This article is part of a series in which OECD experts and thought leaders — from around the world and all parts of society — address the COVID-19 crisis, discussing and developing solutions now and for the future. Aiming to foster the fruitful exchange of expertise and perspectives across fields to help us rise to this critical challenge, opinions expressed do not necessarily represent the views of the OECD.
Join the Forum Network for free using your email or social media accounts to share your own stories, ideas and expertise in the comments.
After several months of consultation, the European Commission submitted in December two new digital regulations: the Digital Services Act (DSA) and the Digital Market Act (DMA). If the DSA aims for better regulation of online content in order to prevent and sanction illegal content, the DMA focuses on the regulation of the digital market to safeguard the rules of antitrust law. At the heart of these two texts are digital platforms, which will be the subject of debates between the European institutions, national parliaments and Member States.
Above all, it is important to understand and define the structure of digital platforms—especially the tech behemoths. The market for digital platforms has been concentrated around a few major players over the past two decades, often grouped under the acronym GAFAM: Google, Amazon, Facebook, Apple and Microsoft. In 2019, the market capitalisation of these five companies reached USD 4 trillion—or twice that of the entire French CAC 40 stock index. While these platforms have developed according to different models, their long-term dominance is due to features common to each. This state of affairs raises serious questions about the effectiveness of competition law. There is a natural trend towards the concentration of markets when platforms benefit from an incontestable power structure.
Without necessarily reducing the structural platforms to GAFAM alone, I want to pay special attention to their shared characteristics and difficulties. It is first necessary to understand the specific business models of structural platforms, each of which has developed independently and so implies a wide variety of models. Google and Facebook's business models are based on advertising; Microsoft's is centred on the sale of software licenses; Apple derives its success from the sale of computer hardware; and Amazon's core business is based on e-commerce and dematerialised “cloud” storage services.
Beyond these differences, each company’s success can be explained largely by common characteristics that have led to their dominant market positions. Digital platforms benefit from significant economies of scale and increasing returns; the digital economy revolves around network effects and multi-sided markets; economies of scope allow platforms to expand easily into related markets; access to data strengthens the competitive advantage of platforms; digital platforms rely on locking strategies; and digital platforms play a "gatekeeper" role in the market.
Existing legislation makes it possible to provide some form of definition for digital platforms, however, “structural” digital platforms are not legally defined as such. This greatly limits the possibility of imposing specific rules upon them.
Competition law in its current state is unable to respond to the many challenges posed by the development of structural platforms, which tend to disrupt it. I consider it necessary that a definition of structural digital platforms be laid down, based on the aforementioned common characteristics. I propose establishing an analytical framework to determine the main criteria for identifying structural digital platforms. From this framework, a list of names of the platforms concerned can be established.
Anti-competitive behaviour by digital platforms can (in the case of certain platforms) already be monitored and sanctioned under common competition law, whether this involves unlawful agreements, abuse of a dominant position or control over the mergers of all the economic players involved. Furthermore, the law of restrictive practices also makes it possible to penalise behaviour relating to significantly unfair advantages. While the competent authorities may have appeared passive at first, the last few years have seen a marked willingness to pursue platforms and this is illustrated in several key rulings. The European Commission notably issued Google with a EUR 8.25 billion fine for abuse of dominant position.
The current competition law framework has a number of limitations, however. Governmental methods for analysis are ill-suited to the specifics of structural digital platforms; sanctions are not always effective because they are insufficiently dissuasive and inadequately matched with available remedies; competition law litigation is excessively long compared to the speed of change within the digital economy; and antitrust law suffers from a legal vacuum, being essentially based on criteria set according to turnover thresholds. Further, these criteria do not make it possible to oversee all the sensitive operations carried out by structural digital platforms. This gap has allowed such platforms to extend their hegemonic position considerably in principal and/or related markets.
In this regard, changes to competition law are necessary. That is why I propose to strengthen competition law in order to improve its effectiveness. Firstly, the competitive market analysis framework used by the authorities should be updated, in particular by putting less importance on the concept of relevant markets in competitive analyses. The effects of networks, conglomerate effects and data ownership should be reflected more in these analyses. I also insist on the need to redefine the notion of consumer well-being beyond the sole criterion of price.
Secondly, from a procedural point of view, there are mechanisms that can be activated for faster and more effective action: developing protective measures to avoid irremediable damage, promoting the use of behavioural remedies and reducing procedural times unanimously deemed to be too long.
Find out more about OECD work on Competition
However, I believe that these changes to competition law will not be sufficient. Schumpeter’s creative destruction has not materialised, and market concentration has reached worrying levels—a sign of competitive dysfunction. Competition law has shown its intrinsic weakness: action can only be taken on a case-by-case basis, through the application of symmetrical and essentially ex post rules. Strong market concentration around a few ultra-dominant players has worrying consequences, not only for economics but also for privacy, digital sovereignty and the democratic guarantees provided by European nations.
Faced with the intrinsic shortcomings of competition law, the establishment of ex ante regulation seems necessary. In addition to common competition law, a law to regulate structural platforms must be developed. Ex ante regulation targets certain sectors of activity which, being conducive to market failure, require greater public intervention than the essentially ex post and casuistic intervention that takes place under common competition law.
I argue for the establishment of ex ante regulation that would apply to all the structural platforms, listed according to the method I defined earlier.
It also seems necessary to provide the regulator with a range of remedies to ensure efficient regulation and succeed where competition law has failed. The objective of regulatory law is to provide the regulator with the necessary tools to supervise structural platforms and prevent breaches of competition law. This asymmetrical regulation establishes a regime of tailor-made obligations, applying to a small category of platforms and with objectives defined by the legislator. These remedies are intended to take various forms and must be applied according to a step-by-step approach.
The toolbox for this regulatory right should make it possible to promote a culture of transparency; to establish obligations and develop technical standards, to facilitate data portability and the interoperability of services; to consider certain data as essential infrastructure and consequently provide for the modalities of access to the latter; to affirm and keep the principle of terminal neutrality alive; and to enshrine a principle of non-discrimination.
Additionally, I argue for the creation of a new monopoly offence for structural digital platforms. This would prevent monopolies from forming by applying sanctions when there is a very serious chance of them becoming a reality, in particular as a result of anti-competitive practices.
Finally, I propose reforming antitrust law via the establishment of a new mandatory notification mechanism for all acquisitions envisaged by digital platforms falling within the scope of the regulator, accompanied, under strictly regulated conditions, with a provision for ex post regulation.
It is also necessary to define the kind of governance that is needed for monitoring and regulating structural digital platforms.
The first step is to designate a national oversight authority. Such an authority would make it possible to identify a specialised entity able to ensure supervision, build an in-depth relationship with stakeholders within the sector and develop expertise on economically and technically complex subjects. To avoid excessive layers of administration, this new body would not necessarily require the creation of a new administrative entity with a separate legal personality. For the sake of good administrative management and consistency of public action, I believe that it would be more appropriate for this new body to work under the auspices of an existing regulator.
Irrespective of the administrative architecture chosen, the oversight authority for the regulation of structural platforms should be endowed with extensive powers (regulatory power, sanctioning power, dispute settlement), in order to use the aforementioned toolbox.
However, this oversight authority would not be called upon to act alone in the digital field. It should therefore be integrated into the ecosystem of authorities involved in this field.
The second step is to build a strategy to increase public skills in, and knowledge on digital issues. A consensus is emerging: strengthen public expertise in digital matters. The digital economy is based on the combination of algorithms and data, so understanding how platforms work requires recruiting for these necessary skills. I propose that a real public strategy be fostered in this area by aligning the recruitment policy for state services and administrative authorities to meet these new challenges.
I want to conclude on the place of France at the forefront of future European developments. I obviously want the asymmetric regulation of structural platforms to be implemented at a European level. Nearly 500 million European users (a substantial number) would then benefit from this regulation. This would also reduce the risk of distorting the rules applicable to structural platforms between different national markets. It is therefore essential that France supports this position during the debates between Member States prior to the development of this new legal framework.
I suggest that national law (in the event that European regulation takes longer to be implemented) spur on the regulation of structural platforms in Europe. Competences would be divided between the French and European authorities based on the principle of subsidiarity—at the heart of European law. Additionally, the regulation of structural platforms will also have to be considered internationally over the long term. Current negotiations show, however, that the international supervision of digital platforms requires a joint political impetus—difficult to reach, but something for which France must work tirelessly.