At the end of 2020, the European Commission (“EC”) published the Digital Markets Act (“DMA”) proposal. The EC claims that its objectives (fairness and contestability) are different from those of existing competition law. It prescribes the designation of so-called gatekeepers and a set of obligations they must follow.
In a report written by Compass Lexecon professionals1, the authors explain that, as it stands, the DMA is no doubt intimately intertwined with competition policy. However, as proposed, it lowers enforcement standards by reverting to the form-based approach prevailing before the 2004 Antitrust Modernisation. Read the full report
Competition enforcement corrects for market failures related to market power. In this way, a sound competition policy ensures that markets work better and remain contestable to the benefit of consumers.
Digital markets may not work well if one or more firms have excess market power. But markets may fail for other reasons, such as information asymmetries or externalities. For example, banks may be reluctant to extend credit to customers with no credit history, leading to adverse selection. Mobile location services can generate positive externalities, such as improving traffic monitoring or contact tracing in a pandemic. But such positive externalities are difficult for the generating firm to appropriate and may result in underinvestment. In the presence of market failures unrelated to market power, contestability is not the primary concern. Making the market more contestable may exacerbate other inevitable market failures.
However, the DMA proposal is focused on contestability and thus market power. This, of course, is also the primary concern of competition policy. But the latter has limiting principles.
Dominance, for example, is well anchored in the economic concept of market power and is a necessary but not sufficient condition to establish an infringement.
A practice is not abusive by nature, but it requires assessing whether and to what extent it is likely to harm consumers. In their assessment, the authorities consider the economic, legal, and regulatory environment. They also consider all factors that impact competition, including sources of market failure other than market power. In a nutshell, the 21st century EU Competition Policy is effects-based, not form-based.
By contrast, the DMA, as it currently stands, has no limiting principles. Moreover, it abstracts from other sources of market failure that may affect digital markets. It presumes that to work well, it is sufficient that digital markets are contestable. Taking this premise as universally valid, it prohibits certain practices that could limit contestability.
The DMA ignores that some of the prohibited practices often correct for other market failures. Tying or interoperability restrictions, for example, may internalize externalities, reduce information asymmetries or incentivize investments. In some cases, where indirect network effects are strong, offering free access to one side of the market may enable the market to exist in the first place. In sum, many, if not all, of the practices “black” or “grey-listed” by the DMA can also make competition work better or directly benefit consumers, particularly in the presence of other market failures.
Regulators are familiar with these trade-offs that fall under the so-called theory of second best. In a market with some “uncorrectable” market failure, actions to mitigate another market failure may decrease overall economic efficiency and harm consumers.
Take a firm that heavily invests in developing a digital platform and succeeds in tipping the market to its advantage. Despite being dominant, such a firm is not automatically prevented from tying or refusing to interoperate, if this is necessary to monetize the service and recoup the sunk investments. Though only one firm may have succeeded, the prospects of acquiring market power incentivize the initial investments of many aspiring rivals. There might be no competition ex-ante without market power ex-post, no matter how contestable the market was.
In our report, we raise concrete concerns but also offer suggestions to address and improve the proposal.
First, the designation of gatekeepers is problematic. A gatekeeper status is not related to the concept of market dominance and does not result from a market-by-market assessment. Self-determination is also complicated because the EC retains the discretion to assign gatekeeper status to a firm not meeting the thresholds. Again, because the gatekeeper concept is merely descriptive, such determination would be arbitrary and impossible to challenge. Also, it is likely, if not virtually certain, that any firm assigned a gatekeeper status would be considered dominant under Article 102, precisely because it is a gatekeeper. But this would short-circuit a proper dominance assessment under Article 102, and the required market delineation. In this way, the proposal, if adopted, will distort Article 102 enforcement, even before any concrete enforcement action under the DMA.
To address this problem, the DMA should anchor the gatekeeper concept on market power and explain how it differs from dominance. This is the same approach taken in ex-ante regulation in the telecoms industry, regarding the designation of operators having significant market power (SMP).
Second, the “blacklist” provisions will inevitably chill procompetitive activity between firms that compete to become the leading platform in a particular market or activity. For example, efforts by one gatekeeper platform (designated as such for one core platform service) to enter a neighboring market where another gatekeeper is active are likely to become less frequent, to the detriment of competition and consumers.
The DMA should consider dropping the blacklist and replacing it with guidelines delineating the mechanism through which certain practices, in the context of digital markets, can result in consumer harm. This, of course, is the approach followed under mergers, Article 101 and Article 102, which made the EC the lead enforcer not just in Europe but worldwide and a model to many other competition and regulatory authorities.
Third, the EC claims that under the DMA it will intervene more quickly than under antitrust rules. We agree that in some cases, antitrust enforcement has been slow, undermining its effectiveness. But the solution is not to lower enforcement standards, or worse, to eliminate the limiting principles that provide indispensable legal certainty for markets to function. Often delays in enforcement are related to resource constraints or excessive risk aversion to legal challenges. On the latter point, the EC directs significant resources to mitigate legal risks. They often request hundreds of thousands of internal documents, making it challenging to find the relevant information. It is now also common in antitrust cases, not only mergers, to request vast quantities of data, merely to estimate market shares across countless permutations of candidate relevant markets – only to frequently leave the market definition open.
Addressing these “operational” problems would speed up enforcement. More importantly, defeats in Court also offer essential guidance on the boundaries of the legality of certain practices. If the EC were less risk-averse, this would not only speed up enforcement but also facilitate the gradual modernization and adaptation of competition enforcement with the support and guidance of the Courts.
Deterrence is the main benefit of antitrust enforcement. But to effectively deter, enforcement decisions must be transparent, proportionate, and evidence-based. This often requires time to investigate and gather the evidence to “build the case”. Rushed intervention under the DMA would undermine this deterrence effect.
Fourth, many of the practices in scope are already covered by existing competition rules. Thus, inevitably, strict enforcement of the DMA will conflict with the ongoing and future enforcement of competition rules. Also, due to their rigidity, the obligations may become quickly outdated. This is the same problem that led DG COMP to move away from a form-based approach to an effects-based approach in enforcing Articles 102 and 101 TFEU over a decade ago. This issue is exacerbated as the proposed DMA generalizes a few specific cases to universally applicable obligations, which also entails considerable risks. The conducts targeted by the obligations are not unequivocally detrimental to consumers and across all markets, so hardcore restrictions might hurt consumers. The proposal lacks procedural safeguards as well as a feedback mechanism for the EC to adjust it. Without these safeguards, the benefits of ex-ante obligations do not outweigh their costs.
In this respect, the solution is clear. First, drop the patchwork of hardcore obligations and provide guidelines on which practices in digital markets are likely to harm consumers and under what circumstances. Second, reinforce and reprioritize the market investigation tool focusing on three principles: (i) obligations to cooperate by all market participants, (ii) non-adversarial focus on identifying all sources of market failure affecting competition in digital markets, (iii) framing remedies constraining conduct by gatekeeper firms in a broader context that may include recommendations to other authorities.
Without the threat of fines, gatekeepers would be more inclined to participate and cooperate at the market investigation stage, leading to more informed enforcement, in turn, reducing false positives and false negatives. Ultimately, the threat of ex-post antitrust enforcement would reinforce the gatekeepers’ incentive to cooperate and resolve concerns ex-ante under a more flexible and adaptable enforcement instrument.
The authors of this report do not support the status quo as regards competition in digital markets. But the DMA echoes a form-based approach to enforcement that has proven ineffective and ultimately detrimental for consumers and businesses. The authors propose updating the DMA to embrace an effects-based approach, supported with solid market investigation powers, including tight and strict deadlines to take enforcement action. Adapting the DMA as suggested in this report will equip the EC with a modern, targeted, and proportionate instrument to address the challenges posed by leading digital platforms across the EU.
1This report was commissioned by Google. The opinions presented here are those of the authors, Miguel de la Mano, Valérie Meunier, Angelos Stenimachitis and Zsolt Hegyesi, who are employees of Compass Lexecon, and do not necessarily reflect the views of Compass Lexecon LLC or its management, its subsidiaries, its affiliates, its other professionals or clients.