Legislators around the world are currently struggling to adequately respond to the new risks that accompany innovative platform-based and data-driven business models. These risks include many problems of economic power – the traditional subject of competition law. However, according to a widely shared perception, a case-by-case enforcement of competition rules will not suffice. This triggered a number of studies, leading ultimately to the adoption of a number of legislative proposals. These include the EU’s Digital Markets Act (DMA) proposal to regulate digital gatekeepers; Germany’s reform of its competition law to endow the Bundeskartellamt with the competence to impose special rules of conduct on undertakings which have been found to be of paramount cross-market significance; and the regulatory regime for digital platforms with strategic market status recently proposed by the UK Digital Markets Taskforce.
On the surface, much of the debate appears to be about legislative technique. Just beneath the surface, fundamental conceptual questions lurk. The most prominent among these questions concerns the goals of these regulations: Are they meant to merely specify the rules of conduct that digital gatekeepers are arguably already required to obey, or should they go beyond what would follow from competition law?
This paper, available here, looks at the relationship between the Digital Markets Act and EU competition law. It suggests that the DMA should be interpreted as a measure of competition policy, if not competition law proper. In particular, it argues that the DMA and EU competition law share the same aim, although the DMA should be read as an effort to recalibrate the goals of EU competition policy: away from the protection of consumer welfare, back towards protecting competition as a process. It follows that the DMA‘s fairness goal, the gatekeeper concept and the lists of obligations imposed on gatekeepers should be read in the light of competition policy.
Section II looks at the goals of the DMA.
The announced goals of the DMA – keeping markets open and any market position contestable – has always been a core goal of EU competition law. Likewise, EU competition law clearly reflects the fairness concern that is also one of the objectives of adopting the DMA. Nonetheless, the Commission sets out to differentiate the DMA from competition law, by holding that the DMA “pursues an objective that is complementary to, but different from that of protecting undistorted competition on any given market, as defined in competition-law terms’. This differentiation is most obvious on methodological grounds. The DMA’s rules of conduct shall apply irrespective of the effects of business conduct on competition, which implicitly must be read as a breakaway from a consumer welfare test. In addition, the DMA avoids the use of established competition law concepts and terminology, like “relevant markets” or “dominance”. This can be explained on grounds of ensuring effective competition in digital markets, with their specific characteristics, but does not imply that the DMA pursues different goals from competition law.
This can be shown by looking at some of the concepts adopted in the DMA. While the DMA purports to ensure “contestable markets”, a closer look reveals that the DMA is concerned with much more than just the protection and promotion of contestability as understood under the ‘contestable markets’ theory. Instead, the DMA’s the DMA’s “contestability” goal is a proxy for the goal to ensure and promote competition in the presence of a gatekeeper in all its relevant respects – competition for the market, competition on the market and competition on other markets. As regards the concept of ‘fairness’, without more precision it is a black box. A competition policy-driven notion of fairness is mainly procedural, tempered by substantive elements. In the EU, it involves a commitment to open markets with undistorted competition, undergirded by a notion that rivalry in the market shall not be decided on the basis of power but based on “competition on the merits”, and subject to control of certain exploitative practices and requirements that a fair share of efficiencies accrue to consumers. A significant part of the obligations to be imposed upon gatekeepers under the DMA appear to relate to these conceptions, focusing on creating a fair competitive process, with some other obligations seeking to control practices akin to exploitative abuses.
To summarise: the DMA protects the same legal interest as EU competition law, namely open markets and a fair and undistorted competitive process. What differs are, first, the thresholds for intervention and, second, the legislative technique. A recognition of the clear and strong basis of the DMA in competition policy – and nothing else – would provide a clear reference point for the interpretation and implementation of the DMA.
Section III considers the fundamental assumptions and choices of the DMA as regards digital gatekeepers.
The most fundamental choice implicit in the draft DMA is the decision to complement EU competition law with a special gatekeeper regulation. It is based on the widespread perception that competition law –whether by the structure of its rules, by its methodologies or by its enforcement structure –is insufficient to address the specific competition problems of digital markets where gatekeepers are present. The special features of digital platforms that tend to raise competition concerns provide the basis for the identification of the gatekeepers to be regulated. In this setting, the DMA starts from the observation that the barriers to entry that can follow from structural features of digital markets are frequently reinforced by (anti-)competitive strategies of the incumbents. In a context where risks to competition follow from a combination of structural features and anti-competitive strategies, the DMA opts for systematically addressing the latter, as is typically the case with competition law.
If the DMA follows a competition policy logic, there are strong reasons to narrow down the concept of “gatekeepers” and to allow for more room for tailoring the rules of conduct to the relevant risk to competition. Much argues for interpreting gatekeeper power as a specific sub-category of market dominance. The gatekeeper position comes with an imbalance in bargaining power, but purely bilateral imbalances of bargaining power or dependencies will not suffice. Rather, it is the economic dependency of many, a platform provider’s broad control of access to a whole market as well as of the functioning of that market, and the particular difficulties of contesting such a market position once it has been attained, that characterises a gatekeeper position. Gatekeeper power, like dominance, is a “position of economic strength” that enables the gatekeeper “to prevent effective competition being maintained”. At the same time, it is a concept that should be extended to oligopoly settings, but in these cases special caution is called for. The focus must be on whether end users are locked into a service, or whether end users tend to multi-home on several platforms where the business users will also be present or are typically able and willing to switch platforms.
Section III also looks at the rules of conduct for gatekeepers.
The designation of a core platform provider as a gatekeeper will result in the applicability of a long number of obligations. Most of the rules of conduct can be linked to a recent competition law case, either at the European or at the Member State level. Nonetheless, they are more than an unprincipled potpourri – jointly, they are meant to address and reduce the systemic risks for competition that results from the special features of platform markets where gatekeepers are present. In light of past experience, it is relatively easy to find a sound competition policy rationale for most, if not all, of these obligations.
What is striking, however, is the regulatory technique. The DMA’s rule-setting approach is characterised by three features: (1) A shift from the broad and general competition law standards to very specific and concrete rules. This shift has received much support, particularly regarding their effectiveness and legal certainty. On the other hand, such rules lack flexibility and will, by definition, prohibit some non-problematic and allow some problematic practices. The DMA’s rules of conduct may therefore need to be adjusted, and a mechanism for doing so is currently foreseen. (2) A shift from a highly context-specific case-by-case approach towards a “one-size fits all” approach. This choice is controversial, since the core platform services to be regulated are very different and some of the prohibitions only make sense for some of them. It is also different from the German (‘case-by-case’) and UK (‘individualised code of conduct’) approaches, and, when coupled with the impossibility to tailor obligations to specific risks, amounts to an important deficiency of the draft DMA. (3) A shift towards per se rules in abuse cases. Given the over-inclusiveness of rules, the introduction of a defence, at least for some of the obligations, would seem appropriate. If a defence were to be introduced, it would not be an “efficiency defence” analogous to the efficiency defences known from competition law. Instead, a “pro-competition” defence is needed: A gatekeeper should be able to argue –and would need to prove–that the overall effect of permitting a prohibited conduct would, given the specific circumstances of its case, be pro-competitive. Merely demonstrating that the practice generates efficiencies would not suffice.
Section IV looks at enforcement and the relationship between the DMA and competition law, while section V concludes.
The draft DMA heavily relies on a centralisation of powers with the Commission when it comes to the implementation of the new regime: for example, it is solely for the Commission to establish gatekeeper status. It is much less clear, however, whether the Commission’s implementation monopoly should extend to the enforcement of the conduct prohibitions. The nature of the obligations suggests that frequently, small and medium business users will be harmed, lodging numerous complaints. The DMA should explicitly encourage private enforcement and empower national authorities to engage in a decentralised enforcement of the rules of conduct imposed on gatekeepers. A decentralised enforcement should be accompanied by strong coordination mechanism to ensure an adequate case allocation and consistent interpretation of the rules.
Merger control in gatekeeper-controlled markets remains a gap: while the DMA’s focus on rules of conduct is right, it should arguably be complemented by a special preventive review of gatekeepers’ acquisitions. At present, the DMA sets out a duty for a gatekeeper to inform the Commission of any intended concentration involving another core platform service provider or a provider of any other digital services, irrespective of whether it is notifiable under the EU Merger Control Regulation (EMCR) or under national merger control, but it does not empower the Commission to intervene. But even if the Commission were empowered to act, he SIEC test remains difficult to apply where, due to the huge uncertainty that comes with the special dynamics of digital markets, the future effects of the acquisition on competition are difficult or impossible to predict.
This paper has been praised and recommended abundantly in academic circles, and it is not hard to understand why. The author, one of the three original experts who wrote the EU Experts’ Report, and also someone who contributed to the German competition law reform, knows the topic in depth, and does a magisterial work of unpicking the DMA, its constituent elements and how it compares to competition law.
I am not fully on-board with the argument that the DMA is pure competition law, however. Of course, I agree that the DMA is heavily inspired by competition law, but it does not follow that it should be interpreted (exclusively) in light of competition law. After all, it is routinely argued that one of the main goals of certain sector regulation is to promote or replace competition, but this does not mean that this regulation falls to be interpreted under the light of competition law (see, e.g. the debates on the EU’s telecommunications regulation).
In addition, I have the impression that the author at times is effectively advocating for concepts with many possible interpretations to be read in line with competition law as much as possible – but this would mean that the DMA can be said to go beyond competition law. More importantly, the differences between the DMA and competition law are implicitly admitted by the author when she, throughout the paper, suggests that the DMA requires amendments (provisions deleted, concepts read down) to be fully aligned with EU competition law. For example, while arguing that ‘gatekeeping power’ is a form of market power, she notes that: ‘in its current version, the DMA appears to envisage a lower threshold of power, [whereby] gatekeepers are “not necessarily dominant in competition-law terms’ before dismissing this as not the correct interpretation.
Ultimately, whether the DMA will be competition law is mainly an academic question, in the best sense of the word – i.e. it is something that will matter for how it is written (which is what the author is currently arguing for) and how it will be interpreted. I expect to see more papers that touch on, even if they do not directly address the topic.