Take it or leave it: Referrals in the EU

There has been quite some talk lately about how the European Commission and national regulators will deal with their respective competences once the Digital Market Act has been set into force. In merger control, referrals in both directions are well established with some very crucial legal and political questions currently in the spotlight. An opportunity to summarize the legal framework, latest trends and to venture an analysis of policy goals – also going forward.

The general rule as to how merger control responsibilities within the European Union are distributed is fairly simple and solely based on objective criteria: If the merging parties exceed certain turnover thresholds of the European Merger Control Regulation (ECMR), a transaction must be notified to the Commission as a one-stop-shop. National regulators within the EU can generally not get a grip on these cases. If the thresholds are not met, national regulators can be competent to review the transaction.

There remains a possibility, however, that the regulator originally competent to review a case is deemed not be best placed to conduct the review. Examples would include a transaction concerning several Member States without triggering the EU thresholds or a transaction triggering the EU thresholds while largely relating to business activities in a particular Member State. To address these situations, EU regulators may reallocate the competency to review a given transaction.

It’s big, but local

Article 9 of the ECMR gives Member States the option to even intervene ask for jurisdiction over transactions which technically are notifiable to the Commission. Such a referral request has to be submitted within fifteen days after a transaction was filed to the Commission. To be successful, the requesting Member State must show that the transaction might affect competition in a market which is not broader than national, and which does not constitute a substantial part of the European Single Market. The Commission will assess whether the national regulator has a point but also whether giving up jurisdiction will lead to a fragmentation of reviews or create legal uncertainty. In light of these criteria, it is quite apparent that the Commission still has wide discretion in practice.

Looking at official statistics, one could get the impression that the Commission’s approach has somewhat shifted since 2020. Whilst during the period from 2010 to 2019 and approx. 60 referral requests overall, roughly one third of them was refused, another third was partially referred and only the last third of cases was entirely referred to the requesting Member State, four out of four cases each in 2020 and 2021 were referred in their entirety. It is probably too early to tell if this is part of a more general trend, so it will be interesting to see how the Commission will decide in the currently ongoing Kronospan/Pfleiderer Phase 2 investigation where the Polish authority has tried to get the case under its roof.

In any event, it can be observed that successful referral requests do often not make the life of the merging parties any easier (and in cases where they might, the parties also have the power to request a referral to Member State level). For example, at least three referrals that were successfully filed by the Bundeskartellamt in the last two years went into a Phase 2 investigation and/or were only cleared subject to remedies on the national level.

It’s small, but all over the place

Mirroring Article 9 in a way, Article 22 allows Member States to send cases to the Commission which do not meet the EU thresholds. Such transactions should affect trade between Member States and threaten to significantly affect competition within the territory of the referring Member State. The test reads like this might be easily established. Indeed, this assumption is also reflected in the numbers. Out of 43 referral requests since 1990, only four were refused and sent back to the national regulator.

Referrals under Article 22 are traditionally feared by transaction parties because of their potential impact on the deal timeline. Whilst Phase I proceedings in various Members States are somewhat predictable, it is far more difficult to foresee the impact of an Article 22 request.

Another twist to Article 22 referrals is the fact that the Commission only gains the competency to review cases in relation to the Member States joining the referral. For EU-wide markets, this might not make a difference. But: National regulators even in countries where merger control thresholds were not triggered can join a referral. A referral can have a huge impact in cases where markets are national or regional, because – depending on which Member States join a referral – certain markets will all of a sudden be scrutinized when they would not have been without a referral. And, Article 22 referrals can lead to split competencies where the Commission takes on a case but national regulators also reviewing the same case do not join a referral. Such cases are rare, though, because in practice national regulators usually align behind the scenes before a referral request is made. (however, see just two paragraphs below regarding Facebook/Kustomer).

Article 22 has been in the spotlight even more since the Commission published new guidelines on its referral mechanisms last year. The Commission announced that it will also accept referral requests for transactions from Member States that fall below the national merger control thresholds – i.e., that were not notifiable anywhere in the EU, at all.

Whilst most regulators like the idea of catching more potential “killer acquisitions” that fall below national thresholds, taking the legal basis for the policy shift into consideration, the feedback on the Commission’s (perceived) new approach has been mixed. France, on one end of the spectrum, is very supportive of the idea and most recently announced to introduce a detection unit designated to these cases. Germany and Austria are prominent examples of a more sceptical view (Facebook/Kustomer has been the most prominent example this far, as reported here). Under regimes in both Germany and Austria, there is a transaction value-based threshold which arguably lowers the numbers of cases not facing merger control scrutiny at all.

Now,everyone is eagerly awaiting the General Court’s decision on the Illumina/Grail transaction, which was the first case referred to the Commission under the new Article 22 mechanism. Regardless of the decision, the policy debate will not be over even after that.

Lessons learnt?

What can be drawn from the above regarding allocating responsibilities going forward? For one, many regulators seem to agree that creativity is appreciated when it comes to getting a grip on cases they deem interesting. This is certainly true for mergers, but regarding the digital world, statements of officials go in a similar direction. Also, the Commission gives the impression to leave national regulators room to operate if they can demonstrate potential antitrust concerns largely relating to national markets. These observations do certainly not resolve the questions circulating around the DMA and corresponding national initiatives. However, they could influence the one or the other opinion in the future.