Counterfactual analysis and restrictions by object: myths and misconceptions
Alfonso’s post on the counterfactual (the best we have published so far this year) made me think of a crucial question that, I believe, it is still very much misunderstood. Often, I hear people say that the counterfactual analysis ‘does not apply’ to restrictions by object, or that it is not relevant in that context. The question is important, as it is crucial to make sense of some major pending issues. I am in fact looking forward to presenting a paper on the counterfactual at the Oxford Antitrust Symposium later this year.
This (relatively widespread) view is, I believe, based on a fundamental misunderstanding about the notion of restriction by object, and about how to go about it in practice. Those who support this view seem to argue, in essence, that the conclusion of the analysis (ie the legal qualification of the agreement under Article 101(1) TFEU) comes before the analysis itself, and not vice versa. To put it graphically (and this is what explains the picture you see above), it is a bit like putting the cart before the horse.
More importantly, this view is contradicted by the case law of the Court of Justice, which makes it abundantly clear that the counterfactual should be considered irrespective of whether the analysis would lead to the conclusion that the agreement is restrictive by object or effect. In other words, what leads to the conclusion that the agreement is restrictive of competition is the evaluation of the counterfactual, and not the other way around.
It makes sense to illustrate the idea by reference to some concrete examples that show that the counterfactual analysis is a step that comes before the legal qualification of the agreement as restrictive by object or effect. Here they are:
- Export prohibitions and Erauw-Jacquery: Typically, an agreement that provides for an export prohibition is restrictive by object. This is not always the case, however. Where the analysis of the counterfactual suggests that the agreement would not have been concluded in the absence of the export prohibition, it is not restrictive of competition, let alone by object. Erauw-Jacquery, which builds on Nungesser, is crystal clear in this regard.
- Joint tendering and objective necessity: Cyril Ritter, an academically-minded Commission official, has recently uploaded a very interesting paper on joint tendering (which he – rightly in my view – sees as a form of joint selling). It is wholly uncontroversial to say that joint tendering can in certain circumstances be restrictive by object (more debatable is whether such agreements are typically by object infringements).
What matters for the purposes of this post, in any event, is an important point that Cyril makes in the paper. Where the agreement is objectively necessary for the parties to take part in the tender, it does not restrict competition, whether by object or effect. As he puts it, ‘there is no competition to restrict’ in the first place, as the parties would not have taken part in the tender absent co-operation between them.
- Price-fixing and collecting societies: The activities of collecting societies provide a wonderful example – if often forgotten – that a horizontal price-fixing agreement is not necessarily restrictive by object. Think about it: we are talking about an agreement whereby all the authors license their works jointly through a common platform. Is it a cartel? Certainly not. Price-fixing and all, this agreement falls outside the scope of Article 101(1) TFEU – and is thus not restrictive by object – where it is necessary for the collecting society to perform its function.
Competition under Article 101(1) TFEU means ‘competition that would otherwise have existed’
The Court clarified in Societe Technique Miniere that the word ‘competition’ in the context of Article 101(1) TFEU ‘must be understood within the actual context in which it would occur in the absence of the agreement in dispute’. In other words, ‘restriction of competition’ under Article 101(1) TFEU means ‘restriction of competition that would have existed in the absence of the agreement’ – as opposed to ‘restriction of competition that could exist in the abstract’.
There is nothing in Societe Technique Miniere, or in subsequent case law, that suggests that a difference must be made in this sense between restrictions by object and by effect. Such a distinction would in any event be artificial. The letter of the Treaty makes it clear that the notion of restriction of competition is a single one. And the Court has repeatedly held (AC Treuhand being a very good recent example) that no differences should be made where the Treaty itself makes no difference.
The object of an agreement cannot be understood without the counterfactual
If the analysis of the economic and legal context suggests that the agreement does not restrict competition that would have existed in its absence, it is very likely that its object is not anticompetitive. The rationale for the agreement (which is after all what the notion of ‘object’ is all about) is most probably a different one.
Allow me to come back to Cyril’s example: if the analysis of the counterfactual reveals that the parties would not have been able to submit tenders individually, the aim of the joint tender cannot be the restriction of competition. Most probably, such an agreement has a pro-competitive purpose, and it is certainly capable of making way for more effective competition (to come back to the expression used by the Court in Gottrup-Klim).
What explains the perpetuation of some myths and misconceptions?
Even though the case law is clear, the view that the analysis of the counterfactual does not apply to restrictions by object is still popular. Why? Because, I think, many people believe that the notion of restriction by object is something that it is not.
For many people, a restriction by object can be established in the abstract, that is, without examining the objective purpose behind the agreement and without considering the economic and legal context. According to this perspective, a price-fixing or market sharing agreement would always be by object infringements. This view is also misguided, and this is something that the Court has repeatedly held. It is true that, as pointed out in Toshiba, the analysis of the abovementioned factors may not be equally detailed in all cases, but this does not mean that the analysis can ever be established in the abstract.