Chillin'Competition

The slides on the Android investigation Alfonso uploaded a few weeks ago got me thinking. In many ways, this is a dream case for an academic. As in Macondo, everything is so recent that many things lack names. Fertile ground for categorisation, which is what I enjoy doing and, more important, what I happen to do for a living. On a related note, and since it has become de rigueur, allow me to disclose that I have absolutely nothing to disclose.

As I see it, the Android investigation is guided by two main themes: modularity and competition for bundles.

Modularity – I see in the case a value chain composed of three main layers: mobile handsets, operating systems and, on top of them, a set of core applications. Interestingly, not all firms follow the same business model. The different strategies relate to the different degrees of modularity each of the firms is willing to accommodate. By modularity I mean the extent to which the three layers are integrated with one another. At one extreme, there is Apple, which does not seem to allow for modularity (its handsets, and only its handsets, feature iOS, which comes with its own set of core applications). At the other extreme, there is Google, which allows for maximum modularity, at least in relative terms (Alfonso tells me that this point was not disputed at the conference). Android is offered to third-party manufacturers and made available – at the very least in theory – with or without Google’s core applications.

It is important to bear in mind this issue when trying to make sense of potential competition concerns. Because different degrees of modularity lead to different business models, one may fall into the trap of comparing apples (unintended) and oranges. In other words, one cannot simply cry ‘predatory pricing’ in relation to the distribution of Android to mobile handset manufacturers without taking into consideration that Google’s business model differs from Microsoft’s. It would be like claiming that Metro and the Evening Standard engage in predatory pricing simply because – unlike The Guardian or The Times – their papers are given away for free to tube users.

At the same time, the fact that Apple, Microsoft and Google allow for different degrees of modularity may make us lose sight of the fact that they are rivals and put competitive pressure on one another. ‘Thank you, Captain Obvious’, more than one reader must have thought, ‘otherwise there would be no complainants and no investigation on Google’s practices’. To which I answer: ‘Sure. But before we jump into any conclusions about bundling claims we need to understand how these firms compete with one another’. Which brings me to the second theme.

Competition for Bundles – Competition in the industry very much reminds me of rivalry in pay TV among cable, satellite and broadband operators. These operators try to attract subscribers by offering a combination of channels to viewers. It would be really awkward to claim that bundling in this context is anticompetitive. This is how competition is organised in the industry, and there is a compelling logic behind it. The fact that it would be awkward to bring such a claim does not mean that nobody has tried to do so. The argument was given a shot in California (there must be something in the water) and of course failed miserably before the US Court of Appeals for the Ninth Circuit.

Apple, Google and Microsoft also compete by offering a bundle of core applications. Several important conclusions follow from this fact, in my view. Dominance (that is, the extent to which Google is subject to effective competitive pressure) should be assessed at the level of the bundle, not at the level of individual applications. In other words, the alleged practices should be understood in a context in which players compete by offering a set of applications, not by offering them à la carte. Requiring Google to offer its applications à la carte ignoring industry dynamics would be as awkward as requiring cable operators to offer TV channels à la carte. Wondering if somebody has given this one a shot? Yes, and no other than Senator John McCain.

Pablo

Some of you will recall that a while ago, in this post, I announced that I’d be speaking about the Android investigation (which I did already, as explained here) and that I would welcome any views anyone could send my way. Geoffrey Manne then sent us the post he’d publish on this topic in Truth on the Market and asked whether it’d be possible for us to republish it for it to have a greater EU audience. We did re-publish it because we thought it was topical and provocative even if it could be seen as one sided, and because it saved us from thinking about what to write for one day.

The post seems to have had more impact than expected. The piece was first subject to criticism by Trevor Soames (Microsoft’s lawyer in the case) in the course of the panel in which I participated (this, by the way, led one of my new associates to comment that he had been surprised at how people actually read what is published on this blog…). And yesterday, ICOMP also published its own equally harsh rebuttal to Manne’s post, purporting to provide information on issues that “many of the ChillinCompetition readers may not be aware” of.

Against this background, in the light of ICOMPs previous legal actions, and in order to preempt any claim that we’re dominant on a competition law blogging market (which we’re not according to revenue-based shares, we can assure you) and should hence be subject to a blogging neutrality principle, we are also linking to ICOMP’s piece and giving it a level of visibility comparable to the one we gave to Manne’s post.

We only have three additional puns comments on this story:

• We are grateful to ICOMP for sending us increased traffic and take no issue with any scraping from content published in Chillin’Competition;

• We frankly have no interest whatsoever in the who-pays-who debate (unless anyone paid us, which very unfortunately isn’t the case) and only regret that academic writing on economically relevant issues these days is suspect (with reason or not) of being paid for. Since we are rather interested in the substance of things, my co-blogger Pablo –who should be free from all suspicion- has committed to writing about some legal issues relevant to the Android investigation before the end of the week (in what seems like a déjà vu, I agreed with this proposed commitment but I may change my mind if a substantial number of you disagree…) [By the way: I’m sure that post will have more readership than his most recent one on two intricate issues such as State aid and string theory J; forgive him, he’s new to this… 🙂 ]

• ICOMP’s piece states the following: “If Manne doesn’t have a basic understanding of European competition rules, why should he be featured on the ChillingCompetition blog?”.                                                  Well, just to clarify a possible misunderstanding: it should be apparent to any usual readers that having a basic knowledge of EU Competition rules was never really a requisite for writing on this blog; we’re the living example.

As promised, I am back with a few more comments on Eventech (more about it here). I am sure the State aid diehards will appreciate. One of the interesting aspects of the case is that there are questions both on the notion of State resources and on the notion of selectivity. As raised by the national court, these are not mutually interdependent, in the sense that the answer given to one of them does not determine the outcome of the other, and vice versa. I did not immediately realise the significance of this aspect of the reference. Looking back at some early landmark rulings, one sees that, typically, questions before the Court either relate to whether a given measure qualifies, in general, as State aid or they assume as undisputed that the contentious measure confers an advantage within the meaning of Article 107(1) TFEU.

What happens when the issue of selectivity is examined in isolation by a national court, as in Eventech? The impression one gets when reading AG Wahl’s opinion is that it may be a source of major confusion. From the national court’s perspective, the question of whether the regulatory advantage conferred upon London taxis is selective depends on whether it favours some firms over others that are in a comparable factual and legal situation. If this is the case, the national court assumes that it would be necessary to determine whether the differential treatment can be justified by reference to the objective of the regulatory regime in question before concluding that it is caught by Article 107(1) TFEU.

Why would it be necessary to identify firms that are in a comparable situation, let alone define the relevant market, to establish whether a measure is selective within the meaning of Article 107(1) TFEU? That is the question I asked myself when reading the opinion. Even though he diligently provides an answer to the national court, paragraph 50 of the opinion suggests that AG Wahl asked himself the question too.

It is clear from decades of case law that a measure can be selective irrespective of whether disadvantaged firms are in a comparable situation to those being favoured by the State. Of course it can. This is what the notion of selectivity is all about. Article 107(1) TFEU applies, inter alia, to measures that favour some sectors of the economy over others, or firms in one region over others. The fact that firms do not compete in the same market is immaterial in this respect. Using market definition as a tool to establish the ‘comparability’ of firms can only add to the misunderstanding. It may be the case that minicabs and London taxis only compete to a limited extent (for pre-bookings, as the opinion suggests). But how can this factor be relevant, if it only reflects the fact that London taxis benefit from even more regulatory advantages than those considered in the case?

It is remarkable that, after so many years, the notion of advantage (of which, as I understand Article 107(1) TFEU, and as the Court held in Altmark and many other cases since, selectivity is an element) can still be a source of major confusion. Which brings me to string theory (and impossibly complex topic that provides, alas, a suitable analogy here). The geekier readers will know that, in the early 1990s, there were five competing versions of string theory. In 1995, Ed Witten surprised everybody at a conference by suggesting that each of these five versions were in fact different ways of looking at a single, unifying theory (which was labelled M-theory).

This is how I feel about the notion of advantage in EU State aid law. The Commission and EU courts look at different dimensions of the notion depending on the needs of the case. Sometimes, the compensation for public service obligations is at stake and one issue acquires prominence. Sometimes, the relevant question is whether the measure is justified by the nature and the logic of the system and it is another aspect that becomes relevant in the analysis. But we do not seem to have a comprehensive and coherent framework to assess the issue systematically. It is therefore inevitable to end up with awkward situations like the one in Eventech, where the national court does not seem to dispute that the measure in question confers an advantage on some firms and at the same time raises questions about the selectivity of the same measure (!?). If only Ed Witten could take a break from Princeton and his equations and teach us some State aid.

Over the weekend I was able to catch up with some readings (notably on two-sided markets in anticipation of this conference), but also on recent case-law that I hadn’t yet had the chance to read. Thanks to this exercise I was able to become aware, among other things, of the content of an ECJ Judgment of 4 September 2014 on which I have read no comment whatsoever. This may be understandable because the case only deals with interesting legal issues, and not with high-stakes matters where the law is seemingly absent, which are lately the only ones grabbing commentators’ attention…

The Judgment at issue – Anonima Petroli Italiana (“API”)– is a preliminary ruling responding to questions posed by an Italian Court in relation to an Italian law pursuant to which the price of road haulage services for hire and reward cannot be lower than minimum operating costs, which are in turn fixed by a body composed mainly of representatives of the economic operators concerned.

The Italian Court asked, in essence, whether any such legislation was compatible with Article 101 read in conjunction with Article 4(3) TEU as well as with the Treaty provisions on free movement of services.

As some of you may recall, the possibility of applying Article 101 to State behavior  pursuant to its joint application with other Treaty rules was born in Inno Attab in 1977 [btw, I just found this little jewel commenting on the earlier case law on the subject]. The reasoning used back then by the Court was that Art. 4(3) TEU (at the time 10 TCE) prohibited Member States from depriving Treaty rules of their effet utile; given that former Article 3.g) (deleted from the Lisbon Treaty at the behest of Mr. Sarkozy; remember?) established undistorted competition as one of the goals of the EU, it was held that Member States could not adopt measures depriving competition rules of their effect utile. This doctrine was considered potentially huge at the time, but never lived up to its promise due in part to the restrictive interpretation endorsed by the Court in the November Revolution of 1993 in the Reiff, Ohra and Meng cases), according to which a State measure could not by itself run counter the Treaty rules in the absence of a certain behavior on the part of the undertakings (unlike, by the way, what happens with Art. 106 as recently re-stated in the Greek Lignite case).

The ECJ’s recent Judgment concludes that by delegating the power to fix minimum tariffs on a committee composed of a majority of representatives of the economic operators who are not bound to observe public interest criteria in their (non-reviewable) decisions, the legislation at issue runs counter the effet utile of Article 101 by preventing undertakings from setting lower tariffs (the Court doesn’t however clarify whether the restriction is “by object” or “by effect”).

My 3 comments:

• The Judgment shows that this doctrine is well alive, even if it isn’t kicking, and this regardless of the elimination of former Article 3.g). A lot could be done with this doctrine if competition authorities took it seriously. But instead of using the well-developed tools at their disposal (like this one or like Art. 106), competition authorities are busy stretching the interpretation of others (like the one of selectivity in State aid, as seen in the recent openings of proceedings in relation to tax rulings).
• Regarding possible justifications, that the ECJ seems to apply the “objective justification” test developed in Wouters, Meca Medina, etc. very naturally and not as anything exceptional, very particularly when it deals with conduct adopted by regulatory or quasi regulatory authorities (albeit not only in those cases, as shown by Pierre Fabre). Some of you may legitimately observe that this fits oddly with the Judgment in Irish Beef, where the ECJ held that “[i]t is only in connection with Article [101(3)] that [other legitimate interests] may, if appropriate, be taken into consideration for the purposes of obtaining an exemption from the prohibition laid down in Article[101(1)]”.
• Quite strikingly, the ECJ does not include a single mention to Article 101(3) (and this despite the fact that the questions referred to it cited Article 101 in its entirety and not to 101(1) alone). It basically states that since the measure falls within 101(1) TFEU then Article 101 prohibits it. In my view, under normal circumstances, and in proper application of the Court’s case law, the Court should have said that it was up to national Courts to assess whether the conduct at issue could benefit from Article 101(3), the assessment of which is mandatory. This error bypassing of Article 101(3), not only at the practical but also at the theoretical level (and on the part of the only institution that took it seriously in the wake of Regulation 1), further confirms the point I made some time ago (and have subsequently cited many times) about the slow death of this provision.

Advocacy is one of the key missions of competition authorities. The benefits of competition are not always well understood, or are not necessarily given the same prominence as other interests in policy debates. What is not emphasised enough is that these same authorities have an equally important role to play when it comes to dismissing unsubstantiated concerns. There are some issues that – sometimes due to effective lobbying – give rise to a great deal of controversy but that turn out to be wholly unproblematic upon closer scrutiny. It would be desirable if, when dealing with these concerns, authorities were explicit about the reasons why they conclude that action is unwarranted and contributed to public discussions by explaining why what looks like a blatantly anticompetitive strategy is sometimes a subtler manifestation of competition on the merits. This is what I call negative competition advocacy.

Recent developments provide a couple of emblematic examples of the areas where negative advocacy is most needed. One is about net neutrality, arguably the catchiest slogan of the decade (which, moreover, was coined and popularised by a most charismatic academic). I have not yet come across a theory that provides a convincing case for net neutrality (which, in essence, means that all Internet services should be treated equally). Unfortunately, sector-specific regulation is not always adopted for the right reasons. What seems clear, on the other hand, is that the concerns underlying net neutrality (in particular, that telecommunications operators block or discriminate against competing services) are, by and large, a non-issue from a competition law perspective. Therefore, I am not surprised that the Commission closed, earlier this month, an investigation into ‘Internet connectivity services’ (which Commissioner Almunia linked to the net neutrality debate in his speeches). After the unannounced inspections carried out in July 2013, the authority has come to the conclusion that ‘the observed practices do not appear to breach EU antitrust law’. I only regret that the press release does not explain more about the reasons behind the decision.

Another issue that makes the headlines every now and then is the use of private labels by supermarket chains (as net neutrality, it is a discrimination-related concern). For some reason, the idea that grocery stores sell products under their own brand and give these brands more favourable treatment is perceived to be deeply problematic (this is at least the impression one gets when reading newspapers). Authorities should explain clearly that there is nothing wrong, per se, in such practices. The speech delivered by the Director General for Competition on 2 October is a very good example of the way forward. Alexander Italianer presented the findings of a year-long study on retail competition in the grocery sector. The study came with good news. Contrary to the widespread view, it would seem that competition is working fine on the whole in these markets. The Director General unambiguously stated that private labels ‘do not appear to hamper choice and innovation’, at least so in ‘moderately concentrated retail markets’. I hope these words (and, more importantly, the hard evidence backing them) will be taken into consideration at the national level, where regulatory instincts (and the lobbying that comes with them) are sometimes very strong.

More examples where negative advocacy could be useful? Think of Apple’s recent decision to stop selling Bose products from its stores. It looks like a refusal to deal. Even worse, it is a disruption that follows Apple’s acquisition of Beats, one of Bose’s competitors. Is this problematic from a competition law standpoint? Not really. I quote Alexander Italianer: ‘practices such as these are only a matter of competition law, if they have an impact on the overall functioning of the market. For instance, if a retailer twists the arm of a supplier in individual bilateral negotiations, then he may well be in the wrong, but it falls beyond the scope of competition enforcement’.

Pablo

Congratulations to Jean Tirole on the Nobel Prize! I cannot imagine anybody being unhappy or disappointed about this year’s pick. Even though he is just 61, I am ready to say the prize comes a little late. I confess I have been waiting for this moment since at least 2009, when it went to Elinor Ostrom and Oliver Williamson. The upside of waiting a few more years is that the award comes solo this time, comme il le fallait.

He is one of these economists who does not need an introduction. He is the author of The Book on Industrial Organisation (le Tirole) and has made fundamental contributions in the field. For instance, his 2003 landmark paper (with Jean-Charles Rochet) on two-sided markets has proved to be enormously influential, as Groupement des Cartes Bancaires and MasterCard show. Competition in Telecommunications, written jointly with the late Jean-Jacques Laffont, is one of my favourite books, and is a good introduction to his work on regulation. I guess I could keep going for a while (for instance, I remember discussing his work on open source software with a colleague the other day).

The best way to start the first post I write after my ‘formal’ induction into the blog is, I tell myself, by referring to one I wrote back in July, and which concerned the preliminary references in Eventech and Kunsten Informatie and Media. The two AG opinions came out during the Court’s hyperproductive month of September. While I still think the two cases are relatively straightforward in terms of substance, it makes sense to write a couple of notes on Eventech (in the form of two separate posts).

You may remember that the key question in that case was whether regulation allowing black cabs (the iconic London taxis) to use bus lanes and excluding the so-called minicabs (private hire vehicles) entails the use of State resources. The answer, in my view, is an obvious no. It is in fact a good textbook example showing that advantages within the meaning of Article 107(1) TFEU are not always granted through the use of State resources. The position taken by AG Wahl therefore comes as no surprise to me. It is rather the orderly analysis of the question that inspired me to write this post.

When going through it, I thought, first, of AG Jacobs’ classic opinion in PreussenElektra. It seems difficult to believe, but many of the issues that now seem straightforward were far from established when that opinion was written back in year 2000 (which, or so I like to think, is not such a long time ago). If a masterful analysis was necessary to put some order into a set of scattered and seemingly contradictory precedents, this must mean that State aid has come of age only relatively recently.

Second, I thought that I had never seen such a careful dissection of the different ways in which State resources may be involved in a case. Eventech had raised several arguments trying to link the regulatory advantage to the use of resources by the State. For instance, the firm argued that black cabs can be said to be exempted from fines for using the bus lanes, which, in turn, leads to a loss of revenue for the State. The opinion is valuable not because this argument, or similar ones, were particularly powerful or persuasive, but due to the way in which they are examined by AG Wahl. To this day, I still have misgivings about the treatment of some issues in landmark cases like Sloman Neptun. From this perspective, the opinion sets a model for the future, in particular to the extent that the Court had seemingly never been confronted with some factual scenarios that were relevant in this case.

I will write something in the coming days about the advantage aspect of Eventech. Before that, I will end on a personal note. I took an Uber cab this summer for a fairly long ride and one of the many questions with which I pestered the driver had to do with this case. Interestingly, he was of the opinion that the regulatory regime in place makes sense. He said that traffic would be unmanageable in London if minicabs were allowed to use bus lanes. And he thought he would rather be subject to lighter regulatory obligations, even if it means giving up some privileges. Thoughtful and sensible, I would say.

Pablo

To all readers,

After a long period of silence – part of it was Commission-imposed – the shareholders of chillincompetition have decided to transform this blog.

From today onwards, the blog moves from joint to sole control. In plain words, I quit, and leave the blog in Alfonso’s hands.

We discussed this together, but the reasons for exit are compelling.

Paraphrasinging BB King: “The Thrill is gone, it’s gone away for good“.

Alfonso’s challenge is now to find a new “bad cop“, to keep a sane degree of political incorrectness on this platform.

Thanks to all for the support in the past years.

And to my good pal Alfonso: “And now that it’s all over, All I can do is wish you well“