Double combo Florence-Brussels: Lundbeck and the new balance between competition law and IP
If Alfonso’s way of dealing with the unexpected – and potentially catastrophic –is to get his thoughts off his chest, mine is to focus, insofar as it is possible, on life as usual. And nothing says ‘life as usual’ more than another post – as if two were not enough – about Lundbeck.
The Brussels School of Competition organised a morning briefing on the case a couple of weeks ago. Together with David Hull and Luc Gyselen, we had a lively discussion on the case and its implications. Even better, the audience engaged with us and did not hesitate to challenge our views. Their slides, and mine, can be found here.
My presentation put Lundbeck in its broader context. This case, like some other recent ones, suggests that the balance between competition law and intellectual property is changing. In the past few years, the Commission has become less deferential to IP regimes.
How is the balance between competition law and IP changing? Back in 2004, the Commission was of the view that there is no potential competition when market entry requires the infringement of an IP right (see para 29 of the old Guidelines on technology transfer agreements).
Lundbeck shows that this view no longer reflects the approach of the Commission. Its decision in the case is based on the idea that potential competition may exist even when entry requires an infringement of an IP right. What is the logic of the new approach? Well, an IP right does not preclude entry if it is not exercised or if, when exercised, it is declared invalid.
This new logic explains Lundbeck. My guess is that it also explains pending cases like Pay-TV. The Pay TV case is unusual. What prevents Sky from offering online content outside the UK is not the agreement with the major studios, but the copyright system. Why would it be a competition law issue, then? Is it not a copyright problem instead? Well, one could argue – à la Lundbeck – that it is a competition law issue if copyright is never exercised against infringing acts.
Testing new approaches is what a competition authority should do. There is nothing wrong with it. If anything, it should be welcome. It would be disastrous if authorities did not seek to respond to emerging challenges. At the same time, new approaches need to be ultimately validated by the Court.
Alas, I am not convinced that the emerging new balance between competition law and IP will win the day. It seems to be at odds with the case law. I believe that paras 473-474 of Lundbeck capture the tension between the new approach and the case law particularly well (Luc Gyselen made a similar point during the event). These paragraphs read as follows:
‘473. The examination of a hypothetical counterfactual scenario — besides being impracticable since it requires the Commission to reconstruct the events that would have occurred in the absence of the agreements at issue, whereas the very purpose of those agreements was to delay the market entry of the generic undertakings […] — is more an examination of the effects of agreements at issue on the market than an objective examination of whether they are sufficiently harmful to competition […].
474. Accordingly, even if some generic undertakings would not have entered the market during the term of the agreements at issue, as a result of infringement actions brought by Lundbeck […], what matters is that those undertakings had real concrete possibilities of entering the market at the time the agreements at issue were concluded with Lundbeck, with the result that they exerted competitive pressure on the latter. […]’
Why am I of the opinion that these paragraphs are at odds with the case law?
The General Court appears to claim that the objective purpose – i.e. the object – of an agreement can be established without considering the counterfactual. I believe the case law is fairly clear in this regard, and it contradicts this view. It is only possible to figure out the objective purpose of an agreement by considering what would have happened in its absence.
- In fact, the Commission has already conceded that a restriction by object cannot be established without looking at the counterfactual. According to the Guidelines on vertical restraints, for instance, an agreement that restricts active and passive selling into a particular territory is not caught by Article 101(1) TFEU when the analysis of the counterfactual suggests that market entry would not have taken place in its absence.
- I also mentioned a venerable precedent, Remia, in my presentation. If you think about it, Remia is a case where the seller of an undertaking receives a payment to stay out of the relevant market. In spite of this fact, the Court held that the non-compete clause may fall outside the scope of Article 101(1) TFEU. Why? The Court understood that, in the absence of the non-compete obligation, the transaction may have never taken place.
- If the counterfactual shows that the agreement does not restrict competition that would otherwise have existed, one can safely presume that it serves a pro-competitive purpose. If the agreement is not capable of restricting competition, how can one claim that it has an anticompetitive object? This insight is apparent from a case like Micro Leader.
I also pointed out that paras 473-474 are in contradiction with other parts of the judgment. The GC examines the counterfactual at length in the judgment. The analysis of the counterfactual is after all indispensable to determine whether there are ‘real, concrete possibilities’ for generic producers to enter the market.
Thus, the GC examines the counterfactual and, at the same time, denies its relevance . Often, a contradiction of this kind suggests that there is something going on with the reasoning. What paras 473 and 474 reveal, first and foremost, is that there are two possible counterfactuals: one in which generic producers lack the ability to enter the market and one in which they are in a position to do so. In this sense, Lundbeck is different from recent cases like Hitachi and Toshiba.
This aspect of the case suggests, in my view, that the agreements are not only different, but more complex than a market sharing cartel. It also suggests that the rationale for the agreements considered in Lundbeck is not necessarily anticompetitive. The ‘by object’ label, as a result, does not seem appropriate (at least if one accepts the principle that the scope of the ‘by object’ label should be interpreted restrictively).
Chillin'Competition 
The reader can be referred to prior comments to prior long (and unconvincibg) posts by Pablo saying roughly the same. With all due respect, repetition of the post does not make an argument more convincing. You read too much in para 29 of the old guidelines. And the issue is often not if a patent is valid (as you always assume) but whether the generic producer can get around it (these are process patents). Why should there be a presumption of infringement, as you submit? It does not make any sense to me.
Spiderman
10 November 2016 at 11:39 pm
Thank you for your comment (and I appreciate that the change in tone). I believe the answer to your question is to be found somewhere above your prior comments (and, indeed, in the very paragraphs of the ruling that I cite above). Rest assured there will be no more posts on Lundbeck!
Pablo Ibanez Colomo
11 November 2016 at 12:14 am
I do not find it. I was referring to test you propose in your slides. Where the counterfactual is presumption of infringement of process patents. I thought you did not like presumptions.
Spiderman
11 November 2016 at 8:17 am
Unfortunately, none of your comments, including this last one, has much to do with the scope and point of my posts/presentation (nor do they relate to the issues of law raised by the case). I suspect it makes no sense to go over and over the same questions here. This said, I would be very happy to discuss a finished paper at length in an appropriate forum. Thanks again for your interest in the blog!
Pablo Ibanez Colomo
11 November 2016 at 10:01 am
I agree it makes limited sense. I would just have hoped that my question was answered since criticising a judgment must also mean proposing a good alternative.
The presumption of infringement that according to you the Court should have followed does not even exist in patent law. So it is unclear how such presumption may draw the right balance with patent laws. Although you often claim to be only interested in the consistency of the case law, your views on precedents are, to me, very selective (and, although you suggest that the judgment is “new”, it seems to me that it is your test which is really new, as it has no basis on prior case law, European or elsewhere). Again, prior cases where the ECJ examined allegations (as opposed to certainties) that competiton was not posible due to legal obstacles (which would render such potential competition unlawful or very unlikely) do not appear to be part of your examination. In cases where it is agreed not to enter a market, your counterfactual is based on certainty of entry, the authority having the burden of proving it (and i would guess you would propose a high standard of proof for that too) On the contrary, you mention cases which are very different, such as Remia.
Anyway, your point that one should be nice with this type of agreements is well understood. Big pharma will be happy to hear you (and you have bright future in participating in future events they organise or sponsor). Consumers (an unlikely sponsor of events) probably less so.
Spiderman
11 November 2016 at 10:16 am
You may not have encountered many of them, and that’s fine. But some people out there just care about the law and about the EU. If they are sometimes critical (and if they take the time to engage with wider audiences), it is precisely because they care.
You appear to suggest that anyone who disagrees with your interpretation of the law is dishonest, a fool, or is trying to please special interests. You are, alas, wrong on that one (and fortunately so).
Pablo Ibanez Colomo
11 November 2016 at 11:26 am
Thanks for the comments, Spiderman. You raise a very interesting issue very much worthy of discussion regarding the burden and standard of proof in relation to the assessment of the counter factual. There still seems to be uncertainty on this key point; we’ll try to address it in detail asap.
I won’t comment on this particular case, but I do need to step in regarding the comments on what may drive Pablo. One of the main reasons why I invited him to join me on the blog was precisely that he is genuinely concerned by the law and disinterested by much of what surrounds it (being in private practice I do care more about the surroundings). I couldn’t think of practically anyone so unconcerned about private or economic interests. Whereas we may sometimes disagree (and note that I have been criticised for behind allegedly hard on the pharma industry on the blog, the contrary of what he’s now accused of), Pablo not only takes the time to do this but always explains his ideas in a reasoned manner, politely (which can be compatible with being provocative) and openly, under his name, willing to discuss and gave criticism as this thread shows. One can certainly disagree (and encouraging constructive debate is what the blog is about), but one should also respect it. Thanks again for the substantive views, which are very pertinent.
Alfonso Lamadrid
11 November 2016 at 12:09 pm
To me as a reader, it is pretty obvious that Pablo’s interest in these questions is purely of an academic nature. The suspicions expressed by Spiderman are rather absurd and unfair.
On the merits: I’m not sure if the validity and the infringement of the process patent is really the crucial aspect when it comes to potential competition and its restriction. It seems that in Lundbeck, the generics already had obtained or taken steps to obtain marketing authorizations and/or concluded supply contracts in anticipation of their entry. Para. 124 refers to considerable investments that had already been made by the generics. This factual threat alone led to potential competition, regardless of Lundbecks prospect of successfully repeling that market entry by virtue of their patents.
The counter-factual is therefore the uncertainty about the outcome and the consequences of future patent litigation. Uncertainty is in itself a competitive restraint, and challenging patents while entering markets could be perceived as a part of competition too. Think of Uber: They entered many markets and certainly exerted a lot of competitive pressure, despite the fact that their business modell faces serious regulatory risks in many countrys. With this legal and economic context in mind, why should the concerted elimination of this uncertainty not be an infringement by object?
After all, if a competitor threatens to block the road to my plants, I am stuck with the remedies offered by law, but certainly cannot respond to the uncertainty of future proceedings with a market sharing agreement?
Spiderpig
11 November 2016 at 5:55 pm
I agree that the innuendos are unacceptable. Certain views may certainly be a better business card to attend the lecture/conference circus, but it is absurd to suggest that Pablo may be moved by such motives.
On the substance: the test of likelihood of entry is to me very narrow. In areas of “clouds” of patents (think about IT, where companies often accuse each other of patent infringements) you may end up considering that what appear to be competitors are not competitors (what are they then?). There would not be any problem with commitments to exit or not enter the market, then. I think there is more to say on how uncertainties about allegations of legal obstacles to competition, which would diminish or negate the possibility of competition, have influenced the analysis in prior case law. I can assure you there is some (in different settings) and going beyond Hitachi (just a few: Sandoz, where export prohibition was said not to be a restriction since such imports were not legally possible, what was dismissed by ECJ as irrelevant, since it would require to examine effects in the context of object; or seamless steel tubes, with identical outcome, or EDF/GDF where there was potential competition even if a period where it was not legally possible since on the facts there was evidence of market entry, probably illegal presence of the market, but the GC did not ignore it, or AstraZeneca on the existence of a bunch of legal obstacles that made future entry unlikely anyway, etc.). In Toshiba the very fact that there was a commitment not to enter the market was seen as evidence that such entry was possible (what is the point otherwise?). Contrary what your slide suggests, that GC did not examine a percentage of likelihood of entry. However, you appear not to grant any probative value to a commitment not to enter a market as evidence that such entry was possible. None at all. I am not suggesting that it is enough, but I do think it is relevant in the examination, and in a way shifts the burden to the companies to explain.
Pablo, I know you always insist that you are not concerned about outcomes (or even facts). But you are not Kelsen and the blog is not Die reine Rechtslehre. There is nothing wrong or impure in being concern about outcomes, let alone about facts. It is an essential part of decision-making. I do not mean deciding the outcome first and selecting the facts and reasoning afterwards, but having a look at the overall picture as well, and the general consequences of different solutions. And outcomes are mainly determined by facts which may explain parts of the reasoning too. I know examining facts take time from writing blogs and articles, but facts are terribly important. Otherwise, it is too easy. One may disagree with the case law and propose changes, but it is better to be transparent about it. I do not have the impression that there is a new balance, very simply this specific type of situation has not been examined before, since generics only became a more widespread reality in Europe at the time when Union pharma law was amended to facilitate generic entry, and it took some time first for originators to develop a strategy, and then for Commission to detect it, and to examine it. What appears to be new to me is your insistence in redefining the “by object” category.
joan
13 November 2016 at 9:48 am
Thanks to Joan and Spiderpig for the great comments:
We may not agree on the substantive analysis, but Joan’s comment identifies all the right questions. This looks, to me, like the appropriate way to frame the discussion.
As to the relationship between outcomes, facts and legal analysis: we do not disagree at all. This relationship – and in particular how facts (may) influence legal analysis and thus the evolution of the law over time – is what drives my interest as a researcher. I could not have put it better!
Pablo Ibanez Colomo
14 November 2016 at 10:30 pm
Dear Pablo,
Short question:
You say (a) there is a long-standing principle according to which EU law does not question of the existence of an intellectual property right, and (b) an agreement is not restrictive of competition by object if it remains within the substantive scope of the IPR. Yet, these points do not entail that the exercise of IPRs cannot raise antitrust concerns (as the exercise of any property right) under certain conditions. It also does not follow that pay-for-delay settlements cannot be found as restricting competition by effect (or even under a object inquiry). However, you seem to suggest the contrary.
To push it a bit further, your reasoning seems to echo the minority in Actavis (‘if the actions of the patentee fall within the scope of the patent they are not subject to antitrust scrutiny’) and favor an antitrust immunity for patent settlements. For you the patent is presumptively valid (so no restriction by object) and in any case the Court should not assess the validy of the patent (‘If the outcome of a case depends on the perceived strength of patent protection, competition law be second-guessing the patent system’). However, this ‘scope of the patent’ argument was rejected by both the GC and the SC. (The GC notes that ‘the fact that an agreement fell within the scope of a patent did not exempt it from an antitrust action’ para 492. Justice Breyer who delivered the majority opinion in Actavis contends ‘to refer, as the Circuit referred, simply to what the holder of a valid patent could do does not by itself answer the antitrust question). The patent here may or may not be valid, and may or may not be infringed’). What do you think about this?
stavros
25 March 2017 at 9:47 pm
Hi Stavros,
I am not suggesting anything of the kind. It is clear from the case law that such agreements may be restrictive by effect (this point was made expicit in Coditel II). This said, effects would have to be established against the relevant counterfactual — and it may turn out that there are no effects at all.
On the scope of the IPR: more than Actavis, it is relevant to discuss the relevant EU competition law case law. I am not the one saying that patents are presumptively valid: it is a point that is uncontroversial as a matter of EU law (see e.g. AstraZeneca). Same is true in relation to the second point you make: it is not me, but the Court of Justice, which has held that EU law should not assess the validity of national IPRs.
Against this background, what the US Supreme Court and/or the General Court have said is less relevant than what the Court of Justice has said in cases like Coditel II, Nungesser, or Erauw-Jacquery (in the latter, the scope of the patent test was even made explicit by the Commission in its submission).
It is in any event an exciting case, and it is good news that the Court of Justice will address the fundamental (and to some extent novel) points of law raised by the pay for delay saga. Thanks for your contribution!
Pablo Ibanez Colomo
27 March 2017 at 12:29 pm