Chillin'Competition

Background

Earlier this week the General Court delivered its ruling in Valve (see here). It addressed some of the legal issues that the Court of Justice did not, and could not, tackle in Canal+ (and which I discussed here). Because the latter case was decided by means of a commitments decision, the ECJ could not ‘definitively’ rule on whether the contentious clauses restrict competition by object (see para 54 of Canal+).

Valve, like Canal+, is about restrictions to the cross-border provision of copyright-protected content within the EU. Both differ from plain vanilla market integration cases in two ways. First, intellectual property is exploited in an intangible manner (as opposed to incorporated in, say, a Hello Kitty t-shirt).

Second, they are not subject to the exhaustion doctrine. As a result (and as the Court held in Coditel I), a copyright holder (or its licensee) may rely on its exclusive rights to preclude the cross-border provision of content.

These two differences can be consequential in individual cases. Because the copyright holder (or its exclusive licensee) can exercise its rights in relation to every communication to the public, cross-border trade may be prevented by the intellectual property system itself (that is, irrespective of whether there is an agreement in place).

In other words (and to use the vocabulary employed in Generics), the intellectual property system may, in some circumstances, represent an ‘insurmountable barrier to entry’ preventing the cross-border provision of services.

We know from the case law (think of Toshiba and E.On Ruhrgas) that, where the absence of (inter or intra-brand) competition is attributable to the regulatory context (as opposed to firms’ behaviour), the agreement under consideration does not restrict competition, whether by object or effect (this case law is elegantly summarised in Advocate General Kokott’s Opinion in Generics).

The (seemingly) straightforward

These precedents suggest that the key question to raise in Valve is whether there were ‘insurmountable barriers’ to the cross-border offer of video games via the Steam platform.

While the outcome of individual cases is never the focus of my analysis, I will note that, as far as I could gather from the public version of the decision (many crucial bits of which were confidential), Valve had received a worldwide non-exclusive licence to distribute the relevant video games via its platform (and therefore, it would appear, the right to distribute the said video games across the whole of the EU).

In such circumstances, it seems very likely (this is at least what the public version of the decision suggests) that any restrictions to cross-border trade would be attributable to the agreements, not to the copyright regime. For the same reason, the qualification of the said agreements as restrictive of competition by object seems in principle inescapable.

Assuming that the agreements (as opposed the copyright regime) are what preclude the cross-border provision of content, there would be no plausible rationale for the restraints other than the limitation of passive sales. This is, almost word for word, what the Commission argued in the case (para 211).

Summing up, the case seems, on the surface, straightforward. Some of the points made about the relationship between competition law and intellectual property, moreover, are wholly uncontroversial (such as the fact that the non-exhaustion of the rights does not preclude, as held in Coditel II, the application of Article 101(1) TFEU).

The open questions: a new relationship with intellectual property?

Other points, however, appear to go further than exisiting case law. They may perhaps be irrelevant in the context of the case, but they are definitely worth discussing. In para 203 of the judgment, the General Court suggests that a restriction of passive selling would be restrictive by object unless and until a national court declares that such passive selling amounts to a copyright infringement:

‘203. […] as is apparent from recitals 305 to 309 and 352 to 354 of the contested decision, the question whether or not […] passive sales by the publishers’ distributors could be the subject of legal proceedings before the competent national courts is not decisive for the purposes of the application of Article 101 TFEU […]’.

In other words, competition law, according to Valve, does not defer to the intellectual property system for as long as an infringement has not been expressly declared. In the same vein, the General Court appears to imply that licensees in different territories are presumed to be potential competitors before a copyright violation is formally established.

If the logic of the General Court were applied in an agreement like the one in Coditel II, the absolute territorial protection granted by virtue of the exclusive licence would be treated as a restriction by object pending a final declaration of an infringement.

It is difficult to overestimate the consequences of this interpretation of Article 101 TFEU. The presumptive validity of intellectual property titles (and, more precisely, the question of whether they raise ‘insurmountable barriers’ within the meaning of Generics) would bear no relevance in competition law analysis. These titles would only come into the picture once a breach has been proved.

The General Court’s arguments draw from the Commission decision in Valve. The latter, in turn, appears to be based on a reading of Generics that turns the specifics of the case into a general rule. Paras 49 and 50 in Generics deal with a very particular factual scenario, namely one where the primary patent had expired and where the relevant process patent did not and could not preclude potential competition.

The Generics case, in other words, was about an instance in which a presumptively valid intellectual property right was not an insurmountable barrier to entry.

The General Court judgment in Valve (just like the Commission decision) suggests the conclusions drawn in paras 49 and 50 of Generics are relevant in a different context, that is, one in which a presumptively valid intellectual property right would constitute such an insurmountable barrier. By doing so, Valve fundamentally changes, in passing, the relationship between competition law and intellectual property.

It remains to be seen whether this expansive interpretation of Generics will be embraced by the Court of Justice (either in Valve or a different case).

What can be said for the time being is that Valve is a manifestation of a wider trend in competition policy. The discipline is becoming less deferential, over time, to intellectual property. I have been tracing this evolution in forthcoming work of mine, which I look forward to sharing with the wider world. In the meantime, I would very much welcome your thoughts.

We are now opening up registrations for the 27th edition of the competition law program at the IEB in Madrid, which will run from January to March 2024. The course will follow a hybrid format (attendees can participate either in person or online). It is taught partly in English and partly in Spanish. Lectures take place in the afternoon (16h to 20h CET) to facilitate the participation of students joining from Latin America and the U.S.

We are proud to have, once again, an impressive panel of international lecturers, including Judges from EU and national courts, officials from the European Commission, the Spanish CNMC and other national competition authorities, as well as top-level academics, in-house lawyers and practitioners.

In addition to the possibility of registering for the full course, it is also possible to register for individual modules or seminars. The modules and seminars in this upcoming edition will be the following:

–Introductory session (12 January- afternoon).

–Module I – Agreements and Restrictive Practices: Vertical and Horizontal Agreements (15-17 January-afternoon). Coordinator: Carmen Cerdá Martínez-Pujalte (CNMC)

–Module II – Cartels and Procedures (22-24 January- afternoon). Coordinator: Isabel López Gálvez (CNMC)

–Seminar 1- Recent Developments in EU Competition Law (2 February- afternoon). Coordinators: Fernando Castillo de la Torre (Legal Service, European Commission) and Eric Gippini-Fournier (Hearing Officer, European Commission)

–Module III- Abuse of Dominance (5-7 February- afternoon). Coordinator: Konstantin Jörgens (Garrigues)

–Module IV – Merger Control (12-14 February- afternoon). Coordinator: Jerónimo Maillo (USP-CEU)

–Seminar 2 –  Private Enforcement of the Competition Rules- New Challenges. (23 February- afternoon). Coordinator: Mercedes Pedraz (Magistrada, Audiencia Nacional)

–Module V- Sector Regulation and Competition (26 February-28 February). Coordinator: Pablo Ibáñez Colomo (LSE, College of Europe)

–Module VI – Public Competition Law: State Aid and Public undertakings (4-6 March- afternoon). Coordinators: José Luis Buendía (Legal Service, European Commission) and Jorge Piernas (Jean Monnet Chair, University of Murcia)

–Seminar 3 – The Digital Markets Act in Practice (15 March- afternoon). Coordinator: Alfonso Lamadrid (Garrigues, College of Europe).

We will also be holding three practical workshops dealing with inspections, distribution agreements and merger control.

Like every year, we are very grateful to the course’s sponsors, namely Araoz y Rueda, Clifford Chance, Compass Lexecon, Cuatrecasas, Garrigues, Gómez Acebo & Pombo, KPMG, Latham & Watkins, MLAB, NERA, Pérez-Llorca and Uría Menéndez.

For additional information, please contact competencia@ieb.es

Over the summer, I discussed Advocate General Rantos’ Opinion in in Case C‑331/21, AdC v EDP (see here for my post). I focused primarily on the notion of restriction by object. The case wonderfully illustrates why formalism fails when identifying ‘by object’ infringements.

While it may be tempting to claim that every market-sharing (or price-fixing) agreement between competitors is caught by Article 101(1) TFEU by its very nature, the case law abundantly shows (most recently in Super Bock) that it is not true. AdC v EDP will add to this body of case law when it comes out.

The Opinion hides a valuable gem that is equally crucial to make sense of the case law, that clarifies some perceived inconsistencies and that, as far as I have been able to see, has not been discussed elsewhere. It is a tricky point that lends itself to confusion and deserves to be addressed in some detail.

Advocate General Rantos is a case study on how restrictions by object are identified in theory and practice. It elaborates on every single key aspect. The anaytical sequence can be summarised around the following points:

• First, the clause is typically the relevant unit of analysis of restrictions by object.
• Second, whether or not the clause in question is restrictive by object is considered in light of the relevant economic and legal context, which certainly includes (if there was any doubt) the analysis of the agreement.
• Third (and this is where the Opinion kicks in), once the contextual analysis reveals that the clause is restrictive by object, it is no longer relevant that other clauses within the agreement are not (following the second step, in other words, there is no second bite of the cherry).

This clarification was particularly relevant in the context of the case. The question raised by the national court concerned specifically a market-sharing clause. We know that even restraints that, at first glance, come across as problematic (including price-fixing and market-sharing itself), are not necessarily restrictive by object.

As already discussed at length in the preceding post, the case law and administrative practice provide many examples showing that the agreement is a key aspect of the evaluation of the relevant economic and legal context. It often sheds light on the object of individual clauses. It may show that such clauses are not anticompetitive and/or that they do not restrict competition at all.

As explained by the Commission in its 2014 Guidance on restrictions by object, a price-fixing clause in an agreement between competitors may not be a ‘by object’ infringement when it is introduced in the context of a (pro-competitive) joint production agreement. In Erauw-Jacquery the contextual analysis led the Court to conclude that the export prohibition in the case (in principle a no-no) was not caught by Article 101(1) TFEU because of the role it fulfilled in the overall agreement.

Sometimes, the analysis of the clause in light of the agreement reveals that it is ancillary to the latter. In Pronuptia, for instance, the Court distinguished between the restraints that were integral to the operation of the franchising agreement (and thus fell outside the scope of Article 101(1) TFEU altogether) and those that were not (such as territorial exclusivity).

However (and this is where Advocate General Rantos’ contribution is particularly valuable), once the analysis of the clause in its economic and legal context reveals that its object is anticompetitive, it does not matter whether other clauses in the agreement are not restrictive by their very nature.

In other words: the agreement informs the contextual assessment of the object of the clause. Once the object is figured out (and it turns out that it is restrictive by its very nature), whether or not other clauses are pro-competitive is no longer relevant.

To some extent, this point is obvious. The parties to an agreement can try and disguise a cartel arrangement in an agreement that has a broader scope and set of aims. The fact that the rest of the agreement would be beyond reproach would not allow them to escape the prohibition (and the fines that come with it).

In other respects, however, it is a point that is worth making. For the sake of simplification, we often refer to the agreement, not the clauses (for very good reasons).

More to the point, Advocate General Rantos’ analysis is helpful because it helps us make sense of what might seem a tension in the case law but that, upon closer inspection, is not.

Thanks to the judgments delivered over the past five years, the picture on restrictions by object is as clear as it has ever been. Landmarks such as Generics and Budapest Bank make it clear that, where a restraint is a plausible source of pro-competitive gains, it is not restrictive by object (see here for an analytical framework I prepared when the second of these judgments came out).

Reasonably, someone may be tempted to reply to the above by asking: has the Court not held, in ANSEAU-NAVEWA and BIDS, that an agreement (or clause) may restrict competitio by object even if it has other, pro-competitive aims? Would this fact not show that the above interpretation is incorrect?

Advocate General Rantos gives us part of the answer to that question and shows that there is no tension between the two lines of case law.

It may well be that an agreement, taken as a whole, has other pro-competitive aims. As explained in the Opinion, however, this point is irrelevant once it has been established that a particular clause is restrictive by object (that is, once it is shown that it has no plausible objective purpose other than an anticompetitive one in light of the relevant economic and legal context).

Issue 5 of this year’s volume of the Journal of European Competition Law & Practice came out a few days ago. It is dedicated to the memory of Professor Valentine Korah.

We pay tribute to her life and achievements in an editorial, which is available for free here. As we point out, every issue of JECLAP (with its blend of academic scholarship and practical application of the law to specific problems) is in a way a celebration of Professor Korah’s approach to EU competition law.

As one would expect from a proper rentree, I have just uploaded on ssrn (see here) a new paper, entitled ‘Form and substance in EU competition law’.

The piece deals with the unlikely comeback of formalism in competition law discussions. Formalism may be back, but the tone and substance of the debates are different from those of yore. During the 1980s and 1990s, formalism was criticised by those that championed the ‘effects-based approach’ and favoured restraint in the enforcement of competition law provisions.

In the past couple of years, formalism has been questioned for the opposite reason. Some leading commentators and practitioners see formalism as an obstacle to effective policy-making. From this perspective, the use of structured legal tests (such as the three Magill conditions or the ‘five criteria’ introduced by the Court in Intel) would hinder authorities’ ability to deal with anticompetitive conduct.

Against this background, the paper seeks to clarify what we mean by formalism (or form-based approach). It seems to me that people refer to different realities when they use the term.

The term was originally used to refer to an approach whereby the (i) categorisation and (ii) lawfulness of practices depend on their form (and their form alone). The meaning changed over time. For some, the ‘by object’ treatment of conduct is synonymous with formalism. For others, the use of structured legal tests (as opposed to unstructured or ‘liquid’ approaches), is also a manifestation of the phenomenon.

One of the conclusions of the paper is that it would be incorrect to conflate ‘by object’ and formalism. The analysis of the restrictive nature of an agreement is not necessarily formalistic. It can also rely on a substance-based approach. In the same vein, the paper explains that the use of legal categories is not inherently (and not always) formalistic. As pointed out by Wouter Wils in his instant classic on Intel, categories are a necessity in any legal order.

Another conclusion is that the Court of Justice has consistently placed substance above form both when it comes to assessing the object of agreements and when defining legal categories. There are two salient (and relatively recent) examples in this regard.

Super Bock confirmed that the form of an agreement does not and cannot determine, alone, whether it is restrictive of competition by object. The so-called ‘object box’ is elegant, intuitively appealing and incredibly useful as a first approximation to the issue, but does not reflect how the Court really goes about it.

After Super Bock (which implicitly overrules Binon), the last remaining pocket of formalism has fallen: resale price maintenance can no longer be said to amount, in the abstract, to a ‘by object’ infringement. The need to consider the ‘economic and legal context’ knows no exceptions.

Slovak Telekom, in turn, illustrates the substance-based approach that the Court follows vis-a-vis the definition of legal categories. According to the judgment, the applicablity of the Bronner test does not depend on formal criteria, but on substantive factors, namely whether requiring a dominant firm to deal with third parties would interfere with its right to property and its freedom of contract.

It would be really wonderful to get your thoughts on the paper. Do not hesitate to contact me. And, as ever, I am delighted to clarify that I have nothing to disclose.