By Kris Van Hove, Argyrios Papaefthymiou, and Margot Vogels1
The present article provides an update of the ongoing process taking place in Europe that is reshaping competition law to bring it up to speed with the digital era. It follows on from a previous CPI Europe article published in October 2020 on the same subject.2 That article set the scene by: (i) identifying the perceived shortcomings of competition law enforcement in the digital sector in Europe; and (ii) discussing the various approaches being considered to tackle them. Since the publication of this first article, various legislative initiatives have been launched that have turned the reflective discourse into concrete action. For instance, in December 2020, the European Commission (“Commission”) published its Proposal for a Digital Markets Act3 (“DMA Proposal”). This proposal seeks to ensure fair and open digital markets by imposing stringent requirements on so-called “gatekeeper” platforms. The inter-institutional deliberations on this key piece of EU legislation are now in their final stretch, with adoption expected sometime in April of this year.
In short, we now have tangible regulatory reforms in the open, at the level of both the EU and national jurisdictions, as well as a flurry of recent enforcement action and judicial developments in “digital antitrust.” It is therefore time to take stock of the concrete regulatory approaches presently being considered to tackle competition law concerns in the tech sector, as well as of the recent developments in the enforcement of “classic” competition law relating to these same concerns. From the “mosaic” of these different yet parallel workstreams, the hope is to shed further light on the precise extent to which a material need for new rules specific to the digital sector actually arises, as well as to assess the adequacy of the approaches that are currently on the table to target the perceived concerns.
II. Self-Preferencing and Interoperability
A. Concerns and Regulatory Response
Self-preferencing can be defined as a practice engaged in by a platform operator that gives preferential treatment to this operator’s products and services when they compete with products and services of other entities using the platform. This type of practice has featured prominently from the outset of the ongoing debate on digital competition policy. Despite continuous controversy among policymakers, practitioners, and academics on the merits of a “self-preferencing” theory of harm, this type of conduct is vigorously scrutinized by competition authorities in Europe and is the behavior most closely associated with competition concerns attributed to the practices of Big Tech companies.
For instance, the Commission’s engagement with self-preferencing nowadays spans from antitrust enforcement to sector inquiries.4 But self-preferencing is also targeted as particularly pernicious behavior under newly proposed as well as existing sector-specific regulation for Big Tech. First, under the DMA Proposal, gatekeepers5 would be prohibited from ranking their own products and services more favorably than those of third parties.6 The DMA Proposal would also create interoperability obligations.7 These obligations are often presented as potential remedies to self-preferencing practices, when the latter are implemented via “closed systems” that indirectly foreclose competitors on one or more levels of the supply chain.
Second, self-preferencing rules have also recently been enacted (e.g. in Germany) or are under consideration (e.g. in Italy) in various national jurisdictions across Europe.8 In Germany and Italy, competition authorities could prohibit an undertaking of paramount significance for competition from treating the offers of competitors differently from its own offers when providing access to supply and sales markets. Finally, in the UK, the Competition and Markets Authority’s (“CMA”) Digital Markets Unit (“DMU”) would be able to impose pro-competitive interventions (or “PCIs”), in the form of “consumer choice and defaults intervention,” on undertakings designated as having “strategic market status” (“SMS”), which is defined as a “substantial, entrenched market power in at least one digital activity, providing the undertaking with a strategic position.”9 These PCIs could be used to address concerns relating to the architecture of a service which influences the consumer’s decision-making.10
B. Enforcement and Judicial Developments in the Fight against Self-preferencing: Luxembourg finally Weighs In
Beyond the legislative and policy initiatives that target self-preferencing with novel tools, the self-preferencing theory of harm is also making headway in the conventional track of abuse of dominance enforcement.
The much-awaited General Court judgment in Case T-612/17 Google and Alphabet v Commission (Google Shopping), issued on 10 November 2021,11 is indisputably the most important recent case law development in the ongoing process of delineating the exact contours of the self-preferencing theory of harm. The case centered on the question of whether Google had abused its dominant position in the market for online general search services to favor its own comparison shopping services over competing services.
The General Court’s judgment is, of course, still subject to appeal before the European Court of Justice. This said, the importance of the judgment cannot be overstated: the very validity of self-preferencing as a self-standing antitrust theory of harm – let alone the particulars of its application in individual factual scenarios – has been the subject of controversy. Critical opinions have invariably been expressed among EU competition law commentators, from the view that self-preferencing is merely a natural – if not intended and efficiency-enhancing – outcome of vertical integration (so that sanctioning self-preferencing would be tantamount to sanctioning vertical integration itself), to the view that, even if some antitrust harm may be identified in what is dubbed “self-preferencing,” this could nevertheless easily fit within existing theories of harm, such as a refusal to supply an indispensable input/a refusal to access an “essential facility.”
The General Court takes a rather straightforward position in the debate on the validity of self-preferencing as a self-standing category of abusive conduct, viewing self-preferencing as, in effect, one of various possible manifestations of “abusive leveraging.” In doing so, the Court has rejected Google’s argumentation to the effect that the conduct at issue should have been examined by the Commission through the prism of the “refusal to supply” theory of harm. The Court did not see, in the conduct at issue, the key constituent elements of a refusal to supply warranting the application of the Bronner12 criteria, namely: (i) an express refusal following a request or otherwise expression of the wish to access the infrastructure at issue, and (ii) the exclusionary effect being principally predicated on this refusal, and not on a separate practice (such as an independent “leveraging abuse”). According to the Court, EU case law indicates that the Bronner criteria do not necessarily apply when assessing the potential abusive nature of conduct beyond the set confines of a refusal to supply, such as the supply of services to competitors on disadvantageous terms.
The most tangible consequence of the Court’s reasoning is that the analysis of whether or not Google’s general search results page is, in fact, an “essential facility” – to which Google could be mandated to provide access subject to the fulfilment of the Bronner criteria (elimination of competition if access is not provided, and indispensability of the “infrastructure” for other market players to be able to compete) – is not a necessary component of the legal test to determine the potential abusive nature of the conduct. In fact, although the Court views Google’s general search results as having “characteristics akin to those of an essential facility […], inasmuch as there is currently no actual or potential substitute available that would enable it to be replaced in an economically viable manner on the market,” it finds that the Commission was not required to establish that the Bronner criteria were satisfied. The Court reasoned that the Commission had identified a form of abusive leveraging that amounted to “‘active’ behavior in the form of positive acts of discrimination in the treatment of the results of Google’s comparison shopping service, which are promoted within its general results pages, and the results of competing comparison shopping services, which are prone to being demoted.”13
In a way, Google got the “worst of both worlds”. Its general search engine is considered “essential enough” to mandate increased scrutiny of any possibly disadvantageous treatment; yet the conduct at issue was not considered a classic refusal to supply, so that the Bronner criteria – which place a rather high threshold to establish a refusal to supply – were not applicable.
In an also very interesting (if somewhat nebulous) refrain, the Court noted that Google, in favoring its own comparison shopping services and simultaneously demoting competing services, engaged in conduct that entailed “a certain form of abnormality.”14 The Court made this finding based on the rationale that the ability to showcase results from multiple and diverse sources is an inherent element of a general search engine’s business model. In conjunction with the rather bold statement that a general principle of equal treatment in EU law is applicable in the context of Article 102 TFEU, the Court concluded that the practice by which Google favored its own comparison shopping service over competing services amounted to a self-standing form of abusive conduct. That conduct consisted in the abusive leveraging of Google’s dominant position in the market for online general search results to provide less favorable treatment to certain search results, based on the origin of the results, that is, based on whether the results originated from Google’s own comparison shopping service or from a rival comparison shopping service. In this context, little relief or legal certainty is derived from the Court’s phrasing that “certain differences in treatment may be considered contrary to Article 102 TFEU when what is at issue are favoring practices established by operators in a dominant position in the internet sector.”15 Although this seems to imply that not all differences in treatment will be considered abusive in breach of Article 102 TFEU, the Court’s judgment provides little in terms of clarity as to where the line might be drawn in this respect.
The Google Shopping judgment unarguably represents a pivotal moment in the broader ongoing discourse on the review of competition law norms as applied to digital markets. The notion of self-preferencing has now been recognized by the EU judiciary (subject to reaffirmation by the Court of Justice) as an independent, self-standing category of abusive conduct, and thus not subject to the specific conditions of the refusal to supply / essential facilities doctrine (i.e. the Bronner test).
What is more, as the General Court pointedly noted in Google Shopping:
“[I]t is for the national courts and authorities to apply Article 102 TFEU uniformly and in accordance with the case-law of the Courts of the European Union, since divergences between the courts and authorities of the Member States as to its application would be liable to place in jeopardy the unity of the EU legal order and to undermine legal certainty.”16
To appreciate the potential impact of this comment, one only needs to consider the number of ongoing antitrust investigations and court cases within the EU at the national level, which revolve around practices whereby a vertically integrated undertaking is providing some form of favorable treatment to a proprietary product / service as compared to the equivalent product / service provided through that undertaking (e.g. a digital platform) by third parties. The necessary deference of national competition authorities and courts towards the approach of EU courts means that, if the General Court’s position is upheld by the Court of Justice, a series of infringement decisions from national authorities can be expected, if not a further proliferation of antitrust complaints against digital “gatekeepers” for similar practices.
C. Takeaways: Unlawfulness of Self-preferencing Becoming a fait accompli
Reading through the rundown of recent intense enforcement action on the basis of the self-preferencing theory of harm, one might reasonably wonder whether an ex ante approach in this respect (as most clearly reflected in the DMA) is truly necessary and/or proportional to the perceived competitive concerns associated with the practices falling within the ambit of self-preferencing. Both the Commission and EU national competition authorities (“NCAs”) as well as the CMA seem to be more than adequately stepping up to the challenge of reviewing and sanctioning this type of practice with ex post enforcement of classic antitrust rules regarding abuse of dominance.
Relatively more critical voices may argue that the Commission is trying, through the DMA, to circumvent the procedural safeguards laid out by EU competition law (notably in Regulation No. 1/2003)17, taking a horizontal stance that practices falling within the broader “self-preferencing” category are presumptively anti-competitive. Such a stance is economically dubious. Self-preferencing may, at least in a multitude of possible manifestations, be viewed as a perfectly valid outcome of vertical integration (without which, in fact, the incentives of firms to vertically integrate, with the corollary efficiencies that may come with such integration, may not arise in the first place).18 In this context, the establishment of what seems an awful lot like a legal presumption that self-preferencing is inherently anti-competitive (thereby also placing the burden on companies that implement practices that may be construed as such, to prove that these practices do not restrict competition), may be considered a step too far, in the direction of dictating to companies what business model they should adopt, in order to accommodate their rivals. That, of course, is not in line with the prime directive of competition law, which is to uphold competition on the merits.
III. Data-related Practices
A. Concerns and Regulatory Response
Data is indisputably acquiring an ever-increasing status of importance within the broader competitive process. This trend is even more prevalent in the realm of digital platforms, whose intensive use generates vast amounts of data, including both personal data (which remain relatively constant over time) and dynamic data (which relate to interactions with content and other users).
Illustrating the pressing nature of the competition concerns arising from data-related practices, the DMA Proposal would impose a comparatively large number of data-sharing obligations aimed at loosening the tight grip that gatekeepers allegedly exert on user data. Primarily, gatekeepers would have to refrain from using, in competition with business users, non-public data generated by the activities of business users on the gatekeeper’s core platform services.19] Moreover, gatekeepers would have to grant a business user effective, real-time access to data provided for, or generated in the context of, the use of the gatekeeper’s core platform services by that business user and the end user engaged with the products or services provided by the business user.20 To allow for easier switching between platforms’ services, the DMA Proposal creates a data portability right for business users (in addition to end users) regarding the data generated through the activities of those users.21 Finally, in a provision designed to address one of the sources of their market power, gatekeepers would be prohibited from combining personal data sourced from their core platform services with personal data from any other services, unless the end user has provided consent.22
Similar rules have been enacted (e.g. in Germany) or are under consideration (e.g. in Italy) at national level.23 For its part, the UK DMU would be able to impose PCIs that should ensure greater end-user control over the end user’s data and that mandate third-party access to data. Additionally, the DMU would be able to oblige SMS undertakings to separate data silos.24
B. Data-related Theories of Harm take Center Stage
Competition authorities in the EU are taking their cue from policy makers, increasingly placing data-related concerns at the epicenter of antitrust enforcement actions. The tempo in this area has been set by the German Facebook case, relating to the social network giant’s data collection practices, and more specifically its combination / integration of data from different sources (e.g. the Facebook Like and Share button functionalities along with the – also Facebook/Meta-owned – WhatsApp and Instagram apps and third-party websites), without the explicit consent of Facebook users to the processing of such combined data. With breach of relevant privacy laws (notably the General Data Protection Regulation or “GDPR”) being at the core of the theory of harm proposed by the Bundeskartellamt, the case’s development into a saga was anything but surprising: on appeal, the Higher Regional Court of Düsseldorf halted the implementation of the NCA’s decision via interim proceedings (after Facebook had applied for suspension of the decision’s execution); the Federal Supreme Court (“Bundesgerichtshof”) went on to reinstate the decision’s immediate implementation; and, in the main proceedings on appeal, the Düsseldorf Court decided in March 2021 to refer seven questions to the Court of Justice, mainly revolving around the interpretation of the GDPR, particularly in the context of (i) the German NCA’s competence to establish breaches of the privacy regulation, and (ii) the manner in, and weight with, which GDPR infringements may play into an antitrust law assessment.25 Facebook is also facing a £ 2.3 billion data-related antitrust class action claim before the UK’s Competition Appeal Tribunal, on behalf of approximately 44 million UK consumers, for an alleged abuse of dominance in the form of unlawful exploitation of consumers’ personal data.26
Competition authorities in other EU jurisdictions are also increasingly following suit with their own investigations into Big Tech’s data-related practices. On 26 November 2021, the AGCM announced the imposition of € 10 million in fines on both Apple and Google for breach of the Italian Consumer Code, in relation to practices involving the collection of data from consumers and the insufficient compliance with the related obligation to adequately inform consumers of said data collection and obtain their prior informed consent.27 Both companies have announced that they will be appealing the Italian authority’s decision. Especially in the case of Apple, this decision probably strikes a particularly sensitive chord, given the company’s proverbial refrain of a privacy-focused comparative advantage vis-à-vis its competitors, ostensibly thanks to its closed-system architecture.
In a similar vein as the AGCM’s infringement decision on Apple’s and Google’s data collection practices, the French data privacy authority (Commission nationale de l’informatique et des libertés or “CNIL”) announced on 6 January 2022 fines of € 60 million and € 150 million on Facebook and Google respectively, for indirectly restricting users’ freedom to reject cookies. The CNIL took issue in particular with the discrepancy between, on the one hand, the ease of accepting cookies on the two companies’ French websites, requiring the push of a single button, and on the other hand, the relatively more complex process of rejecting cookies, which required that a user go through a number of prompts to reach the intended rejection. The CNIL found that this practice was in breach of the French Data Protection Act.
C. Takeaways: Consumer Protection, Privacy and Competition Law Continue to Find Common Ground in Data-related Antitrust Considerations
The enforcement record on data-related practices so far does not always have competition law as its legal basis. We are increasingly seeing infringement decisions against GAFAM companies with regard to data-related practices on the basis of national consumer protection and data privacy legislation. However, in light of the increasing invocation of data-related language by competition authorities to substantiate competitive harm (most characteristic example in this respect being the German Facebook case), these cases, and their adjudication at the appeal level, are likely to influence the evolution of thinking on data-related theories of harm in the competition law space in the near future. This is even more so given that we are now in anticipation of the opinion of the Court of Justice – in its capacity as the highest authority on the interpretation of EU law – on questions that go directly at the core of the privacy/antitrust nexus.
IV. Retail Most-Favored-Nation (“MFN”) Clauses
A. Concerns and Regulatory Response
Most-favored-nation (“MFN”) clauses are traditionally divided into two categories:
(i.) “narrow” MFNs, which provide that the seller may not offer better terms and conditions on its own website (but can do so on other platforms); and
(ii.) “wide” MFNs, which provide that the seller may not offer better terms and conditions on its own website or on any other platform.
In particular, MFNs imposed by providers of online intermediation services, such as marketplaces or price comparison tools, and relating to the conditions under which business users offer goods or services to end users, are commonly referred to as “retail MFNs.” From a competition policy perspective, wide retail MFNs are generally considered more problematic than narrow retail MFNs. At its core, the differentiation traces back to the idea that MFNs may have certain efficiency benefits due to the prevention of free-riding behavior: a provider of online intermediation services will often undertake significant investments (in e.g. advertising and marketing), which would be undermined if, once informed by the intermediation services, the consumer then proceeds to make a final purchase on the seller’s own website because the latter charges a lower price. Although this rationale might, exceptionally, justify a narrow retail MFN, wide retail MFNs are generally considered too disproportionately restrictive to benefit from an efficiency exemption.
The Commission’s draft revised Vertical Block Exemption Regulation28 (“Draft Revised VBER”) excludes wide retail MFNs in favor of online intermediation services from the benefit of the safe harbor. Wide MFNs require a buyer of online intermediation services not to offer, sell or resell goods or services to end users (final consumers or other undertakings) under more favorable conditions using competing online intermediation services.29 As a result of the proposed change, the legality of wide retail MFNs in favor of online intermediation services would be subject to a self-assessment.
The UK may well take a slightly stricter stance on what the CMA prefers to call parity obligations. Unlike the Commission, the CMA has recommended that wide retail MFNs should be viewed as hardcore restrictions regardless of whether they apply to online or offline intermediation services.30
Gatekeepers, for their part, will not have to think twice about concluding wide MFNs in respect of their online intermediation services. According to the Commission, MFNs between gatekeepers and their business users deter the latter from using alternative online intermediation channels, which in turn limits inter-platform contestability, and therefore choice for end users. For this reason, the DMA Proposal will outright ban wide MFNs concluded with the business users of online intermediation services.31 The European Parliament even considers that this prohibition should extend to narrow MFNs.32
B. German Federal Supreme Court Leads the Way against MFNs
The German Federal Supreme Court (Bundesgerichtshof) might be sympathetic to the idea of a rigidly harsh stance against MFNs. On 18 May 2021, the Court confirmed a 2015 decision of the Bundeskartellamt which found that the online hotel reservation platform operator Booking.com had infringed competition law by imposing narrow MFN clauses.33 Through the narrow retail MFNs at issue, Booking.com obtained a commitment from hotel owners operating through its platform to refrain from providing on their own websites lower prices or otherwise better terms than the ones offered on the platform. The German NCA found these clauses to constitute an anti-competitive vertical restraint, in breach of Article 101(1) TFEU and Section 1 of the German Competition Act.
The Bundesgerichtshof disagreed with the OLG Düsseldorf’s key premise that there was a significant free-riding problem that could best be addressed via the narrow retail MFNs. The German Supreme Court highlighted that the main prerequisite for an ancillary restraint to be exempted from the prohibition of Article 101(1) TFEU is that the restraint must be objectively necessary and indispensable for implementing the main activity covered by the agreement, which it did not consider to be the case for the narrow retail MFNs at issue. The Court seemed to consider that the narrow retail MFN was not a fundamental prerequisite for the existence of the distribution agreement itself, as the Bundeskartellamt had viewed it. In the Court’s view, the fact alone that, according to evidence from the Bundeskartellamt’s investigation, Booking.com had not only retained, but actually consolidated its market position after it stopped implementing the clauses at issue, was enough to reject the notion that the clauses were objectively necessary and therefore not an ancillary restraint.
C. Takeaways: Pressure Mounts against Price Parity Clauses of all Shapes and Sizes
The German Federal Supreme Court’s decision in Booking.com has marked the first time that MFNs have been adjudicated at the highest judicial rank within an EU jurisdiction. What is more, the ruling dispels the commonly held view that narrow price parity clauses are “safer” from a competition law perspective than wide MFNs (whereby the supplier – in the Booking.com context, the hotel owner – would be precluded from providing better terms anywhere else, i.e. not only on its own website, but through any other platforms). The key premise behind this differentiation has always been the “free-rider” line of argumentation (which ostensibly might provide some merit to the conditional lawfulness of narrow MFNs). Yet, this reasoning has now been dealt a serious blow by the German Federal Supreme Court. With no case law available in this respect at the highest EU level (i.e. the European Court of Justice), this judgment by the German Federal Supreme Court is the “next best thing”, and is therefore likely to provide the necessary impetus to NCAs in other EU jurisdictions to pursue narrow price parity clauses in a more vigorous manner, at least to the extent that they fall outside the safe harbor of the Verticals Block Exemption Regulation.
In any event, the regulatory mélange of stricter and relatively looser approaches to MFNs in Europe, may inevitably nudge companies to take the safer option of refraining from the implementation of such clauses altogether. This is unfortunate given the efforts the Commission is undertaking to clarify its position with respect to the legality of MFNs in the Draft Revised VBER and the draft revised Vertical Guidelines (“Draft Revised VGL”) 34. More clarity from the European Court of Justice in this respect would be most welcome.
V. Online Sales Restrictions
A. Concerns and Regulatory Response
As a general rule, every distributor must be allowed to sell its products and services over the internet. From a competition law perspective, vertical competition rules assimilate making products and services available online to passive sales, while the use of certain promotional and advertisement strategies over the internet is considered to constitute active sales.35 Furthermore, while restrictions of active sales are generally permissible under certain conditions, restrictions of passive sales are not.36 Accordingly, under the Draft Revised VBER and Draft Revised VGL, restrictions on certain methods of selling or advertising online will be block exempted, as long as they do not amount to an effective ban on online sales.37 As a matter of fact, the Commission proposes to broaden the range of online sales practices which would be block exempted, including some that are currently considered hardcore restrictions. Similarly, the CMA has already indicated in its final Recommendations for a UK Order replacing the retained Vertical Agreements Block Exemption Regulation that it is inclined to treat as active sales certain categories of online sales which are currently assimilated to passive sales.38
Another major change in the Commission’s approach to online restrictions relates to dual pricing: under the revised draft rules, a supplier may validly charge a hybrid buyer different wholesale prices for products to be resold online and those to be resold offline, so long as this difference reflects differences in the costs incurred in each sales channel at retail level and does not have the object of preventing online sales.39
The Commission, however, takes a stricter stance on restrictions on the use of specific advertising channels, such as price comparison websites and advertising on search engines. While a supplier could validly prohibit the use of one specific price comparison tool or search engine without losing the benefit of the block exemption, for example to protect its brand or fight against counterfeiting,40 a ban on using all most widely used advertising services could amount to preventing the use of the internet and, if so, would therefore not be block exempted.41 Likewise, a ban on the use of the suppliers’ trademarks or brand names for bidding to be referenced in online search advertising services would effectively prevent the buyer from selling over the internet and is for that reason considered problematic.42
Finally, the Commission is also stricter as far as online intermediation service providers are concerned. Under the Draft Revised VBER, the Commission categorizes online intermediation services as suppliers. Accordingly, the Draft Revised VGL clarify that online intermediation services cannot in principle qualify as “genuine agents” for the purposes of Article 101 TFEU.
B. The Italian Apple/Amazon Case: Platform Bans in the Spotlight
In an interesting development in antitrust enforcement against online sales restrictions, Italy’s AGCM imposed in late 2021 fines of € 58.6 million and € 114.7 million on Amazon and Apple respectively, for an Article 101 TFEU infringement related to the distribution of Apple products via Amazon Marketplace.43 The Italian NCA’s decision centers around the fact that Apple allowed Amazon itself and a limited number of resellers to distribute Apple-branded and Beats-branded products (Beats being an Apple subsidiary) over Amazon Marketplace. Amazon was thus in fact acting as a hybrid platform, in that it provided intermediation services to Apple resellers, while also acting as an Apple reseller itself. The AGCM noted that the selection process for distributors was arbitrary, without any consideration for objective, qualitative criteria. The Italian NCA thus considered the relevant agreements to be in breach of Article 101 TFEU, as they restricted intra-brand competition by unjustifiably discriminating in favor of a select few resellers.
This case presents a particular novelty in terms of the competition law assessment of platform bans. Such bans are not considered hardcore restrictions of competition under the EU case law in Coty44 and may therefore benefit from the safe harbor of the VBER. Yet, the AGCM in this case considered that the limitations on distribution at issue constituted restrictions of competition by object, given that – according to the authority – the relevant agreements had as their object to hinder the access to Amazon Marketplace of undertakings that were lawfully authorized to resell Apple’s products. The Italian NCA made this finding even though the resellers at issue were still allowed to sell on online platforms / marketplaces in general, except for Amazon.it. The AGCM also identified a cross-border element to the restriction at issue, as it found that the selection process discriminated based on the reseller’s country of establishment, in fact favoring resellers without significant export activities (thus, in effect, favoring Italian resellers). The cross-border restriction element may have accentuated the discriminatory nature of the restriction in the NCA’s view, thus leading to the finding of a “by-object” restriction. Finally, the Italian NCA also conducted a separate effects analysis and found that the online sales restrictions concerned amounted to restrictions by effect that could not be exempted. In any event, the parties are reportedly appealing the AGCM’s decision and it will be interesting to see what the Italian courts will have to say on these aspects of the NCA’s findings.
C. Takeaways: Revised Vertical Rules set to Provide some Clarity on Platform Sales Restrictions
The Commission’s and the CMA’s revision of their respective rulebooks on vertical restraints is expected to provide much-needed clarity on previously grey areas. Key principles that have arisen from European competition authorities’ decisional practice as well case law are now set for crystallization into black-letter law (e.g. platform bans explicitly qualified as non-hardcore restrictions, inasmuch as they do not span into restrictions on the use of online search engines or price comparison tools, which could prevent the effective use of the internet for making sales).
Of course, the revised EU regulatory package on vertical restraints will not alleviate all of the ambiguities that existed under the previous regime (as the AGCM’s decision in Apple/Amazon attests). Therefore, competition authorities and courts in Europe will continue to be tasked with streamlining the interpretation and application of these rules. Having said this, with specific, targeted guidance on the books on all of the hallmark practices that have raised competition law concerns in e-commerce (e.g. marketplace restrictions, price comparison website restrictions, MFNs), the scramble of European competition authorities over the past decade to try to fit novel issues arising from the online reality into an increasingly outdated framework of analysis is hopefully nearing some conclusion.
VI. CONCLUDING REMARKS
Beyond any policy and legal aspects / criticisms of the novel regulatory approaches to digital-specific antitrust in Europe, this piece has provided an overview of the current landscape in classic antitrust enforcement against tech companies’ practices in Europe. As may be deduced from this exposition, enforcement in Europe seems to be robust and rife with new investigations propping up across the region seemingly on a daily basis, by the Commission as well as the EU NCAs and the CMA. While one should not downplay the increasingly complex nature of antitrust enforcement in the digital sector, it is nevertheless a reality that cases dealing with novel questions may take a while to conclude, sparking the unavoidable criticisms regarding the risk of locking in anti-competitive effects until it is too late for any remedies to deliver their intended results. Nevertheless, some optimism may be in order: with classic theories of harm being tested and adapted for the digital economy, European competition authorities seem to be familiarizing themselves with the characteristics of the tech space, and enabling themselves to scrutinize more swiftly the commercial practices implemented in the digital sector.
These trends are intended to generate positive outcomes for consumers in the near future. Given the central role that digital services play in everyday life nowadays, underenforcement of competition law in the digital space is certainly a risk worth serious consideration. However, at the same time, tech companies are facing an increasingly complex regulatory landscape in Europe; and notably, this may, sooner rather than later, affect companies beyond the small GAFAM circle. One important factor in this may be the Commission’s ability to impose certain DMA obligations to providers of core platform services which do not yet enjoy an entrenched and durable position in their operations, but may soon enjoy such a position.45 Additionally, in data-related cases, several NCAs have started relying on sectoral regulatory obligations as standards of conduct in abuse of dominance cases. In light of this ongoing trend, commentators have warned against the possibility that competition authorities or courts may transpose the obligations of the DMA to “dominant non-gatekeepers” through traditional abuse of dominance provisions. The rationale would be that, if data-related regulatory instruments, such as the GDPR, can be used as “benchmarks” for abuse of dominance assessments (see German Facebook case), the DMA could also serve as a similar benchmark for an abuse of dominance assessment by a dominant non-gatekeeper.46
In turn, this could lead to a chilling effect, dampening innovation due to a competition law framework that interferes excessively in digital companies’ business models: why innovate if equal access to the fruits of one’s labour would have to be provided to rivals? So, it is not beyond imagination that any near-term benefits for consumers may be offset by long-term welfare losses, due to missing out on efficiency- and welfare-enhancing practices in the tech space. This is so in particular given that a number of the DMA’s blacklisted practices, such as self-preferencing, do have an efficiency dimension. Indeed, it is hard not to marvel at the concept of the DMA itself, bearing a strong “per se prohibitions” mindset (without the possibility for justification) that is rather foreign in most European competition law regimes (and in fact, is entirely foreign – in principle – to abuse of dominance assessments).47
A lot may depend on whether, in their effort to safeguard consumers’ interests in the digital economy, competition authorities in Europe (including the Commission) overshoot by underestimating classic supply and demand forces. How all of this will play out remains to be seen. With the DMA still being fleshed out, one thing is for certain: no boredom in sight for the digital antitrust afficionados.
1 Kris Van Hove is a competition partner at the law firm Van Bael & Bellis, Brussels. Argyrios Papaefthymiou is an associate in the Antitrust & Competition practice of the law firm Zepos & Yannopoulos, Athens. Margot Vogels is an associate at the law firm Van Bael & Bellis, Brussels. The views expressed in this article are those of its authors and should not be attributed to their respective law firms or clients.
2 Kris Van Hove and Argyrios Papaefthymiou, Revising the Competition Law Rulebook for Digital Markets in Europe: A Delicate Balancing Act, 11 October 2020, CPI Europe column available at: https://www.competitionpolicyinternational.com/revising-the-competition-law-rulebook-for-digital-markets-in-europe-a-delicate-balancing-act/.
3 Proposal for a Regulation of the European Parliament and of the Council on Contestable and Fair Markets in the Digital Sector (Digital Markets Act), 15 December 2020, COM/2020/842 final.
4 In terms of antitrust enforcement, the Commission has ongoing investigations or proceedings against all GAFAM companies on grounds that seem to reflect self-preferencing concerns, including of course the Google Shopping saga but also in the Amazon Marketplace (AT.40462), Amazon Buy Box (AT.40703), Facebook leveraging (AT.40484), and Apple – App Store Practices (music streaming) investigations. The Commission has also launched an investigation into allegations that Microsoft has been force-installing Teams in the Office suite and preventing its uninstallation (EU questions Microsoft rivals over Teams integration in Office, Reuters (8 October 2021), available at https://www.reuters.com/technology/eu-questions-microsoft-rivals-over-teams-integration-office-2021-10-08/). In terms of sector inquiries, the Commission has found in its recently published final report on its sector inquiry into consumer IoT that practices potentially amounting to self-preferencing are present in this space. See Final Report – Sector inquiry into consumer Internet of Things, 20 January 2022, SWD(2022) 10 final.
5 Under the current draft of the “gatekeeper” definition, the term may apply to undertakings that (i) have a significant impact on the internal market, (ii) operate a core platform service which serves as an important gateway for business users to reach end users, and (iii) enjoy an entrenched and durable position in their operations (or can be expected to enjoy such a position in the future). In respect of each of these conditions, the Commission has chosen to create presumptions of satisfaction on the basis of specific quantitative thresholds. See DMA Proposal, Article 3(1).
6 DMA Proposal, Article 6(1)(d).
7 DMA Proposal, Article 6(1)(f).
8 In Germany, see the 10th Amendment to the German Act against Restraints on Competition, approved by the German Parliament on 14 January 2021, available at https://www.d-kart.de/wp-content/uploads/2021/01/GWB-2021-01-14-engl.pdf (“10th Amendment to the German Act against Restraints on Competition”), Section 19a, para. 2, sub-paras. 1 and 5. In Italy, see Report within the meaning of Articles 21 and 22 of the Law of 10 October 1990 as regards competitive reform proposals for the purposes of the Annual Market and Competition Act for the year 2021, 23 March 2021, available at https://www.agcm.it/dotcmsdoc/allegati-news/S4143%20-%20LEGGE%20ANNUALE%20CONCORRENZA.pdf (“AGCM Report on Competitive Reform Proposals”), Article 3bis, para. 2, sub-paras. a and e.
9 Advice of the Digital Markets Taskforce: A new pro-competition regime for digital markets, available at https://assets.publishing.service.gov.uk/media/5fce7567e90e07562f98286c/Digital_Taskforce_-_Advice.pdf (“Digital Markets Taskforce Advice”), para. 12.
10 Id., para. 4.68.
11 Case T-612/17, Google and Alphabet v Commission (Google Shopping), judgment of 10 November 2021, ECLI:EU:T:2021:763.
12 Case C-7/97, Oscar Bonner GmbH, judgment of 26 November 1998, ECLI:EU:C:1998:569.
13 Case T-612/17, Google and Alphabet v Commission (Google Shopping), judgment of 10 November 2021, ECLI:EU:T:2021:763, recital 240.
14 Id., recital 616.
15 Id., recital 180.
16 Id., recital 248.
17 See indicatively Pablo Ibáñez Colomo, The Draft Digital Markets Act: A Legal and Institutional Analysis, 12(7) Journal of European Competition Law & Practice (2021) p. 561.
18 See indicatively Pablo Ibáñez Colomo, Self-Preferencing: Yet Another Epithet in Need of Limiting Principles, 43(4) World Competition (2020) p. 417.
19 DMA Proposal, Article 6(1)(a).
20 DMA Proposal, Article 6(1)(i). Gatekeepers would also have to provide third-party search engines with ranking, query, click and view data in relation to online searches by end users on the search engines of the gatekeeper on the basis of FRAND terms. See DMA Proposal, Article 6(1)(j).
21 DMA Proposal, Article 6(1)(h).
22 DMA Proposal, Article 5(a).
23 10th Amendment to the German Act against Restraints on Competition, Section 19a, para. 2, sub-paras. 4 and 6, and AGCM Report on Competitive Reform Proposals, Section 3bis, para. 2, sub-paras. d, f and g.
24 Digital Markets Taskforce Advice, para. 4.68.
25 See referral decision (in German) at: https://www.justiz.nrw.de/nrwe/olgs/duesseldorf/j2021/Kart_2_19_V_Beschluss_20210324.html, and (in English) at: https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:62021CN0252&from=EN.
26 See GCR, UK class action claim seeks £2.3 billion from Facebook, available at: https://globalcompetitionreview.com/collective-actions/uk-class-action-claim-seeks-ps23-billion-facebook. This is the last in an already trend-forming line of antitrust class action suits against Big Tech companies in the UK, with antitrust class actions against Apple (see Competition Appeals Tribunal, Notice of an application to commence collective proceedings under Section 47B of the Competition Act 1998, Case No. 1403/7/7/21, available at: https://www.catribunal.org.uk/cases/14037721-dr-rachael-kent) and Google (see Competition Appeals Tribunal, Notice of an application to commence collective proceedings under Section 47B of the Competition Act 1998, Case No. 1408/7/7/21, available at: https://www.catribunal.org.uk/cases/14087721-elizabeth-helen-coll) pending with respect to the two companies’ app store practices.
27 Press Release, AGCM, $20 million sanctions against Google and Apple for commercial use of user data (26 November 2021), available at: https://en.agcm.it/en/media/press-releases/2021/11/PS11147-PS11150.
28 Vertical block exemption Regulation, Draft C(2021) 5026 final, 9 July 2021.
29 Draft Revised VBER, Article 5(1)(d).
30 UK competition law: Vertical Agreements Block Exemption Regulation – CMA’s recommendation, available at: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1030880/VABER_Final_RecommendationOctober2021__PVedit.pdf (“CMA’s Recommendations for a UK Order”), paras. 5.93-5.97. The recommendations on MFNs addressed here have been endorsed by the UK Department for Business, Energy and Industrial Strategy in its draft UK Vertical Block Exemption Order. Draft: The Competition Act 1998 (Vertical Agreements Block Exemption) Order 2022, Article 8(2)(f) and Draft Explanatory Memorandum: The Competition Act 1998 (Vertical Agreements Block Exemption) Order 2022, paras. 7.32 – 7.42 available at: https://www.gov.uk/government/consultations/draft-vertical-agreements-block-exemption-order.
31 DMA Proposal, Article 5(b).
32 Parliament’s Amendment 105.
33 An English translation of Bundesgerichtshof judgment available at: https://www.bundeskartellamt.de/SharedDocs/Publikation/EN/Others/BGH_Entscheidung_Booking_EN.pdf?__blob=publicationFile&v=3.
34 Commission Vertical guidelines, Draft C(2021) 5038 final, 9 July 2021.
35 Commission Guidelines on Vertical Restraints, SEC(2010) 411 (“Vertical Guidelines” or “VGL”), paras. 52-53.
36 Commission Regulation (EU) No 330/2010 of 20 April 2010 on the Application of Article 101(3) of the Treaty on the Functioning of the European Union to Categories of Vertical Agreements and Concerted Practices, OJ (2010) L102/1 (“VBER”), Article 4.
37 For instance, taking the view that online marketplaces like Amazon represent only one method of selling online, the Commission considers in the Draft Revised VGL that direct or indirect online marketplace bans should be block exempted irrespective of the distribution system in place and regardless of whether the contract product is considered a luxury product (Draft Revised VGL, para. 194(a)). Similarly, setting quality standards for selling online and requiring that the buyer operates brick-and-mortar shops to be admitted in the supplier’s selective distribution system should also be block exempted (Draft Revised VGL, para. 194(b)). Following the same line of reasoning, a supplier could validly require a buyer to sell at least a certain absolute amount of the contract goods or services offline (Draft Revised VGL, para. 194(c)).
38 CMA’s Recommendations for a UK Order, para. 5.76 and Explanatory Memorandum on the Draft UK Order, paras. 7.25 – 7.31.
39 Draft Revised VGL, para. 195.
40 Draft Revised VGL, para. 325.
41 Draft Revised VGL, para. 192(f).
42 Draft Revised VGL, para. 192(f).
43 See AGCM, Decision no. 29889 of 16 November 2021, available at: https://www.agcm.it/dotcmsdoc/allegati-news/I842%20chiusura.pdf and Press Release, AGCM, Italian Competition Authority: the fines applied to Amazon and Apple for anticompetitive agreements have been redetermined (17 December 2021), available at: https://en.agcm.it/en/media/press-releases/2021/12/I842.
44 Case C-230/16, Coty Germany GmbH v Parfümerie Akzente GmbH, judgment of 2 April 2020, ECLI:EU:C:2017:941, recital 69.
45 The Commission may designate as a gatekeeper a provider of core platform services that does not yet enjoy an entrenched and durable position, but in relation to which it is “foreseeable that it will enjoy such a position in the near future.” DMA Proposal, Article 3(6). In particular, the Commission may declare a limited number of obligations (i.e. Articles 5(b) and 6(1)(e),(f), (h) and (i)) applicable to those gatekeepers. DMA Proposal, Article 15(4).
46 See indicatively on this line of thought, King & Spalding, The Digital Markets Act’s Per Se Prohibitions Increase Legal Risks for Non-Gatekeeper Platforms, available at: https://www.kslaw.com/attachments/000/009/434/original/King___Spalding_–_The_Digital_Markets_Act’s_Per_Se_Prohibitions_Increase_Legal_Risks_for_Non-Gatekeeper_Platforms_–_9_February_2022.pdf?1644955782. As noted in this paper, this would not be unprecedented in the EU as the General Court in Google Shopping relied on the EU Open Internet Regulation (No 2015/2120) in discerning an equal treatment obligation incumbent on Google (see recital 180).
47 For an analysis of the ways in which the DMA marks a shift from an effects-based to a per se approach in the EU competition rulebook on unilateral conduct, see Nicolas Petit, The Proposed Digital Markets Act (DMA): A Legal and Policy Review, 12(7) Journal of European Competition Law & Practice (2021) p. 529.