Article précédent

Unilateral Practices in the digital market: An overview of EU and national case law

Abstract

A frequent starting point of the ongoing debates on a future platform regulation – in the EU in the form of a “Digital Markets Act” (DMA) – is the alleged ineffectiveness of competition law enforcement in the digital realm, and in particular when it comes to “abuse of dominance” or monopolization proceedings against the largest digital platforms. This paper aims to do add to this debate in two ways: In a first part, it provides a rough overview of the competition law cases on unilateral practices in digital markets that have been initiated and partly completed over the last 10 years or so, with a strong focus on cases against large digital platforms. While there is a focus on the EU and its Member States, the overview also looks at relevant cases in other jurisdictions like the U.S., Australia, India, Russia and China in order to give an impression of the global enforcement dynamics. The overview – which is mostly based on the Concurrences database, with only some additional research on our part, which is by necessity selective – does not dive into a discussion of the merits of the cases. Rather, it is meant to systematize the enforcement actions and to provide a clearer picture when, where and why action has been taken on which grounds. A second part strives to draw some tentative conclusions from this overview against the background of ongoing policy debates. Has enforcement indeed been intolerably slow? Does the enforcement panorama indicate what’s special about ensuring undistorted competition in the presence of gatekeepers and why we might need a special regime of platform regulation? Does it tell us something about the optimal scope of such a regulation, and about the interaction of competition law, the law on unfair business terms and consumer protection law in the digital realm? Does it hold insights about what we can expect from public and private enforcement respectively?

I. Introduction

The EU has set out to regulate the largest digital platforms that have come to assume “gatekeeper” positions. The Commission’s proposal for a “Digital Markets Act” (DMA) of 15 December 2020 [1] reacts to a widely shared perception that the existing competition rules and their enforcement over the last 10 years did not suffice to effectively protect competition in digital markets and to ensure that decentralized, independent innovation remains feasible [2].

But did competition law really fail? And if so: in which way(s) and why? Did the Commission and the national competition authorities fail to bring cases? Was the enforcement too slow, or did competition authorities fail to conclude their cases successfully? Is there something in the methodology of applying competition law that systematically hinders authorities to prove anticompetitive conduct to the requisite standard? Are there certain types of practices that have anticompetitive effects but are beyond the scope of the existing competition rules? Or were the remedies ineffective? If the remedies were ineffective: was this due to a failure of the competition authorities to identify the course of action needed, or to constraints implicit in competition law?

The presentation of the major cases on unilateral practices in digital markets that e-competitions Bulletin presents in this special issue cannot answer all these questions. But it provides a useful empirical basis for revisiting some of them. It is an important endeavor: only if we understand the strengths and shortcomings of competition law in addressing failures of digital markets will we be able to take informed decisions on how to improve competition law, complement it by a different set of rules and make the regimes interact smoothly.

Our contribution proceeds in two steps:

A first part (II.) gives an overview of the cases, closed or pending. We distinguish between private enforcement (II.1.) and public enforcement (II.2.-II.4.). There is a focus on the enforcement of EU competition law at EU level (II.2.) and at national level (II.3.), but we also look at competition law enforcement abroad, knowing that this is no more than a rough and incomplete sketch (II.4). We concentrate on enforcement action, but we also give some consideration to sector inquiries and policy contributions.

A second part (III.) is dedicated to some tentative observations on what we can learn from this overview in the ongoing policy debates on the DMA.

II. Overview of cases and enforcement initiatives

1. Mostly public enforcement, little private enforcement

A first brief look at the list of cases confirms the obvious: So far, most of the cases that have challenged allegedly anticompetitive unilateral practices in digital markets have been brought by competition agencies. This is true in particular for the cases against the so-called “Big Tech”, namely Google, Amazon, Facebook and Apple [3].

There are exceptions to this rule: Unsurprisingly, there have been private enforcement efforts notably in the U.S. – but largely unsuccessful so far. In 2013, two class actions against Apple were turned down. In a first case, the plaintiffs alleged that Apple, by monopolizing or attempting to monopolize the software applications aftermarket for iPhones, had harmed competition and injured consumers. Consumer harm was claimed to result from the fact that Apple collects 30% of all app sales intermediated through its App Store, and at the same time prevents iPhone users from downloading any third-party app through other routes. The District Court for the Northern District of California granted Apple’s motion to dismiss [4]. Among other things, the consumers’ standing remained unclear. According to the “Illinois Brick [5] doctrine, only the first party in the distribution chain to suffer antitrust harm has standing to sue. In order to facilitate cartel damage claims, U.S. antitrust law does not allow defendants a pass-on defense vis-à-vis direct purchaser plaintiffs. But at the same time, it bars suits for antitrust damages by customers who do not buy directly from a defendant, because otherwise a risk of double recovery against defendants would result [6]. In another lawsuit, a move by a plaintiff to certify a class of indirect purchasers of the iPod was denied and her antitrust claims were dismissed. The plaintiff had pursued a private damage claim based on an alleged infringement of Sec. 2 Sherman Act: According to her, Apple’s proprietary encryption of music files (called FairPlay) as well as software updates that rendered the music files of the iTunes Music Store and the iPod compatible only with one another amounted to an exclusion of competitors and allowed Apple to monopolize both the market for portable digital media players and the market for music downloads and thereby to inflate music prices while the value of the iPod decreased. Again, the class certification failed: Standing for damages for an iPod overcharge (or a diminution of the iPod’s value) was denied on the Illinois Brick rationale. The damages claim for an overcharge for music downloads was dismissed because the plaintiff had failed to show causal antitrust injury, namely that Apple, if it had not used software updates to thwart competition, would have lowered the price of its music [7].

In 2015, a class action brought by consumers against Google for tying its Android OS to the Google Search App was dismissed on similar grounds: the plaintiffs, who had alleged consumer harm by supra-competitive prices for Android phones, had failed to provide sufficient facts to prove antitrust injury “in the market where competition is being restrained”. The alternative theory of antitrust injury – a restriction of consumer choice and a likely loss of innovation – were found to be “entirely too conclusory and speculative” [8].

As these cases show, consumer class actions did manage to identify potentially anticompetitive types of conduct in principle, but, due in part to their dogmatic limitations under U.S. antitrust law [9], they ultimately failed.

More recently, businesses that consider themselves to be harmed by allegedly anticompetitive conduct of the “Big Tech” have started to file antitrust suits. On 13 August 2020, Epic Games has sued both Apple and Google in the Northern District of California. The claim against Apple is that the company has unlawfully maintained its monopoly on both the iOS app distribution market and the iOS in-app payment processing market by preventing iOS users from downloading apps from any source other than Apple’s App Store, and by imposing a 30% tax on every iOS user’s app purchase [10]. In addition, Apple is said to coerce app developers to exclusively use Apple’s own payment processing platform for all in-app purchases, and to prevent them from informing users about the option of purchasing the same content outside of the app. The parallel complaint against Google is based on the allegation that Google unlawfully monopolized both the market for the distribution of mobile apps to Android users and the market for the processing payments for digital content within Android mobile apps by imposing a tax that siphons monopoly profits for itself for each app or in-app purchase, and by extracting all user data exchanged in such transactions [11]. Furthermore, Google is alleged to have erected contractual and technological barriers that foreclose competing ways of distributing apps from app stores on Android devices, and to tie the distribution of apps through its Google Play Store to the exclusive use of Google’s own payment processing tool, Google Play Billing. On 20 October 2020, the online publisher Genius Media has filed a complaint against Google, alleging anticompetitive conduct in display advertising. In particular, Google’s Ad Server is claimed to impose anticompetitive rules and conduct “that artificially warp the channels through which publishers sell their ad placement inventory” [12]. These cases are still pending. It remains to be seen whether the business partners of the large digital platforms will turn out to be more effective antitrust enforcers than consumers.

There has been a limited amount of private enforcement action also in Europe. The UK High Court’s decision in the Streetmap v. Google case [13] has received much attention [14]. In short, Streetmap – a provider of online mapping services – had contended that, by the visual display of a clickable image from Google Maps, and no other map, at or near the very top of its general search engine result page, Google was abusing its dominant position in the market for online search. The High Court left open whether Google was indeed dominant. Instead, it found that the introduction of the new-style (Google) Maps OneBox was not reasonably likely to affect competition in the market for online maps appreciably [15]; and that the conduct would be objectively justified even if it was reasonably likely to affect competition.

In an earlier case on a somewhat related matter, the Tribunal de commerce de Paris had found that Google had abused its dominant position by offering Google Maps for free with the goal to foreclose competition, thereby engaging in a predatory pricing scheme [16]. But the decision was overturned by the Cour d’Appel de Paris in 2015 [17]. The court clarified that other sources of income, in particular advertising income, must be taken into account in determining whether a given pricing conduct is predatory.

In January 2021, the Landgericht München I issued an interim injunction against Amazon, ordering it not to close a specific seller’s account [18]. According to the court’s initial assessment, Amazon had likely abused its dominant position on the market for online marketplace services for online sellers in Germany by deactivating the account without providing sufficient information and statement of reasons. However, Amazon prevailed in the subsequent main proceeding: The Landgericht München I found that Amazon had shut down the seller’s account based on legitimate reasons, namely the goal to effectively combat the manipulation of product reviews on the platform. This goal would not exonerate Amazon from its obligation to comply with the requirements set out in the P2B Regulation 2019/1150 [19] to provide the seller with clear and fact-based reasons – which it had not done in this case. But as the seller was a repeat infringer, and Amazon had referred to prior infringements of a similar kind, the seller was sufficiently aware of the reproach, and Amazon was justified to deactivate the account without a further statement of the full set of facts. The court referred to Art. 4(5) of the P2B Regulation here, finding that the fundamental decisions underlying the rules of the P2B Regulation should likewise inform competition law, given the partly overlapping goal of the P2B Regulation to ensure a fair, predictable, sustainable and trusted online business environment within the internal market [20].

Private suits in other European jurisdictions have been confined to cases addressed against companies that are not in the “GAFA(M)” league. Whereas the current debate focuses on the anticompetitive potential of strong intermediaries, a case still pending before Italian courts turns on whether undertakings that are dominant in a relevant product or services market are allowed to fend off intermediation. Viaggiare – an online travel agency – sued Ryanair for access to its database and procedures for flight reservations in order to be able to provide intermediation services. In November 2019, the Italian Supreme Court confirmed the dominant position of Ryanair in the relevant markets in the transport sector and sent the case back to the Court of Appeal of Milano to determine whether Ryanair’s refusal to provide access resulted in discriminatory and therefore anticompetitive behavior [21].

The reasons for the scarcity of private litigation against GAFA(M) are obvious: Frequently, businesses will lack the requisite information they would need to sue or the technical skills to competently assess the available information. It is one of the goals of the P2B Regulation 2019/1150 to address these information asymmetries, e.g. with regard to ranking criteria (Art. 5) and self-preferential practices (Art. 7). The draft DMA proposes to oblige gatekeepers to provide advertisers and publishers to which it provides advertising services with information concerning the price paid by the advertiser and publisher (Art. 5 lit. g) and with access to the gatekeeper’s performance measuring tools (Art. 6 lit. g).

But even if the information asymmetries are addressed, the dependency of business users on gatekeepers to reach end users is often so significant that fear of retaliation can deter them from taking action. The relevance of that fear is evidenced by the Commission’s proposition to include into the list of obligations of gatekeepers a provision according to which gatekeepers “must refrain from preventing or restricting business users from raising issues with any relevant public authority relating to any practice of gatekeepers” (Art. 5 lit. d DMA). If such restrictions occur with regard to public complaints, the potential for retaliation in case of private litigation looms even larger.

2. Public enforcement at EU level

Given the hurdles to an effective private enforcement against unilateral practices by digital “gatekeepers” [22], public enforcement is key. As the list of cases presented by this special issue shows, the EU Commission has been by far the most influential competition authority in the digital arena worldwide. Famously, the U.S. agencies had briefly looked into Google’s practices in 2013 [23], but have then ceded action against “Big Tech” until very recently (see below, II.4. lit. a). The Indian and Russian competition authorities have become active from 2015 onwards. The Canadian and Australian competition authorities have moved only recently (see below, II.4. lit. b and c). It was not until 2021 that China began to take significant action against anticompetitive conduct of its Big Tech companies (see below, II.4. lit. h).

The EU Commission, on the other hand, looks back on a continuous enforcement agenda with regard to digital platforms, starting with its Microsoft decision in 2004 [24]. While the Microsoft decision was perceived an outlier at the time, it has proved to be a harbinger of many of the issues and developments that were to follow. Microsoft continued to be on the Commission’s radar for some time [25]. But the focus then shifted first to geoblocking practices utilized by Apple (both with regard to downloads from iTunes and with regard to iPhone repair services), and to restrictions in the terms and conditions of Apple’s licensing agreement with independent app developers who were allowed to use only Apple’s native programming tools and approved languages when writing iPhone apps [26].

Then, from 2010 onwards, Google came into focus [27]. Google has kept the Commission busy for a decade now. The Commission’s three large Google decisions (27 June 2017 – Google Search (Shopping), 18 July 2018 – Google Android, and 20 March 2019 – Google Search (AdSense)) were milestones as they have recognized Google’s gatekeeper position in a number of different markets, as well as the anticompetitive potential that comes with it. But the proceedings have been relatively lengthy, considering the dynamic market setting [28]. Also, the remedies imposed are widely considered to not effectively and comprehensively address the underlying competition problem [29]. New complaints about Google keep coming up. Frequently, these are complaints about strategies that look similar to the strategies observed in the above-mentioned cases but take place in a somewhat different setting or market [30]. On 22 June 2021, the Commission has opened antitrust proceedings against Google to examine whether the company has favored its own display advertising technology services to the detriment of competing advertising technology service providers, advertisers and online publishers [31]. In particular, the investigation focuses on the obligation to use Google’s service DV360 and the Google Ad Manager; the favoring of Google’s ad exchange "AdX" by DV360 and /or Google Ads and the potential favoring of DV360 and/or Google Ads by AdX; the restriction of third-party access to user data which is available to Google; Googles plan to prohibit third-party cookies on Google Chrome and replace them with the "Privacy Sandbox"; and its plans to stop making the advertising identifier available to third parties on Android devices when a user opts out of personalized advertising.

The Commission’s experience with Google and the absence of a possibility to impose, under the existing competition law regime, more forward-looking and comprehensive remedies that do not only provide redress for a past abuse but address the special features of the relevant market such as to effectively protect competition in the future is arguably the most important driver of the EU’s attempt to design a new regulatory regime to complement competition law.

Proceedings against Amazon have been initiated more recently: In 2017, the Commission accepted commitments from Amazon to refrain from enforcing and using MFN clauses that restricted competition in e-book distribution [32]. In 2019, it opened proceedings taking issue with Amazon’s systematic reliance on commercially sensitive non-public business data of independent third-party sellers on its marketplace to the benefit of its own retail business which directly competes with those sellers [33]. According to the Commission’s Statement of Objections, Amazon’s use of non-public marketplace seller data “allows Amazon to avoid the normal risks of retail competition and to leverage its dominance in the market for the provision of marketplace services in France and Germany” [34]. In a second proceeding against Amazon, the Commission investigates whether the conditions and criteria governing the selection of the winner of the “Buy Box” and the eligibility of third-party sellers to offer products under the Prime label lead to preferential treatment of Amazon’s retail business and of marketplace sellers that use Amazon’s logistics and delivery services [35].

In 2020, the Commission has also opened proceedings against Apple again. A first case is driven by a concern that Apple’s terms and conditions, as well as technical measures – namely the restriction of access to the NFC technology embedded in iOS mobile devices – serve to distort competition and reduce choice and innovation regarding mobile payment solutions on iPhones and iPads, where Apple systematically favors ApplePay, i.e. its own proprietary solution [36]. A second set of cases addresses Apple’s allegedly anticompetitive rules for app developers on the distribution of apps via the App Store. The Commission is concerned that the mandatory use of Apple’s own proprietary in-app purchasing system for the distribution of paid digital content such as music, e-books and audiobooks and restrictions on the ability of developers to inform iPhone and iPad users of alternative cheaper purchasing possibilities outside of apps (so-called anti-steering provisions) amounts to an abuse of Apple’s gatekeeper role [37] – a role that follows from the fact that Apple’s App Store is the only app store that iPhone and iPad users can use and that Apple device users typically have a high degree of brand loyalty and do not switch easily. On 30 April 2021, the Commission has sent out a Statement of Objections to Apple on its App Store rules for music streaming providers [38]. Also in 2021, Epic Games has filed an official complaint against Apple with the Commission, complementing its proceedings in the U.S., the UK and Australia (see above) [39].

Notably, Facebook has remained a blank spot on the Commission’s antitrust agenda against “Big Tech” until just recently. Its initial reaction to the Bundeskartellamt’s allegatin of exploitative data bundling practices that violate both the GDPR and German competition law (see below, II.3. lit. g) was sceptical. The proposed DMA seems to reflect a reassessment of the interaction between competition and data protection law in this regard. Its Art. 5 lit. a which prohibits the combination of personal data sourced from a core platform service of a gatekeeper with personal data from any other services offered by the gatekeeper or from third-party services replicates the Bundeskartellamt’s remedy in the Facebook case and would apply irrespective of whether the usual preconditions of an exploitative abuse are met or whether there is any plausible causal link between either the combination of data and anticompetitive effects, or between the undertaking’s market power and its ability to combine data pools [40].

On 4 June 2021, the Commission has opened its first own competition law investigation against Facebook: It proposes to examine whether the company restricts competition by using commercially valuable data from companies advertising their services on Facebook’s social networking platform to directly compete with these companies on “Facebook Marketplace” [41]. More particularly, Facebook could receive precise information on users’ preferences and use that data through the advertisements of online classified ads providers on the social network, and then use these data to adapt Facebook Marketplace. The Commission also plans to investigate whether the way Facebook Marketplace is embedded into the social network qualifies as a form of anticompetitive tying that may foreclose competing online classified ads services.

Finally, the Commission starts to turn its attention to the Internet of Things (IoT): A sector inquiry was opened in July 2020, and a preliminary report was published very recently, in June 2021 [42]. In reaction to the input received from market participants, the Commission has identified four main areas of concern, relating, in particular, to consumer IoT and the role of voice assistants in the wider IoT ecosystems established, inter alia, by Google, Amazon and Apple: (i) exclusivity and tying in relation to voice assistants; (ii) the position of voice assistants and smart device operating systems as intermediaries between users and smart devices or consumer IoT services; (iii) the extensive access to data of voice assistants and smart device operating systems, including information about user interaction with third-party smart devices and consumer IoT services; (iv) lack of interoperability in the consumer IoT sector due to the prevalence of proprietary technologies.

3. Public enforcement in the EU Member States

The parallel public enforcement of European competition law by both the Commission and national competition authorities (NCAs) is a special feature and strength of the European regime (see Art. 3 Regulation 1/2003) [43]. It has also played an important role in the digital context.

Obviously, there have been very different levels of activity by different NCAs in the digital field. A pro-active enforcement agenda has been pursued most notably by the Italian (Autorità Garante della Concorrenza e del Mercato - AGCM), the French (Autorité de la Concurrence) and the German authority (Bundeskartellamt). The UK Competition and Markets Authority (CMA) has significantly contributed to the policy debate mostly in other ways (see below, II.3. lit. i) [44]. These authorities – and to a more limited extent also the Austrian, Dutch and Hungarian competition authorities – have looked into potential anticompetitive conduct in digital markets which the EU Commission had not yet analyzed at the time – like advertising markets. They have tested legally innovative theories of harm – like a theory of “data exploitation”. And they have dealt with complaints by online publishers who have long advocated for an entitlement to remuneration where an online content sharing service provider – most notably Google – makes their content available to the public. Those authorities with a dual competence for both competition law and consumer protection law have explored the overlaps between competition law and consumer protection law. While the decisions taken in the different jurisdictions have not always been completely consistent, they have contributed significantly to the understanding of digital markets, to the identification of the relevant issues and to their discussion. Given this overall beneficial experience of parallel competition law enforcement in the digital realm, the draft DMA’s decision in favor of a fully centralized enforcement model comes as a surprise (see below, III.8).

a) MFNs on hotel booking platforms

The use of so-called “Most Favored Nation”-clauses by digital platforms, and in particular by hotel booking platforms, has become a shared concern of a relevant number of national competition authorities in the EU [45]. Originally, all the large hotel booking platforms that operate in the EU – Booking.com, Expedia and HRS – had inserted clauses into their terms and conditions that obliged the hotels active on the platform to ensure that the prices and terms quoted through the platform would not be higher than those available on any other platforms or distribution channels (“wide MFN”). The first European agency to take action against the use of these “wide MFNs” by hotel booking platforms was the German Bundeskartellamt: In a proceeding against HRS, it found, on 20 December 2013, that such clauses ran counter to Art. 101 TFEU and could not be justified [46].

In 2014, the Italian AGCM opened proceedings against Booking.com and Expedia on similar grounds. But in collaboration with the French and the Swedish competition authority and with the backing of the Commission, they all finally accepted the platforms’ commitment to replace the “wide MFN” clauses with “narrow MFN” clauses – i.e. an obligation of the hotels active on the platform to make sure that the prices and terms quoted through the platform would not be higher than those available on each hotel’s own website – in 2015 [47].

The Bundeskartellamt, by contrast, took a more restrictive approach. Eight months after its peer authorities had accepted “narrow MFNs”, it prohibited, in a proceeding against Booking.com, not only the broad, but also the narrow version of price parity clauses [48]. While that decision was quashed by the Oberlandesgericht Düsseldorf in 2019 [49], the Bundesgerichtshof has recently confirmed the Bundeskartellamt’s view, finding that narrow price parity clauses are not objectively necessary for the online brokerage of hotel rooms and cannot be justified under Art. 101(3) TFEU [50].

In Italy [51], France [52], Austria [53] and Belgium [54], legislative initiatives have been taken to prohibit MFN clauses, including narrow MFNs, in agreements between online travel agencies and hotels.

The different approaches taken re the use of MFNs by hotel booking platforms have repeatedly been taken as evidence that the existing regime of parallel enforcement risks to lead to fragmentation in the European internal market – which may come at high cost in particular in digital markets where platform business models will frequently be pan-European and where their viability may depend on scale [55]. But from a different perspective, the national enforcement action against the use of price parity clauses by hotel booking platforms may rather be seen as an example of the strength of a parallel enforcement regime, with NCAs filling the gaps that a centralized enforcement regime left: the Commission had failed to take action, including against the use of “wide MFNs”. As regards the different approaches taken by the NCAs: According to Art. 11(6) Regulation 1/2003, the Commission retained the possibility to initiate proceedings itself and thereby relieve the NCAs from their competence to apply Art. 101 TFEU so as to ensure a consistent interpretation and application of Art. 101 TFEU throughout the EU.

b) Price parity clauses by Amazon and other platforms

The competition concern with MFNs – or parity clauses – is not limited to hotel booking platforms [56]. In 2013, both the CMA [57] and the Bundeskartellamt closed proceedings against Amazon after Amazon abandoned the use of price parity obligations for sellers on the platform [58]. The case was dealt with under Art. 101 TFEU by both authorities. In 2020, the CMA fined the price comparison website ComparetheMarket for imposing wide MFN clauses on home insurers in breach of Art. 101 TFEU [59].

c) Exclusivity clauses and tying/bundling

In 2017, the Bundeskartellamt – working in close cooperation with the Commission – closed Art. 101 TFEU-proceedings against Audible.com (a subsidiary of Amazon) and Apple on a long-term agreement under which Apple exclusively purchased audiobooks from Audible for sale in the iTunes Store and Audible did not supply any other digital music platform other than iTunes. Proceedings were closed after the parties had agreed to abandon the exclusivity clauses [60].

In a case against Google, the Italian AGCM found that Google, by preventing Enel X Italia from developing a version of its e-car app JuicePass that would be compatible with the Android Auto environment, had violated Art. 102 TFEU [61]. JuicePass is an app that enables the driver of an e-car to make use of a wide range of services in the context of re-charging the vehicle. Google had failed to make available tools for programming that app such as to make it interoperable with Android Auto. Also, the app was not included in the list of apps used by users, thereby limiting consumer choice. Instead, Google favored its own app – Google Maps – that does run on Android Auto. While Google Maps does not yet have the full functionality of JuicePass, its functionalities could be expanded. The AGCM imposed a fine on Google and ordered it to stop the abuse.

National competition authorities have also inquired into other bundling practices. In December 2020, the Bundeskartellamt has opened abuse of dominance proceedings against Facebook with a view to the requirement that users must have a Facebook account in order to use the Oculus glasses [62].

d) The display of press content by Google

Another issue that has been investigated by various NCAs in parallel is Google’s display of excerpts from online press content in search result lists without remuneration for the publishers.

In 2015, the German Bundeskartellamt decided not to initiate proceedings against Google regarding its practice to display snippets of online press content in search result lists without offering a publishing fee to the publishers and regarding Google’s request to the publishers represented by VG Media, a collecting society, to confirm that they agreed to that practice – otherwise Google would have curtailed the display of the relevant websites in the result list in such a way that only the headline, but no small snippets would appear [63]. The Bundeskartellamt left open the exact market definition and the question whether Google actually had a dominant position on any of the relevant markets. Even assuming dominance, it found it unlikely that an abuse could be established. The Bundeskartellamt underlined that a search engine, even if dominant, cannot be subject to a strict non-discrimination requirement with regard to the compilation, ranking and presentation of search results. Given the many criteria, combinations of criteria and their weighing, and given that an abuse control of these choices would directly affect the product design and risk to chill any further development, a search engine must be allowed considerable scope of action. Based on a balancing of the legitimate interests in consideration of the purpose of competition law to protect freedom of competition, the Bundeskartellamt found it unlikely that the publishers’ interest in receiving remuneration for the use of their ancillary copyright would outweigh Google’s interest, given that Google had not meddled, in a self-preferential form, with the criteria of relevance of organic search results.

By contrast, the Autorité de la Concurrence has ordered Google, by way of an interim measure published on 9 April 2020, to enter into good faith negotiations with French publishers and news agencies on a fair remuneration for the re-use of their protected contents. It considers that Google holds a dominant position on the French market for general online search, and that Google’s rejection to enter into negotiations with publishers and news agencies on remuneration for the re-use of their protected contents is likely to qualify as an abuse, namely as an imposition of unfair trading conditions (Art. 102 lit. a TFEU), and possibly as a discriminatory practice within the meaning of Art. 102 lit. c TFEU [64].

So far, the different positions of NCAs on this issue have stirred less debate than their different takes on the use of MFNs. The conflict between Google and the national press tends to play out primarily at the national level. However, Art. 15 DRM Directive 2019/790 [65] now obliges the Member States to provide for an ancillary copyright for publishers for the online use of their press publications by information service providers. While the protection does not apply to acts of hyperlinking, Google has announced to enter into negotiations with publishers on the compensation for the use of snippets, following the implementation of the DRM Directive into German copyright law [66].

This does not appear to put an end to the competition law concerns against Google’s display of media content practices, however. On 4 June 2021, the Bundeskartellamt has announced that it is investigating the Google News Showcase service which enables some German publishers to showcase content more prominently and in greater detail [67]. The service focuses on “story panels”, which were initially integrated in the Google News app and can now also be found on the desktop in Google News and also in the general Google search results. For this service, Google also purchases paywalled content from publishers and offers them to its readers free of charge. The Bundeskartellamt’s investigation follows complaints from Corint Media that the integration of Google News Showcase into Google’s general search function is likely to constitute self-preferencing or an impediment to third-party services of competitors. Furthermore, the authority is examining the contractual terms of the service, in particular whether they make it disproportionately difficult for participating publishers to enforce their ancillary copyright for press publishers (see above).

e) Advertising markets

Despite the huge economic importance of online advertising for the business models in particular of Google and Facebook, and despite the central position that Google occupies in these markets and significant competition concerns regarding its strategies [68], competition law inquiries are only starting.

The Italian AGCM initiated actions against Google as early as 2009, however, namely with a view to the Google News Italia service which collected, indexed and partially displayed news items of online publishers with hyperlinks to the editor’s webpage [69]. Following a complaint by the Italian Newspaper Publishing Federation (FIEG), the agency investigated whether Google was abusing its dominant position in the online advertising market and in the market for intermediation of online advertising by unilaterally determining the degree of visibility of content. The other allegation raised by FIEG related to the fact that editors could not control which content was indexed and made available, but only had the choice of either granting unrestricted access or preventing Google from collecting any information while risking being also excluded from the Google search engine. The investigation was closed when Google agreed to allow publishers to remove or specify content on Google News Italia without being removed from Google search results [70], and to manage AdSense, its advertising solicitation platform, more transparently.

Subsequently, the French Autorité de la Concurrence has been the most active NCA in looking into potential abuses of dominance in this field on a case-by-case basis. The enforcement activities started in 2010, when the Autorité de la Concurrence considered that the suspension of advertisement of Navx [71] and ultimately a termination of its AdWords account based on terms and conditions that were opaque and granted broad leeway to Google including to unilaterally change the rules of the contract potentially amounted to an abuse of dominance, namely a discrimination that was prone to affect competition in the relevant markets. The case was closed when Google accepted commitments to make its AdWords terms more objective and more transparent [72]. In 2013, another complaint against Google’s AdWords practices was rejected [73]. But a complaint by Gibmedia was ultimately successful, although interim measures had been denied [74]: In 2019, the Autorité de la Concurrence found that the suspension of Gibmedia’s Google AdWords account was based on terms and conditions unilaterally set by Google that were unclear in their formulation, such that Google enjoyed broad discretion in their interpretation; that Google interpreted and applied these rules incoherently from case to case, resulting in discriminatory suspensions of some advertisers, but not others; and that the rules were subject to various modifications without advertisers being informed. Although no deliberate and comprehensive strategy to disrupt competition downstream was established, the French competition authority fined Google for an abuse of dominance in the market for online search advertisement, as these practices had damaged sites with low visibility [75]. A similar case, this time based on a suspension of the AdWords account of Amadeus, is still pending. The Autorité de la Concurrence has ordered interim measures, requiring Google to clarify its AdWords terms and conditions on conduct that can lead to a suspension of an advertiser’s account in certain specified respects [76]. On 7 June 2021, the French authority fined Google for having abused its dominant position in the advertising server market for website and mobile applications publishers by granting preferential treatment to its proprietary technologies under the Goolge Ad Manager brand with regard to the DFP ad server (which allows the sale of advertising space), and its SSP AdX sales platform (which organizes the auction processes) [77]. Simultaneously, the Autorité de la Concurrence accepted Google’s commitments to improve the interoperability of Google Ad Manager services with third-party ad server and advertising space sales platform solutions.

A more traditional case – and therefore unusual in the digital advertising setting – has been decided by the Danish Konkurrence- og Forbrugerstyrelsen in 2020: According to the Danish authority, FK Distribution had abused its dominant position in the market for distribution of unaddressed mail (circulars) by tying its sale of distribution of print circulars to the sale of digital circulars by obliging customers of print circulars to also advertise online [78]. FK Distribution was ordered to cease this practice.

The German Bundeskartellamt has launched a sector inquiry into the online advertising sector in 2018. The inquiry is still ongoing [79].

f) Marketplace rules

The rule-setting activity of digital platforms – a core concern of the Autorité de la Concurrence in the online advertising sector – has also been a concern in other contexts and has led to a significant enforcement activity by a number of NCAs.

Both the Austrian Bundeswettbewerbsbehörde and the German Bundeskartellamt have taken a closer look at Amazon’s terms and conditions. The Bundeskartellamt’s proceedings followed up on a large number of complaints by sellers on the platform about, inter alia, non-transparent conditions and abrupt unilateral changes to the general contract terms without sufficient prior notice; about a choice of law and court of jurisdiction clause which provided that Luxembourg was the only court of jurisdiction; about a very far-reaching limitation of liability in Amazon’s favor, combined with an extensive liability of the sellers and an obligation to indemnify Amazon from any third party claims; about Amazon’s unlimited right to immediately terminate contractual relations with sellers or to block them and to suspend sellers’ payment accounts without justification; about Amazon’s rule on returns and reimbursements that make sellers bear the costs; and about Amazon’s practices regarding product reviews that favor Amazon’s own sales. The Bundeskartellamt closed its proceedings after Amazon had amended its general terms of business for sellers on its marketplace in the relevant respects [80]. The investigations by the Austrian Bundeswettbewerbsbehörde addressed very similar concerns, including the abrupt termination of seller-accounts, the obligation to disclose purchase prices or the intransparency of product rankings [81]. Again, Amazon agreed to modify its terms and conditions, and the proceedings were closed [82].

This does not mean that Amazon is “off the hook”: Several proceedings against Amazon are currently pending before the Bundeskartellamt concerning, inter alia, a suspension of sellers who had allegedly offered products at “excessive prices” or other forms of “price control” by Amazon which may serve as a substitute for the former MFNs [83]; and proceedings against agreements between Amazon and brand manufacturers such as Apple that may result in an exclusion of third-party sellers from selling branded products on Amazon [84].

Similarly, the Italian AGCM has initiated an Art. 101 TFEU-proceeding into the ‘brand-gating’ agreement between Apple and Amazon, which prohibits the sale of Apple and Beats-branded products on Amazon by electronics retailers that are not members of Apple’s official program [85].

Apart from Amazon, Apple – and particularly the Apple App Store – has come into focus. The Netherlands Autoriteit Consument en Markt (ACM) has launched investigations into Apple’s App Store practices, including the requirement to use Apple’s payment system for in-app purchases, the 30% commission for all turnover generated through the App Store and through in-app purchases, and its restriction of access to some of the iPhone’s functionalities [86]. The CMA – now no longer part of the European Competition Network (ECN) – has launched an investigation into the terms and conditions applying to the registration as a developer on the App Store, in particular the required agreement to distribute apps only through Apple’s App Store, to use only Apple’s payment system, and to pay the 30 percent ’Apple tax’ [87].

Self-preferencing by vertically integrated platforms has rarely been addressed at the national level so far. But in 2019 and 2020, the Polish Urząd Ochrony Konkurencji i Konsumentów (UOKiK) launched proceedings against Allegro (the operator of the most popular online-shopping platform in Poland, where it also sells products as a retailer), which may have abused its dominant position by favoring its own sales over those of other retailers on the platform [88].

The Dutch ACM has been the only agency to take a closer look at video streaming platforms such as Youtube or Netflix. However, none of these platforms were found to be dominant on either the online advertising market or the online video market [89]. Nonetheless, the ACM discovered unfair terms and conditions in its market study and announced that it would conduct further investigations into such conditions.

g) Data-related practices

As control of and access to data has become ever more central for the functioning of digital markets, data related unilateral conduct has become a focus of attention not only in academic debates [90]. However, for the moment, data-related competition law charges remain the exception, although the attention paid to such practices is increasing.

Arguably the most prominent proceeding in the field so far is the Bundeskartellamt’s Facebook case. It continues to occupy the German courts – and meanwhile also the CJEU. In 2019, less than three years after the proceedings had been opened, the Bundeskartellamt ruled that Facebook had abused its dominant position in the social network market by making the use of its social network conditional on being allowed to collect user data from sources outside of Facebook and to merge it with data collected on Facebook, in an alleged violation of data protection rules [91]. Facebook was therefore prohibited from combining user data from Facebook with data from Facebook-owned services (e.g. WhatsApp and Instagram) and third-party websites. Facebook has challenged the decision before the German courts. The OLG Düsseldorf has meanwhile referred the case to the CJEU, questioning the competence of the BKartA to take position on issues of data protection law [92].

The Bundeskartellamt’s Facebook decision has led to an EU-wide debate on how to conceive of the interface between EU competition law and data protection law. Can a violation of data protection law by a dominant company acting in a data-driven market be considered an abuse of dominance? Under what conditions would an excessive collection and processing of personal data be considered an exploitative abuse [93]? Although the Bundeskartellamt had based its Facebook decision on national “abuse of dominance” law only [94], the case has thereby become an important test case, potentially impacting the evolution of EU competition law, too.

A second line of data related cases is just starting after complaints were filed with the Autorité de la Concurrence and the Bundeskartellamt [95]: According to press reports, Apple has come under competition law scrutiny because of its App Tracking Transparency framework, which requires apps to obtain user permission for tracking through a pop-up window. Along similar lines, the CMA has opened proceedings against Google’s plan to remove third party cookies (TPCs) on its Chrome browser and replace their functionality with a range of “Privacy Sandbox” tools [96].

In June 2021, the CMA has started another data-related case, this time against Facebook. The case partly resembles the Commission’s recently opened Facebook proceedings (see above). The CMA is investigating whether the company abused its dominant position in the social media or digital advertising market through its collection and use of advertising data, in particular by using the data gained from advertising and single sign-on to benefit its own Facebook Marketplace and Facebook Dating services [97]. The Commission and the DMA have announced their intention to closely cooperate.

Furthermore, the Italian AGCM has launched an Art. 102 TFEU investigation into Google’s data access policies on the market for online advertising: Allegedly, Google denies its competitors access to its ID decryption keys and to third-party tracking pixels for the targeting of their display advertising campaigns, while at the same time using data collected through its various applications, in particular through tracking elements enabling its advertising intermediation services, to achieve a higher targeting capability. In such a setting, equally efficient competitors may not be able to compete effectively without being granted access to Google’s vast amount of data on non-discriminatory grounds [98].

h) The interface between competition and consumer protection law

Protecting competition between platforms and ensuring the contestability of a platform’s gatekeeper positions presupposes a forceful protection of informed consumer choice, multi-homing and switching. This has raised awareness of the close interaction between competition and consumer protection law in digital markets. This is also reflected in the enforcement practice of those NCAs that dispose of dual enforcement competencies, i.e. that do not only enforce competition law but also consumer protection law. Seemingly, NCAs with a parallel competence for consumer protection law enforcement frequently prefer to deal with consumer-facing data-related practices under consumer protection law instead of competition law.

The CMA has taken action against online travel agencies on 18 June 2018 under on consumer protection law, based on concerns about the ranking of search results, pressure selling practices (i.e. creating a false impression of room availability), discount claims and hidden charges [99].

Also in 2018, the AGCM, fined Facebook for violating the Consumer Code by misleading consumers into unconscious and automatic data sharing [100]. The Hungarian Gazdasági Versenyhivatal (GVH) fined Facebook in 2019 for advertising its service as free of charge, while Facebook benefitted economically from users’ data [101]. In 2020, it initiated proceedings against TikTok for failure to provide users with sufficient information on the scope of data being collected and processed [102]. The CMA, along with the ACM and the Norwegian Consumer Protection Authority, led 24 other consumer protection agencies in launching investigations into Apple’s App Store on consumer protection grounds in 2020. The concern was that users were not given clear information about the use of their personal data before selecting an app and were thereby unable to compare and choose apps based on the use of personal data [103]. In reaction, Apple has agreed to give clear indications on the personal data used by each app in its App Store.

The Italian AGCM has been another particularly active enforcer of consumer protection law in the digital realm. Inter alia, it has brought a case involving Google, Apple and Dropbox in 2020, alleging unfair commercial practices and unfair terms in contractual conditions [104]. Again, the investigations relate to an insufficient indication of which user data is collected, resulting in the inability of consumers to give informed consent. Furthermore, some contractual terms are under review, such as the operators’ wide-ranging powers to suspend and interrupt the service, or their exemption from liability. Dropbox in particular is accused of failing to provide clear and readily accessible information about the terms, conditions, and procedures for withdrawing from the contract.

i) Additional action by the NCAs

NCAs have also been involved in the broader debates about competition law enforcement in digital markets. The German Bundeskartellamt has had its part in the debates preceding the recent reform of the German competition code (GWB) [105]. It has now initiated the first proceedings under the new § 19a GWB which foresees a possibility for imposing ex ante obligations upon platforms of paramount cross-market relevance. So far, designation proceedings under § 19a(1) GWB have been opened against Facebook [106], Amazon [107], Google [108] and Apple [109].

The CMA has been another highly influential actor in the competition policy debate on digital platforms. The so-called Furman Report [110] has famously been among the first to call for a regulation of digital platforms with a “strategic market status” – an idea that is the basis of the UK Digital Market Taskforce’s proposal for a new pro-competition regime for digital markets [111], but is also reflected in the EU Commission’s proposal for a Digital Markets Act. Furthermore, the CMA has engaged in important market investigations, most prominently into online platforms and digital advertising [112]. The UK market investigation regime has inspired a debate at EU level about whether a “New Competition Tool” is needed to address competition problems that are not addressed effectively by EU competition law, including in the digital field [113]. While the UK is no longer part of the EU, it shares its competition law tradition, and its competition policy ties with the EU remain close. It remains to be seen how a future UK platform regulation regime would interact with the EU’s [114].

4. Public enforcement abroad

Public competition law enforcement in the digital realm – and particularly relating to unilateral conduct of the “Big Tech” – has not been limited to the EU. However, the enforcement agendas outside the EU, have, for the most part, been more confined and stepped up only recently. The year 2020 appears to mark a turning point in many jurisdictions, including the U.S., Australia, Canada and Japan. In 2021, the Chinese State Administration for Market Regulation (SAMR) has announced its intention to intensify its scrutiny of the Chinese platform economy (see below, II.4. lit. h). The Indian and the Russian competition authority have started to engage somewhat earlier.

a) U.S.

The U.S. agencies had an early start: Shortly after the EU Commission, [115] namely in June 2011, the FTC began to review a number of Google’s business practices. Complaints that Google had refrained from licensing standard-essential patents (SEPs) that were needed to produce devices such as smart phones, laptop and tablet computers and gaming consoles resulted in a commitment by Google to grant licenses on fair, reasonable and non-discriminatory (FRAND) terms. Google also agreed to remove restrictions that hampered online advertisers in their management and coordination of ad campaigns across different advertising platforms, i.e. across Google’s AdWords platform and rival platforms. Finally, Google committed not to misappropriate online content from “vertical” search websites that focus, inter alia, on shopping or travel, for use in its own vertical offerings [116]. Like the EU Commission in its Google Shopping case, and partly in cooperation with the EU Commission, the FTC also investigated whether Google had altered its search algorithm so as to demote competing vertical websites in order to reduce or eliminate a nascent competitive threat. But contrary to the EU Commission, the FTC unanimously closed these proceedings in 2013, finding that all changes made to the algorithm could plausibly be justified as innovations that improved Google’s product and the experience of its users [117]. Just recently, internal staff papers and memos that informed this decision have been published by POLITICO [118].

Following the Google investigations, the antitrust authorities refrained from initiating further action for the next seven years. In 2020, the DoJ along with several States, then filed a complaint against Google in the District Court for the District of Columbia under Sec. 2 Sherman Act, alleging that Google forecloses competitors and protects its monopolies in the search services, search advertising, and general search text advertising markets through exclusionary agreements with distributors, and requesting “structural relief as needed to cure any anticompetitive harm” [119]. Among other things, the DoJ complaint zooms in on Google’s attempt to protect its general search and search advertising monopolies by, inter alia, requiring manufacturers and distributors of Android phones and tablets to set Google as the default general search engine and to make its entire suite of search-related apps undeletable, and by engaging into advertising revenue sharing agreements that prohibit the preinstallation of competing search services. The case bears resemblance with the EU’s Google Android case.

The FTC, for its part, has sued Facebook in 2020 for a violation of Sec. 2 Sherman Act, in parallel with U.S. State Attorneys General [120]. The complaints allege that Facebook has harmed competition and maintained its monopoly power in the social networking market through its acquisitions of Instagram and WhatsApp, as well as through anticompetitive terms of access to APIs that require developers not to work with competitors, and through cutting off access for app developers when promising apps have the potential to become competitive threats. The remedies proposed by the FTC include a divestiture of WhatsApp and Instagram, and a prior notice and approval obligation for future acquisitions.

One year earlier, in 2019, the FTC brought charges in the District Court for the District of Columbia against health information company Surescripts for monopolizing the e-prescription routing and eligibility markets by using a series of exclusionary agreements, threats and other exclusionary tactics aimed at preventing users on both sides of each market from using additional platforms, i.e. from multi-homing [121].

There is a growing enforcement activity also at the states’ level. On 25 May 2021, the Attorney General for the District of Columbia filed an antitrust complaint against Amazon in the Superior Court of the District of Columbia for abusing its market power in the U.S. online retail sales market [122]. The claim is that – although Amazon officially abandoned MFNs in the U.S. in 2019 – its “Fair Pricing Policy” still allows it to sanction and even ban third-party sellers if they list their products at lower prices on their own websites or other shopping websites [123]. In addition, Amazon is accused of imposing a complex scheme of unreasonably high fees and extra charges. According to the complaint, these practices result not only in higher prices, but also in less choice for consumers and third-party sellers in the online retail sales market, in a suppression of innovation and in reduced investments in potentially competing online retail sales platforms.

While the Majority staff report of the U.S. House of Representatives on competition in digital markets (2020) focused on general antitrust law reform, the U.S. – like the EU – now appears to shift towards targeted legislative action against the large digital platforms. In June 2021, five far-reaching antitrust bills [124] have been introduced by House members – with some significant bipartisan support. The proposed “Ending Platform Monopolies Act” would provide for the structural separation of platforms if the platform operator owns or controls another line of business and such vertical or conglomerate integration would give rise to an irreconcilable conflict of interest. The proposed “American Innovation and Choice Online Act” would target the ability of online platforms to favor their own products over those of competitors. The proposed “Platform Competition and Opportunity Act“ would make it more difficult for dominant platforms to meet competitive threats through mergers and acquisitions by shifting the burden of proof in merger control proceedings. The “Merger Filing Fee Modernization Act”, would vest antitrust authorities with further resources by adjusting pre-merger filing fees. The “Augmenting Compatibility and Competition by Enabling Service Switching (ACCESS) Act” seeks to lower barriers to entry and switching costs for businesses and consumers through interoperability and data portability requirements.

b) Canada

Like the U.S. agencies, the Canadian Competition Bureau has kept a comparatively low-key profile in the digital field for a long time [125]. It has, however, initiated proceedings under the Competition Act’s abuse of dominance provisions against Amazon in 2020 and invited market participants to provide input [126]. The focus is on restrictions of third-party sellers to offer their products through other retail channels, the ability of third-party sellers to succeed without using Amazon’s “Fulfilment by Amazon” service, and strategies that may influence consumers to prefer products offered by Amazon over those offered by third-party sellers.

The Canadian Competition Bureau has also been an actor in the public policy debate on the role of competition policy in digital markets. In 2017, it published a draft discussion paper on Big Data and innovation, arguing that the application of the Canadian Competition Act must remain case-specific and that the traditional competition law and law enforcement framework is, in principle, fit to address the challenges arising from new business practices in the digital sphere [127]. In its “Strategic Vision for 2020-2024”, the Bureau advocates a “more proactive intelligence-led approach” to competition law enforcement, including in particular the use of new digital tools for enforcement purposes [128].

c) Australia

In December 2017, the Australian Competition & Consumer Commission (ACCC) was directed to consider the impact of online search engines, social media and digital platforms on competition in the media and advertising markets. In June 2019, it has published its “Digital Platforms Inquiry” [129] which has attracted great international attention. Among other things, the ACCC has recommended changes to the Australian merger law and has signaled its intention to investigate and monitor digital platform markets more proactively. It further recommended to inquire into the supply of ad tech services and advertising agencies and to create codes of conduct for designated digital platforms that should govern the relationship between these platforms and media businesses. In addition, the ACCC is currently conducting inquiries into digital advertising services [130] and into digital platform services [131]. In a first Digital Platform Services Inquiry interim report of 28 April 2021, the ACCC has expressed concerns that, as Apple and Google do not only run their respective app stores and control the terms of access to them, but also compete with their own apps, they have both the ability and incentive to promote their own apps over competing ones [132]. Apple and Google should therefore be prevented from using information collected about third-party apps to the advantage of their own competing apps. And app developers should have more information about how their apps are made discoverable to consumers, while consumers should have the ability to change or remove any pre-installed or default apps. Also, Apple and Google should leave developers free choice as regards the payment system to be used for in-app purchases.

On the enforcement side, the ACCC has mainly been active on consumer protection grounds so far: In 2019, it sued Google in the Federal Court of Australia, alleging that Google had engaged in misleading conduct and had made false or misleading representations to consumers about how and when it collected, kept and used their personal information in relation to location data collected through Android mobile devices. On 16 April 2021, the Federal Court largely followed the ACCCs arguments and found against Google [133]. The ACCC also obtained orders in the Federal Court of Australia against Viagogo for making false or misleading representation in relation to the resale of tickets to live music and sports events [134], and against iSelect in relation to its electricity comparison service [135]. On 3 August 2020, the ACCC initiated a second case against Google alleging that Google misled users to obtain their consent to expand the scope of personal data that Google could collect and use (e.g. for targeted advertising) [136].

d) India

The Competition Commission of India (CCI) has started a series of enforcement actions in the digital realm with proceedings against WhatsApp. The alleged abuse was reminiscent of the allegations in the Bundeskartellamt’s Facebook case: The CCI investigated whether WhatsApp abused its dominant position by requiring users to allow Facebook to access their data. In 2017, the CCI ultimately rejected an abuse, however, as WhatsApp provided an opt-out option [137]. A second case was also decided in favor of WhatsApp: In 2020, the CCI found that, prima facie, WhatsApp is dominant in the market for OTT messaging apps through smartphones in India, but that tying the WhatsApp messenger app to the installation of the WhatsApp Pay app did not constitute an – exploitative or exclusionary – abuse of dominance [138]. In order to make use of the WhatsApp Pay app, users needed to register separately. The mere incorporation of the payment option into the messaging app therefore did not seem to influence consumers’ choice. In addition, various players were active in the digital payments market, including Google Pay, Amazon Pay and PayTM, all backed up by big companies and/or investors. In such an environment, it seemed implausible to the CCI that WhatsApp Pay would automatically get a considerable market share only on the basis of its pre-installation. Just recently, WhatsApp’s privacy policy has come under scrutiny again. The CCI has launched an investigation into the new privacy update of January 2021, by which WhatsApp can share some user data with Facebook and its subsidiaries [139].

In 2020, the CCI also turned down a complaint against the acquisition of Wibmo – a leading and arguably dominant technology and service provider for the financial industry, best known for its risk-based authentication and payment security services – by PayU payments – a fintech company that provides payment technology solutions to online merchants. The complainant had argued that this vertical combination would amount to an attempt to monopolize the e-payment gateway market in India, as it would likely result in unfair or discriminatory conditions in availing e-payment processing gateway services in India. The CCI held that the impugned combination and the securing of a dominant position as such cannot qualify as an “abuse of dominance”. Rather, some abusive conduct must be shown. In the case at hand, the alleged foreclosure of competitors by denying them proper access to Wibmo in the downstream market had not been proven [140].

In 2018, the CCI fined Google for abusing its dominant position on the online web search and online search advertising markets in India by engaging in “search bias” practices on its search engine results page, in particular by placing its commercial flight search function prominently on the search results page to the detriment of competitors [141]. In 2019, the CCI found that Google had abused its dominant position in the mobile operating system market by tying Android to the pre-installation of a wide range of its own applications [142].

Also in 2018, the CCI turned down a complaint that Flipkart, a firm that operates an online e-commerce platform for the selling of goods B2C in India, had abused its dominant position in the market for services provided by online marketplace platforms by facilitating discounts and by leveraging its position to enter into another market for manufacturing products through private labels. However, the authority found that Flipkart was not dominant, as a number of competitors – Flipkart and Amazon being the bigger ones – were competing and other players like Paytm Mall had managed to enter the market recently. Even if Flipkart had been dominant, the CCI would not have found an abuse, as the arrangements between the platform and the sellers on the platform did not entail any exclusivity requirements and the terms and conditions were otherwise standard, entailing incentive structures based on objective criteria such as the quality of a product and the volume and value of sales. Given that e-commerce-platforms were still a relatively nascent and evolving model of retail distribution in India, the CCI emphasized that “any intervention in such markets needs to be carefully crafted lest it stifles innovation”.

Along related lines, the authority dismissed a case against Amazon in 2020: Sellers had accused Amazon of deep discounts, preferential treatment of some sellers, and of allowing counterfeit products. The CCI found that Amazon was not dominant on the online fashion retail market in India, however [143].

In a case against Oravel Stays, a booking platform for budget accommodation that functioned on the basis of a franchise model, the CCI found that Oravel Stays, although having the largest budget hotel network in India, was not dominant on the market for franchising services for budget hotels in India, given that hotels retained the choice of business model, all the more given the presence of online travel agencies (OTAs). Had it been dominant, neither its terms and conditions, which allowed Oravel Stays to exert influence on the service standard provided by its partners through its scoring policy, nor the mandatory use of its trademark, were abusive as they had a logical business justification, namely the protection and promotion of the brand [144].

e) Russia

The Russian Federal Antimonopoly Service (FAS) has been among the comparatively proactive competition authorities in digital markets, with a particular focus on Google and Apple. In September 2015, and after an only 7-months long proceeding, it found that Google had abused its dominant position in the Russian market for mobile operating systems by tying the Android OS to the pre-installation of Google applications, including Google Search and the Google Play Store, requiring priority placement of Google’s apps and prohibiting the pre-installation of competing apps [145]. The FAS imposed a fine and ordered the cessation of the abusive practices. The decision was upheld by the Moscow Arbitration Court and the 9th Arbitration Appeal Court, both in 2016 [146]. Further proceedings were initiated in 2016 because Google had failed to execute the FAS injunction. They were brought to an end in 2017 with a settlement agreement approved by the Moscow Arbitration Court by which Google agreed to no longer demand exclusivity of its applications for Android-based devices [147].

Actions against Apple followed in 2017 and 2020. In 2017, the FAS issued a warning in view of excessive repair costs for Apple products, possibly amounting to an abuse of dominance, and requested Apple to open a service center within little more than two months [148]. Three years later, in 2020, and following a complaint from Kaspersky Lab that Apple had declined to grant a new version of its Safe Kids app access to its App Store, the FAS found that Apple had abused its dominant position in the market for distribution of mobile applications on the iOS operating system by restricting the functionalities of third-party parental control applications, and by granting Apple’s Screen time app access to iOS technology capabilities there were not available to third-party developers [149]. The FAS ordered Apple to ensure that in-house apps do not take precedence over third-party apps, that parental control app developers can distribute their applications without loss of important functionalities and to remove provisions that gave it the right to keep third-party apps out of its App Store. In addition, a fine was imposed [150].

In November 2019, the FAS warned Booking.com against using wide price parity clauses on its hotel booking platform, [151] and in September 2020 again against using narrow price parity clauses [152]– in both cases based on a suspected abuse of dominance. Booking.com was asked to drop the clauses.

f) Japan

The Japanese Fair Trade Commision (JFTC) has been among the more cautious competition agencies in the digital realm so far, but it has not been inactive:

In 2017, it closed investigations against Amazon after the online marketplace had abandoned the use of price parity clauses and selection parity clauses in seller contracts [153]. Further proceedings against Amazon were initiated in 2020 after allegations were made that the company discounted products and let sellers cover some of the price differences. The case was closed after Amazon offered commitments, in particular to refund Japanese sellers who had previously had to pay the price differences, as part of the implementation of an “improvement plan” [154].

On 28 February 2020, the JFTC filed a petition for an interim injunction against the online marketplace Rakuten in the Tokyo District Court over its “Shipping Inclusive Program Measures” that promoted the free shipping of all orders over $38 [155]. The petition was withdrawn on 10 March 2020 when Rakuten promised to let the sellers decide whether or not to participate in the program.

g) South Korea

The South Korean Fair Trade Commission (KFTC) has repeatedly taken inspiration from the EU in its competition law enforcement activities [156]. In 2005, the KFTC fined Microsoft for abusing its dominant position in PC operating systems by tying its operating system to its Windows media player and its Windows Messenger [157]. Like the EU Commission, the KFTC ordered Microsoft to offer two versions of its operating system, with and without its proprietary applications.

In 2013, the KFTC cleared Google of charges of an abuse of dominance in the mobile operating system market by tying Android to the preinstallation of its mobile search engine. The authority considered that anticompetitive effects were unlikely, as Goolge’s market share in the mobile search market was 11.7% only and declining [158]. Yet, the KFTC reopened an investigation against Google in 2016, shortly after the Commission’s Statement of Objections in the Google Android case [159]. Currently, the KFTC is investigating whether Google is abusing its dominant position in the app store market by pressuring developers of local mobile game applications to release their games exclusively in its Play Store according to press reports [160].

h) China

The competitive setting in digital markets in China differs significantly from that of almost the entire rest of the world: the main players here are not Google, Amazon, Facebook and Apple, but Baidu (search engine), Alibiba (e-commerce) and Tencent (social networking).

Before the establishment of the Chinese State Administration for Market Regulation (SAMR) in 2018, there was little enforcement activity in China in the “Big Tech” realm. In 2010, the Chinese State Administration for Industry and Commerce initiated investigations into whether Baidu and Tencent are dominant [161], but nothing followed from this initiative. In 2016, the People’s Court in Shanghai Pudong New Area found that Baidu’s unauthorized use of third-party reviews on its own websites violated China’s Anti-Unfair Competition Law [162].

On 7 February 2021, the SAMR has published its “Antitrust Guidelines for the Platform Economy” (Platform Guidelines), signaling a willingness to step up competition law enforcement against abuses of dominance in China’s platform economy. On market definition, the Guidelines suggest that a precise definition of the relevant market may not always be required when investigating monopolistic behavior by a platform, but that it may sometimes be possible to identify anticompetitive conduct directly. Moreover, the Guidelines single out practices which are likely to be abusive [163]. Among them are the so-called “choose one from two” conduct – essentially an exclusive dealing arrangement by which a platform requires users active on its platform not to use competing platforms in parallel. In April 2021, Alibaba was fined for an abuse of dominance because it had prohibited merchants from selling products on rival platforms [164]. Furthermore, the Platform Guidelines suggest that, depending on the specific circumstances, a platform may be considered an essential facility, such that a refusal of access to the platform would amount to an anticompetitive refusal to deal. Based on this theory, TikTok’s parent company ByteDance has filed, on 7 February 2021, an antitrust lawsuit against Tencent with the Beijing Intellectual Property Court alleging that Tencent has blocked users from sharing content from ByteDance’s social networking platforms, thereby abusing its dominant market position and excluding or limiting competition [165].

The Platform Guidelines also mention “Big Data discrimination” as a possible abuse. “Big Data discrimination” may be found where a dominant platform charges different transaction prices or implements different transaction conditions for the same products/services, based on data on consumptions habits, ability to pay etc.; or where the platform implements differentiated rules, standards or payment conditions and transaction methods based on a Big Data analysis. The Guidelines thereby move significantly beyond the widely accepted categories of abuse of dominance.

On 13 April 2021, the SAMR ordered 34 companies, including Baidu and Tencent, to conduct self-inspections and to bring their activities in line with the Platform Guidelines within one month [166].

III. Insights from the case law – some tentative observations

More time and research would be needed to provide a full analysis of the rich case law on unilateral conduct in digital markets that has quickly accumulated across many jurisdictions over the last couple of years and is accumulating ever more quickly as we write. However, even this brief overview calls for some preliminary observations.

1. Changing gatekeeper concerns

To start with, the case overview serves as a reminder that the evolution of the new “gatekeeper” platforms is still a relatively recent phenomenon. In the early 2000s, the competition authorities still struggled with the large telecommunication incumbents that were entering the emerging broadband markets, potentially leveraging their hereditary monopoly power in the fixed telecommunication networks into the internet service provision markets: Internet Service Providers were the potential gatekeepers that competition authorities were particularly concerned about. It was only from 2010 onwards that digital platforms came into view – and this has been the focus of this paper. From 2018 to 2020, many jurisdictions have stepped up their enforcement efforts significantly. In some jurisdictions like the U.S. and China, more cases have been brought within the last year alone than in the last 10 years together.

Very recently, competition law concerns regarding the IoT – and in particular voice assistants – start to come into focus [167]. Also, more and more data-related competition law cases are being brought (see below, 5. lit. a).

Finally, recent cases provide evidence of a new wave of competitive advances among Google, Facebook, Apple and Amazon [168]. Strategic changes in Google’s and Apple’s approaches towards third-party tracking, Facebook’s promotion of its own marketplace and Apple’s attempt to establish and expand its own social networking functions [169] are examples in point. Increasingly, competition agencies around the world will need to grapple with the question whether to allow and promote competition among the “Big Tech” or whether to stick to a commitment for a broader opening of all the large ecosystem providers – for small and medium players, but also potentially for each other.

2. Public enforcement is key

From our sample, we can safely conclude that public enforcement of competition law is key in the digital realm. Private enforcement can be a complement. But the vast majority of cases has been handled by competition authorities. The reasons for this have been highlighted above: Where platform users depend on (fair and non-discriminatory) access to the platform, fear of retaliation may keep them from bringing their case to court. In other cases, the factual issues – e.g. surrounding the precise functioning of a ranking algorithm – may be too complex for private parties, or at least smaller private actors, to bring. Where novel legal issues are involved, the legal uncertainty may keep private parties from going to court.

All this implies that public enforcement is under pressure to function efficiently – in particular where systematic infringements occur. A well-functioning complaint mechanism, effective remedies and deterrent sanctions may be needed to ensure open and competitive markets.

3. Global convergence in the actions against GAFA(M)

From 2010 onwards, Google has experienced the greatest number of competition law actions by far across the different jurisdictions worldwide, followed by Apple and Amazon, more or less on par. Facebook has been the subject of significantly less enforcement action so far, and it has come under scrutiny only rather recently. Microsoft, for its part, has largely vanished from the competition authorities’ enforcement agenda, following the large international Microsoft cases of the early 2000s – in some respects precursors of the platform cases against Google and Apple that were then to come. It is only recently that Microsoft has been realigned with “GAFA”, as evidenced by a growing number of references to “GAFAM” [170].

The antitrust charges against “GAFA” have been manifold. But there are obvious similarities between the cases in various jurisdictions. The charges partly differ between the platforms, but there are marked overlaps between the cases against each of the platforms. This may partly be due to essentially global business strategies pursued by the large digital platforms and the fact that the actions in some jurisdictions, in particular in the EU, have inspired actions in others. Anu Bradford has referred to this phenomenon as the “de jure Brussels effect”, which manifests itself, inter alia, through “copycat litigation”: non-EU jurisdictions benefit from the EU investigations and can intervene based on product market definitions, theories of harm and possibly an effects analysis that have been pre-tested [171]. For example, the EU Commission’s Google decisions [172] seem to have inspired the Google decisions of the Russian, [173] Brazilian [174], Turkish [175], and South Korean competition authorities [176]. At least in the digital sector, the EU Commission now seems to play a role that used to be exercised by the U.S. antitrust authorities in the past. Simultaneously, the large number of parallel proceedings are evidence of a relevant degree of convergence of many competition law regimes worldwide.

4. Length of competition law proceedings – a justification for a regime of plat-form regulation?

In the ongoing public debates, one of the core reasons cited for the need to shift to some sort of “platform regulation” is the impression that competition law enforcement is, by its very nature, overly slow. However, the sample of cases discussed above suggests that this does not apply across all jurisdictions and settings. There are some jurisdictions where competition authorities, once they have decided to take action, move very fast. This is true, in particular, for Russia and China. However, these are simultaneously those jurisdictions where we may suspect that they do not comply with the same procedural standards, do not guarantee the same degree of judicial review and may partly follow a political agenda.

If we focus on competition law enforcement in the EU, a mixed picture emerges [177]. Those who complain about an excessive duration of competition proceedings frequently point to the Google Shopping case, which took about six and a half years from the opening of the proceeding until the publication of the Commission’s decision [178]. Google Shopping was an outlier, however, and apart from novel legal issues, very specific procedural circumstances contributed to the length, including failed commitment negotiations [179].

Looking at the EU Commission’s other enforcement actions in the digital field, and leaving out Google Shopping, the few cases discussed earlier (Microsoft, Microsoft (Tying), Google Android, Google Search (AdSense), Amazon (E-books)) took on average about 2 years and 8 months from the opening of proceedings to the Commission’s final decision. Other cases are still ongoing (Amazon Marketplace, Amazon Buy Box, Apple Mobile Payments, Apple App Store).

An even more nuanced picture emerges when we look at the enforcement activities at Member State level. About half of the national enforcement cases included in this case overview have been decided so far, the other half is still pending. One third of our case sample was only initiated after May 2020, i.e. within the last year. Of the cases decided so far, it took the national competition authorities on average little less than one and a half years from the opening of the proceedings to close the case – about one year less than the Commission. Of the four jurisdictions with particularly high enforcement activity, the French Autorité de la Concurrence and the Italian AGCM were about half a year faster than the German Bundeskartellamt and the UK CMA. What is more, the Autorité de la Concurrence has repeatedly adopted interim measures within about half a year from the opening of proceedings.

In cases where companies offered commitments or voluntarily ceased the practice under review, it took the authorities little more than a year on average to close the case after the opening of proceedings. The proceedings on Amazon’s marketplace rules were even quicker: Both the German Bundeskartellamt and the Austrian Bundeswettbewerbsbehörde closed these cases after about half a year from the opening of proceedings when Amazon changed its terms and conditions. However, other proceedings relating to marketplace rules are still pending, some of them initiated in 2019.

It seems plausible that the length of proceedings is significantly affected by both the factual complexity of a case and the question whether the legal issues are settled or remain controversial. For example, the proceedings on MFNs were closed within on average little more than one year and a half. In comparison, it took the German Bundeskartellamt comparatively long to decide the Facebook case – almost three years.

No general claims can be made so far about the length of consumer protection proceedings on digital matters that overlap with competition law issues. The two cases in our sample that have been decided so far have taken slightly more than one year on average. The majority of these cases is still pending, however.

Against this background, the argument that a shift from competition law towards a regime of platform regulation is needed to speed up enforcement becomes somewhat less persuasive. Rather, an inquiry would be needed into which cases typically take particularly long and why. Also, a closer comparative look would be useful why enforcement action at the national level tends to be quicker and why some authorities – like the Autorité de la Concurrence and the AGCM – are faster than others. Is this due to differences in the cases handled? Are the analytical tools that are part of the handling of competition law cases used in different ways? Does the commitment of a competition authority to the “more economic approach” matter? To what extent can a different handling of the standard of proof be observed, or do different attitudes towards the use of interim measures make a difference?

Furthermore, it is not self-evident that the implementation of and enforcement of obligations under the DMA will be much quicker on a systematic basis. Certainly, the initial implementation phase will take time: A provider of core platform services that meets the thresholds of Art. 6(2) DMA shall notify the Commission within three months after the thresholds are satisfied (Art. 6(3) DMA). Then, no later than 60 days after the notification is complete, the Commission shall designate the provider of core platform services as gatekeeper (Art. 6(4) DMA). The obligations under Art. 5 and 6 must be complied within six months after the gatekeeper is included in the list pursuant to Art. 3(7), see Art. 3(8) DMA. If the gatekeeper implements measures that are not sufficient to ensure effective compliance with Art. 6, the Commission may specify the measures to be implemented within six months from the opening of proceedings (Art. 7(2) DMA). At least in the initial phase of DMA enforcement, these time periods will add up.

Looking at the enforcement of Art. 5 and 6 DMA once the gatekeepers have been designated, the enforcement can again take time. No inquiry into relevant markets and no effects analysis will be needed. But there may be a need to look closely into the functioning of an algorithm or to analyze the conditions of data portability or interoperability in some depth. Also, the speed of enforcement will depend, to a significant extent, on an efficient complaints handling mechanism and on the enforcement resources that are available.

When it comes to the imposition of behavioral or structural remedies in reaction to a gatekeeper’s systematic non-compliance with Art. 5 and 6, these must be based on a market investigation which the Commission shall “endeavor” to conclude within one year (Art. 16(1)), extendable by six months (Art. 16(6)). And an updating of the DMA will require a market investigation which shall be concluded within two years (Art. 17).

This is not to criticize the timelines foreseen in the DMA – decisions on systematic non-compliance or on the updating of the DMA are complex, and a good balance must be struck between avoiding, to the extent possible, error costs on the substance and the speed of intervention. Rather, we would argue that it may not only – and maybe not primarily – be the speed of enforcement that justifies a shift towards platform regulation. It may rather be the systemic nature of the competitive threat, the repeat infringements of similar kind and the inability to address these effectively within a competition law regime that moves case-by-case that may justify a shift towards such regulation (see below, III.5.).

5. Systemic competition concerns where gatekeepers control the access to an ecosystem and/or are vertically integrated – a possible justification for “platform regulation”

Our sample of cases against Google, Apple, Amazon and Facebook consists of different groups of cases: A large number of cases involves settings where a large provider of platform services has the ability and incentive to leverage its advantages from one area of activity to another – either because the core platform service provider controls access to an ecosystem or because it is vertically integrated and competes on the marketplace it operates [180] (lit. a). Another group of cases concerns “marketplace rules” considered to be “unfair” by business users without there being any apparent leveraging strategy involved. Such claims of unfairness are brought forward even where a platform does not control access to an ecosystem or simultaneously competes on the platform (lit. b).

a) Systematic threats to competition in ecosystem and vertical integration settings

Most of the cases brought against Google and Apple – although not all of them – fit into the ecosystem box [181] or may be regarded as attempt to leverage market power from one market onto a vertically related market.

With regard to Google, the focus has been on bundling and tying Google Search and the Google Play Store with Android which is widely regarded to be Google’s most important strategic move in expanding its dominant position in the online search market from the PC environment to the mobile environment [182], on exclusivity arrangements in the advertising market as Google’s most important source of income, and on the favoring of vertically integrated services over competing services. More recently, concerns have turned to Google’s data collection policies, namely its decision to phase out support for third-party cookies while continuing to heavily rely on data collection and the processing of (aggregated) data for the purpose of targeted advertising itself. While these policies may be privacy-enhancing, they can serve to protect Google’s prominent position on online advertising markets at the same time.

The various cases against Google presented in this review address different aspects of the company’s business strategy – the expansion of its ecosystem – and its business model in its heavy reliance on access to consumer data for its use on advertising markets. At the same time, they are a vivid demonstration of the inherent limits of a case-by-case approach in dealing not only with singular abuses, but with a systematic strategy of a cross-market nature. The economic interest in maximizing advertising revenue drives a multi-faceted strategy aiming to expand the ecosystem in varying dimensions and making use of the changing opportunities to reinforce barriers to entry into core markets, in particular advertising markets. The systematic nature of such a business strategy requires a more systemic view of the competition problems than competition law in its traditional form allows for, as well as new tools to react to these systemic problems.

The situation is both similar and different with regard to Apple. Tying cases, a self-preferential treatment of Apple apps and recent charges of Apple discriminating against third parties in their access to data are, again, evidence of an ecosystem strategy. The business strategy nonetheless differs from Google’s in that it is not primarily driven by the aim to monetize data in online advertising markets. Complaints regarding Apple’s terms and conditions for access to the App Store – including the 30 % provision for all sales [183] – may therefore go more to the core of Apple’s business models than to Google’s.

Charges against Amazon have partly focused on attempts to reduce the contestability of its e-commerce-platform by imposing MFNs or functional equivalents and partly on self-preferencing practices, i.e. various ways of favoring its own retail business over that of other sellers on the platform, including through the use of non-public data of competing sellers. The bundling of the use of the platform with the use of “Fulfillment by Amazon” may be viewed as some sort of ecosystem expansion. But so far, the case law has centered around the marketplace and the logistics that go with it – an ecosystem that, compared to Google and Apple, Amazon’s “ecosystem” and ecosystem strategy, seems more confined. However, there are clear signs that this is changing. Amazon’s prominent start into the IoT world, namely with Alexa, Amazon Web Services’ success in the area of cloud computing and its expansion into the areas of video (Prime Video) and Music (Prime Music) seem to indicate that Amazon is about to morph into a full-scale ecosystem of its own kind [184]. The growing awareness of the anticompetitive potential of ecosystem strategies related to consumer IoT is illustrated by the Commission’s ongoing sector inquiry [185].

In all these settings, there is a systemic potential for anticompetitive strategies of various sorts that are difficult to address comprehensively under a competition law case-by-case approach [186], in particular where there is no evident structural remedy. Based on our case overview above, we conclude that it is this systematic ability and incentive to leverage and foreclose, and at the same time the traditional competition law’s confinement to remedying singular abuses, that justifies a special regime of “platform regulation”: Where a gatekeeper has gained a strategic market position of a kind that allows for repeat strategies of leveraging and foreclosure, the competition law approach to remedy singular abuses must be replaced by a regime that allows competition authorities to address the underlying competition problem more systematically and broadly, e.g. by ensuring a sufficient degree of openness and interoperability for competitors.

The need to address the underlying causes of the competition problem includes a need to address imbalances in the access to data where these imbalances would otherwise risk to foreclose competition. Our case sample shows that competition authorities are only starting to consider the strategic potential of gatekeeper control of data. The German Facebook case, the EU Commission’s investigations into Amazon’s access to non-public data of its platform users and into its use of that data for the benefit of its retail arm that competes on the platform [187], the investigations into the parallel change of strategy by both Apple and Google with regard to third-party cookies [188] and finally the recently opened investigations into Facebook’s advertising data practices [189] start to map the scene. The importance of these data-related concerns is likely to grow. Here again, a broader and more overarching approach will likely be needed than case-by-case competition law enforcement can ensure [190].

b) The unfairness of marketplace rules

The other group of cases in our sample concerns marketplace rules – i.e. cases where a platform that has acquired gatekeeper status (ab)uses the dependency of its users to set terms and conditions that are considered to be “unfair”. In the EU, cases concerning an alleged unfairness of marketplace rules – typically vis-à-vis business users – have primarily been dealt with at the Member State level so far. There is not necessarily an obvious link between the alleged unfairness and anticompetitive effects – although there may be one. An imbalance of rights and obligations on platform users can also qualify as an exploitative abuse. But this would normally presuppose a causal link between the dominant position of the platform operator and its ability to extort unfair conditions from its users [191]. In its recent Facebook decision, the Bundesgerichtshof ruled in favor of the possibility of a hybrid abuse which can result from a combination of an imbalance of rights and obligations and plausible anticompetitive effects.

Looking at our case sample, it is unclear whether allegedly unfair marketplace rules justify a special regime of platform regulation. Clearly, systematic problem can arise. Economically, imbalances of rights and obligations can follow from information asymmetries and bounded rationality – well-known problems that underlie regimes of judicial control of contractual boilerplate terms and conditions generally, but are thought to justify such control only in B2C relations in most jurisdictions. Platform providers frequently set customer friendly provisions for consumers but disadvantageous terms and conditions for business users, as the value of an additional business user to the platform is limited, while the value of each additional consumer is much greater. Where a platform has become a gatekeeper, with an entrenched position based on strong network effects, there may be a case for a special control of P2B terms and conditions based on the established rationale for controlling general terms and conditions in contract law.

Given the imbalance of power and the business users’ valid fear of retaliation, public intervention may be justified in such settings. Although this is not a prototypical competition law issue, competition authorities may be well qualified to engage in a review of the terms and conditions as they understand the legitimate business interests on both sides. A plausible argument may be made that terms and conditions that systematically disadvantage platform users on the business side may systematically deprive business users from opportunities to innovate and invest [192]. Depending on the circumstances of the case and the handling of the causality requirements (see above), an exploitative abuse may be established. Where the platform is vertically integrated, discretionary terms and conditions can be used to disadvantage competitors.

In the EU, competition authorities have not failed to meet expectations in this regard so far, however. Many of the relevant cases have been settled rather quickly within the competition law framework.

What is more, specific incidences of an alleged unfairness require will frequently require a careful balancing of interests case-by-case. A fixed list of obligations – like in Art. 5 and 6 of the draft DMA – is not helpful in this regard. In any case, the challenge is different from the challenge to effectively protect against systematic ecosystem, leveraging and foreclosure strategies. The position of “platforms as regulators” [193] is ambiguous and must be assessed in a highly context-specific manner.

6. The hotel booking platform cases have been different

If many of the competition law cases against Google, Apple and Amazon have been driven by concerns about anticompetitive strategies in ecosystem and vertical integration settings, the hotel booking platform proceedings represent a wholly different category of cases. These platforms do not pursue ecosystem strategies nor are they vertically integrated. Rather, the concerns result specifically from the potential of price parity clauses to reduce competition between platforms and to raise barriers to entry for new platform entrants. While these are serious concerns, the overarching systemic risks to competition across markets are not present in these settings. Traditional competition law is perfectly able to deal with these cases and offers a remedy that fits the problem – namely a prohibition of MFNs. Arguably, these and similar platforms should, therefore, not be included in a regime of platform regulation.

This may caution against an overly broad scope of application of a future DMA: Interestingly, no cases on cloud computing services – one of the “core platform services” mentioned in Art. 2(2) lit. g DMA – have been included in our sample: it seems that the market for cloud computing still is rather competitive, although switching may come at a cost. With good reason, streaming services have not been included in the Art. 2(2) DMA list: For the moment, they do not display the type of competition problems that the DMA is meant to address, in particular competition problems related to conglomerate ecosystem strategies or to vertical integration.

7. The adequacy of the lists of obligations in Art. 5 and 6 of the draft DMA in light of the case law

Many of the gatekeeper obligations listed in Art. 5 and 6 of the draft DMA are clearly inspired by the case law compiled in part II of this paper [194]. However, some of the obligations go beyond: Given the specificities of gatekeeper platform’s ecosystem strategies, they strive to address the systemic causes of the competition problems more broadly. This is true in particular for Art. 6(1) lit. h which obliges gatekeepers to ensure the effective portability of data generated through the activity of a business user or end user; for Art. 6(1) lit. i, i.e. the obligation to provide business users with effective, high-quality, continuous and real-time access and use of aggregated or non-aggregated data, that is provided for or generated in the context of the use of the relevant core platform services by those business users and their end users; and Art. 6(1) lit. j requiring gatekeepers to provide third-party providers of online search engines with access on FRAND terms to ranking, query, click and view data in relation to search queries generated by end users on online search engines of the gatekeeper [195]. Interestingly, these obligations are all data-related obligations. Yet, they would not help to deal with the more recent data-related cases – in particular the investigations into Apple’s App Tracking Transparency framework and into Google’s “Privacy Sandbox” (see above) – which shows that any such list of obligations may quickly need an overhaul in a dynamically evolving environment if not backed up by broader principles. Also, given the lack of experience with data-related remedies so far, one may wonder whether the data-related obligations included in the Art. 5 and 6 lists – like the prohibition to combine personal data sourced from one core platform service with personal data from any other service in Art. 5 lit. a DMA – would be effective in restoring competition. With regard to the data portability and interoperability requirements, all will depend on an effective implementation and monitoring regime.

It is equally interesting that – despite the proposed empowerment of the Commission in Art. 10(2) lit. b DMA to update the lists of obligations where, based on a market investigation, it identifies a need for new obligations addressing practices that are unfair – there are few such unfairness provisions in the current Art. 5 and 6 lists. There is only one clear candidate, namely Art. 6(1) lit. k, which requires gatekeepers to apply fair and non-discriminatory general conditions of access for business users to its software application store. Obviously, the existing case law on unfair terms and conditions by platform operators has not incited the Commission to add further obligations to the lists. This shows that the Commission was aware of the mismatch between the necessities of a fairness control of terms and conditions and the regime of platform regulation as set out in the DMA. A regime of fairness control of terms and conditions will need another conceptual and methodological framework. This provides an additional argument for narrowing down Art. 10(2) lit. b DMA to cover only those fairness problems with a clear competition rationale [196].

8. Division of labor between the EU Commission and the NCAs

If we look at the interaction between competition law enforcement at the European and at the national level, the image of a highly cooperative and complementary enforcement regime emerges – including in the digital realm. Parallel public competition law enforcement by the NCAs has not undermined the coherence and consistency of EU competition law. To the contrary, NCAs have assumed a very important role in addressing the new policy issues that have arisen in digital platforms markets. Repeatedly, NCAs have been more willing than the EU Commission to engage in inquiries into markets that had not been thoroughly investigated and well understood before – like advertising markets [197] – or to test cases in areas where an established competition law doctrine has been lacking – as the practice of ecosystem operators to combine data from different sources [198]. The review of marketplace rules has been another area where NCAs have significantly contributed to the evolution of the case law. Overall, the enforcement activity at the national level has filled enforcement gaps and thereby done its share to ensure an effective competition law enforcement in the EU. Notably, the NCAs have also contributed to the digital policy debate – sometimes in cooperation with one another [199].

Against this background, the policy choice underlying the draft DMA to centralize the enforcement of the gatekeeper’s obligations set out in Art. 5 and 6 DMA with the EU Commission should be rethought. It is a core justification of the DMA indeed to avoid legal fragmentation between the Member States. The different stances that some NCAs have taken on the legality of “narrow” MFNs is sometimes cited as an example for the risk of an incoherent interpretation of EU competition law and taken as evidence for a need to centralize. Ultimately, the legality of a “narrow” MFN will, however, depend on a proportionality test under Art. 101(3) TFEU. Among other things, the relevant restriction of competition must be indispensable to the attainment of the efficiency objectives. In its recent judgment on the use of “narrow” MFNs on hotel booking platforms, the Bundesgerichtshof has found that the risk of free riding by users of the platform that the “narrow” MFN was meant to address did not materialize in practice, or just to a very small extent [200]. However, this is a factual issue that can differ from jurisdiction to jurisdiction. The divergence in the hotel booking platform cases may therefore not result from a different interpretation of the law, but rather from different factual settings.

But even if one were to take “narrow” MFNs as an example for diverging interpretations of Art. 101 TFEU, such divergence has been the exception rather than the rule, and it could have been resolved by the EU Commission at any point of time (see Art. 11(6) Regulation 1/2003), or could be resolved in the course of a referral procedure (Art. 267 TFEU). Overall, the European Competition Network (ECN) has contributed to a converging understanding of competition law and its relevance in the Member States. More generally, the decentralized enforcement of EU law has greatly contributed to the growth of a “European legal culture” bottom-up [201]. Promoting centralization out of fear of divergence risks to miss the genius of the evolution of EU law and its anchorage in the Member States.

9. Interaction between competition law, the law on unfair business terms and consumer protection law

As pointed out above, the NCAs have assumed an important role in reviewing the legality of the large digital platforms’ terms and conditions. The Austrian Bundeswettbewerbsbehörde and the German Bundeskartellamt have closely looked at Amazon’s terms and conditions vis-à-vis business users [202]. The French Autorité de la Concurrence has looked into the terms and conditions on advertising markets [203]. And the Italian AGCM has fined Facebook for violating the Consumer Code by misleading consumers into unconscious and automatic data sharing [204].

In the review of the “fairness” of terms and conditions of a digital platform, competition law, national laws on the “unfairness” of terms of conditions, including B2B, consumer protection law and potentially also data protection law can overlap. Looking at the case law, the choice of the legal regime that is applied is sometimes intertwined with the institutional setting. It seems that those NCAs with a parallel competence to enforce consumer protection law have preferred to deal with consumer-facing data-related practices under consumer protection law and not competition law. The German Facebook case is famous for addressing data protection law aspects within the framework of a competition law proceeding. The question whether “unfair” contract terms B2B are challenged on competition law grounds or on the basis of the law on unfair contract terms may likewise depend on whether the applicable law provides for a review of contract terms B2B on fairness grounds and on the enforcement regime.

These overlaps raise the question of what the proper role of competition law should be. EU competition law has not yet answered it definitely. The draft DMA arguably moves beyond the current state of EU competition law, at least when it comes to its Art. 10: The Commission shall be empowered to adopt delegated acts in order to update the gatekeeper obligations laid down in Art. 5 and 6 where it has identified the need for new obligations addressing practices that limit the contestability of core platform services or are “unfair”, i.e. where there is “an imbalance of rights and obligations on business users and the gatekeeper is obtaining an advantage from business users that is disproportionate to the service provided by the gatekeeper to business users”.

The question whether the varied, context-specific problems of the unfairness of contractual terms and conditions can adequately be addressed in the form of a fixed list of “one size fits all” obligations has been already raised (see above, III.5. lit. b). But it may be worthwhile to give more thought to the link between unfair terms and conditions on digital platforms and the protection of competition more generally. In the digital economy, categories of market failure that have traditionally been treated separately and addressed by different areas of the law tend to converge and overlap in novel ways. In particular, issues of market power and dependency and the exploitation of information asymmetries are frequently intertwined [205]. This should be no excuse to give up important conceptual distinctions and to no longer distinguish between competition goals and further-reaching goals of non-competition related fairness [206]. Where the clarity of goals gets lost, the law becomes muddled and unpredictable. Such a state of the law would be unlikely to promote competition or innovation.

— Heike Schweitzer and Frederik Gutmann are grateful to Robert Welker and Philipp Hornung for their thoughtful comments as well as Nicoline Gebler who – based on the Concurrence database – prepared the lists which have guided their analysis.

Note from the Editors: although the e-Competitions editors are doing their best to build a comprehensive set of the leading EU and national antitrust cases, the completeness of the database cannot be guaranteed. The present foreword seeks to provide readers with a view of the existing trends based primarily on cases reported in e-Competitions. Readers are welcome to bring any other relevant cases to the attention of the editors.

Version PDF

Auteurs

  • Humboldt University (Berlin)
  • Humboldt University (Berlin)

Citation

Heike Schweitzer, Frederik Gutmann, Unilateral Practices in the digital market : An overview of EU and national case law, 15 juillet 2021, e-Competitions Unilateral practices in the digital market, Art. N° 101045

Visites 1532

Tous les numéros

  • Latest News issue 
  • Tous les News issues
  • Latest Special issue 
  • Tous les Special issues