Antitrust shrugged? Boycotts, content moderation, and free speech cartels
2023; Taylor & Francis; Volume: 19; Issue: 2 Linguagem: Inglês
10.1080/17441056.2023.2200612
ISSN1757-8396
Autores Tópico(s)Freedom of Expression and Defamation
ResumoABSTRACTAntitrust and free speech may seem to have little in common. Yet, they may start interacting more often as Big Tech undertakings might have incentives to coordinate their content moderation policies and collectively suppress unwanted information. Such coordination might be desirable, but in some cases it may lead to antitrust and free speech concerns. Against this backdrop, the article attempts to provide a framework to analyse this type of cases from the point of view of European Union competition law. It identifies five types of agreements that may be entered into by undertakings and provides outlines on possible ways of approaching them. It concludes that while content moderation is often seen as a free speech issue, antitrust should not shrug off such concerns as out of its scope. Yet, it also suggests that these types of cases are not straightforward and that both more research and vigilance on the part of antitrust authorities might be advisable.KEYWORDS: Free speechantitrustBig Techcontent moderationcartelsboycotthorizontal cooperationnon-economic interestspublic goalsconsumer welfare AcknowledgementsThe views expressed in this text are the author’s own and do not necessarily reflect those of the Polish Office of Competition and Consumer Protection (UOKiK).Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 The first direct link in case law was made in the 1940s in Associated Press, 326 U.S. 1 (1945). This was criticised in Robert Bork, The Antitrust Paradox: A Policy at War With Itself (Bork Publishing 2021) 50, 67, 442. Conversely, it was supported by Robert Pitofsky, see Alec Klein, ‘A Hard Look at Media Mergers’ The Washington Post (29 November 2000) accessed 7 January 2023. Later on, the link was discussed by e.g. Maurice Stucke and Allen Grunes, ‘Antitrust and the Marketplace of Ideas’ (2001) 69 Antitrust Law Journal 249; Hillary Greene, ‘Antitrust Censorship of Economic Protest’ (2010) 59 Duke Law Journal 1037; Tim Wu, The Master Switch: The Rise and Fall of Information Empires (Vintage Books 2011) 303, quoting Associated Press. More recently, it was analysed by Daniel Crane, ‘Collaboration and Competition in Information and News During Antitrust’s Formative Era’ (2020) Knight First Amendment Institute accessed 7 January 2023; Evelyn Douek, ‘The Rise of Content Cartels’ (2020) Knight First Amendment Institute accessed 7 January 2023; Gregory Day, ‘Monopolizing Free Speech’ (2020) 88 Fordham Law Review 1315. For a more European perspective, see Jan Polański, ‘The Marketplace of Ideas and EU Competition Law: Can Antitrust Be Used to Protect the Freedom of Speech?’ (2022) 2022 YSEC Yearbook of Socio-Economic Constitutions 199; Cezary Banasiński and Marcin Rojszczak, ‘The role of competition authorities in protecting freedom of speech: the PKN Orlen/Polska Press case’ (2022) 18 European Competition Journal 424.2 Douek (n 1).3 They do so within the Global Internet Forum to Counter Terrorism.4 “Coordination” in this article is meant as a neutral term that includes both collusion and cooperation.5 The quote alludes to Bork (n 1) 8.6 Timothy Garton Ash, Free Speech: Ten Principles for a Connected World (Atlantic Books 2017) 73.7 For a discussion of this concept see generally Grunes and Stucke (n 1); Polański (n 1).8 Friedrich Hayek, ‘The Use of Knowledge in Society’ (1945) 35 The American Economic Review 519; Garton Ash (n 6) 75.9 This problem has been subject to both theoretical and practical discussion. For instance, long before the rise of Big Tech, Habermas saw the growing role of business in exercising free speech as a form of “re-feudalization”, i.e. a process of restricting the “public sphere” in which public opinion is shaped, see Jürgen Habermas, Structural Transformation of the Public Sphere (Blackwell Publishers 1989) 181. In case law, albeit rather American than European, the issue was discussed in the context of attempts to apply free speech defences in relation to non-government actors, see e.g. Marsh, 326 U.S. 501 (1946), but later on was labelled as the “death of the public forum in cyberspace”, see Dawn Nunziato, ‘The Death of the Public Forum in Cyberspace’ (2005) 20 Berkeley Technology Law Journal 1115.10 Day (n 1); Polański (n 1).11 Tim Wu, ‘Is the First Amendment Obsolete?’ (2018) 117 Michigan Law Review 547.12 Douek (n 1) 11.13 See generally, Herbert Hovenkamp, ‘Whatever Did Happen to the Antitrust Movement?’ (2018) 94 Notre Dame Law Review 583.14 See, however, Section 4.2.1 and Posner’s views.15 See generally, Giorgio Monti and Jotte Mulder, ‘Escaping the Clutches of EU Competition Law: Pathways to Assess Private Sustainability Initiatives’ (2017) 42 European Law Review 5, 635.16 Content can be suppressed in various ways, which are not going to be discussed in more detail here, e.g. outright content removals, user bans, visibility reduction.17 Amazon’s Twitch provides an example. Twitch (a video games streaming platform) deletes content violating its dress code guidelines, see Tom Phillips, ‘Twitch updates its nudity and attire policy with some very specific new guidelines’ Eurogamer accessed 7 January 2023. The rationale is that some streamers attempt to raise their popularity on “sex sells” principles, which brings viewership, but goes against the long-term interests of Twitch and its reputation among streamers, viewers, and advertisers. Likewise, fake news may decrease the utility of a social media platform.18 For a classic perspective, see Milton Friedman, ‘A Friedman doctrine: The Social Responsibility Of Business Is to Increase Its Profits’ The New York Times (13 September 1970) accessed 7 January 2023. For a contemporary one, see e.g. Jean Tirole and Roland Bénabou, ‘Individual and Corporate Social Responsibility’ (2010) 77 Economica 1.19 Garton Ash (n 6) 52 makes a similar observation in relation Mark Zuckerberg and Steve Jobs.20 See Douek (n 1) 9 on self-regulation calls in the EU and suggestions that public regulations will be adopted otherwise.21 Jean Tirole, ‘Socially Responsible Agencies’ (2022) 13 accessed 7 January 2023. Furthermore, pushing undertakings to act collectively by regulatory threats might paradoxically be in the interest of political actors, since regulating speech in liberal democratic countries is difficult. The situation here might be similar to surveillance, where much information is collected by corporate actors, but can be passed to government actors. On this last issue see e.g. Wu (n 1) 249.22 This may concern e.g. advertisements and advocacy for Web3 as it might be at odds with the interests of Web2 undertakings, see Thibault Schrepel, ‘The Complex Relationship between Web2 Giants and Web3 Projects’ (2023) accessed 10 January 2023.23 See examples in Douek (n 1) 18, 20.24 Douek (n 1) 10.25 This scenario is loosely based on Signal, a privacy-oriented instant messaging app, allegedly encountering obstacles in advertising its services on Facebook (which owns WhatsApp), see Sam Shead, ‘Facebook and Signal are fighting over an ad campaign. Here’s why’ CNBC accessed 7 January 2023. The scenario extends this story by assuming that e.g. Meta (WhatsApp) agreed with Alphabet (Google Hangouts) that they will boycott competitor’s advertising (both Meta and Alphabet are present on the online advertising market). Signal’s role as a free speech forum is limited due to lack of public groups, yet e.g. Telegram offers them.26 For instance, on the technical level, during the 2020 US presidential elections, social media were able to block from circulation news reports concerning Hunter Biden and contents of his laptop, as they were considered to be fake news, see Katie Paul, ‘Twitter, Facebook restrict users’ dissemination of New York Post story on Biden’ Reuters accessed 7 January 2023.27 See e.g. Tiffany Hsu and Eleanor Lutz, ‘More Than 1000 Companies Boycotted Facebook. Did It Work?’ The New York Times accessed 7 January 2023.28 Schrepel (n 22).29 It also needs to affect trade between the Member States. While this might result in non-applicability of Article 101 TFEU, national laws typically replicate its wording – for this reason, this issue will not be further discussed.30 Case C–226/11, Expedia (ECLI:EU:C:2012:795), para 37.31 See e.g. Stefan Enchelmaier, ‘Restrictions ‘by object’ after Generics, Lundbeck, and Budapest Bank: are we any wiser now?’ (2022) Journal of Antitrust Enforcement, jnac020 .32 See generally Richard Whish, Competition Law (6th edn, Oxford University Press 2008) 126.33 Case C–67/13 P, CB (ECLI:EU:C:2014:2204), para 50.34 Opinion of Advocate General Kokott, case C–8/08, T-Mobile (ECLI:EU:C:2009:110), para 47.35 See generally Whish (n 32) 124.36 Case AT.40178 (Car Emissions).37 See generally Richard Whish and David Bailey, Horizontal Guidelines on purchasing agreements: Delineation between by object and by effect restrictions (Publications Office of the European Union 2022).38 See also Greene (n 1) 1057.39 Commission Staff Working Document, ‘Guidance on restrictions of competition “by object” for the purpose of defining which agreements may benefit from the De Minimis Notice’, SWD(2014) 198 final.40 Such an interpretation seems to be supported by Nicolas Petit, ‘Windows’ (Chillin’ Competition, 8 March 2010) accessed 7 January 2023.41 Whish and Bailey (n 37).42 Greene (n 1).43 Ibid.44 For a detailed discussion, see Crane (n 1).45 This was then quoted in a number of other cases as a source for a position that antitrust is “supportive” of First Amendment values, see Polański (n 1) 211–12.46 References provided in (n 1).47 It can be argued, however, whether this is a valid dimension of competition or rather simply a way to ensure compliance with law. Douek (n 1), for instance, appears to see it as the former.48 Case T-30/89, Hilti (ECLI:EU:T:1991:70), para 182.49 For instance, case C–228/18, Budapest Bank (ECLI:EU:C:2020:265).50 One also needs to prove that the restriction in question was “appreciable”, which does not make the case any clearer.51 See also Greene (n 1) 1097.52 Interesting signalling in that regard took place in the US in 2019. First, it was signalled that free speech can be seen as a quality improvement, see Makan Delrahim, ‘“ … And Justice for All”: Antitrust Enforcement and Digital Gatekeepers’ accessed 10 January 2023. Then, that free speech should not be assessed outside the scope of the consumer welfare standard, see Makan Delrahim, ‘The Future of Antitrust: New Challenges to the Consumer Welfare Paradigm and Legislative Proposals’ accessed 10 January 2023.53 Tirole (n 18) 4. See also: Jan Polański, ‘A Positive Program for Antitrust? Enforcement in Times of Political Tides’ (2022) 45 World Competition 237.54 Richard Posner, ‘Free Speech in an Economic Perspective’ (1986) 20 Suffolk University Law Review 1.55 In case T–451/08, Stim (ECLI:EU:T:2013:189), the General Court dealt with cultural diversity, which is more similar to free speech, but regrettably the case does not offer much guidance on the issue at hand.56 For instance, World Press Freedom Index.57 Greene (n 1) 1087 appears to concur.58 A different option could be to think about it in a way akin to the “Akzo rule” (case C-62/86, Akzo, ECLI:EU:C:1991:286) used in 102 TFEU investigations. Akzo’s logic can be simplified as: (a) if you price above ATC, you are not engaged in predation; (b) if you price below ATC, but above AVC, you are engaged in predation when additional evidence is presented; (c) if you price below AVC, you are presumed to be acting abusively, but evidence to the contrary is admissible, see Whish (n 32) 733. Think also of case C-8/08, T–Mobile (ECLI:EU:C:2009:343), para 27: “while the intention of the parties is not an essential factor in determining whether a concerted practice is restrictive, there is nothing to prevent the Commission (…) from taking it into account”. Since the discussion here is theoretical, and we merely look for some point of reference that could be used as a “model”, imagine that there is an agreement between Big Techs to counter fake news. Imagine that during elections a newspaper uncovers a controversial story that could harm one of the candidates; the story is re-printed by other newspapers. There is additional communication between the Big Techs that the dissemination of the story needs to be stopped, but there is evidence that this decision was not motivated by thorough analysis of the accuracy of the story, rather an understanding that the story could be harmful to one of the candidates. The effect of such an agreement on competition could be of limited character, yet it became more difficult for consumers to make up their minds at a critical time – hence, even if we assume that generally such a practice is not a by object restriction (if it had been, that would mimic Akzo’s Scenario C), one could consider whether the evidence of undertakings’ motivation could tip the conclusion towards a by object restriction (this mimics Akzo’s Scenario B). The example above is loosely based on Big Techs’ reaction to New York Post’s article about Hunter Biden, see Paul (n 26).59 Tim Wu, ‘After Consumer Welfare, Now What? The “Protection of Competition” Standard in Practice’ (2018) Competition Policy International 6 accessed 10 January 2023.60 ‘Sustainability Sandbox’ accessed 7 January 2023.61 Such evidence might be thus the “intent” factor mentioned in T-Mobile (n 58), para 27, i.e. something that is not legally required, but possibly of use in specific cases. See also (n 58).
Referência(s)