The growth of digital platforms has focused the minds of policymakers and competition authorities across the world on whether stricter enforcement is required against “self-preferencing.” Competition authorities are taking an increasingly expansive approach. While the abuse of dominance regime provides competition authorities with a great deal of flexibility, their treatment of self-preferencing as a standalone abuse raises fundamental questions and leaves them unanswered: over and above the gating question of “dominance,” where is the dividing line between “competition on the merits” and unlawful commercial practices, and how does the “special responsibility” of dominant firms play into the steps that dominant platforms need to take to ensure they are not crossing that line? There has been increasing pressure to use ex ante regulation to tackle self-preferencing. However, here too many unanswered questions remain: is regulation necessary in light of the flexible approach being taken with antitrust enforcement? Which markets (and which companies within those markets) should be subject to such regulation?
By Rod Carlton & Rikki Haria1
I. SELF-PREFERENCING IS IN THE SPOTLIGHT IN THE DIGITAL ECONOMY
In principle, an undertaking’s expansion into a vertically-related or complementary business area is more likely to have pro-competitive effects than be anti-competitive. Stakeholders commonly agree on the benefits for competition and consumers, including the scope for generating significant efficiencies and reducing transaction costs for consumers.2
There has, however, been renewed focus by policymakers and regulators across the world on whether a stricter approach is required in the digital economy. This has, in part, been triggered by the rapid expansion of digital platforms and technology companies into new business areas as they seek to adapt to new technological developments and retain the attention of people and businesses. For example, Google has acquired a range of companies to expand its operations along the entire ad tech stack and enter into a range of new business areas, such as smart home technology and healthcare.3 Similarly, Amazon has expanded its operations to cover virtual assistants (i.e. Alexa), groceries, music and video streaming, and pharmaceuticals.
Similarly, many digital platforms are used by large numbers of independent sellers of goods and services. For example, Amazon has over 2.5 million active sellers selling their products on its marketplace,4 Apple and Google each make millions of apps available to consumers on their app stores,5 and millions of businesses advertise on Google and Facebook.
In this context, “self-preferencing” has emerged as a key area of focus for policymakers and regulators. Self-preferencing involves a digital platform giving preferential treatment to its own products and services when they are in competition with the products and services provided by other companies. Concerns have been raised that this type of behavior may enable large digital platforms to protect their market position in their existing business areas or enter into new business areas in an “anti-competitive” manner, ultimately leading to worse outcomes for consumers and businesses.
The most high-profile investigations into self-preferencing conduct in Europe are:
- Google Shopping: in 2017, the European Commission imposed a €2.42 billion fine on Google for practices that amounted to an abuse of Google’s dominant position in the general internet search markets by stifling competition in comparison shopping markets. In particular, the Commission held that: “by giving prominent placement only to its own comparison shopping service and by demoting competitors, Google has given its own comparison shopping service a significant advantage compared to rivals.”6 The Commission considered that such practices stifled “competition on the merits” in comparison shopping markets. Google has appealed this infringement decision to the EU’s General Court and judgment is pending.
- Amazon Marketplace: in 2019, the European Commission opened an investigation into Amazon. Amazon has a dual role whereby: (i) it sells products on its website as a retailer; and (ii) it provides a marketplace where independent sellers can sell products directly to consumers. The Commission’s press release announcing this investigation notes: “When providing a marketplace for independent sellers, Amazon continuously collects data about the activity on its platform. Based on the Commission’s preliminary fact-finding, Amazon appears to use competitively sensitive information – about marketplace sellers, their products and transactions on the marketplace.”7 The investigation is ongoing. If proven, the Commission considers that the practices under investigation breach EU competition rules on anti-competitive agreements between companies (i.e. Article 101 TFEU) and/or the abuse of a dominant position (i.e. Article 102 TFEU). Following a news story in April 2020 reporting that Amazon employees had used data about independent sellers on its marketplace to develop competing products,8 this is likely to be a focal area for the Commission’s investigation.
- Apple: in 2019, Spotify filed a complaint with the European Commission in relation to Apple’s App Store complaining about a 30 percent fee that Spotify and other digital services are required to pay on purchases made through Apple’s payment system. As part of this complaint, Spotify noted: “apps should be able to compete fairly on the merits, and not based on who owns the App Store. We should all be subject to the same fair set of rules and restrictions—including Apple Music.”9 While the Commission has not publicly opened a formal investigation against Apple, it has reportedly sent information requests to Apple and third parties on this issue.
This article explores two issues that are resulting in considerable legal and regulatory uncertainty for digital companies (and practitioners advising on self-preferencing conduct):
- The “innovative” approach being taken by competition authorities when assessing self-preferencing conduct.
- The push for ex ante regulation to deal with self-preferencing in the digital economy.
II. THE “INNOVATIVE” APPROACH BY COMPETITION AUTHORITIES WHEN ASSESSING SELFPREFERENCING
Self-preferencing by digital platforms is, in effect, a form of “leveraging” behavior. Enforcement action against leveraging behavior is not a novel issue under the competition rules relating to abuse of dominance. Indeed, it is a concept that underpins many different categories of “recognized” abuses. For example, refusal to supply, margin squeeze, discrimination and tying – all of which are “recognized” categories of abuse of dominance – to some degree involve the leveraging of market power from a market where a company is dominant to a vertically-related or neighboring market where that company may not be dominant. By way of example:
- In Commercial Solvents, which involved a refusal to supply, the Court stated: “an undertaking being in a dominant position as regards the production of raw material and therefore able to control the supply to manufacturers of derivatives, cannot, just because it decides to start manufacturing these derivatives (in competition with its former customers) act in such a way as to eliminate their competition.”10
- In Tetra Pak II (which involved a number of different practices, including tying), the Commission held that an infringement of Article 102 TFEU could occur where the abuse occurred on a neighboring market to the market where a company was dominant (i.e. to gain an advantage in the neighboring market).11
- Abusive discrimination can occur where a vertically-integrated company that is dominant on the upstream market engages in discriminatory conduct that places competitors on the downstream market at a competitive disadvantage (i.e. vis-à-vis its own downstream operations). This has been well-explained by AG Wahl in MEO12 and by Ofcom in its recent infringement decision against Royal Mail relating to the supply of bulk mail delivery services.13
This takes us to the Commission’s Google Shopping infringement decision,14 which is currently under appeal. As noted above, this case involved self-preferencing behavior: Google’s algorithms in the general search space allegedly demoted results of rival comparison shopping services, while these algorithms were allegedly not applied in the same way to Google’s own comparison shopping service. The Commission concluded that this resulted in traffic being increased to Google’s comparison shopping services, not due to “competition on the merits,” but because Google had systematically and unduly positioned its own services at the top of its general search results.
In explaining the legal basis for its infringement decision (in response to arguments by Google that the Commission had not established Google’s general search services constituted an “essential facility” in line with the refusal to supply case-law), the Commission held: “It is not novel to find that conduct consisting in the use of a dominant position on one market to extend that dominant position to one or more adjacent markets can constitute an abuse […] Such a form of conduct constitutes a well-established, independent, form of abuse falling outside the scope of competition on the merits.”15
While “leveraging” behavior has underpinned the reasoning behind other “categories” of abuse (as noted above), it is debatable whether – based on existing case law – “leveraging” conduct is a well-established standalone abuse in and of itself.16
That said, as is clear from the language of Article 102 TFEU, there is no exhaustive or closed list of conduct that constitutes an abuse of a dominant position. The mere absence of a recognized “category” of abuse relating to leveraging or self-preferencing conduct does not legally preclude the Commission (and EU Courts) from holding that a company’s conduct is abusive. Irrespective of whether the General Court rules in favor of the Commission, this case illustrates the Commission’s willingness to take a more expansive approach to tackle self-preferencing behavior and utilize Article 102 TFEU in a more flexible way when dealing with large digital platforms and technology companies. The question is whether competition authorities’ treatment of self-preferencing as a standalone abuse is a legitimate enforcement innovation in response to the emergence of innovative digital business models.
The expansive approach taken by the Commission creates considerable legal uncertainty. Previous cases involving “leveraging” behavior have been linked to other categories of abuse, where there are helpful criteria (and guidance) on what divides permissible behavior from abusive behavior. Take “refusal to supply” cases as an example: the case-law has evolved over time to clarify that access to the upstream product must be “indispensable” for a rival to compete in the downstream market for a competition law problem to arise under Article 102 TFEU. In policy terms, “indispensability” (as opposed to mere convenience or desirability) aims to safeguard investment incentives for innovation in products or facilities. And in terms of legal compliance and justiciability, “indispensability” aims to provide an objectively verifiable litmus test – both for owners of facilities and access seekers to those facilities. Similarly, in its ongoing appeal of the Google Shopping decision, Google has provided the following example: “a dominant undertaking’s practice of setting low prices cannot, on its own, be considered abusive. It would only be abusive if an additional feature departing from competition on the merits were identified that could be classified as predatory pricing.”17
However, the approach being taken by the European Commission (and other competition authorities) to tackle self-preferencing as a standalone abuse seems to rest on “competition on the merits” as the relevant litmus test. In the context of the ongoing appeal in the Google Shopping case, the Commission has reportedly articulated a three-step test for determining when the conduct of a dominant company treating its own service differently to the way it treats a rival’s similar service can be abusive: (i) the difference in treatment has a significant impact on an important parameter of competition (e.g. traffic in the case of Google Shopping); (ii) the significant impact is capable of having an anti-competitive effect; and (iii) the dominant company has no objective justification for the difference in treatment.18
While this may make sense from a theoretical perspective (and underpins the Commission’s willingness not to be tied to “categories” of abuse and focus more on whether conduct has anti-competitive effects), this three-step test provides little guidance to dominant platforms on the line between permitted conduct and prohibited conduct. This is exacerbated by recent comments from the Commission suggesting it should be subject to a lower standard of proof to demonstrate conduct in tech markets is anti-competitive.19 This uncertainty may result in potentially dominant digital platforms taking an unduly conservative approach to their commercial conduct. This in turn risks stifling innovation, particularly in light of the recognized efficiencies and pro-consumer benefits that have been delivered by firms in the digital economy.
Seen also from the perspective of legal advisers to independent businesses who use digital platforms to compete in downstream or neighboring markets, the Commission’s approach makes it very difficult to assess on the facts whether their client has a compelling complaint. Third party complaints to competition authorities need to be cogently argued and evidenced in order to stand any chance of gaining traction, let alone priority. A vague legal test risks deterring potential complainants from allocating the significant resources and time required to press their case with the authorities.
In light of the above, the challenge for competition authorities is to publish practical guidance that more clearly delineates what constitutes anti-competitive self-preferencing, similar to the approach used for other recognized “categories” of abuse (such as refusal to supply, margin squeeze, predatory pricing, and rebates).
From the perspective of practitioners, one can only hope that the General Court’s judgment in Google Shopping and any decisions stemming from the Commission’s ongoing investigations (such as Amazon Marketplace) will provide a coherent and clear framework for companies to assess self-preferencing behavior.
III. THE PUSH FOR REGULATION TO DEAL WITH SELF-PREFERENCING IN THE DIGITAL ECONOMY
There has been increasing pressure amongst regulators and policymakers to use ex ante regulation to tackle self-preferencing behavior. As noted by Professors Ennis and Fletcher, a number of expert reports (some of which were government-commissioned) “give weight to the important ‘gatekeeper’ or ‘bottleneck’ role that can be held by digital platforms.”20 This is used as part of the justification for some of these expert reports to recommend ex ante regulation to complement ex post antitrust enforcement for issues such as self-preferencing. For example:
- Report of the Digital Competition Expert Panel commissioned by the UK Government (the “DCEP Report”):21 the DCEP Report recommends establishing “a digital platform code of conduct, based on a set of core principles [which] would apply to conduct by digital platforms that have been designated as having a strategic market status.”22 Some of the principles suggested for this code of conduct relate to self-preferencing, namely: (i) “business users are provided with prominence, rankings and reviews on designated platforms on a fair, consistent, and transparent basis”23 – the DCEP Report considers that it would be inconsistent with this principle for a platform to give “undue preferential prominence on its webpages to its own integrated services”;24 and (ii) “business users are provided with access to designated platforms on a fair, consistent and transparent basis”25 – the DCEP Report considers that a behavior that is inconsistent with this principle is “an online marketplace […] excluding or suspending rival sellers from its platform to give its own product or service an advantage.”26
- Special Advisers’ report to the European Commission (the “Special Advisers’ Report”):27 while the Special Advisers’ Report does not go far as recommending ex ante regulation, it seeks to achieve a similar outcome through reversing the burden of proof for ex post antitrust enforcement. The Special Advisers’ Report notes: “In a market with particularly high barriers to entry and where the platform serves as an intermediation infrastructure of particular relevance, we propose that, to the extent that the platform performs a regulatory function, it should bear the burden of proving that self-preferencing has no long-run exclusionary effects on product markets. The dominant platform would then need to prove either the absence of adverse effects on competition or an overriding efficiency rationale.”28
Regulators and policymakers have, in turn, been discussing more fundamental changes to the legal framework applicable to the digital economy, including how to tackle self-preferencing behavior. For example:
- Inspired by the DCEP Report, in its Interim Report for its market study into online platforms and digital advertising, the UK Competition and Markets Authority (the “CMA”) has proposed an enforceable Code of Conduct (potentially applying only to online platforms with “strategic market status”) is set up to govern such platforms’ relationships with advertisers, publishers, consumers and other third parties. The CMA has indicated that this Code of Conduct could cover self-preferencing behavior. For example, a principle being considered by the CMA under a Code of Conduct is that “platforms should not give particular prominence or preferential terms to their own services in terms of how they are presented to users on the core platform, or in terms of how they design the information provided to customers and users.”29
- The German Federal Ministry for Economic Affairs and Energy has proposed an amendment to the Act against Restraints of Competition. The draft proposal provides that the Bundeskartellamt would be able to designate a firm as being an “undertaking of paramount significance for competition across markets” on the basis of specified criteria, namely its dominance on one or more markets, its financial strength or access to other resources, its vertical integration and its activities on related markets, its access to data and/or the importance of its activities for third parties’ access to supply and sales markets. The Bundeskartellamt could then prohibit such an undertaking from engaging in a variety of specified activities, unless they can be objectively justified, with the burden of proof lying with the undertaking in question. Such specified activities include self-preferencing – i.e. “treating the offers of competitors differently from its own offers when providing access to supply and sales markets.”30
- Margrethe Vestager, Executive Vice-President for the European Commission’s digital agenda and the Commissioner for Competition, has also suggested that ex ante regulation may be needed for “digital gatekeepers.” Such regulation would set out clear-cut prohibitions and obligations on digital gatekeepers.31 The Commission is consulting with stakeholders on what such regulation should look like – as part of this, it is engaging experts to conduct a study on concrete issues posed by digital gatekeepers and whether regulation is required to tackle such issues, with the terms of reference for this study showing that self-preferencing is a key issue being considered.32
However, there remain a number of unanswered questions for policymakers to consider before seeking to tackle self-preferencing through ex ante regulation.
First, is ex ante regulation actually needed to tackle self-preferencing behavior? In light of the innovative approach being taken by the Commission when investigating self-preferencing behavior through ex post competition enforcement (as set out above), it remains unclear why ex ante regulation is required and what additional type of conduct such regulation would seek to prevent.
Second, if the desire to impose regulation stems from the perceived need to ensure regulators can act quickly to resolve self-preferencing behavior in markets that are susceptible to “tipping” and enduring market power, this begs the question: which markets (and which companies within those markets) should be subject to such regulation? Common to most of the reports and policy proposals noted above is the concept that a regulatory prohibition on self-preferencing would not apply to all dominant companies, but instead to a subset of companies (some of which may not be dominant in a competition law sense) that hold some form of “strategic market status” or important “gateway” or “bottleneck” position. However, there remains considerable uncertainty around how regulators will determine in an evidenced, consistent, proportionate, and fair way which companies fall within such definitions.
For example, the DECP Report noted that: “a key component of this system is to develop a clear legal test for the characteristics of a company’s market position above which regulatory powers are appropriate – termed in this review a strategic market status. This needs to be carefully designed to identify where companies operating platforms are in a position to exercise potentially enduring market power, without granting an excessively broad scope and bringing within the bounds of regulation those companies who are effectively constrained by the competitive market.”33 As part of this, the DECP Report states: “this should be along the lines of identifying digital markets where strategic market status may materialise due to characteristics including significant direct or indirect network effects, limited offsetting effects of multi-homing and differentiation, and significant sources of non-contestability.”34 Looking at social media, content and messaging platforms, for example, the CMA’s Interim Report itself acknowledges extensive multi-homing by users, while market developments demonstrate ample evidence of the frequent entry and expansion of differentiated services (e.g. Snapchat, TikTok, Zoom, and Telegram). Given these factors, it remains unclear that any regulatory regime prohibiting self-preferencing should apply to platforms whose core business is providing social media, content and messaging services. Moreover, if the competition concern lies in tackling situations where a company is an important “gateway” to other businesses, this could be relevant to many different types of digital businesses (e.g. app stores, marketplaces, classifieds websites / apps, etc.), and even non-digital markets.
Third, who bears the burden of proof in relation to self-preferencing behavior? Many of the proposals noted above in effect imply a blanket ban on self-preferencing save to the extent that a dominant player can prove that any such conduct is objectively justified. Given the very limited circumstances in which competition authorities have accepted an objective justification defense in abuse of dominance investigations, this places a very high burden of proof on dominant platforms. Further consideration is needed as to whether this legal framework would fairly reflect the pro-competitive benefits and efficiencies that come from vertical integration, expansion into complementary business areas, and innovation by digital platforms.
There are no easy answers to these questions. And dominant firms undeniably have a “special responsibility” not to behave in an anti-competitive manner. But given the importance of these unanswered questions, further debate and consideration are needed to ensure policymakers do not adopt regulatory approaches that harm competition and innovation in the digital economy.
1 Rod Carlton (partner) and Rikki Haria (senior associate) are members of Freshfields Bruckhaus Deringer’s Antitrust Competition and Trade Group. The views expressed in this article are the authors’ own.
2 For example, see: (i) the OECD Competition Committee’s 2019 session on “Vertical mergers in the technology, media and telecom sector” in which the Secretariat’s Background Note states: “Vertical mergers are traditionally presumed pro-competitive, as they are generally driven by efficiency-enhancing motives”; and (ii) the European Commission’s “Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings,”  OJ C265/6, paragraphs 13-14: “The integration of complementary activities or products within a single firm may produce significant efficiencies and be pro-competitive […] integration may also decrease transaction costs and allow for a better co-ordination in terms of product design, the organisation of the production process, and the way in which the products are sold.”
10 Joined Cases 6 and 7/73, Istituto Chemioterapico Italiano S.p.A. and Commercial Solvents Corporation v. Commission, EU:C:1974:18, paragraph 25.
11 92/163/EEC, Commission Decision of July 24, 1991.
12 Case C-525/16, MEO — Serviços de Comunicações e Multimédia SA v. Autoridade da Concorrência, Opinion of Advocate General Wahl, December 20, 2017, paragraphs 77 to 78.
13 CW/01122/01/14, Ofcom, Discriminatory pricing in relation to the supply of bulk mail delivery services in the UK, August 14, 2018, paragraphs 5.39 to 5.45.
14 Case AT.39740, Google Search (Shopping), June 27, 2017 (the “Google Shopping” decision).
15 Paragraph 649 of the Google Shopping decision.
16 For example, Competition Law & Policy Debate, Theories of self-preferencing and duty to deal – two sides of the same coin?, Bo Vesterdorf, February 2015.
19 Margrethe Vestager recently commented: “if the bar for the requisite standard of proof with respect to anticompetitive conduct is set too high, it may result in under-enforcement to the detriment of consumers” and that enforcers should strive to “find the right balance between accuracy and administrability” (American Bar Association Antitrust Virtual Spring Meeting, April 17 – May 1, 2020).
20 Sean F. Ennis & Amelia Fletcher, Developing international perspectives of digital competition policy, March 31, 2020.
21 Report of the Digital Competition Expert Panel (Jason Furman, Diane Coyle, Amelia Fletcher, Derek McAuley and Philip Marsden), Unlocking Digital Competition, 2019.
22 Page 9 of the DCEP Report.
23 Box 2.A on page 61 of the DCEP Report.
24 Box 2.B on page 64 of the DCEP Report.
25 Box 2.A on page 61 of the DCEP Report.
26 Paragraph 2.36 of the DCEP Report.
27 Jacques Crémer, Yves-Alexandre de Montjoye and Heike Schweitzer, Competition policy for the digital era, 2019.
28 Page 66 of the Special Advisers’ Report.
29 CMA’s Interim Report in its online platforms and digital advertising market study, December 2019, Appendix I, pages I11 to I12.
30 For an unofficial English translation, see https://www.d-kart.de/wp-content/uploads/2020/02/GWB10-Engl-Translation-2020-02-21.pdf.
31 American Bar Association Antitrust Virtual Spring Meeting, April 17 – May 1, 2020.
33 Paragraph 2.116 of the DECP Report.
34 Paragraph 2.115(i) of the DECP Report. This is also recognized in a 2019 report by the Stigler Center (Scott Morton, Bouvier, Ezrachi, Jullien, Katz, Kimmelman, Melamed & Morgenstern, Committee for the Study of Digital Platforms, Market Structure and Antitrust Subcommittee, Stigler Center for the Study of the Economy and the State, 2019): ““bottleneck power” describes a situation where consumers primarily single-home and rely upon a single service provider, which makes obtaining access to those consumers for the relevant activity by other service providers prohibitively costly” (page 32).