The More Economic Judicature: How the General Court has Recalibrated the Merger Gauge
By Dr. Tilman Kuhn, & Prof. Dr. Stefan Thomas (White & Case LLP & Eberhard Karls University)1
In its first ever judgment dealing specifically with the interpretation of the “significant impediment to effective competition” (“SIEC”) test of Article 2(3) of the EU merger regulation2 (“EUMR”) concerning unilateral effects below the level of dominance, the European judicature has possibly opened a new chapter in EU merger enforcement policy roughly 16 years after the introduction of the test. Four years after the European Commission (“EC”) blocked Hutchison’s attempt to acquire O2’s mobile telecoms business in the UK, the General Court (“GC”) has quashed the EC’s prohibition decision, finding fault with its application of the SIEC test in key respects.3 The decision will have far-reaching implications beyond the telecoms sector. Our hypothesis is that it will require substantial adjustments to the EC’s handling of unilateral effects analyses in the future. The judgment bears on the standard of proof, on the concept of “closeness of competition” and “important competitive force,” and it finally goes to the heart of the SIEC test by demanding – to a certain degree – the integration of efficiency analysis into the theory of harm (as opposed to exclusively treating efficiencies as a mere “defense”). In essence, the GC has not rejected the more economic approach in unilateral effects analysis. Quite the opposite is true. It has advanced a more economic approach itself and thereby revealed static weaknesses in the Commission’s merger enforcement architecture.
Standard of Proof
According to the GC, the EC must demonstrate a “strong probability”4 of harmful non-coordinated effects. In particular, the GC finds that the pertinent standard of proof is stricter than the EC’s “more likely than not” approach on the basis of a “balance of probabilities.”5 In particular, the General Court held that “the more a theory of harm advanced in support of a significant impediment to effective competition put forward with regard to a concentration is complex or uncertain, or stems from a cause-and-effect relationship which is difficult to establish, the more demanding the Courts of the European Union must be as regards the specific examination of the evidence submitted by the Commission in this respect.”6 The General Court found that the EC’s theories of harm did not meet this standard in a number of respects, as further discussed below.
The Concepts of “Closeness of Competition” and “Important Competitive Force”
To establish that non-coordinated effects arising from a concentration may result in an SIEC, according to recital 25 of the EUMR, two cumulative conditions must be satisfied. The transaction must result in (i) the elimination of important competitive constraints that the parties had exerted upon each other; and (ii) a reduction of competitive pressure on the remaining competitors.7
Accordingly, in addition to keeping up the long-standing dominance case-law, the EC’s 2004 Horizontal Merger Guidelines (“HMG”)8 introduced an analytical framework with particular relevance for the analysis of unilateral effects below the level of dominance, focusing on the closeness of competition between the parties,9 and on whether the transaction removes an important competitive force (i.e. a competitor that is more important than its market share would suggest).10
Yet, the EC’s application of the “closeness of competition” and “important competitive force” criteria had become somewhat disconnected from the way the HMG discuss them. Their application in concrete cases was almost impossible to predict. That was, principally, for two reasons. First, the EC’s practice did not provide clarity on whether an SIEC was only likely where the merging parties were the “closest” competitors, “particularly close,” or just “close.” To put it differently, it was unclear, whether in a 4-to-3 merger, the merging parties could escape the finding of an SIEC only if they were the most distant (rather than just not the closest) competitors. Second, the EC has expanded the use of the criterion beyond differentiated product markets (as envisaged by the HMG), and – most notably in several mobile telecommunications cases – relied on the concept in homogeneous product markets. Especially in 4-to-3 cases, it was difficult to discern any clear principles.11
An illustrative example was Dow/DuPont. The EC demanded the divestiture of DuPont’s cereal herbicides business because, in the EC’s view, Dow and DuPont were considered “important and close” competitors in this space. Moreover, the purchaser of that divestiture business, FMC, was required to sell the business further, since it, too, was considered an important and close competitor.12 In the mobile telecoms cases, the EC typically found that all four players were close competitors. Yet still, there was an ambiguity in the use of this argument. In Hutchison 3G UK/Telefonica UK, the EC decision underlying the CK Telecoms judgment, the EC used it as a key argument to block the merger,13 whereas in T-Mobile NL/Tele 2 NL, with the same finding, it cleared the deal unconditionally.14
Similarly, the concept of an important competitive force, as enshrined in para. 37 of the HMG, had been diluted in the EC’s practice.15 That, too, is illustrated by the case at hand. There, the EC held that one of the merging parties was either such an important competitive force, or in any event an important competitive constraint.16 Effectively, both paradigms seem to have become vastly interchangeable, with little guidance on how to distinguish them from each other and from the general standard of SIEC.
In CK Telecoms, the GC held that the EC had confused the concepts of “significant impediment to effective competition” (Article 2(3) EUMR), “elimination of an important competitive constraint” (Recital 25 EUMR), and “elimination of an important competitive force” (para. 37 HMG).17
Regarding the closeness of competition, the GC insinuated that the concept may be less useful in homogeneous product markets,18 and it held that the EC has only determined that all four oligopolists were close competitors, but not that the merging parties were “particularly close,” but that in fact other players were each party’s closest competitors.19 The mere allegation that they were “relatively” – but not “particularly” – close was, in the GC’s view, not sufficient to find that the merger eliminated important competitive constraints, which the parties had previously exerted on each other. If the EC’s approach were to be endorsed, any 4-to-3 merger would, as a matter of principle, have to be blocked.20 That would not be a convincing policy.
In the same vein, the GC found that the EC had misapplied the concept of an important competitive force. The GC voiced concerns over the idea that “the mere decline in the competitive pressure which would result, in particular, from the loss of an undertaking having more of an influence on competition than its market share would suggest is sufficient, in itself, to prove a significant impediment to effective competition.”21 Such an “additional and alternative” concept22 would, according to the GC, considerably broaden the scope for the EC to intervene, given that “any elimination of an important competitive force would amount to the elimination of an important competitive constraint which, in turn, would justify a finding of a significant impediment to effective competition.”23 The GC particularly criticized the EC for the decision’s deviation from the EC’s own case-law regarding the criteria or thresholds at which it had previously found an oligopolist to be an important competitive force.24
On the Integration of Efficiencies into Quantitative Unilateral Effects Analysis
The judgment will also affect the use of quantitative horizontal merger analysis in how it defines the role of efficiencies as an element of the EC’s theory of harm. Albeit merely constituting a small part of the entire reasoning, the impact of the judgment’s passage in paras. 279 et seq. cannot be underestimated. It will make the use of Upward Pricing Pressure (“UPP”) more challenging for the Commission. And the judgment nudges the EC into, finally, resolving what the notion of “significant” in terms of the SIEC test actually means.25 What the GC essentially rules is that, when the EC steps outside the comfort zone of the dominance criterion, it cannot limit its analysis to the restriction of competition between the merging parties. It must put equal weight, in those cases, on the assessment of whether and to what extent the merger will bring about efficiencies. To put it in more general terms: Any quantification of harm presupposes the quantification of the non-existence of benefit, for if the latter is eschewed, the conclusion is flawed.
Any horizontal merger reduces competition between the merging firms, which will precipitate a unilateral upward pricing pressure, if everything else is ignored. That in itself, however, will not produce any meaningful information on whether to block or clear a merger, since it would be tantamount to effectively considering as harmful any horizontal merger. This, in turn, would ignore that fact that mergers generally bring about efficiencies to some extent.26 That, by the way, is the key reason why horizontal mergers, unlike horizontal cartels, are not prohibited per se.27 Therefore, it is absolutely mandatory to include into the merger equation a variable for efficiency considerations. To that end, there are basically two conceptual ways of doing this.28 One can either require an upward pricing pressure to be of a certain amount for it to be considered “significant” in terms of the SIEC test.29 Or, alternatively, one factors into the UPP analysis a certain amount of efficiencies that constitute a concomitant downward pricing pressure, such as a reduction in marginal cost of, e.g. 10 percent. The latter approach was suggested by Farrell & Shapiro in their original conception of the UPP test.30 Both ways effectively orbit around the same problem, which is to define a point at which a merger is likely to produce more harm than benefit.
Now, the problem with the EC’s application of UPP analysis in the past was the following: While the Commission recognized that efficiencies can serve as a downward pricing pressure, it refused to include any default efficiency parameter into the UPP equation, or to apply a de minimis threshold, hence effectively relying on upward pricing pressure void of any efficiency credit. The EC considered the burden of proof for offsetting efficiencies to rest entirely on the merging parties. The high standards for an efficiency defense, as stipulated in the EC’s HMG,31 however, were usually found to not have been met.
This way of dealing with upward pricing pressure has been criticized for warping the standard of Article 2 EUMR.32 If the agency relied on an upward pricing pressure without any de minimis threshold or efficiency credit, and without the dominance threshold being triggered either, it would effectively presume any horizontal merger to be harmful without more. The adage would be: “unless the merging parties demonstrate offsetting efficiencies, any horizontal merger will be deemed harmful for the mere fact that it reduces competition among the parties, thereby resulting in an upward pricing pressure.” That, however, does not reconcile with the law of Article 2 EUMR, which does not set up a presumption of competitive harm.
The GC, now, effectively criticized just that in paras. 279 et seq. The Luxembourg judges rightly observed that the EC cannot confine itself to demonstrating some amount of upward pricing pressure and infer from this that the merger is harmful. The GC therefore rightly reasoned that the EC, in its decision, confused two types of efficiency considerations: these within a UPP framework and those outside of it. While the latter are a defense, the lack of efficiencies as part of a UPP analysis is an element of the theory of harm. That means that, when the Commission wants to rely on UPP as an element of its theory of harm, it must assess whether there is a lack of potentially offsetting efficiencies. Since horizontal mergers usually yield some degree of efficiencies, this question cannot be just left open, or be considered a mere defense for the parties.
Conclusion and Outlook
The GC’s judgment is convincing. It puts the Commission’s enforcement practice on a tighter leash, especially with respect to the standard of proof, the closeness of competition, the concept of “important competitive force,” and last but not least, regarding the consideration of efficiencies within the UPP framework.
What will it mean for the future? The EC will need to make some important decisions, which it has shied away from until now. It will need to establish more transparent criteria on how to gauge the closeness of competition and the importance of a firm as a competitive force. This will, eventually, determine whether those concepts are viable economic tests in the first place or merely rhetorical operations. Also, the EC will need to set out principles on how to deal with the efficiency credit within the quantitative unilateral effects assessment. It must, either, define a default efficiency credit to be used within the UPP-equation. Or, alternatively, it has to resign itself to the fact that, when relying on UPP, it must demonstrate a lack of offsetting efficiencies. The latter, moreover, requires the EC to define the standard of proof for the demonstration of a lack of efficiencies. While the EC, in its HMG, makes far-reaching stipulations on the standard of proof for efficiencies as a defense, there is no guidance on the standard of proof for the absence of efficiencies as an element of a theory of harm.
The bottom line is that any unilateral theory of harm, outside the single firm dominance criterion, will rest on less firm ground from now on. Policy predictions? The GC judgment might eventually lead to a walk away from unilateral theories of harm outside the dominance criterion in oligopoly settings. Coordinated effects theories might be resurrected.33 Is this, after all, the beginning of the swansong for quantitative unilateral effects theories in horizontal mergers in Europe? Or is it a new chapter of a “more economic approach 4.0”? The ball is in the EC’s court.
1 Partner, White & Case LLP and Law Faculty Eberhard Karls University, Tübingen, respectively. Dr. Kuhn’s views are strictly personal and do not represent the view of White & Case LLP or any of its affiliates or clients.
2 Council Regulation 139/2004 of January 20, 2004, on the control of concentrations between undertakings, 2004 O.J. L24/1.
3 GC, Judgment of May 28, 2020, CK Telecoms UK Investments Ltd. v. Commission (“CK Telecoms”), Case T-399/16 EU:T:2020:217.
4 CK Telecoms, para. 118.
5 Ibid. para. 118.
6 Ibid. para. 111.
7 Ibid. para. 96.
8 Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, 2004 O.J. C31/3.
9 HMG, paras. 28-30.
10 HMG, paras. 37-38.
11 See for more details Tilman Kuhn, The 15th anniversary of the SIEC test under the EU Merger Regulation – where do we stand? Part 1, ZWeR 2020, Vol. 1, 1-51 (10-19).
12 Dow/DuPont, Case COMP/M.7932, Commission decision of March 27, 2017, paras. 795, 797, 801, 807 (“In conclusion, the Commission considers that constraints imposed by competitors are limited.”) vs. FMC/DuPont Divestment Business, Case COMP/M.8435, Commission decision of July 27, 2017, paras. 58 et seq.
13 Hutchison 3G UK/Telefonica UK, Case COMP/M.7612, Commission decision of May 11, 2016, paras. 438, 463.
14 T-Mobile NL/Tele2 NL, Case COMP/M.8792, Commission decision of November 27, 2018, para. 720.
15 See for more details Tilman Kuhn, The 15th anniversary of the SIEC test under the EU Merger Regulation – where do we stand? Part 1, ZWeR 2020, Vol. 1, 1-51 (19-22).
16 Hutchison 3G UK/Telefonica UK, Case COMP/M.7612, Commission decision of May 11, 2016, paras. 325, 777.
17 CK Telecoms, para. 173.
18 Ibid. para. 240.
19 Ibid. paras. 242, 245.
20 Ibid. para. 249.
21 Ibid. para. 171.
22 Ibid. para. 172.
23 Ibid. para. 173.
24 Ibid. e.g. paras. 183, 186, 187.
25 On that see Stefan Thomas, The Known Unknown: In Search for a Legal Structure of the Significance Criterion of the SIEC Test, 13 Journal of Competition Law & Economics, 346–387 (2017).
26 CK Telecoms, para. 277.
27 See Stefan Thomas, The Known Unknown: In Search for a Legal Structure of the Significance Criterion of the SIEC Test, 13 Journal of Competition Law & Economics, 346, 368 (2017).
28 On that see Stefan Thomas, The Known Unknown: In Search for a Legal Structure of the Significance Criterion of the SIEC Test, 13 Journal of Competition Law & Economics, 346, 361–72 (2017).
29 As a conceptual matter, the notion of “single firm dominance” is an emanation of that idea. It defines a level of monopolization, which implies unilateral effects to the extent that their harm will most likely not be offset by efficiencies. The GC judgment concerns the application of the SIEC test outside the dominance criterion, which is where quantitative methods, such as UPP analysis, have become more and more popular in the Commission’s decisional practice.
30 Joseph Farrell & Carl Shapiro, Antitrust Evaluation of Horizontal Mergers: An Economic Alternative to Market Definition, 10 B.E.J. Theoretical Econ., 1, 13, 19 n.40 (2010): “Using the illustrative efficiency parameter of 10%.”
31 HMG, paras. 76 et seq.
32 Stefan Thomas, The Known Unknown: In Search for a Legal Structure of the Significance Criterion of the SIEC Test, 13 Journal of Competition Law & Economics, 346–387 (2017).
33 Thereby ending a foray into the troubled waters of quantitative unilateral effects analysis which has started with the implementation of the SIEC test after the Airtours defeat, on that see Simon Baxter & Frances Dethmers, Collective dominance under EC merger control – after Airtours and the introduction of unilateral effects is there still a future for collective dominance, 27 E.C.L.R. 148, 159 (2006); see also Tilman Kuhn, The 15th anniversary of the SIEC test under the EU Merger Regulation – where do we stand? Part 1, ZWeR 2020, Vol. 1, 1-51 (22-28).