Relaxing whilst doing Competition Law is not an Oxymoron

Case C‑23/14, Post Danmark A/S v Konkurrencerådet: first thoughts on the judgment

with 17 comments


As announced by Alfonso, here I am to share some thoughts on today’s ruling in Post Danmark II. It is impossible to do full justice to it in a blog post. It is incredibly rich and some of the points can only be meaningfully developed in independent posts that I will try to publish in the coming weeks.

I would mention for the time being that the ruling is valuable, first and foremost, for the many issues that it clarifies. The effort in this sense is apparent (and very successful). There is also a clear attempt to bring together different strands of the case law and to achieve overall consistency (both within Article 102 TFEU and across provisions). I find it much easier to explain Article 102 TFEU case law to my students after the ruling.

It is useful to distinguish, as I have done in past entries, between issues of principle (relating to Article 102 TFEU as a whole) and issues that are specific to rebate schemes.

Issues of principle

Abuses by object and by effect: The ruling further confirms that the object/effect divide exists in the context of Article 102 TFEU. As the law stands, some practices are prima facie prohibited as abusive irrespective of the context and effects they produce (this category includes exclusive dealing and loyalty rebates). Other practices are only prohibited insofar as they have exclusionary effects (such as ‘margin squeeze’ practices and standardised rebate schemes).

Threshold of effects: While it has been clear for a while that some practices are only abusive where they have exclusionary effects, the relevant threshold has remained elusive. In line with the Advocate General, the Court seems to hold that the relevant threshold is one of likelihood. Post Danmark I is paraphrased in the ruling to state that the relevant question is ‘whether the conduct of the dominant undertaking produces an actual or likely exclusionary effect, to the detriment of competition and, thereby, of consumers’ interests’ (para 69). Para 67 (‘only dominant undertakings whose conduct is likely to have an anti-competitive effect on the market fall within the scope of Article 82 EC‘) is equally eloquent.

Effects on what?: The case law was not entirely clear as to the effects to consider when examining the lawfulness of potentially abusive practices. Some rulings gave the impression that the impact of a practice on the contractual freedom of the customers of a dominant firm was sufficient for Article 102 TFEU to come into play. Post Danmark II suggests that this is not enough. The impact of the practice on the competitive process needs to be considered. The assessment of all the circumstances relates to whether the practice ‘makes market entry very difficult or impossible for competitors of the undertaking in a dominant position’ (para 31). It is now credible to claim that the notion of ‘effects’ has a single meaning across competition law provisions (I will develop this thought further in a separate post).

Issues specific to rebate schemes

Scope of the prima facie legality rule: As I explained in my paper, the rebate scheme applied by Post Danmark was based on the volume supplied. It was therefore reasonable to infer from Hoffmann-La Roche and Michelin I that it was presumptively lawful under Article 102 TFEU. The ruling (para 28) clarifies that the prima facie legality rule only applies to rebate schemes that are (i) based on volume and (ii) that relate to individual orders. Only this category of rebate schemes is presumed to be justified by the cost savings made by the firm. The Court seems to attach importance to categories but redefines them.

Assessment of ‘all the circumstances’: According to the Court, it is necessary to consider all the circumstances to determine whether a system of standardised rebates has exclusionary effects. Such factors include the position of the dominant supplier and the conditions prevailing on the relevant market (para 30); the rules and criteria governing the grant of the rebate (para 32), and in particular the retroactive nature of the scheme and the length of the reference period. Beyond the specific issues, the judgment proposes a true effects-based analysis. The regulatory context and the features of the relevant market seem to play a crucial role in the assessment (as they should).

More on the analysis of effects:

  • The expressions ‘fidelity-building’ and ‘loyalty-inducing’ seem to be absent from the judgment. These expressions were used in rulings that suggested that a restriction in the commercial freedom of the customers of the dominant firm is sufficient to establish an abuse. This conspicuous absence suggests, in line with what I pointed out above, an attempt to bring together different strands of the case law and to propose a true analysis of effects.
  • Precisely because the Court engages in a careful analysis of the impact of the practice on the competitive process, the references to de minimis and the appreciability of the practice seem to be of little theoretical or practical relevance.

Relevance of the ‘as efficient competitor’ test: Unsurprisingly, the Court holds that the relevant case law does not mandate the ‘as efficient competitor test’ (para 57). Equally unsurprisingly, it holds that the purpose of the Guidance is merely to set the priorities of the Commission as a competition authority (para 52). More interesting is para 58, where it states that the above ‘ought not to have the effect of excluding, on principle, recourse to the as-efficient-competitor test in cases involving a rebate scheme for the purposes of examining its compatibility with Article [102 TFEU]’. In the specific context of the case, the test is not deemed to be relevant. I am inclined to agree with the Court. You may remember that I said ‘never say never’ in a previous post. Well, Post Danmark II is a good example showing that sometimes even rebate schemes that do not amount to below cost pricing can be exclusionary (the regulatory context is crucial in this regard and supports this conclusion, in my view).

Objective justification: In line with Post Danmark I, the Court confirms that an objective justification is available to dominant firms, and that it may relate to the efficiency gains made through the award of the rebate scheme (para 48). This is the aspect of the ruling that is likely to prove more controversial. Because the scope of the legality rule is defined so narrowly, it is not obvious to see how an efficiency defence will succeed (but I may be wrong). In this regard, the ruling seems to be in line with Michelin II (which helps explain, in turn, why an objective justification has not been put forward by dominant firms in some recent cases). I will develop this point elsewhere.

This seems enough for today. As usual, I would love to read your views on the ruling!

Written by Pablo Ibanez Colomo

6 October 2015 at 5:07 pm

Posted in Uncategorized

17 Responses

Subscribe to comments with RSS.

  1. […] There’s now an excellent analysis of the Case by Pablo over at Chillin’ Competition:… […]

    • Vaya birria de sentencia!

      jesus alfaro

      8 October 2015 at 11:38 am

  2. Dear Pablo,

    Congratulations for your thorough and rich analyzis.

    Always a challenge to be quick and first to comment.

    To your excellent post, I would add:
    – Article 102(3) defense is sedimented, but there remains a risk of arguing efficiencies in the dark, given the asymmetry between the liability – presumed or qualitative – and justifiability – to be proven and quantitative – prongs of Article 102 TFEU
    – The dismissal of an appreciability threshold is a concern. Not only because it assumes that a dominant position weakens competition in itself, but almost equally because this proposition is based on the very structuralist idea that this is due to “very large market share”. This is nowhere best expressed than at §40.
    – Loyalty rebates are in a category of their own, distinct from quantity rebates and from those in the proceedings, as confirmed at §28. The Intel appeal with CJEU will be key.
    – The Court draws a somewhat interesting distinction between the probability and appreciability of anticompetitive effects. It assumes appreciability, but seems to suggest that a reasonable degree of plausibility must be proven to apply Article 102 TFEU. It is certainly a commendable suggestion that probability matters.
    – Guidance paper language is taken on board in the narrative of the theory of harm: “suction”; “contestable” and “non contestable” share
    – On the AEC test, the Court does not follow the radical proposition of the AG to abrogate the test; It says it is a “tool amongst others” §61. The question remains of how this will operationally work, for instance when the Commission does not resort to the AEC test, but the parties invoke it in defense. Moreover, the notion of “tool amongst other” suggest that the Commission shall use a tool. Not necessarily the AEC, but a tool… Is qualitative analyzis sufficient to that effect?
    – The AEC test receives implicit legitimation for three types of pricing practices: selective prices, predatory prices (we knew this one) and margin squeezes (§55)
    – Last, but not least: I have been told that the Guidance Paper had disappeared from DG COMP’s website!!!!! True or hoax? I quickly searched, and did not find it. If true, what is it we shall read in this?
    Happy to discuss this on twitter, @CompetitionProf

    Again, congratulations.


    Nicolas Petit

    8 October 2015 at 3:56 pm

  3. Dear Nico,

    Thanks so much for your thoughtful comment. It provides a great opportunity to elaborate on some issues.

    – Efficiency defence: I agree with you. The problem is that the Court seems to presume that cost savings are only made in respect to individual orders (this is at least what I infer from para 28). I struggle to see, against this background, how an efficiency defence will succeed in practice. We know well that it is efficient to use rebates and exclusive dealing to sell one’s products. Unfortunately, we also know that simply claiming that a practice is a plausible source of efficiency gains is insufficient for an objective justification to be accepted (in the same vein, plausibility is not likelihood).

    – Appreciability and de minimis; As I say in the post, I do not see it with concern. It is only relevant in relation to ‘by object’ abuses (i.e. exclusive dealing and loyalty rebates). If the Court holds that it is necessary to examine the likely exclusionary effects of a standardised rebate scheme in light of the features of the market and the regulatory context, I do not see how the issue of appreciability would make a practical difference. It would be implicitly (and necessarily) addressed through the analysis of ‘all the circumstances’ (and now we know that such an analysis is not simply confined to the position of customers as Michelin II suggested; it is a true and proper effects-based analysis).

    – I fully agree with your observation about the AEC test. It is clear that the Court is not opposed to it and even suggests that it might be an appropriate proxy for the analysis of ‘all the circumstances’. My guess is that it will become the main tool de facto (other proxies are not operational). The exclusion of the test is confined to the specific circumstances of the case: a firm protected by a statutory monopoly in a recently liberalised industry. The Court is not doing anything different than sector-specific regulation (this is interesting in itself, and relevant for some aspects of my own research).

    If I had a twitter account, I would use it! 😉 Thanks a lot again!

    Pablo Ibanez Colomo

    8 October 2015 at 5:28 pm

  4. Thank you so much Pablo. On the efficiency defense, your point is all the more valid given that there are many other efficiencies that the Court does not consider, and which come in the form of reduced transaction costs, overcoming of enforcement problems in case of incomplete contracts, incentives to make relation-specific investments, etc.
    If I may add something that is, I believe of crucial importance: after Post Danmark, dominance analyzis has perhaps never been so important, and scholars may want to reinvest this deserted area of the law. Let me explain. There is good ground to take the Court on the word when it equates dominance with a market in which competition is “very weak”. This, in turn, may invite a reflection on the characterization of dominance; In positive competition enforcement, dominance is often characterized through indirect proxies (MS, barriers to entry, etc.) and the assessment tends to disdain empirical data and contemporary performance indicators. But if, after all, dominance means a situation of “very weak” competition, then this dismal competitive situation shall be witnessed in the market, and the various symptoms of SMP (or monopoly power) should be readily observable (supra-competitive pricing, rent-seeking conduct, etc). The bottom line is that if the dominance assessment is rigorous, then there is less concern to have with the hard standard of liability set for abusive conduct. And the other take away is that economic theory has much to tell us on this.

    Nicolas Petit

    9 October 2015 at 8:52 am

  5. I agree with the second part of Nico’s last post, this has been my view for years now: “it’s the dominance, stupid!”
    Jesus Alfaro, unclothing the emperor since…forever!


    9 October 2015 at 9:15 am

  6. While I agree that the Court’s efficiency assumptions are oddly narrow, I don’t see why this case would preclude the Commission or a NCA from taking all efficiency arguments into account. It’s a prejudicial question judgement; the Court has assumed there are no further efficiencies, but hasn’t decided it. (Still less decided it for all future cases.) So the net effect is simply to put some of this stuff in the justification bucket (par. 47-49) rather than in the “is it capable of being exclusionary in the first place?” bucket. I’m not sure how much that matters in practice.

    I was wondering about par. 58 though. Since when is it necessary for the application of the equally efficient competitor test that the emergence of such a competitor is in some sense practically possible?

    Martin Holterman

    9 October 2015 at 10:35 am

  7. Thanks to all for your comments!

    The issue of dominance is quite interesting. I would have two comments on it:

    – I am tempted to say that dominance is the factor that best exemplifies the shift in the Commission enforcement priorities since the late 1990s. BA and Michelin II were very controversial decisions not only because of the ostensible absence of exclusionary effects but because it is open to question whether the firms were dominant in the first place. After the Discussion Paper, the Commission seems to have re-focused its enforcement efforts towards cases in which the issue of dominance is, as a rule, far less controversial. This trend is not always given the attention it deserves.

    – Dominance and Post Danmark II: Some commentators argue that establishing the likely exclusionary effects of a practice in abuse cases is unnecessary because the firm is dominant. As the market structure is already weakened, the argument goes, one can assume that the practice is likely to have such effects. Post Danmark II is very valuable because it confirms that this interpretation of the case law is not correct. The fact that the firm is dominant does not mean that it is not necessary to show an anticompetitive effect (at least as far as ‘by effect’ practices are concerned). The same point was made in Deutsche Telekom, TeliaSonera and Post Danmark I. An interesting point, which is also relevant for the Intel case: the Court suggests in para 40 that the question of whether a dominant firm is an ‘unavoidable trading partner’ cannot simply be assumed to exist by virtue of its position; it needs to be established in light of the features of the relevant market and the regulatory context.

    Objective justification: Martin, you ask whether it might make a difference in practice. In light of past experience, it clearly does. Just take a look at Michelin II and BA. The prevailing view at the time was that efficiency gains had to be transaction-specific (which is what para 28 of Post Danmark II suggests again) and that, for instance, loyalty rebates could never be objectively justified on economies of scale grounds (this is explained very well by Luc Gyselen in his wonderful piece on rebates). It is therefore not surprising that objective justifications were not put forward in subsequent cases like Intel.

    Pablo Ibanez Colomo

    9 October 2015 at 5:33 pm

  8. Dear Nicolas, Pablo and others,

    First of all I would like to say that I had the great opportunity to met Nicolas at the Swedish Competition Authority’s Pro and Cons Conference.

    I am in no position to question my peers in this forum. However, I think that the ECJ has actually thrown a curveball. This is because they follow the same line of argument that was established in TeliaSonera and Post Danmark I, i.e. first establish the form of the abuse and then establish effects. If the latter is not established then the dominant undertaking has in fact not committed an abuse. As regards to the Post Danmark II case in, which the ECJ establishes that the AEC-test is not a pre-condiction for establishing the form of the abuse (under thoughts specific circumstances that are present in the case). However, when it comes to anti-competitive effects, the ECJ states that in order to demonstrate potential effects in the market, the rebate scheme has to foreclose a competitor that it as lest as efficient as the dominant undertaking (para. 66). My analysis might be wrong here, but from what I can understand, the ACE-test has to be applied indirectly.

    Best regards,



    11 October 2015 at 2:10 pm

  9. […] Judgment which nevertheless also included a nod to the more economic approach zeitgeist, as discussed here). He was also rapporteur in another interesting case decided on the same day, Cableuropa, the first […]

  10. Nico: isn’t 60 eur too expensive for a morning briefing?


    14 October 2015 at 11:49 am

  11. Luca: nice way to ask for a rebate :). Which deserves a reply in United Brands terms: we hardly break even, given the costs charged for room renting, breakfast, travel expenses, etc. + our rate hardly compares to those charged by other organisations. Check the November “conference about the bigger picture” :). Early bird rate is 1650€ for a day and a half. + this organization gets money from sponsoring speakers.

    Nicolas Petit

    14 October 2015 at 9:59 pm

  12. OK, I understand 🙂


    15 October 2015 at 10:27 am

  13. […] was delighted to give a presentation on Post Danmark II in the Lunch Talk organised by the GCLC in Brussels earlier today. My slides can be found here. I […]

  14. Hi Pablo,

    One more comment on Post Danmark II, but I believe a quite important one. Paragraph 28 of Post Danmark II advances a definition of loyalty rebates which is more narrow than the definition given by the General Court in Intel. It talks of a financial advantage “coupled with an obligation for, or promise by, purchasers to obtain all or a given proportion of their supplies”. This definition makes perfect sense, and understands the distinction between exclusivity obligations (loyalty rebates falls within this category) and exclusivity options (loyalty rebates do not fall within this category). In Intel, the General Court did not make this distinction, and proposed to treat exclusivity options (if you buy X%, you get a Y% discount) as loyalty rebates on their face w/o consideration whether the price advantage created economic incentives conducive to de facto exclusivity. With this, it seems that exclusivity options shall be treated under the economic approach — and possibly AEC test — described in the judgment — and possibly Guidance Paper.

    Have fun at the conference today!


    Nicolas Petit

    19 November 2015 at 10:19 am

  15. To put it even more simply, the Intel Court had put de facto exclusivity discounts in the per se prohibition box. This was illogical. How can you prove de facto exclusivity w/o looking at the facts??? Post Danmark II restores – and this is most welcome – some common sense, and says that de facto exclusivity rebates must be proven under an assessment of all circumstances.

    Nicolas Petit

    19 November 2015 at 4:25 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 )

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: