The Illumina Opinion: Article 22, Antitrust and the Rule of Law. The Devastating Critique of Advocate General Emiliou in the Illumina/Grail case
Introduction
On March 21st Advocate General Emiliou handed down his Opinion in Joined Cases C-611/22P and C-625/22P, respectively Illumina Inc. v. European Commission and Grail LLC, and Illumina Inc. v. European Commission. One can argue that the Emiliou Opinion is an exercise in throwing the proverbial kitchen sink at the position of the European Commission and the judgment of the General Court in respect of the scope of the first subparagraph of Article 22(1) EUMR. Nevertheless, despite the almost C.S. Lewis and J.R.R. Tolkien feel of the authors debating in The Eagle and Child (and ninthly!), the critique of the Advocate General is devastating.
In a detailed forensic textual and interpretative analysis of the arguments of the Commission and the General Court the Advocate General comprehensively demolishes the case of both institutions. Reading his opinion, it is difficult to come to any other conclusion than that the first sub-paragraph of Article 22(1) EUMR does not permit National Competition Authorities (NCAs) who already have operating merger review authority in their state to refer cases outside that review authority to the European Commission.
The Opinion provides a detailed literal, historical, contextual, and teleological analysis of the first subparagraph of Article 22(1) to come to that conclusion. It takes the view that notwithstanding its literal wording the scope of the referral mechanism is profoundly limited. There was no enforcement gap that Article 22 EUMR was designed to close. Nor was it the intention of the Union legislature to create a ‘competence sandwich’ where neither the European Commission nor the NCAs had jurisdiction over cases below their thresholds but where if a NCA was requested to make a referral of cases below its own thresholds, the Commission would gain jurisdiction.
Hopefully, the Court of Justice of the European Union (CJEU) will follow the Opinion in its final judgment. There are in addition two particular issues which are raised by the Advocate General which should weigh heavily in the minds of the Luxembourg judges.
First, he makes a compelling argument underpinned by the provisions, objectives, history, and logic of the structure EUMR that sustaining predictability and legal certainty in merger control has to be at the core of any system of European merger control. This is particularly so in the case of the EUMR, as it provides atypically for world-wide suspension of the closing of any deal subject to its jurisdiction. In such a context the Commission and General Court’s broader approach to the scope of Article 22(1) cannot but undermine predictability and legal certainty. This lack of predictability and certainty is reinforced by the absence of available procedures at either EU or national level to deal with cases below national thresholds leaving merging parties in a ‘legal no man’s land’.
Second, there is the underlying rule of law question. One can understand why the Commission would want to review so called killer acquisitions. There is now a considerable body of literature that persuasively argues that such acquisitions, with very small or even no turnover, may have a damaging effect on competition on the market [1]. The way to review those acquisitions however is not however to run a coach and horses through the EUMR undermining the institutional balance between the Commission and the NCAs, the subsidiarity principle, the one stop shop principle and legal certainty. As the AG argues if the Commission takes the entirely legitimate view that it needs to review ‘killer acquisitions’ it should, as provided by the procedures of the EUMR seek a formal legislative extension of the jurisdiction of the Regulation from the Union legislature.
This article is in four parts. Part two provides an overview of the history of the case, part three progressively works through the argument of the opinion setting the argument of the Advocate General in respect of his view of the literal, historical, contextual and teleological interpretation of the first subparagraph of Article 22 (1). And part four offers a conclusion.
History
This saga commences with Illumina acquiring Grail in September 2020. Illumina publishes a press release on 21st September 2020 announcing the acquisition. The acquisition was not notified under the procedures of the EUMR as the thresholds set out in Article 1 were not met. And nor was a Member State merger regime notified as no EU Member State thresholds were met either. This is perhaps not surprising that Grail did not generate any revenue in any EU Member State or elsewhere in the world.
Following a complaint received in December 2020, the European Commission in February 2021 informed the NCAs of the EEA states of the acquisition and invited them in accordance with Article 22(5) EUMR to make a referral. This letter informed the NCAs of its prima facie opinion that the Illumina/Grail concentration appeared to satisfy the conditions laid down in Article 22(1) EUMR and invited the NCAs to submit a referral request.
Subsequently, the Commission informed the representatives of Illumina and Grail of the invitation letter and the likelihood of a referral request being made under Article 22(1). A referral request was initially received by the Commission from the Autorité de la Concurrence Française on 9th March 2021 and then was followed by several more referral requests. On 31st March 2021 the Commission published its Guidance on the Application of the Referral Mechanism Set out in Article 22.
By decision of 19th April 2021 the Commission accepted the referral requests of the NCAs. The Commission found the referral requests had been submitted within the time limit of 15 working days, found that the concentration at issue satisfied the criteria for referral under Article 22(1) EUMR and rejected as unfounded the arguments of Illumina and Grail concerning an alleged breach of their rights of defence and other general principles of law.
On 28th April 2021 Illumina supported by Grail applied to the General Court seeking annulment of the decision to accept a referral from the NCAs. However, the General Court by judgment under appeal on 13th July 2022 dismissed the action. It came to the conclusion that a literal, historical, contextual and teleological interpretation of the first subparagraph of Article 22(1) EUMR supported the view that Member States may request the Commission to examine a concentration which does not have a Community dimension, even where they have no competence to review such a concentration under national law. The General Court found that, drawing in particular on Recital 11 of the EUMR, that Article 22 pursues different objectives one of which is,
‘permitting effective control, as a “corrective mechanism” of all concentrations which are capable of significantly impeding effective competition in the internal market and falling outside the scope of the merger control rules between the European Union and the Member States because the turnover thresholds have not been exceeded.’ (Judgment under appeal, para 177)
Illumina and Grail appealed to the Court of Justice on 22nd and 30th September 2022 respectively.
The Appeal to the CJEU and the Opinion of the Advocate General
The Opinion follows item by item the argument of the General Court in respect of its literal, historical, contextual, and teleological interpretation and analysis. With each item the Advocate General analyses and then dismisses the argument of the Court.
Literal Interpretative Analysis: ‘Bare Reading is Bare Feeding’
As the Advocate General says the key issue in this case is whether the first subparagraph of Article 22(1) enables the Commission to review a merger referred to it by Member State authorities, where the latter lacks any competence to review it since the merger in question falls below the thresholds set out in their national legislation on merger control.
It is worth first starting with the relevant text from Article 22:
‘One or more Member States may request the Commission to examine any concentration as defined in Article 3 that does not have a Community dimension within the meaning of Article 1 but affects trade between Member States and threatens to significantly affect competition within the territory of the Member State making the request’.
One can see at first sight that a literal interpretation of the first subparagraph of Article 22(1) would appear to support the Commission’s and General Court’s contention that NCAs could refer merger cases to the Commission which fell below the relevant merger thresholds of national merger law.
However, even the General Court accepted that the relevant text was not conclusive. In addition, the AG identified at least two textual elements in the provision that raised some further doubt as to the interpretation of the Commission and the General Court.
First, the title of the provision of Article 22 is entitled ‘Referral to the Commission’. The AG suggests that the title concerns in principle cases that are actually or potentially before the NCAs and are then referred to the Commission. And that interpretation would be in line with the legal maxim nemo dat quod non habet (no one can give what they do not have).
Second, under the first subparagraph of Article 22(1) EUMR, one of the conditions that has to be satisfied for the Commission to able to review mergers that fall below the thresholds set out in Article is that the merger in question ‘threatens to significantly affect competition within the territory of the Member State or States making the request’ (AG’s emphasis). That wording appears consistent with the original purpose of Article 22(1) to review mergers that could distort competition in a Member State that does not have a merger control system. And the wording is consistent with the 1997 and 2004 amendments to strengthen the one-stop shop process and avoid multiple national filings.
However, the wording of Article 22(1) becomes less obvious if that provision is interpreted by the General Court to constitute ‘a corrective mechanism’ to permit merger review ‘in the internal market’. This was the view of the Court that the scope of this provision of the EUMR included filling a jurisdictional gap left by transactions not within either EU or national thresholds. The AG asks if that is so why did the EU legislature refer only to restrictions of competition occurring at Member State level? He argues that given the argument of the General Court should not Article 22 (1) have referred to restrictions of competition within the internal market? And more fundamentally, why would the Commission need a referral from a Member State’s authority altogether, if the competition problem is at EU level?
Not unreasonably therefore the AG took the view that these textual elements cast some doubt on the allegedly straightforward interpretation of the first subparagraph of Article 22(1). Such a provision, which after all is a single subparagraph of an article in a regulation, is to some extent unclear and not self-contained and for which the old English adage ‘bare reading is bare feeding’ appears pertinent. Accordingly, in order to determine the exact meaning and scope of Article 22(1) it is necessary to resort to the other methods of interpretation, historical, contextual and teleological deployed by the CJEU
Historical Interpretative Analysis: The Lapalissian Truth
Turning first to a historical interpretation the AG observes that the General Court relied upon a series of documents in support of its interpretation of the provision which were all authored by the Commission and post-dated the 1989 enactment of the then European Community Merger Regulation (hereafter ECMR). As the Opinion notes in the course of the oral hearing, the Commission was asked whether its broad interpretation of what is now the first subparagraph of Article 22(1) EUMR was already present in the ECMR as enacted in 1989. The Commission confirmed that that was indeed the case.
Given the response of the Commission, it is difficult as the Opinion says to carry out a historical assessment without examining documents pre-dating the ECMR’s enactment. It is also clear from the General Court’s ruling that the Court itself also only relied on documents authored by the Commission and none from the Council, even though the Council is one of the organs of the Union legislature. No reliance was in fact made on any preparatory documents of any EU institution in undertaking the General Court’s historical assessment.
Not surprisingly therefore the AG took the view that the Court’s historical assessment was problematic given none of the documents were authored by the Council and/or pre-dated the 1989 enactment of the ECMR.
He went on to observe that the historical documents relied upon by the General Court do not in fact support a broad interpretation of the first subparagraph of Article 22(1). The documents relied upon do no more than set out the Lapalissian truth that the provision applies to concentrations with cross-border effects that do not meet the thresholds laid down in Article 1. Nothing he argues in those documents illuminates directly or indirectly, the question at the heart of the appeal: whether or not the provision permits Member States which have a national merger control system to refer cases that do not fall within that system.
In fact, the AG argues on closer examination the very documents referred to in the judgment appear to contradict the General Court’s findings and hence corroborates the interpretation put forward by the appellants. The AG underlines here that extrapolating one or more specific passages from a document and then drawing inferences from those that are inconsistent with the real content of the document as a whole is an error of law.
The Opinion then proceeds to fillet a series of passages in documents relied on by the General Court and question the reliability of the interpretation that the Court drew from them. For instance, the 2001 Green Paper indicated that one of the reasons for which the referral mechanism set out in Article 22 was underused consisted of the ‘technical differences in national merger control procedures, notably concerning the event triggering the notification and the rules concerning timing of notifications’ (AG’s emphasis). Clearly, no such consideration would have been relevant if Article 22 permitted Member States to refer concentrations to the Commission regardless of whether a notification at national level was triggered.
Or, in the 2003 Commission Proposal which indicated that the main weakness of the referral provisions (Articles 9 and 22) is the fact that they could only be used after a merger had been notified to either the Commission or the NCAs, as the case may be. Furthermore, the Proposal goes on to say that the possibility for the Commission to invite Member States to make a referral request was limited to cases that had already been notified.
The AG argues that the error made by the General Court in its historical assessment of Article 22 becomes even more evident if other relevant documents, most notably some of the travaux préparatoires, including those authored by the Council are examined. He notes that not a single document, among the vast amount of travaux préparatoires relating to the original version of the ECMR points to the referral mechanism contained in Article 22 (3) to (5) ECMR as having the ‘corrective‘ objective mentioned by the General Court. Again, the Commission was asked in the oral hearing to point to any such document and was unable to do so.
Contextual Interpretative Analysis: The Missing Context
In respect of the contextual analysis the AG provided a compelling argument that the General Court had overlooked aspects of the legal context which did support the appellants case. He noted that Recital 15 states that under Article 22 EUMR, the Commission acquires the ‘power to examine and deal with a concentration on behalf of a requesting Member States or requesting Member States’.
As the AG observes the language of the recital is hard to reconcile with a provision that-according to the Commission and the General Court-gives competence to the Commission to review certain mergers that affect competition in the internal market. He pertinently asks, if the problem is in the internal market, why should the Commission act in the interests of, in iieu of or in the name of a national authority, a fortiori one that is not competent to review the merger in question?
This concern over the language contained in Recital 15 is reinforced by the wording of Article 22(5) EUMR. It reads,
‘pursuant to paragraph 3 the Commission shall take only the measures strictly necessary to maintain or restore effective competition within the territory of the Member State at the request of which it intervenes’.
It is difficult here to sustain the argument of the General Court that Article 22 does indeed have a broad correction function attributed to it when such an express limit is placed on the powers of the Commission when acting under that provision.
Finally, the AG observes that the original referral mechanism set out in Articles 22(3) to 5 ECMR was initially conceived as a temporary mechanism. Article 22(6) ECMR provided that,
‘paragraphs 3 to 5 shall continue to apply until the thresholds referred to in Article [1(2)] have been reviewed’.
Hence the EU legislature in 1989 was of the view that the referral mechanism was destined to become obsolete once the experience on the ground permitted the legislature to make the appropriate adjustments to the turnover thresholds. Clearly, such a provision in the original ECMR would be utterly meaningless if the referral mechanism was as the Commission argued meant from the very inception of the regulation to catch concentrations falling below national thresholds.
Overall the AG took the view that with respect to the contextual interpretation is that while there are elements of context going in both directions, those pointing to a narrower scope for the first subparagraph of Article 22(1) are more numerous and relevant than those pointing to a broader scope.
Teleological Interpretative Analysis: Dealing with the ‘Competence Sandwich’
The Opinion then turned to the teleological interpretation of the first subparagraph of Article 22(1) EUMR. The General Court had relied heavily here on the preamble of the EUMR and had come to the conclusion that the gap filling objective attributed by the Court to the referral mechanism was consistent with the objective of the Regulation.
The AG however raised two issues with this interpretation of the preamble. He accepted that clearly that the objective to ensure effective control of concentration is the very raison d’etre of the regulation. However, that is not the only objective. In fact, Article 2 EUMR refers to ‘concentrations within the scope of this Regulation [to] be appraised in accordance with the objectives of this Regulation’.
He then went on to argue that the objective of permitting effective control of concentrations goes hand in hand with the pursuit of other objectives. These include first the establishment of a system in which jurisdiction is shared between the Commission and the NCAs. A second objective is to realise at EU level, an efficient system based on the ‘one stop shop’ principle, wherein the Commission has sole jurisdiction to review mergers notified under the EUMR which need no additional filings at Member State level, and the national authorities can no longer apply their national competition laws to those transactions. A third objective is to establish an efficient and predictable system capable of offering legal certainty to the undertakings concerned.
The AG noted that the third objective is common to every system of merger control. Every merger system seeks to strike a balance between effective scrutiny of competition and avoidance of unnecessary costs and delays for both the merging parties and the public administration itself. He argued that it is impossible to overemphasise the importance that predictability and legal certainty have, especially for the merging parties,
‘undertakings that are potentially subject to notification and suspension obligations need to know, with a relatively high level of confidence, whether their proposed deal will be subject to antitrust scrutiny and by which authorities, and when a definitive answer from those authorities may be expected’. (para 194 of the opinion)
The importance of legal certainty and predictability is even more vital at EU level. This is not just because within the European Union, the Union and national enforcement authorities co-exist but atypically the EUMR imposes on merging parties a world-wide bar on closing (AG’s emphasis). This means in principle that the implementation of a notified transaction must be suspended in its entirety until the Commission adopts a final decision. As the AG observes,
‘the costs and risks imposed on the merging parties are, consequently, even more significant, and those undertakings must thus be in a position to take appropriate precautions in that regard.’ (para 195)
In light of the above the AG while accepting that the effectiveness of the system i.e. the capacity to catch harmful mergers, is the primary objective of the EUMR, argues that it cannot be achieved at the expense of a satisfactory pursuit of other objectives of the regulation. He goes on to argue that the references in the preamble to ‘effectiveness’ cannot be deployed to maximise the scope and purpose of the EUMR to the point that their reach goes beyond the clear intentions of the EU legislature, upsetting the carefully devised balance between the various objectives of the regulation.
The AG argues therefore that against this background, the gap-filling objective of Article 22 advocated by the Commission and endorsed by the General Court is not consistent with the other objectives discussed above. The interpretation of the first subparagraph of Article 22(1) EUMR sits at odds with the objectives of the regulation and the balance between those objectives that the EU legislature sought to attain.
Underpinning his argument the AG points to the ‘competence sandwich’ that would follow from the General Court’s interpretation of the first subparagraph of Article 22(1). As the AG says this appears hardly consistent with what the CJEU’s description of the regulation as,
‘based on the principle of a precise allocation of competences between national and [EU] control authorities’. (para 199)
Perhaps most fundamentally the procedure that would result from a broad interpretation of the first subparagraph of Article 22(1) EUMR would hardly be efficient, predictable, and capable of ensuring legal certainty to the parties. That interpretation creates a situation that unless the merging parties take positive action to inform 30 NCAs of the existence of what is a non-notifiable merger those parties cannot have any legal certainty as to whether the Commission will not at some point in the future be asked to review the merger.
It is true as explained above the merging parties can notify all 30 NCAs. However, the reality of the situation is that there is no procedural regime at either or EU or national level which applies to such mergers. Only after the Commission has accepted the referral do some provisions of the EUMR apply. As the AG observes,
‘the period before that is a sort of legal ‘no man’s land’ in respect of which there is very little clarity and predictability’ (para 209)
That lack of predictability is enlarged by questions as to who is entitled to trigger the informal procedure? Should it include third parties? The Commission also takes the view that the 15 day working period for assessment only runs from when the NCAs have sufficient information to carry out any analysis under Article 22 EUMR. What level of detail is required from the merging parties to constitute sufficient information? And when will the parties know they have provided sufficient information? In addition to what language should information be communicated in respect of this informal procedure.
Unsurprisingly given the above the AG took the view that the teleological interpretation of the first subparagraph of Article 22(1) EUMR carried out by the General Court is erroneous, as it is inconsistent with a number of objectives that the merger control system established by the EUMR seeks to pursue, and is capable of upsetting the balance between those objectives envisaged by the EU legislature.
Systemic Issues Arising from a Broad Interpretation of the First Subparagraph of Article 22(1) EUMR
The Advocate General finally turned to a number of systemic issues when taking account of various general principles of EU law.
At the outset, the Opinion stressed that the General Court’s reading of the first subparagraph of Article 22(1) EUMR gives rise to a very significant extension of scope of the EUMR and of the Commission’s jurisdiction.
‘In one fell swoop, by means of an original interpretation of Article 22 EUMR, the Commission gains the power to review almost any concentration, occurring anywhere in the world, regardless of the undertakings’ turnover and presence in the European Union and the value of the transaction, and at any moment in time, including well after the completion of the merger’. (para 216)
The Commission indicated before the Court of Justice that it intended a future in which the provision would be applied conservatively. However, AG argued that the Commission’s position cannot but raise concerns in various regards.
For example, it is doubtful that the broad interpretation of the first subparagraph of Article 22(1) is consistent with the principle of institutional balance, deriving from Article 13(2) TEU which requires each of the institutions to exercise its powers with due regard to the powers of the other institutions.
That broad interpretation also appears to raise the potential for conflict with the principle of territoriality of EU law. It follows from the judgments of Intel and Gencor where the application of EU competition law to the conduct of undertakings is legitimate, regardless of where it takes place, so far as that conduct has foreseeable, immediate and substantial effects in the European Union.
A further concern is comity. Whilst the contours of the principle and its legal limitations are hazy nevertheless one may at least derive at the very least a general requirement for states to consider before claiming jurisdiction, the impact on third states. This is particularly so where the European Union has a rather weak domestic connection and foreign states have a stronger territorial link with a case.
Fourth the claim the AG took the view a broad interpretation of the first subparagraph of Article 22(1) EUMR conflicted with the principles of equality and proportionality. He pointed out that a broader interpretation created a situation where undertakings with limited or non-existent sales in the European Union would end up in a situation considerably worse than that of undertakings with significant activities in the Union.
The latter would benefit from either the procedural safeguards of European or national law. Those procedural safeguards provide time limits and filing requirements by which undertakings can plan and prepare their business arrangements. The former, however, would have no means to predict the fate of their merger unless they file in the EEA, no less than 30 informal notifications, and even then many aspects of the procedure remain uncertain.
Finally, the very broad scope given by the General Court to the first subparagraph of Article 22(1) EUMR when that provision is an exception to the principal thresholds set out in Article 1 of that Regulation goes against the well-accepted principle of interpretation according to which exceptions to, and derogations from, the general scheme or the general rules of a legal instrument are to be interpreted strictly so that the rules are not negated.
Conclusion
If one stands back and looks at the Opinion of the Advocate General as whole one is left with a strong sense of the entire illegitimacy of the project to expand the scope of the first subparagraph of Article 22(1) EUMR. As the AG says in ‘one fell swoop’ the Commission would gain the power to review almost any concentration in the world no matter the turnover or value of the transaction. That breath-taking grab of jurisdiction should surely alone have made the Commission let alone the General Court think twice before supporting that expansion of Article 22(1). While a ‘bare reading’ of the first subparagraph can potentially absorb such an expansion of the provision, when one considers the broader literal, historical, contextual and teleological issues, that reading cannot be sustained. How can two EU institutions have successively supported such a substantial power grab given the lack of substantive legal underpinnings to the proposition and the damage to legal certainty and predictability that would follow?
The underlying story here appears to be a willingness to bring the tech platforms under Union surveillance at whatever price to the integrity of the Union’s legal order. This approach is reminiscent of the scene in Bolt’s play a Man for All Seasons with the Commission and the General Court playing the part of John Roper, the tech platforms being embodied by Richard Rich, the onetime counsellor and ultimate betrayer of Thomas More, and Advocate General Emiliou playing the role of the Lord Chancellor himself (hopefully with a better fate than that Lord Chancellor). Roper argues for the immediate arrest of Rich. More asks on what grounds, there is no legal basis for arrest? Roper says he would cut down every law in England to go after Rich. More firmly opposes Roper, the ‘land is planted thick with laws’ there for the protection of the devil Rich and ‘my own safety’s sake’.
More and Emiliou are correct in warning on this willingness to tear up the law to go after an adversary. Such practices undermine the legitimacy of the Union’s legal order. They are in fact likely to provide additional allies to the tech platforms that the Commission is trying to target and there is also fundamentally no need to take this expansive interpretation of Article 22 (1). Given the concern surrounding so called killer acquisitions, it is likely that the Union legislature would approve an extension of the EUMR to reach such types of concentrations.
[1] For a review of the arguments and literature see, Start Ups, Killer Acquisitions and Merger Control (2020) OECD, Paris.
