– Présentations PPT : Accessible aux abonnés Concurrences+ (voir ci-dessus)
– Vidéo : Accessible aux abonnés Concurrences+ (voir ci-dessous)
– Audio : Accessible aux abonnés Concurrences+ (voir ci-dessus)
– Synthèse : Accessible aux abonnés Concurrences+ (voir ci-dessus)
– Retranscription : Accessible aux abonnés Concurrences+ (voir ci-dessus)
– Articles Concurrences (Cliquez sur Voir plus ci-dessous) Consultez la section "Prochaines Conferences" pour vous inscrire aux futurs webinaires.
Paul Lugard presented the panelists and introduced the discussion. He stressed that different platforms display different features and one should be mindful of those distinct features when assessing, for example, network effects, switching costs, multi-homing, etc. He also emphasised the challenge of adequately factoring efficiencies generated by the platform into the analysis and the role of economics. Guillaume Loriot’s opening remarks focused on the European Commission’s initiatives related to digital markets and platforms. The Commission is considering adopting a new competition instrument, alongside the existing competition rules and the EU Merger Regulation. A parallel initiative consists of wider regulatory intervention under the Digital Services Act, which would have a broader scope than competition law. Regarding the adoption of the new competition tool, Mr Loriot noted that authorities have, in the first instance, been making full use of their existing tools, including interim measures. Following the special advisors’ report in 2019, the decision was taken to launch a public consultation on a potential New Competition Tool, inspired by regulatory instruments that already exist in the UK and other European countries, to deal with structural issues that are not tackled effectively or adequately under the current antitrust rules. Mr Loriot emphasised that while the consultation is agnostic as to sector, the debate surrounding the New Competition Tool has focused on digital markets. The structural issues are twofold : first, situations where a market failure has resulted in a structural lack of competition. One example concerns situations where there are entrenched dominant positions, which are not likely to be challenged because of high entry barriers and other features of digital platforms. Another example is the situation of pricing transparency (algorithms) and possible alignment in some e-commerce markets, which Article 101 may not effectively address. The second category of structural issues that may be tackled with the new competition tool is where there are structural risks to competition, because of monopolisation trends notably in tipping markets. This may especially be the case with platforms, due to network effects, a lack of multi-homing and other well-known features such as data accumulation. According to Guillaume Loriot, the aim of the NCT would be to deal with those issues in a more holistic way. However, Mr Loriot noted that the basic characteristics of competition enforcement would remain i.e., a case-by-case and facts-based approach, with due process safeguards. Where issues are detected, the Commission could propose remedies (behavioural or structural), or legislation, subject to judicial review. Separately, the Commission’s parallel workstream – the Digital Services Act – would regulate online platforms characterised as “gatekeepers” on the basis of concepts going beyond competition law and rather relying on the fairness of the trading relationships between platforms and businesses. Finally, Mr Loriot highlighted the challenges of competition enforcement : the collection of evidence, the demonstration of effects, and the assessment of efficiencies. Maureen Ohlhausen shared her views on the European Commission’s initiative from the perspective of the Federal Trade Commission (FTC) and summarised actions undertaken in the U.S. in relation to the digital industry. Although the debate is similar in the U.S. and in the EU, the U.S. has different approaches and different tools. The FTC held a series of hearings regarding antitrust and consumer protection law. The importance of data and privacy were highlighted during these discussions, and these issues have been intermingled with antitrust law. The FTC created a Technology Enforcement Division within the agency, in the Bureau of Competition, to build expertise. The FTC also conducted a 6(b) study (an information-gathering exercise), pursuant to which Apple, Amazon, Google, Facebook and Microsoft were asked to provide data on acquisitions that took place over the past ten years and which were not reportable under the Hart-Scott-Rodino Act, to determine whether the Hart-Scott-Rodino rules should be changed. The FTC sought to explore whether traditional antitrust is sufficient to tackle issues related to acquisitions of nascent competitors before they grow into threats or acquisitions of adjacent businesses. In the interim, the FTC issued a new proposed rule for reportability of transactions, which would apply economy-wide not only target tech companies. Also, earlier this year the FTC and the Department of Justice issued new vertical merger guidelines, which reflect the existing practice of the agencies, but were potentially issued to emphasise competition concerns with vertical mergers. Regarding the possibility of changing antitrust enforcement to be more aggressive, Ms Ohlhausen noted that to be durable, any new theories would need to be tested in court, the result of which could be determined by the balance of non-interventionist judges on the bench. While the debate to date has focused on the five biggest tech companies, platforms are very common throughout the economy and any changes to antitrust laws will have effects beyond this industry. Regarding privacy, Ms Ohlhausen noted that there has been a move towards a federal privacy bill and that this could also address concerns raised in the antitrust debate. Finally, Ms Ohlhausen highlighted the differences between antitrust and regulation ; antitrust aims at allowing and protecting competition on the merits, and is not intended to, and should not be used to predict or dictate certain market outcomes. David Foster focused on economies of scope, which are a characteristic of digital markets. The concept of economies of scope is of pivotal importance to designing regulation. Economies of scope are created by platforms that innovate and expand organically in adjacent services to make the most of efficiencies and synergies. So, seeing this kind of development should be a feature of the well-functioning operation of markets. This positive effect can be contrasted with anti-competitive foreclosure based on self-preferencing, which is not desirable. However, it is very difficult for regulators to tell these situations apart for a dominant platform ; for example, to distinguish a positive synergistic complementary merger from a so-called “killer acquisition”. Mr Foster then explained that all of the existing models of regulation have not been developed to deal primarily with economies of scope but with economies of scale. Those economic models are not fit for purpose considering the underlying economics of platforms. The main challenge of regulation is to distinguish synergies and spillovers. Synergies are efficiencies that are obtained from organic expansion and occur within the firm. In contrast, spillovers occur outside the firm and are efficiencies that are obtained from the ecosystem by attaching to other businesses and borrowing ideas from them without being under common ownership. On this topic, he recommended reading Capitalism Without Capital by economists and technology experts Jonathan Haskel and Stian Westlake. Good regulation requires identifying when innovation and product improvement occur best inside a closed architecture ; in that case, deep integration, self-preferencing from platforms and complementary acquisitions should be allowed and supported. If the products are developed and the benefits of competition are best obtained in an open ecosystem benefiting from spillover effects, then refusal to interoperate or self-preferencing from platforms and complementary acquisitions may raise the potential risk of foreclosure. Mr Foster stressed that there is however no easy way to determine whether markets, platform models, or products produce synergies or spillovers. The assessment requires careful analysis, and more ‘contemplative’ regulatory solutions. In this regard, market investigations – a kind of regulatory R&D – should be encouraged rather than ex ante rulemaking. Finally, regarding the agility of competition law, Mr Foster commented that although competition law investigations are often slow-moving, the impact of each case stretches beyond the case itself, the true effects of which may not be observed until many years later. Brinkley Tappan, who previously worked for the U.S. Department of Justice (DoJ), explained that the DoJ has been looking into market-leading online platforms, especially in retail and social media. The DoJ has also been investigating Google specifically for more than a year. Ms Tappan noted that the public will not have more insight about this investigation unless and until the courts have a case before them, as the DoJ has always been very tight-lipped about investigations. Public reports reveal that the investigations have been focusing on market power and activities in search and ad tech. A number of State Attorneys General, led by Texas, are also reportedly looking at the same markets. The main concerns relate to the lack of interoperability, the practice of self-preferencing, and lack of transparency. Privacy rules are also likely relevant, for example, whether privacy is a sufficient pretext for limited interoperability. The question is whether those issues, assuming that there is a problem, are effectively dealt with through competition enforcement, or whether regulation is needed. Ms Tappan suggested that proceeding on parallel tracks has benefits ; both competition enforcement and regulation have their strengths. The work of competition authorities has contributed to collecting information about the facts and the economic effects resulting from the conduct of dominant players. Competition enforcement also provides a vehicle for a wide array of structural and behavioural remedies. Regulation could play a complementary role ; it can affect positive change more quickly than can be achieved through competition investigations. Information that will be gathered in competition investigations will influence the shaping of regulations. The adoption of regulations might also push competition enforcers to sharpen their theories. Eliana Garces focused her comments on data, certain features of which are critically important in the context of regulating digital platforms. First, data is not a price. Users do not pay with data for services. Users provide data that is combined with other data to produce services and create value that is then offered back to users. The service cannot be separated from the underlying data. Second, data is not an essential facility. Data hoarding cannot be used to foreclose competitors because different data can be used for the same service. There are many ways to match a product with an audience, which compete with each other and rely on different kinds of data. Third, data must be processed, and processes are company-specific. Ms Garces stressed that access to data does not allow the recipient to replicate a service. More data does not result in a better service, rather, the process must be improved, and improving data processing is an integral part of innovation. There is debate regarding data-sharing and the idea of limiting the use of data to the purpose for which it was collected. This is counter to the nature of technology businesses and risks jeopardising technological evolution. On the other hand, companies are not fundamentally opposed to data sharing, because the source of value is not found in the data itself. That said, data-sharing is technically difficult if one wants to avoid privacy risks and preserve business integrity. Fourth, the value of the data collected and processed by the platform reverts to all of the users of the platform. Businesses that operate on Facebook benefit from Facebook’s data and data processing to target audiences, to build marketing campaigns, to establish relationships with customers, etc. Limiting access to or use of data would therefore have an impact on all the users of the platform. Ms Garces emphasised that any solution should support the creation of value and preserve the integrity of the businesses and that of the users as well as the ability to innovate and to differentiate. For example, Facebook’s competitors did not gain traction thanks to data ; rather, they offer differentiated products based on, for example, the ability to curate, the ability to react, or how users interact with one another. Oliver Bethell emphasised the importance of defining goals and measurable desired outcomes when it comes to debating amendments to competition law. Also, certainty as to the application of competition law rules over time is critical. Mr Bethell noted that the application of new regulatory frameworks has varied over time, in terms of the markets within its scope for example. Ex ante regulation risks casting the net too wide, with the result that certain products might not be introduced to the European market. Some errors can be anticipated and avoided by providing guidance on practices that will generally be considered acceptable, as well as on appropriate objective justifications. Mr Bethell stressed that experimentation and evaluation for the purpose of informed regulation require resources. Similarly, experts such as data scientists must be included in the discussion to move from broad notions towards more concrete positions. He also explained that regulation can be regarded in some cases as frictions, prohibitions and restrictions on innovation. However, regulation could also provide lubrication : well-formed regulation could allow companies like Google to have a clearer idea of what they can do.