With the recent Autotalks referral request the European Commission’s new approach to the application of Article 22 of the EU Merger Regulation (EUMR) becomes ever more a regular feature of EU merger law. Nevertheless the Commission’s assertion of extended jurisdiction under Article 22 remains unlawful, unconstitutional, and ultimately unsustainable. Even though its new policy was supported by the EU General Court in the Illumina case the judgment is unconvincing and inadequate. It also involves an underlying constitutionally stunning proposition: that without the explicit consent of the Union legislature, the national competition authorities (NCAs) should be able to transfer merger cases to the Commission which they themselves had no national competence to assess. In other words, the EUMR is able to be deployed by the NCAs to grant the Commission jurisdiction over cases which national parliaments have never empowered NCAs to review, and for which the EUMR does not otherwise grant the Commission jurisdiction. The EUMR is in effect being used as a mechanism to subvert both national and EU competence. Clearly the principal objective of this assertion of jurisdiction is to be able to call in so-called ‘killer acquisition’ cases involving the major tech platforms. Whilst this policy goal may be understandable, such an expansion of jurisdiction at the least requires, it is submitted, an express legislative amendment to the EUMR.
European Union Merger Control: Its True Structure and Purpose.
When the original European Community Merger Regulation (ECMR) 4064/89 came into force in September 1990 the objective was clear. It was to create a ‘one-stop shop’ Community merger regime for the largest merger cases which had significant cross-border effects.  As a proxy for significant cross-border effects the ECMR deployed turnover thresholds to determine cases which had a ‘Community dimension’ and fell within the regulation’s scope. The thresholds defining Community dimension were always considered an imperfect but acceptable proxy to identify the merger cases which were likely to have the most significant cross-border effects.
By setting merger thresholds, with cases crossing those thresholds being deemed to have a Community dimension, the ECMR created an exclusive merger jurisdiction for the European Commission. This was expressly an ex-ante system. It sought to assess a proposed merger before it was implemented: the ECMR provided that such mergers were not to be concluded before the Commission has taken a decision to either clear the merger, clear it with conditions or prohibit the merger. The Regulation also provided merging parties and potentially interested parties, as well as the Member States, with a series of procedural rights, express tight deadlines, a right to be heard, and a right of appeal.
In addition, the ECMR contained two mechanisms for transferring cases between the Commission and NCAs. Under Article 9 the Commission could transfer a case with a Community dimension back to a requesting Member State where the case had all the attributes of affecting a distinct market located substantially in that state.
Article 22 by contrast provided a mechanism by which NCAs could transfer cases to the Commission. This was famously known as the ‘Dutch clause’ because at the time of the enactment of the ECMR, it was principally the Netherlands which had advocated for such a mechanism as it did not have its own national merger control regime (although other Member sStates such as Belgium and Italy who joined the Dutch advocacy efforts also did not have merger regimes in 1989).
Following the coming into force of the ECMR almost all Member States (save in fact now Luxembourg) adopted national merger controls. This however created a new challenge for European merger control: a merger below ECMR thresholds could often result in multiple national filings. This ‘multiple filing’ issue was seen as wasteful of resources and something that in principle that ECMR should be capable of resolving via its one-stop shop regime. An amendment to the ECMR adopted in Regulation 1310/97 provided Article 22, by an amendment to Article 22(3) with a new purpose: permitting joint requests by two or more Member States to refer cases to the European Commission.
Subsequently in the major reform of EU merger control system in 2004 with the enactment of the EU Merger Regulation (EUMR) 139/2004, the referral mechanisms were further refined. The EUMR provided for a pre-notification regime under Article 4 which permitted merging parties to approach the Commission to request that an EU merger case be transferred to a Member State (Article 4(4)); or where a case was reviewable in three or more Member States, for the case to be transferred up to the Commission (Article 4(5)). Member States retained their capacity on grounds of the existence of a distinct market to have cases transferred back to them under Article 9, and the capacity to refer cases to the Commission under Article 22.
A Step too Far: The Commission’s Assertion of Wider Article 22 Jurisdiction.
In its March 2021 Communication on the application of the Article 22 referral mechanism the Commission claimed a broad jurisdiction to be able to seek to call in deals which fell below its own merger thresholds and those of the EU Member States (hereafter the Guidance). It took the view that the EUMR contained a broad ‘corrective mechanism’ to the application of the jurisdictional thresholds which brought cases under EU review:
“the wording, legislative history and purpose of Article 22 of the Merger Regulation, as well as from the Commission’s enforcement practice, that Article 22 is applicable to all concentrations, not only those that meet the respective jurisdictional criteria of the referring Member State” (para 6).
In its Guidance the Commission also made it clear that a major objective of this assertion of jurisdiction was to seek to call in cases which currently have limited or even no turnover, but which may subsequently play a significant role on the market. It identified such developments as being particularly significant in the digital economy, where services regularly launch with the aim of building up a significant user base and/or commercially valuable data inventories, before seeking to monetise the business (para 9).
The Commission accepted it was changing current practice by taking the view that it could seek to call in cases over which NCAs themselves did not have jurisdiction. However, it argued that such a change did not require modification to the relevant provisions of the EUMR (para 21).
The Commission also indicated that the fact a transaction has closed does not preclude a Member State from requesting a referral. It did indicate that it would not generally accept a referral if made more than six months after closing. However, two other factors set out in the Guidance indicate that the six-month rule is not in fact much of a limitation. First, it indicated that the six months would run from when material facts about the concentration had been made public in the EU. Given that the acquisition of many start-ups or early development businesses may involve limited or no public disclosure, this ‘public domain’ exception may well result in cases being called in well after closure of a deal and their integration into the business of the acquirer. Second, the Commission also indicated that a later referral may be appropriate in any event given potentially significant competition concerns (paras 26-27). It is not unreasonable therefore to read the Commission’s Guidance as having created a form of ex-post system of merger control.
In procedural terms where the Commission becomes aware of a case that it considers as meeting the relevant criteria for a referral it will contact the Member State(s) concerned and invite a referral request. The Commission will seek to inform the merging parties that a referral request is being considered who may decide to take appropriate measures such as delaying completion of the transaction.
Such a referral request must be made at most within 15 working days of the concentration being made known to the Member State concerned. According to the Guidance ‘made known’ should be interpreted as implying sufficient information to make a preliminary assessment as to the existence of the criteria relevant for the assessment of the referral. Once a referral request has been made the Commission will inform the other NCAs. Other NCAs may join the initial request. At the latest 10 working days from the expiry of the original 15- working day period the Commission will decide whether to examine the case (para 28).
If one stands back for a moment and considers this assertion of jurisdiction in the March 2021 Guidance, it is difficult to be left other than stunned. As explained above the EUMR was developed as an ex-ante one-stop shop merger control system to deal with the most significant cross-border merger cases. This system provided tight deadlines, transparency, due process and legal certainty. The Guidance by contrast seeks to create a very different merger control system by administrative fiat. It asserts jurisdiction to call in cases which are not subject to national merger control law, and over which the Commission itself does not have jurisdiction. By definition, these cases are not the most significant merger cases with cross-border effects for which the EUMR was established and the European Commission given jurisdiction by the Union legislature. More, as explained above, the Commission’s Guidance also substantially abandons the ex-ante system of review for these cases not within national merger thresholds, and replaces it with an ex-post system. With an ex-post system, with no final closure date for application of Article 22, legal certainty is undermined. Merging parties cannot be sure where and when the Commission may seek to assert jurisdiction. The security of any deal which is outside the jurisdiction of the Commission or the NCAs will be questionable.
The fact that the referral system created by the Guidance can apply to an unlimited class of cases which fall outside EU and national merger control systems creates further difficulties. As national merger law cannot suspend cases outside of their own jurisdiction which are being considered for a referral request, the Commission in the Guidance is left to suggest that the merging parties may like to consider taking their own steps to delay the deal. Equally, the lack of a formal notification procedure under national law leaves merging parties and the NCAs in some confusion as to when a deal is ‘made known’ to the relevant NCA. At what point is sufficient information deemed to be available to the NCA to trigger the initial 15 working day time period? Do the merging parties have to contemplate, on hearing that the Commission is asking NCAs to make a referral request, making a substantial submission to all NCAs who may be conceivably interested in the case? The lack of predictability and certainty and confusion here stems from the fact that there are no national merger procedural rules upon which either the merging parties or the NCAs can rely.
The General Court in Illumina
The first firm that was subject to the Commission’s new interpretation of Article 22 was Illumina in its attempt to acquire Grail. It sought to challenge the Commission’s interpretation of Article 22 as set out in its March 2021 Communication before the General Court (Case T-227/21, Illumina v. Commission). It failed. That case is now on appeal to the Court of Justice of the European Union (CJEU). Nevertheless, the weakness, inconsistency and incoherence of the General Court’s judgment reinforce the arguments above as to the unlawfulness, unconstitutionality and unsustainability of the Commission’s interpretation of Article 22.
One has to first concede the point to the General Court that a literal reading of Article 22 of the EUMR does not indicate any limitation on the application of its referral power of a NCA. Article 22 reads as follows,
“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 or Member States making the request”.
However, any reading of a specific provision has to take account of the objective of the legislation, its historical context and the place of that provision in relation to other provisions of the relevant instrument.
As already explained above the Commission’s new interpretation of Article 22 runs wholly counter to the object and purpose of the EUMR. It undermines the concept of the one-stop shop; creates a form of ex-post review which the EUMR sought to avoid; and as a consequence creates significant legal uncertainty for economic operators since it leaves merging parties without the due process rights that apply in other merger cases. Furthermore, as also indicated above, Article 22 itself was adopted to deal with a specific issue, the lack of merger control review in a handful of Member States. Subsequently it was refashioned to deal with the challenge posed by multiple merger filings in some cases. At no point was there ever any discussion of a broad general power to refer cases outwith the jurisdiction of Member States themselves.
The General Court in its Illumina judgment fails to effectively address any of these concerns.
The General Court did seek to undertake an historical analysis of the EUMR legislation. In this however, it also failed. In order to undertake such an analysis one would have to consider the travaux préparatoires of the Union legislature in respect of the ECMR and EUMR back to the origins of Regulation 4046/89. The Union legislature consists of what is now the Council of the European Union, and the European Parliament, with the European Commission making legislative proposals which the legislature then considers, amends, or rejects. The General Court instead relies solely on three documents to seek to underpin its support for the Commission’s interpretation of Article 22. These are the 1996 Commission Green Paper on Review of the Merger Regulation, the 2001 Commission Green Paper on Review of the Merger Regulation and the 2003 Proposal for a Council Regulation on the Control of Concentrations Between Undertakings (para 97). So the General Court did not consider materials from the Council or Parliament, the bodies that actually adopted the legislation, nor any materials from 1989 when the ECMR was first adopted. Why the General Court fails to undertake a fuller analysis of the travaux préparatoires is unclear. However, seeking to interpret Union legislation solely on the basis of Commission Green and White papers is neither credible nor convincing.
The intentions of the Member States are all the more important in this specific case as the central provision under discussion is Article 22. As explained above, Article 22 originally came into being because of the demands of Member States seeking a means of referring merger cases to the Commission as they did not themselves have merger control regimes. Therefore, the debate within the Council over this issue at the time of the enactment of Regulation 4064/89, and subsequently in respect of Regulation 1310/97 and with the latest iteration of the EUMR Regulation 139/2004, would surely be critical to any correct interpretation of Article 22. However, the relevant documentation and the discussion by the Council members is entirely absent from the judgment of the General Court.
When one considers the actual travaux préparatoires it is difficult to see how the judgment of the General Court can be sustained (see rapport du Groupe ad hoc Conseillers économique en date du 9 décembre 1988 au Comité des Représentants Permanents Doc 10189/89/EC (RC 36) at page 9). It is clear from the discussion in the Council in the run-up to enactment of Regulation 4064/89 that the focus of the Member States was solely to provide a mechanism to permit States without a merger control jurisdiction to refer cases to the Commission. There was no intent to create a system of general jurisdiction (page 3).
Equally, amendments to the Merger Regulation in 1997 and 2004 were focussed solely upon providing a means to deal with the challenges surrounding multiple merger filings. This view of the limited purpose of the EUMR in respect of Article 22 is reinforced by a further consideration of the rest of the text of that provision and Recital 15. That text heavily suggests that Article 22 was only supposed to apply to cases where the NCAs did already have jurisdiction under national law. For example,
Article 22(2) provides that, “all national time limits relating to the concentration shall be suspended until, in accordance with the procedure set out in this article, it has been decided where the concentration shall be examined“.
Article 22(3) goes on to provide that following an Article 22 request, “the Member State or States having made the request shall no longer apply their national legislation on competition to the concentration“.
In addition, Recital 15 of the EUMR provides that when a referral request has been made by one Member State, “Other Member States which are also competent to review the concentration should be able to join the request“.
The General Court faced with these provisions proceeds to seek to minimise, obfuscate and offer profoundly unconvincing explanations as to why these provisions do not mean what they say. Fundamentally, the difficulty here for the General Court is this: if there was a general power to deal with referrals from the NCAs over which the NCAs did not themselves have jurisdiction, why would Article 22(2) and (3) exist at all?
In reality the relevant EUMR provisions all reflect the limited construct and purpose served by Article 22 and are wholly inconsistent with any broad general power to extend EUMR jurisdiction willy nilly to any case catching the Commission’s attention. And indeed there is not the slightest hint in the travaux préparatoires of an intention to empower the Commission in that way.
Conclusion: Jurisdictional Overreach and Judicial Failure
It is a legitimate policy goal to seek to regulate ‘killer acquisitions’, particularly start-ups acquired by dominant tech platforms. However, seeking to do so by manipulating the scope of the EUMR via administrative fiat is not the way to do it, even if the General Court is unwisely supportive. The Commission should propose legislation to amend the EUMR to bring such acquisitions within its scope or consider proposing free-standing legislation.
Given the argument laid out above it is likely that the CJEU will reverse the judgment of the General Court in Illumina. However, even if the CJEU follows the argument of the General Court the consequences of this unsustainable and unconstitutional assertion of jurisdiction will not go away. When will a case below even NCA thresholds actually ever close? What legal certainty can merging parties have? How are procedural rights protected in a case where no national or EU laws initially apply? These issues, amongst many others, will be endlessly litigated in the courts.
Second, and perhaps most significantly, the constitutional issues cannot be washed away. It may well have seemed a good idea to assert jurisdiction against killer acquisitions amongst antitrust regulators in the Madou Tower and in some of the Member States. And an unwise General Court helpfully backed the regulators. However, in the broader world of national media, national parliaments and ultimately national governments, this assertion of jurisdiction is likely to be seen as a constitutional abomination.
In essence, this assertion of jurisdiction involves the Commission obtaining the power to review cases referred from Member States which those Member States’ own NCAs have no jurisdiction themselves to review. For many governments, parliaments, populist parties and media this is a frankly outrageous abuse of power by the Commission, aided and abetted here by the General Court. The only reason this has not so far become a major political issue is that it has remained shrouded in obscurity within the antitrust community. As the Illumina case approaches the CJEU that obscurity is likely to be lifted.
 Perhaps the best overall view of the objective and purpose of the ECMR at close to the moment of its creation can be found in the article of the late Lord Brittan in his 1991 article, The Development of Merger Control in EEC Competition Law, in Competition Policy and Merger Control in the Single European Market, Grotius Publications, Cambridge 1991.