research and innovation, or the development of new and improved methods of production and distribution. Unilateral conduct by dominant undertakings also increases allocative efficiencies whenever it reduces prices or increases output.

    There is widespread consensus among economists that efficiencies are to be measured in terms of consumer surplus. This methodology explicitly underlies the Guidelines on the application of Article 81(3) and, less explicitly, the Guidelines on the assessment of horizontal mergers.

    As a consequence, it is nowadays well established under EC competition law that static and dynamic efficiency gains are to be taken into account when assessing both horizontal and non-horizontal mergers and agreements. Similarly, under US law, under both Section 1 and 2 Sherman Act, efficiencies are factored into the analysis of courts and antitrust agencies of the net effect of the transaction at issue. In contrast, the scope for taking account of efficiencies under Article 82 seems at best very limited in a number of cases, non-existent in other cases and, on the whole, unclear. First, while case law under Article 82 (e.g. Michelin II, Telemarketing) postulates that an objective justification prevents a finding of an abuse of a dominant position, that notion appears to be interpreted in an extremely limited manner (see e.g. Hoffmann-La Roche; see also the Commission decision in Microsoft). Second, in addition to the (few) cases where the proffered objective justifications were considered - but rejected -, a number of practices, in particular in the field of rebates, are subject to a per se analysis. As a result, the treatment of efficiencies under Article 82 is cumbersome, at best.

    We recommend that in its guidelines, the Commission expressly take account of all types of efficiencies for the following reasons:

    • There is no economic support for a per se approach to the analysis of Article 82. On the contrary, there is consensus among economists that (unilateral) price- and non-price conduct of dominant firms may produce both pro- and anticompetitive effects. The ambiguous nature of conduct of dominant firms militates in favour of a full appreciation of the (positive and negative) effects on consumers. While it may perhaps be justified to treat some restrictive agreements, such as horizontal price fixing and market sharing agreements, as per se violations under Article 81 because it is obvious that they will produce anticompetitive effects, a similar rationale does not apply in relation to conduct assessed under Article 82.
    • Consideration of efficiencies in the assessment of conduct under Article 82 merely reflects the role of undistorted competition as a means towards the achievement of the Treaty objectives as set out in Article 2. As indicated above, conduct which generates dynamic or static efficiencies should not be deemed abusive unless it is demonstrated that the impact of this conduct on competition will result in consumer harm outweighing these efficiencies.
    • Additionally, a full recognition of the role of efficiencies under Article 82 would be consistent with those (limited) Commission precedents and case law of the Community courts that support a weighing of the anticompetitive effects and claimed efficiencies. Conversely, a failure to acknowledge efficiencies would create unnecessary confusion as to the significance and scope of the efficiency defense as may be derived from those cases.

    Furthermore, we believe that efficiencies in Article 82 cases should be assessed pursuant to the following principles:

    • While efficiency claims under Article 81 are a defense against a finding of infringement under the conditions set out in Article 81(3), the assessment of efficiencies under Article 82 is an integral part of the finding of abuse. This is obvious from the wording of Article 82 (which does not include any provision mirroring Article 81(3)) in the light of Articles 2 and 3 of the Treaty. Therefore, it is for the authority investigating an alleged infringement of Article 82 to support any affirmative finding of abuse by evidence that the conduct at issue is not justified by efficiencies, in particular in those instances where the dominant company proposes a prima facie efficiency justification.
    • The recognition of the principle that conduct of dominant firms may enhance efficiency would bring EC practice in line with the litigation-oriented framework under Regulation 112003. Indeed, while per se rules and limited efficiency defenses may have some benefits from a public enforcement point of view (by reducing enforcement costs), they send the wrong signal to the business community and will create much bigger overall costs by deterring efficient conduct or undertakings. There is no sound reason why national courts should be bared from evaluating business justifications in an attempt to arrive at an appreciation of the overall impact of the positive and negative effects of dominant undertaking behaviour. Typically, under such an approach a dominant undertaking is allowed to assert and substantiate that its conduct enhances efficiency, after which the burden of proof shifts to the plaintiff to rebut that claim as being unfounded or disproportionate.
    • Efficiencies should be assessed in the same manner in all cases of alleged abuse. There is no support for the proposition that conduct restrictive of competition by companies with very high market shares (including monopolies) is unlikely to be justified by efficiency gains. Indeed, there is no correlation between market structure, on the one hand, and price competition or innovation, on the other. Under the Treaty rules, conduct that is adopted by companies with very high market shares and generates efficiencies to the benefit of consumers must be permitted even if it may lead to the elimination of (presumably less efficient) competitors.

    On refusal to license intellectual property rights

    It is a well established principle under EC competition law that inroads on the rights of intellectual property holders are only allowed under exceptional circumstances. The underlying reason for this approach is the need to preserve companies' incentives to engage in research and development and other ventures aimed at generating innovative products and services. ICC believes that the future Guidelines on Article 82 should clearly signal that the Commission intends to adhere to the criteria developed by the Community court. In this respect, ICC believes that patented and non-patented technical technology should be treated on the same footing and that the requirement that the refusal to license prevents the appearance of new goods or services be clearly set out.

    5. The Lisbon Agenda

    One of the Lisbon Agenda objectives is sustainable economic growth. It refers to stimulation of competitiveness and innovation as one of the policy tools. An economics-based approach to the application of Article 82, particularly as it takes efficiencies into consideration, is likely to promote competitiveness and growth. Moreover, by focusing on the effects on consumer welfare rather than on forms of conduct, such approach will, provided appropriate guidelines are issued, improve the regulatory environment in which undertakings operate, contribute to reduce their regulatory burden and thus allow them to become more competitive and innovative, while safeguarding consumer welfare.

    Document n° 225/623
    12 December 2005


    FOOTNOTES

    1. M. Monti, "European Competition Policy for the 21st Century, in B. Hawk (ed) 2000 Fordham Corporate Law Institute ch. 15 at 257.

    2. N. Kroes, Preliminary Thoughts on Policy Review of Article 82, Speech at Fordham Corporate Law Institute (23 September 20).

    3. E.g. Hoffman-Laroche, [1979] ECR461, para 91.

    4. E.g. AKZO, OJL1985, L 375/1 para 81.

    5. E.g. Atlantic Container Lines, judgment of the CFI of 30 September 2003 n.y.r., para. 1460

    6. See also Jacobs AG in Bronner. "...the primary purpose of Are (821 is to prevent distortion of competition - and in particular to safeguard the interests of consumers."[1998] ECR1-7791 at. 7811

    7. Continental Can, [1973] ECR215, para. 25

    8. Ibid. para. 26

    9. See e.g., Report by the Economic Advisory Group on Competition Policy ("EAGCP") (July 2005) (advocating on "economic-bed approach" to Article 82 "based on the assessment of anti-competitive effects of business behaviour").

    10. OJ 2002 C 465/6.

    11. OJ 2002 C 31/5.

    12. OJ 2004 C101/2

    13. The ECJ has defined dominance as "a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by giving it the power to behave to an appreciable extent independently of its competitors, customers and ultimately consumers". (United Brands v European Commission, [1978] ECR 207, para. 65)

    14. Hoffman-Laroche, supra n. 3.

    15. See for example, Robert C. Lind and Paul Muysert "Innovation and Competition Policy Challenges tor the New Millennium", (20031 ECLR 87; Christian Ahlborn, David Evand and Jorge Padilla, .. Competition Policy in the New Economy: Is European Competition Law up to the challenge?" in LECG Global Competition Policy, Economic Issues and Impact 2004.

    16. Commission Notice on the definition of relevant market for the purposes of Community competition law (OJ 1997 C372/5).

    17. Paras 1 and 2 of the Market Definition Notice.

    18. As was recalled by the CFI in European Night Services (199) ECR II-3141 at paras 135 and 136, the assessment of whether an agreement has restrictive effects requires that account be taken of the actual conditions in which the agreement functions, in particular the economic context in which the undertakings operate, the products or services covered by the agreement and the actual structure of the market concerned.

    19. The name is derived from the US Supreme Court case United States v. DuPont 351 US 377, (1956). There the court held that cellophane was part of a wider market including other flexible wrapping materials without realising that the price of cellophane was already at such a level that consumer were prepare to switch to other products which wold not have ben regarded as substitutes had the price ben at competitive levels. Fora detailed discussion of the cellophane fallacy see Bishop and Walker, The Economics of Competition Law (2 nd . ed) para 4.34 - 4.46 and the OFT's Discussion Paper 2, The roe of market definition in monopoly and dominance enquires (July 2001), para 2.25 et seq.

    20. Case 1V/26.699-Chiquita OJ 1976, L95/1

    21. Ibid. para II.A.2.

    22. United Brands v. Commission [1978] ECR 207 at para 31.

    23. Cases IV/30.787 and 31.488 -Eurofix-Bauco v. Hilfi, OJ 1988 L 065/19

    24. Ibid. para 61: examples are different technical possibilities of the various systems; the fact that certain characteristics differ radically and the fact that local building regulations prohibit the use of nailguns for certain applications.

    25. Hilti v. Commission, [1991] ECR II 1439 at para 73.

    26. Hugin Kassaregister AB and Hugin Cash Registers Ltd. v. Commission, [1979] ECR 995, paras 5 to 7: the Court examine the "category of clients who require (spa) part" and concluded that since there was a specific demand for Hugin's spare parts, those parts were not interchangeable with spare parts for cash registers of other makes.

    27. Ibid, par. 69 and Hilti AG v. Commission, [1994] ECR I-667, para. 13. The courts first narrowly defined the market for the main product, in which Hilti was found to be dominant The market for consumables (ie nails compatible with Hilti equipment) was defined as a separate market from the market for the equipment for which they were intended, in which Hill was also found to be dominant, although there were other players in the market

    28. Tetra Pak International SA v. Commission, [1994] ECR II-755, para. 79 to 85 where the Court dismissed the argument based on a commercial link between the machines for packaging liquid foods and the packaging itself.

    29. This wold also be the case where the spare part or consumables are protected by patents or other intellectual property rights, for example in the case of the Hilti cartridge strips (compatible with the Hilti tools) for which Hilti held patents.

    30. XXVth Report on Competition Policy, pa. 86 and 87, p. 41 and 42. This decision concerned the manufacturer of computer printers and toner cartridges for those printers.

    31. XXIXth Competition Report p.169-170. This decision concerned toner cartridges and photocopiers. The complainant alleged that there was a market for empty toner cartridges compatible with Ricoh photocopiers, an argument that was dismiss by the Commission since powder and cartridge had to be considered as a single product

    32. Hoffman La Roche v Commission, [1979] ECR 461 at para. 90; Michelin v Commission (Michelin I), [1983] ECR 3461 at para. 73; Irish Sugar v Commission, (1991) ECR 1-2969, pa. 114 and 188; Michelin v Commission (Michelin II) n.y.r. at para. 98; British Airways v Commission, (n.y.r.) at para. 247 and 271.

    33. Michelin II and British Airways supra n. 33.

    34. Michelin II. supra n. 33, para 98-110.

    35. CFI judgment of 23 October 2003 in Van den Bergh Foods (nyr).

    36. Bronner, [1998] ECR 1-7791.

    37. Hoffman La Roche supra n.33

    38. Ibid.at para.90.

    39. Ibid at para.90.

    40. Supra n.33

    41. Ibid. para 71.

    42. Michelin II supra n.33

    43. Ibidatpar-a.59.

    44. Ibid. at para. 95: "a quantity rebate system in which there is a significant variation in the discount rates between the lower and higher steps, which has a reference period of one year and in which the discount is fixed on the basis of total turnover achieved during the refer period, has the characteristics of a loyalty-inducing discount system".

    45. It should also be noted, as was pointed out ibid. at 96, that the aim of any competition on price and any discount system is to encourage the customer to purchase more from the same supplier.

    46. Ibid, at para 60, "In determining whether a quantity rebate system is abuse, it will therefore be necessary to... investigate whether, in providing an advantage not base on any economic service justing it the rebates tend to remove or restrict the buyer's from to choose his sources of supply, to bar competitors from access to the market, to apply dissimilar conditions to equivalent transactions with other trading parties or to trengthen the dominant position by distorting competition".

    47. See Van den Bergh. supra n. 36.


    ICC Comments on the European Commission discussion paper on the application of Article 82 of the Treaty to exclusionary abuses

    Executive Summary

    General Comments

    The International Chamber of Commerce (ICC) welcomes the opportunity to comment on the European Commission's ("the Commission") Discussion Paper. We hope that our comments will be helpful. We welcome the opportunity to expand upon our comments, if necessary.

    In the currently contemplated guidelines there are positive aspects, mainly in the central concern to enhance consumer welfare and to protect competition and not competitors. We recommend that such a welfare-based approach be more supported in the overall design of these guidelines. Consumer welfare does not necessarily equate to the public interest more generally, and short term consumer welfare and the wider public interest may differ. The interrelationship between the purposes of antitrust and the wider public interest should be made more explicit.

    Consistent with a welfare-based approach would be a clearer acknowledgement that as specific business conduct may simultaneously give rise to (short term) efficiency gains and (longer term) negative effects, the reviewing agency necessarily must take account of both effects in its (initial) finding of abusive behavior. Moreover, as the contemplated guidelines are likely to be relied upon by a large number of decision makers, including national court, it would be helpful if the guidelines would make clear that - in ex ante assessments of conduct under Article 82 - the finding of an abuse of a dominant position is subject to a rigorous standard of proof, relating to the successive future chain of events ultimately giving rise to the negative effects on consumers required under Article 82. In ex post reviews, a key element in the evaluation is the causal connection between the alleged abuse and those negative effects.

    Section 3: Market definition in Article 82 cases

    We invite the Commission in particular to further consider the risk of market definitions that are artificially narrow, in particular with regards to new technologies which relevant markets are, more than any other, likely to be excessively segmented.

    Section 4: Dominance

    The stress on market shares in the evaluation of dominance (paragraphs 29-33) appears in clear contrast with the conclusions of the modern theory of market leadership: market leaders have larger market shares exactly when they are constrained by effective and potential competition since, in this case, they adopt more aggressive (pricing and investment) strategies which expand their market shares. In other words, there is not necessarily a positive correlation between the presence of larger market shares and a dominant position and, especially in highly dynamic markets, there is not unambiguous theoretical support for a statement saying that "(m)arket share is only a proxy for market power' (paragraph 32). As a recent DG Competition's study on Article 821 has correctly pointed out, "the case law tradition of having separate assessments of dominance and of abusiveness of behavior simplifies procedures, but this simplification involves a loss of precision in the implementation of the legal norm. The structural indicators which traditionally serve as proxies for 'dominance' provide an appropriate measure of power in some markets, but not in others", as indeed in high-tech and New Economy industries (e.g., computer hardware and software, online businesses, mobile telephony and biotechnology).

    Section 5: Framework for analysis of exclusionary abuses

    In particular, we would encourage the Commission more fully to ensure that the interests of consumers are always paramount of those of competitors, to move even further away from form-based rules and presumptions towards a more economics- and fact-based approach, and to expand the avenues through which account may be taken of the efficiency-enhancing effects of challenged conduct.

    Section 6: Predatory pricing

    The Discussion Paper substitutes the standard Areeda-Turner test based on average variable cost ("AVC") with the average avoidable costs ("AAC"), a sort of average marginal (or incremental) cost of the extra output to serve the predatory sales. Unfortunately, the AAC can be higher than the right theoretical concept whenever it accounts for fixed costs. Moreover, the AAC can be much more difficult to measure than the AVC since it is almost always impossible to precisely define which costs are sustained for a given output and isolate the extra output (supposedly the predatory output) from total output. Finally, there are well-known conditions, as in the presence of network externalities, under which pricing below marginal cost is a normal competitive strategy for a market leader. Hence it would be better to substitute the concept of AAC with that of average variable cost, in line with the traditional economic interpretations of the Areeda-Turner test.

    Section 7: Single branding and rebates

    Overall, the Discussion Paper contemplates a more flexible approach than in the past. It appears to depart from a per se prohibition and make the assessment of rebates conditional on the existence/likelihood of foreclosure effects. In principle, the Commission intends to conduct an analysis of the market conditions in order to show that foreclosure effects are at least likely. ICC also welcomes the introduction of an efficiency defence that dominant companies can use in order to justify their rebate systems. However, several passages in the Discussion Paper seem to cast doubts on a genuine change of approach.

    Section 8: Tying and bundling

    While the Discussion Paper purports to adopt a more balanced approach that takes into account that tying and bundling can be pro-competitive, we are concerned that this approach is not carried through into the details of the analysis. A close reading suggests that certain older presumptions against tying remain embedded in the analysis, which, taken together, risk perpetuating the current situation in which tying and bundling are viewed as suspect unless proven otherwise. In our view, this would be a mistake, and we urge the Commission instead to adopt an approach that would better reflect that basic principle that tying is generally pro-competitive.

    In addition to our overarching concern that the proposed analysis fails to take account of the quite common benefits of tying, our specific concerns include: (i) the proposed "distinct products" analysis; (ii) the discussion of the "market foreclosure effect"; and (iii) the treatment of the efficiency defence.

    Section 9: Refusal to supply

    1. Controversial Issues
    2. The Section of the Discussion Paper on Refusal to Supply seems to start from the existing case-law, but still raises many controversial policy issues that, ICC submits, warrant further consideration by the European Commission, such as necessary or sufficient conditions, different thresholds, indispensable input and foreclosure effect.

      The thresholds to argue efficiencies and objective justifications seem to be too high to be realistically successful in practice. Furthermore, the Discussion Paper fails to acknowledge that an input may become indispensable simply as a result of a company's superior business performance. ICC submits that a duty to deal/supply should not be imposed simply because consumers prefer the dominant company's products.

    3. Refusal to Licence IPRs
    4. In setting out the exceptional circumstances where refusal to licence an IPR may constitute an abuse, the Discussion Paper starts from the principles and approach well-established in the case-law of the Court of Justice (notably and most recently, IMS Health). However, it then fails to give guidance on some key issues still left open by IMS Health and, in some instances, expands the scope of potential compulsory licensing to cover cases beyond the requirements of exceptional circumstances set out in IMS Health, thus potentially having a chilling effect on incentives to invest and innovate.

    Section 10: Aftermarkets

    We recall the comments made in our submission dated 12 December regarding aftermarkets2. In particular, we suggest that the Commission examine the competitive links between products and systems at the stage of market definition. The Commission would thus recognize, in line with economic analysis, that main products and their spare part or consumables should, in appropriate cases, be considered as systems which, together with other systems against which they are in competition, constitute a single relevant product market.

    We believe that the complex, multi-step analysis of aftermarkets set forth in the Discussion Paper would be both unnecessary and counterproductive. The Discussion Paper appears to acknowledge that harm to customers through actions by a supplier of aftermarket products and services is a limited concern. The only example provided is one in which a supplier adopts a "policy change" with respect to aftermarket products or services. (paragraphs 261-262). We submit that it is preferable to address this limited concern regarding "installed based opportunism" through private contracts rather than by attempting to apply Article 82 to single-brand aftermarkets and treating a "policy change" as a potential abuse of dominance.

    We hope that our comments on DG Competition's Discussion Paper will be helpful. We welcome the opportunity to expand upon our comments, if necessary.

    Document n° 225/627
    7 April 2006


    ICC Comments on the European Commission discussion paper on the application of Article 82 of the Treaty to exclusionary abuses

    Prepared by the Commission on Competition

    General Comments

    The International Chamber of Commerce (ICC) welcomes the opportunity to comment on the European Commission's ("the Commission") Discussion Paper. We hope that our comments will be helpful. We welcome the opportunity to expand upon our comments, if necessary.

    In the currently contemplated guidelines there are many positive aspects, mainly in the central concern to enhance consumer welfare and to protect competition and not competitors. We recommend that such a welfare-based approach be more supported in the overall design of these guidelines. Consumer welfare does not necessarily equate to the public interest more generally, and short term consumer welfare and the wider public interest may differ. The interrelationship between the purposes of antitrust and the wider public interest should be made more explicit.

    Consistent with a welfare-based approach would be a clearer acknowledgement that as specific business conduct may simultaneously give rise to (short term) efficiency gains and (longer term) negative effects, the reviewing agency necessarily must take account of both effects in its (initial) finding of abusive behavior. Moreover, as the contemplated guidelines are likely to be relied upon by a large number of decision makers, including national court, it would be helpful if the guidelines would make clear that - in ex ante assessments of conduct under Article 82 - the finding of an abuse of a dominant position is subject to a rigorous standard of proof, relating to the successive future chain of events ultimately giving rise to the negative effects on consumers required under Article 82. In ex post reviews, a key element in the evaluation is the causal connection between the alleged abuse and those negative effects.

    Section 3: Market definition in Article 82 cases

    The approach developed in the 1997 Communication to which the Commission's document refers3 solves most of classical cases covered by Article 82.

    Nevertheless, we believe that improvements can be made and we refer to the suggestions made in our previous comments on the reform of Article 82.4 We invite the Commission in particular to further consider the risk of market definitions that are artificially narrow, in particular with regard to new technologies which relevant markets are, more than any other, likely to be excessively segmented.

    An over-subjective definition of the market according to the criticized abuse should also be avoided as much as possible. One should rely on objective criteria. In this respect, it is regrettable that the Commission plans on too often setting aside the SSNIP test, which offers companies a certain predictability and assists competition authorities in evaluating the market under a dynamic and realistic perspective. The Commission correctly identifies a central problem of market definition in relation to dominant companies and notes in its discussion of the SSNIP test and the cellophane fallacy that the test is inappropriate. However, there does need to be some test and no alternative is proposed. This demonstrates the weakness of the current position and the unpredictability of the law.

    Moreover, the application of a test premised solely on product characteristics may well result in an overly narrow market definition. This could lead to erroneous findings of dominance in the overly narrow market.

    Section 4: Dominance

    Following a traditional definition, Section 4 of the Discussion Paper associates dominance with "a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by affording it the power to behave to an appreciable extent independently of its competitors, its customers and ultimately of the consumers" (paragraph 20). Such a definition requires "a leading position on that market" compared to the rivals (paragraph 22) and the lack of "effective competitive constraints" (paragraph 23) in the process in which "the undertaking and the other players act and inter-act on the market"( paragraph 23).

    Given the positive emphasis put on an economics-based approach to competition policy, it is important to notice that this definition of dominance is clearly associated with two situations examined by economic analysis: the pure monopoly, as an extreme case of dominance, and the market leadership where the dominant undertaking faces some competitors, which is clearly the most interesting case. It should be noticed that, according to standard economic analysis, a market leader can really act independently of its rivals (so as to satisfy the above condition for dominance)5 only when the number of competitors is exogenously set and further entry is impossible, while a market leadership constrained by effective competition and potential entry cannot be associated with dominance: in this case, modem economic theory tells us that leaders tend to be aggressive (pro-competitive) in their pricing and investment strategies, conquering larger market shares in a way that has nothing to do with dominance as defined above, and which is also beneficial to consumers.6

    As a consequence of the approach of the Discussion Paper, it would be better to eliminate a certain ambiguity in the statement at paragraph 27 which says that "the fact that an undertaking is compelled by the pressure of its competitors' price reductions to lower its own prices is in general incompatible with [...] the existence of substantial market power' and hence with dominance. In particular:

    1. this should be always true and not just "in general", since in this case the market leader is constrained by effective competition and cannot act independently from it, as the definition of dominance would require;
    2. this should be extended to any other form of aggressive competition that is not only competition on prices, but also competition on quantities or on alternative forms of strategic investments.

    Hence, the fact that an undertaking is compelled by the pressure of its competitors' aggressive strategies to adopt aggressive (pricing and investment) strategies should be always incompatible with dominance.

    The emphasis on market shares in the evaluation of dominance (paragraphs 29-33) appears in clear contrast with the modern theory of market leadership: market leaders have larger market shares exactly when they are constrained by effective and potential competition since in this case they adopt more aggressive (pricing and investment) strategies which expand their market shares. In other words there is not necessarily a positive correlation between the presence of larger market shares and a dominant position and, especially in highly dynamic markets, there is no unambiguous theoretical support for a statement saying that "[m]arket share is only a proxy for market power" (paragraph 32). As a recent DG Competition's study on Article 827 has correctly pointed out, "the case law tradition of having separate assessments of dominance and of abusiveness of behavior simplifies procedures, but this simplification involves a loss of precision in the implementation of the legal norm. The structural indicators which traditionally serve as proxies for 'dominance' provide an appropriate measure of power in some markets, but not in others", as indeed in high-tech and "new economy" industries (e.g., computer hardware and software, online businesses, mobile telephony and biotechnology).

    Finally, the part on dominance clearly refers to competition in the market, while it is hardly useful to evaluate cases where competition for the market takes place. In these cases, typical of the New Economy, competition is dynamic and innovators conquer large parts of a market, so that any static analysis of market shares cannot say anything about dominance. In other words, a market can be currently dominated by a single firm, but if many other firms which are not even active in this market are investing in R&D to enter into it, as it happens in many high-tech sectors, this market is substantially competitive in a dynamic sense. Nevertheless, any leader in such a competitive winner-takes-all market would be always characterized as dominant by the static and market-share-based approach of the Discussion Paper.

    Moreover, modern economic theory tells us that in these dynamic sectors market leaders, as long as they are constrained by effective competition in the market for innovations, invest more than their competitors and hence are more likely to remain leaders.8 In this sense, statements saying that "high market shares, which have been held for some time, indicate a dominant position" can be true in some sectors, but not in high-tech sectors with competition for the market. That is, in dynamic markets, incumbents, with rare exceptions, are under permanent threat of entry and must continue to innovate if they wish to maintain this incumbency. In conclusion, the general impression is that there is an excessive reliance on market shares to evaluate dominance, and that this can be highly misleading especially for dynamic markets.

    We agree that market share is logically a criterion of limited significance within a system where the definition of the relevant market is questionable as a result of the weakness of the SSNIP test as an aid to market definition.

    The part on barriers to expansion and entry (paragraphs 34-40) concerns a concept which is far from unambiguous in economic theory. The definition of these barriers as "factors that make entry impossible or unprofitable while permitting established undertakings to charge prices above the competitive level" (paragraph 38) applies to legal barriers but not to other factors which are sometimes seen as barriers. For instance, high fixed costs of production and R&D or investments needed to develop network externalities or learning by doing advantages, do not make entry impossible: the correct definition in these cases would be that these factors endogenously limit entry or endogenously determine how many and which firms profitably enter. The difference is not just in the definition but also in the economic consequence, since modem economic theory has shown that when entry is impossible market leaders may behave in an anti-competitive way, but when entry is constrained by technological or demand conditions they (always) behave in a pro-competitive way even if such factors limit entry and the market leaders obtain high market shares.

    It should also be noted that barriers to entry can be cumulative, which is a point not covered in the Discussion Paper. It should also be noted that legal barriers may have effects long after their formal removal, as in the case with post patent right protection.

    Section 5: Framework for analysis of exclusionary abuses

    Section 5 of the Discussion Paper sets out the basic analytic framework that the European Commission intends to use in analyzing exclusionary abuses under Article 82. We welcome the Discussion Paper's statement at the outset that the essential objective of this analytic framework "is the protection of competition on the market as a means of enhancing consumer welfare and of ensuring an efficient allocation of resources."9 We likewise agree that "the purpose of Article 82 is not to protect competitors from dominant firms' genuine competition based on factors such as higher quality, novel products, opportune innovation or otherwise better performance, but to ensure that these competitors are also able to expand in or enter the market and compete therein on the merits, without facing conditions which are distorted or impaired by the dominant firm."10

    Despite these welcome pronouncements, we have some concern that they are not fully carried through into certain aspects of the analytic framework. In particular, we would encourage the Commission more fully to ensure that the interests of consumers are always paramount of those of competitors, to move even further away from form-based rules and presumptions towards a more economics- and fact-based approach, and to expand the avenues through which account maybe taken of the efficiency-enhancing effects of challenged conduct. We address each of these issues in turn.

    1. Promoting interests of consumers over competitors

    The analysis of whether an undertaking has engaged in abusive conduct under Article 82 should ultimately turn on the conduct's actual effects on efficiency and consumer welfare. Thus, if the pro-consumer benefits of a dominant undertaking's conduct are significant, it should be immune from liability even if it disadvantages certain competitors. As we noted in our December 2005 Comments, inventing better products or more efficient methods of distribution, reducing prices or offering better terms of trade, and more quickly adapting to changes in the market can disadvantage rivals and maybe even cause them to exit the market. Yet these forms of conduct often also enhance efficiency and consumer welfare.

    This focus is particularly important with respect to fast-moving markets such as those commonly found in high-tech and "new economy" industries (e.g., computer hardware and software, online businesses, mobile telephony and biotechnology). These industries are often characterised by massive R&D investments, strong reliance on IPRs and other intangible assets, network effects, high fixed sunk costs and low marginal costs. Competition in these markets is dynamic in the sense that competition often takes place for the market in a "winner-takes-all" race. Leading firms in these markets might enjoy high market shares yet be subject to massive competitive pressure to constantly create better products at lower prices due to threats from innovative competitors and potential entrants. Undertakings that hold a significant share of the market at any given point of time may see this share decrease rapidly and significantly following the development and supply of a new and more attractive product by an actual or potential competitor.

    In certain respects, the analytic framework set forth in the Discussion Paper provides grounds for optimism that the Commission is moving toward a stronger focus on consumer welfare. Yet others aspects of the framework suggest that competitors' interests will at times trump those of consumers and force dominant undertakings to forego competitive behaviour that in fact would generate efficiency gains or promote consumer welfare. For example:

    • In spelling out the concept of foreclosure, the Discussion Paper states that "it is sufficient that the rivals are disadvantaged and consequently led to compete less aggressively."11 This proposition gives cause for concern. First, this statement is not consistent with standard economic theory which has made clear that an aggressive behaviour of the market leader inducing a less aggressive competition of its competitors is not sufficient to create any harm to consumers (actually the net effect is typically the opposite happens).12 The inconsistency of this statement is even more clear when it is claimed that "[r]ivals may be disadvantaged where the dominant company is able to ... reduce demand for the rivals' products" (paragraph 58) which is really what any aggressive or pro-competitive strategy would do. Putting together the two sentences, we are told that in order to establish foreclosure it would be sufficient that the strategy of the dominant firm reduces demand for the rivals' product: but this amounts to banish any pro-competitive strategy by market leaders. Moreover, the above statement could arguably support the conclusion that a dominant undertaking in a market characterized by network effects could be guilty of abuse if it is able to attract new customers on the basis of a new, superior technology. This view is contrary to the basic principle that dominant undertakings should be permitted--and indeed encouraged--to compete aggressively on the merits. Allowing a finding of abuse merely where competitors are "disadvantaged" would penalise dominant undertakings for engaging in a wide range of conduct that is ultimately pro-competitive. In our view, this aspect of the analytic framework should be revised to clarify that conduct by a dominant undertaking would be deemed to be an abuse only if its net effect is to harm consumer welfare.13
    • The analytic framework posited by the Discussion Paper is particularly troubling in dynamic markets where competition is often for the market. Competitors in these markets invest vast amounts in research and development with the hope of winning a large portion of the market. Some succeed while others, ultimately, do not. Dynamic markets are unique in this way, in that there is no sustainable market equilibrium with a number of players of differing sizes co-existing in the market. Rather, there are successive innovation races, resulting in "winners and losers" as par of a Schumpterian "gale of creative destruction". The analytic framework presented in the Discussion Paper runs the risk of interfering with this natural competitive process and therefore, inefficiently obstructing the workings of a dynamic market.
    • On a related point, the analytic framework seems to rest on an assumption that, because conduct that harms competitors may perhaps decrease consumer welfare in the longer term, any efficiencies generated by such conduct should be discounted.14 In our view, such an assumption is unwarranted. Accurately predicting the magnitude of long-run harm to competition or consumers resulting from conduct that is otherwise efficiency-enhancing is almost always a difficult and uncertain undertaking. Such predictions are particularly unreliable with respect to dynamic markets and run the serious risk of under-estimating the capacity of rivals and new entrants to exert competitive pressures through product innovation or other means. Accordingly, we would urge the Commission not to assume long-term harm to consumers from immediate impact of the conduct on one or more competitors, but rather to examine, in each particular case, whether there is any evidence supporting the view that the impact on competitors will cause long-run harm to consumers and whether such harm, if any, exceeds the short-run increases in consumer welfare and both short- and long-run efficiency gains attributable to the dominant undertaking's conduct.
    • The Discussion Paper states that the Commission may at times prohibit the use of price discounts where doing so will "protect competitors that are not (yet) as efficient as the dominant company."15 In our view, there is no economic justification for barring dominant undertakings from decreasing prices simply in order to protect less efficient rivals-- particularly since such a prohibition will mean that these rivals will face even less competitive pressure to become more efficient. This condition also places dominant undertakings in the untenable position of having to guess what level of rival inefficiency will be used to judge whether the dominant undertaking's own efficiency-enhancing conduct is lawful.
    • The Discussion Paper also states, in its discussion of the "meeting competition defence", that a dominant undertaking has an obligation to weigh "the interests of its competitors to enter or expand" into the market when deciding upon alternative courses of action, and that dominant undertakings can only benefit from this defence if they prove there was no less anti-competitive alternative.16 In the real world, the best businesses are focused on advancing the interests of their customers, not their competitors--which, of course, is one sign of a competitive market. Thus, most dominant undertakings will be ill-equipped to evaluate which of various possible options will least disadvantage their competitors. We would therefore recommend that this requirement that dominant undertakings weigh the interests of competitors be dropped from the analysis.
    • The Discussion Paper states that, where an undertaking holds a market share above 75%, any pro-competitive efficiencies generated by the conduct in question will be given lower priority than the conduct's impact on competitors.17 In our view, undertakings--whether dominant or not--should never be under an obligation to place the interests of their competitors over those of consumers. Such a rule will end up protecting less efficient rivals and restricting the behaviour of dominant undertakings in a manner that undermines Article 82's purpose of promoting efficient markets and consumer welfare. Protecting rivals against competition in this manner will also reduce their incentives to compete aggressively.
    • We have some questions about the Discussion Paper's statements regarding presumptions of abuse at paragraph 60. If, as the first sentence postulates, certain exclusionary conduct "is clearly not competition on the merits," "clearly creates no efficiencies" and "only raises obstacles to residual competition,"18 such conduct will almost certainly be abusive and we cannot imagine why a "presumption" is necessary. If, however, this statement is meant to signal that the Commission intends to look to the form of challenged conduct in making an initial assessment of abuse, and that it will then fall to the dominant undertaking to rebut that presumption through factual evidence, we would disagree with this approach for the reasons already noted and described in more detail below. Also, we would note that, in the interests of legal certainty and business guidance, it would be more helpful if the Discussion Paper were to set out circumstances in which abuse cannot be found, rather than, as in paragraph 60, cases in which the Commission will necessarily assume that an exclusionary abuse has occurred.

    In sum, we would recommend that any final Article 82 guidelines move even further away from the more traditional focus on protecting competitors and instead assess whether the conduct is likely to promote or impede efficiency and benefit or harm consumers.

    2. Greater reliance on economics-based approach

    In the past, EU competition policy has been criticised for focusing more on the form of unilateral conduct than on its actual effects in the marketplace. As we noted in our December 2005 Comments, there is broad consensus among economists that (unilateral) price- and non-price conduct of dominant undertakings may produce both pro- and anti-competitive effects. The ambiguous nature of conduct of dominant undertakings militates in favour of a full appreciation of the (positive and negative) effects on consumers. It is therefore vital that the framework for analysis under Article 82 provides for a rigorous, economics-based examination of the market context in which unilateral conduct occurs.

    For instance, the Commission should clarify that, despite the references to the "form and nature" of conduct in the general discussion of exclusionary abuses,19 whether market foreclosure will be found to exist will ultimately turn on the likely or actual effects of the conduct in the marketplace. Also, while we commend the Commission for placing less reliance on per se rules and irrebutable presumptions of market foreclosure and abuse, the Discussion Paper retains elements of this approach. For example, in several places, certain forms of conduct or market shares will make it "highly unlikely" that some legal determination will result.20 We would urge the Commission to lessen its reliance even on these quasi-per se rules and to adopt a more thoroughgoing, economics- and effects-based analysis that focuses on increasing consumer welfare and is based on sound economic theory of the behaviour of market leaders and on solid empirical analysis.21

    Furthermore, we are not persuaded that an approach based on the weighing of pro- and anticompetitive effects will decrease legal certainty. As we noted in our December 2005 Comments, much of the current uncertainty about the boundaries between permissible and prohibited business practices results from a form-based approach to certain pricing practices and the difficulty inherent in such an approach in determining whether new kinds of economic activity should be regarded as being of one type of form or another. Form-based approaches lack consistent and rigorous analysis of the concrete effects of a given practice and often have the effect of condemning profit-maximizing conduct that benefits consumers. The uncertainty that results from the condemnation of conduct that may not have any significant impact on competition or that may benefit consumers creates added risks for business, which itself reduces efficiency, and deters undertakings from applying business practices (e.g. certain pricing schemes) which in fact increase competition and are beneficial for consumers.

    In sum, we would urge the Commission to make it clear that unilateral conduct whose benefits to efficiency or consumers outweigh its negative impact on competitors is not an abuse, whatever its form and regardless of the degree of market power of the undertaking concerned.

    3. Expand opportunities to take account of efficiencies

    Unilateral conduct that enables an undertaking to operate more efficiently normally results in direct consumer benefits because it allows the undertaking to increase output and/or lower its prices. Unilateral conduct can also promote dynamic efficiency by freeing up resources for increased research and innovation, or to develop improved methods of production and distribution. Indeed, the consideration of efficiencies in the assessment of conduct under Article 82 reflects the role of undisturbed competition as a means towards the achievement of the broader Treaty objectives set out in Article 2.

    For these reasons, conduct that generates efficiencies should not, in our view, be deemed abusive unless it is demonstrated that the impact of this conduct on competition will result in consumer harm outweighing these efficiencies. While the Discussion Paper acknowledges that promoting efficiency is one of the primary objectives of Article 82,22 the framework for analysis itself actually provides relatively limited scope for taking efficiencies into account. This manifests itself in a variety of ways;

    • Burden of proof. The Discussion Paper indicates that, consistent with existing practice, it will fall on dominant undertakings to prove the extent to which their conduct was justified on grounds of efficiency.23 As noted in our December 2005 Comments, the final burden of proving efficiencies should be placed on the authority investigating the alleged abuse because, in stark contrast to the bifurcated approach under Article 81, the assessment of efficiencies is an integral par of the assessment whether any given conduct amounts to "abuse" under Article 82. More importantly, bringing efficiencies into the analysis only as an affirmative defence will send the wrong signal to the business community. It means that investigations will often have moved quite far along before efficiency considerations fully come into play. Placing the burden of proof on competition authorities, by contrast, makes more sense as they are likely to be in a better position to obtain relevant evidence from the dominant undertaking as well as other market participants (such as consumer organizations) on whether challenged conduct promotes efficiency--and have the expertise and resources to undertake such an inquiry. Accordingly, we believe it is for the authority investigating an alleged infringement of Article 82 to support any finding of abuse by evidence that the conduct at issue is not justified by efficiencies, in particular in those instances where the dominant undertaking proposes a prima facie efficiency justification.24 The legal burden of proving an infringement of Article 82 must always rest on the authority or part alleging the infringement, in line with the legal framework of Article 82 (which differs from Article 81) and the express wording of Article 2 of Regulation 112003. There is a need to distinguish between the legal burden of proof and the evidentiary burden. Only the latter may shift to the dominant company once the part or the authority alleging the infringement has proved its existence to the required legal standard.
    • Narrow scope of efficiency defence. To assert a successful efficiency defence under the proposed analytic framework, dominant undertakings will be required to show that their conduct was "indispensable" in order to achieve the resulting efficiencies and that "competition in respect of a substantial portion of the products concerned (was) not eliminated."25 To meet the first of these conditions, the defendant must "demonstrate that there are no other economically practicable and less anti-competitive alternatives to achieve the claimed efficiencies."26 This condition means that liability could be imposed even on conduct whose efficiency and consumer benefits far outweigh its adverse effect on competitors simply because there exists an alternative that would have disadvantaged rivals less. We wonder whether such rule would have any economic justification and any basis in commercial reality. Wouldn't it at most, merely provide an excuse for rivals to second-guess the business decisions of their dominant competitors? The second condition is equally troubling. There seems to be an inherent contradiction between the "no -- elimination -- of -- competition" prong of the analysis, which is taken from Article 81(3), and the very concept of dominance. As a result, would efficiency claims by dominant undertakings, particularly those with high market shares, not be systematically given short shift because of the difficulty of satisfying this condition? In essence, would dominant undertakings not be effectively required to place the interests of competitors and the competitive process over the interests of efficiency and consumer welfare?
    • Virtual exclusion of efficiency defence for monopolies. The Discussion Paper seems to suggest that, where a dominant undertaking holds a market share above 75%, the protection of competitors will be given priority over efficiency. In our view, efficiencies should be assessed in the same manner in all cases, regardless of the defendant's market share. Under the Treaty, and consistent with the goals of Article 82 as described by Commissioner Kroes, undertakings that generate pro-competitive efficiencies that benefit consumers should not be penalised regardless of the level of market share or potential impact on less efficient competitors. Moreover, the Discussion Paper introduces a concept of market position "approaching that of a monopoly" (paragraph 92), for market shares above 75%, for which no economic analysis is presented. Moreover, does economic analysis justify any separate treatment for undertakings with high market shares? As shown by the modem economic theory, market leaders tend to have higher market shares exactly when they face effective competitive pressure which induces them to adopt aggressive (pricing and investment) strategies and hence to expand their market shares in a pro-competitive way27. Under these conditions, exceptionally high market shares (but not monopolistic ones) can be due to relevant scale economies or to the existence of "learning by doing" or network effects. The existence of these high market shares should not exclude the undertaking from using the efficiency defence.

    In conclusion, we would urge the Commission to clarify that conduct by a dominant undertaking, regardless of its form and irrespective of the undertaking's market share, could be deemed abusive only if the efficiency gains or consumer benefits generated by such conduct were outweighed by the negative effects of such conduct on the competitive process and consumer welfare.

    Section 6: Predatory pricing

    Predatory pricing is defined in the Discussion Paper as "the practice where a dominant company lowers its prices and thereby deliberately incurs losses or foregoes profits in the short run so as to eliminate or discipline one or more rivals or to prevent entry by one or more potential rivals thereby hindering the maintenance or the degree of competition still existing in the market or the growth of that competition" (paragraph 93) . The Discussion Paper uses a number of cost benchmarks in order to assess whether "predatory pricing" by a dominant undertaking has actually taken place.

    Pricing below average avoidable cost ("AAC") gives rise to a rebuttable presumption that the pricing is "predatory". Average avoidable cost is the average of the costs that could have been avoided if the undertaking had not produced a discrete amount of extra output (this extra output is usually the amount allegedly subject to abusive conduct). Apparently, this principle is supported by the idea that pricing below marginal cost could only have a predatory purpose, but the concept of marginal cost is difficult to measure. This is in line with the standard economic theory and antitrust doctrine coming from Areeda and Turner (who noticed that "the incremental cost of making and selling the last unit cannot readily be inferred from conventional business accounts, which typically go no further than showing observed average variable cost ("AVC"). Consequently it may well be necessary to use the latter as an indicator of marginal cost".28 However, the Discussion Paper substitutes the standard Areeda-Turner test based on AVC with the MC, a sort of average marginal (or incremental) cost of the extra output to serve the predatory sales. Unfortunately, the MC can be higher than the right theoretical concept whenever it accounts for fixed costs. Moreover, the MC can be much more difficult to measure than the AVC since it is almost always impossible to precisely define which costs are sustained for a given output and isolate the extra output (supposedly the predatory output) from total output. Finally, there are well known conditions, as in presence of network externalities, under which pricing below marginal cost is a normal competitive strategy for a market leader. Hence it would be better to substitute the concept of MC with that of average variable cost, in line with the traditional economic interpretations of the Areeda-Turner test.

    According to the Discussion Paper, where pricing is above MC, but below average total cost ("ATC"), predation cannot be presumed. ATC is the average of the variable and fixed costs incurred by a company. Pricing above ATC is in general not considered predatory, but according to the virtually unanimous economic literature, it would be better to state explicitly that pricing above ATC is never predatory since it cannot lead to foreclosure of 'as efficient' competitors.

    In certain sectors, the Commission uses a long-run average incremental cost benchmark ("LAIC"), instead of AAC. This is usually the case in industries where fixed costs are high and variable costs very low. In these cases, the LAIC benchmark is used as the benchmark below which predation is presumed. The same considerations as above hold also here: there are not economic justifications for a change of standard from AVC to LAIC. Moreover, we believe that the LAIC standard is inconsistent with business reality because it requires companies to price to cover average sunk fixed costs that are unrecoverable: this approach ignores the economic reality that, when businesses decide how to price a product, they do not consider costs that are "sunk" or "unrecoverable," even if not a single product is sold.

    In the Discussion Paper, a dominant undertaking may, even if the price is below the relevant cost benchmark, rebut a finding of predatory pricing by providing a "justification" for its pricing behaviour (this is a departure from earlier case-law, where pricing below AVC was considered to be abusive per se). The Discussion Paper lists several examples of possible justifications (paragraph 131), including an issue of re-start up costs or strong learning effects, the need to sell off perishable or obsolete stock and the justification that the low price is a short-run loss minimising response to changing conditions in the market (including those resulting from a dramatic fall in demand leading to excess capacity). We suggest that besides learning effects, the proposed guidelines should include network effects, whose theoretical role in justifying aggressive pricing is very similar to that of learning by doing and is recognized by standard economic theory.

    Finally, the scope of objective justifications should not be unduly restricted to few "acceptable" justifications in terms of productive efficiency. The possibility should be left to provide any objective justification satisfying the necessity and proportionality requirements.

    Dominance itself should not be sufficient to establish the likelihood of recoupment, particularly in technology markets. For example, looking forward one or two years in the dominance inquiry is not sufficient to undertake a proper assessment of recoupment where significant uncertainty abounds regarding not only cost and demand but the existence of potential entrants. It is entirely possible that a firm may be dominant in the sale and/or distribution of a given product, yet be constrained by entrants with highly disruptive technologies which require greater than one or two years to mature and be successfully commercialised.

    Section 7: Single branding and rebates

    Overall, the Commission's approach seems more flexible than in the past. It appears to depart from a per se prohibition and make assessment of rebates conditional on the existence/likelihood of foreclosure effects. In principle, the Commission intends to conduct an analysis of the market conditions in order to show that foreclosure effects are at least likely. ICC also welcomes the introduction of an efficiency defence that dominant undertakings can use in order to justify their rebate systems. This is an improvement to the "old" position where the only possible efficiency defence consisted of economies of scale linked to the adoption of volume rebates.

    However, several passages in the Discussion Paper cast doubts on a genuine change of approach. The introductory chapter contains several 'statements of principle' about potential 'negative' effects of rebates that seem more in line with the 'old' approach of a per se prohibition. At paragraphs 148 and 149, the Commission alleges that "the dominant position already enables the dominant company to prevent effective competition to be maintained or to emerge in the market and it thus becomes particularly important to protect the limited degree of competition still existing in the market and the growth of residual competition... Where the dominant company applies a single branding obligation to a good part of its buyers and this obligation therefore affects, if not most, at least a substantial part of market demand, the Commission is likely to conclude that the obligation has a market distorting foreclosure effect and thus constitutes an abuse of the dominant position." At paragraph 139, the Commission also mentions that rebates are likely to foreclose competitors when they maintain or strengthen the dominant undertaking's position thereby "hindering the maintenance of growth of residual or potential competition." Later on, the Discussion Paper is even clearer when mentioning at paragraph 158 that "([i]n case the threshold(s) are formulated on terms of percentage of total requirements of the buyer or an individualized volume target, the Commission will normally presume that they are set at that level to hinder customers to switch and purchase additional amounts and thus to enhance loyalty."

    Article 82 pre-supposes the existence of a dominant undertaking. By essence, the adoption of rebate systems by dominant undertakings will likely have a 'distorting' effect on competitors, at least in the short-term, and will likely impact on limited residual competition. Under the above-mentioned principles, most rebates applied by dominant undertakings will be deemed abusive without the need of much further market analysis. At paragraph 146, the Discussion Paper gives examples of situations where rebates are unlikely to have foreclosure effect. However, it is difficult to understand how these markets could give rise to dominance issues in the first place. The paragraph mentions that rebates will not have foreclosure effects if "competitors are competing on equal terms for all the customers" One may wonder how this statement may be reconciled with the necessary prerequisite of dominance that supposes the dominant undertaking's 'power to behave independently of competitors.'

    ICC would like to stress the following points:

    • First, 'distorting' effect on competitors does not necessarily mean 'abusive. Findings of abuse should be based on a longer-term market assessment that should take into account competitors' likely response to the rebate system, customers' ability to switch and long- term benefits for end-users. The Discussion Paper refers to some of these factors but fails to give sufficient guidance on how it intends to apply them.
    • Second, ICC has problems understanding why it would be abusive for dominant undertakings to try to 'maintain or strengthen' market shares through the adoption of rebates. There is a fundamental ambiguity throughout the Discussion Paper in admitting that dominant undertakings are allowed to compete (even aggressively) and, at the same time, considering that dominant undertaking's strategies to maintain or gain market share are likely to be abusive. Competition is generally about increasing market shares to the detriment of competitors. The Discussion Paper seems to assume that, once in a dominant position, undertakings should stop expanding and 'freeze' their commercial behaviour to whatever is strictly necessary to meet competition. This is difficult to reconcile with a night to aggressively compete. In addition, such a position is highly unrealistic. The Discussion Paper should clarify that, in principle, dominant undertakings should be allowed to compete aggressively on rebates since this may lead to long-term aggressive price competition. Rebates would become abusive in limited defined circumstances when competitive strategies are not 'on the merits' or involve predatory or other anticompetitive behaviour resulting in likely foreclosure effects, for example, "full line forcing".
    • Third, the Commission should give some benchmark on the size of the tied market that is likely to involve negative effects. In addition, the Discussion Paper mentions that one way for dominant undertakings to avoid tying a significant part of the demand is to selectively apply the rebates to some customers. However, in that case, dominant undertakings may be caught between a rock and a hard place, on one end the loyalty-enhancing effect of fidelity rebates, on the other the risk of price discrimination. In this respect, ICC welcomes comments recently made by a Commission official that 'real' price discrimination cases should be limited to wide price differences with significant distorting effect and would appreciate if the draft guidelines could include and clarify that aspect.

    1. Conditional rebates on all purchases

    ICC is not certain that the Commission's approach in assessing the likelihood of an abuse is based on a well-structured operational test. The Discussion Paper mentions that rebates are likely to be abusive when:

    1. they apply to all purchases (including past purchases) made within the reference period;
    2. the threshold is set at a level that induces switching customers to buy additional quantities from the dominant undertaking;
    3. competitors' required share exceeds the commercially viable amount per customer, as calculated by the Commission on the basis of the dominant undertaking's ATC and the effective price of the last slice of the rebate, as follows. If the basic price of the dominant undertaking is p, the percentage rebate is r, total sales are S+X, of which S is the threshold above which rebates start and X are the extra sales beyond the threshold, then the effective price for the same extra fraction of sales is given by the difference between total price with the rebate p(1-r)(S+X) and the total price of the threshold quantity without rebates pS, divided by the extra quantity X:

      where we define x=X/(S+X) as the fraction of extra sales. When this average price is below the average total cost ATC entry is foreclosed. Assuming that ATC is constant and equal to C, the necessary condition can be easily derived. A competitor with the same average total cost of the dominant undertaking could profitably enter only selling at least the fraction:

      which is defined as the required share.

    4. the rebate system ties a significant part of buyers; and
    5. there is no clear indications of significant entry or customers' switching.

    As a general comment, ICC considers that the system proposed by the Commission is far too complex and seems to leave room for a large margin of error and uncertainty. ice considers that only factors (iv) and (v) are reliable indicators of the existence of foreclosure effects. However, although the Discussion Paper refers to these elements in several passages, it does not give much guidance on how the Commission intends to apply them. ICC would welcome some examples/benchmark in order to help companies to quantify the degree of foreclosure that is likely to be held abusive (notably when the Discussion Paper specifies that "a significant part" of customers should be tied by the rebates). On the other hand, factors under (i), (ii) and (iii) seem highly unreliable for the following reasons:

    • They seem to be based on three artificial assumptions: (i) a dominant undertaking's demand is automatically divided between an inelastic and elastic part (ii) buyers automatically switch once the tied market share is exceeded; and (iii) buyers are automatically 'sucked in' the rebate system once they get close to the threshold. These assumptions are artificial because they suppose that buyers always react in the same way to price variations without taking into account competitors' response or demand fluctuations. In addition, the Discussion Paper focuses on buyer's reactions to prices only without taking account of their reactions to competition on quality and innovation.
    • The rebates under discussion are substantially equivalent to a simple quantity discount, which, as is well known, has a welfare enhancing role.29 The implication is that similar rebates should never be considered abusive when the percentage rebate is small enough since they have similar effects to simple quantity discounts.
    • Any kind of fidelity rebate can have a pro-competitive role in the sense that it creates a further dimension of competition (the non-linear price schedule) and it can represent a more aggressive pricing strategy: hence an additional minimal condition for rebates to be abusive should be that competitors are not able to propose similar rebates or different ones (with different thresholds),30 but this issue is absent from the list of conditions.
    • It also seems questionable from an economic point of view to calculate the effective price on the last slice of the rebate since one may argue that the threshold is exceeded by all the customers' purchases, and not only by the last slice.
    • The theoretical formulation of point (ii) is largely unrelated or inconsistent with standard economic theory and it is affected by a dangerous theoretical problem. The derivation of the formula for the required share is valid assuming that the ATC is constant. However, in general the average total cost is not constant and typically U-shaped, so that undertakings producing different amounts have different ATCs and there is usually a minimum ATC associated with a certain scale of production. Suppose that ATC depends on the fraction of sales x according to a general relation c(x) - the particular case where this is constant is assumed by the Discussion Paper.31 Now, according to the reasoning of the Discussion Paper, foreclosure would require:

      In general, this is not equivalent to a cut-off rule for which the competitors' required share exceeds the commercially viable amount per customer. For instance, if we are in the range for which ATC is increasing (which is likely to be the relevant one if we are referring to a dominant firm with a large market share) we actually have the opposite result.32 In general, applying mechanically the cut-off rule suggested in the Discussion Paper is theoretically inconsistent and can lead to completely ineffective conclusions whenever the market is characterized by a minimum efficient scale.33

    • The calculation method proposed by the Discussion Paper is far too complex to be applied by business people in their daily practice. When negotiating rebates with customers, one may wonder how dominant undertakings could reasonably project whether or not the last slice of the rebate will exceed their ATC at the particular point in time when the customer precisely buys the last slice. In most cases, dominant undertakings will not be able to determine whether their ATC is below the effective price because they will lack information on their competitors' "commercially viable shares/ required shares". As a result, dominant firms wishing to ensure compliance with Article 82 may elect not to make use of conditional rebate programmes even when they would enhance consumer welfare.
    • The calculation method also opens the door to legal uncertainty. According to the Commission, calculating the competitors' "commercially viable test" is the crucial element of the equation. If the CVT is inaccurate, would the whole equation have any value in terms of assessing potential abusive conduct or foreclosure effect? In several parts of the Discussion Paper, the Commission seems to admit that it is extremely difficult to calculate CV with accuracy. (paragraph 157 refers to the need of 'revising' the CV, paragraph 163 admits that there must be cases where "entry or expansion of competitors is in effect not limited to the amount assessed by the Commission as the commercially viable share."; paragraph 164 mentions that "where it is not possible to establish accurately the required share...., the Commission may use cost data of apparently efficient competitors".) This raises a number of questions. In most cases, it will be difficult for the Commission to find competitors as efficient as the dominant undertaking. In the alternative, the Discussion Paper does not specify on what criteria these competitors will be considered "as efficient". Last, the reference to competitors that are "as apparently efficient" leaves significant room for discretion and flexibility and thus legal uncertainty.
    • There is a risk that the selected operational test may tip the burden of proof quite heavily against the dominant undertaking. Under the "new approach", presumed abusive and foreclosure effects will be mathematically calculated by the Commission on the basis of an equation, the main parameters of which (required share, commercially viable share and sometimes ATC) will be unknown to or difficult to establish by the dominant undertaking (see paragraph 163.)

    The Commission should also clarify in what order it intends to apply the above listed tests. At paragraph 164, the Discussion Paper suggests that the Commission will investigate the performance of the dominant undertaking and competitors when information on costs is not available. ICC believes that this comparison should be done before any cost calculation is made since it may be a more reliable indicator of the existence of foreclosure effects. In general, ICC suggests that, before entering into a cost analysis, a more pragmatic approach could be for the Commission to identify whether the rebate system involves some significant foreclosure effects, e.g., by comparing the evolution of the dominant undertaking's market share and competitors' market shares before and after the adoption of the rebate system, determining the percentage of total customers tied by the rebate, and identifying customers' switching ability.

    This is particularly troubling in innovative markets and even more so in so-called "pure-play" innovation companies (companies that engage in research and development activities but neither produce nor distribute the resulting products). It is critical for these companies to find ways to incentivise their downstream partners to grow the market for the technology in question, solve potential underinvestment problems resulting from the inability to appropriate the know-how typically transferred to manufacturers and distributors to cover the fixed costs of their investments in research and development.

    Loyalty rebates are particularly important to a pure-play innovation company and ultimately for end-consumers. Consumers may benefit, for example, from a manufacturer having a low marginal input cost, when that low cost is passed on to the consumers. This in turn will provide the manufacturer with an incentive to expand sales by competing on price. Additionally, loyalty rebates may facilitate efficient recovery of fixed costs. In general, consumers will face higher prices where an innovator needs to charge higher prices - resulting in lower volume - in order to recover fixed research and development costs. A loyalty rebate scheme allows the innovator to charge a relatively high price for the non-contestable share of the market, where demand is relatively inelastic, while charging a lower price (after loyalty rebates) for the contestable part of the market, where demand elasticity is higher. The company can simultaneously profit from a higher margin on the infra-marginal units without losing volume at the margins.

    2. Conditional rebates on incremental purchases

    ICC would welcome some specific examples or guidance showing in practice how this type of rebates is likely to have foreclosure effect. The wording of the Discussion Paper is vague and would leave significant flexibility in assessment. It should be kept in mind that conditional rebates on incremental purchases are exactly equivalent to quantity discounts whose welfare enhancing role is well known in economic theory.34

    For example, if the threshold is set in terms of a percentage of total requirements or individualized volume targets, the Discussion Paper refers to a presumption of abuse when the resulting price is predatory (however, predation in that case is not based on MC but on ATC) and the tied demand is important enough to create a foreclosure effect. If the threshold is set in terms of a standardized volume target, the rebate system will normally not have a loyalty-inducing effect with some exceptions, e.g., when it targets customers that are of particular importance for the possibilities and expansion of competitors.

    None of these concepts are quantified or illustrated by examples. Therefore, they remain highly subjective. In particular, importance of customers is a flexible concept that varies from one customer to the next.

    3. Unconditional rebates

    ICC understands that this type of rebates is generally unproblematic except if the dominant undertaking targets some important customers. As mentioned above, ICC considers that the 'importance' of customers is a subjective concept, which should be clarified. Otherwise, it will give competitors significant flexibility to argue that unconditional rebates are abusive by targeting an 'important' customer.

    4. Efficiencies

    Although ICC welcomes the introduction of an efficiency defence in the context of Article 82, the examples contained in paragraphs 172-176 seem to introduce a very limited defence. As regards the first example, it is extremely difficult for dominant undertakings to quantify which amount of cost efficiencies is specifically linked to a specific percentage of the rebate grid. The two other examples (rebates applied to large retailers and relationship-specific-investment) refer to specific situations and offer limited guidance in practice.

    Section 8: Tying and bundling

    We welcome the Discussion Paper's recognition that tying and bundling are often pro-competitive and its movement away from the per se approach to these practices reflected in prior case law. Indeed, economists today generally acknowledge that tying often produces positive efficiencies and consumer benefits.35 The pro-competitive effects of tying are particularly pronounced in the case of technical tying (when companies innovate by linking formerly separate technologies or products, efficiencies often emerge through improved performance and quality), but they also emerge because tying is often used as an aggressive strategy which leads to lower pnces.36

    While the Discussion Paper purports to adopt a more balanced approach that takes into account that tying and bundling can be pro-competitive, we are concerned that this approach is not carried through into the details of the analysis. A close reading suggests that certain older presumptions against tying remain embedded in the analysis, which, taken together, risk perpetuating the current situation in which tying and bundling are viewed as suspect unless proven otherwise. In our view, this would be a mistake, and we urge the Commission instead to adopt an approach that would better reflect that basic principle that tying is generally pro-competitive.

    In addition to our overarching concern that the proposed analysis fails to take account of the quite common benefits of tying, our specific concerns include: (i) the proposed "distinct products" analysis; (ii) the discussion of the "market foreclosure effect"; and (iii) the treatment of the efficiency defence. We discuss each in turn.

    1. Distinct Products Test

    While we fully agree with the Discussion Paper's emphasis on consumer demand (and independent supply to the extent that it reflects that demand) in assessing whether a tying arrangement might be abusive, we are concerned that the Discussion Paper places too much emphasis on consumer demand for the tied product. Such demand does not shed light on whether there exist distinct products for the purposes of tying analysis, which uses the distinct products test as a proxy for determining whether the tying arrangement produces efficiencies. In other words, while there is clearly consumer demand for shoelaces, this should not mean that shoes and shoelaces are distinct products for the purposes of tying analysis. This issue can only be addressed by asking whether there is consumer demand for shoes without shoelaces.37 In sum, whether or not consumer demand exists for the tied product is the wrong question; the correct question is whether there is any significant consumer demand for the tying product without the tied product. Unless the analysis focuses on this question, there is a danger that the mere existence of consumer demand for the tied product may prevent the emergence of efficient tying arrangements and end up protecting suppliers of tied products at the expense of consumers and innovation.

    We are also concerned that, in the case of technical integration of two products that were previously distinct, the distinct products test itself may not be helpful for understanding market dynamics because, by definition, this test is backward-looking. As a result, the Discussion Paper's proposal that consumer demand be considered in such cases is particularly troubling.38 A better approach in these cases would be simply to ask whether the undertaking integrating the previously distinct products can make a plausible showing of efficiency gains. Since technical tying is normally efficient, market-leading undertakings would be able to continue producing innovative products benefiting consumers without running afoul of the prohibitions on tying.

    2. Market Foreclosure Effect

    1. In Genera
    2. The Discussion Paper also provides that a tying arrangement would be prohibited if it "is likely to have a market distorting foreclosure effect that would result in harm to consumers."39 Yet the description of what constitutes such foreclosure is vague. Market foreclosure effects are described as conduct that has "the capability, by its nature, to foreclose competitors from the market." Total foreclosure is not necessary - it is enough if competitors are "disadvantage."40 The Discussion Paper goes on to state that a tying practice will be presumed to result in market-distorting foreclosure where it ties a "sufficient part" of the market, but fails to provide guidance as to the meaning of "sufficient."41

      Under this vague foreclosure standard, any tying arrangement that has the effect of reducing demand for a competitor's product could be deemed to have a prohibited foreclosure effect, irrespective of the benefits to consumers or whether these effects are solely the result of competition on the merits. Unless clearer guidance is provided on the degree of foreclosure that is presumed to give rise to anti-competitive effects, undertakings will be left in a state of uncertainty in assessing tying arrangements. To give undertakings as much concrete guidance as possible, it would be helpful to have more precise indication of the degree of foreclosure that is considered to be abusive.

      The Discussion Paper fails to provide clear guidance on the effect of bundling by competitors of the dominant undertaking on the analysis of market foreclosure. At one point, it suggests that bundling is less problematic if competitors also offer bundles.42 At another point, it indicates that the foreclosure effect might be greater if others in the industry also bundle.43 We believe this inconsistency should be resolved in favor of the former position: the fact that other undertakings in the market also offer bundles is a presumption that bundling generates efficiencies and meets consumer demand - if not, bundling by the dominant undertaking would provide competitors with a great opportunity to differentiate their offerings and make them more attractive to consumers. Additionally, the dominant undertaking ought to be able to compete with bundles offered by its competitors.

      Finally, the Discussion Paper's treatment of "commercial usage" in the context of market foreclosure does not reflect the economics of tying. According to the Discussion Paper, the sale of a tied product by a dominant undertaking may be an abuse, even when it is standard commercial practice.44 Furthermore, the fact that a competitor ties may add to the foreclosure effect.45 As mentioned above, the Discussion Paper overlooks that in practice the customary nature of bundling is evidence that such tying generates efficiencies, or that there is no demand for the unbundled product. If there were sufficient customer demand to make the supply of the unbundled product profitable, competitors of the dominant undertaking would most likely avail themselves of this business opportunity.

    3. Foreclosure by Mixed Bundling
    4. While we agree with the Discussion Paper's general approach to determining when a discounted or "mixed" bundle might give rise to foreclosure, we disagree with the specific test proposed. The Discussion Paper provides the following guidance on the point at which a mixed bundle might give rise to foreclosure: "[c]ompetitors are foreclosed if the discount is so large that efficient competitors offering only some but not all of the components, cannot compete against the discounted bundle."46 The Discussion Paper then indicates that such foreclosure will exist unless "[t]he incremental price that customers pay for each of the dominant company's products in the bundle ... cover[s] the long run incremental costs of the dominant company of including this product in the bundle"47

      We believe that the long-run incremental costs standard is inconsistent with business reality because it requires companies to price bundles to cover sunk fixed costs that are unrecoverable. This approach ignores the economic reality that, when businesses decide how to price a product, they do not consider costs that are "sunk" or "unrecoverable," even if not a single product is sold.

      The Discussion Paper's reliance on long-run incremental costs to measure foreclosure is based on the assumptions that an automobile factory can be converted into a semiconductor plant and that a steel worker can be retrained to become a software engineer without cost. These assumptions do not reflect economic reality. The impracticality of making economic decisions based on "long run" analysis was perhaps best articulated but John Maynard Keynes, who said "in the long run, we are all dead."

      We believe that a more appropriate cost standard in this case would be marginal costs ("MC'') or at least Average Avoidable Costs ("AAC''). When business people decide whether or not to make a marginal sale at a particular price, they generally consider the marginal cost of making that sale. We note that the Discussion Paper uses AAC as the appropriate measure of cost in its predatory pricing guidelines48 and that same reasoning supports AAC as the appropriate measure of cost in the mixed bundling context. Indeed, the only time the Discussion Paper uses the long-run average incremental cost measure in the predatory pricing context is when addressing pricing by monopolies that are (or were) established by law.49

    5. Standard of Proof
    6. We are concerned that the standard of proof the Commission is required to meet to establish harmful foreclosure effects is too low, particularly in light of the fact that the analysis of foreclosure effects can be speculative in nature. In the case of tying, actual market foreclosure effects are not required by the Discussion Paper- it is enough that such effects are "likely"50 to occur. In other words, the mere risk of foreclosure can result in a finding against a dominant undertaking. In merger cases, the European Court of Justice ("ECJ") has held that the Commission must put forward convincing evidence to block a merger, as the Commission is trying to predict the future effects of the merger on a market.51 As the analysis of market foreclosure effect under Article 82 will often entail a prediction of future effects, the Commission should set a similarly high standard of proof for tying cases.

      In establishing foreclosure in a tying case, the Commission must address a chain of causation that is similar to that involved in a conglomerate merger case, which, in the words of the ECJ, are "dimly discernable, uncertain and difficult to establish"52 Establishing foreclosure not only requires the Commission to predict what will happen in the future if the tying practice continues, but requires it to establish that the dominant firm has the ability and the incentive to leverage its dominant position on the tying product's market to foreclose competition on the tied product's market. A standard of proof that requires convincing evidence will help ensure that companies will not be deterred from bringing new products to market as a result of concerns about remote, potential foreclosure effects.

    3. Efficiencies

    As discussed above, we are concerned about both the burden of proof placed on the dominant undertaking as well as the standard of proof that it must meet to establish the existence of efficiencies. Procedural rules that create presumptions against the dominant. undertaking are particularly out of place in the case of tying and bundling practices, which are recognized to be pro-competitive in most cases.

    We are also concerned that the Discussion Paper fails to acknowledge that bundling can be used to create value for consumers in markets that experience network effects or in multi-sided markets. In fact, with regard to network effects, the Discussion Paper indicates that foreclosure effects of a bundle may be greater when there are network effects.53 In such markets, bundling is a valuable strategy to gain broader distribution of the products or service that is subject to network effects. And the broader the distribution, the greater the value produced for all consumers. This is particularly true when the product or service in question has a low (or no) marginal costs, because the supplier can costlessly include the product or service in bundles with other products. In this respect, the Discussion Paper appears to advocate an interpretation of Article 82 outlawing business practices that create wealth for society and large consumer benefits.

    Similarly, we believe that it should be acknowledged that bundling can generate efficiencies in multi-sided markets, i.e. markets where products or service must be matched with other products or service to have value. Newspapers exist in multi-sided markets. Newspapers are sold to readers, but they also sell advertising space to advertisers. The reader is not only a "customer" of the newspaper, the reader is also a supplier of "eyes" that the newspaper sells to advertisers. The complex business models resulting from multi-sided markets often require bundling practices because the consumption on one side of the market is being "sold" on the other side of the market, and piece-meal consumption on one side of the market breaks down the interdependent ecosystem.

    In conclusion, while we recognize and welcome the shift in the Discussion Paper away from rigid, per se rules and presumptions, we would urge the Commission to pursue this further, including by more fully taking account of the pro-competitive benefits of tying and bundling and by expanding the avenues through which these benefits may inform the analysis.

    Section 9: Refusal to supply

    The Discussion Paper distinguishes between termination of an existing supply relationship and de novo refusals to start supplying an input.

    The European Commission would introduce a rebuttable presumption that continuing a supply relationship is pro-competitive (paragraph 217). Four conditions are set out for an abuse to occur in case of termination of an existing supply relationship:

    • The behaviour must amount to termination (including delaying tactics, excessive pricing, unfair terms, margin squeeze);
    • Dominance must be established in an "upstream" input market (but it could also be a distinct market where access is needed to link with another market, e.g., to interface information);
    • There must be a likely market distorting foreclosure effect; and
    • There are no objective justifications and no efficiencies (for example, lack of commercial assurances to fulfill obligations or plans to integrate downstream).

    The threshold is higher for an abuse to occur in cases of de novo refusals to start supplying an input. The following five conditions need to be met:

    • Behaviour is a refusal to supply (incl. delaying tactics, excessive pricing, unfair terms);
    • Dominance must be established in an "upstream" input market (but it may also be a captive, potential or hypothetical market);
    • The input is indispensable (in the sense that there are no real or potential substitutes in the market and it is impossible or extremely difficult or expensive for competing companies to duplicate the input);
    • There must be a likely market distorting foreclosure effect; and
    • There are no objective justifications and no efficiencies (for example, no commercial assurances to fulfill obligations, lack of capacity constraints, unreasonable cost increase in access).

    Controversial Issues

    The Section of the Discussion Paper on Refusal to Supply seems to start from the existing case-law, but still raises many controversial policy issues that ICC submits warrant further consideration by the European Commission:

    • Necessary or sufficient conditions: It is not clear whether the conditions for finding an abuse are necessary or simply sufficient. The Discussion Paper qualifies that "normally" those conditions must be fulfilled; therefore, it seems to leave open the possibility that on a case-by-case basis the Commission could identify other criteria beyond those listed above. This significantly undermines legal certainty and potentially leads to open-ended cases of intervention.
    • Different thresholds: The Discussion Paper does not explain the basis for the view that, in general, continuing a supply relationship should be presumed to be pro-competitive. We consider that this should be the result of a case-by-case assessment of the economic circumstances of each case and not the subject of a legal presumption. Several Member States have specific provisions on issues of economic dependency outside the scope of competition law; any concerns arising out of the termination of existing supply relationships could be adequately dealt with in that context rather than reversing the burden of proof by introducing a pro-competitive presumption and lowering the threshold for competition law intervention. We submit that the threshold for intervention should be the same as for cases of de novo refusals to supply and, therefore, the requirement that the input is indispensable should also be added for termination of existing supply relationships.
    • Objective justifications: In paragraph 224, the Commission states, in the context of an objective justification for a termination of a supply relationship, that "the dominant company may also argue that it is terminating the supply relationship because it wants to integrate downstream and itself perform the downstream activities. In such a situation it falls upon the dominant company to show that consumers are better off with the supply relationship terminated". This presumption the Commission stipulates can hardly be reconciled with established principles of competition law. In a market system of free competition, even a dominant company must, at least in principle, be allowed to freely decide upon its sales strategy and distribution system. If it decides to change its distribution policy, e.g., to terminate existing distribution contracts and to establish a direct sales organisation, it is its own choice for which it bears responsibility54. Competition law is' not meant to guarantee an existing distributor relationship once and for ALL. As long as the supplier does not act in order to discipline a specific distributor and as long as the necessary termination periods A3 observed (depending on the given set of facts, the length may vary), there is no reason to intervene by means of Article 82. Therefore, it should not "fall upon the dominant company to show that consumers are better off with the supply relationship terminated" as the Commission proposes. If there be a presumption at all, it should be in favor of the company's freedom to decide upon its distribution strategy. Only in the case where the terminated dealer can show that he was disciplined or discriminated, the supplier might be required to justify the termination.
    • Indispensable input: The definition of the indispensable input does not address the necessary economic analysis that should be carried out to decide whether duplication of the input is impossible, difficult or expensive for any competitor or for "as efficient" competitors. ICC submits that the Commission should clarify that the focus of the analysis should be on whether a second, substitute product can be created by "as efficient competitors, rather than whether any competitors will in fact make the investment to create it. This approach would be consistent with the Oscar Bronner case and with the Commission's objective of protecting competition on the merits.
    • Foreclosure effect: The standard for intervention by the Commission is not fully developed. In particular, the Discussion Paper does not give sufficient guidance on the degree of the likely anti-competitive foreclosure effects in the market. It states that the market distorting foreclosure effect should not be understood to mean the complete elimination of all competition, but it does not specify to what extent competition in the downstream market should be affected for an abuse to be found. Furthermore, there is no mention of the economic analysis that should be carried out of the effects of the abuse, in particular to assess the degree of efficiency of excluded competitors. ICC therefore submits that the Commission should clarify that the foreclosure effect should be substantial and at least amount to the creation of dominance in the downstream market (in terms of price increases or output reduction) resulting from the exclusion of "as efficient" competitors.
    • Finally, the thresholds to argue efficiencies and objective justifications seem to be too high to be realistically successful in practice. Furthermore, the Discussion Paper fails to acknowledge that an input may become indispensable simply as a result of a company's superior business performance. ICC submits that a duty to deal/supply should not be imposed simply because consumers prefer the dominant undertaking products.

    Refusal to Licence IPRs

    Compulsory licensing of intellectual property rights is a very sensitive and controversial area under Article 82 and, therefore, deserves particular attention. As already explained in ICC's comments submitted to the European Commission on 12 December 2005, ICC believes it is important to preserve companies' incentives to engage in research and development and other ventures aimed at generating innovative products and services. ICC welcomes a number of pronouncements in the Discussion Paper that appreciate the benefits of the IPR regime and IPR protection. The European Commission acknowledges that:

    • The indispensable input is often the result of substantial investments enabling significant risks (paragraph 235);
    • In order to maintain the incentive to invest and innovate, the dominant company must not be unduly restricted (paragraph 235);
    • There is no general obligation for the IPR holder to license the IPR (paragraph 238);
    • The very aim of the exclusive right is to prevent third parties from using the IPR without consent (paragraph 238); and
    • Refusal to license an IPR does not in itself constitute an abuse (paragraph 239).

    In setting out the exceptional circumstances where refusal to license an IPR may constitute an abuse, the Discussion Paper starts from the principles and approach well-established in the case-law of the Court of Justice (notably and most recently, IMS Health). H9wever, it then fails to give guidance on some key issues still left open by IMS Health and, in some instances, expands the scope of potential compulsory licensing to cover cases beyond the requirements of exceptional circumstances set out in IMS Health, thus potentially having a chilling effect on incentives to invest and innovate.

    1. Exceptional Circumstances

    The Discussion Paper sets out that the refusal by a dominant company to license access to an IPR could be considered abusive when the five conditions for de novo refusal to start supplying an input are satisfied, AND "the refusal to grant a licence prevents the development of the market for which the licence is an indispensable input, to the detriment of consumers". The threshold for intervention in cases of refusals to license IPRs is therefore higher than in other cases of refusals to supply. In summary, the conditions under the Discussion Paper are as follows:

    • The behaviour can be properly characterised as a refusal (again, including cases of constructive refusals such as delaying tactics in supplying, imposing unfair trading conditions, or charging excessive prices for the input);
    • The company refusing to license must be dominant in the market where input is provided;
    • The input must be indispensable (i.e., it must not be possible to turn to any workable alternative technology or to "invent around" the IPR - the Discussion Paper mentions as examples cases where the technology has become the standard or where interoperability is necessary);
    • The refusal is likely to have a negative effect on competition;
    • There is no objective justification; AND
    • The additional condition is that the refusal prevents the development of new goods or services and for which there is a potential consumer demand.

    2. Unresolved Issues

    The Discussion Paper does not give guidance on some of the key issues left open by IMS Health:

    • Dominance in an upstream market: The dominance requirement as set out in paragraph 227 of the Discussion Paper broadens the scope of potential reach of compulsory licenses for IPRs that have no commercial or independent use (i.e. that are not marketed separately), but are only used as an input in other commercial products or services. Under IMS Health, there must be two identifiable markets as a necessary condition for IPR compulsory licensing. The Discussion Paper states that it is sufficient to identify a "captive", "potential" or even a "hypothetical" upstream market, and that "such is the case where there is actual demand for the input on the part of the undertakings seeking to car out the activity for which the input is indispensable" . This broad construction can lead to a greater number of compulsory licensing of IPRs (provided the other conditions are met) by covering IPRs that are only used as an input without the need to identify a distinct product or service that . would be sold or licensed separately. Furthermore, there is no reference or explanation in the Discussion Paper of the qualification given by the Court of Justice in IMS Health that the potential market must at least correspond to an identifiable "stage of production". Finally, there is no economic assessment of the conditions under which holding an IPR could amount to. market power, which should be the correct framework of analysis without any presumption that holding an IPR may automatically give rise to market power. ICC submits that without further qualification, such a potentially broad application of Article 82 could have a negative impact on incentives to invest in developing IPRs and investing in new production processes and research.
    • "New product" requirement: There is no explanation of the requirement that the refusal to license must prevent the appearance of new goods or services. The Discussion Paper says that the company requesting the licence should not limit itself to the duplication of goods/services already offered. However, it does not provide any guidance on the criteria to identify or define a "new" product. ICC submits that the Commission should clearly specify that it must be a new kind of product (rather than just an incremental or minor improvement of an existing product) that must expand the market rather than steal sales. In this respect, it would helpful to clarify, consistently with IMS Health, that the new product should satisfy consumer demand that is not satisfied by existing products.

    3. Concerns

    Despite having some deference for IPRs in a number of welcome pronouncements as explained above, the Discussion Paper does not fully carry them through and goes significantly beyond the exceptional circ~mstances for compulsory licensing set out in IMS Health. For example, the Discussion Paper advocates the position that a refusal to license is abusive if it is likely to have a negative effect on competition, while in IMS Health, the EC) required an elimination of competition. ICC is concerned about the following sections that may carry the risk of reducing the incentives to invest and innovate in the long term:

    • For follow-on innovations. the additional condition that the refusal prevents the development of new goods or services is not necessary. Paragraph 240 of the Discussion Paper states that "a refusal to licence an IPR protected technology which is indispensable as a basis for follow-on innovation by competitors may be abusive even if the licence is not sought to directly incorporate the technology in clearly identifiable new goods and services. The refusal of licensing an IPR protected technology should not impair consumers' ability to benefit from innovation brought about by the dominant undertaking's competitors." This goes much further than the exceptional circumstances set out in IMS Health and the statements of principle in the Discussion Paper. This would be a worrying departure from the established principles of the European case-law, because it effectively means the introduction of open compulsory licensing to competitors for a myriad of IPRs. Furthermore, the Discussion Paper does not define what could amount to "follow-on innovation" and does not explain why intervention is required in this area to bring benefits to consumers. Finally, inefficient competitors may effectively have the possibility to free- ride on the investments and risks taken by a dominant undertaking. For all these reasons, companies may be deterred from investing and innovating in the first place, with a potential much bigger negative impact on consumers in the medium-long term.
    • For refusal to supply information needed for interoperability. the Discussion Paper (paragraphs 241-242) says that (a) leveraging market power from one market to another may be an abuse of a dominant position and (b) it may not be appropriate to apply the same high standards for intervention even if such information may be considered a trade secret. The Commission does not develop the framework for assessing where/how such leveraging may occur, nor does it substantiate why trade secrets do not deserve the same high standards for protection. Again, such a broad policy intervention could have chilling effects on incentives to invest and innovate and could ultimately end up protecting inefficient competitors that may free ride on the risks and investments of the dominant undertaking, therefore in contradiction with the Commission's objective of protecting competition on the merits.

    Section 10: Aftermarkets

    We recall the comments we made in our submission dated 12 December regarding aftermarkets55. In particular, we suggest that the Commission examine the competitive links between products and systems at the stage of market definition. The Commission would thus recognize, in line with economic analysis, that main products and their spare parts or consumables should, in appropriate cases, be considered as systems which, together with other systems against which they are in competition, constitute a single relevant product market.

    As the Discussion Paper notes, it is common for the supplier of such equipment to have "a very strong position" in the sale of "secondary" products and services used with its own brand of equipment (paragraph 253). Indeed, undertakings with smaller positions in the primary equipment market may have even larger shares of their brand's aftermarket because third-part suppliers and other primary market firms typically focus on the most successful equipment brands since those brands provide the largest aftermarket revenue opportunity. As a result, there is a risk that undertakings with quite modest positions in the primary market would be viewed as dominant in the aftermarket if the assessment were to be focused only on an aftermarket consisting of products and services for their individual brand of equipment.

    We believe that the Discussion Paper is correct in emphasizing that the "secondary markets" should not be viewed in isolation since "the actual degree of market power of the supplier (in the aftermarket) ... may be constrained by competition in the primary market." (paragraph 246). As the Discussion Paper explains, "competition in the primary market may make price increases in the aftermarket unprofitable due to its impact on sales in the primary market, unless prices in the primary market are lowered to offset the higher aftermarket prices." (paragraph 246). This fundamental insight regarding the key relationship between the primary market and any related aftermarkets means that a separate examination of a single brad aftermarket under Article 82 is rarely, if ever, appropriate.

    The Discussion Paper appears to accept this conclusion for "customers who may buy the primary product in the future" since competition in the primary market will protect such customers (paragraphs 254-259)56. However, the Discussion Paper draws a distinction between "future customers" and "prior purchasers" on the basis that "competition in the primary market does not protect customers who have already bought the primary product." (paragraph 254). We believe that the distinction in the Discussion Paper between "future customers" and "prior purchasers" is misguided57. Since every "prior purchaser' was, by definition, a "future customer' before it acquired the primary product, competition in the primary product market also protects this subset of customers. In addition, as noted in the Discussion Paper, the "prior purchasers" are also protected by the supplier's interest in its reputation with respect to its aftermarket pricing and practices because its reputation will affect its future sales of the primary product as well as its future sales of other equipment that requires afermarket products and services (paragraph 262).

    We believe that the complex, multi-step analysis of aftermarkets set fort in the Discussion Paper58 is both unnecessary and counterproductive. The Discussion Paper appear to acknowledge that harm to customers through actions by a supplier of aftermarket products and services is a limited concern. The only example provided is one in which a supplier adopts a "policy change" with respect to aftermarket products or services (paragraphs 261-262).

    However, such a change is likely to take place only in very unusual circumstances - where both (1) the entire primary market is declining or the particular supplier has decided to exit or is losing market share and (2) the relevant supplier is not engaged in other equipment markets and thus would not be deterred by the impact of the "policy change" on its reputation (paragraph 262). Even in those very limited situations, there would be no harm to customers if the customers utilized the primary market competition to protect themselves by contract when they purchased the equipment (paragraph 263). We submit that it is preferable to address this limited concern regarding "installed based opportunism" through private contracts rather than by attempting to apply Article 82 to single-brand aftermarkets and treating a "policy change" as a potential abuse of dominance59. Otherwise, there is a risk that suppliers will be deterred from adopting more open and flexible afermarket policies in the first place if future changes in those policies will subject them to a risk of costly investigations, fines, and private damages actions for violation of Article 8260 .

    There is a risk that the Discussion Paper's focus (for example, paragraph 247) on customers who have already purchased the primary product will lead to an over-restrictive analysis on the basis of alleged "lock-in". First, the supplier would need to be able to discriminate against the so-called "locked-in" customers so as not to prejudice sales in the primary market. Secondly, the practical possibility of switching to a different "system" would need to be analysed and not just by reference to up-front purchase costs. This latter point is relevant, for example, in markets where the customer already owns and uses different (competing) systems, for example machinery used with consumable products, and can switch between them whenever the price of the consumables for one system is increased without the need to make a further capital investment (which may be significant in comparison with the increase in the price of the consumables) (see paragraph 249).

    Moreover, the supplier risks losing sales in the primary market going forward if, having acquired a sufficiently large "installed base" to make discrimination worthwhile, it then increases prices in the secondary market to customers who genuinely are locked-in, for example because switching costs are too high. The supplier may, however, then suffer reputation damage which reduces future demand in the primary market. This point appears to be missing from the discussion in paragraph 254 - see also paragraph 261 which postulates a change in policy by the supplier--


    FOOTNOTES

    1. Patrick Rey (Coordinator), Report by the EAGCP 'An EamomÙ" Approach to Article 82 :. July, 2005.

    2. Pages 9-11.

    3. Paragraph 12 and following of the Discussion Paper

    4. ICC, Comments on the Reform of the Application of Article 82 of the EC Treaty (12 December 2005), pp.7 to 11.

    5. And potentially it can implement anti-competitive strategies, that is engage in abusive conduct.

    6. See Franco Modigliani (1958), "New Developments on the Oliopoly Front', Journal of Political Economy, 66, 3, June, pp. 215-32, and Federico Etro (2006), "Aggressive Leaders", Rand Journal of Economics, V01. 37, Sprig.

    7. Patrick Rey (Coordinator), Report by the EAGCP 'An Economic Approach to Article 82', July, 2005.

    8. See Etro, "Innovation by Leaders", Economic Journal, Vol. 114,281-310 (2004).

    9. Discussion Paper, paragraph 54 (emphasis added).

    10. Ibid. (emphasis added).

    11. Ibid., paragraph 58.

    12. As pointed out by well established economic doctrine (Drew Fundenberg. and Jean Tirole, 1985, "The Fat Cat Effect, the Puppy Dog Ploy and the Lean and Hungry Look", American Economic Review, 74 , May, pp. 361-68), an aggressive behaviour of the market leader can lead to more aggressive competition by a competitor (generally under competition in prices) or to a less aggressive one (typically under competition in quantities) with positive consequences for the consumers in the first case and only ambiguous ones in the second. Moreover, when entry of competitors is endogenously taken into account (which should be the relevant case), an aggressive behaviour of the leader does not affect each single competitor but can reduce entry, with net effects for consumer welfare and allocation of resources which are always positive (Etro, 2006). Hence, an aggressive behaviour of the market leader inducing less aggressive competition of the competitors is not sufficient to create any harm to consumers or to deteriorate the allocation of resources.

    13. See, e.g., Commissioner Neelie Kroes, Preliminary Thoughts on Policy Review of Article 82, at 3 (23 Sept. 2005) (stating that, in the analysis of exclusionary conduct under Article 82, "ultimately the aim is to avoid consumers' harm").

    14. See,, e.g., Discussion Paper, paragraphs 54-60.

    15. Discussion Paper, paragraph 67.

    16. Ibid., paragraphs 82, 83.

    17. Ibid., paragraphs 91-92.

    18. Ibid., paragraph 60.

    19. Ibid., paragraphs 58, 59.

    20. See. e.g., ibid, paragraphs 30, 90-91.

    21. The recent DG Competition's study on Article 82 (Rey et al., Report by the EAGCP 'An Economic Approach to Article 82,'" July 2005) correctly emphasizes the need of solid theoretical and empirical foundations in the antitrust procedure: "a natural process would consist of asking the competition authority to first identify a consistent story of competitive harm, identifying the economic theory or theories on which the story is based, as well as the facts which support the theory as opposed to competing theories. Next, the firm should have the opportunity to present its defense, presumably to provide a counter-story indicating that the practice in question is not anti-competitive, but is in fact a legitimate, perhaps even pro-competitive business practice."

    22. See, e.g., ibid, paragraph 4 ("With regard to exclusionary abuses the objective of Article 82 is the protection of competition on the market as a means of enhancing consumer welfare and of ensuring an efficient allocation of resources.") (emphasis added).

    23. Ibid., paragraphs 77, 79.

    24. See footnote 20.

    25. Ibid., paragraph 84.

    26. 1bid., paragraph 86.

    27. See the discussion on Dominance, above.

    28. See Philip Areeda and Donald Turner (1975), "Predatory Pricing and Related Practices under Section 2 of the Sherman Act", Harvard Law Review, 88, pp. 637-733. See also F. Etro (2006), "Competition Policy: Toward a New Approach", European Competition Journal, Vol. 2, April, in press.

    29. For instance see Massimo Motta, Competition Policy. Theory and Practice, Cambridge: Cambridge University Press, 2004 (Chapter 7).

    30. This requirement is also strengthened by the general statement for which "only conduct which would exclude a hypothetical 'as efficient' competitor is abusive" (paragraph 63).

    31. This relation is motivated as follows. What is behind the discussion on rebates is a market with many customers each one buying a number of products. As long as the total number of customers is given and they have similar demand, x is directly related to total demand, on which a proper average cost function depends.

    32. Formally if c(x) is U-shaped, the relation holds for required shares within a closed set and not above a threshold. Hence a mechanic application of the formula may lead to derive the upper bound of the set rather than its lower bound) leading to completely wrong conclusions in every case!

    33. Or whenever the market is not a natural monopoly) that is, in every relevant case for our purposes.

    34. Again, see Motta (2004, Chapter 7).

    35. DO Competition's own study on Article 82 notes that cases of anti-competitive tying are "relatively scarce." See Rey et al., Report by the EAGCP 'An Economic Approach to Article 82,' My 2005, at 39.

    36. The Chicago school has advanced efficiency rationales in favour of bundling with positive (or at worst ambiguous) consequences on welfare, including production or distribution cost savings, reduction in transaction costs for the customers, protection of intellectual property, product improvements, quality assurance and legitimate price responses. The post Chicago approach has shown that, when the bundling firm has some market power, bundles can have a predatory purpose (Michael Whinston, 1990, "Tying, Foreclosure and Exclusion", American Economic Review, 80, pp. 837-59), but in general, tying should be submitted to a rule-of-reason standard. However, more recently, the modem theory of market leaders has emphasized that bundling by the incumbent I) is just an aggressive (pro-competitive) strategy of the incumbent for a competitive tied product market, 2) may not have a specific entry deterrence purpose, and 3) may increase welfare even without taking efficiency reasons into account. Technically, bundling works as a commitment device to be aggressive, that is to produce more for the secondary market and hence to be able to adopt a lower price. As a consequence, the leader can exploit larger scale economies, reduce the average price level for the consumers and hence increase welfare (see Etro, 2006).

    37. See Commission notice - Guidelines on Vertical Restraints, paragraph 216.

    38. Ibid., paragraph 187 (noting that analysis will consider "whether consumer demand has shifted as a consequence of the product integration so that there is no more independent demand for the tied product").

    39. Ibid., paragraph 183.

    40. Ibid., paragraph 58.

    41. Ibid., paragraph 188.

    42 . Ibid., paragraphs 195,202.

    43. Ibid., paragraph 197.

    44. Ibid., paragraph 182.

    45. Ibid., paragraph 197.

    46. Ibid., paragraph 189.

    47. Ibid., paragraph 190.

    48. Ibid., paragraph 108.

    49. Ibid., paragraphs 124-26.

    50. Ibid., paragraph 183.

    51. CaseC-12/03, Commission v. Tetra Laval, Judgment of 15 Feb. 2005 (not yet reported).

    52. Discussion Paper, paragraph 44.

    53. Paragraph 199.

    54. Langen, Bunte, European Competiton Law, 10 th. Edition 2006, note 168 to Article 82.

    55. Pages 9-11.

    56. Customers can protect themselves by comparing the lifecycle costs of the products thereby taking into account both the initial cost of the primary product and the anticipated aftermarket costs over the useful life of the product. Often customers use long-term maintenance contracts or other contractual guarantees regarding the lifecycle cost of the product to protect themselves against the impact of subsequent changes in aftermarket prices or policies. Sophisticated customers (referred to in the Discussion Paper as "professional buyers" (paragraph 258)) are more likely to engage in this practice than individual consumers. However, consumers may use extended warranties or other contractual arrangements to protect themselves. Even where customers lack information or sophistication to compare lifecycle costs or to negotiate contractual protections, competition among suppliers will normally protect purchasers. Since suppliers can be expected to understand the long-term revenue opportunity flowing from the sale of the primary product, they are likely to compete aggressively in pricing their primary products in the expectation of obtaining a stream of aftermarket revenues from the primary equipment sale. See paragraph 259.

    57. Of course, many cu~tomers are repeat purchasers of the primary product and thus are both "prior purchasers" and potential "future customers". A customer may well be able to protect itself with respect to the impact of "policy changes" with respect to its prior purchases when it negotiates for its next purchase of the primary product. See paragraph 254, footnote 146.

    58. The first step in the approach set fort in the Discussion Paper is to determine whether there is a separate single-brad aftermarket. This focuses only on customers that have already purchased the primary product and asks whether it is possible for such customers (1) to switch to the secondary products provided by other primary market suppliers or (2) to switch to another brand of the pómary product in order to defeat an attempt by the supplier of the primary product to increase prices of its secondary products or services. In many cases, this step will lead to defining a single-brand aftermarket. Only in the second step of the proposed approach -- determining "dominance" in the single-brand aftermarket -- is the impact of competition in the primary market taken into account. Here, the Discussion Paper appears to require a separate assessment be made of "customers who have already bought the primary product" (paragraph 254) and appears to treat ease of entry as the only factor that would keep a supplier with a "very strong position on the aftermarket" for its own brand of equipment from being viewed as "dominant" for this group of customers. While suggesting that "the weaker the position of the supplier in question on [the primary] market" the "less likely it is that the supplier in question can be considered dominant on the aftermarket" for its brand of equipment (paragraph 260), in fact the analysis in the Discussion Paper leaves such suppliers exposed to a finding of dominance and of abuse of dominance if they "decide to change policy and raise prices in the aftermarket or restrict the possibilities of other suppliers in the aftermarket." (paragraph 261). In sum once such a supplier has begun to deal with others in connection with the aftermarket for its brand of equipment, the analysis in the Discussion Paper indicates that any attempt to "change policy" will expose the supplier to Article 82 claims.

    59. See Carlton, "A General Analysis of Exclusionary Conduct and Refusal to Deal - Why Aspen and Kodak Are Misguided," 69 Antitrust L.J. 659, 679-680 & & 39 (2001); Klein, "Market Power in Antitrust: Economic Analysis After Kodak," 3 Sup. Ct. Econ. Rev. 43, 47-63 (1993).

    60. For example, a supplier might be wiling to supply parts to third-part service firms or to train their personnel in order to make its primary products more attractive to customers because of the presence of a number of service alternatives, including alternatives located close to potential customers. Such an approach might assist a small supplier in entering the primary market or in expanding its sales in the primary market However, if taking such an approach to aftermarket par and training could not be altered in the future without violating Article 82, it is possible that the supplier would be deterred from pursuing that approach. The concern that efficient conduct that benefits consumers will be deterred by an analysis that places importance on "policy changes" is similar to the problem noted with the different standards set forth in the Discussion Paper for "termination of an existing supply relationship" and "refusal to star to supply".


    STATEMENT
    of the
    UNITED STATES COUNCIL
    FOR INTERNATIONAL BUSINESS

    Submission to the Directorate-General for Competition on the
    application of Article 82 of the Treaty to exclusionary abuses



    March 30, 2006




    UNITED STATES COUNCIL FOR INTERNATIONAL BUSINESS
    1212 Avenue of the Americas, New York, New York 10036. (212) 354-4480


    Table of Contents

    Introduction

    1. Framework for Analysis
      1. Forms and Rules v. Economic Principles
      2. Enforcement Should Focus on Consumer Welfare and Not Harm to Competitors
        1. Focus on Consumer Welfare
        2. Exclusionary Abuses
        3. Efficiencies
    2. Determining When Conduct is Exclusionary
    3. The Role of Dominance
      1. Market Share is Only a Starting Point for Analysis
      2. High Market Share is Not an Accurate Indicator of Dominance
      3. Market Definition
    4. Burdens and Levels of Proof
      1. Burden of Production
      2. Issues Involving the Standard of Proof
        1. Standards Should Remain Constant and Not Shift in Response to Market Share
    5. Collective Dominance
    6. Predatory Pricing
    7. Bundling and Tying
      1. Tying and Bundling are Competitive Strategies Designed to be Pro-consumer, and Should be Considered Non-abusive Unless Proven Otherwise
      2. An Economics-Based Test of Tying and Bundling Would Incorporate These Elements
    8. Refusals to Deal
      1. Consumer Welfare Requires an Approach Grounded in Economics
      2. Intellectual Property Rights
    9. Aftermarkets

    COMMENTS ON THE REFORM OF THE
    APPLICATION OF ARTICLE 82 OF THE EC TREATY

    The United States Counsel for International Business (USCIB) welcomes this opportunity to submit comments on the Reform of the Application of Article 82 of the EC Treaty. USCIB works to advance the global interests of business both at home and abroad. It is the American affiliate of the International Chamber of Commerce (ICC), the Business and Industry Advisory Committee (BIAC) to the OECD, and the International Organization of Employers (JOE).

    As a globally-oriented business organization, USCIB has a particular interest in how Article 82 is applied in the context of international business, and is able to draw on a rich and varied range of perspectives from different sectors on these issues. It is the belief of USCIB that there are significant efficiencies to be gained, both in Europe and throughout the global economy, through the application of sound economic principals to issues of competition regulation. While necessarily general in nature, it is hoped that the following comments will assist in the identification of priority areas for further advances in competition policy by the Directorate-General for Competition.

    1. Framework for Analysis
      1. Forms and Rules v. Economic Principles.

    There are few, if any, unilateral practices that can be condemned as anticompetitive without consideration of their actual impacts on the market in which they operate. Cases involving single firm conduct are necessarily about both the dynamics of the firm engaging in the conduct in question and the dynamics of an entire market. Most will involve vigorous competition, which competition laws should be designed to encourage. Consumer welfare is harmed when pro-competitive conduct is chilled or condemned. A full understanding of the overall economic impact of a given business activity is necessary to avoid mistakes in enforcement that ultimately have a negative impact on consumers, businesses and the market. Adhering to sound economic rationales for enforcement results in more uniform enforcement and more predictable decision making.

    Enforcement should continue to push in the direction of increased objectivity, transparency and administrability.l Regardless of the type of conduct at issue - whether tying, rebates, or refusals to deal - the question of whether a business has engaged in anti-competitive conduct can only be answered by an analysis of the conduct's impact on the undertaking, on efficiency, and most significantly, on consumer welfare. While concepts such as market share and market power provide useful starting points in the analysis of market dynamics, standing alone they cannot answer whether unilateral conduct violates Article 82.

    Keeping in step with advances in economic thinking, competition enforcement has moved away from a focus on the "form" of challenged conduct in favor of a more flexible and context specific economic analysis of competing interests.2 USCIB applauds this development. Form-based regulatory regimes in which certain types of lawful behavior become unlawful once a firm passes over some threshold of market share are difficult to justify and even more difficult for businesses to comply with.

    For example, a rule that creates a presumption of market power above certain levels of market share deters pro-competitive conduct simply because of the risk it imposes on dominant firms to take any action at all. While per se rules may reduce enforcement costs, this purported benefit in fact imposes a much greater cost on society as a whole. This is because such inflexible rules are necessarily over-inclusive and thus have the effect of deterring efficient conduct that would otherwise increase overall consumer welfare. The challenge in exclusion cases is to determine how the law should treat conduct that can have both efficiency benefits and exclusionary harms. As a result of the difficulty of distinguishing when the line between aggressive pro-competitive conduct and conduct that is harmful to competition is crossed, standards governing such conduct need to be sufficiently flexible and adaptive to be able to incorporate continuing advances in our understanding of economics.

    The act of competition enforcement should be a fact-intensive inquiry that requires consideration of sophisticated economic evidence. There are basic economic principles that should be applied in a uniform manner to cases involving single firm conduct. For example, a competition law that seeks to maximize consumer welfare should take as its underlying principle that government intervention should be modest and undertaken only when the rules are clear and understandable so that uncertainty about the rules does not inhibit competitive and entrepreneurial forces that competition regulation is intended to encourage. The rules adopted should give clear notice to affected paries so that they will know what is required of them. Additionally, enforcement decisions should turn on tractable factual issues, and like cases should be treated alike.

    The criterion by which competition rules should be judged is how well the rules deter welfare-reducing conduct without reducing welfare-enhancing conduct. This is a very fact-intensive consideration. Closer adherence to the general principle that enforcement exists to foster competition, not competitors, combined with a greater effort to publicize the economic rationale behind enforcement decisions (including enforcement actions not taken) will result in more uniform and economically accurate decision-making, as well as greater transparency to the business community. In sum, these core principles help make the law and its enforcement more predictable, thereby furthering robust conduct by economic actors, and thus in turn promoting competition objectives.3

      1. Enforcement Should Focus on Consumer Welfare and Not Harm to Competitors.
        1. Focus on Consumer Welfare

    USCIB welcomes the discussion paper's statement in paragraph 4 that "(w)ith regard to exclusionary abuses the objective of Article 82 is the protection of competition on the market as a means of enhancing consumer welfare and of ensuring an efficient allocation of resources." Competition law should prohibit conduct only where the net impact on competition, taking into account any benefits of efficiency gains and consumer benefits, is nonetheless negative. Under Article 82, harm to competitors is not a sufficient condition for enforcement.4 If the efficiencies generated by a dominant firm's conduct outweigh their negative effects on competition such that the net effect advances consumer welfare, the conduct should not be considered abusive.5

    How efficiencies and negative impacts are to be identified, measured and balanced is a matter of some debate. In addressing these questions, however, it is important that policy makers be mindful as to how the adopted approach can be fashioned onto a set of rules and regulations that can be efficiently administered. In addition, the rules and regulations must not impose such high costs of compliance that they destroy the incentives to create the efficiencies they are intended to foster and protect. USCIB believes that a sound and administrable policy begins with adoption of the proposition that a dominant firm's conduct should be viewed as abusive only when its net effect is to harm consumer welfare. As Judge Learned Hand6 once admonished: "(t)he successful competitor, having been urged to compete, must not be turned upon when he wins."7 When a dominant firm wins business from a customer that was formerly served by a competitor, the competitor is excluded from that transaction, yet such competition is entirely beneficial. We believe that the means of reducing the risk that consumer welfare might be harmed though over-enforcement requires close adherence to the general principle that enforcement exists to foster competition, not competitors. It is often difficult to distinguish between firms that achieve and maintain their dominant position as a result of being aggressive, innovative competitors that benefit consumers, from those who engage in conduct that retards competition. Pro-competitive conduct that benefits consumers and the business engaging in it may hinder the efforts and prosperity of that business' competitors.8

        1. Exclusionary Abuses

    Paragraph 1 of the discussion paper defines exclusionary abuses as "behaviors by dominant firms which are likely to have a foreclosure effect on the market, i.e., which are likely to completely or partially deny profitable expansion in or access to a market to actual or potential competitors and which ultimately harm consumers." Because enforcement is appropriate only in instances where consumer welfare is harmed, the definition of exclusionary abuses must necessarily be limited to where the extent of the foreclosure is sufficient in the context of the relevant market to harm competition and where that harm is brought about by illegitimate means. The objective of any business enterprise is to expand its position in the marketplace to the maximum amount possible. It is in this manner that a company maximizes the return on investment of its stakeholders. To accomplish this goal, businesses are compelled to out-perform their competitors through superior efficiency, innovation and price. Far from being a threat to competition, companies that are most successful in one or more of these pro-competitive behaviors are by definition the most likely to become dominant firms. Such firms necessarily engage in aggressive competitive practices, practices that should be encouraged regardless of who wins or loses as long as the outcome is not the impairment of competition itself.

    When a dominant firm wins the business from a customer formerly served by a competitor, the competitor is excluded from that transaction, yet such competition is entirely beneficial. Therefore, it is often difficult to distinguish between firms that achieve and maintain their dominant position as a result of being aggressive, innovative competitors and those who engage in conduct that retards competition. This task is made even more difficult by the fact that conduct that benefits both the business engaging in it and, ultimately, the consumers served by it may also act to hinder the efforts and prosperity of competitors.9 For these reasons, it is critical that it be made clear that the mere fact that competitors may be harmed is insufficient to establish liability. There must be harm to consumers that was predictable at the time the conduct occurred.

        1. Efficiencies

    Efficiencies, recognized but only briefly in the discussion paper,10 should be a central consideration in determining whether conduct is abusive. Efficiencies occur where unilateral conduct results in positive effects that benefit consumers, such as improved quality or reduced prices. These efficiencies may be of either a qualitative or quantitative nature. For production or distribution, it can reduce costs and hence prices. Efficiencies are also created as a result of greater research and innovation. All of these things are beneficial to consumer welfare, and, thus, conduct that results in efficiencies should not be deemed abusive. A failure to recognize the consumer benefits from the creation of efficiencies would result in over-enforcement, which in turn would result in innovation-deterring uncertainty as to when otherwise pro-competitive conduct becomes abusive.11

    1. Determining When Conduct is Exclusionary

    Standards for exclusionary conduct should continue to be determined by taking into account sound economic theory, with particular focus on developing an administrate test that is intelligible enough to provide guidance to businesses seeking to compete aggressively while conforming their conduct to the law. Indeterminacy in competition laws creates uncertainty for businesses. Uncertainty for businesses creates a risk that they will not undertake pro-competitive, pro-consumer activities for fear of becoming embroiled in costly, lengthy litigation. A workable. definition of exclusionary conduct under Article 82 must not be over-inclusive and must also be readily administrable.

    Exclusionary conduct is behavior that excludes competitors on some basis other than efficiency, to the detriment of consumers. Mere market exclusion or serious harm to competitors should not be enough to establish liability, absent a further showing of abnormal methods of competition or competition not on the merits. Stated another way, conduct is unlawful if it would be unprofitable for the acting firm but for both the exclusion of rivals from a market and the resulting market power that would enable the dominant firm to recoup its prior period losses. Conversely, conduct is not exclusionary if the conduct would be profitable for the acting and would make good business sense even if it did not exclude rivals and thereby create or preserve market power for the acting firm.

    A standard for whether behavior is appropriate must consider the economic reality confronting the acting firm at the time the decision to take a given course of action is taken. Moreover, that standard must be the same at the time the challenged conduct is undertaken as it is when the conduct is scrutinized, possibly years later. An efficient standard provides for adequate deterrence while reducing the risk of false positives. Ideally, the standard and the means of its implementation will achieve the following goals. First, the standard must condemn only conduct that clearly could be anticipated to generate incremental costs for the acting firm that exceed the incremental revenues or cost savings that the conduct creates for the firm.. It should condemn only conduct that, viewed ex ante, reduces welfare on the basis of the short-term, non-transitory consequences of the conduct. As the European courts have noted in the analogous context of conglomerate mergers, the Commission should be required to show that the conduct of the dominant firm would be likely to result in harm to competition "in the relatively near future." It should not condemn conduct on the ground that it might lead to future increases in market power and a resulting welfare reduction. Such a standard would present a greatly reduced risk of false positives. To the extent that the Commission wishes to consider long-term effects on the market, the Commission should adhere to a higher evidentiary standard, given that showing long-term effects are often "dimly discernible, uncertain and difficult to establish."12 Therefore, the Commission should be required to show clear and convincing evidence that the long-term effects will likely result from the behavior in question.

    Second, an effcient standard directs attention to the nature of the defendant's conduct,not just to market-wide effects, many of which are unanticipated and beyond a defendant's control. In this manner, an efficient standard ensures that firms are entitled to reap the fruits of their "skill, foresight and industry," even if those fruits include market power, and condemns only conduct that is not competition on the merits.

    Third, an efficient standard ensures that the competition laws condemn only conduct from which an anticompetitive intent can unambiguously be inferred. One way it can accomplish this is by condemning only conduct that makes no sense apart from exclusion of competitors and resulting market power. It should not condemn conduct that makes good sense from the firm's perspective, regardless of the resulting market power simply on the ground that the conduct may have the effect of creating market power.

    Finally, the efficient test is administrable by enforcement agencies and courts, and provides simple and meaningful guidance to firms to enable them to know how to avoid liability without steering clear of procompetitive conduct. Firms would be able to comply with the law simply by determining whether their contemplated conduct would make good business sense even if the conduct did not increase their market power.

    1. The Role of Dominance
      1. Market Share is Only a Starting Point for Analysis.

    The Discussion Paper takes an important step towards a more economic approach to Article 82 by defining dominance as having "substantial market power."13 As the Discussion Paper explains14, an undertaking has substantial market power "only if it is capable of substantially increasing prices above competitive levels for a significant period of time." Market share alone is never a sufficient indicator of dominance since it is necessary in each case to examine the "link between the position of economic strengths held by the undertaking and the competition process" in the context of the relevant market.15 Standards that, either overtly or covertly, run contrary to these fundamental principles should be suspect, as more often than not they result in discouraging leading firms from engaging in conduct that benefits consumers at the expense of a competitor's market share. Placing undue reliance upon market share II determining dominance and making enforcement decisions deters firms from engaging in aggressive competition. Why should a firm engage in activities that are beneficial to consumers, such as rebates and discounts, when the increased market share potentially gained by being more efficient serves only to attract the punitive attention of competition enforcers? Additionally, given the tremendous variation that exists between and among markets, reliance on market share thresholds to evaluate abuse risks being substantially over-inclusive to the detriment of consumer welfare. This is especially true in dynamic markets, where competition is generally "winner-takes-most." In these markets, market shares are a particularly poor proxy for market power since even incumbents with very high. market shares are almost always constrained by a permanent threat of entry, mandating innovation.

      1. High Market Share is Not an Accurate Indicator of Dominance.

    There is no economic basis for discounting a firm's pro-competitive, welfare-enhancing, efficient behavior simply because it holds a particular percentage of the relevant market. The fact that a firm has a large portion of the relevant market simply does not translate ipso facto to the possession of dominance in the market by that firm.16 As the DG Discussion Paper recognizes,17 to avoid being over-inclusive at the expense of consumer welfare, a great number of market factors must be considered before it can be concluded that a firm possesses market power. Factors such as low barriers to entry, changes in technology, or unusually in the possession of buyers restrict the behavior of even dominant firms. However, where market share thresholds trigger presumptions of market power, the risk of false positives is extraordinarily high to the detriment of consumers, as even an above-cost price cut can result in liability, even though any exclusionary effect either reflects the lower cost structure of the alleged predator, and so represents competition on the merits, or is beyond the practical ability of a judicial tribunal to control without courting intolerable risks of chilling -legitimate price cutting.

    What is of significance in assessing dominance, as the DG Discussion Paper recognizes18is an analysis of a firm's exercise of market power over time. Obviously, where a firm is required to modify its behavior in a relevant market as the result of the activities of its competitors, it cannot truly be said to possess market power.19 This is because the firm in such circumstances cannot act without regard to whether the conduct is efficient or harms consumer welfare over an extended period of time. Similarly, as noted in the portion of the Discussion Paper dealing with rebates, evidence that other firms are able to compete for all or most of the demand in the relevant market is inconsistent with a finding of dominance even if the leading firm has a very high market share.20 In addition, evidence that competitors are expanding or are able to expand their operations, or that new firms are entering or are able to enter a market, is inconsistent with a finding of market power, and hence market dominance.21

    The Discussion Paper suggests that in certain cases, market share alone can lead to a presumption of dominance.22 Given the emphasis on economic analysis suggested throughout the Discussion Paper, these sections should be modified to acknowledge that in certain circumstances, be indicative of market power, but insufficient to establish a per se presumption.

    While high market shares are not presumptively indicative of market power, it is nonetheless appropriate to establish a safe harbor based on market shares.23 For example, in United States v. Aluminum Company of America,24approved and adopted by the U.S. Supreme Court in American Tobacco Co. v. United States,25Judge Learned Hand created the widely accepted rule of thumb that to find monopolization, "it is doubtful whether 60 or 64 percent would be enough; and certainly, 33 percent is not."26 Supreme Court cases have suggested that absent special circumstances, a defendant must have a market share of at least fifty percent before he or she can be guilty of monopolization.27 Thus, as a matter of law, absent other relevant factors, U.S. court have found that a 55 percent market share will not prove the existence of monopoly power.28 For market shares above that level,29 a full-scale economic analysis of the economic justifications from the firm's perspective and of the effects of the firm's behavior on consumers, not competitors, is called for. By adopting a market share "safe harbor," businesses can be spared the enormous expense of being forced to defend single firm conduct when there is virtually no likelihood that such conduct could have harmed consumers.

      1. Market Definition.

    As the DG Discussion Paper points out,30 to establish whether a business possesses market power, it is necessary to define the relevant market in which said power allegedly operates. Additionally, competitors selling reasonable substitutes for the product or products at issue must be accurately identified and, since what constitutes a reasonable substitute is a question of demand, the relevant customers must also be accurately identified. Thus, for example, it would be inappropriate to define a relevant market by reference to a single firm's intellectual property rights (IPRs), whether patents, copyrights, or unpatented trade secret technology. Indeed, the United States Supreme Court in Illinois Tool Works, Inc. v. Independent Ink, Inc., recently recognized that "Congress, the antitrust enforcement agencies, and most economists have all reached the conclusion that a patent does not necessarily confer market power upon the patentee. Today, we reach the same conclusion, and therefore hold that, in all cases involving a tying arrangement, the plaintiff must prove that the defendant has market power in the tying product."31 The development of economic methods for accurately addressing these questions is a process that is rapidly developing and rules and regulations must be flexible enough to accommodate newer methodologies as they become economically feasible to apply. Therefore, mechanical application of any test must be avoided. In particular, the Commission should not rely solely on methods that focus exclusively on product characteristics, which may lead to overly narrow market definitions, and concomitantly erroneous findings of dominance.

    1. Burdens and Levels of Proof
      1. Burden of Production.

    No matter how well designed or empirically based a standard for determining whetherconduct is exclusionary, the benefits of such a standard can be dissipated or even eliminated by the procedures through which it is applied. For example, placing on dominant firms the burden and risk of being able to prove that the efficiency gains of their conduct outweigh any negative effects on competition, rather than requiring the par seeking enforcement to rebut any prima facie efficiency claims as an essential part of plaintiffs case, will discourage dominant firms from engaging in efficiency enhancing conduct, regardless of the standard selected.32 There is no sound reason why enforcement agencies should not carry the burden of evaluating business justifications when attempting to assess the overall positive and negative effects on the marketplace of the behavior in question. It should be the burden of the authority investigating the alleged infringement of Article 82 to support any finding of abuse by empirical economic evidence that the conduct can credibly be shown to cause substantial harm to consumer welfare, and only then should the firm have a burden of coming forward with justifiable efficiencies that can be balanced against the demonstrable harm.

      1. Issues Involving the Standard of Proof.

    Standards that are unjustifiably high, such as requiring that a firm demonstrate that the conduct in question was "indispensable" in order to realize the claimed efficiencies, or that there are no other economically practicable and less anticompetitive alternatives to achieve the same efficiencies, should be avoided.33 While a business may consider multiple courses of action, it is unrealistic to expect firms to expend the resources necessary to perform studies to eliminate the possibility that there might be alternative courses of action that might have less impact on competitors. Imposing such a burden on businesses institutionalizes the very inefficiencies this process is trying to avoid, and places the protection of competitors above creating efficiencies in the market. Indeed, imposing such costs eliminates the incentives for firms to undertake innovative measures to increase efficiencies or reduce prices in the first instance.

    In addition to the criteria of indispensability, the criteria regarding the elimination of competition in respect of a substantial portion of the products concerned seems to be an unjustifiably high standard. At least if this criteria is interpreted as in Article 81(3) it will be hardly possible for a dominant company to demonstrate that "competition in respect of a substantial portion of the products concerned is not eliminated." Finally, the presumption that above a market share of 75% efficiencies will no longer justify otherwise abusive behavior is at odds with an economics-based approach. It will make it impossible for a number of companies to engage in efficiency-enhancing pro-competitive behavior.

        1. Standards Should Remain Constant and Not Shift in Response to Market Share.

    As previously noted, market share is not a substitute for an empirical understanding of the market. It is only a staring point for analysis. As such, there is little if any justification for applying different standards to the same conduct when engaged in by firms having differing market shares.34 Where firms with high market shares generate efficiencies that benefit consumers, their conduct is no different than that of any other actor in the market and their conduct must be permitted, even if it means the elimination of less efficient competitors. 35 While it could be argued that in the context of predatory pricing, a larger market share provides an incentive for preclusive conduct, the instances where recoupment may successfully be had are extraordinarily rare and thus do not provide a justification for standards to shift in response to the degree of market share.

    1. Collective Dominance

    The Discussion Paper notes that Article 82 is also applicable where two or more undertakings together hold a dominant position.36 Applying Article 82 to situations where two or more undertakings function in the market as a single firm because of "ownership interests or other link in law"37 is unlikely to pose problems for undertakings. Indeed, as the Discussion Paper confirms, to date Article 82 has only been applied ''with respect to exclusionary abuse of a collective dominant position" in "situations where there were strong structural links between the undertakings holding the dominant position."38

    USCIB questions the proposal in the Discussion Paper to expand the application of Article 82 to reach oligopolistic markets in which there are no structural or legal links among the undertakings.39 It would be extremely difficult to advise firms as to how to avoid Article 82 violations when their economic self-interest may lead each firm in an oligopolistic market to pursue similar independent actions. In addition, it is unclear what remedial steps would be appropriate to address ongoing independent behavior involving "conscious parallelism" taken by firms in an oligopolistic market. The failure of the attempts by U.S. antitrust agencies in the 1970s and 1980s to deal with "shared monopolies" suggests that the Commission should avoid expanding the concept of collective dominance beyond the reach of the current case law.

    1. Predatory Pricing

    As the Discussion Paper notes, the "lowering of prices, the directly visible part of predation, is also an essential element of competition. By lowering its price and/or improving the quality of its products a company competes in the market. This is competition that benefits its consumers and that a competition authority wants to defend and protect."40 Two important aspects of the Discussion Paper's approach to predatory pricing should be revised because they threaten to chill pricing initiatives that will benefit consumers.

    First, pricing at or above average total cost (ATC) should not provide a basis for a claim of predatory pricing. The Discussion Paper correctly observes that prices that are above a firm's ATC "are in general not considered to be predatory because such pricing can usually only exclude less efficient competitors."41 Rather than create uncertainty that might chill pricing behavior by successful firms that will benefit consumers in order to make it possible to address the rare "exceptional situation,"42 the Commission should encourage lower prices by making pricing at or above ATC a safe harbor that will not expose firms to Article 82 scrutiny.

    Second, the Discussion Paper fails to focus appropriately on recoupment. Recoupment should be a critical element of any predatory pricing claim, since consumers will benefit overall from lower prices unless the firm engaging in below cost pricing is able to recoup all of its losses on a net present value basis. It is therefore not sufficient to presume a "likelihood of recoupment" from the fact that a firm holds a dominant position and, consequently, that there are likely to be barriers to entry into the relevant market.43 The existence of barriers to entry is necessary for the dominant firm to recoup its losses but is not sufficient to establish that recoupment would occur. The recoupment assessment should take into account the magnitude of the likely losses, the level of increased prices following foreclosure and the period of time during which those prices would need to be charged, the time value of money, and the prospects for innovation affecting the ability to recoup as well as the prospects for entry prior to recoupment of the losses on a NPV basis.

    1. Bundling and Tying
      1. Tying and Bundling are Competitive Strategies Designed to be Pro-consumer, and Should be Considered Non-abusive Unless Proven Otherwise.

    Bundling and tying are ubiquitous to the marketplace and are now considered to provide significant benefits to both producers and consumers. Thus, a per se approach is not appropriate and results in over-enforcement and a reduction reduces activities that would result in increases in consumer welfare. It is now a well-accepted economic principle that, more often than not, bundling and tying results in lower production, transaction and distribution costs, lower prices, and greater convenience and utility for consumers.

    Distinguishing instances of anticompetitive tying or bundling from instances of procompetitive tying or bundling is especially difficult.44 "[T]here are decent theoretical reasons for concern that vertical restraints can have anticompetitive consequences," yet that outcome will occur "probably in only a small minority of cases in which they are employed. Yet even in suspicious cases there invariably are multiple possible reasons for a challenged practice - no responsible student of competition policy is about to suggest that bundling, discounting, exclusive dealing, volume discounts, consumer rebates, or even tying should be presumptively unlawful - and sorting out the reasons in particular cases will often be very difficult. It is easier to conjecture anticompetitive (reasons) for such practices than it is to determine the practices' actual or even (in contrast to cartel cases) likely economic consequences."45

    Market-leading companies should be able to continue producing innovative combinations of products benefiting consumers without running afoul of the prohibitions on tying unless the competition authority can rebut the innovating firm's prima facie case of efficiency gains.46 When companies combine formerly separate products, consumer welfare is usually increased as firms realize the efficiencies involved. These efficiencies may be the result of greater product functionality or the elimination of double marginalization, or simple convenience. Such tying or bundling may also lead to system-based competition, which may create an even more innovative and competitive market than component-based systems, as the markets for computer systems, home theaters, and cell phones aptly demonstrate. As a result, any rule that did not create broad safe harbors would run a substantial risk of deterring pro-competitive conduct to the detriment of the consumer.

    One such safe harbor is suggested in the Discussion Paper in the context of bundling. The Paper offers a safe harbor based upon analysis of whether "the incremental price that customers pay for each of the dominant company's products in the bundle (covers) the long-run incremental costs of the dominant company of including th[e] product in the bundle."47 Assuming that this safe harbor is sufficient, then for mixed-bundle discounts or rebates that fall outside the safe harbor, the Commission should then continue the analysis by demonstrating (1) a likelihood of recoupment and (2) a likelihood of the creation of substantial market power in the relevant market for the "bundled" product in order to show that discounting through mixed bundling constitutes an abuse of dominance. Absent such a showing, mere exclusion of a competitor should not be found sufficient to establish a finding of anticompetitive bundling. Moreover, as discussed in Section B following, a somewhat broader safe harbor may further serve to limit false positives.

    As is the case in other areas of single firm conduct, it can be difficult to distinguish between vigorous competition and anticompetitive acts, especially where the alleged act results in the lowering of prices to consumers. It is therefore essential that the regulation of bundling and tying practices focus not on their effects on competitors, but on the welfare of customers in the market. This can only be determined through an extensive analysis of the market, the conditions of entry and the constraints the threat of entry places on market participants. Evidence of harm to competitors is ambiguous at best with regards to the positive or negative impact of the conduct on the market. Because erroneous enforcement may discourage pro-competitive practices that provide benefits to the market, entities challenging such conduct must be required to demonstrate anticompetitive consequences directly resulting from the alleged conduct.

      1. An Economics-Based Test of Tying and Bundling Would Incorporate These Elements.

    An analysis of bundling should not proceed unless a defendant's price discount brings the firm's price below its cost.48 Even when a company engages in below-cost pricing, the firm still should not be found liable without substantial proof that the firm can and will recover its discounts because of a reduction in competition.49 Indeed, in markets where businesses can move in and out, the short-term benefits to consumers in the form of lower prices may more than offset the remote risk that the seller will ultimately succeed in driving all rivals from the market.

    Because the harm over-enforcement can cause to consumer welfare is significant in this area, the ideal test is one that greatly reduces the risk of enforcement by being administrable by competition authorities while being easily and predictably applied by businesses. It would create a safe harbor for which a business can qualify using its own readily available data, thus not diminishing the effects of efficient conduct as a result of compliance costs. It must be designed so that it condemns only conduct that generates incremental costs for the defendant that exceed the incremental revenues or cost savings that the conduct creates for the defendant, and thus condemns only conduct that is not "competition on the merits," while allowing firms to reap the fruits of their skill, foresight and industry. The adoption of economically sound, administrable and predictable standards provides straight-forward and meaningful guidance to firms to enable them to know how to avoid competition liability with data readily available to them at the planning stage. An efficient standard should move business behavior in a pro-competitive direction without imposing excessive decision-making costs or chilling aggressive competition.

    1. Refusals To Deal
      1. Consumer Welfare Requires an Approach Grounded in Economics.

    Similarly, an empirical, economically based test for exclusionary conduct would be particularly beneficial for application to cases involving refusals to deal. It is well established that firms, even dominant firms, generally have the right to decide whom to supply, including whether to supply at all.50 It is also widely recognized that forcing dominant firms to grant access to their inputs can deter innovation, both by discouraging dominant firms from investing in innovation in the first instance, and by encouraging smaller rivals not to innovate but instead to "free ride" on the innovations of others.51 The United States Supreme Court, echoing these principles, recently observed that compelling firms who have established an advantage "to share the source of their advantage is in some tension with the underlying purpose of competition law, since it may lessen the incentive for the monopolist, the rival, or both to invest" in ways that promote consumer surplus.52

    A firm's dealings with third parties and its prior dealings with rivals provide a baseline for evaluating its challenged conduct. Where a firm is willing to deal with its retail customers on certain terms (such as a certain price), claiming that its refusal to deal with a rival on those terms constitutes anticompetitive conduct makes no economic sense. However, absent discriminatory dealing or departures from prior profitable courses of dealing, decisions by either courts or regulatory agencies to enforce sharing distort the incentives to innovate and should therefore be avoided.53

    The Discussion Paper appears to make the troubling suggestion that there is no need to identify an actual downstream market, and would deem the existence of a potential or hypothetical market sufficient for a showing of anticompetitive effect.54 Read literally, the effect could be to require dominant firms to share every technological advance made to improve production processes, even absent the existence of an existing market for such technology, even without the presence of any effort at leveraging. This would be particularly troubling in an innovation market, where technological advances are the primary competitive advantage.

    As in bundling and tying cases, reducing the occurrence of over-enforcement in cases involving refusals to deal while being efficient and administrable requires the consistent application of sound economics. In order not to suppress conduct that would be beneficial to consumers, appropriate standards must be adopted that condemn only conduct that is not "competition on the merits," while allowing firms to reap the fruits of their skill, foresight and industry by being able to predict the likely consequences of their actions. Meaningful guidance must be provided to firms to enable them to know how to avoid liability using data that is readily available to them at the planning stage, and that the conduct, if challenged, will be evaluated under the same efficient standard that applied at the time the company decided to engage in the conduct.

      1. Intellectual Property Rights

    It is a well established principle that the rights of intellectual property holders are to be respected in all but most exceptional circumstances.55 In fact, there is no economic reason why cases involving intellectual property rights should be treated any differently than any other case involving a refusal to deal.56 The purpose of intellectual property law in the first instance is to provide businesses an incentive to invest in research and development activities aimed at generating new products and services. Thus, intellectual property rights are of vital importance to promoting consumer welfare. The adoption of rules and standards that create uncertainty as to when a company may be required to license its intellectual property will have a chilling effect on investment in research and development, to everyone's detriment.57 This is particularly true in markets that are already subject to governmental regulation. Such regulation tends to significantly reduce the likelihood of major antitrust harm. The additional benefit to competition of adding another layer of legal process will tend to be small, whereas the risk of false positives is high.58

    Moreover, the Discussion Paper does not provide clear guidance with regard to refusals to license IPRs, and runs the very real risk of over-deterrence. For example, while rightly acknowledging that the refusal to license an IPR should only be considered an abuse in "exceptional circumstances,"59 the Paper goes on to state that such circumstances may be present where the potential licensee "intends to produce new goods or services not offered by the owner of the right and for which there is a potential demand." However, there is no definition of precisely what constitutes "new goods or services" for the purpose of application to Article 82. This leaves open the possibility of an overly broad definition, and hence, potential over-enforcement. The Commission could utilize this opportunity to clarify the definition of "new goods or services," a standard developed in the IMS Health licensing case a few years ago, but still without substantive content.60

    Similarly, there is no justification in law or economics for the proposition that trade secrets should be entitled to less protection under Article 82 than other forms of intellectual property. If trade secrets are provided less protection than other forms of intellectual property, the net effect will be less innovation and competition in the market, not more. This is simply because the protection of trade secrets enables firms to recover the investments they make in the research and development that are necessary for the firm to be able to meet the competitive pressures of its rivals, who are themselves investing in research and development for the same reason. Thus, as is the case with other forms of intellectual property, uncertainty as to the ability to recover the costs of the research and development necessary to create innovative trade secrets acts as a disincentive, to the detriment of consumer welfare. From the other perspective, there is little incentive for risking the loss of your own investment in research and development that may fail to yield the desired results when you have the option of free-riding off of the efforts of a rival. For these reasons, sound economics requires that trade secrets be protected the same as any other form of intellectual property, and that the rules and regulations impacting intellectual property rights not create ambiguity with regards to the extent of their protection.

    1. Aftermarkets

    As the Discussion Paper notes, it is common for the supplier of capital equipment to have "a very strong position" in the sale of "secondary" products and services used with its own brand of equipment.61 As a result, there is a risk that undertakings with quite modest positions in the primary market would be viewed as dominant in the aftermarket if the assessment were to be focused only on an aftermarket consisting of products and services for their individual brand of equipment.

    The Discussion Paper is correct in emphasizing that the "secondary markets" should not be viewed in isolation since "the actual degree of market power of the supplier (in the aftermarket)... may be constrained by competition in the primary market."62 As the Discussion Paper explains, "competition in the primary market may make price increases in the aftermarket unprofitable due to its impact on sales in the primary market, unless prices in the primary market are lowered to offset the higher aftermarket prices."63 This fundamental insight regarding the key relationship between the primary market and any related aftermarkets means that a separate examination of a single-brand aftermarket under Article 82 is rarely, if ever, appropriate.


    FOOTNOTES

    1. International Chamber of Commerce, ICC Comments on the Reform of the Application of Article 82 of the EC Treaty, Doc. No. 225/623 (12 Dec. 2005) at 3-4, 16.

    2. See John Vickers, Abuse of Market power, 115 Economics Journal F244, F245-47 (2005); John Vickers, Competition and Economics Policy, October 3, 2002 at 3-6.

    3. International Chamber of Commerce, ICC Comments on Selected Issues for Study by the US. Antitrust Modernization Commission, Doc. No. 225/621 (1 Sep. 2005) at ~ 6.5

    4. European Commission, DG Competition Discussion Paper on the Application of Article 82 of the Treaty to Exclusionary Abuses (Dec. 2000) ("DO Discussion Paper") at ¶ ¶ 57, 58, 60, 94, 134.

    5. Id. at ¶ ¶ 54-60, 67, 91, 92.

    6. Billings Learned Hand (1872-1961), a senior judge of the U.S. Circuit Court of Appeals for the Second Judicial Circuit, was responsible for several landmark decisions in the early development of competition law. His decisions, particularly those involving the charge of monopoly, set judicial precedence in the United States for decades.

    7. United States v. Aluminum Co. of America, 148 F.2d 416, 430 (2nd Cir. 1945).

    8. DG Discussion Paper at ¶ ¶ 58, 82, 83, 96.

    9. ICC Comments on Selected Issues for Study by the U.S. Antitrust Modernization Commission at ¶ ¶ 4.4, 6.1.

    10. DO Discussion Paper at ¶ ¶ 130-134; see also 206.

    11. DO Discussion Paper at ¶ ¶ 87, 88, 91, 92.

    12. See Commission v Tetra Laval, Judgment or is Feb. 2005, Case C-12/03. (ECJ 2005), at ¶ 44.

    13. DG Discussion Paper at ¶23.

    14. DG Discussion Paper at ¶ 24.

    15. DG Discussion Paper at ¶ 23.

    16. ICC Comments on the Reform of the Application of Article 82 of the EC Treaty at 5-6.

    17. DG Discussion Paper at ¶ ¶ 28-29.

    18. DG Discussion Paper at ¶ 30.

    19. DG Discussion Paper at ¶ 27.

    20. DG Discussion Paper at ¶ 146 & fn. 92.

    21. DG Discussion Paper at ¶ 40.

    22. See, e.g, Paras. 29-31.

    23. ICC Comments on the Reform of the Application of Article 82 of the EC Treaty at 5-6.

    24. United States v. Aluminum Company of America, 148 F.2d 416 (2d Cir.1945).

    25. American Tobacco Co. v. United States, 328 U.S. 781 (1946).

    26. United States v. Aluminum Co. of America, 148 F.2d 416, 424 (2nd Cir. 1945).

    27. Illustrative Supreme Court cases include Otter Tail Power Co. v. United States, 410 U.S. 366 (1973) (91% market share); United States v. Grinnell Corp., 384 U.S. 563 (1966) (87%); International Boxing Club of New York, Inc. v. United States, 358 U.S. 242 (1959) (81%); United States v. American Tobacco Co., 221 U.S. 106 (1911) (86%); Standard Oil Co. v. United States, 221 U.S. 1 (1911) (90%); United States v. United Shoe Machinery Corp., 110 F. Supp. 295 (D. Mass. 1953), aff'd per curiam, 347 U.S. 521 (1954) (75%); United States v. Paramount Pictures, Inc., 334 U.S. 131 (1948) (70%); United States v. Pullman Co., 50 F. Supp. 123 (ED. Pa.1943), aff'd per curiam, 330 U.S. 806 (1947) (100%).

    28. United States v. Dentsply Int'l, Inc., 399 F.3d 181,187 (3d Cir. 2005); Fineman v. Armstrong World Indus., 980 F.2d 171, 201 (3d Cir. 1992).

    29. ICC Comments on the Reform of the Application of Article 82 of the EC Treaty at 5.

    30. DG Discussion Paper at ¶ 12.

    31. Illinois Tool Works, Inc. v. Independent Ink, Inc., 126 S.Ct. 1281, 1293 (2006).

    32. DO Discussion Paper at ¶¶ 77, 79, 84, 86, 91.

    33. Id. at ¶¶ 84, 86.

    34. Id. at ¶¶ 91, 92.

    35. ICC Comments on the Reform of the Application of Article 82 of the EC Treaty at 18-19.

    36. DG Discussion Paper at ¶ 43.

    37. DG Discussion Paper at ¶ 45.

    38. DG Discussion Paper at ¶ 76.

    39. DG Discussion Paper at ¶ 46. Any anti-competitive agreements between undertakings can be addressed under Article 81.

    40. DG Discussion Paper at ¶ 94.

    41. DG Discussion Paper at ¶ 127.

    42. DG Discussion Paper at ¶ 128.

    43. DG Discussion Paper at ¶ 122.

    44. Id. at ¶¶ 183-206.

    45. Richard Posner, Vertical Restraints and Antitrust Policy, 72 U. Chi. L. Rev. 2229, 240-41 (2005).

    46. ICC Comments on Selected Issues for Study by the Us. Antitrust Modernization Commission at ¶ 8.0.

    47. D.G. Discussion Paper at ¶ 190.

    48. Brook Group, LTD v. Brown & Williamson Tobacco Co., 509 U.S. 209, 222-23 (1993).

    49. Id. at 224-26. Any recoupment analysis should also take into account the impact of the time value of money, that is, that the amount of recoupment has to be larger in absolute terms than the loss from pricing below cost, since the recoupment will by definition occur in a later period.

    50. Id. at ¶¶ 207, 213,218.

    51. Brief for the United States, et al, as Amici Curiae Supporting Petitioner, Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004) (No 02-682), 2003 WL 21269559 at * 13-20.

    52. Verizon Communs., Inc.,540 U.S. at 407.

    53. DG Discussion Paper at ¶ 210.

    54. D.G. Discussion Paper at ¶ 227.

    55. D.G. Discussion Paper at ¶ 239; but see ~ 240.

    56. Illinois Tool Works, 126 S.Ct. at 1293.

    57. ICC Comments on the Reform of the Application of Article 82 of the EC Treaty at 19.

    58. Trinko, 540 U.S. at 407-408, 411-15.

    59. D.G. Discussion Paper at ¶ 239.

    60. Also troubling is the suggestion in Paragraph 240 that dominant firms could be forced to supply a license to their rivals for IPR technology that is indispensable for follow-on innovation even if the technology is not sought for direct incorporation in a product or service. The anomalous result would be that the dominant firm's rivals could pick and choose its technologies on the notion that such technologies could be useful at some indeterminate later date to develop a follow-on innovation, thereby eliminating any incentives for innovators in order to protect "free-riding" rivals engaged in exploitation of the technological innovator. This paragraph should be deleted from the Paper.

    61. DG Discussion Paper at ¶ 253.

    62. DG Discussion Paper at ¶ 246.

    63. Id

  • Updated June 25, 2015

    Was this page helpful?

    Was this page helpful?
    Yes No