Illumina/Grail : flawed originalism and the judicial hunch

Abstract

In competition law, where economic analysis and pragmatism reign, methods of interpretation generally seem to have distinctly marginal relevance. Quite rarely do antitrust or merger decisions depend on whether the applicable provisions are interpreted ‘textually’ or ‘purposively’.1 Yet the Court of Justice of the European Union’s (EU) appellate judgment in Illumina/Grail,2 overturning the General Court of the EU,3 almost reads like a manual on methods of legal interpretation. Based on a literal, historical, contextual, and teleological (purposive) reading of Article 22 of the EU Merger Regulation (EUMR),4 the Court of Justice held that a national competition authority may not request the European Commission to examine a merger that does not meet the relevant national merger thresholds. The judgment is a setback to the Commission’s attempts to work around the limits of turnover-based thresholds for the assessment of so-called ‘killer acquisitions’. This contribution makes three main claims. First, while the Court of Justice goes to great lengths to demonstrate that its own historical, contextual, and teleological interpretation of Article 22 is better than that of the General Court, from a strictly legal perspective, its judgment is no more or less convincing than that of the General Court. At important points, the reasoning of the Court of Justice ties itself in knots, and the judgment is equally flawed in its attempt to establish the ‘original meaning’ of Article 22 EUMR. Secondly, the Illumina/Grail judgment’s reliance on the various policy objectives of the EU Merger Regulation, particularly its effectiveness, predictability, and legal certainty, in my view fails to persuasively show that its own decision better ensures predictability and legal certainty than the General Court’s judgment. Although the Court of Justice’s judgment is based on inconclusive and partly contradictory arguments, its conclusion is not obviously wrong either. So, what to make of Illumina/Grail? My third claim is that the judgment seems primarily shaped by a judicial intuition, or ‘hunch’, that there is something fishy about the Commission’s interpretation of Article 22. It seems, more specifically, that the Commission was unable to explain and justify its new understanding of the alleged original meaning of Article 22: before the Illumina/Grail merger, for several years, the Commission had instead ‘developed a practice of discouraging Member States from requesting under Article 22 the referral of transactions for which they did not have jurisdiction’.5 Though the General Court showed that the literal meaning of Article 22 essentially endorses the Commission’s novel viewpoint, there is something uncanny about how the Commission exercised administrative discretion to take back control over killer acquisitions and other potentially problematic mergers that nonetheless do not meet European and national notification thresholds. The hunch is that this cannot be within the scope of the Commission’s administrative discretion. And although all of the Court of Justice’s arguments are either inconclusive or can be turned on their head (or both), the outcome of the case is entirely attuned to this hunch. Originalism is the predominant method of constitutional interpretation in American law. In Europe, most jurisdictions eschew a myopic focus on one particular method of interpretation, rather recognizing the utility of various methods such as literal, contextual, historical and teleological interpretation. Even in Europe, however, interpreting statutes is often aimed at finding the original intention of the legislature, though it is usually believed that this intention may be identified not only through preparatory documents, but also through scrutiny of the text, context, and purpose of the provision. In this sense, statutory interpretation in many jurisdictions including EU law is frequently a search for the ‘original meaning’ of the provision. This search for original meaning may include the question of how the legislature may have intended the provision to apply to changing, possibly even unforeseen, circumstances. Thus, the core question is the following: how did the EU legislature intend Article 22 to apply to below-threshold mergers? Through a literal, historical, contextual, and teleological investigation of Article 22, both the General Court and the Court of Justice were trying to answer this question. In this methodological sense, they were on the same originalist page. One of the starting points of this inquiry is whether we should look at the original meaning of the initial adoption of the Article 22 referral mechanism in Regulation 4064/89, or whether, instead, the analysis should focus on the original meaning of Article 22 of the succeeding Regulation 139/2004. The General Court in particular held that the meaning of Article 22 had changed since the adoption of Regulation 4064/89.6 As the number of national merger control systems increased after 1989, according to the General Court, Article 22 no longer applied only to referrals by Member States without any merger control; it also became a means to strengthen merger control of below-threshold concentrations with cross-border effects and to ensure the ‘one-stop-shop’ principle in EU merger control.7 At the hearing before the Court of Justice, however, the Commission was asked precisely whether the meaning of Article 22 had changed since 1989, which the Commission answered in the negative.8 The Court of Justice was not convinced either.9 More generally, in the eyes of the Court of Justice, the General Court’s originalism was flawed. Arguably, the Court of Justice did not dismiss outright the historical, contextual, and teleological arguments on which the General Court had relied upon. Rather, the Court of Justice considered them inconclusive. In turn, the Court relied on other contextual and teleological arguments that, in its view, were dispositive to reach another outcome. For example, the General Court had relied on several Commission documents from the 1990s and early 2000s to cast light on the alleged intention of the legislature in regard to the proper interpretation of Article 22.10 It had also referred to several contextual factors, including, but not limited to the fact that the notification thresholds in Article 1 of the EU Merger Regulation are ‘without prejudice to […] Article 22’,11 the fact that Article 22(1) does not expressly require the national competition authority requesting a referral to the Commission to be itself competent to assess the merger,12 and the fact that the Commission may inform any Member State that a concentration, in the Commission’s view, fulfils the criteria of Article 22(1).13 According to the Court of Justice, these factors may be relevant but do not conclusively establish the conclusions of the Commission and the General Court.14 The reason is that the General Court had failed to also take into account other contextual factors. These other contextual factors include, notably, that Article 22 differs from Article 4(5) EUMR, which provides for another referral mechanism for concentrations that do not have a European dimension on the basis of the EUMR’s notification thresholds, but which are capable of acquiring such a dimension under specific conditions.15 In contrast, mergers referred to the Commission under Article 22 do not have, nor do they acquire, a European dimension; the Commission rather replaces the national referring authority or authorities, which according to the Court of Justice implies that the referring authority or authorities must be competent in the first place.16 Like the contextual factors taken into account by the General Court, however, this argument—although indeed it supports Illumina and Grail’s position—does not conclusively establish the right interpretation of Article 22. It is not clear why the argument based on Article 4(5) EUMR carries more weight than the contextual factors relied upon by the General Court. The same applies to the Court of Justice’s argument that Article 22 should be interpreted in light of Article 1(4) and (5) EUMR, which enable the Council to revise the EUMR’s notification thresholds.17 The fact that the Council is competent to revise the jurisdictional scope of the EUMR in general surely does not irrefutably establish that an exception, which has already been deliberately included in the same instrument by the same legislature, should be interpreted more narrowly than its express formulation suggests. A similar critique applies to the Court of Justice’s teleological interpretation. The Court rejects the General Court’s reliance on recital 11 of the EUMR, which describes the various referral mechanism in the EUMR as a ‘corrective mechanism’. The General Court inferred from this recital that Article 22 intends to remedy deficiencies in the merger control system.18 The Court of Justice counters that recital 11 was only inserted in Regulation 139/2004 and was not included in its predecessor, Regulation 4064/89.19 In other words, that recital cannot support a finding of the ‘original intent’ of the EU legislature regarding the Article 22 referral mechanism. But that does not prove the contrary proposition either. The Court of Justice moves on and observes recital 15 of the EUMR, which states, among others, that when a national authority refers a merger under Article 22, ‘[o]ther Member States which are also competent to review the concentration should be able to join the request’ (emphasis added).20 It follows, says the Court, that the initial Member State to request the referral of a merger to the Commission must be competent.21 However, the core case to which Article 22 undisputably applies is a situation where a Member State does not have any national merger control rules and wishes the Commission to assess a particular concentration.22 In such a situation, the Member State requesting the referral obviously is not competent to scrutinize the concentration simply because it has no merger control system at all. On this point, the reasoning of the Court of Justice is apparently contradictory. Interestingly, while the Court of Justice rejects on originalist grounds the teleological arguments made by the General Court, the Court’s claim that it has not been established that Article 22 is ‘intended to remedy deficiencies in the control system inherent in a scheme based principally on turnover thresholds’23 is unsupported by any reference to legislative history. Absence of evidence is not evidence of absence: this truism poses a major challenge for any attempt to find the original meaning of Article 22 (and, arguably, to originalism as a method of interpretation in general). Even more remarkably, the Court of Justice glances over the apparent contradiction between, on the one hand, the claim that Article 22, read in light of recital 11, only intends to correct the allocation of competences between the Commission and the national authorities,24 and, on the other hand, the undisputed fact that Article 22 intends to apply to a situation where the referring Member State does not have any national merger control rules.25 How can Article 22 correct the allocation of competences between the Commission and a national authority which lacks any competence to assess a concentration because its Member State does not have a system of merger control? In other words, up to paragraph 201, none of the Court’s arguments proverbially knock down the General Court’s judgment. Let us consider a simple thought experiment that reverses the respective positions of the EU courts: suppose the General Court had invoked all of the contextual and teleological reasons that the Court of Justice relies on, to conclude that Article 22 should be interpreted as precluding a referral below the national notification thresholds. The Court of Justice could have dismissed any of these as inconclusive in exactly the same way that it dismissed the General Court’s actual reasoning, and could have relied on exactly those arguments provided by the (actual) General Court’s judgment to conclude that a referral below national notification thresholds is possible. The core of the Court of Justice’s reasoning, in fact, seems to be in paragraphs 202 to 218, where it relies on ‘a number of objectives which [the EUMR], taken as a whole, seeks to pursue’,26 namely the ‘one-stop-shop’ principle, a clear allocation of tasks between the Commission and the Member States, and more generally the ‘effectiveness, predictability and legal certainty that must be guaranteed to the parties to a concentration’.27 These objectives were quite clearly salient for the adoption of the EUMR, but as such they are inconclusive in regard to the proper interpretation of Article 22. This is why the Court of Justice criticizes the General Court, not for neglecting these objectives, but for ‘upset[ting] the balance’ among them.28 While predictability and legal certainty are legal principles, they are vague ones that are always balanced against other objectives. The exact balance among them often remains unsettled by legislation itself, which makes this primarily a policy question. Indeed, in paragraphs 202 to 218, we mostly find arguments of policy, not originalism. And even though the Court throws in a reference to the principle of institutional balance as part of EU constitutional law,29 such is no longer part of the ambition to conclusively establish the original meaning of Article 22. From paragraph 202 onwards, originalism is left behind. Thus, the search for the ‘real’ or ‘original’ intent of the EU legislature, for all the intellectual heavy lifting in both EU Courts’ judgments, ends quite disappointingly. Both courts did not conclusively establish the original meaning of Article 22. Like the major debates over what the Framers of the US Constitution actually intended, the quest for the original meaning of Article 22 is probably a dead end.30 If the Court of Justice’s originalist reasoning is equally flawed as the General Court’s, what about the policy arguments that the Court relied on in paragraphs 202 to 218—the ‘effectiveness, predictability and legal certainty that must be guaranteed to the parties to a concentration’?31 The Court connects effectiveness, predictability, and legal certainty to the requirements of ‘sound administration’ and ‘the business world’.32 For the Court, it follows that any exceptions to the turnover-based thresholds for notification should be interpreted narrowly. In paragraph 209, the Court almost goes as far as to say that turnover-based thresholds are indispensable for legal certainty. It would be for the EU legislature to amend these thresholds, not for the Commission to circumvent them through Article 22.33 The gap in the reasoning of the Court is that, despite the Court’s strong emphasis on the ‘cardinal importance’ of turnover-based thresholds34 and the prerogative of the legislature to amend them,35 it is precisely the legislature that expressly provided for multiple referral mechanisms deviating from these turnover-based notification thresholds. According to the plain meaning of Article 22, there is no requirement that the concentration meets any national notification threshold, let alone a turnover-based one. It is also not obvious that the interpretation by the Court of Justice results in a greater degree of legal certainty than the General Court’s judgment. The Court may have been anxious about a radical expansion of the Commission’s powers, in the words of Advocate General Emiliou, to review almost any concentration, occurring anywhere in the world, regardless of undertakings’ turnover and presence in the European Union and the of the and at any in including after the of the However, as the Advocate General the scope of Article 22 is limited by criteria that a concentration must meet in to be referred to the it must between the Member States, and it must to competition within the of the Member State or States the One may not be convinced that these criteria are to the jurisdictional scope of the EUMR. On the other hand, the of for the national competition authority and the Commission to demonstrate that the concentration competition within the Member the request cannot be dismissed as The Court of Justice these criteria in its strong support of turnover-based thresholds, in the of their in Article 22. Secondly, as other have Illumina/Grail Member States to for to scrutinize concentrations below their notification the Court of Justice’s in fact, the Commission already to Member States to make of such A of these would Illumina/Grail and may the predictability and legal certainty at the core of the Court’s to an expansion of national requirements and to those provided by Article the Court of Justice refers to the for national competition authorities to scrutinize concentrations below national notification thresholds under Article the However, this predictability and legal certainty to the of an Article 22 In to the and more even than the of powers, the of Article may take several years, which the outcome and are to The Court also that Member States may also their national notification the Court’s to the legislative is particularly however, is usually in regard to the interpretation of vague or As Article 22 is A in legal has on the factors that may judicial the more of this are the of the so-called American legal including and as as pragmatism in for example, by Though these as a there are and methodological among for relies on economic analysis as one of the in In contrast, in an in US the of and in judicial describes the as or that a seems and to make decision that seems to the according to and general of view, as the that to from the is right in general is the of a of into the and of and make no attempt to answer this question. In any judicial hunch may a better of Illumina/Grail than the legal and policy arguments invoked in the judgment. is the hunch Illumina/Grail? It is essentially that a major in the law and practice of Article 22 referrals simply should not be by a of of the in to particular policy objectives. The Commission’s on to have been particularly by three main of a Commission to scrutinize the of originalist support for the Commission’s and the of a for the Commission’s of about Article 22 As to the first as Advocate General that one by means of an original interpretation of Article 22 EUMR, the Commission the to review almost any concentration, occurring anywhere in the world, regardless of undertakings’ turnover and presence in the European Union and the of the and at any in including after the of the the Commission’s was that that not be the case as the Commission has no in that frequently and with in that This answer may be but it is the wrong More specifically, this is a wrong answer to a in search of a interpretation of Article 22. The better by the that its under Article 22 remains by the criteria of that as in the Advocate General that 15 may be to those and that this interpretation of Article 22 may scrutiny of a concentration that not to have and effects within the referring Member what this is problematic may be in light of the equally for the assessment of a concentration by the However, the Court of Justice to the provided by Article 22 and apparently them The by the Court of Justice is that and as provided by turnover-based notification thresholds, are of ‘cardinal for ‘the business It is to with this viewpoint, even though it is an rather than an originalist or strictly legal The namely of originalist support for the Commission’s also clear from the Advocate On the Commission was apparently on the exact historical of its interpretation of Article 22: The Commission was asked at the hearing whether the alleged scope of is the first of Article 22(1) EUMR was already there in the original in when that provision was in or when the new EUMR was in The Commission without that such a scope was there from the that in Article as in the of this was also a wrong Advocate General that such is the the of documents the Regulation is less the with which the Commission had that the scope of Article 22 had been from the was not expressly by any of the legislative documents from either before the adoption of Regulation or between and of express originalist support of not problematic in legal However, about an to scrutinize mergers not any of the national notification thresholds a hunch to the could be by other but this probably The Commission apparently could This to the third that to have the Commission’s namely the of a for the Commission’s of about Article 22. The reasons for this policy are and no but can it also be by a legal The for a legal argument would be the in the is interpretation to However, the Court of Justice is generally not to itself to a literal interpretation even where the is or seems to be in its the Commission in regard to its practice that the of national for merger control in almost all Member States, the in the discretion to it by the Merger Regulation, a practice of discouraging referral under Article 22 from Member States that did not have original over the at (emphasis The clearly aimed to demonstrate that the Commission’s practice was not by any of law. However, this was not by originalist argument either. it that the Commission was simply unable to answer about the basis of its in Thus, it was unable to the hunch that there was something fishy about this what the Commission could in to salient but inconclusive legislative was the plain meaning of Article 22 and the of able to take control over killer These are not But on the we a expansion of administrative about whether such expansion was intended by the legislature, and a clear between this expansion and policy of the Commission This does not and it probably a judicial hunch that to Illumina/Grail is a first of all because of the methodological and of both the General Court’s and the Court of Justice’s The judgment is also however, because of the with which the Court of Justice rejects the General Court’s reasoning as without a clearly of reasoning In whether this judgment better ensure predictability and legal certainty remains to be but the of and Article would make this If the outcome in Illumina/Grail is principally the of a judicial this is not to the judgment most are at partly by intuition, and it may not be a However, Illumina/Grail and the judicial hunch also to a in the between the General Court and the Court of Justice, namely the number of in which the General Court is This number seems may be many reasons for such and they would be to Illumina/Grail the question of how the Court of Justice whether the General Court has made an of and what exactly is an of in a like EU competition law, where law, and policy are The Court of Justice may have considered the General Court’s reasoning inconclusive and but is that to of an of there is no to be simply because the answer to the originalist question of what the legislature to is are not because we are but we are only because we are reads the by US Court Justice This is for all This however, should not and from the important question of whether they with a judgment either because it is or because a judgment would have been in the law, policy, and hunch are to