Chillin'Competition

People sometimes ask me how much time I devote to the blog. And I almost invariably answer by saying that the correct question is how much I benefit from the blog.

I have just uploaded a draft paper on ssrn (see here). This is a piece I would never have written had it not been for the blog. It has been inspired by the many discussions on pay-for-delay we have had in the past few years.

The comments I received were invariably thoughtful, and were sometimes outright passionate. I learnt a great deal from them. Something I learnt was that, for better or worse, the message I wanted to convey did not always come across as clearly as I would have wanted it to. The paper just poured out of me in an attempt to bring clarity to the issues.

These comments also gave me the – right or wrong – impression that I had something to contribute to the debate. There is a rich line of case law addressing the relationship between Article 101(1) TFEU and intellectual property that had gone virtually unnoticed.

Of these cases, two of them – Nungesser and BAT (Toltecs-Dorcet) – provide, in my view, the structure to analyse the lawfulness of pay-for-delay agreements. Other crucial cases include Coditel II, Ottung and Micro Leader.

One of the points I wanted to clarify is that I am not interested in the outcome of individual cases. This is something I have often emphasised on the blog. In spite of this, many commentators assumed that I was challenging the outcome in Lundbeck.

That was never my intention. I will say more: it is very difficult for a moderately attentive reader of the Lundbeck decision to avoid the impression that the agreements at stake in the case were restrictive by object and deserved a fine. But again: I am interested in principles, not outcomes.

Another point that I felt needed clarification: some commentators thought I endorsed Roberts’ dissent in Actavis. Nothing further from the truth. The more I think about theat dissent, the more I am persuaded that it is intellectually indefensible.

Stephen Breyer’s majority opinion, which rejected the prima facie unlawfulness of genuine settlements and proposed a case-by-case approach, comes across as more reasonable (more on this below).

As rightly observed by one of our commentators, John Roberts’ dissent seems to be based on the idea that, since patents are presumed valid, we can presume that, in the context of a dispute, they have been infringed. But there is no such thing as a presumption of infringement.

Main points raised in the paper

On the notion of (potential) competition and restriction by object

So how does my paper go about the analysis of pay-for-delay agreements? I thought the right approach was to go step by step: I start from the fundamentals and then progressively move to the specifics of Lundbeck and Paroxetine.

Thus, I discuss the fundamental notions at stake in these cases (competition, potential competition and restriction by object); I move on to the discussion, in general, of the relationship between Article 101 TFEU and intellectual property to finally explain the controversy behind Lundbeck and Paroxetine.

On the fundamentals, I believe the case law is consistent and unequivocal since the 1960s. The Court defined the notion of competition early on, in Societe Technique Miniere.

It is clear from this case and subsequent ones that the notion of competition under the Treaty should be understood as meaning ‘actual or potential lawful competition which would have existed absent the practice’.

One of the implications is that the evaluation of the counterfactual is of course relevant when assessing restrictions of competition (whether by object or by effect). I recently addressed this question on the blog (see here).

Another implication is that the evaluation of the object and/or effect of a practice is a case-specific inquiry. Restrictions by object do not exist in the abstract, and can never be established without considering the relevant economic and legal context (both AG Kokott – see here – and AG Bobek – in his Opinion discussed here – have recently emphasised the importance of context).

On the relationship between Article 101(1) TFEU and intellectual property

There is a wealth of case law addressing the relationship between Article 101 TFEU and intellectual property.

Nungesser provides the principle. The Court distinguished between two scenarios: (i) one in which the scope of the agreement does not go beyond the range of acts that the right holder would be able to authorise or prohibit by virtue of its intellectual property and (ii) one in which the agreement allows the right holder to obtain protection that it would not have been able to obtain by enforcing its intellectual property.

Agreements falling under the first (i) scenario are not necessarily restrictive of competition, and may sometimes (e.g. Coditel II, Micro Leader) fall outside the scope of Article 101(1) TFEU altogether. Agreements falling under the second (ii) scenario (e.g. Ottung) may restrict competition, and are often restrictive by object.

The principle set out in Nungesser was applied to intellectual property settlements in BAT (Toltecs-Dorcet). The Court, uncontroversially, concluded that the settlement at stake in the case was not a genuine one, and thus pertained to the second (ii) scenario.

Indeed, BAT had sought to obtain via the settlement what it would not have been able to obtain via the enforcement of its (dormant) trade mark. The object of the agreement was the restriction of competition.

On the application of the principle to pay-for-delay agreements

The framework described above fits pay-for-delay cases particularly well. One can distinguish between settlements that address a genuine patent dispute and those that do not. The object of the former would not be the restriction of competition; the latter are likely to infringe Article 101(1) TFEU by its very nature.

Lundbeck shows how this framework would work in practice. The originator itself admitted during the proceedings that its process patents did not block all possibilities of entry. Accordingly, it sought to obtain by means of the agreements a degree of protection that its patents would not have afforded. The conclusion that the agreements in Lundbeck are restrictive by object is in principle inevitable as a result.

Why the controversy, then? Again, it is not about the outcome. The reasoning of the Commission in Lundbeck, and of the CMA in Paroxetine, suggests that a settlement addressing a genuine dispute would also be restrictive by object.

It follows from this reasoning that there would be a prima facie infringement of Article 101(1) TFEU even when it has not been shown, to the requisite legal standard, that generic producers need to rely upon a patented process to enter the market.

As I have suggested in the past, the approach in Lundbeck and Paroxetine heralds a new relationship between competition law and intellectual property (these are not the only cases, hints at this new relationship can be found elsewhere).

This new approach is not obvious to square with the principle whereby EU law does not question the existence of the rights, but only their exercise. It is also difficult to square with the principle whereby registered titles are presumed valid.

Exciting times ahead! I very much look forward to your comments.

As ever, I am delighted to clarify, in accordance with the ASCOLA declaration of ethics, that I have nothing to disclose.