To quote this paper: A. Portuese, “The DMA and the EU’s French Presidency: The Road To Precaution and Tensions”, Competition Forum, 2021, n° 0029, https://competition-forum.com.

The Digital Markets Act (DMA) proposed by the European Commission in December 2020 appears likely to be adopted early next year during the French Presidency of the European Union (EU) . Indeed, on November 23, 2021, the European Parliament’s Committee on Internal Market and Consumer Protection approved the DMA following compromises and 1,199 amendments . On November 25, 2021, the Competitiveness Council of the Council of the EU unanimously approved the DMA (as well as the Digital Services Act ) labeled as to its “general approach” . On November 30, the European Parliament issued its DMA report. The plenary vote is due to take place on December 15. The approved text will thus become the Parliament’s mandate for negotiations with EU governments during the French Presidency of the EU in the first semester of 2022. Despite domestic presidential elections in April, the EU’s French Presidency will thus be critical for adopting the DMA especially in ways to minimize harmful unintended consequences. The French Presidency has already identified the DMA as one of the priority pieces of legislation for its EU Presidency.

These unintended consequences are considerable since the DMA regulates how innovative companies can compete in the market. The DMA does so by singling out a handful of mostly large U.S. tech companies, imposing a wide range of obligations and prohibitions on them while leaving their direct competitors exempt from such obligations . The resulting uneven playing field would generate unfair competition rather than a stronger competition on the merits.

This article discusses the extent to which the DMA reflects portrays a precautionary logic to regulate innovative companies (1). It then discusses the legislative evolution of the DMA and how the latest version reinforces the precautionary obligations imposed on the designated platforms (2). It concludes with a discussion of the EU’s French Presidency’s role in finalizing a highly problematic text in early 2022 (3).

1. The Precautionary Nature of the DMA

The DMA represents a paradigmatic shift from ex-post administrative and judicial enforcement of competition rules toward an ex-ante regulatory regime of regulation of competition. Traditionally, competition rules have laid down broad principles and rules leading to the prosecution of infringers on a case-by-case basis. Antitrust rules largely have remained part of a tort liability regime: Only harmful conduct will lead to remedial consequences.

In contrast, the DMA rules are ex-ante, meaning there are prohibitions and obligations before any harm or infringement of competition rules has ever taken place. Indeed, the Commission stated that “the current proposal minimizes the detrimental structural effects of unfair practices ex-ante, without liming the ability to intervene ex-post under EU and national competition rules.”

Consequently, the DMA is antitrust legislation since it complements ex-post enforcement of EU and national competition rules. This contradicts many officials who tried to convince the public that the DMA is a sector-specific regulation, not a broad antitrust regulation. On the contrary, it is broad, and it concerns antitrust matters.

Most importantly, the DMA is an ex-ante regulation that portrays the precautionary principle’s characteristics. This principle of regulation revolves around four main elements: i) market uncertainties do not preclude regulatory interventions; ii) a reversed burden of proof; iii) timely interventions to prevent irreversible situations; iv) hypothetical harms justify interventions. These elements underpin the DMA.

1.1. Uncertainties Do Not Preclude Regulatory Interventions

The precautionary principle operates in situations of scientific uncertainties regarding the consequences of the market actors’ conduct. But the precautionary principle commands that even in situations of scientific or market uncertainties with respect to the effect of a practice, regulators should not refrain from intervening. This approach materializes with the DMA, which enables regulators to prohibit practices whose effects remain more ambivalent than what these blanket prohibitions may suggest.

For example, one of the main practices prohibited by article 5 of the DMA consists, for the so-called “gatekeeper”, to “combine data from its core platform services with personal data from any other services offered by the gatekeeper.” For example, this would lead Amazon to no longer be able to combine data from its marketplace to offer better and more personalized service on, say, its Prime Video platform. This prohibition would prevent Amazon Prime Video from vigorously competing with incumbents such as Netflix or Hulu at the expense of competition and consumers. This practice can also prohibit Google from using the data it accumulates from its search engines activities to improve its Google Pay instruments. This would prevent Google from competing with incumbent financial institutions at the expense of competition for cheaper prices on financial instruments and at the expense of stronger financial innovation. Consequently, the uncertainties surrounding the pro- and anticompetitive effects of the practices, which are nevertheless subject to the DMA’s blanket prohibitions, are genuine.

And yet, the Commission claims to have sufficient knowledge and experience to prohibit these practices, contrary to its own Impact Assessment Report, which concluded that “for some of the practices [prohibited by the DMA] no decision or judgment is confirming its effects on the market.” Furthermore, the DMA states that “The list of obligations foreseen by the proposal has been limited to those practices (i) that are particularly unfair or harmful, (ii) which can be identified clearly and unambiguously to provide the necessary legal certainty for gatekeepers and other interested parties, and (iii) for which there is sufficient experience.” More specifically, Recital 33, mostly unchanged from European Parliament’s amendments, reads:

“Therefore, the obligations should correspond to those practices that are considered unfair by taking into account the features of the digital sector and where experience gained, for example in the enforcement of the EU competition rules, shows that they have a particularly negative direct impact on the business users and end-users.”

Contrary to the European Commission’s claims that these prohibitions have been articulated from extensive regulatory experience, almost none of the prohibitions and obligations of the DMA’s article 5 and 6 results from enforcement experience . Instead, the only substantial experience pares down to self-preferencing prohibitions with the Google Shopping judgment from the General Court, albeit such judgment is not final as it is under appeal to the European Court of Justice .

Consequently, the competitive and innovative effects of the many practices prohibited by articles 5 and 6 of the DMA remain uncertain at best, controversial at worst. Be that as it may, following a precautionary logic, the practices which often underpin innovative efforts may justify regulatory interventions on the sole basis that they are performed by innovative “gatekeepers” and not by small European SMEs. The negative bias toward innovation whenever such innovations come from the targeted market actors generates a precautionary reaction from regulators against such innovations, irrespective of their economic desirability.

1.2. Reversed Burden of Proof

An essential element of the precautionary principle is the reversed burden of proof: Everything is prohibited unless the market actor demonstrates a convincing case to not be subject to such prohibition. In other words, the precautionary principle is the economic translation of the criminal defect that considers individuals guilty unless proven innocent. The precautionary principle’s motto–‘better safe than sorry’–justifies prohibitions irrespective of their costs as long as the market actors cannot demonstrate the tangible benefits of their actions.

The DMA reverses the burden of proof against gatekeepers who can hardly, if ever, rebut those presumptions. Indeed, the shift from ex-post judicial enforcement of competition toward ex-ante regulatory rules inherently materializes the reversed burden of proof: Target conduct is prohibited for gatekeepers unless they heroically manage to reverse the burden of proof. Such prowess may be heroic given the heights of the threshold to effectively rebut the presumption.

To illustrate the broader reversed burden of proof that the DMA instills, Recital 23 provides a good example of such reversal. This recital deals with the limited possibility for gatekeepers to rebut the presumption that they are designated as “gatekeepers” under the DMA whenever they meet the quantitative thresholds laid down in the DMA. To effectively rebut this presumption, Recital 23 of the Commission’s proposal stated that:

“Providers of core platform services which meet the quantitative thresholds but are able to present sufficiently substantiated arguments to demonstrate that, in the circumstances in which the relevant core platform service operates, they do not fulfil the objective requirements for a gatekeeper, should not be designated directly, but only subject to a further investigation. The burden of adducing evidence that the presumption deriving from the fulfilment of quantitative thresholds should not apply to a specific provider should be borne by that provider.”

After the European Parliament’s amendments, the reversed burden of proof conditional to “sufficiently substantiated arguments” has morphed into a much higher threshold. The new recital reads now:

“Providers of core platform services should be able to demonstrate that, despite meeting the quantitative thresholds, due to the exceptional circumstances in which the relevant core platform service operates, they do not fulfil the objective requirements to qualify as a gatekeeper only if they can present sufficiently compelling arguments to demonstrate this. The burden of adducing compelling evidence that the presumption deriving from the fulfilment of quantitative thresholds should not apply to a specific provider should be borne by that provider.”

Thus, the reversed burden of proof comes with increased evidentiary thresholds: The company must no longer merely advance “sufficiently substantiated arguments” to rebut the presumption but must present “exceptional circumstances” to substantiate “sufficiently compelling arguments.” These amendments will reach their stated objectives: To make the rebuttal of the reversed burden of proof virtually impossible for the market actor so that the precautionary regulatory framework becomes incontestable.

This reversed burden of proof also applies to the wide range of obligations and prohibitions in the DMA. These obligations reverse the burden of proof instead of traditional competition rules since the ex-post judicial enforcement of competition rules requires the competition authority to present its arguments in the first instance. The DMA’s obligations apply without a reversed burden of proof: The “gatekeepers” need to demonstrate that specific obligations do not apply to them. However, the scope for rebuttal is again extremely limited. And again, the European Parliament’s amendments further reduced the scope for rebuttal, hence further strengthening the reversal of the burden of proof inherent to the DMA.

The strengthening of the reversal of the burden of proof following the European Parliament’s amendments is illustrated in two ways. First, the European Parliament has reduced the breadth of admissibility of efficiency considerations possibly brought forward by “gatekeepers.” In other words, it has become harder (if not impossible) for gatekeepers to justify their practices and to suspend the application of obligations and prohibitions because of the pro-competitive and pro-innovative effects of the targeted conducts. Second, the European Parliament has greatly reduced the possibility of “anti-circumvention” by extensively amending Article 11 of the Commission’s proposal: Gatekeepers will not only be able to escape Article 5 and Article 6’s obligations and prohibitions, but they will also be subject to an unlimited range of additional obligations.

On the first ground of the relevance of efficiencies for gatekeepers to justify compliance with the DMA, Article 7 paragraph 1 of the Commission’s proposal stated:

“The measures implemented by the gatekeeper to ensure compliance with the obligations laid down in Articles 5 and 6 shall be effective in achieving the objective of the relevant obligation. The gatekeeper shall ensure that these measures are implemented in compliance with Regulation (EU) 2016/679 and Directive 2002/58/EC, and with legislation on cyber security, consumer protection, and product safety.”

This version would implicitly assume that compliance is met without precluding the gatekeeper from bringing forward efficiency arguments in demonstrating compliance with the DMA’s obligations and prohibitions. However, the European Parliament’s amendments have explicitly minimized, if not eradicated, the relevance of efficiency arguments. Indeed, the text now requires “full compliance,” meaning that efficiency arguments have no role in relieving the gatekeepers’ accountability. Indeed, the new Article 7 paragraph 1 reads:

“The measures implemented by the gatekeeper to ensure full compliance with the obligations laid down in Articles 5 and 6 shall be effective in achieving the objective of the relevant obligation and the objectives of this regulation, namely safeguarding contestability and fairness for business users as well as end-users. The gatekeeper shall be responsible for and be able to demonstrate full compliance with those obligations (‘accountability’), in particular when defending its measures on the grounds of efficiency. Within six months after its designation and in application of Article 3(8), the gatekeeper shall notify those measures to the Commission and shall provide the Commission with a report describing those measures in a detailed and transparent manner and demonstrating how they ensure compliance with those obligations. The gatekeeper shall ensure that they are implemented in compliance with Regulation (EU) 2016/679, Directive 2002/58/EC and Regulation XX on a Single Market for Digital Services, and with legislation on cyber security, consumer protection and product safety.”

Beyond the complexity of the language this amendment uses, it clearly appears that the gatekeepers are accountable for demonstrating “full compliance” with the DMA’s obligations and prohibitions, and that “defending its measures on the grounds of efficiency” does not prevent the gatekeepers from remaining accountable for “full compliance.” In other words, the shift from mere “compliance to “full compliance” reveals that efficiency arguments lose relevance in assessing the gatekeeper’s accountability to enforce the DMA’s obligations and prohibitions. As a result, the efficiency considerations (i.e., cost-savings arguments, pro-competitive justifications, innovation rationale, etc..) are inevitably bound to play a much lower role in the Commission’s assessment of the gatekeepers’ accountability and enforcement of the DMA.

Consequently, with a reduced role for efficiency considerations in enforcing DMA’s obligations and prohibitions, the European Union drifts toward rules of per se illegality regarding new regulatory requirements for gatekeepers. Therefore, the shift with the DMA from ex-post judicial enforcement of competition to ex-ante regulatory rules on competition leads to a weakening (if not destruction) of the role of efficiency considerations in favor of rules of per se illegality.

On the second ground of anti-circumvention, the automaticity of the DMA’s obligations and prohibitions which characterizes the reversed burden of proof, appears stronger after the European Parliament’s amendments. Indeed, Article 11 of the Commission’s proposal was short and stated:

“A gatekeeper shall ensure that the obligations of Articles 5 and 6 are fully and effectively complied with. While the obligations of Articles 5 and 6 apply in respect of core platform services designated pursuant to Article 3, their implementation shall not be undermined by any behaviour of the undertaking to which the gatekeeper belongs, regardless of whether this behaviour is of a contractual, commercial, technical or any other nature.”

On the contrary, the new Article 11 paragraph 1 regulating anti-circumvention for gatekeepers of DMA’s obligations and prohibitions now states that:

“1. A gatekeeper shall ensure that the obligations of Articles 5 and 6 are fully and effectively complied with.
1a. While the obligations of Articles 5 and 6 apply in respect of core platform services designated pursuant to Article 3, a gatekeeper, including any undertaking to which the gatekeeper belongs, shall not engage in any behaviour regardless of whether is of a contractual, commercial, technical or any other nature, that, while formally, conceptually or technically distinct to a behaviour prohibited pursuant to Articles 5 and 6, is capable in practice of having an equivalent object or effect.
1b. The gatekeeper shall not engage in any behaviour discouraging interoperability by using technical protection measures, discriminatory terms of service, subjecting application programming interfaces to copyright or providing misleading information.”

The detailed paragraph 1 of Article 11 intends to limit the ability of gatekeepers to act in a way that may contradict the DMA’s spirit. Indeed, the new article 11 paragraph 1 explicitly not that the gatekeepers not only are subject to the wide range of obligations and prohibitions of article 5 and 6 of the DMA, but most creatively and radically, the gatekeepers are subject to prohibitions of any conduct which “is capable in practice of having an equivalent object or effect” of the conduct prohibited in article 5 or 6. Borrowing from well-known language of the internal market law, the new version of the DMA’s article 11 paragraph 1 generates a considerable (and potentially unbounded) expansion of the obligations of articles 5 and 6.

For instance, article 5 paragraph 1 of the DMA prohibit gatekeepers from requiring business users or ending users to “subscribe to or register with any other core platform services as a condition for being able to use, access, sign up for or registering with any of their core platform services identified according to that Article.” This would mean that, say, Amazon may no longer require to have an Amazon Prime account to access Amazon Prime Video: The core platform service (i.e., Amazon marketplace) cannot justify the gatekeeper (i.e., Amazon Inc.) to require consumers to sign up to benefit other core platform services (such as, say, Amazon Prime Video). This unbundling between integrated services disincentivizes innovations designed to enter adjacent markets and compete with incumbents (i.e., Netflix).

Most importantly, the new anti-circumvention article prohibits practices that are an “equivalent effect” of articles 5 and 6 of the DMA prohibitions. What if Amazon does not “require” sign up but, say, offers discounts for Amazon Prime users to subscribe to Amazon Prime Video? Would this practice have “an equivalent effect” of requiring users to sign up through an Amazon Prime account to access Amazon Prime Video given that the economic incentive would strongly (depending on the magnitude of the discount) incentivize users to sign up with an Amazon Prime Account? Could big rivals such as Netflix, Disney+, Hulu, and others complain successfully against Amazon as infringing on the DMA’s obligations based on the claim that its practice has an “equivalent effect” of the prohibited practice of Article 5 paragraph 1 of the DMA irrespective of the pro-competitive and pro-innovative effect of Amazon’s practice?

The answer is likely positive: The vagueness of the notion of “equivalent effect” expands the obligations and prohibitions included in articles 5 and 6 of the DMA to an unlimited extent. As a consequence, these articles impose per se rules of illegality despite the pro-competitive and pro-innovative effects of the prohibited practices. Beyond harming consumers and lowering (rather than fostering) competition, the new anti-circumvention article of the DMA strengthens the reality that the reversed burden of proof is an essential part of the inherently precautionary nature of the DMA.

1.3. Preventative Interventions Before Irreversible Situations Arise

The essence of the precautionary principle is to intervene quickly before the hypothetical harm may ever materialize: It supposes a willingness for the regulator to intervene in a timely and preventative manner. Correspondingly, the DMA rests upon the belief that antitrust officials need to intervene before the hypothetical harm ever materializes. But ex-post enforcement of competition rules is, per definition, a posteriori the materialization of the harm: They are part of a tort liability system, not part of a preventative regulatory mechanism. The DMA adopts the precautionary logic by endorsing the likely need to intervene “in a timely and effective manner” as stated in the Commission’s proposal:

“by addressing unfair practices in respect of core platform services operated by gatekeepers at Union-level, the functioning of the internal market will be improved through clear behavioural rules that give all stakeholders legal clarity and through an EU-wide intervention framework allowing to address effectively harmful practices in a timely and effective manner.”

Furthermore, the Commission considers that “Article 102 TFEU does not always allow intervening with the speed that is necessary to address these pressing practices in the most timely and thus most effective manner.” Consequently, the DMA “provides for timely intervention for all the identified problematic practices, while allowing for some of these a regulatory dialogue for implementing measures by the designated gatekeeper.” Even companies which are not yet “gatekeepers” but who are merely considered to be “foreseeable gatekeepers” will be subject to “early interventions” with preventative measures. Recitals 26 and 27 of the Commission’s proposal make this clear when they state that:

“A particular subset of rules should apply to those providers of core platform services that are foreseen to enjoy an entrenched and durable position in the near future […] Undertakings can try to induce this tipping and emerge as gatekeeper by using some of the unfair conditions and practices regulated in this regulation. In such a situation, it appears appropriate to intervene before the market tips irreversibly […] However, such an early intervention should be limited to imposing only those obligations that are necessary and appropriate to ensure that the services in question remain contestable and allow to avoid the qualified risk of unfair conditions and practices.”

The amended Recital 26 further reinforces the unavoidability of the early interventions under the DMA: The last sentence now reads as “in such situation, intervention may be necessary before the market tips irreversibly in favor of the largest competitor and adversely affects other competitors.” Thus, early interventions for mere “foreseeable gatekeepers” are no longer merely “appropriate” but “necessary”. The discretionary power of the Commission not to intervene early is thus further reduced to make interventions inevitable.

Furthermore, it appears clear that the DMA embodies an obsolete “big-is-bad” approach: Market tipping in favor of a small competitor is not a concern for the DMA. Still, market tipping in favor of a large competitor constitutes a ground for early interventions even though such large competitor remains a mere “foreseeable gatekeeper.” Irrespective of the innovation costs and consumer harm possibly imposed by such early intervention against foreseeable gatekeepers whenever innovation successes lead to market tipping, the amended DMA empowers the Commission to intervene broadly to slow down innovations whenever these innovations upset the dystopian view of the equality among competitors.

Such “early intervention” for “foreseeable gatekeepers” is equivalent to precautionary measures concerning the alleged risk of “market tipping”–namely, a risk that market structure becomes unbalanced in favor of few winners-take-all platforms. These timely (or “early”) regulatory interventions to limit the way companies compete are part of the “precautionary measures” that Vice-President Margrethe Vestager has advocated for many years and which materialize with the DMA. In the context of interim (temporary) measures, back in 2017, Spanish MEP Ramon Luis Valcarcel Siso asked Vestager, in a question entitled “Applying precautionary measures in antitrust cases,” whether some temporary measures (also designated as precautionary measures) could be imposed to ensure timely regulatory interventions. Vestager replied that:

“The Commission recognises that the speed and timely nature of an intervention, if necessary, may be crucial in antitrust cases. For this reason, the Commission carefully analyses in each case whether the imposition of interim measures is appropriate. This analysis is based on Article 8 of Council Regulation (EC) No 1/2003, as well as on the case law of the Court of Justice on interim measures. In addition, the Commission is particularly attentive to lessons that can be drawn from national competition authorities in the European Competition Network as regards the use of interim measures. The Commission will not hesitate to decide on interim measures in suitable cases.”

The first interim measures to be applied in twenty years occurred for the Broadcom case in 2019 . Now, through the DMA’s Article 22, the Commission appears determined to intervene early and “timely,” not merely with temporary (interim) measures, but most controversially with permanent measures. As confirmed by its amended version, the DMA supports the institutional belief expressed in the commissioned reports that any imperfection of the market structure reveals risks of market tipping, which, in turn, justifies early regulatory interventions toward (foreseeable) gatekeepers . The risks of false positives and the deterrence of firms to innovate through disruption are largely ignored—these precautionary measures designed market dynamism whenever such dynamism results from disruptors’ first-mover advantages .

Any successful innovation leads to a winner-takes-all phenomenon inherent to first-mover advantages. Moreover, these innovations are not necessarily meant to monopolize a market but rather ways to disruptively outcompete rivals by creating new markets, new methods of competition. Irrespective of these fundamental characteristics of capitalist changes, the DMA approaches innovations, and its inevitable imbalance in market structure from a precautionary perspective, thereby discounting innovation arguments for the sake of preventative interventions . The amended DMA proposal further reinforces the precautionary logic of the regulation.

1.4. Hypothetical Harms Suffice To Regulate Innovative Practices

A fundamental element of the precautionary principle is to intervene in the absence of harm: The mere risks, however remote or hypothetical such risk is, justify regulatory interventions based on the precautionary principle. The same logic underpins the DMA and its amended version. It is no longer consumer harm – as predicated by the long-traveled consumer welfare standard – which justifies interventions on competition matters. With the DMA, no harm grounds the numerous obligations and prohibitions – neither consumer harm, nor competitor harm, nor even harm to the competitive process.

The key notion of the DMA is not ‘harm’, but ‘risks’: The virtually unlimited number of obligations and prohibitions that the DMA imposes on (foreseeable) “gatekeepers” rests upon, according to the European Commission, two categories of risks: The “structural risks for competition” and the “structural lack of competition.” Indeed, the European Commission started its design of the DMA’s proposal on the belief that:

“While structural competition problems can arise in a broad range of different scenarios, they can be generally grouped into two categories depending on whether harm is about to affect or has already affected the market:
 Structural risks for competition refer to scenarios where certain market characteristics (e.g., network and scale effects, lack of multi-homing and lock-in effects) and the conduct of the companies operating in the markets concerned create a threat for competition. This applies notably to tipping markets. The ensuing risks for competition can arise through the creation of powerful market players with an entrenched market and/or gatekeeper position, the emergence of which could be prevented by early intervention. Other scenarios falling under this category include unilateral strategies by non-dominant companies to monopolise a market through anticompetitive means.
 Structural lack of competition refers to a scenario where a market is not working well and not delivering competitive outcomes due to its structure (i.e., a structural market failure). These include (i) markets displaying systemic failures going beyond the conduct of a particular company with market power due to certain structural features, such as high concentration and entry barriers, consumer lock-in, lack of access to data or data accumulation, and (ii) oligopolistic market structures with an increased risk for tacit collusion, including markets featuring increased transparency due to algorithm-based technological solutions (which are becoming increasingly prevalent across sectors).”

Similarly, as part of the background work done by the European Commission performed ahead of the DMA’s proposal, the European Commission published in June 2020 an “Inception Impact Assessment” for a “Digital Services Act package: Ex ante regulatory instrument for large online platforms with significant network effects acting as gatekeepers in the European Union’s internal market” . The Inception Impact Assessment considered that:

“Many innovative digital firms and start-ups find it difficult to bring innovative solutions, including innovative alternatives to these large online platforms, to the consumer, in particular in view of the existence of an increasing number of ‘online platform ecosystems’ that these large online platforms operate. This raises a risk of reduced competition and dynamism and consequently reduced choice for consumers and business users in the long-run and their ability to take full advantage of the digital single market.
An ability of a small number of large online platforms to comparatively easily enter adjacent markets, since they benefit from the use of data gathered from one area of their activity to improve or develop new services in these adjacent markets, increases a risk of these adjacent market also tipping in favour of these platforms to the detriment of innovation and consumer choice.”

Gatekeepers’ obligations of the DMA thus result from the “long-run” risks of reduced competition based on the mere presence of digital ecosystems and based on market tipping in adjacent markets. Therefore, it is the risks to the structure of competition–in other words, the departure from an idealized structure for market competition–which merely grounds regulatory interventions irrespective of the benefits these gatekeepers can generate to consumers.

Consequently, in the absence of actual or even potential harm to consumers or competition, the European Commission adopts a precautionary approach to antitrust by adopting the notion of hypothetical risks as a seemingly sufficient economic basis for regulatory interventions through the DMA . Indeed, the mere “risk” associated with the network effects of online intermediaries, irrespective of the benefits these intermediaries generate, suffices to impose the DMA’s obligations and prohibitions. Recital 13, for instance, reads:

“In particular, online intermediation services, online search engines, operating systems, online social networking, video sharing platform services, number-independent interpersonal communication services, cloud computing services, and online advertising services all have the capacity to affect a large number of end-users and businesses alike, which entails a risk of unfair business practices.”

Also, Recital 32 of the DMA details the purpose of the DMA, which is:

“To safeguard the fairness and contestability of core platform services provided by gatekeepers, it is necessary to provide in a clear and unambiguous manner for a set of harmonised obligations with regard to those services. Such rules are needed to address the risk of harmful effects of unfair practices imposed by gatekeepers, to the benefit of the business environment in the services concerned, to the benefit of users and ultimately to the benefit of society as a whole.”

The mere hypothetical risk of an unbalanced market structure thus justifies interventions. Otherwise, the “fairness and contestability” will allegedly be impaired. In that regard, the language of the precautionary principle – namely, a “risk of serious and irreparable damage” has remained unchanged from the Commission’s proposal after the European Parliament’s amendments as Article 22 illustrates:

“In case of urgency due to the risk of serious and immediate damage for business users or end-users of gatekeepers, the Commission may, by decision adopt in accordance with the advisory procedure referred to in Article 32(4), order interim measures against a gatekeeper on the basis of a prima facie finding of an infringement of Articles 5 or 6.”

The mere risks to the structure of competition, only perceived in the DMA as an atomized market structure, justify the DMA’s obligations and prohibitions.

In conclusion, the DMA proposal was a precautionary instrument applicable to antitrust matters . The amended version of the DMA further reinforces the characteristics of the precautionary principle. We now turn to the extent to which the legislative process of the DMA has also reinforced the stringency of the DMA concerning the regulation of innovative practices by (foreseeable) gatekeepers.

2. The Legislative Evolution of the DMA: A Confirmation, Not a Revolution

The EU Council’s “general approach” has further increased DMA’s obligations’ stringency from the European Commission’s proposal. For instance, the Council’s general approach shortens deadlines for the designation of gatekeepers, prevents further circumvention risks, and adds a new obligation as part of article 6 . Moreover, the general approach adopted unanimously on November 25 by the Council will constitute the basis for the “Presidency to start negotiations with the European Parliament.” The Council’s Presidency will be held by France beginning on January 1, 2022 . This Presidency during domestic presidential elections will prove crucial for the final version of the DMA, as discussed in the next section.

The European Parliament has increased the number of “core platform services” subject to the DMA’s ambit. The Commission’s proposal listed the following “core platform services” as part of the DMA’s remit of regulating gatekeepers:

(a) online intermediation services;
(b) online search engines;
(c) online social networking services;
(d) video-sharing platform services;
(e) number-independent interpersonal communication services;
(f) operating systems;
(g) cloud computing services;
(h) advertising services, including any advertising networks, advertising exchanges, and any other advertising intermediation services, provided by a provider of any of the core platform services listed in points (a) to (g).”

Now, the amended article 2 not only defines “core platform services” by listing the specific instances of services covered by the DMA’s obligations and prohibitions, but it also defines “core platform services” as being “a widespread and commonly used digital service that a platform service provider provides intermediates between business users and end users or within either group and.” This definition reveals that “core platform services” pares down to the role of (popular) intermediary of online platforms.

However, the intermediary function of digital platforms as efficient matchmakers generates considerable efficiencies and benefits to consumers and innovation. This intermediary function does not necessarily generate antitrust concerns, contrary to the DMA’s assumptions in digital markets . Most notably, the amended DMA expands the DMA’s scope. To the initial list of core platform services, the European Parliament has added:

• web browsers;
• digital voice assistants and virtual assistants;
• software as a service.

This expansion of the scope of the DMA, despite the lack of evidence, generated concerns . It also reveals the precautionary nature of the amended DMA with an ever-increasing ambit on competitive markets. These three additional markets covered by the DMA not only controversially expand the scope of the DMA, but they also presuppose that these markets portray a structural lack of competition with entrenched gatekeepers without any evidence of such assumptions discussed from the beginning of the legislative process.

Another considerable change made in the amended proposal concerns the quantitative thresholds for the designation of gatekeepers. The amended version of Article 3 elaborated by the European Parliament raises the quantitative threshold from 6.5 billion euros to 8 billion euros as annual European turnover in the last three financial years . This increased threshold has been explicitly designed to prevent European tech companies from falling within the DMA’s ambit, thereby targeting U.S. tech companies exclusively (and possibly one or two Chinese tech companies). In that regard, comments by the MEP Andreas Schwab, Rapporteur of the DMA, confirm the protectionist bias of the DMA concerning foreign tech companies and the clear desire to exempt European tech companies from the DMA’s obligations and prohibitions .

The European Parliament had also endorsed the controversial theory of “killer acquisitions” when they explicitly added this vague concept into Recital 10, which now states:

“systematic mergers and acquisitions should have a clear and legal threshold to put an end to killer acquisitions where big companies buy start-ups and growing companies to suppress any possible competition. A special attention should be given to takeovers in important sectors such as health, education, defence and financial services.”

Accordingly, article 12 is modified to include a notification system for any acquisition envisaged by the designated gatekeepers not only of digital companies (as provided in the Commission’s proposal) but of any company of any size (as now amended by the European Parliament). Therefore, if Google wants to acquire a coffee shop in San Francisco for its employees, Google will have to notify the European Commission. The absurdity of such measure is further expanded with the many procedural requirements added in paragraph 2 of Article 12, and also by the very fact that paragraph 3 of that article now reads: “The information gathered pursuant to this Article may be used in parallel competition cases, especially for purposes of merger control.”

In other words, any acquisition project by any gatekeeper would lead to the European Commission amassing a considerable amount of information about the gatekeeper, which can only be used at its detriment for potential future competition cases. In that context, how can the gatekeepers not be better deterred from entering into any acquisitions in the first place? Beyond the obvious harm to the gatekeeper’s ability to further innovate and compete in new markets, such formidable deterrence equating a de facto prohibition of all mergers for all gatekeepers would harm (European) tech entrepreneurs who perceive acquisitions as a viable growth strategy .

The European Parliament has explicitly mentioned that end-users (i.e., consumers) and business users must be protected by the DMA’s prohibitions and obligations imposed on gatekeepers . However, this apparent focus on consumer welfare is not shared with the Council’s general approach. On the contrary, the general approach goes on considering that “In certain circumstances, the notion of end-users should encompass users that are traditionally considered business users, but in a given situation do not use the core platform services to provide goods or services to other end-users, such for example businesses relying on cloud computing services for their purposes.” The extent to which the DMA is designed to ensure consumer welfare rather than mere competitor welfare remains subject to the upcoming negotiations.

The additional obligations, expansion of the scope of the DMA while reducing the number of companies possibly designated as gatekeepers through the increased quantitative thresholds for their designation all contribute to a precautionary perspective of the disruptive competitive exerted by the large (U.S.) tech companies: The amended DMA imposes more stringent obligations for a fewer number of companies. The asymmetric regulation distorts competition between the designated gatekeepers and their rivals, undermines their innovation capabilities at the expense of business users and consumers, and may hurt U.S. tech competitiveness without creating any substantial competitiveness for the European tech industry.

3. The Adoption of the DMA: Precaution and Tensions Over Competition and Innovation

The legislative process of the DMA appears to follow an inevitable destiny: As European politicians get involved in the assessment of the European Commission’s proposal, the reactions against the U.S. superstar firms generate protectionist reactions at the expense of European consumers and businesses, but at the benefit of short-term rent-extraction from a radical regulatory proposal .

Despite radical U.S. antitrust bills mimicking the DMA, the latter remains a unique and controversial piece of legislation, possibly upsetting already stretched transatlantic relationships. Indeed, Sean Heather, Senior Vice President of the U.S. Chamber of Commerce, writes that the DMA abandons “competition law as a tool to reign in anticompetitive behavior, opting instead to stitch together regulatory straitjacket to be selectively applied to a handful of American companies […] Europe’s proposed DMA is inconsistent with the collaborative, non-discriminatory, and plurilateral approach the EU seeks from the United States.”

Also, the DMA possibly violates the intellectual property rights and trade secrets of U.S. tech companies, thereby raising serious concerns from the U.S government. The Biden Administration has expressed these concerns recently in a document reviewed by Reuters, which states that “DMA would require gatekeepers under certain circumstances to provide competitors with information that may be protected by intellectual property and trade secret law” .

Given these rising tensions and regulatory divergences, Brussels and Washington launched on December 7, 2021, a Joint Technology Competition Policy Dialogue as part of the newly launched transatlantic EU-U.S. Trade and Technology Council . But these dialogues and councils appear fruitless concerning the DMA since the European Union has already signaled that the DMA is outside the scope of any transatlantic discussions .

The rising transatlantic tensions grow from the EU’s assertive antitrust approach to U.S. tech companies and the U.S.’s claim of the DMA’s being a protectionist tool . The upcoming French Presidency of the EU Council may have a considerable influence in minimizing, or on the contrary, exacerbating, these tensions. Given the recent miscommunication between France and the U.S. , President Macron may be eager to further reinforce the assertive antitrust approach that the DMA embodies. Indeed, as part of building a stronger “sovereign Europe,” President Macron emphasizes the concept of so-called “digital sovereignty.” And the DMA in targeting U.S. tech companies in the vain hope to help EU tech companies appears to contribute to the strengthening of such “digital sovereignty.”

Consequently, the adoption of the DMA (together with the DSA) remains one of the top priorities of President Macron for his EU’s Presidency . Given the assertiveness of President Macron on digital matters, it is highly likely that the more stringent version of the DMA as amended by the European Parliament will prevail. The Council’s general approach made few changes, mostly involving procedural changes and the involvement of national competition authorities. The more aggressive stance adopted by European parliamentarians will presumably be the preferred option of President Macron.

Therefore, as the adoption of the DMA may end up being based on the most extreme version of the proposal, transatlantic tensions may not ease. The joint competition dialogue may instead be short-lived in a time when such a transatlantic front is direly needed, given Chinese tech superpowers.
In conclusion, the legislative process of the DMA is at odds with transatlantic, and more broadly, geopolitical, considerations and odds with a more innovation-driven approach to digital competition. Rather, the newly assertive European approach favors precaution over innovation, sovereignty over transatlantic cooperation, and protectionism over free digital trade. As a result, European consumers, as well as U.S. tech companies, will be harmed.

European tech companies will not necessarily scale up since the DMA does nothing to address the problem of fragmented digital single market and the problem of the risk-aversion of European entrepreneurs. In contrast, Chinese tech companies may continue to thrive unlimitedly at the expense of the strength of Western tech companies and the democracies these companies belong to.

Dr. Aurélien PORTUESE
DIRECTOR OF THE SCHUMPETER PROJECT ON COMPETITION POLICY, INFORMATION TECHONLOGY AND INNOVATION FOUNDATION