A comment on Case T-79/12 Cisco Systems and Messagenet v European Commission (Microsoft/Skype)
On 15 February 2012, Cisco Systems and Messagenet appealed the Commission’s decision authorizing the purchase of Skype by Microsoft. On 11 December 2013, the General Court rendered its Judgment dismissing the application for annulment.
As many readers of this blog will know, I was one of the lawyers representing the applicants, and was personally very involved in the judicial phase of the case, which I very much enjoyed. For the past 5 months I’ve read some succinct comments about and I think that there are many genuinely interesting things about it that might so far have been overlooked.
Whereas I –biased as I am- have issues with most of what’s in the Judgment (and particularly with what isn’t there), I’ve decided to try to get rid of any bitterness (some irony will be inevitable, I’m afraid) and approach it in a hopefully constructive way, leaving a myriad factual case-specific issues aside, and focusing only on selected matters of general relevance to any competition lawyer.
So instead of re-arguing the case –which would be of little use at this time- my intention is to shed light on some aspects of the Judgment which otherwise not attract the attention they deserve. I’ll touch on 6 selected issues, and will offer some personal views as a conclusion.
Needless to say, my opinions are, aside from non-objective, exclusively attributable to myself, not to anyone else, notably clients and colleagues, and neither Cisco nor Messagenet have anything to do with this post.
1) The Court ruled that the standards of proof and review applicable to Phase I (Art.6) decisions are identical to those applicable to Phase II (Art. 8) decisions
Whereas we argued that the merger should be annulled regardless of how the Court interpreted the applicable standards of proof and review, we also claimed that the standard of proof must necessarily be higher in the case of Phase I decisions because the Commission has to prove that the case couldn’t objectively give rise to “serious doubts” (which is the applicable legal test according to Art 6 of Regulation 139).
This interpretation, now held wrong, was fairly uncontroverted in academia (see e.g. the contributions EUI’s 2009 workshop on standard of proof in competition law), and had been formulated previous cases. In her Opinion in Impala AG Kokott went even further and explained that a “beyond reasonable doubt” standard applied to Phase I decisions “to compensate for the fact that at that stage the investigation of a concentration is merely a summary one” (…) “[a]t that stage, serious doubts as to the compatibility of a concentration with the common market will only prevented its being cleared to quickly and force the Commission to make a more extensive investigation in a formal procedure”. A test of absence of doubts also governs the initiation of in-depth reviews in the State aid domain, and the Court has established in that context that this test requires a review that “will, by nature, go beyond simple consideration of whether or not there has been a manifest error of assessment” on the Commission’s part (for more on this see, e.g. cases T-73/98, para 47 and T-119/02, para 77).
The Judgment in this case nonetheless states that “the standard of proof is no higher for decisions adopted under Article 6 of Regulation 139/2004 than those adopted under Article 8 of that regulation” (para 46). The Court then goes on to explain that even if we correctly argued that the Commission has no discretion as regards the initiation of Phase II whenever it has serious doubts, the Institution “enjoys a certain margin of discretion” to carry out the “complex economic assessments” required in merger cases (para. 49), and that therefore the standard of review for both Phase I and Phase II is the same: that applied to complex economic assessments (limited judicial review).
What the Court is effectively saying in paras 46 to 49 is that even if the notion of serious doubts is an objective one, the Commission has discretion to have doubts or not. In my mind, this would mean that the alleged objectivity of the concept is meaningless, but perhaps there’s a different reading, which I don’t yet grasp. Even if the standard of review is the same for Phase I and Phase II decisions, it seemed intuitive to me that what has to be proved in one case (no serious doubts) and the other (compatibility or incompatibility with the internal market) is different. By rejecting this previously uncontroversial interpretation I think the Court has importantly -rightly or wrongly- expanded the Commission’s margin of discretion in merger cases.
2) Unless I’m missing something in para. 67 the Court explains that competitive assessments in most Phase I decisions are not to be taken seriously because they do not assess the “real” relevant market.
“The applicants therefore base their complaint relating to market power held by the new entity on an incorrect assumption, in so far as the Commission did not define the existence of a specific market for consumer video communications on Windows based PCs. The Commission did not therefore establish in the contested decision that operators present on the narrow market could act independently of the competitive pressure from other means of consumer communications, such as services offered on other platforms or other operating systems. In addition, the applicants did not themselves submit any evidence or study to support the conclusion of the existence of such a narrow market. By contrast, they merely criticised the factors put forward in the contested decision in order to qualify the significance of market shares”.
What this paragraph says isthat the fact that the Commission chose to assess the market for video communications on Windows based PCs was irrelevant, and that we could only have challenged this assessment if we proved that the market was the real one (!). This is quite astonishing may perhaps be a bit surprising to some, because what we were challenging was precisely the conclusion that “the proposed transaction does not give rise to any competition concerns even on the narrowest possible definition of the relevant product market”. The market might have been hypothetical, but its assessment was the only one contained in the decision and therefore the only one that could be appealed.
Unless I’m wrong (again, let me know if you see it differently) what this means that from now onwards any party wishing to appeal a Phase I merger decision should not challenge the assessment actually carried out by the Commission, but will need to prove that the assessment of the “narrowest possible market” corresponds to a real market, which will almost never be the case! In other words, from now onwards the Commission could get immunity from Court review by carrying out assessments of markets whose definition is left open.
3) On the irrelevance of market shares in dynamic markets
The few paragraphs that have so far received public attention are the ones concerning the irrelevance of high market shares. In para 69 the Judgment states that “the consumer communications sector is a recent and fast‑growing sector which is characterised by short innovation cycles in which large market shares may turn out to be ephemeral. In such a dynamic context, high market shares are not necessarily indicative of market power”.
In fact, I agree with this statement. Market shares in these markets are “not necessarily indicative of market power”; they provide an indication which may be disproved by other factors. My problem with this is they do provide an indication, and even if it can be disproved by looking at countervailing factors, I still struggle to see those here.
In any event, there are a few paras in this section (mainly paras 79 to 84) that that are potentially quite troublesome for enforcement, particularly in technology and communication markets. No wonder these will from now onwards be cited by any company with large market shares.
4) On the irrelevance of network effects in a non-interoperable communications market
Paragraph 76 also marks –in my view- a change in the way network effects are assessed in EU competition law by stating that “the existence of network effects does not necessarily procure a competitive advantage for the new entity”.
This may seem at odds with all past Commission precedents, mainstream economics, regulation of other communication markets, the Commission’s soft law on market definition, 102 and mergers, as well as with Skype’s own repeated statements in official public submissions claiming that “the scale, global distribution and growth of our user base provide us with powerful network effects, whereby Skype becomes more valuable as more people use it, thereby creating an incentive for existing users to encourage new users to join. We believe that these network effects help us attract new users and provide significant competitive advantages”.
You may recall that the Decision’s argument to rebut the role of network effects was that users “make the majority of their voice and video calls to the small number of family and friends that make up their so called “inner cicle” (4-6 people) and that “it is not difficult for these groups to move between communication services”. This peculiar argument was endorsed by the Court. As I’ve repeatedly said over the past two years, I may well Skype the most with my wife, girlfriend (J), mother and best friend, but I would assume that my best friend has in turn a different mother, girlfriend and wife (or so I’d like to think…); in other words, groups of people are interconnected and do not communicate in movable autarkic nodules. On this point, the Judgment simply repeats (thereby endorsing) the Commission’s argument at the end of para 52 (“the network effects to which the concentration might give rise would be diluted by the fact that users tend to communicate in small restricted circles and use a range of operators. Those factors demonstrate the ease with which user groups switch to other communications services”). [On multi-homing, note that the “range of operators” meant the two merging parties –otherwise they couldn’t have a 90% market share- as openly acknowledged in footnote 52 of the decision].
4) On the identification of competitive constraints.
A paragraph that could also prove important for various markets where companies rely on others’ technology (and for private label products) is para. 72, which dismisses the claim that Facebook (the second largest player with an overwhelming 10% of the market, whose video call service runs on Skype, which has Microsoft as a shareholder and which interoperates with Skype) would not be an effective competitor with this reasoning.
“The only factor that they put forward in support of that argument is that Facebook is a licensee and strategic ally of Skype, which cannot use Skype’s software to offer services in competition with the paid services of Skype, called SkypeOut, which make it possible to, inter alia, call fixed or mobile telephone numbers and to conduct video calls involving more than two persons. However, they do not submit that that agreement prevents Facebook from offering its video communications services to consumers who might decide to switch away from the new entity if it decided to exert any market power.
So, being a “strategic ally”, using the same technology and the existence of a non-compete agreement do not indicate mitigated competitive vigor. Note taken.
5) On switching, statement of reasons and the comparison with the Microsoft (and Google) abuse cases
[Click here to continue reading]
Paras 73 and 79-80 of the Judgment validate the Commission’s point about downloading being able to offset network effects by pointing out that “contrary to the situations that formed the basis of the Commission’s previous decisions which are relied upon by the applicants, and as stated by the intervener, there are no technical or economic constraints which prevent users from downloading several communications applications on their operating device, especially as the software concerned is free, easy to download and takes up little space on their hard drives”.
This is important because in para 96 the Court rejects the claim that the Commission deviated from its own precedents (previous Microsoft cases on downloading and multi-homing). The Court states only that “unlike the earlier decisions, the case in point is not characterized by the presence of technical or economic constraints preventing users from downloading several” applications at the same time, and that “there can therefore be no question of any change in policy for which the Commission ought to have stated reasons in the contested decision”.
However, as most of you may recall, the absence of technical or economic constraints was precisely the argument used by Microsoft to defend itself in prior cases (both in 2004 and particularly in 2009). Both the Commission and the Court based their case on the idea that, technical and economic constraints aside, network effects and user’s inertia also play a very important role. Don’t take my word; instead read, e.g. paras 47 and 48 of the 2009 Microsoft commitment decision.
By forgetting about these, the Court rightly or wrongly shifts away from those precedents, from arguments inspired in behavioral economics, and providing –at a very convenient moment- an unbeatable defensive argument for Google (go now tell Google that competition in search is not one click away or that people can’t download apps other than those allegedly bundled to Android!) Indeed, there are no technical or economic obstacles to using another search engine nor to downloading apps in mobile phones. Paradoxically, complainants in the search investigation against Google (including the intervener in this case) nevertheless complain that… network effects and inertia also need to be considered! Ah, isn’t competition law fascinating…
6) On the analysis of the cross-market effects
In our view, the main consumer harm arising from the merger affected the enterprise communications market (by granting Microsoft’s Lync -already the preeminent and arguably dominant enterprise UC solution, fueled also by the dominance of Outlook and Office- exclusive access to the great majority of users of consumer UC solutions). As publicly revealed in the course of the proceedings, this was the main business motivation for Microsoft to invest $ 8.6 billion in acquiring Skype.
Despite the “conglomerate” label, this is a fairly easy to grasp cross-market effects theory for a two-sided market: if enterprises want to access consumers and company A gains a dominant position vis à vis consumers, then it’s likely that it’ll manage to tip the enterprise side in its favor too.
This should have been the heart of the case, but it wasn’t. That’s why this aspect the case is more relevant to those interested in judicial review than to those interested in substantive issues. What happened was that, despite adequately summarizing our theory of harm in para 213 of the Decision, the Commission’s rebuttal (in 215 and 221) was -aside from internally inconsistent- totally unrelated to the concern expressed (again, don’t take my word; you can easily read them and verify it yourselves).
In view of the fact that the reasoning in the decision was invalid, the Judgment ignores it and includes a new reasoning (contained in paras 120 to 121) of which you will find absolutely no trace in the decision. I always thought that the “manifest error of appraisal test” was justified precisely for the Court not to substitute the Commission’s assessment for its own, but again I might have been wrong.
What I think explains this case is that the Commission did, in good faith, not see a prima facie competitive problem arising out of the merger, took the decision to clear it in Phase I, and then attempted to solve the difficulties it encountered by throwing out to the wall every conceivable argument and then see what stuck, including some that were not solid or that were inconsistent with previous practice.
The Court’s reaction may be explained similarly. Whereas I was hoping to have a more traditional sort of judicial review (one merely looking at whether the Commission’s conclusion –no serious doubts even in the narrowest possible relevant market- was right and rightly justified), the members of the Court probably also had an intuition that, leaving the decision aside, the merger would not give rise to real life problems and that, ultimately, its clearance was justified.
The problem is that saving the decision could only be done, as we’ve seen, by -factual issues aside- (i) adopting a questionable approach regarding the standard of proof applicable to Phase I decisions (ii) minimizing the importance of large market shares in technology markets; (iii) minimizing the importance of network effects as barriers to entry, even in a communication market without interoperability (precisely the textbook situation where they’d be relevant!); (iv) rightly or wrongly deviating –in my view- from the traditional stance towards “switching” in online markets; and (v) providing a new substantive assessment (paras 120 and 121, regarding conglomerate/cross-market effects) when the Decision’s wasn’t obviously valid.
Ends may perhaps justify means, but means create precedents.
Not that it’s of any consolation, but the Commission, and Microsoft too, have reasons to be as unhappy as we were with how things turned out. Ironically, the real winner of this case wasn’t even in the Courtroom.