Persistent myths in competition law (I): ‘the pro-competitive aspects of an agreement can only be considered under Article 101(3) TFEU’
EU competition law keeps busy a growing number of officials, practitioners and academics. It is fascinating that, with so many professional lives devoted to it, with so many specialised conferences and journals, there are still many myths that refuse to die. Some received ideas are incredibly powerful. They keep being repeated as uncontroversial truths even when they are patently at odds with the case law.
I thought it would be a good idea to do something more than simply express my fascination: identify what these myths are and explain why they are unfounded.
The idea of writing a series of posts on myths in EU competition law came when I thought about the statement you find in the title. To this day, some people remain convinced that the pro-competitive aspects of agreements can only be evaluated under Article 101(3) TFEU. It is perhaps the most popular myth around.
According to this view, claims about the ways in which the agreement could create (or improve the conditions of) competition would have no role to play under Article 101(1) TFEU.
This mantra may be very popular. For better or worse, it is unambiguously contradicted by a consistent line of case law.
The pro-competitive aspects of an agreement play, and have always played, a central role under Article 101(1) TFEU
The abovementioned line of case law dates back to the very early days. In Societe Technique Miniere, the Court already ruled that an agreement that is ‘really necessary’ for a supplier to enter a new market is not restrictive of competition, whether by object or effect.
Another one of my favourite examples is Asnef-Equifax. You will remember that the case was about an information exchange system available to credit providers. The point of the system was to allow the said providers to figure out the likelihood of repayment by potential borrowers.
The Court concluded that the system was not restrictive by object. As part of the reasoning, it noted that information exchange mechanisms of that kind ‘are in principle capable of improving the functioning of the supply of credit’ (paras 46, 47 and 55) and ‘of increasing the mobility of consumers of credit’ (para 56). I do not believe the reasoning can get much more explicit than this.
A similar example (and another all-time favourite) is Gottrup-Klim. A group of competitors decide to buy an input in common. Cartel arrangement? Could be (buyers’ cartels do exist and are sanctioned as such), but not in the specific circumstances of this case. The joint purchasing agreement at stake in Gottrup-Klim was different from a cartel and thus not restrictive by object. Crucially, the Court noted, in this regard, that the agreement in question may ‘make way for more effective competition’ (para 32).
I could go on for a while, but it makes sense that I mention two relatively more recent cases.
Pierre Fabre is arguably as explicit at Asnef-Equifax. The Court held that clauses in an agreement that would otherwise be restrictive by object fall outside Article 101(1) TFEU if there is an ‘objective justification’ for them (para 39 of Pierre Fabre).
How do we figure out if an agreement is objectively justified within the meaning of Pierre Fabre? By ascertaining (para 40) whether the agreement in question aims at the ‘attainment of a legitimate goal capable of improving competition’ (in other words: by identifying its pro-competitive aspects).
Our last station in this overview has to be Cartes Bancaires. Just remember (para 74) that the crucial factor to rule out the ‘by object’ characterisation was the fact that the Court accepted that the contentious clauses could be understood as a means to tackle free-riding issues.
If the pro-competitive aspects of the agreement count under Article 101(1) TFEU, what is the role of Article 101(3) TFEU?
The above examples make it sufficiently clear that the pro-competitive aspect of an agreement are not only relevant, but crucial under Article 101(1) TFEU – and have always been. If there is anyone disputing this conclusion in spite of the examples given, please leave us a comment. I would really love to know your thoughts.
I believe I understand why the myth is so persistent in spite of the overwhelming evidence to the contrary. Article 101(3) TFEU is the forum in which the pro-competitive aspects of an agreement are balanced against its anticompetitive aspects. Would Article 101(3) TFEU not be completely devoid of purpose if the former were also considered in the first paragraph of the provision?
To which my answer is: no. And this, for two reasons.
First, the exercise conducted under, respectively, the first and the third paragraphs of Article 101 TFEU, is different. For the same reasons, the pro-competitive aspects of the agreement are considered for very different purposes too under each paragraph.
In the context of Article 101(1) TFEU, the pro-competitive aspects of an agreement are indispensable to figure out its object. Put differently: it is simply impossible to evaluate the objective purpose of an agreement (i.e. what it can objectively achieve, irrespective of the subjective intent of the parties) without considering whether it is a plausible source of pro-competitive gains.
This is something that the Court understood from the outset, and that has emerged as a pillar of the case law.
Second, the intensity of the analysis is very different under the first and the third paragraph. The pro-competitive aspects of an agreement are considered in two stages.
Asnef-Equifax, Pierre Fabre and Cartes Bancaires (as well as others like Delimitis or Pronuptia) all tell you that the analysis of the pro-competitive aspects of an agreement is relatively superficial under the first paragraph.
Under Article 101(1) TFEU, the question is whether the agreement is a plausible source of pro-competitive gains. A cursory analysis suffices, and no quantification is needed. The General Court in MasterCard (2012 judgment) made a similar point (para 80).
The analysis is much deeper under Article 101(3) TFEU (once the burden of proof shifts). Then, it is for the parties to show that it is likely that the pro-competitive gains generated by the agreement will weigh more than the anticompetitive effects resulting from it.
As can be seen, there is a role for the two paragraphs: they fulfil a different role and the intensity of the analysis is also very different.
If you ask me: the case law makes a lot of sense.
It makes sense for the the Court to say (as it does): if the agreement is a plausible source of pro-competitive gains, then it can only be prohibited if the authority or claimant establish that it is likely to have appreciable effects on competition. And it makes sense to engage in a full-blown balancing of the pro- and anticompetitive once the latter are shown.
It is also sensible that the burden of proof shifts when the agreement is not a plausible source of pro-competitive gains. Want to claim that a cartel is on balance pro-competitive? Well, experience and economic analysis tell us that such a claim is implausible, but the possibility of claiming the implausible exists under Article 101(3) TFEU. Just think of BIDS, another key judgment where this point is made.
Chillin'Competition 
Thank you Pablo, this is clearly right. I would add last year’s judgment by the UK Competition Appeal Tribunal as a useful source, since it discusses the relationship between art. 101(1) and art. 101(3) in some detail. https://www.catribunal.org.uk/sites/default/files/2018-10/1279_Ping_Judgment_CAT_13_070918.PDF
The only thing I would say is that this whole mess cannot be properly understood without recognising that the concept of “objective justification” under art. 101(1) is really a holdover from the time before art. 6 of Regulation 1/2003, i.e. before national courts had the power to apply art. 101(3). I would imagine that at some point in the future the ECJ is going to start cutting down on objective justifications (and on ancillary restraints).
Martinned
13 May 2019 at 3:22 pm
We borrow this for the US, starting labeling as pro-competitive agreements which are ‘only’ pro-consumer. I would go further and say that pro-competitive elements of anticompetitive agreements can only be demonstrated within 101(1). All benefits explicated in 101(3) are either pro-growth, pro-consumer or pro-competitors.
O
13 May 2019 at 4:20 pm
Thank you Pablo! If only competition authorities would agree that it is indeed a myth that pro-competitive aspects can only be considered under article 101(3)….
Charlotte Forno
14 May 2019 at 8:51 am
Dear Pablo,
I do not think Asnef-Equifax has anything to do with the point you are trying to illustrate. The acid test for restriction by effect in paras 58 and 59 of that judgment (see also 61 and 62) was whether the system was “capable of revealing the market position or the commercial strategy of competitors,” or “liable to reduce uncertainty as to the risks of competition” what could be the case in highly concentrated markets. This was not a concentrated market, so no problem. Access had to be non-discriminatory as well (para 60). If it had been a market where the exchange of info system was “capable of revealing the market position or the commercial strategy of competitors” the ECJ would have been likely (or even certain) to rule that the system was restrictive by effect, whatever the benefits you describe and that the ECJ indeed mentioned (only) at the beginning of its reasoning. To put it differently, in the absence of anticompetitive effect the benefits which may have resulted were irrelevant, in law.
I have not checked the other precedents. I hope they are better chosen.
Joan
14 May 2019 at 9:31 pm
Dear Joan,
We do not disagree at all. That is a very accurate description of what the analysis of effects entails. There is no doubt that the benefits of the agreement are not relevant when it comes to the evaluation of these effects.
But please note that I was discussing, in my post, the object of the exchange. As you know, the Court concluded in Asnef-Equifax that the exchange was not restrictive by object. At the object stage, the pro-competitive benefits played a crucial role.
Thanks for your participation
Pablo Ibanez Colomo
15 May 2019 at 6:21 am
Thanks you, Pablo. This is very interesting. If it’s not too much to ask for, a post clarifying whether the myth also applies to 102 (where the object/effect dichotomy is not as clear) would be very interesting!
Henry
15 May 2019 at 9:28 am
Thanks for the idea, Henry! And also thanks for reading us! It is definitely something that fits with the scope and spirit of this series of posts.
Pablo Ibanez Colomo
15 May 2019 at 12:27 pm
I see. The general title of the post was a bit more ambitious. If benefits only make the analysis move to the “effects” box, but under the “effects” box the benefits are irrelevant to rule out the application of art 101(1) TFEU, I am not sure this takes us very far. As regards Asnef-Equifax, your analysis would (implicitly) suggest that the ECJ would have said that there was a “by object” restriction if it had not found the benefits, but I have reread the judgment and nothing in it suggests that. I cannot imagine why the ECJ would have ruled that the exchange of info was a “by object” restriction, even in the absence of benefits. The difference between object and effect is not always clear-cut and a number of considerations (not just benefits) have played a role in the analysis, depending on context. You may be destroying a myth that never existed.
Joan
15 May 2019 at 1:11 pm
Joan: you suggest that I might be destroying a myth that has never existed. I take this as meaning that you agree with me that the pro-competitive aspects of an agreement are relevant when evaluating whether it is restrictive by object. Great that we agree!
The above said, the myth has well existed and, I strongly suspect, is still alive and kicking. Take a look, just to give you a good example, at the amicus submissions of the Commission in the French MIF case before the Cour de Cassation. See here: http://ec.europa.eu/competition/court/mif_29102012_fr.pdf and here: http://ec.europa.eu/competition/court/mif_17022015_fr.pdf
In the first, for instance, the submission goes: ‘en vertu d’une jurisprudence bien etablie, l’economie de l’article 101, paragraphe 1, du TFUE ne permet pas d’ecarter la qualification de restriction de concurrence par objet au motif que l’accord en cause presenterait un bilan economique positif ou produirait des avantages economiques’ (with an emphasis on ‘ou produirait des avantages economiques’).
So, at the very least in some Commission quarters, it looks like there are some believers. In case you were wondering: the ‘jurisprudence bien etablie’ mentioned in the submission is confined to a single judgment that never made it to the ECJ.
Thanks a million again for your active involvement!
Pablo Ibanez Colomo
15 May 2019 at 6:02 pm
Honestly, I think you read too much in that amicus, in the case law, and in my comments as well. Context is often lost. The case law suggests that benefits may have played, in some cases, a role in the analysis if there was an object. In some cases. In some circumstances. In situations were perhaps anyway the ECJ could have reached the same result (Asnef-Equifax, where the Commission pleaded that there was no object by the way). However, you turn a factual assertion in some cases into a normative statement and there you lost me. The case law as never said that as soon as there are some benefits there is no object, I am afraid. I would guess that the amicus simply means that. It is not “enough” to show benefits to escape the object box. This does not exclude that in given cases they may have played a role in the overall factual assessment of the ECJ.
Joan
16 May 2019 at 8:18 pm
Actually, I think that, in a way, the procompetitive aspects of a practice can potentially also be taken into account in the effects analysis, not only to examine whether the “by object” label is appropriate.
This will not be true of all procompetitive considerations, but it will be the case when a restraint does not restrict but enhance competiiton (which is different from a restriction that generates efficiencies or benefits, which you indeed consider under 101(3)). Say, for example, a non-compete clause in a franchising agreement, which altogether falls outside the scope of Art. 101, regardless of 101(3).
Also, doesn’t the ancillary restraints doctrine posit precisely that some restraints may be procompetitive overall and that this exempts them from the very scope of the rules? And what about Wouters? Or the General Court’s Judgment in Cartes Bancaires RENV, which did consider “lors de l’analyse des effets” the question of whether there would be more or less competition in the relevant market absent the restraint (paras. 123)
I guess one could construe a taxonomy of different “procompetitive” aspects. Perhaps different people fit different things in there.
Alfonso Lamadrid
15 May 2019 at 6:33 pm
Thanks, Alfonso!
I do not see things exactly that way, but we do not appear to disagree on the fundamental points.
In the context of ancillary restraints, the pro-competitive nature of the agreement plays a role. It plays the role of ruling out the by object qualification – Pronuptia is a wonderful example in this regard. The pro-competitive nature of the whole transaction is sufficient to conclude that the agreement, by its very nature, is not restrictive.
The above is in line with the points I made in the post and the comments.
You are absolutely right to say that ancillary restraints fall outside the scope of Article 101(1) TFEU altogether, insofar as they are not restrictive by object or effect.
Why is that? Because the analysis of the restrictive effects of a measure takes into acount ex ante considerations. The evaluation, in other words, considers the conditions that would have existed in the absence of the practice.
Absolutely crucial point (and relevant across the board). But saying that the restrictive effects consider both ex ante and ex post factors is not the same as saying that the pro-competitive dimension plays a role. It simply means taking seriously the lesson from Societe Technique Miniere, according to which competition means ‘competition that would have existed in the absence of the practice’.
This will be the topic of another post. Myth: ‘restrictive effects only take ex post considerations into account’. I look forward to your comments then!
Pablo
Pablo Ibanez Colomo
16 May 2019 at 11:32 am
Your myth is simply wrong.
It would suffice to read para. 18(2) of the Commission’s 2004 Guidelines on 85(3) to see what are the acceptable conditions under which an agreement cannot fall under 101(1) at all. The Commission applies this consistently, but the MasterCard amicus in France you cite was not such a case, as the MIF in question was not objectively necessary in that sense, as the ECJ has held.
More importantly, I think your myth tries to bring through the back door the (by now dead) rule of reason analysis in 101(1).
À disappointed reader
1 June 2019 at 3:11 pm
Thank you very much for taking the time to share your thoughts on the blog!
It would be wonderful if you could also share with us your views on the case law discussed in the post, and on how it can be reconciled with your position.
We certainly agree on something: the ‘rule of reason’ is a concept foreign to the EU system. There is no room for it under Article 101(1) TFEU. As I explain in the post, the appropriate forum to balance the pro- and anticompetitive effects of an agreement is Article 101(3) TFEU.
Pablo Ibanez Colomo
2 June 2019 at 10:54 am
Dear Pablo,
Many thanks for your always insightful and inspiring contributions! I was going through the comments and on one point have to agree with Joan, re Asnef-Equifax. Paras. 55 and 56 are in fact not relevant to/informative about the assessment of “by object” cases. The Court makes clear, in para. 48, that it considers the register not to amount to a “by object” restriction (“As registers such as that in issue in the main proceedings do not thus have, by their very nature, the object of restricting or distorting competition within the common market within the meaning of Article 81(1) EC…”). Whether it is a “by effect” restriction is, according to the Court, “for the national court to determine”. Then follows para. 49, which is the start of a series of considerations (including 55 and 56, and up to para. 63) which give guidance to the national court on the “by effect” assessment to be made, with the Courts own (interim) conclusion in para. 61 being: “an information exchange system such as the register is not, in principle, liable to have the effect of restricting competition within the meaning of Article 81(1) EC.” So this is really not about any by object assessment. All the best, Alvaro Pliego Selie
Alvaro Pliego Selie
14 June 2019 at 6:27 pm
Thanks, Alvaro, for the extensive comment!
The crucial point is that the pro-competitive aspects of the agreement are the reason the exchange in Asnef-Equifax is found not to restrct competition by object.
Paras 55-56 build on para 47. In para 47, the Court concludes that the exchange at stake in the case is capable of improving the conditions of competition, and as such not restrictive by object.
You are right that the Court goes on to provide guidance about the analysis of effects, but this fact does not deny the idea I express in the post, ie that the pro-competitive nature of the exchange rules out its qualification as a ‘by object’ infringement.
Pablo Ibanez Colomo
16 June 2019 at 11:36 am
[…] Persistent myths in competition law (I): ‘the pro-competitive aspects of an agreement can only be … EU competition law keeps busy a growing number of officials, practitioners and academics. It is fascinating that, with so many professional lives devoted to it, with so many specialised conferences and journals, there are still many myths that refuse to die… Pablo Ibanez Colomo (Chilling Competition) […]
Blog O Blogs May - Competition Policy International
2 September 2019 at 6:16 pm