NEW PAPER | EU merger control between law and discretion: when is an impediment to effective competition significant?
A new paper, entitled ‘EU merger control between law and discretion: when is an impediment to effective competition significant?’ is available on SSRN (see here). It is an attempt to make sense of the different approaches to define the substantive test under Regulation 139/2004 and of the choices made by the General Court in its landmark CK Telecoms ruling (the appeal against which is currently pending before the Court of Justice). Your comments on the draft would be most welcome (as usual, nothing to disclose).
The fundamental point the paper makes is that the Commission, in its Guidelines and administrative practice, has interpreted the substantive test in merger control in a way that it would be fulfilled, in principle, pretty much always and everywhere (at least in relation to transactions involving actual or potential competitors). This is the key reason, it is submitted, behind the annulment of the Commission decision in CK Telecoms (if anything, the standard of proof and the role of quantitative evidence come across as side issues that follow from this core question).
Suffice it to take a look at Hutchison 3G UK/Telefonica UK (Three/O2), which is at the origin of the ruling. The crucial passage is to be found in paragraphs 310-314 and 320-326 of the decision, where the Commission advances its interpretation of Article 2 of Regulation 139/2004. According to the authority, the compatibility of horizontal transactions with the internal market is to be evaluated in light of an unstructured set of factors, each of which may or may not be relevant in the context of a particular case.
Thus, the understanding advanced by the Commission in its decision would give it the leeway to decide the exact criteria against which compatibility of a given transaction would be assessed and how the said factors are weighed against one another. What is more, the decision suggested that an ‘important competitive force’ need not stand out from rivals.
If this approach to the substantive test were to be embraced, any undertaking in a given market would qualify as an ‘important competitive force’ (something noted by the General Court itself in CK Telecoms, para 174). For the same reason, virtually any transaction involving actual or potential competitors could be declared to be prima facie incompatible with the internal market. After all, horizontal mergers lead, by definition, to the elimination of a source of competitive pressure.
Two consequences follow from the interpretation of Article 2 of Regulation 139/2004 favoured by the Commission in Three/O2. First, it is an understanding of the substantive test that does not make it possible to draw a clear line between significant and insignificant impediments to effective competition.
Second, an approach that relies on an unstructured set of criteria (the relevance and weighing of which cannot be anticipated by the parties) is not only a source of legal uncertainty (as noted by the General Court in CK Telecoms, para 175) but amounts, for all intents and purposes, to giving de facto discretion to the administrative authority. One should note, in this regard, that the case-by-case evaluation of the likely effects of concentrations is the privileged realm of ‘complex economic assessments’ in relation to which the authority enjoys a margin of appreciation.
It is submitted that an interpretation of Article 2 that gives the Commission such leeway is not obvious to square with Regulation 139/2004 and, more generally, the division of powers between the administrative authority and the EU courts. Pursuant to Article 2 of the Regulation, the substantive test is an issue of law.
In addition, the approach favoured by the Commission would not allow the EU courts to review administrative action in an effective way. If the substantive test were interpreted in a manner that does not impose meaningful boundaries defined ex ante, judicial control would be confined, in practice, to ‘manifest errors of assessment’. Such a reality would not be easy to reconcile with the fact that the interpretation of issues of law is subject to full review.
This is the background against which the CK Telecoms judgment must be understood. The General Court crafted a substantive test that defines a set of clear legal boundaries to administrative action and that makes it possible to draw the line between significant and insignificant impediments to effective competition. These legal boundaries can be found in the Table below.
CK Telecoms is no different from Airtours in the above regard. In the latter, the Commission had also defined the notion of collective dominance in a way that the test would be satisfied in pretty much any oligopolistic market (by focusing on the incentives to collude and neglecting the ability to do so on a sustainable basis, the authority identified a set of conditions that would rarely ever, if at all, fail to be met).
As was true of Airtours, the interpretation of Article 2 embraced in CK Telecoms ensures that the EU courts can perform their core function.
Chillin'Competition 
Dear Pablo,
Thanks for the post and the stimulating paper. About the Commission’s interpretation of Article 2, you write: “If this approach to the substantive test were to be embraced, any undertaking in a given market would qualify as an ‘important competitive force’ (something noted by the General Court itself in CK Telecoms, para 174)”.
But this is a clear misreading of both para. 174 and the Commission’s position. The Commission argued that an important competitive force “does not need to stand out from its competitors in terms of impact on competition” (para. 169) but also that it “must have more of an influence on competition than its market share would suggest” (para. 170), arguably in line with its Horizontal Merger Guidelines.
By no means this implies that “any undertaking in a given market would qualify as an ‘important competitive force’”. According to the GC, the Commission approach “would allow it to treat as an ‘important competitive force’ any undertaking *in an oligopolistic market* exerting competitive pressure” (quoting from para. 174). I find it surprising that you omitted “in an oligopolistic market” because this is the fundamental point of that conclusion. This also undermines your claim that the Commission’s interpretation would entail unfettered discretion in relation to *all* horizontal mergers because they always lead to the elimination of a source of competitive pressure. Nowhere did the Commission claim that in non-oligopolistic markets any undertaking is an important competitive force.
Now, some might argue that the purpose of the SIEC test is not to allow the Commission to block, subject to an efficiency defence, any horizontal merger in oligopolistic markets. But it seems to me plain wrong that the Commission’s interpretation would allow it to block any horizontal merger in *any* market. This is neither what the Commission argues, nor what the GC held, nor what either or the two implies. And it seems to me this is key to the scope of the Commission’s discretion and the relevance of legal certainty under the EUMR.
Justin
8 July 2021 at 7:17 pm
Thanks, Justin, for your comments, which are invariably thoughtful and move the discussion forward.
I would definitely agree with your point if there were a meaningful and operational way to tell apart oligopolies and non-oligopolistic markets and if real-world firms did not have at least some degree of market power. Since there is not, I very much struggle to see how it can be the fundamental point of the discussion.
In the same vein, the exercise of at least some degree of competitive pressure is the inevitable consequence of undertakings operating on the same relevant market (you will have seen that both post and paper refer to horizontal mergers). For the same reason, it can be assumed as a given when a transaction involves actual or potential competitors.
Thanks again so much!
Pablo Ibanez Colomo
8 July 2021 at 8:55 pm
Dear Pablo,
Many thanks for this post and for sharing this interesting and thought-provoking paper that I enjoyed a lot.
While reading your paper, I was particularly struck by two points:
A key assumption of your paper is that in non-coordinated effects cases ‘economic analysis does not readily provide the basis for a structured framework against which the likely impact of transactions can be established’ (10)? I wonder how this statement accounts for more than 30 years of economic analysis of unilateral effects in differentiated product markets (see for instance Willig 1991, Shapiro/Farrell 2010 to name but a few). This literature identifies the degree of substitutability between the products sold by the merging parties (aka closeness of competition) and pre-merger margins as essential proxies for the analysis of unilateral effects. Both factors inform merger control in differentiated product markets since the 1992 US Merger Guidelines. They are clearly set out in the HMGs (para. 28) and have been a central element of the Commission’s analysis of mobile mergers since 2006.
Your paper also contends, ‘that quantitative evidence is, as such, insufficient to establish, to the requisite legal standard, that a concentration is likely to lead to a significant impediment to effective competition.’ (p. 30) This point disregards that the Commission in O2/Three did not rely exclusively (‘as such’) on quantitative evidence to support the finding of a SIEC, but considered it as ‘further support’ of its qualitative evidence (see for instance para. 886 of the decision and para. 248 of Annex A). This proposition also raises the question of why the Commission should ever carry out a quantitative analysis (e.g. UPP analysis/demand estimation) in merger proceedings after CK Telecoms if it could no longer rely on it as evidence to support the finding of a SIEC. Moreover, one might wonder whether this strict evidentiary standard also applies to quantitative evidence submitted by the merging parties to demonstrate that their merger does not result in a SIEC (e.g. merger simulations showing substantial efficiencies). Further, would you also suggest that this principle extends to quantitative evidence used in other areas of competition law, say exclusionary pricing practices under Art. 102 TFEU? This might sit uneasily with your endorsement of the AEC test (at least in form of a price-cost test) for finding an abuse under Art. 102 TFEU (https://chillingcompetition.com/2021/05/10/why-article-102-tfeu-is-about-equally-efficient-rivals-legal-certainty-causality-and-competition-on-the-merits/ ) .
I have recently published a working paper that, inter alia, questions the law v discretion account of CK Telecoms that you advanced in your initial post on the judgment and that forms the core of this article. (pp 9-15, 27-28, 50-52 and 59-62 notably address this point). http://competitionpolicy.ac.uk/documents/8158338/34046344/2021-07+Working+Paper.pdf/93c5fea3-56e1-f41e-550f-cc48496f580e
Thank you again for sharing your thoughts on the CK Telecoms case.
Elias Deutscher
16 July 2021 at 12:06 am
Dear Elias,
Thanks for sharing your thoughts on congratulations on the thorough and thoughtful paper, which I enjoyed reading!
On the first point, you have rightly identified the relevant criteria to ascertain the likely effects of a horizontal merger. My claim is a different one, and does not follow logically and inevitably from your accurate summary of the issue: the question is whether these criteria provide the basis for a structured analytical framework, and one that allows to tell apart significant and insignificant impediments. The truth is that they do not (and the Three/O2 decision itself acknowledges just as much).
On the second point, I believe it is absolutely essential to distinguish the substantive test (what needs to be proved) and the standard of proof (whether the what has been proved to the requisite legal standard). This is something on which Andriani Kalintiri wrote on this very blog. It is also important to distinguish between positive and normative claims.
The blog post to which you refer addresses the what, not the whether. What is more (and this is also crucial), the post is not about the AEC test (which is a proxy), but the principle, enshrined in the case law, whereby only the exclusion of equally efficient rivals is a concern under Article 102 TFEU. The blog post distinguishes between the said principle and the AEC test. More than an ‘endorsement’, it is an attempt to capture the essence of the consistent case law of the past few years.
Thanks so much again! Our readers will have benefitted immensely from your contribution.
Pablo Ibanez Colomo
16 July 2021 at 8:25 am