Background…

In December 2022, the European Commission launched a public consultation on the regulation to implement the Digital Markets Act, including how the DMA will be enforced procedurally. Among the issues the regulation covers are parties’ rights to be heard, firms’ deadlines to submit documents to the Commission, access to those documents and to the Commission’s case file, and how confidentiality will be protected.

However…

While reasonable people may disagree about the merits of digital-markets regulation, appropriate procedural rules that safeguard parties’ rights and create legal certainty are essential. The timing, background, and content of the Implementing Regulation, however, all raise legitimate concerns and underscore broader issues in the DMA.

Read the full explainer here.

As 2023 draws to a close, we wanted to reflect on a year that saw jurisdictions around the world proposing, debating, and (occasionally) enacting digital regulations. Some of these initiatives amended existing ex-post competition laws. Others were more ambitious, contemplating entirely new regulatory regimes from the ground up.

Read the full piece here.

Bullet Points

• Starting with the Intel ruling, the European judiciary has slowly crafted a coherent framework for Article 102 enforcement.
• However, doubts persist concerning the exact scope of Intel and whether it is truly the lodestar that some make it out to be.
• It is arguably still unclear whether “non-price” conduct, such as self-preferencing, should be assessed under the general framework laid out in Intel, which concerned price-rebate schemes.
• With this in mind, the upcoming Google Shopping ruling will likely be the bellwether that reveals the true legacy of Intel and the future direction of European competition law.

Introduction

Competition cases routinely hinge on the fundamental distinction between conduct that anti-competitively serves to exclude competitors, on the one hand, and competition on the merits that may lead firms to exit the market, on the other.^[1] While even first-year law students intuitively understand this critical distinction, it can prove challenging to distinguish between the two in real-world cases. The reason is simple: anticompetitive foreclosure and competition on the merits both ultimately result in the same observable outcome: namely, that rivals exit the market. In order to draw the line, policymakers must infer both the root causes and the effects of firms’ market exit.

Against this backdrop, it is becoming increasingly clear that the 2017 Intel ruling marked a crucial turning point in the enforcement of Article 102 TFEU.[2] The ruling’s powerful legacy notably looms large over other recent court cases, such as the European Court of Justice’s (“ECJ”) ruling in Servizio Elettrico Nazionale and Others, as well as the General Court’s (“GC”) Intel Renvoi and Qualcomm judgments.^[3] In these Intel-inspired rulings, the European judiciary appears to have settled largely on a workable effects-based standard that sorts the wheat from the chaff in all Article 102 TFEU cases. It does this, notably, by looking at the effect that a firm’s behaviour has on “as-efficient competitors”, while also creating an administrable standard of proof to govern such proceedings.

This is not to say that Article 102 TFEU case law is entirely settled—far from it. For instance, it is arguably still unclear whether “non-price” conduct, such as self-preferencing, should be assessed under the general framework laid out in Intel, which concerned price-rebate schemes. To wit, the GC’s Google Shopping judgment marks a clear departure from the Intel framework under the justification that “[i]n the present case, the practices at issue are not pricing practices”.^[4] Despite nominally looking into the effects of Google’s “self-preferencing”, the GC eschewed key aspects of the Intel framework, such as the effect of Google’s behaviour on “as efficient competitors”.[5] This and other aspects of the ruling led to criticism that the GC failed to establish a clear boundary between anticompetitive self-preferencing and permissible instances of firms favoring their own products. As Elias Deutscher has observed:

Although the Court considered various pathways to determine the legality of self-preferencing, it failed to articulate a clear legal test that establishes limiting principles as to when self-preferencing by a dominant firm violates EU competition law.[6]

With this in mind, the pending Google Shopping appeal ruling offers the ECJ a unique opportunity to settle the debate. Namely, the ECJ may confirm that the lessons from the Intel strand of case law apply across the board, and reaffirm the dividing line between anticompetitive and benign conduct, including self-preferencing under Article 102 TFEU. This may prove a politically fraught endeavor, as the Digital Markets Act (“DMA”) will essentially outlaw the same behaviour—at least for so-called “gatekeeper” firms.^[7] It is, however, vital to safeguard the internal consistency of competition enforcement under Article 102 TFEU.

     I.        Intel and the analysis of foreclosure

The history leading up to the Intel cases is well-known. In 2009, the European Commission (“commission”) found that rebates granted by Intel to certain original equipment manufacturers (“OEMs”) foreclosed its competitor AMD from the market.^[8] During the administrative procedure, the Commission claimed it did not need to take into account the effects of the practice, as the exclusivity rebates in question were automatically illegal under Article 102 TFEU.^[9] The Commission nevertheless carried out an “as efficient competitor” (“AEC”) test “for the sake of completeness”.^[10] In a nutshell, the AEC test inquires whether a rebate scheme is able to exclude competitors from the market that are at least as efficient as the dominant firm.

During the subsequent annulment proceedings, Intel claimed the Commission had failed to apply the AEC test correctly. The GC dismissed Intel’s appeal on grounds that its conduct was per se illegal. The Commission was thus not required to assess the effects of the exclusivity rebates.^[11]

Intel took the case to the ECJ, which overturned the GC’s ruling and found that—while the Commission can, under certain circumstances, rely on a presumption of illegality—exclusivity rebates are not automatically illegal under Article 102 TFEU.^[12] In other words, the anticompetitive effects of a practice must always be assessed, regardless of how the burden of proof is allocated.

The ECJ laid out the standard of proof that parties must meet in such cases. It held that:

138. [W]here the undertaking concerned submits, during the administrative procedure, on the basis of supporting evidence, that its conduct was not capable of restricting competition and, in particular, of producing the alleged foreclosure effects.…

139. [T]he Commission is not only required to analyse, first, the extent of the undertaking’s dominant position on the relevant market and, secondly, the share of the market covered by the challenged practice, as well as the conditions and arrangements for granting the rebates in question, their duration and their amount; it is also required to assess the possible existence of a strategy aiming to exclude competitors that are at least as efficient as the dominant undertaking from the market (see, by analogy, judgment of 27 March 2012, Post Danmark, C?209/10, EU:C:2012:172, paragraph 29).^[13]

The underlying standard appears rather straightforward: if and when a defendant raises such an objection, the Commission must show that the firm’s conduct is capable of excluding “as efficient” competitors. The focus on AECs is repeated in the following paragraph:

140. That balancing of the favourable and unfavourable effects of the practice in question on competition can be carried out in the Commission’s decision only after an analysis of the intrinsic capacity of that practice to foreclose competitors which are at least as efficient as the dominant undertaking.^[14]

The ECJ’s ruling sparked numerous debates. Scholars notably questioned how it would be interpreted by the lower court when it reexamined the case. The key question was whether Intel should be read mostly as a procedural ruling—in which case, the main problem was merely that the Commission and GC did not appropriately respond to certain of Intel’s claims, something that could be corrected upon reexamination^[15]—or whether, instead, the ECJ had ultimately outlined a substantive framework to assess (at least) rebates under Article 102 TFEU.^[16]

While the finer points of its so-called “renvoi” ruling are beyond the scope of this article, the GC appears to have opted for a maximalist interpretation of Intel. It examined the effects of Intel’s rebates in great detail, following each step of the ECJ’s framework. As it explained:

125. Having regard to the wording of paragraph 139 of the judgment on the appeal, the Commission is, as a minimum, required to examine those five criteria for the purposes of assessing the foreclosure capability of a system of rebates, such as that at issue in the present case.^[17]

Following this logic, the GC examined several aspects of Intel’s rebates under a detailed-effects analysis, including their duration, market coverage, and age, as well as their effect on AECs (three of the five criteria set out in Intel).^[18]

The Intel framework was also central to the GC’s more recent Qualcomm ruling, which appears to confirm and develop insights from Intel.[19] In the judgment, the GC quashed a EUR 1 billion fine the Commission imposed on Qualcomm under Article 102 TFEU. Under scrutiny was an agreement pursuant to which Apple received payments conditional upon sourcing all of its LTE chips from Qualcomm. According to the Commission, such “exclusivity payments” were capable of foreclosing competition, in that they reduced the incentives of a major purchaser of baseband chipsets to switch to competing producers.[20]

The GC ultimately overturned the Commission’s decision on both procedural and substantive grounds. It notably found that the Commission had carried out an improper analysis of anticompetitive effects. According to the Court, the Commission failed to show that Apple would not have sourced all its LTE chips from Qualcomm absent the impugned agreement. In fact, at the time of the “exclusive payment” agreement (2011-2015, and half of 2016), Qualcomm was the only company capable of satisfying Apple’s demand for LTE chips.[21] Accordingly, Qualcomm’s decision to buy exclusively from Qualcomm could easily be attributed to “competition on the merits”, rather than anticompetitive conduct.[22]

The GC reiterated that the Commission must consider all the relevant facts of the case, as well as the supporting evidence submitted by Qualcomm that its conduct could not restrict competition and, in particular, could not have the alleged foreclosure effects.[23] As the Court put it:

Indeed, if such conduct is to be characterised as abusive, that presupposes that that conduct was capable of restricting competition and, in particular, producing the alleged exclusionary effects, and that assessment must be undertaken having regard to all the relevant facts surrounding that conduct (see judgment of 30 January 2020, Generics (UK) and Others, C?307/18, EU:C:2020:52, paragraph 154 and the case-law cited).[24]

The upshot is that Qualcomm’s exclusivity payments were incapable of excluding competitors, as there were no competitors to exclude in the first place.[25] By failing to make the link between the relevant factual circumstances and its theory of harm (i.e., the alleged lessening of Apple’s incentives to switch to a competitor to source all its LTE chipset requirements for iPhones), the Commission had fallen afoul of a line of case law, epitomized by Post Danmark, establishing that competition-law analysis cannot be purely hypothetical.[26]

Throughout its judgment, the Qualcomm court makes numerous references to Intel and its key finding that anticompetitive behaviour must be capable of excluding as efficient competitors.[27] For instance, in paragraph 356 of the judgment, the GC states that:

the Court must examine all of the applicant’s arguments seeking to call into question the validity of the Commission’s findings as to the foreclosure capacity of competitors that are at least as efficient, relating to the practice in question (see, to that effect, judgment of 6 September 2017, Intel v Commission, C?413/14 P, EU:C:2017:632, paragraph 141).[28]

The judgment also highlighted the importance of identifying the difference between anticompetitive conduct and competition on the merits. The Commission had conflated the two, and it had cost it the case. As the GC found:

The fact – which was not properly taken into account in the contested decision – that Apple sourced LTE chipsets from the applicant, and not from the applicant’s competitors, in the light of the absence of alternatives fulfilling its own technical requirements could fall within competition on the merits, and not an anticompetitive foreclosure effect resulting from the payments concerned.[29] (emphasis added)

It had also failed to show how the conduct in question prevented competitors that were as efficient as Qualcomm from developing products that fulfilled Apple’s requirements, but “that Apple’s incentives to switch to the applicant’s competitors for all its LTE chipset requirements had been reduced”.[30]

Taken together, the two Intel rulings and Qualcomm mark a clear rejection of the forms-based approach that initially enabled the Commission to ignore the economic arguments put forward by firms like Intel and Qualcomm. At least as far as rebates are concerned, the rulings marked the end of a policy of excessively deferential judicial review that turned a blind eye to, arguably, important errors of economic analysis.

   II.        The Intel reasoning is not confined to the realm of price-related conduct

The preceding paragraphs raise an important question that goes to the very foundations of European competition law. Since its adoption in 2017, there have, broadly speaking, been two competing views concerning the ECJ’s Intel ruling.

The first view held that Intel was a contained statement of the law applying to rebates.^[31] For instance, Marc Van der Woude, a judge at the GC, has argued that the Google Shopping case was correctly decided, among other things, because the Intel case law had little bearing on it. As reported by Mlex:

He said… [t]he Intel case involved the application of a framework set out by the EU’s top court, while “Google is a new phenomenon.”

Van der Woude said… that the Commission did not always need to delve into economics. “It depends on the theory of harm you will develop as a regulator.”.^[32]

Conversely, the second view contends that the Intel court was actually fleshing out a framework that undergirds all European competition law under Article 102 TFEU.^[33] According to this interpretation, Intel is but the latest in a series of rulings, including Post Danmark and Cartes Bancaires, that seek to bring economic clarity to European competition law.^[34]

There are important reasons to believe this second view is most likely the correct one. First, the GC’s Intel Renvoi and Google Shopping rulings appear to imply that Intel is not merely a rebate-specific rule. The Intel Renvoi ruling, while ostensibly dealing with the topic of rebates, ultimately rests on what appears to be a more generalized framework of presumptions and effects analysis:

124   [A]lthough a system of rebates set up by an undertaking in a dominant position on the market may be characterised as a restriction of competition, since, given its nature, it may be assumed to have restrictive effects on competition, the fact remains that what is involved is, in that regard, a mere presumption and not a per se infringement of Article 102 TFEU, which would relieve the Commission in all cases of the obligation to conduct an effects analysis.^[35]

Second, the GC’s Google Shopping ruling adds two important pieces to this puzzle. The court notably reaffirms the idea that there are no “per se” or “by object” infringements under Article 102 TFEU, suggesting that the Shopping case should be assessed under an effects-based framework similar to the one laid out in Intel:

435    [U]nlike Article 101 TFEU, Article 102 TFEU does not distinguish forms of conduct that have as their object the prevention, restriction or distortion of competition from those which do not have that object but nevertheless have that effect.

Google Shopping also contains several important references to Intel. This is a clear sign the GC believes Intel is relevant outside the realm of rebates:

[W]here, in order to classify a practice in the light of the provisions of Article 102 TFEU, the Commission attaches real importance to an economic analysis, the Courts of the European Union are required to examine all of the arguments put forward by the undertaking penalised concerning that analysis (see, to that effect, judgment of 6 September 2017, Intel v Commission, C?413/14 P, EU:C:2017:632, paragraphs 141 to 144).[36]

157    Thus, not every exclusionary effect is necessarily detrimental to competition. Competition on the merits may, by definition, lead to the departure from the market or the marginalisation of competitors that are less attractive to consumers from the point of view of, among other things, price, choice, quality or innovation (see judgment of 6 September 2017, Intel v Commission, C?413/14 P, EU:C:2017:632, paragraph 134 and the case-law cited).[37]

Along similar lines, Advocate General Nils Wahl had, in his Intel opinion, unequivocally endorsed the notion that effects analysis undergirds all of Article 102 TFEU enforcement—and not just rebate cases.^[38] The AG notably surmised that “ an abuse of dominance is never established in the abstract”, and that courts should consider “the legal and economic context of the impugned conduct”. That AG Wahl subsequently uses predatory pricing (in a rebates case) to illustrate his point is further confirmation of this broad reading.^[39]

Finally, and perhaps most dispositively, the ECJ’s recent Servizio Elettrico Nazionale and Others ruling, while pertaining to non-price conduct, is replete with references to Intel and to the notion that behaviour is only anticompetitive if it is capable of foreclosing competitors that are at least as efficient[40]:

En revanche, ainsi que la Cour l’a déjà souligné, ladite disposition ne s’oppose pas à ce que, du fait d’une concurrence par les mérites, disparaissent ou soient marginalisés sur le marché en cause des concurrents moins efficaces et donc moins intéressants pour les consommateurs du point de vue, notamment, des prix, du choix, de la qualité ou de l’innovation (arrêt du 6 septembre 2017, Intel/Commission, C?413/14 P, EU:C:2017:632, point 134 et jurisprudence citée).[41]

As if this reference to Intel—the passage is originally from the Post Danmark ruling, which also concerned rebates[42]—was not sufficiently clear, the judgment explicitly states that a practice’s effect on “as-efficient competitors” is a key part of the legal test for both price and non-price restrictions of competition. According to the court, it is this criterion, among others, that differentiates competition on the merits that forces less-efficient firms to exit the market from anticompetitive foreclosure that causes as-efficient ones to exit, too:

La pertinence de l’impossibilité, matérielle ou rationnelle, pour un hypothétique concurrent aussi efficace, mais n’étant pas en position dominante, d’imiter la pratique en cause, aux fins de déterminer si cette dernière repose sur des moyens relevant d’une concurrence fondée sur les mérites, ressort de la jurisprudence relative aux pratiques tant tarifaires que non tarifaires. [The case has not yet been translated to English][43]

This interpretation finds further support in the opinion issued by AG Rantos, who writes that:

[T]he case-law of the Court, in my view, confirms that exclusionary conduct of a dominant undertaking which can be replicated by equally efficient competitors does not represent, in principle, conduct that may lead to anticompetitive foreclosure and therefore comes within the scope of competition on the merits.

…

[A]s regards exclusionary practices not related to pricing – such as, for example, refusal to supply – the case-law seems to confirm the relevance of the test as regards the possibility of replication, inasmuch as a dominant undertaking’s decision to reserve for itself its own distribution network does not constitute a refusal to supply contrary to Article 102 TFEU when a competitor is able to create a second distribution network of a comparable size. In other words, there is no abuse if inputs refused by the dominant undertaking can be duplicated by equally efficient undertakings by purchasing them from other suppliers or developing them themselves.[44]

The Servizio Elettrico Nazionale and Others ruling thus appears to confirm the substantive importance of Intel; clearly, it is not just a procedural ruling. It also appears to confirm Intel’s significance for Article 102 TFEU enforcement beyond rebates. Indeed, the numerous important references to Intel and its analytical framework would suggest that the case marks a clear turning point for European competition law. Moreover, the court explicitly acknowledges that the Intel framework also applies to “non-price” conduct.

All of this reinforces the sense that all Article 102 TFEU cases are ultimately subject to an effects analysis, with the implication that the Commission and courts cannot summarily disregard economic arguments put forward by the parties. None of this is entirely new, of course. Several scholars have observed that other cases, like Post Danmark and Cartes Bancaires, already went a long way toward bringing European competition law more in line with economic analysis.^[45] However, the Intel cases crystallize this trend in a way that increasingly looks like a general framework for all Article 102 TFEU cases.

 III.        The Implications of Intel as a General Principle of European Union Competition Law

Subjecting all Article 102 TFEU cases to the Intel framework has far-reaching implications, including for ongoing cases such as Google Shopping. At its heart, the Intel framework ultimately seeks to ascertain whether conduct that can eliminate competitors is also anticompetitive. As the ECJ has explained in cases such as TeliaSonera and Post Danmark:

Not every exclusionary effect is necessarily detrimental to competition […]. Competition on the merits may, by definition, lead to the departure from the market or the marginalisation of competitors that are less efficient and so less attractive to consumers from the point of view of, among other things, price, choice, quality or innovation.^[46]

In the Intel proceedings, the AEC test was one of the relevant tools used to establish whether that was indeed the case. However, under different circumstances, other tools may well be more useful in separating conduct that disadvantages or even eliminates competitors—and is also anticompetitive—from conduct that has essentially the same effect but is not. The Court of Justice says this much in its Servizio Elettrico Nazionale and Others ruling:

[L]’importance généralement accordée audit test, lorsque celui?ci est réalisable, n’en démontre pas moins que l’incapacité qu’aurait un hypothétique concurrent aussi efficace de répliquer le comportement de l’entreprise dominante constitue, s’agissant des pratiques d’éviction, l’un des critères permettant de déterminer si ce comportement doit être considéré comme étant ou non fondé sur l’utilisation de moyens relevant d’une concurrence normale.[47]

Several recent cases fall short in this respect, making them vulnerable to challenges under Intel. Google Shopping is a prime example. In its ruling, the GC appears to suggest that it is anticompetitive for a dominant platform to favor its downstream services if doing so reduces web traffic to rivals:

445. [T]he Commission was fully entitled to conclude… that those practices had led to a reduction of that traffic for almost all competing comparison shopping services and, secondly, that those practices had led to an increase in traffic to Google’s comparison shopping service…. and it may be concluded that the Commission established actual effects that are more or less pronounced, depending on the country, but in any event significant.

But unlike the detailed effects analysis of Intel, the GC’s approach is tantamount to a per se condemnation of so-called “self-preferencing”. Indeed, favoring one’s own services necessarily puts competitors at a disadvantage, some of whom might even exit the market. As Pablo Ibáñez Colomo has written:

If the notion of anticompetitive effects were equated with a competitive disadvantage, then self-preferencing would become, de facto and by definition, prima facie unlawful^[48].

This forms-based analysis withers under the Intel framework. Indeed, in what would later turn out to be a direct contradiction of the upper court’s Servizio Elettrico Nazionale and Others ruling, the GC consciously ignores whether foreclosed rivals are as efficient as the dominant company:

538. The use of the as-efficient-competitor test is warranted in the case of pricing practices (predatory pricing or a margin squeeze, for example)…. In the present case, the practices at issue are not pricing practices.

We believe this is an unduly narrow reading of Intel. In Intel, the ECJ carefully distinguished the AEC test from the notion of “less attractive” or “less efficient” rivals. While the former is rightly confined to the realm of pricing practices, the latter offers a framework for deciding a much broader range of competition cases.

Few would argue it is desirable for less attractive websites to be displayed as prominently as more attractive ones, or that less efficient firms should be shielded from market exit if they cannot profitably compete in an auction (as is the case for rivals complaining about the Google Shopping remedy).^[49] Focusing on efficiency (or attractiveness, to use the ECJ’s words) is critical to draw the line between market exit that occurs due to consumer demand and that stemming from an anticompetitive strategy that harms consumers. The standard outlined in Google Shopping ignores this distinction.

This error has important consequences. Practices whereby some firms—including dominant ones—give a leg up to their own products or services often result from competition on the merits.^[50] This is especially true when one considers the Servizio Elettrico Nazionale and Others finding—quoted in full above[51]—that a dominant firm’s conduct is more likely to result from competition on the merits when non-dominant firms deploy similar strategies. Unfortunately, this very argument was rejected by the GC in Google Shopping:

Secondly, even if, as Google indicates in the application, ‘Bing’s product ads must link to pages where users can purchase the offer’, that does not address the competition concern identified. What is at issue in the present case is not Microsoft’s conduct via its Bing search engine, which, moreover, is not in a dominant position on the market for general search services, but Google’s conduct. The fact that Bing’s ads also link internet users to merchants does not preclude Google’s conduct from being anticompetitive.[52]

To be clear, nothing in the case law we have discussed suggests that non-dominant firms implementing the same practice as a dominant firm is dispositive proof that the conduct belongs to competition on the merits. It does, however, suggest that the GC was wrong to dismiss Google’s claim out of hand. Indeed, as explained above, whether non-dominant firms also implement a practice is part of the appropriate analysis to determine whether it amounts to competition on the merits. Ultimately, however, the clearest dividing line is whether the dominant firm’s conduct is capable of excluding as-efficient rivals.

This has important ramifications for policy. Allowing firms to exploit advantages obtained on the market—as opposed to government-granted privileges—may be precisely what stimulates those firms and their rivals to compete and innovate in the first place.^[53] Self-preferencing is also at the core of certain franchising agreements, which have the advantage of allowing the rapid expansion of a successful brand.^[54] In the specific case of Google, the “Google Shopping Box” may provide a better experience by giving users a more direct answer, instead of forcing them to scour through hyperlinks and retype their query.^[55] All of these scenarios may be procompetitive.

In his opinion in Intel, AG Wahl argued that:

Experience and economic analysis do not unequivocally suggest that loyalty rebates are, as a rule, harmful or anticompetitive, even when offered by dominant undertakings. That is because rebates enhance rivalry, the very essence of competition^[56].

If there remains insufficient empirical evidence to condemn an age-old practice like rebates without properly examining its ability to foreclose competitors, then policymakers should be particularly careful when drawing the boundaries for such comparatively novel practices as self-preferencing in digital markets.

In short, the Google Shopping ruling appears incompatible with the effects analysis that underpins Article 102 TFEU. It essentially rests on a per se prohibition that does not distinguish procompetitive and anticompetitive exclusion, rather than the efficiency analysis imposed by Intel. It thus provides no way to draw a line in the sand separating self-preferencing practices that disadvantage competitors and are also anticompetitive from instances in which dominant undertakings favor their own (usually downstream) products to the disadvantage of competitors but are not anticompetitive.

At the time of writing, Google has appealed the GC’s ruling.[57] It is more than likely that these distinctions will be pivotal to that case. Among other things, Google argues that:

The General Court erred in upholding the Decision despite the Decision’s failure to identify conduct that deviated from competition on the merits.… The General Court’s additional reasons as to why Google did not compete on the merits are legally invalid. [58]

Given what precedes, it goes without saying that the ECJ’s conclusions on this point will be pivotal for the future of European competition law. In light of its Intel and Servizio Elettrico Nazionale and Others rulings, the ECJ’s upper court has essentially two options.

The first would be to affirm the lower court’s distinction between price and non-price conduct. This would be unfortunate. From a purely legal standpoint, it would amount to no less than a soft renunciation of Servizio Elettrico Nazionale and Others and a pruning of Intel. It would further convey that the court’s Article 102 case law has no clear direction and is not guided by clear overarching principles. From a more substantive perspective, this first path would create an unfortunate distinction between price-related conduct—assessed under a single unified framework—and the rest of Article 102, in which formalistic distinctions determine what idiosyncratic criteria under which types of conduct are ultimately assessed.

The second option would be for the court to confirm that the potential exclusion of as-efficient competitors is a key part of the legal test for all conduct under Article 102. Not only would this conclusion safeguard the consistency of Article 102, but it would also give lower courts and competition authorities some guiding principle to assess whether novel conduct—such as self-preferencing—does or does not stem from competition on the merits. It would be insufficient for the Commission to prove (a) that a vertically integrated undertaking favors its own downstream products and (b) that this reduces the incentives of consumers to buy competing products. Accordingly, this second outcome would likely tee up so-called renvoi proceedings, where the General Court would reassess whether the Commission adduced sufficient evidence of anticompetitive foreclosure.

In short, there is every reason to believe that, when it is finally decided, the Google Shopping appeal will be a huge milestone for European competition enforcement.

Conclusion

European competition law has come a long way. From early cases like United Brands and Hoffman Laroche—widely derided for their lack of economic literacy—the European judiciary has slowly crafted a coherent framework for Article 102 enforcement.

Although many failed to recognize it at the time, it is increasingly clear that the 2017 Intel ruling was a major catalyst of this upheaval. The ruling indicates that, for any given conduct under EU competition law, there needs to be a way to discern competition on the merits from anticompetitive conduct. In most cases, this includes asking whether the conduct under investigation is capable of excluding as-efficient competitors. However, as we have explained throughout this article, this judicial revolution is not yet complete. Doubts persist concerning the exact scope of Intel and whether it is truly the lodestar that some (including us) make it out to be.

The upcoming Google Shopping ruling will likely be the bellwether that reveals the true legacy of Intel and the future direction of European competition law. The ECJ has an opportunity to clarify when self-preferencing is illegal under Article 102 TFEU. To do so, it must look at the broader implications and recognize that self-preferencing practices that disadvantage some competitors may nevertheless still be procompetitive, and that the two must be differentiated.

Such an outcome might, however, prove somewhat bittersweet. With the Digital Markets Act (“DMA”) just around the corner, careful assessment of effects under the Intel framework will largely be a thing of the past in most digital markets. For instance, both Article 102 and the DMA cover “self-preferencing practices”, but under Article 102, a prohibition is subject to different standards, a higher burden of proof, and certain limiting principles that the DMA does not possess. By its own admission, the DMA cares little about the distinction between procompetitive and anticompetitive conduct, or between competition on the merits and foreclosure, which it replaces with concepts such as fairness and contestability:

The purpose of this Regulation is to contribute to the proper functioning of the internal market by laying down rules to ensure contestability and fairness for the markets in the digital sector in general, and for business users and end users of core platform services provided by gatekeepers in particular.[59]

This does not make Intel and its legacy any less important. With the DMA in force, it will be even more important for competition policymakers to understand the difference between self-preferencing under Article 102 TFEU, which is not per se prohibited and which aims to protect competition to the benefit of consumers, and self-preferencing under the DMA, which aims to protect downstream competitors (“fairness”). By the DMA’s own admission, these are two different things. The DMA’s standards should not be unduly grafted unto competition law, even if it would be politically convenient to interpret longstanding EU competition-law principles in this still nascent and uncertain light.^[60]

The stage is thus set for the European Court of Justice to put the finishing touches on the Intel framework, thereby bringing much-needed clarity and consistency to Article 102. This would complete a remarkable turnaround for a body of law that has often been criticized for its lack of unity and economic proficiency. Whether or not this will ultimately prove to be a swan song—with the DMA attracting the bulk of enforcement—is another question.

^[1] Case C 209/10 Post Danmark, EU:C:2012:172, para 22; Case C?52/09 TeliaSonera, ECLI:EU:C:2011:83, para 43.

[2] Case C-413/14 P Intel v Commission, ECLI:EU:C:2017:788.

^[3] Case C-377/20 Servizio Elettrico Nazionale and Others, ECLI:EU:C:2022:379 (“Servizio Elettrico Nazionale and Others”); Case T-286/09 RENV Intel v Commission, ECLI:EU:T:2022:19 ; Case T-235/18 Qualcomm v Commission, ECLI:EU:T:2022:358.

^[4] Case T-612/17 – Google Shopping, ECLI:EU:T:2021:763.

[5] Id., para 538 (“[A]s regards the arguments summarised in paragraph 514 above, according to which the Commission failed to demonstrate that competing comparison shopping services that had experienced difficulties were as efficient as Google, when in fact they are not, the Commission is correct in maintaining that it was not required to prove this. The use of the as-efficient-competitor test is warranted in the case of pricing practices (predatory pricing or a margin squeeze, for example), in order, in essence, to assess whether a competitor that is as efficient as the dominant undertaking allegedly responsible for those pricing practices, and which, in order not to be driven immediately from the market, would charge its customers the same prices as those charged by that undertaking, would have to do so at a loss and accentuating that loss, causing it to leave the market in the longer term…. In the present case, the practices at issue are not pricing practices.”).

[6] Elias Deutscher, ‘Google Shopping and the Quest for a Legal Test for Self-preferencing Under Article 102 TFEU’ (2022) 6 European Papers, 1345-1361; see also Pablo Ibañez Colomo, ‘Self-Preferencing: Yet Another Epithet in need of Limiting Principles’ (2020) 43 World Competition, 417, 443.

^[7] Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the digital sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828 (Digital Markets Act) [2022] OJ L 265; on the potential overlaps and conflicts between EU competition law and the Digital Markets Act, see Giuseppe Colangelo, ‘The European Digital Markets Act and Antitrust Enforcement: A Liaison Dangereuse’ (2022) ICLE White Paper. A revised version of the paper is forthcoming in European Law Review, Available at SSRN: https://ssrn.com/abstract=4070310 or http://dx.doi.org/10.2139/ssrn.4070310.

^[8] Intel (COMP/C-3/37.990) Commission Decision of 13 May 2009.

^[9] Id., paras 920-24; 1760-61. See e.g., at 1760: “thirdly, the as efficient competitor analysis conducted in section VII.4.2.3 is not relevant for the purpose of deciding whether the Commission should impose a fine or for determining its level as it does not relate to the existence of the infringement or to the question whether it was committed intentionally or by negligence, or to its gravity within the meaning of Article 23(2)(a) of Regulation (EC) No 1/2003 and of the Guidelines, in particular points 19 to 23 thereof”.

^[10] Id., paras 1002-1153.

^[11] Case T-286/09 Intel v Commission, 145 and 152.

^[12] Case C-413/14 P Intel v Commission.

^[13] Id.

^[14] Id.

^[15] See, e.g., Ruppert Podszun, ‘The Role of Economics in Competition Law: The “Effects-Based Approach” After the Intel Judgment of the CJEU’ (2018) 7 Journal of European Consumer and Market Law 57, 64 . “Intel does not reconceptualise Article 102 TFEU. It is a case on procedural fairness. It does not call for an effects-based approach in competition law”.

^[16] See, e.g., Pablo Ibáñez Colomo, ‘The Future of Article 102 TFEU after Intel’ (2018) 9 Journal of European Competition Law & Practice 293.

^[17] Case T-286/09 RENV Intel v Commission.

^[18] Id., para 128 et seq.

[19] Qualcomm v Commission.

[20] Qualcomm (Case COMP/AT.40220) Commission Decision of 24 January 2018.

[21] Id., para 405-417.

[22] Id., para 414

[23] Id., paras 354, 396.

[24] Id., para 355.

[25] Id., para. 410. (“The undisputed fact that, on the relevant market, there was no technical alternative to the applicant’s LTE chipsets for a very large part of Apple’s requirements during the period concerned is a relevant factual circumstance which must be taken into account when analysing the capability of the payments concerned to have foreclosure effects, since the Commission found that capability in the light of Apple’s total requirements for LTE chipsets and, in particular, in the light of the reduction of Apple’s incentives to switch to the applicant’s competitors for all its requirements.”).

[26] Id., paras 397, 412, 415; see also Case C?23/14 Post Danmark, EU:C:2015:651, para 65-68.

[27] For instance, id., para349-356.

[28] Commission v. Qualcomm, para 356.

^[31] Nicholas Hirst, “Intel, Google Shopping Rules Don’t Conflict, Top EU Judge Says”, (Mlex, 31 March 2022), available at https://mlexmarketinsight.com/news/insight/intel-google-shopping-rulings-don-t-conflict-top-eu-judge-says.

^[32] Id.

^[33] See, e.g., Ibáñez Colomo (2018) 293. “The Court of Justice (hereinafter, the ‘Court’ or the ‘ECJ’) introduced an important clarification that will have a significant impact on the analysis of abusive practices under Article 102 TFEU…. Intel makes two fundamental contributions to our understanding of the notion of abuse.”

^[34] Post Danmark; Case C?67/13 P Cartes Bancaires, ECLI:EU:C:2014:2204.

^[35] Case T-286/09 RENV Intel v Commission, para 124.

[36] Google Shopping, para 131.

[37] Google Shopping, para 157.

^[38] Opinion of AG Wahl in Case C-413/14 P Intel v Commission, ECLI:EU:C:2016:788, para 73. “In this section, I shall explain why an abuse of dominance is never established in the abstract: even in the case of presumptively unlawful practices, the Court has consistently examined the legal and economic context of the impugned conduct. In that sense, the assessment of the context of the conduct scrutinised constitutes a necessary corollary to determining whether an abuse of dominance has taken place. That is not surprising. The conduct scrutinised must, at the very least, be able to foreclose competitors from the market in order to fall under the prohibition laid down in Article 102 TFEU.”

^[39] Id., para 78. “As I see it, the analysis of ‘context’—or ‘all the circumstances’, as it is termed in the Court’s case-law —aims simply but crucially to ascertain that it has been established, to the requisite legal standard, that an undertaking has abused its dominant position. Even in the case of seemingly evident exclusionary behaviour, such as pricing below cost, context cannot be overlooked. Otherwise, conduct which, on occasion, is simply not capable of restricting competition would be caught by a blanket prohibition. Such a blanket prohibition would also risk catching and penalising pro-competitive conduct” (footnotes omitted for ease of reading).

[40] Servizio Elettrico Nazionale and Others, paras 45, 46, 51, 73, 74 & 86.

[41] Id. para 45.

[42] Post Danmark, para 22.

[43] Id., para 79.

[44] Opinion of AG Rantos in Case C-377/20 Servizio Elettrico Nazionale and Others, paras 69-74.

^[45] See, e.g., Jao Cardoso Pereira, ‘Groupement des Cartes Bancaires: Reshaping the Object Box’ (2015) 18 Competition and Regulation 265, 266. “This judgment shows how economic analysis plays an increasing role in shaping the boundaries of competition policy. In this judgment, the ECJ has taken insights from recent advances in the economic analysis of two-sided markets to reach the conclusion that the pricing measures adopted by the Groupement could not be treated as a restriction of competition by object.” See also, Dirk Auer & Nicolas Petit, ‘Two-Sided Markets and the Challenge of Turning Economic Theory into Antitrust Policy’ (2015) 60 The Antitrust Bulletin, 447.

^[46] Post Danmark, para 22.

[47] Servizio Elettrico Nazionale and Others, para 82.

^[48] Pablo Ibañez Colomo, ‘Self-Preferencing: Yet Another Epithet in need of Limiting Principles’ (2020) 43 World Competition, 417, 443.

^[49] Dirk Auer, “Case Closed: Google Wins (for now)” (Truth on the Market, 19 November 2021), available at  https://truthonthemarket.com/2021/11/19/case-closed-google-wins-for-now/.

^[50] See. e.g., Geoffrey Manne, “Against the Vertical Discrimination Presumption” (2020) 2-2020 Concurrences 1. “The notion that self-preferencing by platforms is harmful to innovation is entirely speculative. Moreover, it is flatly contrary to a range of studies showing that the opposite is likely true. In reality, platform competition is more complicated than simple theories of vertical discrimination would have it, and there is certainly no basis for a presumption of harm.”

[51] Servizio Elettrico Nazionale and Others, para 79.

[52] Google Shopping, para. 354.

^[53] Forcing those firms to share such advantages with rivals’ might encourage them to slack off and free-load. See, in this respect, Opinion of AG Jacobs’ opinion in Case C-7/97 Oscar Bronner, EU:C:1998:264. For a more detailed discussion, see. e.g., Dirk Auer, ‘Appropriability and the European Commission’s Android Investigation’ (2017) 23 Columbia Journal of European Law 647.

^[54] Case 161/84 Pronuptia, EU:C:1986:41, para 15.

^[55] Jakob Kucharczyk, “When Product Innovation Becomes a Competition Law Infringement: Preliminary Thoughts on the Google Shopping Decision”, (2017) 1 European Competition and Regulatory Law Review 193; see also Geoffrey Manne, “The Real Reason Foundem Foundered” (2018) ICLE Antitrust & Consumer Protection Research ProgramWhite Paper 2018-2, available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3384297.

^[56] Opinion of AG Wahl in Case C-413/14 P Intel v Commission, para 90.

[57] Appeal brought on 20 January 2022 by Google LLC and Alphabet, Inc. against the judgment of the General Court (Ninth Chamber, Extended Composition) delivered on 10 November 2021 in Case T-612/17, Google and Alphabet v Commission (Case C-48/22 P)

[58] Id.

[59] Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), point (7) of the preamble.

^[60] Margethe Vestager has recently claimed the Apple Pay investigation would “inform the future application of the Digital Markets Act. It will set a precedent with regard to the analysis of the security concerns, and a recipe for effective and proportionate access to NFC for mobile payments.” See European Commission, “Remarks by Executive Vice-President Vestager on the Statement of Objections sent to Apple over practices regarding Apple Pay” (Brussels, 2 May, 2022), available at https://ec.europa.eu/commission/presscorner/detail/en/SPEECH_22_2773.

 

With just a week to go until the U.S. midterm elections, which potentially herald a change in control of one or both houses of Congress, speculation is mounting that congressional Democrats may seek to use the lame-duck session following the election to move one or more pieces of legislation targeting the so-called “Big Tech” companies.

Read the full piece here.

As it has before in its history, liberalism again finds itself at an existential crossroads, with liberally oriented reformers generally falling into two camps: those who seek to subordinate markets to some higher vision of the common good and those for whom the market itself is the common good. The former seek to rein in, temper, order, and discipline unfettered markets, while the latter strive to build on the foundations of classical liberalism to perfect market logic, rather than to subvert it.

Read the full piece here.

Background…

In recent years, a set of politicians, academics, regulators, and activists have called for more stringent antitrust scrutiny of companies perceived to be threats to democracy simply by virtue of their size. Today, the brunt of these criticisms target so-called “big tech” companies (see here, here, and here).

But…

There is little evidence to support the contention that more concentrated markets lead to less democratic outcomes. Furthermore, the fears of private economic power expressed by advocates for more forceful antitrust tend to elide the genuine threat of state coercion. This is at odds with key principles of liberal democracy, such as the rule of law and freedom from arbitrary state interference.

Read the full explainer here.

Early Morning

I wake up grudgingly to the loud ring of my phone’s preset alarm sound (I swear I gave third-party alarms a fair shot). I slide my feet into the bedroom slippers and mechanically chaperone my body to the coffee machine in the living room.

Read the full piece here.

Executive Summary

Intermediaries may not be the consumer welfare hero we want, but more often than not, they are one that we need. Policymakers often assume that intermediaries and centralization serve as a cost to society, and that consumers are better off when provided with “more choice.” Concrete expression of this view can be found in regulatory initiatives that aim to turn “closed” platforms into “open” ones (see, in Europe, the Digital Markets Act; and in the United States, the Open App Markets Act and the American Innovation and Choice Online Act). Against this backdrop, we explain that, as with all economic goods, intermediation involves tradeoffs. Intermediaries emerge when it would otherwise be too difficult (or too costly) for groups of users to meet and interact. There is thus no guarantee that government-mandated disintermediation — such as that contemplated in the European DMA and the U.S. AICOA bill — will generate net benefits in a given case. The ongoing Epic v Apple proceedings are a good example of why it is important to respect the role of intermediaries in digital markets, and the unique benefits intermediation can bring to consumers. The upshot is that intermediaries are far more valuable than they are usually given credit for.

Read the full issue brief here.

Executive Summary

Digital advertising is the economic backbone of much of the Internet. But complaints have recently emerged from a number of quarters alleging the digital advertising market is monopolized by its largest participant: Google. Most significantly, a lawsuit first filed by the State of Texas and 17 other U.S. states in 2020 alleges anticompetitive conduct related to Google’s online display advertising business. The U.S. Justice Department (DOJ) reportedly may bring a lawsuit similarly focused on Google’s online display advertising business sometime in 2022. Meanwhile, the United Kingdom’s Competition and Markets Authority undertook a lengthy investigation of digital advertising, ultimately recommending implementation of a code of conduct and “pro-competitive interventions” into the market, as well as a new regulatory body to oversee these measures. Most recently, a group of U.S. senators introduced a bill that would break up Google’s advertising business (as well as that of other large display advertising intermediaries such as Facebook and Amazon).

All of these actions rely on a crucial underlying assumption: that Google’s display advertising business enjoys market power in one or more competitively relevant markets. To understand what market power a company has within the market for a given type of digital advertising, it is crucial to evaluate what constitutes the relevant market in which it operates. If the market is defined broadly to include many kinds of online and/or offline advertising, then even complete dominance of a single segment may not be enough to confer market power. On the other hand, if the relevant market is defined narrowly, it may be easier to reach the legal conclusion that market power exists, even in the absence of economic power over price.

Determining the economically appropriate market turns importantly on whether advertisers and publishers can switch to other forms of advertising, either online or offline. This includes the specific ad-buying and placement tools that the Texas Complaint alleges exist within distinct antitrust markets—each of which, it claims, is monopolized by Google. The Texas Complaint identifies at least five relevant markets that it alleges Google is monopolizing or attempting to monopolize: publisher ad servers for web display; ad-buying tools for web display; ad exchanges for web display; mediation of in-app ads; and in-app ad networks.

As we discuss, however, these market definitions put forth by the Texas Complaint and other critics of Google’s adtech business appear to be overly narrow, and risk finding market dominance where it doesn’t exist.

Digital advertising takes numerous forms, such as ads presented along with search results, static and video display ads, in-game ads, and ads presented in music streams and podcasts. Within digital advertising of all kinds, Google accounted for a little less than one-third of spending in 2020; Facebook accounted for about one-quarter, Amazon for 10%, and other ad services like Microsoft and Verizon accounted for the remaining third. Open-display advertising on third-party websites—the type of advertising at issue in the Texas Complaint and the primary critiques of Google’s adtech business—is a smaller subset of total digital advertising, with one estimate finding that it accounts for about 18% of U.S. digital advertising spending.

U.S. digital advertising grew from $26 billion in 2010 to $152 billion in 2020, an average annual increase of 19%, even as the Producer Price Index for Internet advertising sales declined by an annual average of 5% over the same period. The rise in spending in the face of falling prices indicates that the number of ads bought and sold increased by approximately 26% a year. The combination of increasing quantity, decreasing cost, and increasing total revenues is consistent with a growing and increasingly competitive market, rather than one of rising concentration and reduced competition.

But digital advertising is just one kind of advertising, and advertising more generally is just one piece of a much larger group of marketing activities. According to the market research company eMarketer, about $130 billion was spent on digital advertising in the United States in 2019, comprising half of the total U.S. media advertising market. Advertising occurs across a wide range of media, including television, radio, newspapers, magazines, trade publications, billboards, and the Internet.

An organization considering running ads has numerous choices about where and how to run them, including whether to advertise online or via other “offline” media, such as on television or radio or in newspapers or magazines, among many other options. If it chooses online advertising, it faces another range of alternatives, including search ads, in which the ad is displayed as a search-engine result; display ads on a site owned and operated by the firm that sells the ad space; “open” display ads on a third-party’s site; or display ads served on mobile apps.

Although advertising technology and both supplier and consumer preferences continue to evolve, the weight of evidence suggests a far more unified and integrated economically relevant market be-tween offline and online advertising than their common semantic separation would suggest. What publishers sell to advertisers is access to consumers’ attention. While there is no dearth of advertising space, consumer attention is a finite and limited resource. If the same or similar consumers are variously to be found in each channel, all else being equal, there is every reason to expect advertisers to substitute between them, as well.

The fact that offline and online advertising employ distinct processes does not consign them to separate markets. The economic question is whether one set of products or services acts as a competitive constraint on another; not whether they appear to be descriptively similar. Indeed, online advertising has manifestly drawn advertisers from offline markets, as previous technological innovations drew advertisers from other channels before them. Moreover, while there is evidence that, in some cases, offline and online advertising may be complements (as well as substitutes), the dis-tinction between these is becoming less and less meaningful as the revolution in measurability has changed how marketers approach different levels of what is known as the “marketing funnel.”

The classic marketing funnel begins with brand-building-type advertising at the top, aimed at a wide audience and intended to promote awareness of a product or brand. This is followed by increasingly targeted advertising that aims to give would-be customers a more and more favorable view of the product. At the bottom of this funnel is an advertisement that leads the customer to purchase the item. In this conception, for example, display advertising (to promote brand aware-ness) and search advertising (to facilitate a purchase) are entirely distinct from one another.

But the longstanding notion of the “marketing funnel” is rapidly becoming outdated. As the ability to measure ad effectiveness has increased, distinctions among types of advertising that were once dictated by where the ad would fall in the marketing funnel have blurred. This raises the question whether online display advertising constitutes a distinct, economically relevant market from online search advertising, as the Federal Trade Commission, for example, claimed in its 2017 review of the Google/DoubleClick merger.

The Texas Complaint adopts a non-economic approach to market definition, defining the relevant market according to similarity between product functions, not by economic substitutability. It thus ignores the potential substitutability between different kinds of advertising, both online and offline, and hence the constraint these other forms of advertising impose on the display advertising market.

If advertisers faced with higher advertising costs for open-display ads would shift to owned-and-operated display ads or to search ads or to other media altogether—rendering small but significant advertising price increases unprofitable—then these alternatives must be included in the relevant antitrust market. Similarly, if publishers faced with declining open-display ad revenues would quickly shift to alternative such as direct placement of ads or sponsorships, then these alternatives must be included in the relevant market, as well.

If advertisers and publishers are faced with a wide range of viable alternatives and the market is broadly defined to include these alternatives, then it is not clear that any single firm can profitably exercise monopoly power—no matter what its market share is in one piece of the broader market. Similarly, it is not clear whether “consumers” (e.g., advertisers, publishers, or users) have suffered any economic harm.

With a narrow focus on “open display,” it is quite possible that Google’s dominance can be technically demonstrated. But if, as suggested here, “open display” is really just a small piece of larger relevant market, then any fines and remedies resulting from an erroneously narrow market definition are as likely to raise the cost of business for advertisers, publishers, and intermediaries as they would be to increase competition that benefits market participants.

Read the full issue brief here.