Chillin'Competition

(Guest post by Marcos Araujo Boyd)

On 6 October 2021, the Grand Chamber of the CJEU issued its much-awaited decision in respect of the legal entities against which follow-on claims may be made (link here)

This decision that will be remembered for reasons way beyond the liability of subsidiaries in a follow-on cartel claim. As suggested by the appointment of a Grand Chamber, the Court soon realised that it would have to reconsider a particularly convoluted area of EU competition law:  the theory of the undertaking or economic unit and its relationship with legal persons and groups of companies. The answer provided may be ranked alongside Hydrotherm, Viho, Dansk Rørindustri , Confederación Española de Estaciones de Servicio , Akzo Nobel or Skanska in its importance on the construction of the undertaking as a legal concept in EU competition law.

The discussion follows the order of the arguments in the judgment starting from its paragraph 31, after having dealt with procedural and admissibility issues. For a fuller discussion on the context of the judgment, the reader is invited to check my previous guest post on AG Pitruzzella’s Opinion.

The opening statements: On Giant’s Shoulders (paras 32 to 37)

After rephrasing the three initial questions as whether a victim may indifferently sue a parent company sanctioned by the Commission or an affiliate provided both entities constitute an economic unit, paragraphs 32 to 37 of the judgment recall the jurisprudence of the Court on private enforcement. The arguments feature Skanska prominently, stressing that the determination of the liable entity is a matter of EU law only, and abundantly noting the link between public and private enforcement.

While there is little new in this section, these references usefully reveal the reluctance of the Court to heed to the temptation of facilitating private claims under the principle of effectiveness or, as suggested in para 52 of AG Pitruzzella’s Opinion, by admitting that national courts affirm it without that being required by EU law. Rather, the Court builds its arguments on private enforcement over the strong shoulders of public enforcement, dismissing by implication potential divergences between the two tools. That perspective enables it to rely on its rich public enforcement case-law in the sections that follow.

The Centrality of the Notion of Undertaking (paras 38 to 44)

Following the discussion on public and private enforcement, the Court moves on to a second constitutional stepping stone, constituted by the notion of undertaking. Quoting earlier jurisprudence, the Court depicts it as ‘an autonomous concept of EU law’ that designates ‘the perpetrator of an infringement (…), who is liable to be punished’ and ‘the entity on which the Commission may impose a fine’, contrasting it with other concepts such as companies or legal persons, and observes that this notion is employed both in primary and secondary legislation, especially the Damages Directive 2014/104. It then recalls the jurisprudence on the notion of undertaking by recalling Imperial Chemical Industries, Confederación Española de Estaciones de Servicio, the 2009 and 2017 Akzo Nobel judgments and Knauf Gips before moving on to the application of the principle of personal responsibility to the undertaking and not to legal entities, an apparent oxymoron that has, not without some reaction from various AGs, featured prominently in EU competition law since ETI. That recollection ends with a surprising, yet reiterated, principle of EU competition law used in public enforcement since at least Siemens Österreich whereby the joint and several liability amongst the entities of a single economic unit applies ipso iure or automatically, no decision to that effect being actually needed, an argument that resonates differently in the context of private enforcement than when discussing a case where the separate entities have been identified in a decision following a procedure. That is, in any event, inevitable given the logic of parallelism between public and private enforcement already noted.

The consequence of the above is clear: upward and downward liability are placed on equal footing, both resulting from the very nature of the undertaking as defined in EU law, and not as a result of control or agency theory. But thar is not the end. Keep reading.

Undertakings within Groups of Companies? (paras 45-50)

After reaching the above conclusion, the Court moves on to a correction required by the problem identified already by AG Pitruzzella: the link that should exist between the legal entity against which the claim is made and the undertaking that is initially liable. It will be recalled that the AG had proposed to require an involvement by the subsidiary on the specific economic activity under consideration, for example, by selling the goods object of the cartel (see para 57 of the Opinion).

Quoting the AG’s Opinion, the Court follows its logic with a significant twist. Taking conglomerate groups as an example, the Court argues that groups of companies may contain various ‘economic units’ (or undertakings, although the judgment avoids that term in this context). This would be the case where the groups are active in ‘several economic fields having no connection between them’. It even notes that, in those conglomerate groups, ‘the same parent company may be part of several economic units made up (…) of itself and of different combinations of subsidiaries all belonging to the same group of companies’, thereby affirming that a group of companies, all linked by control, may actually contain several separate ‘economic units’.

This logic is used to solve the absurdity that a subsidiary ‘could be held liable for infringements committed in the context of activities entirely unconnected to its own activity and in which they were in no way involved, even indirectly’ (para 47). However, its impact is far reaching beyond the matter at hand, changing the notion of undertaking as hitherto regarded by supplementing the presence of ‘control’ with ‘sharing an economic field’. Wow.

It is difficult to overestimate the relevance of this logic. From now on, groups of companies may, at least where their activities substantially differ, be understood to integrate various economic units or undertakings. Worldwide turnovers used in the calculation of sanctions may need to be determined separately for each economic entity. One can not help but to note the divergence this represents with the notion of undertaking as used in the EUMR, where conglomerate groups would remain to be a single ‘undertaking’. It might even be wondered if Article 101 could apply to agreements between separate ‘economic units’ of a conglomerate group, a door that might have appeared to have been closed in Ecoservice not that long ago.

Other questions to be clarified in future cases will look at what standards may be used to tell an activity from another. Sumal has been cautious in presenting this in the context of conglomerate groups where the legal entities act in ‘several economic fields having no connection between them’, That said, it will be interesting to follow what intensity of ‘connection’ is relevant for these purposes.

That said, the answer given by the Court adequately resolves the problem of inverse or downward liability in a consistent way which is firmly anchored on the notion of undertaking. It also provides a hook to resolve the inconsistency resulting from Recital 22 of the Merger Regulation, which appeared to recognise the existence of separate undertakings within public conglomerates, an option arguably unavailable for private ones.

What About Rights of Defence? (paras 51 to 67)

Read the rest of this entry »

The post published on Monday (see here) sought to explain the tensions that tend to emerge in relation to the interpretation of Article 101(1) TFEU. The experience of the past few years shows that the analysis of the objective necessity of a practice and/or of the ancillarity of some clauses are occasionally conflated with the evaluation of the restrictive object of an agreement.

As the ISU judgment shows, it is sometimes claimed that, unless an agreement is objectively necessary to attain a pro-competitive aim, it is restrictive by object. This interpretation of Article 101(1) TFEU is at odds with decades of consistent case law, but it is still relatively commonplace. I was reminded of this understanding of the provision when the German Federal Court of Justice announced its ruling on Most Favoured Nation Clauses (MFNs) earlier this week.

The document of the judgment is not yet available, but a press release explains the key points here. The case concerned the so-called ‘narrow’ MFNs. Even though these clauses have less restrictive potential than the so-called ‘wide’ ones, the German BGH ruled that they are restrictive of competition under Article 101(1) TFEU and do not fulfil the conditions set out in Article 101(3) TFEU.

While we wait for the judgment, there are three salient aspects of the press release that are worth discussing at this stage:

• First, how the case was framed and approached, which is most unusual in contemporary competition law: at least in light of the press release, it looks like the discussion revolved around the application of the ancillary restraints doctrine and of Article 101(3) TFEU.
• Second, the BGH may have departed from the case law of the Court of Justice on a key point: contrary to what the press release appears to suggest, the pro-competitive aspects of an agreement are relevant (and must be considered) under Article 101(1) TFEU.
• Third, the BGH did not refer the case to the Court of Justice for a preliminary ruling. To the extent that the judgment may have departed from the case law on the abovementioned point, it raises the issue of the fragmentation of the EU legal order.

Ancillarity, object and Article 101(3) TFEU

The press release issued by the BGH suggests that the compatibility of narrow MFNs with Article 101 TFEU revolved around two questions: whether the clauses can escape the prohibition in accordance with the ancillary restraints doctrine and whether they fulfil the conditions defined in Article 101(3) TFEU.

There are two key intermediate steps that are nowhere to be found in the press release: first, whether the narrow MFNs are restrictive of competition by object and, second, whether they have, or are likely to have, anticompetitive effects. The judgment jumps (or so it seems from the information available) from ancillarity to Article 101(3) TFEU.

This approach comes across as unusual for a number of reasons. To begin with, the BGH seemingly accepted that ‘narrow’ MFNs may be explained as a means to address free-riding. This point would have been central to the assessment of whether the clauses are restrictive of competition by object. Suffice it to remember that free-riding considerations determined the outcome in, inter alia, Cartes Bancaires.

There are, however, no traces of the question of whether ‘narrow’ MFNs amount to a restriction by object in the press release. There is a chance that, just as in ISU, the ancillary restraints doctrine was conflated with the evaluation of the object of the clauses.

Judgments like Cartes Bancaires and Budapest Bank show that the above are two separate inquiries. More precisely, the fact that the ancillary restraints doctrine does not apply in a given case does not mean that the agreement is necessarily restrictive of competition, let alone by object. It simply means that the assessment under Article 101(1) TFEU has to move to the object and, if needed, the effect stages.

The case law makes it clear that an agreement may escape the ‘by object’ qualification irrespective of whether it is objectively necessary within the meaning of the ancillary restraints doctrine. There was no objective necessity test in Delimitis, Asnef-Equifax or Cartes Bancaires. However, in all these cases, the Court expressly ruled that the object of the agreement was not anticompetitive and that an analysis of its effects was necessary.

The BGH judgment, just like ISU and the CMA decision in Ping (to mention another salient example in which the same error of law was committed), suggests that some ideas about the operation of Article 101(1) TFEU are not easily abandoned, even when they are clearly at odds with the case law. Several factors may potentially explain this reality. In this sense, the judgment suggests that how cases are framed and argued before courts may well contribute to it.

The role of pro-competitive effects under Article 101(1) TFEU

The press release is notable in that it appears to suggest that, according to the BGH, the pro-competitive aspects of the MFN clauses (namely, the fight against free-riding) can only be considered under Article 101(3) TFEU, where they would be weighed against any anticompetitive effects.

On this point (and while we wait for the judgment), it is important to emphasise that, after Generics, there should be no doubt that the pro-competitive aspects of an agreement are relevant under Article 101(1) TFEU. More precisely, the Court made it clear that, as elements of the economic and legal and economic of an agreement, they ‘must’ be considered under the first paragraph of the provision (Generics, para 103).

What is more, the Court expressly ruled that the above should not be construed as meaning that there is such thing as a rule of reason under Article 101(1) TFEU (Generics, para 104). Thus, and contrary to what the press release appears to suggest, there was a role for the pro-competitive aspects of ‘narrow’ MFNs under Article 101(1) TFEU in the case (just as there was in Cartes Bancaires and Budapest Bank) and that such role does not necessitate any balancing assessment.

Suffice it to remember, in this regard, the centrality of the pro-competitive aspects of the agreement in Budapest Bank. The fact that the agreement is capable of leading to pro-competitive or at least ambivalent gains is, the Court confimed, a key consideration to establish a restriction of competition under Article 101(1) TFEU (Budapest Bank, paras 82-83).

The risk of legal fragmentation and the role of the Court of Justice

Regulation 1/2003 sought the decentralisation of EU competition law. The flurry of enforcement across the EU attests to the resounding success of this landmark piece of legislation. At the same time, some developments show that with decentralisation comes (inevitably) a real risk of legal fragmentation.

The risk of the fragmentation of the EU legal order originates from two main sources. On the one hand (the BGH judgment hints at this) some interpretations of Articles 101 and 102 TFEU are not abandoned, even when they are at odds with a plain reading of the case law.

There are, on the other hand, developments at the national level that are exploring (when not stretching) the outer boundaries of the EU legal order. Just last week, for instance, the Italian Competition Authority adopted a decision in a dispute involving Enel and Google that substantially expands the ‘exceptional circumstances’ case law (see here) and ventures into uncharted territory.

It was always clear that, in a decentralised model, the Court of Justice would play a central role in preserving the uniformity of the EU legal order. The evolution of the enforcement regime suggests that, as things stand, it will be called upon to intervene more frequently that some might have anticipated in the wake of the adoption of Regulation 1/2003.

The question of the legal basis might be the single most important legal and political issue when it comes to the design and adoption of the future Digital Markets Act (as anticipated here). Yet, it has been largely overlooked by external observers.

The desire to come up with a new tool that would enable enforcers to operate free from the constraints imposed by the competition case law may have obscured the fact that EU legislators also operate under legal constraints. These constraints are not the bread-and-butter of most competition lawyers (much less economists) so it is only understandable that many commentators might have missed it. Unsurprisingly, these legal concerns have not escaped the Commission’s internal reviewers.

Press reports (see here) suggest that the Commission’s proposal may have been delayed and somewhat watered down due to internal “legal basis concerns”. These were to be expected. In my view, it is impossible to properly understand or discuss the upcoming DMA proposal absent some context about the legal basis debate.

This should hopefully help you understand what is happening, and anticipate what might happen going forward:

What is the issue?

The EU can only legislate within the confines of the powers conferred upon it by the Treaties. Each legislative proposal needs to have a “legal basis” in the Treaty. Different Treaty provisions provide the legal basis for legislative action in each area. The choice of the legal basis will determine the legislative procedure (including who acts as legislator) and the possible scope of the act.

In the field of competition law, Art. 103 TFEU enables the Council, on a Commission proposal, to legislate in order to “give effect to the principles set out in Articles 101 and 102”. That legal basis, however, would not allow the EU to go beyond Arts. 101 and 102, which was precisely the idea here. So this legal basis would, at least in itself, be insufficient.

The Treaty contains a specific Protocol (Protocol 27, on the internal market and competition) indicating that, should the EU need new powers to protect competition in the internal market, it can resort to Art. 352 TFEU. This is a legal basis envisaged precisely for situations like the one at issue here. It was actually the legal basis used for the adoption of the merger regulation, which was also adopted to fill gaps in Arts. 101 and 102 TFEU. Art. 352 TFEU would, evidently, appear to be the appropriate legal basis for the creation of new and far-reaching powers. BUT, it requires unanimity among Member States, and it would not enable the Parliament to act as co-legislator. Either to bypass the unanimity requirement or because the Commission wants to involve the Parliament, the Commission prefers an alternative legal basis: Art. 114 TFEU. It seems clear that this will be the legal basis of the proposed Regulation.

What the Commission can (and cannot) do under Article 114 TFEU

Art. 114 TFEU enables the EU to adopt “measures for the approximation of the provisions laid down by law, regulation or administrative action in Member States which have as their object the establishment and functioning of the internal market”. Under Art. 114 TFEU, the proposal would be dealt with under the ordinary legislative procedure (with Council and Parliament acting as co-legislators), and it would not require unanimity among Member States. The choice of this legal basis nonetheless carries some constraints.

There is plenty of EU case law on Art. 114. This case law makes clear that measures which do not harmonize national rules (e.g. because they aim at introducing new legal forms and/or leave unchanged the different national laws in existence) cannot be adopted under Art. 114 TFEU. And even when reliance on Art. 114 as a legal basis is justified, any legislative measure should comply with the principle of proportionality (i.e. the measure must be appropriate for attaining the objective pursued and not go beyond what is necessary to achieve it).

The Commission is obviously aware of this (awareness may be particularly acute within the Legal Service and the Regulatory Scrutiny Board). To rely on Art. 114 TFEU as the legal basis and avoid the unanimity requirement, the Commission needs to argue that the proposed legislation’s primary aim is not about protecting competition beyond Arts 101 and 102 (which was widely assumed to be the goal of the proposal, as also suggested by the digital reports and the myriad webinars on this new legislation) but to harmonize national rules. In recent weeks the Commission has made all possible efforts to emphasize this message (for two examples, see here and here; there are many more).

For this to be correct, the proposal would need to (i) show a risk of likely regulatory fragmentation liable to distort competition; (ii) be designed to prevent regulatory fragmentation within the internal market; and (iii) not go beyond what is necessary to attain that goal.

What to look for in the proposal

At this stage it is unclear how the adoption of the DMA would prevent Member States from adopting or maintaining in force divergent national rules (e.g. German plans to amend national competition rules). In fact, some competition authorities such as the CNMC have warned (see here) that the new EU rules themselves could cause, rather than prevent, regulatory fragmentation. To validly rely on Art. 114 TFEU, the DMA would need to set limits for Member States to deviate from the DMA’s provisions. Keep an eye peeled for that.

The Commission has some discretion to choose the specific harmonization method, and this may include preventive measures and even imposing obligations upon private parties. To validly rely on Art. 114 TFEU, however, the proposal would also need to ensure that the eventual provisions of the DMA do not go beyond what is necessary to address appreciable distortions arising from regulatory fragmentation. The proportionality analysis may be particularly relevant in relation to both the definition of the DMA’s scope and the articulation of the more far-reaching measures proposed (e.g. “blacklisted” practices). The Nordic competition authorities and the CNMC, among others, have also expressed doubts about the proportionality of such an approach, so it will be interesting to see how this will be framed in the upcoming proposal.

What can the European Parliament do?

The Parliament is arguably the Institution with the greatest interest in operating under Art. 114 TFEU to preserve its role as co-legislator (under Art. 352 TFEU the Parliament would only be able to give or deny its consent to the Proposal). The Parliament, understandably, has a history of pushing for Art. 114 TFEU in the face of doubt. The need to remain within the confines of this provision might rule out the most far-reaching amendments that some may have had in mind. If anything, the Parliament would need to ensure the proportionality of the Act by reference to its stated aim.

And then what?

Until the proposal is out and these details (among others) are clear, it would not be prudent to express an opinion about the chances that it might withstand Court scrutiny, if it ever comes that.

[DISCLAIMER: As explained here, I advise a number of companies that could be affected by the new rules. The views in this post are exclusively mine and while they may coincide with the advice I have provided to clients, they do not in any way reflect any client’s views].

I have just uploaded on ssrn (see here) a paper discussing the Court of Justice’s ruling in Generics (Paroxetine). It is a case note commissioned by the Common Market Law Review earlier this year. Hopefully it provides a clear and useful overview of the judgment and its implications. There is also a cameo appearance of Budapest Bank, which was issued shortly afterwards and could not be left out of the analysis.

Before I say anything about the piece: I am really grateful to Fernando Castillo de la Torre (Legal Service), who shared his comments on a previous version and definitely improved the end product. Thanks so much again!

I have written several (almost certainly too many) posts on a legal issue which I continue to find fascinating. But I tell myself that it is worth emphasising the points that follow.

Pay-for-delay and minimisation of Type I and Type II errors

The more I think about it, the more I tell myself that Generics (Paroxetine) achieved an optimal balance to minimise the risk of both Type I and Type II errors.

It is true that the Court defines the notion of potential competition in a relatively broad manner. It is sufficient to show that there are ‘real concrete possibilites’ of entry (as opposed to the likelihood of entry). What is more, potential competition can be established by proxy, on the basis of indirect indicia.

On the other hand, the Court is careful to point out that pay-for-delay agreements are not always restrictive by object, and that it is necessary to consider the specific circumstances of each case to come to conclusions about whether a given settlement is prima facie unlawful irrespective of its effects.

Intellectual property and insurmountable barriers to entry

Part of the interest of the preliminary reference in the case came from the fact that the UK Competition Appeal Tribunal openly invited the Court to think in probabilistic terms about intellectual property titles. Instead of seeing them as presumptively valid exclusive rights, the tribunal suggested that the probability of them being declared invalid could be incorporated in the assessment.

The Court did not follow the path suggested or implied in the reference. This is not surprising, considering that there is a consistent line of case law, dating back to the very early days of Consten-Grundig, according to which EU law does not question the existence of intellectual property rights, but only their exercise.

What I found interesting (but also not surprising) is that the Court declares that, in the specific circumstances of the case, intellectual property rights were not deemed an ‘insurmountable barrier to entry’.

The question for future rulings is that of when intellectual property will be deemed an ‘insurmountable’ barrier. The Court suggests it is an issue to be decided on a case-by-case basis in light of the relevant economic and legal context.

On restrictions by object

On the notion of restriction of competition by object, I have already explained that Generics (Paroxetine) will be remembered as a seminal ruling, together with Budapest Bank.

The two confirm that the identification of the object of an agreement is a context-specific inquiry. In addition, they (in particular Generics) made it explicit that the pro-competitive aspects of a practices are part of this assessment.

Another point that I have addressed at length is the counterfactual: in the two judgments, the conditions of competition that would have prevailed in the absence of the practice play a role in the evaluation of the object of the agreement. Insofar as this is the case, it seems difficult to argue that the counterfactual is merely confined to the assessment of effects.

The analysis of effects under Article 102 TFEU

Finally, Generics (Paroxetine) is particularly valuable in the contributions it makes to the clarification of the notion of anticompetitive effects under Article 102 TFEU.

In this regard, the Court confirms that the evaluation of this question under Article 102 TFEU is not different from that undertaken under Article 101 TFEU or indeed merger control (which is not only welcome but natural).

Accordingly, it is necessary to assess effects by reference to the market as a whole. By the same token, the impact of the practice would need to go beyond that consequences it has for the freedom of action of individual undertakings.

In reality, the Court’s approach is not any different from that laid down in Delimitis (an Article 101 TFEU case) and sketched in Post Danmark II and Intel (where the Court placed an emphasis on the coverage of the practice as a factor).

Following two of my latest posts – a though experiment on smartphones and cameras, and an update on Slovak Telekom[1]– some of you have contacted me about the central argument I develop in them.

My main point in both is that the remedy is not an afterthought that is irrelevant when establishing an infringement. The remedy – or, more precisely, what a finding of liability would entail – is central to determine whether there is a breach in the first place.

This statement is true as a matter of positive law – Van den Bergh Foods encapsulates the essence of the case law – and is also true from a normative standpoint – as the post on smartphones and cameras sought to explain.

The messages I have received following these posts are certainly sensible. How can the remedy determine the applicable legal test? Is it not getting the case backwards? Should we not establish an infringement first and then figure out the way to remedy it?

These questions, no doubt reasonable, are based on a fundamental assumption, which is that the finding of an infringement and a remedy are two separate steps, independent of one another.

I do not believe this assumption reflects the reality of the interaction between legal tests and remedies. More importantly, I do not believe the Court treats the infringement and the remedy as separate steps. They are rather two sides of the same coin.

The remedy reliably tells us what a case is about, and what the legal test is

A finding of infringement and a remedy are so closely intertwined that the single most reliable way to tell whether there is a competition law breach is to inquire about how to bring the alleged breach to an end.

In other words: to understand what a firm has done (and whether what it has done is prohibited), the best approach is to ask the claimant what the alleged infringer would need to do to comply with the law.

This point can be illustrated by reference to refusal to deal scenarios. These cases are not always easy to spot, and the lines are sometimes blurred between them and other legal categories, including tying and ‘margin squeeze’.

How to tell apart one category from the other? Inquire what the endgame would be. If the endgame is one in which a court or authority mandates (directly or indirectly, implicitly or explicitly) shared access to an input or an infrastructure, then the case raises the issues that are typical of refusal to deal cases. For the same reason, indispensability would be an element of the legal test.

Suppose, conversely, the remedy is a different one. For instance, an obligation on the firm not to require exclusivity from its customers. If so, we would be in Hoffmann-La Roche and Intel territory (where nobody has ever thought of making indispensability an element of the legal test).

These statements are hardly revolutionary. It is, in essence, what the Court held in Van den Bergh Foods and then in TeliaSonera. In Van den Bergh Foods, the firm argued that indispensability was an element of the legal test. Both the General Court and the Court of Justice, in light of the endgame, concluded that the case belonged in the Hoffmann-La Roche/Intel family, and rejected the argument.

The remedy is routinely considered when calibrating the legal test

The history of competition law also teaches us that the remedy (that is, what the case would involve in practice) has been routinely considered when designing the appropriate legal test.

In particular, courts tend to craft strict legal tests where a finding of infringement would demand the administration of a regulatory-like remedy (such as mandating access obligations and setting the terms and conditions under which access is to be granted).

Again, this idea is hardly revolutionary, and is part of our law. Why is indispensability part of the legal test in refusal to deal cases? Magill and Bronner seem to reflect a deep awareness of the implications of a finding of infringement.

Think of Magill. Again, the question is: what is the endgame of ruling that a refusal to license copyright-protected information is abusive? Imposing a duty to license on the right holder. By definition, there is a tension between such an endgame and the logic of intellectual property systems (the very point of which is the right to refuse to license).

It is not a surprise, against this background, that the Court confined to ‘exceptional circumstances’ the instances in which Article 102 TFEU can be relied upon to impose a duty to license (thereby minimising the tension with intellectual property systems).

The uncertainties and complexities of administering a remedy have also been mentioned as factors to consider when crafting a legal test. In his celebrated article on essential facilities, Phillip Areeda – the ultimate centrist in our field – expressed the view that ‘[n]o court should impose a duty to deal that it cannot explain or adequately and reasonably supervise’ and went on to warn against the imposition of regulatory-like obligations.

The EU experience shows that competition authorities do not fare much better than courts when it comes to the administration of these proactive remedies. The failure of the media player remedy in Microsoft I often comes up in these discussions. The principles-based approach to the question in Google Shopping and Android, and the resulting state of limbo (which appears to continue), another one.

Big Tech and legal tests: the question that will not go away

The discussion above gives, hopefully, a sense of what is going on, from a legal standpoint, in relation to Big Tech. Many of these cases demand regulatory-like intervention, and this, more frequently and more intensely than in the past.

It would seem that some competition authorities are more confident than they used to about their ability to redesign products and to tweak business models (other authorities, like the CMA, are more cautious, as the Final Report on digital advertising shows).

This increased confidence tends to trickle down into the interpretation of competition law provisions. The link between the nature of the remedy and the legal test, which acted as a limit on intervention, is now questioned (with the institutional implications that follow, in particular in a decentralised system). Fascinating, and potentially transformative, times.

---

[1] Speaking of which, AG Saugmandsgaard Øe’s Opinion is out on 9 September.

The CK Telecoms (Three/O2) judgment has revived the interest in issues relating to the standard of proof in EU merger control (and competition law at large). It adds valuable insights on the question. As I read it for the first time, I was, yet again, reminded that there is some tendency by commentators to conflate two separate questions: (i) the standard of proof and (ii) the substantive legal test (and more precisely the applicable threshold of effects – capability, likelihood, certainty).

Our friend Andriani Kalintiri (King’s College London) knows these issues inside out, which is why I immediately thought of inviting her to present them in a post. The judgment, by the way, is very much in line with her own take on the standard of proof.

I leave you with her analysis, which hopefully will complement mine. If hungry for more, take a look at her superb book, from which I have learnt a great deal (on evidence, presumptions and judicial review). It is particularly timely now that discussions on presumptions and the burden of proof are all the rage. Thanks again, Andriani!

There is only one way to start this post. First, with a heartfelt ‘thank you’ to Pablo and Alfonso for the kind invitation! And second, with a confession: that the ‘misconception’ this note hopes to illuminate and caution against – i.e. the tendency to conflate substantive legal tests and standards of proof – has been the cause of many headaches during my research! Indeed, in the EU Courts’ antitrust and merger judgments, one often comes across various references to the ‘plausible’, ‘potential’, ‘probable’, ‘likely’ or ‘actual’ effects that a practice or arrangement must give rise to, in order to be prohibited. Are these references, I was wondering, indications of the applicable ‘standard of proof’ or of something else, and why?

BACK TO THE BASICS

To answer this, let me begin with a simple yet important reminder: competition enforcement takes place under conditions of ignorance and uncertainty (as does law enforcement in general). In an ideal world, courts, authorities, and businesses would possess perfect information and would always make correct decisions about the proper meaning of the antitrust and merger rules and their application. In real life, however, information is incomplete and resources, time and our cognitive capacity are limited. Yet, decisions must still be made, and the rules must be enforced.

Among other mechanisms, legal tests and standards of proof are what allows ‘the show to go on’. On the one hand, the legal tests developed by the EU Courts specify the scope of the vaguely worded antitrust and merger rules by setting out the conditions that must be satisfied for a practice to be prohibited – for instance, where the conduct has caused harm or where there is a stronger or lesser chance of such harm occurring. On the other hand, the standard of proof indicates the threshold, falling short of certainty, that the party with the burden of persuasion must surpass for the evidence to be accepted as proof of an allegation.

SO SIMILAR, YET SO DIFFERENT

­In practice, however, these two thresholds are commonly conflated – for instance, it is sometimes said that different ‘standards of proof’ apply to different types of unilateral conduct. Based on what I have said, a key – albeit not the only – explanation for this tendency may be already obvious: that when we think about competition legal tests and standards of proof, we consciously or subconsciously employ probabilistic language – i.e. we use words such as ‘plausible’, ‘potential’, ‘probable’, ‘likely’ and so on, in relation to both of them. For instance, we may say that a practice will be prohibited when it is ‘probable’ or ‘likely’ to foreclose competition. Or that an allegation is proved when it is ‘more likely than not’.

Despite any illusion of unity though, there are important differences between the two. For one, they serve distinct functions in that they answer different questions. The key question competition legal tests are ultimately preoccupied with is: ‘what level of harm is required for a conduct to be deemed unlawful?’ On the other hand, standards of proof address a different issue – i.e. ‘what level of evidence is sufficient for an allegation to be accepted as true in the eyes of the law?’

Most importantly, the considerations that inform their design – i.e. what the substantive legal test and the standard of proof should be – are different, too. As a starting point, both aim to minimise false convictions and false acquittals since either error entails costs. While, however, these costs are not entirely disassociated (at least in the longer term), they are not identical either – and nor are the factors that underpin their balancing. This discussion is particularly complex, but in rough terms the following may be noted.

As far as ‘erroneous’ legal tests are concerned, their primary costs are chilling procompetitive behaviour and encouraging anticompetitive conduct (including in both instances, the respective harm to competition and consumers). Their balancing in designing the ‘optimal’ legal test – and threshold of effects – largely depends on the nature of the conduct in question in the light of current knowledge about it, and the cost of enforcement as well as business compliance. This explains why the substantive legal test varies practice to practice – cartels, for instance, are not treated in the same way as refusals to supply, for which the required threshold of effects is much higher, given the implications for firms’ incentives to innovate.

By contrast, the primary costs of an ‘erroneous’ standard of proof are the undue interference with the freedom and rights of the defendant and of harm to society. Their balancing in designing the ‘optimal’ standard of proof largely depends on fairness considerations linked to the seriousness of the consequences involved for the person concerned and the specific features of the enforcement model, including any inequality of arms. This explains why the standard of proof is ‘static’ – the same standard of proof (i.e. ‘firm conviction’) applies to cartels and refusals to supply, and rightly so, since the consequences for the undertakings involved and the nature of the enforcement proceedings are the same.

MUCH ADO ABOUT NOTHING?

This discussion is not purely academic for several reasons, but I will confine myself to two.

First, conflating substantive legal tests and standards of proof may lead to confused discussions about what these are or should be. Two examples illustrate this risk. On the one hand, recent reports on digital competition policy suggested, among others, modifications in the burden and standard of proof; more accurately though, these proposals arguably concern the substantive legal test, and should be assessed as such.[1] On the other hand, in merger control it is often argued that the applicable standard of proof is – and should be – the ‘balance of probabilities’ due to the absence of fines and the prospective nature of the analysis. As I have explained elsewhere though, the case law suggests – and rightly so, in my opinion – a much higher standard of proof. The recent judgment in CK Telecoms seems to confirm this: as the General Court noted, ‘the standard of proof (…) is therefore stricter than that under which a significant impediment to effective competition is “more likely than not”, on the basis of a “balance of probabilities” (…)’, although it is ‘it is less strict than a standard of proof based on “being beyond all reasonable doubt”’ (para 118).

Second, substantive legal tests and standards of proof are equally important to the correctness and legitimacy of enforcement. Indeed, ‘correct’ legal tests may not compensate for shortcomings in the design of the standard of proof. Think, for instance, of cartels – the fact that they are rightly subject to a ‘by object’ prohibition could not ‘save the day’, if the standard of proof were too low – say, equivalent to the balance of probabilities (or similar threshold of belief) given the operation of the presumption of innocence. The opposite is also true: ‘high’ standards of proof may not rectify inappropriately formulated substantive legal tests. Think, for instance, of Google Shopping Service: even if the Commission has successfully established a ‘firm conviction’ that the elements of the substantive legal test – as it understands it – are met in the light of the evidence, this will not mean much, if the latter is eventually found to be incorrect.

BACK TO THE FUTURE

Ultimately, it is important to appreciate that substantive legal tests and standards of proof are different rules. Of course, the stricter the substantive legal test, the more difficult it will be to discharge the standard of proof – and vice versa, the stricter the standard of proof, the more difficult it will be to prove in specific cases the constituent elements of the substantive legal test. However, they remain distinct yet equally important for an authority’s decision to be not only lawful but also legitimate. In this regard, the discussion is not about semantics – rather, words such as ‘plausible’, ‘potential’, ‘possible’, ‘probable’, ‘likely’ and so on, may lead to very different perceptions of the applicable threshold of effects or standard of proof, depending on what they actually refer to, and we should thus use them more consciously and accurately to avoid confusion and misunderstandings.

And a final remark, for the avoidance of any doubt: I have no interest to disclose – other than a deep academic fascination with the topic!

---

[1] For example, in the Report on Competition Policy for the Digital Era, it is noted (p 4) that ‘(…) competition law should try to translate general insights about error costs into legal tests. The specific characteristics of many digital markets have arguably changed the balance of error cost and implementation costs, such that some modifications of the established tests, including allocation of the burden of proof and definition of the standard of proof, may be called for.’

Online platforms are all over the news these days. Today we have learnt that the European Commission has opened a formal probe into Amazon’s practices. I could write a post, but I have the sense that the one I prepared on the preliminary investigation a while ago still captures well my thoughts on the case.

I would simply add a question for discussion: is the data supplied to Amazon by merchants not just an element of the cost of doing business via the marketplace (i.e. part of the price they pay to access the platform)? If so: is this aspect of the case not simply about alleged excessive pricing (with all the consequences and implications that follow)?

More to the point of today’s post, large tech companies have made the headlines on both sides of the Atlantic following the French Senate’s green light to the so-called Digital Services Tax (which, it seems, is known as ‘taxe Gafa’ in France).

I have been asked to share my views on the State aid dimension of this tax (a topic that Alfonso covered on the blog). Last week I was interviewed by Bloomberg Tax (see here), together with Alfonso’s partner José Luis Buendía Sierra.

On Monday of this week, I took part in a radio programme (Knowledge@Wharton), run by The Wharton School at Penn and offered via SiriusXM Radio. I was interviewed together with Ruth Mason (University of Virginia) and Andrea Matwyshyn (Penn State). It was fascinating, no less because this case has sparked an academic interest in EU State Aid Law in the US. You can access the interview here.

What are my views on the State aid dimension of the Digital Service Tax (or ‘taxe Gafa’)? From a substantive perspective, I believe it is difficult to argue that this tax is not vulnerable to challenge on State aid grounds.

If one pays attention to recent administrative practice, the most reasonable conclusion is that it is more likely than not that the Commission would conclude that it is caught by Article 107(1) TFEU. In this sense, the French Senate’s position, which insisted on the notification of the measure, comes across as sensible and prudent (and in the spirit of Articles 107 and 108 TFEU).

Does it mean that the measure is necessarily State aid? No. Even though it appears to be explicitly targeted at some firms (it is known as ‘taxe Gafa’ for a reason), one could try and make the argument that the targets are not in a comparable factual and legal situation as everybody else.

Perhaps. But there is, at present, no operational test to determine whether undertakings are in a comparable factual and legal situation. This is one of the fundamental issues I emphasised in my presentation at our State aid workshop a month ago. I struggle to see why and when two groups of undertakings are likely to be deemed in a comparable situation, and when they are not.

Against this background, my sense is that it would be desirable to get some guidance from the Court on this point. And perhaps this case (together with the pending disputes on the Polish Retail Tax and the Hungarian Advertising Tax) provide the ideal context to do so.

We have always cautioned against the application of different rules and standards only to certain companies. That is a trend that appears to be in vogue these days. In some cases, this approach may threaten the consistency of the law and breach general principles. In other cases, the application of ad hoc rules would be in direct breach of EU Law itself.

Here is a perfect example that affects two areas that are of particular interest to the European Commission and to myself: tech and fiscal state aid. [Disclosure: I have examined these issues following a request from our friends at CCIA, who will also post this piece on their own blog. Below is my reasoning, you can make your own conclusion]

You have probably read on the news about the Commission plans to create a “digital service tax” (“DST”). This is a plan that was reportedly abandoned recently due to the opposition of some Member States (see here). Absent a harmonized EU DST, some Member States (namely France, Italy and Spain) have adopted or announced the adoption of their own unilateral DSTs. The declared objective of these taxes is to ensure that the provision of digital services is taxed in the jurisdictions where users contribute significantly to the process of value creation.

This is not the place to discuss the political opportunity or possible consequences of these initiatives nor to challenge wider proposals to reform tax systems (which may make sense), so let’s assume that these are legitimate policy initiatives. We will focus not on the idea, but on its execution.

As currently designed, it is obvious to me that the different national DST proposals would involve the granting of State aid (which doesn’t necessarily mean they are illegal, only that they need to be notified by the Commission for clearance).  Think about it:

First, the case law makes clear that a specific tax addressed only to certain undertakings may imply a selective advantage to other companies not subject to the tax. The key question is whether all the companies in the same legal and factual situation, having regard to the objective of the system, are treated in the same way. [See e.g. the Court’s recent Judgment in ANGED, where the CJEU considered that regional environmental taxes that applied to supermarkets but exempted commercial malls despite having a similar negative impact on the environment constituted a selective advantage to malls].

This means that one needs to verify whether the scope of the tax at issue is consistent with its declared objective or whether, on the contrary, it exempts other companies that should have been subject to the tax.

In the case of the DSTs, the current proposals do not target all activities where users create value, but only some, namely online advertising services, online intermediation services and data transfer services. This excludes any other services where users may also create value.

Second, the scope of the proposals is further delineated in the light of a high worldwide revenue threshold (both the French, Spanish and Italian proposals set a 750 million worldwide revenue threshold). But the case law also makes clear that crafting objective thresholds in such a way that de facto exempts domestic companies from the application of a tax can also amount to State aid.

The clearest examples are the recent Commission decisions regarding Hungarian and Polish turnover taxes on certain activities, now before the CJEU. In three separate decisions (here, here and here), the Commission considered that these taxes constituted State aid as they were based on particularly high turnover thresholds and were effectively designed to tax almost exclusively large companies (usually foreign companies) and not smaller ones (almost always domestic companies), in a way that was unrelated to the objectives sought by the Member States. [Evidently, the State aid nature of a measure does not depend on whether it is granted by Hungary and Poland or by France, Spain and Italy…]

Third, revenue-based taxes do not appear to be consistent with the declared objective of taxing digital services in the place where users contribute to the process of value creation. This is because users can create value in many ways regardless of the company’s turnover. Actually, this is precisely the reason why many people would like to see a change in merger notification thresholds only based on turnover.

Public statements from national politicians saying that SMEs and national companies have nothing to fear (see e.g. here, here –the French Government actually calls this the “GAFA tax” – or here) would appear to confirm that these initiatives have been crafted with addressees in mind, not just to tax digital services where users contribute to the process of value creation (a tax with a company nickname sounds a tiny bit selective…). This is a sort of Marxist (I mean Groucho) view of fiscal policy: “these are my principles, but to some companies we’ll apply others”. This is not just me saying it: the Danish, Swedish and Finish governments (not known for being soft on taxes) have also spotted the problem and opposed even the EU proposal (see here).

By treating differently companies that are in a comparable situation, unilateral measures such as the French, Italian or Spanish DSTs would, in my view, constitute clear State aid. It’s therefore surprising that none of these pro-European Member States have notified the projects so that the Commission can assess their compatibility.

In the light of the Polish and Hungarian precedents, one could anticipate that –regardless of its political views- the Commission will in any event ultimately intervene, or be forced to intervene.

To nobody’s surprise, the Commission has announced, today, the adoption of a decision concerning Android. The stakes in the case are so high that the outcome was known well in advance. The only open question related to the amount of the fine. But even then, it looked like a given that it would be the highest ever. Which turned out to be correct.

When the stakes are so high, corporate strategies tend to dominate the landscape and the discussions.

Some aspects of these strategies are beyond reproach. It makes sense for companies to hire the very best lawyers – a category that definitely includes Alfonso, who has always been open about his involvement in this case.

Alfonso, by the way, will not be blogging again on Android, no less because he is likely to have very little time in the coming months (bye bye summer holidays and bye bye paternity leave, I guess).

I regret other aspects of these corporate methods. Companies tend to make loads of noise when their interests are at stake. Big Food, for instance, has perfected a strategy of confusion that gives many people the impression that beef and cheese are perhaps healthy after all (they are not).

There has certainly been a lot of noise recently in the competition law community – coming from all over the place.

The ongoing cacophony is a real pity, since there is a lot of exciting legal stuff beneath the noise. And it is worth discussing it.

For those interested in the law bit of competition law and policy, here are some thoughts.

According to the press release, the Android decision finds that three main practices amount to an abuse of a dominant position:

• The tying of Google Play to other applications: Google does not license its applications a la carte. In particular, the Commission takes issue with the fact that Google Play Store is not available as a stand-alone product. For a comparable practice, think of pay TV providers preventing users from cherry-picking the channels to which they subscribe.
• Compensation for exclusive pre-installation: According to the press release, mobile phone manufacturers are given financial incentives for exclusively pre-installing Google Search. This arrangement makes me think of a supermarket chain receiving compensation as consideration for prime shelf space – or perhaps an online store receiving compensation for placing some products as default choices.
• Android Forks: If mobile phone manufacturers choose to offer the Google version of Android, they may not offer rival versions of Android. It is like McDonald’s requiring its franchisees not to run their own burger joints (or a Burger King restaurant) in parallel.

The second of these practices is perhaps the least exciting one. The only intriguing question is perhaps whether the exclusive pre-installation amounts to an exclusivity obligation a la Hoffmann-La Roche (which is what the press release seems to imply) or to a practice falling elsewhere along the spectrum of schemes that can have a fidelity-building effect. Either way, the legal framework is firmly in place after Post Danmark II and Intel.

The two other practices, on the other hand, raise more fundamental issues. So much so, in fact, that this case may mark the evolution of EU competition law. Allow me to explain.

Tying: how products are sold vs how products are made

Some people will argue that the application of Article 102 TFEU in relation to the tying aspect of the decision cannot surprise anyone. And it is a reasonable point to make. After Microsoft I, any tie-in that gives a distribution advantage to the dominant firm’s tied product amounts to an abuse – which is another way of saying that tying is presumptively abusive under Article 102 TFEU.

What can be exciting around this case, then? Well, the fact that, in some respects, the setting is different from that found in traditional tying scenarios. Inevitably, the remedy is also different.

Traditional competition law in general, and tying in particular, typically interferes with how products are sold. By the same token, competition law is generally wary not to second-guess how products are made.

What do I mean by this distinction? I mean that, absent exceptional circumstances, competition law is not there to tell companies how to run their business. The point of competition law is to ensure that companies have the ability and the incentive to thrive in the marketplace using the strategies of their choosing.

Competition law is agnostic about whether companies vertically integrate or sell their products through third parties, whether they choose selective distribution over franchising or (more to the point) make money through advertising (like a free-to-air TV channel) or through subscriptions (like HBO). For the same reasons, authorities dislike telling firms what prices they should charge.

How is Android different from traditional tying cases?

It is obvious to everyone why Play is tied to Search (and why Search is given a distribution advantage). It is through this mechanism that a company like Google makes money. Thus, if the tying of content and advertising is made unlawful, Google will have to find new ways to make money – or perhaps reinvent Android as a non-profit entity.

The remedy in the case is likely to lead to a fundamental rethink of Android. This issue does not arise in traditional tying cases. If a firm like Coca-Cola is not allowed to engage in tying, it can carry on making money the same way it used to. Not even Microsoft had to change its business strategy – this said, the Media Player remedy failed, which is not an unimportant factor in this context.

To sum up: Android will inevitably lead to more intrusive intervention than usual. And the potential unintended consequences of second-guessing firms’ strategies are universally acknowledged in the competition law community (and have often informed legal analysis).

Against this background, the open question, I guess, is whether, and to what extent, this difference should be reflected in the law.

Is this factor irrelevant from a legal standpoint? If it is not irrelevant, how does (or should) the law adapt to the increased intrusiveness? What are the closest precedents at which the remedy hints?

The reach and scope of competition law intervention may vary significantly depending on how these questions are answered.

Android forks and the legal status of non-compete obligations

What I say above can also be extended to the issue of Android forks. As explained above, obligations relating to this matter are like the sort of non-compete obligations found in franchising agreements or in those seeking to protect the goodwill around a business (think of Remia). From this perspective, one could argue that they are reasonable.

Is it not sensible for a company to prevent free-riding and to make sure that it does not create competition to itself when licensing its products and services? One could point to the Guidelines on technology transfer agreements to suggest that, indeed, it is. In relation to these agreements, Valentine Korah consistently emphasised that competition cannot be examined from an ex post perspective alone.

One could also argue, equally reasonably, that dominant firms have a special responsibility. I agree that they do. This point, however, does not say anything about the relevant legal test. Under what conditions are dominant firms precluded from taking measures against free-riding? Can they avoid creating competition to themselves when licensing their products? How are these considerations integrated in the legal framework?

These are questions, again, for which there is no clear-cut answer in the case law – Article 102 TFEU case law, that is. In that sense, Android looks like a good opportunity to evaluate and clarify the status of these business strategies – and similar ones raising the same issues.

Conclusions

These are not the only questions in which I am interested. But it gives you an idea of the sort of major points to which I will jump when the decision is made available. The telecoms lawyer in me is also intrigued by some aspects of market definition mentioned in the press release (in particular the reference to indirect constraints). And I am curious to know how the notion of effects is defined.

If you are interested in making sense of the law (as opposed to making noise) too, I would very much welcome your thoughts.