Ten years of effects- Based approach in EU competition law

© Groupe De Boeck s.a., 2013

EAN 978-2-8027-3882-4

ISSN 2294-5571

Cette version numérique de l’ouvrage a été réalisée par Nord Compo pour le Groupe De Boeck. Nous vous remercions de respecter la propriété littéraire et artistique. Le « photoco-pillage » menace l’avenir du livre.

Pour toute information sur notre fonds et les nouveautés dans votre domaine de spécialisation, consultez notre site web :

www.bruylant.be

Éditions Bruylant

Rue des Minimes, 39 • B-1000 Bruxelles

Tous droits réservés pour tous pays.

Il est interdit, sauf accord préalable et écrit de l’éditeur, de reproduire (notamment par photocopie) partiellement ou totalement le présent ouvrage, de le stocker dans une banque de données ou de le communiquer au public, sous quelque forme et de quelque manière que ce soit.

Foreword

The Global Competition Law Centre (GCLC) is the College of Europe’s research centre in the field of competition law and economics. It was established in January 2004, so as to provide a discussion forum for academics, practitioners, in-house lawyers, economists and enforcement officers. The GCLC is now well known for its regular lunch talks as well as its evening policy talks, its ad hoc conferences and its annual conferences on competition law. In the context of those activities, it publishes working papers, reports as well as books containing the proceedings of it’s annual conference. Rather than merely commenting on past events, the GCLC’s ambition is to try to set itself apart from other conference organisers by being as much as possible "forward looking", i.e. anticipating developments as well as participating actively in discussions on the shaping of competition policy. The unique contribution of academics and practitioners allows the GCLC to combine "fundamental research" with "applied research", by keeping the necessary independence and where needed even promoting sometimes thorough reconsideration of "accepted truths".

The present book encloses the proceedings of the GCLC’s Seventh Annual Conference. In many ways, this conference was the logical consequence of the previous ones, which were devoted respectively to due process and judicial review in competition cases in the specific light of the "modernisation" context. However, the focus this time was not so much on procedural rules or on judicial control but on substantive rules. After roughly a decade of "effects-based approach" of competition rules, the idea was to assess how substantive "modernisation" has worked, and whether it can be refined, clarified or even improved.

Many basic questions have been raised during the conference which are discussed in the present book:

— First, an "effects-based approach" requires a definition of what is a "restrictive effect" on competition. So many years after the treaties were signed it is still hard to find a clear line on how to define this basic concept. Is it sufficient e.g. to demonstrate a restriction on the freedom of action of the parties as some old judgments seem to indicate? Or does the restriction of the structure of competition matter, a view still often expressed by the Court of Justice? Alternatively is a potential effect on one isolated parameter of competition sufficient (as repeatedly stated in other EU Courts’ judgments or by the Commission)? Or finally does a more global assessment of the negative effect on the consumer as a whole need to be established as required by the Court of Justice in other cases? As can be seen, this discussion turns really on the "DNA of the system", on its very foundations. At the same time it is all but theoretical. Any practitioner knows that the proper definition of the concept of "restriction of competition" itself determines the advice given in the multitude of situations which characterise the "real life" problems encountered by firms in their ordinary course of business. Similar questions arise in this regard not only under Article 101 TFEU but also under Article 102 TFEU as well as in merger control.

— Moreover, and beyond the concept of "anticompetitive effect", one also needs to discuss the concept of "anticompetitive object" as a possible exception to, or alternation of, the effects-based approach. Recent judgments in Pierre Fabre, T-Mobile or GlaxoSmithKline illustrate the difficulties and uncertainties of the concept. Can quasi "per se" violations of the law (so-called "object restrictions") at all be established in competition law independently of any proper "effects-based analysis" and when? And a contrario can "safe harbours" also apply indistinctively to the full range of commercial practices, without the need to carry out an in-depth effects analysis in some cases? In other, more fundamental words, how can law and economics best be reconciled? And how can the goal to achieve the "right decision" in terms of economic outcome be combined with the legitimate need for "legal certainty"? Clearly, it is not possible, for each advice given, to engage in in-depth economic analysis. Knowing moreover that even economics are all but an "exact science", that economists themselves often disagree even on fundamental questions, that there is not "one religion" in competition policy, that competition law in any event applies differently according to the sector concerned, its specificities, the structure of the markets and the characteristics of each undertaking, that hardcore restrictions in one sector will therefore be pro-competitive in another, that competition law is not immune from political ideologies etc., all this raises complex questions and delicate issues of predictability of the law. But it also raises difficult questions of burden of proof, of standard of proof and of nature and value of presumptions which the various chapters of the present book address in depth. Again the discussion is not specific to Article 101 TFEU but arises equally under Article 102 TFEU or in merger control.

— Further, once effects and object have been defined, it is still necessary to look at the analytical framework, possible theories of harm, evidence, causality and defences. This raises practical issues both for the lawyer and the economist, the enforcer, the plaintiff and the defendant. And it raises the fundamental question of the place of efficiencies and objective justifications in competition law.

— Finally, the practical implications of the effects-based approach for enforcers, judges and companies are examined. The Copernican change in the application of competition rules, which started around the turn of the millennium, raises also many other questions. What practical problems have been encountered by those applying it every day, be it as an enforcer, judge or party? Just to quote one, what is the value of precedents elaborated pre-2000, i.e. at the time where competition law was largely "form-based" whilst it is now "effects-based"? Obviously, again this is not a mere academic exercise. And it is all the more important as competition law is at the centre of market economy and as there is obviously a significant cost to any error, both type I and type II errors.

It is our hope and belief that the present book will bring its own contribution to the development of an improved system of competition rules. Each and every chapter brings new original and constructive ideas which deserve wider consideration. We have been particularly fortunate to have in this regard an exceptional number of outstanding professionals and academics who enthusiastically took part in the initiative and contributed their experiences and ideas to the conference and the present book. We wish very much to thank each of them very warmly for their contribution. We thank wholeheartedly the European Commission staff for their high profile responses, written contributions and participation in our discussion forum. Further we wish to thank our many sponsors without whom the GCLC would not exist, as well as the Rector of the College of Europe for his continuous support for the GCLC’s initiatives.

Last but not least, our gratitude goes to Professor Nicolas Petit for his unfailing support in preparing this conference and to Tarik Hennen for assisting us through the logistics and organisation of the conference. Both are very much the kingpins of the whole organisation and nothing could have be done without their selfless commitment to the GCLC. And at the same time, we would like to thank Harriet Dykes and Vera McManus, trainee solicitors at Hogan Lovells, Brussels, for their linguistic review of some of the chapters of the present book.

Unfortunately, although our hope was to publish a book that should have been as comprehensive as possible in reproducing the full extent of the particularly rich discussion at the conference, for a number of reasons it was not possible in the end for a very few participants to produce a written report. However, the great majority of them were able to provide within rather short time limits highly valuable contributions which we are proud to publish herewith.

JACQUES BOURGEOIS

AND DENIS WAELBROECK

Part One

The effects-based

 approach under Article 101 TFEU and its

 paradoxes: modernisation at war with itself?

Damien M.B. GERARD

¹

In a retrospective essay published in 1992 on his 30-year long career as an antitrust lawyer in Brussels, Don Holley wrote as follows: "It was in 1962 that EEC competition lawyers began asking themselves in earnest by what criteria a contractual restriction should be judged under Article 85(1) of the EEC Treaty. They are still asking the question".² Twenty years later, the question is still there:³ what is a restriction of competition under what has become Article 101 of the Treaty on the Functioning of the European Union (TFEU), that is according to which criteria an agreement between undertakings should be deemed restrictive of competition?

The present contribution aims to assess the extent to which the advent over the past ten years of a so-called effects-based approach in the enforcement of EU competition law has modified (if not clarified) the notion of restriction of competition in relation to agreements and other collaborative arrangements. To that effect, it first attempts to capture the transformation induced by the move toward the effects-based approach (part I) and then assesses the consequences thereof for the enforcement of Article 101 TFEU (part II). The thrust of the argument lies in the fact that the substantive modernization of EU antitrust enforcement is a welcome evolution but one that is not immune of some significant paradoxes. Hence it wonders: is not modernization a policy at war with itself?

I. The advent of the effects-based approach

A central hypothesis underlying this edited volume is the move, around the turn of the century, toward an effects-based approach in EU competition law enforcement including in, the competitive analysis of agreements and other concerted practices between undertakings (i.e., coordinated practices). One may speculate at length about the genesis of that transformation and the milestones that have contributed to reshaping the notion of restriction of competition under Article 101 TFEU. A rule of thumb, for example, is that there is a two to three decades lag between the development of ideas in economics and their impact on the formulation of competition policy.⁴ It is well known in that regard that the end of the 1970’s had witnessed the emergence of a new approach to antitrust harm based on neoclassical microeconomic theory, generally associated with the University of Chicago,⁵ which supported that various horizontal and vertical relationships may not necessarily result in a reduction of total welfare. Other (complementary) currents in economic theory, such as new institutional economics and game theory models,⁶ have also contributed to the reassessment of the harmful potential of agreements and other types of coordination. So did the refinement of econometric tools, insofar as they facilitated the empirical modelling and testing of economic relations.

In turn, a number of developments at EU level led to a greater awareness of the tools of economic analysis and to an overall reappraisal of the objectives of competition policy over the years, including prominently the introduction of merger control in 1989 and the strengthening of the transatlantic dialogue on the occasion of transactions subject to review both in the US and the EU.⁷ Likewise, the progressive structuring of the transatlantic competition law group further supported the dissemination of novel ideas and approaches,⁸ and expanded subsequently into a dynamic epistemic community at global level. The establishment of the Chief Economist office in the aftermath of the resounding annulment of negative merger decisions participated to the same movement.⁹ Eventually, the move toward an effects-based approach in EU competition enforcement is probably best captured as the substantive component of a greater transformational process known as modernization,¹⁰ which also entails procedural and institutional dimensions.

A. 

The Effects-based Approach as the Substantive Dimension of Modernization

Substantive modernization can be understood as a shift in the underlying rationality of EU competition law from the static protection of the freedom to trade of competitors to the dynamic promotion of a competitive process conducive to efficiency gains and contributing as a result to the welfare of consumers. In practice, it entails a move from a form-based to an effects-based approach in EU competition law enforcement. In line with the above discussion, the term economic approach is often used as a synonym for effects-based because the basic premise of the effects-based approach is indeed of an economic nature (in the neo-classical sense): it aims to explore the implications of coordinated practices as outcomes of rational choices by profit/utility-maximizing economic actors.¹¹ Moreover, the move to an effects-based approach has been made possible by the growing sophistication of the relevant models and tools of economic analysis, thereby overcoming prior scepticism at the possibility of weighing the actual welfare effects of agreements and other coordinated practices. In the EU, though, the previous form-based approach was also encouraged by the design of the procedural enforcement framework, that is the conscious choice made at the time of the negotiation of Regulation 17/62 for a compulsory prior notification system for those "agreements, decisions and concerted practices of the kind described in Article [101](1) of the Treaty".¹² Thus, substantive modernisation is inherently linked to the procedural modernization introduced by Regulation 1/2003.

The procedural dimension of modernization was indeed marked by the transition from a prior notification to an exception system for the review of coordinated practices.¹³ Originally, by granting the Commission a monopoly on the application of the then equivalent to paragraph (3) of Article 101 TFEU, which forms an integral part of the determination of a restriction of competition under that provision, Regulation 17/62 also triggered a broad and lenient interpretation of the scope of paragraph (1) thereof, that is of the notion of agreement. The immediate consequence of the notification requirement was the filing of tens of thousands of (mostly innocuous) agreements with the Commission,¹⁴ thereby triggering their formal treatment according to uniform principles and categories. Conversely, the move toward an exception system unlocked the enforcement framework and enabled, in theory, the rebalancing of the analysis between the first and third paragraphs of Article 101 TFEU. It also freed resources and enabled the development of a proactive and targeted enforcement policy focused on those most damaging practices and, more importantly, on the specifics of each case. There is a second aspect to the procedural modernisation process, however, which is often underestimated: the emergence of and increasing reliance on negotiated procedures in the enforcement of EU competition law (i.e., leniency, settlement, commitments proceedings considered against the background of significant fines). That phenomenon has profound implications for the enforcement process because it alters the incentives of the relevant actors and affects the credibility and legitimacy of an effects-based approach conducted at EU level by an integrated agency subject to limited judicial review.¹⁵

The third dimension of modernization is institutional in the broad sense of the term: the decentralization of the enforcement of Article 101 TFEU that resulted from the move toward an exception system and the abolition of the Commission’s monopoly over the application of Article 101(3) TFEU, in effect formalized the transition toward a form of network antitrust enforcement in the European Union. To be operational and deliver consistent outcomes, the European Competition Network (ECN) needed a substantive consensus at the outset and a rigorous common language; the welfare-consensus relying on the language of economics appeared well suited to the task.¹⁶ The relevant network is of course broader than the ECN and the EU; it extends to the whole antitrust community, which was originally transatlantic and has grown global in recent years. The mainstream consensus prevailing today within that community (with sometimes important nuances in the way it is implemented) is that free markets can contribute to consumer welfare by promoting competition as a process geared toward efficiency.¹⁷ To deliver on that promise, the process is subject to rules designed to tackle restrictions of competition which distort the incentives to achieve efficient outcomes. As a reflection of that consensus, the modernized approach in the EU entails that those agreements liable to have an "appreciable adverse impact on the parameters of competition on the market, such as price, output, product quality, product variety and innovation", are caught by Article 101(1) TFEU.¹⁸ The method used to assess that impact aims to verify economic assumptions by means of observable facts (i.e., effects-based), and is made possible by a cascade of legal process rules including evidence and other rules regulating the conduct of antitrust proceedings.

In view of the above, the three dimensions of the modernization process appear intrinsically interrelated and have all contributed to a radical transformation of EU competition law enforcement over the past ten years. The remainder of this section focuses on the substantive component of modernization and how it has emerged over time but the other dimensions will continuously loom in the background of the discussion.

B. The Progressive Formulation of the Effects-based Approach

Historically, the emergence of the effects-based approach can be traced to the combination of two phenomena or, put otherwise, consists in a synthesis between two phenomena: (i) the emergence of the neo-classical consensus of competition as an efficiency-driven process; and (ii) the development of the EU antitrust enforcement process, which was virtually inexistent at EU level before the mid-1960s. In chronological order, this second phenomenon came first as the European Court of Justice attempted to clarify step by step the scope of Article 101 TFEU (i.e., of its predecessors Art. 85 EEC and Art. 81 EC) and the powers of the Commission in conducting what was and still is the only vertically integrated (from policy to enforcement) area of EU competence.

Thus, in a stream of cases starting with Société Technique Minière in 1966, the ECJ established a number of evidentiary principles that are still considered valid law today including the counterfactual method, the ancillary test and the need to consider the "position on the market" of the parties to the agreement.¹⁹ In a number of those cases, the ECJ took a progressive stance concerned with the commercial reality of agreements, which was sometimes noticeably different from that of the Commission and emphasized progressively the need to consider agreements in the broader context of the "competitive forces operating on the relevant market".²⁰ However, the reach of those cases was limited by the procedural framework in place at the time for, as discussed above, the ex ante notification requirements induced a formalistic treatment of cases based on uniform rules and categories. Moreover, in the framework of a nascent European (market) integration project where the main concern was to prevent private practices from replicating the effects of those state barriers to trade that were progressively being brought down, the prevailing goal underlying EU competition enforcement remained the protection of the freedom to trade of market actors,²¹ influenced heavily by fairness considerations.

That approach was subject to growing criticism, notably in view of developments occurring on the other side of the Atlantic under the influence of the Chicago School of (law and) economics, which reached its peak in the mid-1990s, i.e., upon completion of the EU single market. The seminal Sytem Failure article published by Barry Hawk, a frequent commuter between the US and the EU, in the Common Market Law Review of 1995, conveys with particular candour the accumulated frustrations at the rigidity of the application of then Article 85 EEC.²² Almost 20 years after the US Supreme Court judgment in Continental/Sylvania,²³ which subjected all vertical restraints to "the traditional rule of reason standard,²⁴ Hawk lamented the EU policy toward vertical restraints, which he viewed as the underbrush of detritus of the fast-growing forest of EU competition law.²⁵ Generally, he voiced concerns at the: (i) stubborn adherence to the definition of a restriction on competition as a restriction on the ‘economic freedom’ of operators in the marketplace; (ii) overly broad application of art. 85(1); (iii) deficiencies in economic analysis; and (iv) lack of consideration for market power which should be the threshold issue".²⁶ As noted, these views had become widely shared at the time and had gained particular traction in the years following the introduction of merger control in 1989.²⁷ In truth, they had even permeated the ranks of the Commission.

Among others, an economist in the policy directorate of what was still DG IV, David Deacon, acknowledged openly by then the "emergence of a consensus in economic thinking that appears […] not to be incompatible with the [case law of] the European Courts and which, together with fundamental changes in distribution and logistics, call for more market analysis and a less regulatory approach" to vertical restraints.²⁸ The acknowledgment of the formation of that new consensus, as well as the observation of transformations in distribution structures and business environments triggered by the EU market integration process and the opening-up of economies on a global scale, marked the start of a long-lasting process of policy reforms initiated under the tenure of Commissioner Van Miert and Director-General Ehlermann, and continued by their successors. That process was also shaped by parallel discussions ongoing in other circles, such as the OECD or at the level of national competition authorities.²⁹ Eventually, the reform agenda was kicked off by the publication in January 1997 of the Green Paper on Vertical Restraints in EC Competition Policy,³⁰ which acknowledged upfront the need for a broad policy review that was to translate in a greater emphasis put on the analysis of the economic impact of agreements under what was to become Article 101 TFEU.³¹

The Green Paper on Vertical Restraints was groundbreaking for it aimed to translate the new economic consensus and the new reality of the Union into new policy directions, starting with the recognition that "anti-competitive effects are only likely where interbrand competition is weak and there are barriers to entry", i.e., the focus should lie on market power and horizontal competition "at either producer or distributor level".³² That policy statement had two corollaries: "vertical restrains cannot be considered per se violations of competition law and the enforcement process ought to concentrate on the impact on the market rather than the form of the agreement".³³ In expressing those views, the Commission was departing from established precedents of the EU Court of Justice, none other for example than Consten & Grundig.³⁴ To appreciate the scope of the transformation proposed by the Green Paper on Vertical Restraints, it is worth quoting directly from that judgment: (i) "[T]he absence in the contested decision of any analysis of the effects of the agreement on competition between similar products of different makes [i.e., interbrand competition] does not, of itself, constitute a defect in the decision; (ii) the efforts of [the] dealers are stimulated by competition between distributors of products of the same make" [i.e., intrabrand competition]; and (iii) no further considerations, whether of economic data (…) or [otherwise], and no possible favourable effects of the agreement in other respects, can in any way lead, in the face of the abovementioned restrictions, to a different solution under article 85(1).³⁵ The contrast is striking. And yet the new policy direction had to accommodate the reality of a Union whose foundations lie in the internal market.³⁶ The Commission was therefore keen to point to the existence of an EU exception in the form of a principled stance against absolute territorial protection (or ATP").³⁷ Arguably, the recent Premier League and Pierre Fabre judgments can still be read along the lines of that EU exception.³⁸

The new direction was further articulated in the Vertical Block Exemption Regulation and its accompanying Guidelines adopted in 2000,³⁹ which created a presumption of legality for vertical agreements entered into between entities holding a market share below a certain threshold and excluded the status of per se violation for those agreements falling outside the scope of the block exemption because "vertical restraints often have positive effects" on competition.⁴⁰ That presumption derived from a number of premises restated in the Guidelines, as follows:

• "The protection of competition is the primary objective of EC competition policy, as this enhances consumer welfare and creates an efficient allocation of resources. […] Market integration is [an] additional goal of EC competition policy" (para. 7);

• "The Commission will adopt an economic approach which is based on the effects on the market; vertical agreements have to be analysed in their legal and economic context" (para. 7);

• "[C]ompetition concerns can only arise if there is insufficient inter-brand competition. This will limit the scope of application of Article 81 to undertakings holding a certain degree of market power" (para. 6).

In retrospect, the 2000 Guidelines on Vertical Restraints constitute probably the clearest policy embodiment to date of the reception of the new neoclassical consensus in EU competition law enforcement, together with their companion Horizontal Cooperation Guidelines of January 2001.⁴¹ Indeed, the Horizontal Guidelines also presented market power as the starting point of an effects analysis of agreements between competitors with the view to assessing their ability to "affect negatively prices, output, innovation or the variety or quality of goods and services.⁴² In turn, the Horizontal Guidelines addressed the object/effect dichotomy in relatively clear-cut terms:⁴³ price fixing, output limiting or market sharing agreements are presumed to have negative market effects — they constitute therefore restrictions by object. However, the Guidelines emphasized, many horizontal agreements do not include object restrictions and must therefore be assessed on the basis of their effects or market impact. Generally, both the Vertical and Horizontal Guidelines marked a departure from the economic freedom approach that was still relied upon just a few years before. The 1993 Report on Competition Policy, for example, viewed the exclusive nature of a contractual relationship between a producer and a distributor […] as restricting competition since it limits the parties’ freedom of action in the territory covered.⁴⁴ In contrast, one of the core features of the Guidelines was the recognition that not every agreement which restricts the freedom of action of the participating undertakings, or one of them, necessarily falls within the prohibition of Article 101(1)".⁴⁵

The publication in April 2004 of the Article 81(3) Guidelines completed the substantive modernisation of Article 101 enforcement policy.⁴⁶ They aimed to facilitate the self-assessment by businesses of the compliance of their contracts and other arrangements with then Article 81 EC under the new exception system. They were also to serve as guidance for national authorities and courts endowed from then on with the power to apply Article 81 EC in full. In effect, the Article 81(3) Guidelines make the synthesis mentioned before between, on the one hand, the transformation in the economic consensus and, on the other hand, 40 years of case-law by the EU courts.⁴⁷ In essence, they embody a rebalancing of the competitive analysis between paragraphs (1) and (3) of Article 81 EC/101 TFEU: the stated objective is now to determine the "net effect of agreements on the basis of a method using market power as a key benchmark and relying on the counterfactual method as the basic analytical framework.⁴⁸ The Article 81(3) Guidelines also contain a rare clarification of a defining feature of EU competition law, namely the prevalence ultimately" of the process of competition over potential efficiency gains.⁴⁹ This is the essence of the fourth requirement of Article 101(3) TFEU: EU law does not trust monopolies to deliver efficiencies.

Eventually, the modernized version of Article 101 therefore displays "all the necessary elements of a rule of reason", as one of the authors of the 81(3) Guidelines put it,⁵⁰ "where the anticompetitive aspects of agreements are analysed under [paragraph](1) and the pro-competitive elements are analysed and balanced against the anti-competitive elements under [paragraph](3)". It is a structured rule of reason allocating the burden of proof between the Commission and the defendant(s), but still a rule of reason with the consequence that certain considerations are relevant both to the assessment of the restrictive and of the pro-competitive aspects of agreements, whether they relate to market analysis or proportionality⁵¹. First, under paragraph 1, the Commission bears the burden of articulating a theory of harm in economic welfare terms, that is of likely impact on prices, output and innovation (and quality and choice, if they can form stand-alone grounds), and of supporting it by means of an in-depth market analysis (backed by empirical data) focused on the parties’ market power and therefore on the likelihood of the theory translating into actual effects. Second, under paragraph 3, the defendant bears the burden of substantiating the efficiency gains alleged to derive from the agreement, which means evidencing: (i) the nature of the claimed efficiencies and the extent to which they would benefit consumers; (ii) the link between the agreement and the efficiencies, which must be proportionate; (iii) the likelihood and magnitude of each claimed efficiency gain; and (iv) how and when each claimed efficiency gain would be achieved. Eventually, the likely anti-competitive effects of the provision are weighted against the potential efficiency gains, knowing that at the margin (i.e., in case of doubts) competition trumps efficiency. The threshold question under paragraph 1 is market power. The threshold question under paragraph 2 is proportionality. But on both side of the 101 equation, assumptions and deductions must be backed by market-specific "detailed, robust and compelling" data and facts.⁵² This is the effects-based approach applicable under Article 101 TFEU, though, paradoxically, as explained below, it may be only scarcely applied or even applicable.

C. The Ambition of the Effects-based Approach

As noted, the substantive modernization of Article 101 enforcement aims to promote empirical inquiries into the net welfare effects of agreements entered into by economic actors benefiting from a certain level of market power. It can be contrasted with a pre-modern formal approach based on the lenient justification of limitations put by an agreement on the freedom to trade of the relevant (third-) parties. From the mid-1990s, it became clear that the completion of the EU internal market combined with the opening-up and integration of markets on a global scale led to a series of transformations in business environments, including within the firm, and prompted companies to enter into complex collaborative arrangements whose outcome has become increasingly difficult to predict. In that context, the ex ante vetting of agreements by means of formal analyses based on a priori theories of harm appeared all the more inappropriate and prone to generate errors, of all types. Hence, the modernization of EU competition law was supposed to entail a qualitative leap forward and a greater sophistication of the competitive analysis of agreements, notably to limit Type I and Type II errors, i.e., false positive (excessive credulity) and false negative (excessive scepticism). Ten years later, one may wonder whether it did succeed in that ambition. Many cases decided since the early 2000s testify of a general rationalization of the competitive assessments and it appears, indeed, that concrete cases endeavoured to correct past errors.

The 2007 MasterCard decision, for example, which is probably the paradigmatic effects-based case to date⁵³, clearly endeavoured to adjust the Commission’s previous approach toward Multilateral Interchange Fees (MIF) in the payment card sector, and can therefore be viewed as an attempt to redress a Type I error. While in its Visa II decision the Commission had stated that "an interchange fee agreement can in principle contribute to economic and technical progress within the meaning of Article [101](3) of Treaty",⁵⁴ it emphasized in MasterCard that there was "no presumption that MIF’s in general enhance the efficiency of card schemes as there is no presumption that they do not fulfil the conditions of Article 81(3) of the Treaty and are therefore illegal.⁵⁵ Rather it underlined that [a] MIF may be used by banks to achieve efficiencies as well as to extract rents, so that the Commission’s conclusion on the efficiencies of a MIF will depend on the concrete [empirical] evidence brought forward by the parties and, conversely, cannot be determined in a general manner by economic theory alone".⁵⁶ Eventually, the Commission decided that the MasterCard MIF did not meet the conditions of paragraph 3 of Article 101 TFEU, whereas it had reached the opposite, positive, conclusion in Visa II⁵⁷. The case against Visa was subsequently revisited and led to the adoption of a commitment decision in 2010.⁵⁸

In turn, the correction of Type II errors is typically the province of the EU Courts, in the exercise of judicial review over Commission decisions. Even though the role of the EU Courts in the substantive modernisation process is rather paradoxal, as explained below, the 2006 judgment of the General Court in O2 could be construed as an attempt to address a Type II error, irrespective of the exemption granted initially to the roaming agreement between T-Mobile and O2, pursuant to Article 81(3) EC.⁵⁹ Indeed, the applicant in that case argued that the roaming agreement examined by the Commission did not entail a restriction of competition in the first place and thus fell outside of the scope of Article 81(1) EC altogether. The General Court upheld the applicant’s plea for, in its view, the Commission decision "suffers from insufficient analysis, first, in that it contains no objective discussion of what the competition situation would have been in the absence of the agreement, which distorts the assessment of the actual and potential effects of the agreement on competition and, second, in that it does not demonstrate, in concrete terms, […] that the provisions of the agreement on roaming have restrictive effects on competition, but is confined, in this respect, to a petitio principii and to broad and general statements".⁶⁰ The sharp conclusion of the General Court was not appealed by the Commission, which might be viewed as an acknowledgment of the pre-modern character of its initial assessment.⁶¹ However, if the effects-based approach has tightened up the competitive assessment of agreements conducted under Article 101 TFEU, it is not immune of paradoxes.

II. The paradoxes of the effects-based approach

The move toward an effects-based approach under Article 101 TFEU has been a long and necessary journey,⁶² but not one immune of paradoxes. These side-effects of the substantive modernization process need to be addressed in order to ensure the successful completion thereof and its sustainability over the longer term. The remainder of this contribution endeavours to discuss five apparent paradoxes, which are interrelated. The first one is a drop in enforcement activity, beyond cartels. The second one consists of the increase in enforcement costs, both for the Commission (decision costs) and defendants (compliance and justification costs). The third paradox lies in the fact that competitive assessments rely more than ever on general presumptions, categories and abstract guidance found in Block Exemption Regulations and Guidelines. The fourth one pertains to the position of the EU Court of Justice, which has turned from a progressive into a conservative actor in the realm of EU antitrust enforcement. The fifth paradox builds on all previous ones and begs the question: isn’t the substantive modernization of Article 101 TFEU fundamentally a process at war with itself?

A. Paradox #1: Drop in enforcement activity, beyond cartels

A review of all decisions adopted pursuant to Article 81 EC/101 TFEU between the first of January 2000 and the first of January 2011 reveals that, excluding hardcore cartels, the Commission has issued altogether 18 infringement decisions and 10 commitment decisions. This is in addition to 6 negative clearance decisions and 18 exemption decisions adopted under Regulation 17/62. Among the 28 infringement and commitment decisions, 14 related to vertical restraints and 14 to horizontal restraints. Among the 18 infringement decisions, 10 are posterior to the entry into force of Regulation 1/2003 but