Chillin'Competition

The amount of commentary to which the Intel ruling has given rise in less than six months is quite extraordinary. There is every reason to welcome the debate and exchange of views. Given the interest in the topic, I thought it would make sense to develop the ideas I sketched in a post published back in June. Click here to download the paper. Your comments would be most welcome (e-mail: P.Ibanez-Colomo@lse.ac.uk).

I started my LLM in Bruges in September 2003. A couple weeks into the programme, Michelin II came out. I remember vividly the discussions we had in class about it. Even for a postgraduate student with no experience it was easy to understand the significance of the judgment and its implications. 11 years have passed, I am now the one teaching LLM students, and the relevant case law continues to generate considerable controversy. It seems clear that the debate touches upon some fundamental questions in EU competition law.

The lack of clarity – in spite of the frequent and lively discussions – about the issues that are really at stake is what prompted me to write the paper. I argue that the controversy, while being of major importance, is far more limited in its nature and scope than commonly assumed. Upon closer scrutiny, it seems to me that the case law on exclusive dealing and rebates is criticised not so much because there is a disagreement about the objectives of EU competition law, but due to the tensions (or ‘frictions’, as I call them in the paper) that have arisen in the case law.

What the controversy is NOT about: This endless controversy has nothing to do with the so-called ‘more economic approach’ or ‘more economics-based approach’. I never liked these expressions. They suggest that economic analysis has been introduced by the European Commission from the top down in an attempt to change EU competition law. Nothing could be further from the truth. As I show in the paper. EU courts have always relied upon mainstream economic analysis to shape EU competition law (just think, among the many examples, of Woodpulp II or AKZO) and have frequently taken the lead in this sense. Tetra Laval and Airtours, both of which unambiguously embraced mainstream positions, are indispensable to make sense of the policy shifts observed in the course of the past decade.

Similarly, this debate is NOT about the objectives of EU competition law. As I explain at length in the paper, the current approach to exclusive dealing and rebates is not the only conceivable one under the Treaty. The prohibition, absent an objective justification, of exclusive dealing and loyalty rebates does not follow logically from the fact that the objective of EU competition law is to create a system of undistorted competition. Likewise, there is nothing in the said objective that is inimical to the use of efficiency considerations to shape the law, as rulings like AKZO, Deutsche Telekom, TeliaSonera and Post Danmark show.

How I see this controversy: EU competition law evolves and is shaped from the bottom up in an incremental manner. As a result, tensions in the case law are inevitably bound to arise. Some rulings may hint at a particular logic, while others at a different one. There is nothing wrong with that. Quite to the contrary, it is the very reason why law is fascinating as a scholarly discipline. Over time, doctrines are refined, and contradictions between cases addressed, through the interaction between (and disagreement between) judges, officials, lawyers and scholars. Nothing else is going on in relation to Intel. As I see it, the real (and only) issue at stake is that the case law on exclusive dealing and rebates is difficult to reconcile with Article 101 TFEU and Article 102 TFEU case law applying to the same or comparable practices. 

• Rules and standards in Article 102 TFEU case law: A first inescapable fact is that the ECJ requires evidence of an anticompetitive effect in some cases but not in others. Exclusive dealing and rebates conditional upon exclusivity are prohibited absent an objective justification. Intel does not depart from previous cases in this regard. Such practices can be said to be subject to a (qualified) prohibition rule. Other practices, including refusals to deal, are subject to a standard. The ECJ made it clear in Deutsche Telekom that an anticompetitive effect is necessary to establish that a margin squeeze is abusive within the meaning of Article 102 TFEU case law. Similarly, the selective price cuts at stake in Post Danmark were not deemed abusive by their very nature by the Court. As a result, action under Article 102 TFEU was made dependent on evidence of exclusionary effects.
• Making sense of the case law: It is necessary to make sense of these two lines of case law, i.e. why the Court required evidence of an anticompetitive effect in some cases but not in others. What becomes clear from the outset is that the justifications most commonly advanced to explain the case law on exclusive dealing and rebates are unconvincing. It is often claimed that a qualified prohibition rule is appropriate on a market were competition is weak due to the presence of a dominant firm. Competition was also weak in Deutsche Telekom and Post Danmark (these firms, let us not forget it, were the incumbents operators in a recently liberalised industry), but this fact did not preclude the Court from requiring evidence of exclusionary effects. It is also argued by some authors that the relative rigidity of a prohibition rule is appropriately tempered by the availability of an objective justification. Again, an objective justification was available in the two abovementioned cases (Post Danmark is quite eloquent in this regard), but the Court still made it clear that it was necessary to show an anticompetitive effect to establish an abuse.
• Abusive practices ‘by their very nature’: As I understand the case law, the reason why the Court requires evidence of exclusionary effects in some rulings but not in others has to do with the fact that some practices are considered to be abusive ‘by their very nature’ (in the same way that some agreements are deemed to restrict competition by object). Pricing below average variable costs is deemed abusive because it is presumed to be driven by an anticompetitive intent. The practice is understood to have no plausible explanation other than the exclusion of competition. Exclusive dealing and rebates conditional upon exclusivity are also presumed to be abusive by their very nature since Hoffmann-La Roche. Other practices, like refusals to deal and margin squeezes, are not seen in the same light. This makes sense. After all, even dominant firms are entitled, as a matter of principle, to deal with whom they please.
• In light of experience and economic analysis, a standard seems more appropriate for exclusive dealing and loyalty rebates: In Groupement des Cartes Bancaires, the Court emphasised the importance of experience and economic analysis when determining whether an agreement is restrictive of competition by its very nature. If we apply the same framework to potentially abusive practices, it becomes clear, in my view, that exclusive dealing and rebates should be assessed in the same way ‘margin squeezes’ and selective price cuts a la Post Danmark are. It would be wrong to assume that exclusivity, whether it is achieved directly or indirectly, has exclusionary effects. This is one of the reasons why Michelin II and British Airways were so controversial at the time. The market share of the dominant firm had declined, and rivals had been able to thrive in spite of the practices. Likewise, there is no reason to presume that exclusivity is driven by an exclusionary intent. Exclusivity is implemented by firms that lack the ability to foreclose competition (the market share of the supplier in Delimitis was around 6-7% and around 4-5% in Brasserie de Haecht). In British Plasterboard and Van den Bergh Foods, the General Court seemingly accepted that exclusivity has pro-competitive motivations, even when implemented by dominant firms. In his meticulous analysis of Michelin II, Massimo Motta identified several factors that were incompatible with an exclusionary strategy (for instance, some of the rebate schemes were similar in their nature and purpose to those examined by the Court in Delimitis). If exclusive dealing and rebates do not always, or almost always have exclusionary effects and are not always, or almost always driven by an anticompetitive intent, it would make sense to make them subject to a standard. It would make sense, in other words, to conclude that they are not abusive ‘by their very nature’. This would bring Article 102 TFEU case law in line with Delimitis, where the Court came to the same conclusion in relation to Article 101 TFEU.