Relaxing whilst doing Competition Law is not an Oxymoron

The notion of abuse after the Android judgment (Case T‑604/18): what is clearer and what remains to be clarified (II)

with 4 comments

Following my first post on the Android judgment, in which I addressed some of the issues that have become clearer after the General Court’s analysis, here is the second part, discussing those where further clarity will be needed. As ever, I really look forward to your comments (and, as ever, nothing to disclose).

What needs to be clarified after the Android judgment

The point at which an anticompetitive effect can be said to exist.

As mentioned in the first post, the judgment is valuable in that it confirms what an effect is not. It is not just a competitive advantage, and it is not merely a limitation of a firm’s freedom of action. If these factors alone are not enough, what is the point at which an anticompetitive effect exists? How to differentiate it from a mere competitive disadvantage? This is one of the issues where the Android judgment fails to shed enough light.

Suffice it to focus on the MADA aspects of the decision to illustrate this idea. The judgment covers at length the main competitive advantage identified by the Commission, namely the ‘status quo bias’ from which Google’s products would benefit and which would derive from the pre-installation of these products in smartphones.

Claims of ‘status quo bias’ raise a number of factual questions, in particular whether rival applications can be pre-installed alongside Google’s (thereby nullifying any competitive advantage) and how pervasive pre-installation was.

What matters, in any event, is that the ‘statu quo bias’ is as such insufficient to show that a practice leads, or has the potential to lead, to foreclosure. Decades of experience – including in technology markets – show that even a significant competitive advantage does not necessarily result in exclusion and that it may well be compatible with thriving competition (one needs to look no further than the Microsoft/Skype decision).

In this regard, the judgment does not offer anything by way of an approach to distinguish between advantage and effects within the meaning of the case law. It touches upon the issue, but ultimately leaves it open and unaddressed (see para 565, which ends up speaking of a ‘significant competitive advantage’). This is one of the blind spots that will hopefully be explored in the near future.

This point is not a minor or esoteric one. The development of a comprehensive toolkit to evaluate anticompetitive effects (and tell them apart from mere competitive advantages) is a necessity if the assessment is to be meaningful. In the absence of effective analytical tools, establishing foreclosure would be a mere formality, not an actual informed scrutiny of the impact of potentially abusive conduct in the relevant economic and legal context. It would be sufficient, in practice, to point to a competitive advantage.

As things stand, a toolkit exists, but only in relation to some practices. As a result, distortions will persist unless the issue is directly addressed by the EU courts. If gaps remain in the case law, the evaluation of anticompetitive effects will only be meaningful for some categories of conduct, not others. The meaning of foreclosure – and the nature and depth of the scrutiny – would therefore vary depending on the practice. It would be difficult to justify or rationalise this reality.

As far as price-based conduct is concerned, for instance, the EU courts already rely on an operational toolkit (including the ‘as efficient-competitor test’). As a result, the potential impact on competition can be meaningfully ascertained. In addition, the coverage of the behaviour has emerged, since Intel, as a helpful tool with regard to both pricing and non-pricing strategies, and it has proved particularly illuminating in some recent cases.

The Android judgment shows that an operational toolkit is missing in relation to other conduct, including tying. As a result, there is some way to go to achieve consistency within Article 102 TFEU case law (and, indeed, EU competition law at large – for instance, it has long been acknowledged in EU merger control that a significant competitive advantage is insufficient to establish foreclosure).

The benchmark against which effects are assessed

One of the reasons why this case is so fascinating is that the but-for world (the famous counterfactual) is not straightforward to establish. This is so for at least two reasons.

First (and as mentioned above), the practices at stake in the case create competition, in the sense that they sustain an ecosystem that provides opportunities for rivals to expand their business. Second, it is far from clear that alternative monetisation strategies would have given rivals more opportunities to thrive (look no further than the iPhone to realise what may happen when other business models are relied upon).

In the specific circumstances of the case, accordingly, it may well be that, in the absence of the contentious behaviour, there would have been less, not more, competition. If this is so, it would not be possible to argue that the practices have exclusionary effects. If it appears that they allow for more competition relative to the counterfactual, they fall outside the scope of Article 102 TFEU.

Fascinating as this issue may be, it was avoided by the General Court. Even though it concedes that the Android platform is pro-competitive, in the sense that it opens opportunities for rivals, it fails to consider the counterfactual.

In paras 587-596, the judgment merely points out that the Commission decision only challenged some aspects of Google’s monetisation strategy, not the business model as a whole. Which, if there was any doubt, does not address the question of what the conditions of competition would have been in their absence.

The attributability of anticompetitive effects

It is well-established case law that the effects of a practice must be attributable to the contentious practice for Article 102 TFEU to come into play. In other words, there must be a causal link between the behaviour and the impact on competition. In many respects, Android provides an ideal scenario in which to test the question: are Google’s market shares the result of its conduct or are they the consequence of the fact that its products are superior?

Inevitably, these questions emerge in the case, but, again, there is no definitive answer from the General Court. It is worth reading paras 546-558 and 570-584 in this regard. The judgment describes the evolution of market shares and explains that other competing applications are not downloaded by users. Whether or not the observable consumer behaviour can be attributed to the conduct is not something that is addressed directly.

In fact, the General Court notes at some point (para 574) that the figures outlined in the judgment are in themselves sufficient to establish harm. In this sense, the judgment suggests that attributability may be established by proxy. It remains to be seen whether this is enough to show a causal link between conduct and effects.

Written by Pablo Ibanez Colomo

3 November 2022 at 6:57 pm

Posted in Uncategorized

4 Responses

Subscribe to comments with RSS.

  1. Dear Pablo,
    This is a very selective reading of the judgment. All the more remarkable since it was delivered more than a month ago, so you had time to read it in detail.
    Some of your queries are factual issues on which the judgment thoroughly examines the available evidence. Since those issues are mainly factual, it is unclear why you are so interested in them, since you always claim you only care about the law, not the outcome of the cases. This case appears to be the exception.
    “Whether or not the observable consumer behaviour can be attributed to the conduct is not something that is addressed directly.” This sentence is really strange. The GC does address the issue, just not in the way you like/prefer. The capability to restrict competition is examined in the legal and economic context, as you always say. The behaviour of consumers (for example, tendency not to download something else if an app is already in the smartphone, a -factual!- tendency that you appear to contest, since you do not share, for example, what the GC says at 574) is a relevant context in order to examine if the conduct has the capability to restrict competition. Whether or not such behaviour is dictated by Google’s conduct, the regulatory context, cognitive bias or laziness/inertia, is irrelevant, in my view, to examine causality. But it is relevant for examining the potential impact of the conduct, which must be examined in the relevant context. There is no requirement that the context of a conduct must be also determined by the conduct. And there is no requirement that the effect must be exclusively attributed to the conduct (as you now suggest). If a conduct contributes to the effect (by making it more probable, for example) one can argue that there is a causal link already.
    In your seemingly very narrow view of the practices prohibited by competition law (where context apparently only serves to negate an infringement, but rarely to confirm it), purely theoretical alternatives appear to be enough to escape competition enforcement. The GC disagrees, for example, at 428, which makes a point of principle: “A distinction must be made in this respect between theoretical competition assumptions and the practical reality, where the competitive alternatives to which Google refers appear to have little credibility or real impact due to the ‘status quo bias’ arising from the MADA pre-installation conditions and the combined effects of those conditions with Google’s other contractual arrangements, including RSAs.”
    The GC also addresses your alternative explanation based on the alleged superior quality of Google products, even if you say nothing in your post about the relevant parts of the judgment.
    As to foreclosure, the case law does not require the elimination of all competition (that’s only a condition under Bronner), and plenty of judgments have upheld decision where some competitors remained in the market. There is nothing to clarify here, the case law is clear on that. Reinforcing a dominant position is unlawful, even if it does not lead to a monopoly. Your view of “foreclosure” is akin to applying Bronner across-the-board (you appear to suggest as much when you regret that “the meaning of foreclosure – and the nature and depth of the scrutiny (sic) – would therefore vary depending on the practice”).
    For the rest, the idea that the ecosystem would not have been created in the absence of the abuse is rejected by the GC, with very specific reasoning. Your view appears to be that since Google created Android, it can do as it wishes with the system.


    3 November 2022 at 10:32 pm

    • I really appreciate your taking the time to share your thoughts! I am sure our readers will appreciate an alternative view, too.

      If you take a look at the post again, you will see I am not discussing the facts of the case, but the underlying legal issues (what we mean by effect and how we assess it).

      The factual issues (for instance, whether other apps can be pre-installed or whether there were effective alternatives to pre-installation) are neither addressed nor challenged. And I certainly do not take views on whether or not the ecosystem would have been created in the absence of the practice, which is another factual question.

      A thought on an important point: I do not believe anybody is under the impression that foreclosure means ‘elimination of all competition’ (including, to be sure, myself). What should be clear to anyone is that foreclosure does not mean ‘significant competitive advantage’, either. The latter point of law is the one I address, not the former.

      Thanks so much again!

      Pablo Ibanez Colomo

      4 November 2022 at 9:26 am

      • There is little point in continuing this conversation, since the case was not simply based on the existence of a ‘significant competitive advantage’ (as you post and comment suggest) but much more. Anyway, your issue with this point of the judgment (echoing one of Google’s arguments in the proceedings) is already replied in a specific section of the judgment, in particular, at paragraph 564: ‘As to the alleged conflating of competitive advantage and anticompetitive foreclosure, it must be pointed out that this complaint is based on a misreading of the contested decision, from which it is apparent that it establishes, first, the existence of an advantage linked to the MADA pre-installation conditions that cannot be offset by competitors, and, second, the anticompetitive effects of that advantage.’ I know the judgment is long, but we all must read it in full.


        5 November 2022 at 1:20 pm

      • Once again, thanks so much for sharing your thoughts!

        It looks like we agree on a great deal. We both agree that the case is not simply about a ‘significant competitive advantage’.

        As I explained in my previous post, it is clear from the judgment that a significant competitive advantage is not enough to show foreclosure (another aspect on which we seem to agree).

        The point (admittedly modest) I make in this second post is that the line between significant competitive advantage and foreclosure is not clear. Whether or not the two concepts were conflated in the original decision is the sort of case-specific issue in which I am not interested.

        The notion of ‘significant competitive advantage’ was not created by Google. As you certainly know, it was central to the Commission’s analysis in the Microsoft saga.

        Thanks again so much, also on behalf of the Chillin’ community!

        Pablo Ibanez Colomo

        5 November 2022 at 8:56 pm

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: