European Commission’s approach to Article 22 referrals in jeopardy after AG opinion

Background

On 26 March 2021, the EC published new guidance on the referral mechanism set out in Article 22 EUMR (the “Article 22 Guidance”),¹  addressing the circumstances in which a Member State may request that the EC accept referral of a concentration over which the Member State does not have jurisdiction (i.e., while there is a national merger control regime, the concentration does not satisfy the jurisdictional thresholds).

Article 22, the so-called “Dutch clause”, was originally included to fill the enforcement gap for member states that did not have national merger control regimes, such as the Netherlands (i.e., to ensure that such a member state without the power to review a concentration that threatened to be anticompetitive could request the EC to review it). As a result, the EC has historically discouraged referrals of concentrations over which NCAs did not have jurisdiction because the thresholds were not met (rather than because there was no national merger review regime).

Essentially, the Article 22 Guidance is intended to deal with a perceived enforcement gap - to enable review of so-called “killer acquisitions”, those concentrations where “the turnover of at least one of the undertakings concerned does not reflect its actual or future competitive potential” (e.g., start-ups or recent entrants with significant competitive potential that have yet to generate significant revenues, which represent actual or potential important competitive forces). The EC’s view is that recent market developments mean that concentrations that could significantly affect competition have escaped review, particularly in the pharmaceutical and digital sectors (although nothing prevents it from being applied in other sectors).

The first application of this approach to Article 22 referrals was triggered when the French Competition Authority (FCA) sought to refer Illumina’s USD$8bn acquisition of GRAIL. Since then the EC has accepted referral requests in three further merger cases where the transaction was not notifiable at the EU level or in any Member State.

Illumina, a global developer, manufacturer and supplier of next generation genetic sequencers, consumables and related services, has significant European turnover. However, GRAIL, a U.S. company developing blood-based cancer tests for sequencing, had supplied no products, generated no revenue, and had not employed personnel to prepare for regulatory approvals or distribution in the EEA.

On 19 April 2022, the EC accepted the Article 22 EUMR referral.²  The proposed concentration did not meet the EUMR turnover thresholds and was not subject to the jurisdiction of any Member State. Nonetheless, the EC found that it could affect trade within the single market and threatened to significantly affect competition within the territory of France, and that a referral was appropriate, as its (lack of) turnover did not reflect GRAIL's competitive significance.

Illumina sought to annul the EC’s decision to accept the referral on the basis that the EC did not have the right to accept an Article 22 EUMR referral from a Member State that did not have jurisdiction under its merger control regime over the proposed concentration. The GC judgment was delivered on 13 July 2022,³  supporting the EC’s position. The AG disagrees with the GC’s findings regarding the meaning and scope of Article 22 EUMR. For the reasons stated below the AG concludes that the first ground of appeal is well founded and that the GC judgment should be set aside.

The first ground of appeal on the scope of Article 22(1) EUMR

The wording of Article 22(1) EUMR does not allow for a definitive conclusion

The AG agrees with the GC that the language of the provision wording is not so obvious that the court should refrain from considering interpretations beyond the literal interpretation. He acknowledges that the provision states that “any concentration” can be referred regardless of whether there is a national merger control system. However, he then points to two textual elements which lead him to doubt the EC position. First, the term “referral” in the title suggests that the provision concerns cases that are actually or potentially before an NCA. Second, Article 22(1) EUMR requires that the referred merger “threatens to significantly affect competition within the territory of the Member State or States making the request”. He concludes that this language makes it difficult to interpret the provision as a corrective mechanism to catch mergers that “are likely significantly to impede effective competition in the internal market”, as the GC did. The AG stresses that every provision of EU law must be placed in context and interpreted in light of EU law as a whole.

The legislative history of Article 22(1) EUMR supports a narrow interpretation

Based on the legislative history of the EUMR, the GC concluded that NCAs can refer concentrations irrespective of their national merger control rules. The AG criticizes the GC’s reliance on documents authored by the EC and post-dating the adoption of the original EUMR⁴  in 1989. In his view this is even more remarkable in the context of the EC’s position (repeated during the oral hearing) that the alleged broad scope of the provision was already there at initial adoption of the provision in 1989. This should have led the GC to assess the pre-1989 preparatory documents for the EUMR. The AG then points out that the provision had been controversial during the legislative process and had been added by the Council late in the day, making it reasonable for the GC to also rely on Council documents.

The AG went on to conclude that the documents the GC did rely on do not support its conclusions. Rather, they contradict them when read in their entirety. In the AG’s view the decisive question is whether Article 22(1) EUMR “permits Member States which have a national merger control system to refer cases that do not fall within that system”. He points out that the Article 22 mechanism was conceived for those member states without a merger control regime, and that an EC Green Paper from 2001 acknowledged that the potential scope for use had become limited (following the adoption of merger control rules in the remaining member states)⁵ . The GC failed to take this passage into account even though it contradicts the assumption that NCAs should be able to refer any concentration.

The GC characterised the broadly interpreted provision as a “corrective mechanism” to ensure effective review of all concentrations with significant effects on competition in the EU, including those concentrations that would otherwise escape review at either EU or national level. The AG does not find any support in the legislative history of the EUMR for this conclusion. He emphasises that at the time of the EUMR’s adoption it was clear to all involved in the legislative process that certain transactions would escape ex-ante review by the EC. Finally, the AG also points out that the EC itself has used the term “post-notification referrals” in the context of Article 22 EUMR, for example, in the 2005 Notice on Case Referral.

Article 22(1) EUMR needs to be interpreted in context

When analysing whether other provisions in the EUMR support the GC’s findings, the AG disagrees with the GC’s reference to the second subparagraph of Article 22 (2) EUMR. He emphasises that the interpretations of Article 22(1) and 22(2) EUMR do not need to be aligned, and that there would be no adverse consequences in terms of legal certainty and predictability if any Member State could join a referral request. The AG then points to Recital 15 EUMR which states that Article 22 EUMR enables the EC to deal with concentrations on behalf of requesting Member States. The limited scope of the referral mechanism and the boundaries of the EC’s powers are supported by the wording of Article 22(5) in the original EUMR, which stated that “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”.

Interestingly, the AG questions the need to have a corrective mechanism through interpreting Article 22 EUMR broadly, since the EUMR already contains a simplified mechanism to replace turnover thresholds with other criteria, including transaction value, if necessary (Article 1(4) and (5) EUMR). In addition, the original EUMR foresaw that the referral mechanism was only to be temporary and could become obsolete once a review of the thresholds had taken place. He concludes that this also suggests that a narrower scope for the referral mechanism was intended.

The objective of effectiveness must be balanced against other EUMR objectives

The AG notes that the objectives of Article 22 (1) EUMR are to (i) permit review where the referring Member State does not have merger control (introduced in the original EUMR) and (ii) strengthen the “one-stop-shop” principle (introduced in 1997 and reinforced in 2004). The GC identified Recital 11 EUMR as identifying a third objective, namely that the rules on referrals should operate as an effective corrective mechanism. The AG disputes this interpretation, concluding that Recital 11 was only meant to address the allocation of competences between the EC and the NCAs and address the issue of multiple filings.

The AG then turns to the GC’s conclusion that the EUMR’s objective to ensure effective control of concentrations supports a broad interpretation of Article 22(1) EUMR. He concludes that this objective must be balanced with other objectives, including allocation of competences between the EC and the NCAs, establishment of a one-stop shop and provision of legal certainty and predictability for undertakings. The AG points out that the latter is of particular importance, and that the interpretation supported by the EC and the GC can upset the balance between it and the other three objectives. He elaborates on the grave practical implications for legal certainty regarding concentrations that do not meet any merger control thresholds in the EU. In his view the GC’s interpretation would have the effect of increasing filings while reducing legal certainty as to whether and when the EC might intervene. The AG concludes that making informal notifications to all NCAs to “start the clock”, as the EC suggested, is not feasible.

The GC’s conclusions are inconsistent with general principles of EU law

The AG referred to several general principles of EU law to support his conclusions. First, the principle of institutional balance requires that the institutions exercise their powers with due regard to the powers of the other institutions. The EC’s position would devalue both the thresholds in the EUMR and in national law by making any merger notifiable. If a revision of the thresholds is required this should be a matter for the EU legislature. The AG also notes that the principle of international comity could be infringed because the referral policy could lead to the EU claiming jurisdiction in cases without adequate nexus to its territory. Finally, he questions whether undertakings with limited or no sales in the EU would be in a worse position that undertakings with more significant EU activities which benefit from the one-stop-shop system. This might infringe the principle of equality. It is worth noting that, contrary to the EC, the AG does not see the need to extend the scope of the EUMR, since mergers which do not meet the thresholds still fall within the scope of Articles 101 and 102 TFEU.

The second and third grounds of appeal

The GC had examined Illumina’s second ground of appeal, namely that the referral request was submitted out of time, noting that a referral request must be made within 15 working days of a concentration being “made known” to the Member State if no notification of that concentration is required. In the case at issue, the invitation letter, sent by the EC on 19 February 2021 to inform the Member States of this concentration, constituted the concentration being “made known”. Therefore the GC held, and the AG agreed, that the referral request was made in due time. While the EC made a procedural error by letting a period of 47 days elapse between receipt of the complaint on 7 December 2020 and the 19 February 2021, the appellants did not substantiate the argument that the outcome of the procedure might have differed without this error.

Finally, the AG concurs with the GC’s conclusions regarding the alleged breaches of the principles of legitimate expectations and legal certainty. The GC recalled that, to rely on the principle of legitimate expectation, the party concerned should establish that it has received precise, unconditional and consistent assurances, from authorised, reliable sources, which would lead it to entertain well-founded expectations. Because Illumina had failed to demonstrate this, it could not rely on the EC’s prior decision-making practice.

Final Thoughts

If the ECJ follows the AG, the EC will be faced with a dilemma.

In 2021 it concluded (following its review of the EUMR) that “the turnover-based jurisdictional thresholds of the EU Merger Regulation, complemented with the referral mechanisms, have generally proved effective in capturing significant transactions in the EU internal market” and that “the absence of complementary transaction value-based thresholds had not in itself significantly impaired the effectiveness of the existing jurisdictional thresholds”.⁶  It explicitly stated that there was no need to introduce value-based thresholds (i.e., the approach taken in Germany and Austria to ensure jurisdiction over concentrations involving highly valued companies with limited turnover).⁷

In other words, before it adopted the “new” expansive approach to Article 22, the EC concluded that the jurisdictional thresholds in the EUMR did not need to be revised. However, this conclusion came with the caveat that the historic policy of discouraging Article 22 referrals by member states without jurisdiction was “limiting its effectiveness”. As such, it appears that the EC’s conclusion was made at least in part on the assumption that the change in approach to Article 22 would obviate the ned to amend the EUMR. Article 14 of the Digital Markets Act, for example, reflects an assumption that Article 22 EUMR will enable the EC to review planned concentrations (in the digital sector or that enable the collection of data) that it is “informed” about by “gatekeepers”.

Given that a law that will create a merger review regime in Luxembourg has been presented to the Luxembourgish parliament, the window for a member state to use Article 22 to refer concentrations (whether as a result of the EC being “informed” of the deal or otherwise) taking the “historic” approach is narrowing. Once that law is on the books, Article 22 will become moot (under the historic interpretation).

This will leave the EC with the possibility of relying on ex post review of concentrations under Articles 101 and 102 TFEU (as noted by the AG), an approach that does not appear to have been a major factor in the EC’s 2021 conclusion that there was no need to amend the EUMR’s jurisdictional thresholds. That said, the EC’s thinking on the utility of the TFEU in the context of mergers may have recently evolved although it seems unlikely that it would consider being limited to ex post review to be sufficient in the context of “killer acquisitions” in particular.

As a result, if the ECJ follows the AG, the EC may revisit its 2021 conclusion that the EUMR thresholds do not need to be revised. The EC was reluctant to take this approach in 2021, and it is not clear that the EUMR would be amended now without reviving the industrial policy debates that the EC sought to avoid then.⁸

Finally, as noted above, even if the ECJ takes the approach proposed by the AG, the UK CMA’s broad jurisdictional discretion would remain intact. The CMA can open an investigation relating to transactions over which it believes it has jurisdiction without having to take a definitive position on jurisdiction until the end of its Phase 1 review. This means that it can issue an Initial Enforcement Order (IEO) that prohibits implementation of a transaction for the duration of that Phase 1 review.