Chillin'Competition

The meaning and scope of Article 101(1) TFEU is much clearer than it was just a decade ago (see here for a discussion). This relieving reality is in no small part a consequence of the steady stream of cases dealing with pay-for-delay practices in the pharma sector.

Rulings addressing pay-for-delay issues (and variations thereof) have provided the core test to establish whether a practice amounts to a restriction by object (since Generics, the issue is whether the only plausible explanation for it is the restriction of competition) and emphasise that, as far as Article 101 TFEU is concerned, anticompetitive effects will be assessed against the relevant counterfactual (a point on which I elaborated here).

More generally, the pay-for-delay saga illustrates that context is everything: as recently explained by Advocate General Emiliou in Tondela (here), a restriction of competition can never be established in the abstract. Whether it is a by object or by effect infringement, it is necessary to consider the circumstances surrounding the agreement.

One could have thought that these cases could not provide more clarity on the analytical framework to assess restrictions of competition. The Court judgment in Teva, however, has showed that the saga still has much potential to shed light on Article 101(1) TFEU.

Assessing the object of a practice in its context is not the same as evaluating its effects

The ruling in Teva is particularly helpful in that it engages with a line of reasoning that has become commonplace in recent years. According to some commentators, the contextual assessment of restrictions by object is problematic insofar as it blurs the line between object and effect.

I have never been persuaded by that criticism, which does not reflect the reality of the case law. Assessing the object (the rationale, that is) of a practice in the relevant economic and legal context is fundamentally different from establishing its actual or potential effects in the market (or markets) in which it is implemented.

It is certainly true that, once the anticompetitive object of an agreement is established, it is not necessary to ascertain its effects (something that the Court briefly reminds in Teva, paras 121-124, when dealing with the second plea raised by the appellants). It does not follow from this fact, however, that a contextual analysis is not required.

More generally, nowhere is it written that proving the object of an agreement should ignore the surrounding circumstances, or that it should be easy or straightforward to do so. The idea that the ‘object’ category is something of a ‘shortcut’ is foreign to EU law. It is borrowed from US antitrust law, which is fundamentally different from its European counterparts in more ways than one.

The appellants in Teva sought to advance a variation on this argument. They claimed, more precisely, that the General Court had conflated object and effect by, inter alia, introducing a counterfactual assessment, thereby making it impossible for the parties to the agreement to discharge their burden of proof.

The Court summarily dismissed these claims. When ascertaining the rationale, or object, of an agreement, it makes sense to consider the set of incentives it creates (paras 77 and 78). I would say more: how a practice influences the incentives of the parties will often provide a vital clue about what it seeks, objectively speaking, to achieve. That assessment does not entail an evaluation of the effects of the agreement in any way, contrary to what the parties claimed.

In the same vein, the Court concludes, the General Court did not require the parties to show what would have happened in the absent of the agreement. The relevant question remained that of whether the contentious transactions could be plausibly explained other than as a means to restrict competition (para 100). To the extent that it did, the General Court’s assessment remained firmly within the boundaries defined by Generics.

The unit of analysis is the clause, but context matters

The preceding case law had been clear in stating that the relevant unit of analysis is the clause, not the agreement as a whole (and this includes the crucial issue of severability). Examples abound. Arguably the canonical one is Pronuptia, where the Court distinguishes between the clauses that are ancillary to the operation of the franchise and those that are not (which, moreover, were found to restrict competition by their very nature).

The Court’s long-standing position is the only reasonable one. It would be too easy for the parties to avoid a finding of a ‘by object’ infringement if they were allowed to invoke the non-restrictive nature of the agreement taken as a whole. In effect, it would empty the ‘object’ category of its substance. For instance, a standard-setting agreement could be used as a vehicle to conceal a cartel agreement.

What Servier pointed out, and the Court mentions again in Teva (para 68) is that the analysis of the clauses cannot be undertaken in isolation. It may be the case that the clauses, taken together, are different aspects of a common strategy. If that is true, it is only reasonable that they are analysed together. By the same token, it is futile for the parties to try and argue that a given clause, examined in isolation, is not restrictive by its very nature. Such an interpretation would, again, entail a fundamental misunderstanding of the nature of the analysis under Article 101(1) TFEU.

At the end of the day, the Court’s clarification in Servier and Teva is another way of saying that context is everything at the ‘by object’ stage. The contextual analysis must by definition commence by looking at the agreement of which the clauses are a part. It is the first starting point, before shifting to factors such as the nature of the products or the dynamics of competition within the sector. The first step of the contextual assessment is likely to provide valuable hints about their nature.

Consider the example of price-fixing. There is a difference between a ‘naked’ price-fixing clause and one that is inserted in an agreement that pursues a non-restrictive aim. For instance, the Commission has long held (and rightly so, since the Court itself held just as much in Gottrup-Klim) that price-fixing clauses in the context of a joint purchasing agreement are not necessarily restrictive by object (see here). It has recently provided a good example that applies to SEP licensing.