Chillin'Competition

Yesterday I was invited to take part in a round table discussion by the UK’s Digital Competition Expert Panel. Philip Marsden – a beloved speaker at our Chillin’ events – was there in his capacity as member of the panel, together with Jason Furman and Derek McAuley.

Whenever I discuss digital issues, I tend to emphasise the importance of legal certainty and consistency, and yesterday was not an exception. These are arguably the most important factors in ensuring the best outcomes for competition and innovation.

And bearing their importance in mind is particularly crucial when anxiety – from stakeholders and from politicians – is on the rise. If policy-making is driven by anxiety, it tends to be more volatile – and thus less consistent and less predictable.

These were my musings when I read that the Bundeskartellamt has started an investigation into Amazon’s practices (see here). The press release suggests that there is a significant overlap with the European Commission’s investigation. The central concern appears to relate to Amazon’s dual role as a marketplace and as a competitor to the very merchants that use the marketplace – the famous ‘conflict of interest’.

The Bundeskartellamt’s press release is surprising. Not because it signals that common carrier antitrust is on the rise (remember this post?), but because it appears to be impossible to reconcile with the authority’s position in relation to other matters, in particular online marketplace bans.

Our readers are certainly familiar with the Bundeskartellamt’s position in relation to online marketplace bans. This position has even been made public by the authority in many ways, including in a paper published earlier this year. This paper advances a heterodox and narrow interpretation of the Coty judgment (on this, see here and here). According to the Bundeskartellamt, online marketplace bans are prima facie unlawful outside the specific factual confines of Coty (that is, luxury products).

The policy behind the Bundeskartellamt’s position seems clear: online retailers should have as many opportunities as possible to reach consumers, and these opportunities must include marketplaces.

The economic consequences are also clear: the authority’s strict stance against online marketplace bans amounts to subsidising online marketplaces, in particular Amazon and eBay. No wonder these two firms have been overtly championing the same stance.

With the new investigation comes the obvious question: why would a competition authority subsidise online marketplaces and simultaneously time take action against them?

I, for one, cannot make sense of the two together.

By subsidising online marketplaces, the Bundeskartellamt would be strengthening their position, thereby worsening the very concern raised in the investigation announced this week. One could accept that, potentially, one of the positions advanced by the authority could make sense if considered in isolation. But I do not see how it is possible to make sense of the two together.

I would say more. None of the two concerns raised by the Bundeskartellamt is warranted, and the fact that the two are examined together is evidence of it. If there is something that Coty and similar cases (including those investigated in Germany) reveal, it is that online marketplaces are not in any way indispensable to reach end-users (and certainly not indispensable within the meaning of Bronner/IMS Health).

It is so clear that online marketplaces are not indispensable that many manufacturers, in many industries, take active steps to ensure that their products are not sold through them.

And if online marketplaces are not indispensable, what is exactly the problem with Amazon’s conflict of interest? What would justify imposing common carrier obligations to a service that is not comparable to a telecommunications or an electricity network?

These are the questions that I hope will be discussed explicitly and extensively in the coming months.

In the meantime, I will try to make sense of the factors that may drive the Bundeskartellamt’s behaviour, which amounts to pursuing two mutually contradictory policies. For the time being, I can think of the following explanations:

• Missing the forest for the trees: the Bundeskartellamt’s (commendable) mantra has always been to keep markets open. If, however, the rationale behind certain policies is not acknowledged, and if the inherent trade-offs of intervention are not considered, there is a risk that inconsistent policies are pursued when action aimed at opening markets is taken. These policies may make sense when examined in isolation, but not when one looks at the broader picture.
• Online retailers favoured at everyone else’s expense?: There is a clear set of winners from the Bundeskartellamt’s policy stances: online retailers. These would benefit from guaranteed access to online marketplaces; in addition, online marketplaces would be regulated to their advantage.
  Perhaps the German authority’s implicit goal is to favour this category of operators (some of which will be SMEs). If so, this goal appears to go against the German competition law tradition. German competition law, as I understand it, has always (rightly) sought to protect competition as a process and has always (equally rightly) avoided micro-managing markets and picking winners. In this sense, formulating policy to favour a particular category of stakeholders seems to be against this tradition.
• Two problems are better than none: There is an advantage in pursuing mutually contradictory policies. It keeps an authority busy. As described by Wouter Wils in one of his many great pieces, the desire to show a high level of enforcement activity may sometimes explain the behaviour of officials (and an organisation at large).

We greatly enjoyed yesterday’s conference and wanted to thank you all once again for making it possible, whether you were able to make it there or not.

The feedback we have received so far has been excellent. You can see some of it on our Twitter profile.

The atmosphere was great and confirmed that there are lots of genuinely nice people in our area of practice. Perhaps the legal rankings have a point in having something positive to say about everyone in the business… 😉

If you were not there, you unfortunately missed the memorabilia which, to top last year’s mugs, included Chillin’Competition t-shirts and classical meme calendars. Btw, the image at the beginning of the post was taken by the Commissioner herself and shared on her Twitter account, so it’s ©Vestager.

We will soon post all pictures as well as videos of the brilliant panel discussions and presentations.

In the meantime we simply wanted to say: THANK YOU!

As is customary the day before our conference, we would like to express our gratitude to the conference sponsors. Their support is what enables us to offer this conference for free (and the food, and the drinks, and the gifts…)

A big thank you to all of them!

For several months now every conference and journal in our small world has featured a panel and a few articles on the Intel Judgment, on how to interpret it and on the impact it will have on enforcement and on judicial review going forward. But the most authoritative view has been written by the very Judge in charge of the case.

Judge da Cruz Vilaça was President of Chamber at the Court of Justice of the EU until a few weeks ago, when his term ended. He had also been the first President in the history of the General Court. He is one of the people that has most influenced the course of competition law in recent years, and also (now that he’s not at the Court we can say it) one of the nicest people in the world of EU Law (with a talent for surrounding himself with competent and nice people too). Focusing only on the Intel Judgment doesn’t do justice to his contribution to EU Law and to competition law. But doing justice to that would require much more than a blog post.

Back in August, Judge da Cruz Vilaça published a piece –that has only now come to our attention- titled “The intensity of judicial review in complex economic matters—recent competition law judgments of the Court of Justice of the EU”. It has been published in the Journal of Antitrust Enforcement.

This is a reflection on the role of the Court in competition cases, and in particular on how Intel fits within that case law and on how it is to be understood and interpreted. It does not represent the view of lawyers trying to push things in a given direction or of academics. It is an objective account by the most authoritative person possible. And it engages with academic commentary and is kind enough to cite our blog posts at different points..

We recommend you to read the full piece, but here are just some ideas that stand out:

Intel as an important yet limited step: “the two EU courts have progressively strengthened the conditions of their review and developed a more intense scrutiny in competition cases (…) [the Intel] judgment constitutes an important—albeit limited—step made by the Court in strengthening its judicial control as regards the application of Article 102 TFEU”.

Evolution in the case law: “The way the Court necessarily operates helps to explain why one may sometimes feel that the Court did not always follow an unambiguous line of reasoning in judgments concerning the same field of competition law. It may simply happen that time have not yet come for certain developments (…) But in a fast moving world, judges must also permanently listen to changes in the social, economic, technical, and political environment. How could it be different in the digital age, in a time of globalization and in a knowledge and information-based society?”

On the lessons to be learnt by the Commission: “the regulator would be well advised in the future not to expect any special indulgence for relying on a presumption of any kind of infringement per se of Article 102 TFEU when seeking to prove the anticompetitive character of such a system. Indeed, it would be odd to submit the European Commission to a stricter test when it carries out an analysis of all the circumstances than if it simply had relied on a per se infringement approach!”

Consumer welfare as a useful reference point: “not every exclusionary effect is necessarily detrimental to competition, as a result of which less efficient and less attractive competitors may be forced to leave the market. I believe such statements are useful for clarifying what the Court considers to be the major goal of EU competition rules, which the Commission has articulated as being ‘to protect competition on the market as a means of enhancing consumer welfare and of ensuring an efficient allocation of resources” As Professor Jones put it, ‘the case-law of the EU courts does not unambiguously endorse’ that view. Rather it sometimes also placed emphasis on the interests of competitors or individual undertakings. I do not think this a serious defect: although perhaps not always 100 per cent consistent, the case-law places in general sufficient emphasis on consumer welfare as a goal”.

Making life harder for rivals is not the same as foreclosure [you will have heard me say this a few times..]: “Of course, the objective (or the result) of the competition process is to make life harder for any operator on the market, forcing it to be more efficient—which in general does not happen with monopolist”.

It is all about correctly allocating the burden of proof: “the ECJ also made it clear that the balancing of favourable and unfavourable effects on competition of the practice in question can be carried out only after the analysis has shown the intrinsic capacity of that practice to foreclose competitors that are at least as efficient as the dominant undertaking. I think this is nothing more than the correct apportionment of the burden of proof between the Commission and the defendant. The specificity of a system of rebates that includes an exclusivity clause is that once the Commission has discharged its duty to establish the likelihood of anticompetitive conduct, the hurdles will be particularly high for the latter when trying to rebut the Commission’s finding”

(…)

“the dividing line between margin of appreciation and duty to review is sometimes very thin, which makes it more difficult to maintain the consistency of the case-law in all circumstances. Courts, however, have an arsenal of legal weapons available to them in this challenge. The most obvious are the rules on the burden of proof. A fine example of how to use them is given in paragraphs 66 and 67 of the judgment of the General Court in Intel”.

[By the way, that last bit makes exactly the same point as that made by General Court Vice-President in the abstract of an article to be published soon in JECLAP. He also says that, following Cartes Bancaires and Intel, the key to striking the right balance between the Commission’s margin of discretion and the Court’s in-depth review “can be found in the burden of proof that rests upon the Commission pursuant to Article 2 of Regulation 1/2003”. We couldn’t agree more]

***

The article contains much food for thought. And, incidentally, as frequent readers of the blog will have already noticed, it confirms the understanding of the case that we had been positing here and elsewhere for a while..

The excessively generous eulogy that Pablo wrote a few days ago, together with my inactivity on the blog over the past few months weeks, have led some to suspect of my demise.

The absurd number of extremely kind messages that I’ve received over the past few days even made me doubt it…  But I’m happy to confirm that reports in this regard have been greatly exaggerated:

After a few months locked in the office for different and very good reasons (piece of advice: don’t try to combine that with the additional sleep deprivation linked to having a 4 month old baby!), I’m now starting to see the light again  and hope to be back blogging regularly very soon.

In the meanwhile, and beyond work, we have made some conference outings to start regaining the habit:

• On 11 October I was in Helsinki for a great conference organized by Ilkka Leppihalme. Director General Laitenberger, Richard Whish, Jorge Padilla and John Rattlif were the other speakers. Richard and John used some slides to accompany their excellent presentations and kindly gave green light for us to post them on the blog (thanks!). Here they are:

Richard Whish- Recent Cases on Vertical Agreements and Excessive Pricing

Kilpailuoikeus 2018 – EC Competition Cases before the European Courts – John Ratliff

I won’t share my slides for now because I might use them again in the future and they’ll lose the surprise factor 😉 But here is a teaser.

• On 24 October I participated at another very good conference put together by Lexxion’s CoRe, the VUB and CCIA on “Competition Policy in Online Markets”. A video with the highlights of the event is available here. While we are not (yet) subject to must-carry remedies, here is a succinct recap of the event published in a “competing” blog;
• And today I spoke on the impact of the Intel Judgment at this LeadershIP conference in Brussels. There was much to discuss and little time, but my main message was one of optimism (and I think realism) and trust regarding EU Courts, who are often the target of what I think is (mostly) unwarranted criticism. Will leave the details for a separate post;
• My efforts to catch up with Pablo re conference appearances were doomed, as he has been not only taking care of the blog but also actively travelling around (see here).

Future events

Looking to the future now, there are several must-attend events coming up:

-First of all, we are very excited to be hosting the 4^th Chillin’Competition Conference next Tuesday (20 November). With the help of some overbooking we will try to release some additional seats for those in the waitlist. If you did not get a ticket (apologies), we will try to make sure you are well informed of what goes on. If you registered but cannot make it, please let us know so that we can get someone else in.

-The very day after, on 21 November, there will be an interesting debate on procedural issues in Luxembourg (“Due Process in Competition Law Enforcement: A Comparative Perspective”). More info is available here. I won’t be able to attend, but if any of you can and is willing to report on it, please drop us a line.

-On 26 November I will be speaking at the AEDC’s annual conference (that is the Spanish Association for the Defence of Competition) in Madrid. The programme is here: Programa Jornada Anual AEDC 2018.  The conference is sold out but will be streamed live.

-On 17 January Commissioner Vestager will be hosting her conference “Shaping Competition Policy in the Era of Digitalisation”. Registrations are now open and here is all the relevant info (again, stay tuned, because we hear additional speakers may be announced soon…)

There is not a single legal test in EU competition law.

Some practices are deemed prima facie lawful, and other practices prima facie unlawful irrespective of their effects. In between, there is conduct subject to a (standard) analysis of effects. As I explained recently on the blog, there is yet another category of practices; these are subject to an ‘enhanced effects’ analysis. By ‘enhanced’ I mean that it is necessary to provide, at the very least, evidence of indispensability before intervening.

I like to think of the applicable legal tests in EU competition law as discrete points along a spectrum.

This is what the spectrum looks like under Article 102 TFEU:

It makes sense to zoom into the right end of the spectrum, which is a bit crowded. It looks like this:

This is what the spectrum looks like under Article 101 TFEU (ATP stands for ‘absolute territorial protection’ in case you are wondering):

One question that keeps me busy (and which I discussed, inter alia, in Ithaca and Vilnius) is why and how the Court of Justice chooses the applicable legal test in individual cases. Why is some behaviour deemed prima facie lawful? Why is the enhanced effects analysis chosen for some practices but not for others?

The question is important, and the answer not immediately obvious. For instance, leveraging conduct is not subject to a single legal test under Article 102 TFEU. ‘Leveraging abuse’ is not a legal test. These words do not say anything about the conditions under which a given practice is deemed abusive.

In some instances (e.g. tying), leveraging is prima facie unlawful irrespective of its effects. In other instances (e.g. refusal to deal), it is subject to an enhanced effects analysis. What explains the disparate treatment of the various leveraging strategies?

I found the question interesting for other reasons. For instance, the Court has been criticised for applying different legal tests to refusals to deal, on the one hand, and ‘margin squeeze’ practices, on the other. The latter are subject to a standard effects analysis whereas the enhanced effects analysis applies to refusals to deal. Could it be that the Court has a point in spite of the criticisms? Spoiler alert: I believe it might.

***

Why are some practices prima facie unlawful under both Articles 101 and 102 TFEU?

The case law suggests that a practice is deemed prima facie unlawful irrespective of its effects – i.e. it is a by object infringement – where:

• It is presumed to serve no purpose other than the restriction of competition; and
• It is presumed to be capable of having restrictive effects on competition.

Intel (in the context of Article 102 TFEU) and Murphy (in the context of Article 101 TFEU) make it clear that firms can always provide evidence to rebut the presumption of capability.

Why are some practices prima facie lawful? I can identify two instances in this sense:

• Where the practice is objectively necessary to attain a pro-competitive aim (think of Metro I, Coty, Pronuptia and so on as clear examples); or
• Where the effects would not be attributable to the practice. For instance, the effects of above-cost price cuts are not attributable to the dominant firm, but to rivals’ inefficiency.

And finally: why is it that sometimes the Court of Justice has required an enhanced effects analysis? And why, for instance, is the enhanced effects analysis required in refusal to deal but not in ‘margin squeeze’ cases?

It all has to do, I believe, with the remedy. One needs to turn the analysis on its head and start by figuring out what intervention might entail in practice.

The standard effects analysis applies where the remedy is reactive in nature – a one-off obligation not to do something. These are instances in which a cease and desist order can effectively deal with the issue (no monitoring, no complexity with the implementation and/or with compliance).

The enhanced effects analysis applies where the remedy is proactive – that is, where it amounts, directly or indirectly, to imposing positive obligations on firms (a requirement to do something). These are instances in which the remedy comes with all sorts of potential difficulties (related either to the design, the implementation or the monitoring of the obligations).

The Microsoft I case (and, more precisely, the obligation to supply interoperability information on FRAND terms) is a good example of proactive enforcement. Determining the fair price of interoperability information is incredibly complex. So much so, in fact, that the Commission left it for the firm to figure out.

Against this background, one can understand why the Court may have deemed it justified to differentiate the legal treatment of refusals to deal and ‘margin squeeze’ practices.

A ‘margin squeeze’ can be effectively addressed through reactive intervention. Dealing with an abusive refusal to deal, on the other hand, is much more complex. One has to determine (directly or indirectly) a wholesale price, and then monitor compliance with the remedies on a lasting basis.

In cases like Commercial Solvents, Magill or IMS Health, the Court may have felt that the complexity of the remedy justified raising the substantive bar. The remedy, accordingly, is not an afterthought, but a key question that informs substantive analysis.

Ordering a company to resume supplies with a rival, to start licensing its intellectual property or to change the design of its products (as in the Internet Explorer case) can go wrong for several reasons – this is where I generally crack my joke about Richard Whish being probably the only person with a version of Windows without the Media Player.

If this is true (which it is), it makes sense to confine to exceptional circumstances the instances in which these remedies may apply.

I would say more: looking at the remedy to identify the appropriate legal test is a much more meaningful exercise than all other attempts to distinguish between categories of practices.

Instead of using more or less artificial labels and discussing whether one label is more appropriate than another, it makes sense to look at what remedial action may involve in practice, and infer the legal test from this exercise.

As ever, I look forward to your thoughts!