Restrictions of competition by object under Article 101(1) TFEU: chapeau bas, Prof Wahl!
Note by Alfonso: Advocate General Wahl’s Opinion in Groupement de Cartes Bancaires out on Friday, and its take at clarifying the object-effect conundrum is remarkable. Pablo Ibañez Colomo offers his views on the Opinion below:
Advocate General Wahl’s opinion in Groupement des Cartes Bancaires v Commission (published last Friday, and available in French and in Greek only for the time being) is a model of lucidity and flexible thinking. It is also very much in line with an article of mine on the subject, but that is plain irrelevant. What matters, and what makes this opinion remarkable, is that it manages to capture the logic underlying the existing case law addressing the boundaries between restrictions by object and by effect. Many commentators and some advocates general have tried in the past few years to identify the elusive factors that should be considered when establishing whether an agreement restricts competition ‘by its very nature’. Paragraph 56 of the opinion sets out a formula that is, in my view, more accurate and elegant than any previous attempt (the fact that I am forced to read it in French for the moment probably adds to the latter):
‘Ne devraient donc être considérés comme restrictifs de concurrence par objet que les comportements dont le caractère nocif est, au vu de l’expérience acquise et de la science économique, avéré et facilement décelable, et non les accords qui, au vu du contexte dans lequel ils s’insèrent, présentent des effets ambivalents sur le marché ou qui sont porteurs d’effets restrictifs accessoires nécessaires à la poursuite d’un objectif principal non restrictif de concurrence’.
In other words, what really matters is whether, given the context in which it is concluded, an agreement is a plausible source of efficiency gains. Thus only those agreements that have no credible redeeming virtues are understood to restrict competition by object. A careful reading of the relevant case law shows, in my view, that this is the ‘default methodology’ (which is the expression I use in my article) – or, if one prefers, ‘l’appréciation plus standardisée’ (as Advocate General Wahl writes in his opinion) – followed by the ECJ when it examines the nature of agreements under Article 101(1) TFE. The methodology changes, and rightly so, when market integration as an objective is directly at stake in a case (as is true of agreements restricting parallel trade).
From Societe Technique Miniere to Pronuptia and Delimitis, and from Remia to Wouters and Asnef-Equifax (to mention just a few landmark rulings), the ECJ has followed the same approach, which revolves around an analysis of the rationale behind the agreement. The Court typically seeks to identify the reasons why two or more firms would introduce some restraints in an agreement. If it appears that such restraints are a plausible means to achieve legitimate business objectives, it concludes that the agreement does not restrict competition by its very nature. In Groupement des Cartes Bancaires, the parties to the agreement claimed that it was intended to address free-riding issues and therefore that it did not have a restrictive object. In light of the relevant case law, the question in these proceedings is whether this story is a credible one given the nature of the agreement and the context in which it was concluded.
The opinion is notable for other reasons, of which I mention a couple:
– It is sometimes claimed that the category of ‘object restrictions’ captures those agreements that can be presumed to have anticompetitive effects (the famous speed-limit analogy and variations thereof). This interpretation of the notion is problematic insofar as it sits at odds with the principle, well established in the case law, whereby an agreement may restrict competition by its very nature irrespective of the effects it produces. Advocate General Wahl emphasises, in this same vein, the importance of distinguishing between the analysis of the nature of the agreement and the analysis of its effects. If the question of whether an agreement restricts competition by object depends on its presumed effects, the two would be confused. The rulings mentioned above indeed confirm that the two are separate steps and that the Court has been careful not to mix them (and has rightly reacted when the General Court has done so, as in Glaxo Spain – also discussed in the opinion).
– The opinion shows that, when confined to its role, the use of economic analysis can be very useful and, more importantly, wholly uncontroversial. Advocate General Wahl does not rely on economic analysis for normative purposes (that is, to state how the law should be, or to claim that the case law is misguided), but as a tool (among others) to make sense of a legal issue. Economics is used in the opinion, in other words, as a guide – a code – to decipher a complex reality. I hope this opinion contributes to a more fluid dialogue between disciplines. I was pleased and surprised to even find a reference to Rochet and Tirole’s ground-breaking work on two-sided markets – which, as you all know by now from Alfonso’s last post, is ‘the single most important and fascinating subject in contemporary antitrust (and beyond)’.
Lastly, I will also mention that writing this post brings very good memories of a great seminar (and even better post-seminar!) to which Luis Ortiz Blanco and Alfonso invited me last year and in which I had the chance to discuss these questions with some luminaries from the Commission.
Chillin'Competition 
I respectfully disagree: the speed limit analogy is useful precisely because you can punish something without effects (!) simply because there is a high likelyhood of such effects occurring (speeding is not a problem in itself,the problem is that there is a high likelyhood that an accident would happen,hence speeding is prohibited in the first place). To quote from a perfectly analogous US example on the per se rules (from justice stevens):
“The per se rules in antitrust law serve purposes analogous to per se restrictions upon, for example, stunt flying in congested areas or speeding. Laws prohibiting stunt flying or setting speed limits are justified by the State’s interest in protecting human life and property. Perhaps most violations of such rules actually cause no harm. No doubt many experienced drivers and pilots can operate much more safely, even at prohibited speeds, than the average citizen.”
Asimo
31 March 2014 at 10:05 pm
Hi Asimo,
One can interpret the notion of restriction by object in two ways:
– One can claim that the notion encapsulates a presumption of anticompetitive effects, as you say.
– One can say instead that the notion captures those agreements that are unlikely to have pro-competitive effects. If these agreements have effects at all, these will be negative.
A close analysis of the case law makes it quite clear, I would say, that the Court follows the second interpretation. Advocate General Wahl seems to share this view. When examining whether an agreement has a restrictive object, the Court does not examine whether it is likely to have anticompetitive effects, but whether it has a pro-competitive dimension. Exchanges of information illustrate these ideas very well. T-Mobile rules out the role of effects when establishing an object restriction; John Deere makes it clear that exchanges of information sometimes have a pro-competitive dimension and thus not necessarily impact negatively on competition in such cases; finally, Asnef-Equifax provides an example of an exchange of information which, because of its pro-competitive dimension, is not restrictive by its very nature and can even fall outside Article 101(1) TFEU altogether depending on the market context.
I would add that the interpretation you favour is not operational. You say that the notion refers to agreements that are very likely to have negative effects on competition. All fine, but the question would remain: what criteria do you use to identify these agreements in the first place? Formal criteria like ‘price-fixing’ or ‘market sharing’ do not mean anything in themselves. If an agreement is a source of efficiency gains it will be found not to restrict competition by object even when it has a price-fixing component (think of joint production agreements or of research and development agreements that provide for the joint exploitation of the results).
Pablo Ibanez
1 April 2014 at 11:57 am
You raise a very interesting debate: whether the qualification as a restriction by object is determined by pro- or by anticompetitive effects. I would be highly surprised though if the Court will choose either way. Why would it? By taking one way the Court would foreclose itself to the other way. Less wriggle room. Moreover, by focusing predominantly on pro-competitive effects, the Court would be compelled to label an agreement that has little (or even no) potential anticompetitive effects, but also no pro-competitive effects, a restriction by object. And that would be hard to defend. First, practitioners would point out that it does not square with the text of 101 TFEU. Second, it would also be out of line with the US approach, as Asimo points out (which at least partly inspired 101 TFEU). Third, it would be quite unfair to citizens (employees), who ‘feel’ that a real potential of anticompetitive effects may be unjust – but may have less intuition about a lack of pro-competitive effects. And, fourth, it does not fit into the current catch-all approach the Court has followed in the past. See e.g. Allianz Hungary :
“38. The Court has, moreover, already held that, in order for the agreement to be regarded as having an anti-competitive object, it is sufficient that it has the potential to have a negative impact on competition, that is to say, that it be capable in an individual case of resulting in the prevention, restriction or distortion of competition within the internal market.”
The potential negative impact on competition is judged along with all other factors, if I read it correctly – “the content of its provisions, its objectives and the economic and legal context of which it forms a part”, but also “the nature of the goods or services affected [and] the real conditions of the functioning and structure of the market or markets in question” and also, if the national courts woul like to, “the parties’ intention”. No factor seems to be off-limits for leading into the objects box. The lack of any pro-competitive effects may really matter, of course. Just like other factors.
stefan
2 April 2014 at 2:34 pm
Pablo: if you assess welfare enhancing effects under 101 (1) as you propose (or as you say the case law does) what do you then leave for 101 (3) ?
Aside from the undeniable expansion of the “object box”, doesn’t the problem rather lie in the practical unavailability of 101(3)?
Alfonso Lamadrid
2 April 2014 at 11:41 am
Hi Alfonso,
You do not assess the pro-competitive effects under Article 101(1) TFEU, you assess whether the agreement has a pro-competitive justification. In other words, the Court evaluates whether the agreement has the potential to lead to efficiency gains.
And yes, the Court has consistently followed this approach. My favourite example is Delimitis (paras 10-12, where it examines the business rationale underlying exclusive dealing arrangements). But think of Pronuptia (paras 13-15) or Asnef-Equifax (paras 46-48). If you read my article, as you should, you will find many more examples of what is a very marked trend in the case law 😉
Groupement des Cartes Bancaires is a very good example too. Advocate General Wahl tries to understand the logic behind the contentious restraints and finds the free-riding justification quite compelling. Providing for such restraints was reasonable. As a result, he did not find it justified to conclude that the agreement was restrictive by its very nature.
Pablo Ibanez
2 April 2014 at 12:12 pm
Granted, but doesn’t the fact that even object restrictions can (theoretically) benefit from Art. 101(3) mean that a 101(1) assessment by definition never rules out that they “may have the potential to lead to efficiency gains”?
I agree with you, and with the AG, on the fact that the Court has resorted to the approach you outlined in your article (which I of course read, on a train back from Luxembourg actually), but don’t you think it may have done it in view of the little flexibility and practical unavailability of 101 (3)?
Alfonso Lamadrid
2 April 2014 at 3:02 pm
To Alfonso: If an agreement is not a plausible source of efficiency gains, it will most probably fail to meet the conditions set out in Article 101(3) TFEU. Does it mean that Article 101(3) TFEU is only relevant for agreements that restrict competition by effect? Well, it is clear to anyone that cartels (the foremost object restrictions) would most probably fail to meet the four conditions of the provision. But do not forget that parallel trade restrictions are treated differently (they may lead to efficiency gains, but they are deemed to restrict competition by object because they run counter to the market integration objective). Article 101(3) TFEU would be very relevant in such cases.
And I do not think the Court necessarily followed that approach in light of the limited availability of Article 101(3) TFEU in practice. The structure of Article 101(1) TFEU demands that a boundary be drawn between object and effect restrictions, and, as far as the first is concerned, it makes intuitive sense to focus on the rationale underpinning the agreement.
To Stefan: Thanks for your comment. Do not forget that we are not discussing a theoretical possibility or an issue raised for the first time before the Court. We are examining decades of case law. The Court has repeatedly considered the pro-competitive dimension of an agreement when examining whether it is restrictive by object. Advocate General Wahl simply follows a well-established approach and identifies the logic underpinning this default approach.
Agreements with no pro-competitive justifications have the potential to impact negatively on competition. If they have effects at all, these effects will be negative. I do not see any contradiction between the approach described above and the idea that the mere potential of an agreement to restrict competition is sufficient to establish an object restriction.
Finally, you say that ‘by focusing predominantly on pro-competitive effects, the Court would be compelled to label an agreement that has little (or even no) potential anticompetitive effects, but also no pro-competitive effects, a restriction by object. And that would be hard to defend’. I respectfully disagree with this view. Cartels are inherently unstable. Firms have incentives to collude and to cheat. What if a group of firms meets to discuss future prices (say, a la T-Mobile) but the potential effects on prices of this meeting fail to materialise because the parties never adopt a common course of conduct? An exchange of information of this kind (future prices, individual firms, non-public information) is clearly an object restriction. I fail to see anything new, controversial or ‘hard to defend’ in this conclusion. This is well established. The fact that the negative effects that such exchange has the potential to produce are never manifested does not change anything in this regard.
Pablo Ibanez
2 April 2014 at 3:54 pm
I don’t see the problem in saying that agreements that are anticompetitive by object are those which necessarily produce anticompetitive effects. Is this really at odds with the case law saying that an agreement may be anticompetitive by reason of its object independently of the effects it produces? Doesn’t that case law refer to effects actually produced ? As opposed to those that are not actually produced but would nevertheless necessarily be produced by such an agreement? Isn’t the point to tell parties that they can’t defend themselves by showing that no actual harm was done in the case at hand?
This leads us to the speeding analogy. I don’t believe it is a very apt one, but or a different rerason than Pablo. It is a “brightline” rule, as opposed to one leaving discretion to the judge (e.g. a rule prohibiting “dangerous” driving, leaving it to the judge to determine what is “dangerous” and what isn’t). But the “object” box is far from being a “brightline” rule; based on the meaning that the Commission thinks it has (which cannot even be summarized easily since it encompasses all sorts of different behaviour, old and new), it leaves a great deal of discretion to whoever is applying it (and correlatively makes it very difficult for parties to predict what “object” means). Of course, the analogy is a lot better suited in the US where “per se” restrictions are clearly identified.
This is also where I would tend to disagree with Pablo; it is true that “price-fixing” is not necesarily as clear-cut a category as “no driving above 50mph” but the US antitrust practice seems to show (in my opinion) that the limitative list of “per se” infringements provide good level of legal certainty. It is also easy enough to implement, both by parties and agencies (and maybe easier than a test based on the possible redeeming pro-competitive effects of the agreement). It seemed to me that this is what AG Wahl is getting at in the last sentence of para 55.
ag
2 April 2014 at 5:59 pm
Hi ag,
I really should not go on, but let us makes this my last comment on this topic (until my next comment). I tell myself that the importance of the topic justifies it.
It is pretty clear to me that, in paras 55-56 of his opinion, Advocate General Wahl explains that the object category is inappropriate for agreements that have ambiguous (ambivalent) effects on competition. Where the agreement has both pro- and anticompetitive effects, it only makes sense to apply Article 101(1) TFEU following a careful assessment of its negative impact on competition. Where, on the other hand, an agreement is not a plausible source of pro-competitive effects, it makes sense prohibit it under Article 101(1) TFEU even absent anticompetitive effects. This is how I interpret the last sentence in para 55. Because no efficiency gains can be expected from the agreement, one can safely jump to the conclusion that it is contrary to Article 101(1) TFEU ‘dans un souci d’economie procedurale’.
The key question, which you do not address in your message, relates to the criteria one should use to identify the agreements that fall under the object category. And here again, Advocate General Wahl’s opinion is invaluable. Agreements that do not have ambiguous effects on competition (in the sense that they do not have a pro-competitive justification) are identified based on experience and economic analysis, not based on formal categories (price-fixing, market sharing and so on), which do not mean anything in themselves. When we say ‘price-fixing’, we really mean ‘an agreement between competitors that can only be plausibly explained as an attempt to eliminate price competition among them and which has no credible redeeming virtues’.
On object restrictions and effects. In a sense we agree. Agreements with no pro-competitive virtues can only have, if at all, negative effects on competition. In a sense we disagree. The fact that such agreements can only be expected to impact negatively on the competitive process does not mean that this is the key factor to qualify them as ‘object restrictions’. The key factor is the rationale, the very point of the agreement. Why would two or more firms conclude such an agreement?
Pablo Ibanez
3 April 2014 at 8:33 am
[…] while ago I wrote a post and engaged in some follow-up comments on the issue of restrictions by object. But since Alfonso is busy these days and has shown some […]
More on AG Wahl and restrictions by object: ‘pay-for-delay’ settlements as a case study | Chillin'Competition
22 May 2014 at 12:18 pm