ITIF Search

The Digital Markets Act: European Precautionary Antitrust

The Digital Markets Act: European Precautionary Antitrust
May 24, 2021

The European Commission has set out to ensure digital markets are “fair and contestable.” But in a paradigm shift for antitrust enforcement, its proposal would impose special regulations on a narrowly defined set of “gatekeepers.” Contrary to its intent, this will deter innovation—and hold back small and medium-sized firms—to the detriment of the economy.


The Digital Markets Act (DMA) arbitrarily distinguishes digital from non-digital markets, even though digital distribution is just one of many ways firms reach end users. It should assess competition comprehensively instead of discriminating.
The DMA’s nebulous concept of a digital “gatekeeper” entrenches large digital firms and discourages them from innovating to compete, and it creates a threshold effect for small and mid-sized firms, because it deters successful expansion.
This represents a paradigm shift from ex post antitrust enforcement toward ex ante regulatory compliance—albeit for a narrowly selected set of companies—and a seminal victory for the precautionary principle over innovation.
By distorting innovation incentives instead of enhancing them, the DMA’s model of “precautionary antitrust” threatens the vitality, dynamism, and competitive fairness of Europe’s economy to the detriment of consumers and firms of all sizes.
Given its fundamental flaws, the DMA can only be improved at the margins. The first steps should be leveling the playing field with reforms that apply to all firms, not just “digital” markets, and eliminating the nebulous “gatekeeper” concept.
Authorities in charge of market-investigation rules need to be separated from antitrust enforcers; they need guidance and capacity for evidence-based fact-finding; and they should analyze competition issues dynamically, focusing on the long term.


The “Digital” in the Digital Markets Act

The Nebulous Concept of Gatekeeper

The Gatekeepers’ Core Obligations—Article 5

Specifiable Obligations—Article 6

Precautionary Antitrust in the DMA




Updated regulations for the Internet have been looming—and were finally unfurled in Brussels in December 2020. “So, for the world’s biggest gatekeepers, things are going to have to change,” warned EU vice president Vestager.[1] The EU announced its aims to create “digital traffic lights to stop certain practices and allow others to proceed better” in a debatable metaphor that conveys its belief that web traffic should be regulated like road traffic.[2] The EU’s regulatory proposals will shape how tech companies compete, innovate, and interact with market actors in digital markets. Like the General Data Protection Regulation’s (GDPR) extraterritorial effects, the European regulatory proposals will determine the competition rules for European digital players and non-European ones whenever they operate in Europe and potentially outside.[3]

Together with the Digital Services Act (DSA), which updates the sensible E-Commerce Directive of 2000, EU commissioners Breton and Vestager presented the Digital Markets Act (DMA).[4] Deemed “pretty aggressive” even by proponents of a heavy regulatory overhaul of the digital markets, the DMA constitutes a radical change in regulating digital innovation and competition.[5] Against decades of improvement in antitrust knowledge and practice, the DMA introduces per se prohibitions of practices for a narrowly targeted set of companies—the so-called “digital gatekeepers.” These prohibitions are blacklisted practices enforced through ex ante interventions.

The DMA represents a paradigm shift from ex post analysis of antitrust liability wherein arguments are debated in courts toward ex ante regulatory obligations wherein the administration ensures compliance. The DMA exhibits the logic of the precautionary principle to competition rules at the expense of innovation.

The Commission attempted to introduce the blacklisted practices without any evidence of economic harm during the negotiations on the Platforms to Business Regulation in 2019.[6] Many member states blocked this attempt because these prohibitions violated the fundamental principle of competition on the merits.

When these proposals were introduced, the Commission provided no basis for member states to be convinced by any ex ante regulations against digital gatekeepers. And many member states remained highly skeptical of the evidentiary elements available when the DMA was introduced in 2020.[7] Despite some member states’ skepticism, the act ambitiously aims at ensuring a

contestable and fair digital sector in general and core platform services, with a view to promoting innovation, high quality of digital products and services, fair and competitive prices, as well as high quality and choice for end users in the digital sector. This … can only, by reasons of the business models and operations of the gatekeepers and the scale and effect of their operations, be fully achieved at Union level.[8]

To keep digital markets fair and open to competition, the DMA rests upon two pillars. The first pillar is a list of “do’s and don’ts for big digital gatekeepers.”[9] The second pillar is a “harmonized market investigation framework in place across the single market” to “investigate certain structural problems in digital markets” and to “take action to make these markets contestable and competitive.”[10] These rules substitute the initially announced “new competition tool” and grant the Commission increased leeway to regulate and update gatekeepers’ regulatory oversight.[11]

The DMA has narrowly confined the targets to the core platform services, or more specifically, digital gatekeepers, which are the popularly vilipended GAFAM companies (i.e., Google, Apple, Facebook, Amazon, and Microsoft).

The potential negative impact of the DMA on both digital markets and disruptive innovation cannot be overestimated. It has the power to deter innovation, distort competitive forces, and shape relationships among digital players however regulators see fit. Moreover, the implications of the DMA are significant given its inevitable extraterritorial implications.[12] Again, the paradigm shift from ex post antitrust liability to ex ante regulatory compliance represents a resurgence of the precautionary principle concerning tech companies at the expense of innovation.[13]

A confusingly consensus position has emerged concerning the necessity of an additional digital regulation at the EU level. This consensus is both bewildering and troubling. The DMA runs the risk of stifling innovation, harming consumers, and derailing the competitive process it aims to protect.

Ironically, the DMA aims to create “the right innovation incentives.”[14] This report critically assesses the DMA from an innovation perspective. Section one raises concerns regarding the definition of the “digital economy.” Section two questions the very notion of gatekeeper. Section three argues that the DMA is illustrative of precautionary-infused regulations at the expense of innovation. Sections four and five study the newly created obligations derived from Article 5 and Article 6 of the DMA, respectively, with section five specifically discussing how the DMA embeds a precautionary approach to competition matters—namely, precautionary antitrust. Finally, the conclusion summarizes the assessment and offers recommendations.

The “Digital” in the Digital Markets Act

The fundamental premise of the DMA is that the digital sector has peculiar characteristics that need to be addressed by a specific economic regulation. This section argues that digital markets should not be subject to a different competition regulation than non-digital markets. The DMA applies vertically to the digital sector. But if it is to be implemented, it should apply horizontally to all sectors of the economy, including purely digital firms, to create a fair level playing field and avoid regulatory threshold effects between rival companies.

The European Digital Strategy acknowledges that “many European companies—and [small and medium-sized enterprises] in particular—have been slow at taking up digital solutions, and therefore have not benefitted from them and missed opportunities to scale up.” The Strategy proposes, without any logic or evidence that “Europe needs to continue to act and decide independently and reduce over-reliance on digital solutions created elsewhere.”[15] But if digital solutions for small and medium-sized enterprises (SMEs) created elsewhere are superior to digital solutions created in Europe (and aren’t buyers of these solutions best positioned to make these decisions), then why would the EU want to limit best-in-class solutions for its own firms?

Invoking “European technological sovereignty,” the European Digital Strategy also identifies the need to ensure that “competition rules remain fit for a world that is changing fast, is increasingly digital and must become greener.”[16] Again, while addressing climate change is key, there is no logic or evidence that competition policies are appropriate or needed to address energy issues in the digital space.

The Commission thereby assumes that current rules are inadequate for the digital age and that the digital sector needs tailor-made competition rules—and as such, the DMA focuses on regulating the “digital sector across the Union where gatekeepers are present.”[17]

Yet, this is a new and bold position that is out of step with much EU thinking. In their advice for the “New Competition Tool,” the influential Economic Advisory Group on Competition Policy (EAGCP), which advises the European Commission, made clear that specific regulatory tools applied with a narrow scope to the digital industry were inappropriate because non-digital sectors may have the same competition concerns identified in the digital sector, thereby advising a broad scope for such a tool to preserve a level playing field essential for fair competition to take place.

Indeed, in their “Recommendation 2,” EAGCP stated, “We see a strong case for a New Competition Tool with a broad scope within and across sectors…. [because the] market features … surveyed … could in principle apply in any sector of the economy, we similarly see no benefit to limiting its applicability across sectors.”[18]

However, the DMA’s obligations are not imposed with a horizontal scope (i.e., to all companies across all sectors), but rather a vertical one (i.e., to all companies within the so-called digital sector).[19] This hinders competition and innovation because rival companies may be subject to different obligations, based not only on whether they are deemed digital but on their size (even if a smaller company is also a digital gatekeeper in a smaller market niche).

In addition, when the Commission consulted National Competition Authorities (NCAs) ahead of proposing the DMA, most argued that ex ante rules should not be applied to digital actors only: “As regards the scope of application, most respondents considered that such a tool should be applicable to all markets. Most respondents that expressed a view also indicated that the tool should not be limited to only markets/sectors affected by digitization.”[20]

The DMA not only rejects such advice but diverges from U.S. legislative proposals that do not single out the digital sector.[21] Yet, the DMA proposal itself points to the problems inherent in this approach, stating that “unfair practices and lack of contestability lead to inefficient outcomes in the digital sector in terms of higher prices, lower quality, as well as less choice and innovation to the detriment of European consumers.”[22] It then concludes that

although some of these phenomena specific to the digital sector and core platform services are also observed to some extent in other sectors and markets, the scope of the proposal is limited to the digital sector as there the problems are the most pressing from an internal market perspective. Weak contestability and unfair practices in the digital sector are more frequent and pronounced in certain digital services than others.[23]

This assertion proceeds from two implicit, and contested, rebuttable claims: The digital sector can easily be distinguished from the non-digital sector; and the digital sector has a greater probability of abuse of market power.

The Digital Sector Cannot Easily Be Distinguished From the Non-Digital Sector

Contrary to both shared beliefs and DMA’s fundamental misconceptions, there is no such thing as a digital sector or a digital market. Many industries, not just Internet platforms, are increasingly being transformed by digital technologies. Firms in more and more industries compete against one another to reach consumers through multiple business channels.[24] Digital usually represents only one channel among many others. It often minimizes costs and is usually highly competitive.

Firms using data is the “relatively straightforward cost-focused approach,” but does it imply that data-driven companies with digital solutions are necessarily digital companies evolving in digital markets?[25] The thin line between digital and non-digital companies is impossible to draw unless one grasps the notion that digital is a distribution channel—namely, an innovative business model—not a market.

Digital is a distribution channel—namely, an innovative business model—not a market.

The DMA defines the digital sector as the “sector of products and services provided by means of or through information society services.”[26] Information society services are defined as any “service referred to in point (b) of Article 1(1) of the Directive (E.U.) 2015/1535.”[27] In that Directive, there are “any service normally provided for remuneration, at a distance, by electronic means and at the individual request of a recipient of services.”[28] This article further details essential concepts that delineate the scope of the DMA:

  • “At a distance” means that “the service is provided without the parties being simultaneously present”[29]
  • “By electronic means” suggests that the “service is sent initially and received at its destination by means of electronic equipment for the processing (including digital compression) and storage of data, and entirely transmitted, conveyed, and received by wire, by radio, by optical means or by other electromagnetic means”[30]
  • “At the individual request of a recipient of services” finally specifies that “the service is provided through the transmission of data on individual request”[31]

Technological innovation means that this definition now applies to many sectors of the economy, including banking, entertainment, insurance, real estate, information (including newspapers), health care, retail, legal services, and even manufacturing. Digital transformation also suggests that digital competition disciplines non-digital market actors.[32] Jacques Crémer, one of the authors of the influential report that led to the DMA, recently recognized, “We have to adapt these laws to digitizing the economy. One of the problems is that everything is becoming a digital platform: lots of firms, even if they are not mainly digital, can have a digital part that is dominant in an important sector of the economy.”[33] When everything is digital, what is digital?

Within a given industry, digital players and analog firms compete to reach end users through distinct means, each having its own particular costs and benefits.[34] To illustrate, when Amazon decided to expand its business from selling books and music online to many any other catalog categories, it did not create a digital market. It entered (and admittedly disrupted) both traditional bookshop markets and well-identified music industry markets.[35]

Let us assume that Amazon is the only online retailer of books and music, but has only 30 percent market share for each of the two markets. Does this make Amazon a monopoly in the digital market for books and music? Obviously not, because there is no such thing as an online bookshop market distinct from the off-line bookshop market. The same is true for the music industry market, like any other category of products.[36] The only relevant market for antitrust purposes is the product market, encompassing online and off-line distribution channels irrespectively.

The French National Competition Authority (Autorité de la Concurrence) clearly adopted this position in the case of the merger between Fnac and Darty in 2016.[37] In the two French retailers’ competition assessment, the French Competition Authority considered that off-line and online markets were part of the same relevant market for antitrust purposes.[38] Indeed, the sales taking place through digital distribution channels greatly impact off-line sales in stores, and vice versa. Consequently, the competitive constraints come to play for products irrespectively of the distribution channels—be they in-store or digital.

“Digital” is not a market but a distribution channel: it is a different (often innovative) way of reaching end consumers in well-known markets. Thus, competition takes place in that product market, not on the digital channel.

This conclusion is interesting, as it underlies the increasingly irrelevant distinction between digital and non-digital distribution channels. More particularly, the blurring of the distinction between these distribution channels is encapsulated in the interesting concept of “phygital”:[39] the digital becomes physically embedded and, reversely, the physical assets are mixed up with digital strategies. The only viable strategy appears to be the omnichannel strategy.[40] This strategy optimally responds to patterns of consumption by consumers, such as “research online, purchase off-line” (ROPO) and “showrooming.”[41] As an illustration, one can see that Amazon is opening physical stores. Simultaneously, successful brick-and-mortar companies are expanding their digital distribution channel, along with the data accumulation that comes with such a channel.[42]

So what is the ratio of physical versus digital sales necessary for a company to be considered a digital company? For example, if Amazon were to increase its off-line sales, could it transform itself into a non-digital company under the proposed DMA? Regulating companies according to fixed, rigid categorizations runs the risks of misapprehending the business realities—let alone the business dynamics inherent to highly volatile and dynamic markets.

Indeed, the DMA could hinder the widespread adoption of digital technologies in each industry for two reasons. First, it creates extra regulatory costs for the industry actors that have adopted—let alone created—digital means to operate in the industry. These extra costs may undermine the competitiveness of these digital actors. Akin to the GDPR’s effects on helping big tech companies over other companies, the DMA creates extra regulatory costs that raise both existing barriers and barriers to entry—thereby preserving the current situation with the existing big tech incumbents. The dynamics of once-rapidly changing markets is perverted so that digital innovation may slow down.[43] Second, the reduced competitiveness of the digital actors as opposed to the non-digital actors in the same industry deters the adoption of digital disruption, thereby stifling technological innovation.[44]

Box 1: Data-Sharing Obligations and the Creation of an Uneven Level Playing Field

The regulatory obligation to share data, grant access, and encourage innovation among rivals makes it cheaper (if not free) for firms to copy market leaders’ innovations. The regulation makes imitation more attractive at the expense of innovation.[45] Thus, the interaction between regulation and innovation yields a negative impact since excessive and artificially created competition stifles innovation.

In other words, innovation laggards benefit from the regulation, enabling them to imitate the innovation leaders that are thus deterred from innovating at subsequent stages in order to avoid further regulatory-driven free-ridership problems.[46] Regulation-created rivalry artificially generates competition at the expense of innovative market leaders because asymmetric regulation requires disclosure and access be given to rivals—which thus gain a strategic and decisive advantage through regulation, at no cost.

This risk of free riding is present in the DMA in multiple instances. For example, the DMA states that “the gatekeepers should therefore be obliged to ensure access under similar conditions to, and interoperability with, the same operating system, hardware or software features that are available or used in the provision of any ancillary services by the gatekeeper.”[47]

This obligation, laid down in Article 5 and 6 of the DMA, overlooks the innovation dynamics resulting from the initial creation and subsequent innovations, designing an operating system with proprietary services attached to it as an incentive to create the operating system in the first place.

More practically, what would be the innovation incentives for Apple if the company were prevented from favoring its own proprietary apps (e.g., iMessage, Maps, Safari, etc.) through either preinstalled or prominent placements in the App Store? Admittedly, the prescribed equal access would prevent self-preferencing and equally undermine essential proprietary assets and services of the company’s innovations. The obligation deters innovation both at the upstream level (e.g., updates on operating systems, as Apple’s ability to appropriate its innovations would decrease) and at the downstream level (e.g., updates and creation of Apple’s apps would be hindered since the expected benefits derived from these investments would decrease).

Similarly, should Android OS be granted equal access to app developers without Google being able to self-preference its apps, Android OS’s freely and openly licensable characteristics would be put at risk. Indeed, Google may recoup its investments and innovations through a more traditional, chargeable business model.

In both instances, innovation laggards would benefit from the regulatory obligations, whereas the innovation leaders would reduce their investment levels. The overall impact on innovation and competition would by no means be guaranteed to be positive.

Consequently, the two-level playing field generated by such asymmetric regulation prevents fair competition.[48] This outcome clashes with the very objectives laid down in the regulation, which include fostering a fair competition. The DMA risks indirectly favoring less digitally innovative firms, encourages a race to the (innovation) bottom where less disruptive firms are insulated from intense competitive restraints thanks to the extra regulatory costs imposed only to more-digitally innovative companies.[49]

Thus, it is detrimental for the DMA to create an unlevel playing field between rivals that are digitally advanced and those that are less innovative but also exempt from the DMA’s obligations. Consequently, not only will the DMA’s obligations punish digital investments and innovation as opposed to traditional businesses but it may also treat digital-specific abuses of market power that may constitute instances of digital disruptions. We now discuss these potential false impressions.

To impose a specific regulation such as the DMA on digital actors while not imposing it on non-digital actors of the same industry means the competition regulators wish to enforce an asymmetric regulation at the expense of a level playing field in each industry.

Box 2: Contestability of Digital Markets or Core Platform Services?

The concept of “market” is used confusingly in the DMA. Indeed, not only are digital markets a disputable notion, as we have demonstrated, but the very concept of market in the DMA reveals an overlap with the concept of “core platform services.” This latter notion is instrumental in designating gatekeepers.

However, it remains unclear whether the DMA wants to increase the contestability of digital markets in general or increase the contestability of the core platform services more precisely. Indeed, the DMA states, “Fairness and enhanced contestability in the digital sector would result in higher productivity, which would translate into higher economic growth. The promotion of greater contestability of core platform services and digital markets is also of particular importance in increasing trade and investment flows.”[50] (Emphasis added.)

Does the DMA want to increase core platform services’ contestability as separate from the digital markets’ contestability increase? What does “greater contestability of core platform services” as opposed to “digital markets” imply?

The too-narrow focus on increasing the contestability of core platform services rather than digital markets—an already narrow objective—raises doubts about which other objectives are to be pursued. Indeed, core platform services’ contestability suggests that the DMA is exclusively designed to uproot the digital gatekeepers’ market positions in favor of other digital actors. Such an objective would be equivalent to an artificial selection of firms destined to replace current digital gatekeepers. Current digital gatekeepers seem condemned to be replaced by other digital players. It is therefore wrong to assume that current digital gatekeepers enjoy “unassailable” market positions.

By achieving greater contestability of core platform services, the DMA helps non-market leaders replace current digital gatekeepers. Such an objective does not convey clear benefits regarding consumer welfare and innovation, especially if the replacing firms provide services that may correspond less to consumer preferences. This objective also incentivizes rent-seeking behaviors of free riding at the expense of innovation incentives. To illustrate, does greater contestability of core platform services mean Google’s search engine needs to be subject to greater contestability by promoting Microsoft’s Bing or the French search engine Qwant? To pursue such an objective would ignore consumer preferences, and would inevitably deter Google’s innovation pace without increasing other firms’ innovation capabilities.

Consequently, it appears that the confusion between the promotion of greater contestability of core platform services as opposed to digital markets reveals the real intention of the Commission in this respect is to replace current digital gatekeepers with other digital players irrespective of consumer benefits and innovation. Consequently, the confusion over the notion of “digital market” reveals that the DMA’s essential notion is instead the alleged unassailable market position of gatekeepers.

Digitally Enabled Abuses or Digital Disruptions?

The digital distribution channel enables companies to disrupt sectors by offering lower prices and higher quality for products and services, thereby disrupting traditional markets by indirect entries. And they don’t impose that disruption; rather, consumers choose it freely.

Thanks to disruption, less innovative firms will be left behind.[51] At the same time, those that are unable to embrace digital disruption will be disrupted. This is a positive development that is not something to be slowed, particularly given the EU’s productivity growth crisis.[52] What is perceived as digitally enabled abuses often pare down to digital disruption—something that should be supported, not sanctioned.

Essentially new and disruptive, the enabling of digital disruption can easily be attributed to abusive practices, particularly by lagging competitors and civil society actors that inherently resist change. The convenient monopoly explanation proves useful and effective for inexperienced, dramatically new practices that are waiting to be more widely adopted by less innovative firms. First-mover advantage also comes with first-mover suspicion of disruptive practices.

One example of a practice that is assessed much more negatively online than off-line is that of default setting. Whenever a gatekeeper is considered to be enforcing default settings, such conduct is often seen as be abusive. Although it is commonly accepted that default/preinstallation settings on entry points such as handsets and browsers are not strictly exclusivity requirements, many believe that default settings can have remarkably similar effects, given consumer default bias. In his paper endorsing the DMA, Alexandre de Streel provided the example of the Google Android decision for Google Chrome and Google Search set up by default on smartphones. He raised concerns about Google paying Apple to preinstall Google as the default search engine on the iPhone/iPad.[53]

Consumer default bias can be nothing less than consumer default choice. Default settings are also standard business practices that foster asset-specific investments, minimize consumers’ search costs, and develop complementarities.

The notion of consumer default bias is both unsubstantiated and debunked by the digital markets’ evolution.[54] Contractually enforced default settings are often arrangements that minimize transaction costs among contracting parties such that innovation and competition can be maximized. Default settings can deliver fiercer competition by providing bundles of complementary products.[55] Competition does not decrease, but rather strengthens. Innovation is not deterred but made possible through contractual certainty and predictability.

The Google Android case is a prime example. Absent the Google search engine and Google Chrome being preinstalled on Android OS smartphones, Google would have never entered the smartphone market and thus would not have innovated in that sector and become a strong competitor to Apple. And absent Google Android’s default setting, Apple would have perhaps enjoyed a much stronger position in the smartphone market.

The DMA is both confusing and incomplete in this matter. First, it states, “The mere offering of a given product or service to consumers, including by means of pre-installation, as well the improvement of the offering to end users, such as price reductions or increased quality, should not be construed as constituting a prohibited barrier to switching.”[56] However, the DMA later criticizes self-preferencing: “[C]ertain software applications or services [can be] pre-installed by a gatekeeper. To enable end user choice, gatekeepers should not prevent end users from uninstalling any pre-installed software applications on its core platform service and thereby favour their own software applications.”[57]

But the issue is not about being unable to uninstall preinstalled software. The majority of people who don’t use these applications simply move them to a folder of seldom-used applications. The only competitive issue would be if the firm prevented users from downloading competitive applications which, in the Google Android decision, it was not the case.

In addition, to distinguish between “software applications or services” and the rest of the products and services in the economy is dubious. Why would default settings, in general, be accepted as pro-competitive but seen as anticompetitive whenever they concern gatekeepers’ software services? Companies preinstall products regularly, from autos that come with tires and radios to lights with light bulbs to products with installed batteries.

And what about the criticism of the Google search engine as a default setting, given it is not a software application but rather just a website with low switching costs for consumers? Indeed, consumers can easily download another free search app.

Moreover, to what extent can the preinstallation of software be objectively justified when required as part of an extensive free provision of services, such as in the case of Google Play Store preinstalled as part of the free provision of Android OS to manufacturers? Furthermore, how can providing the Google search engine as a default setting be considered problematic when rivals can also achieve default settings, such as Qwant as the default search engine on French administration computers?[58]

The DMA’s stance is both overinclusive and underinclusive. It prohibits pro-competitive behaviors while exempting behaviors similar to those that are prohibited, thereby creating detailed delineation of practices in rapidly changing markets.

Even though default settings and linked products are almost considered consumer welfare enhancing, assume for a moment that they can have anticompetitive effects in the off-line world. In such a case, the particular focus made in the DMA suffers inconsistency in addressing anticompetitive effects in general. Default settings are inherent to business life.[59] For example, when people buy a car, the default setting is to have the radio and tires installed. Mandating choice at the expense of default rules may alter consumers’ “choice architecture” without providing tangible benefits.[60] It is widely accepted that “well-designed product or service defaults benefit both company and consumer by simplifying decision making, enhancing customer satisfaction, reducing risk, and driving profitable purchases.”[61] Indeed, if the goal is consumer welfare, these default settings clearly do that.

Default settings, while neither good nor bad, are designed to efficiently provide consumers with the products and services they expect at a satisfactory quality level. As such, the Commission itself acknowledged precisely this in its Google Android decision of 2018:

The reason why pre-installation, like default setting or premium placement, can increase significantly on a lasting basis the usage of the service provided by an app is that users that find apps pre-installed and presented to them on their smart mobile devices are likely to “stick” to those apps … Users are unlikely to look for, download, and use alternative apps, at least when the app that is pre-installed, premium placed and/or set as default already delivers the required functionality to a satisfactory level.[62]

Consumers stick with default settings precisely because they are satisfied with the quality of the provision of services they receive. The belief that default settings only lead to anticompetitive consequences is only partially true. In reality, default settings are widely accepted and typically produce satisfactory outcomes in the off-line world. Unfortunately, the DMA endorses such a partial view. Therefore, to prohibit default settings only in digital markets while allowing them off-line would not only hurt consumer welfare but generate unfair competition.

Another example of a practice that, when conducted off-line, is praised but despised when conducted online, is charging low prices. Charging low (or no) prices should be acclaimed by antitrust enforcers. However, low prices are loathed by antitrust enforcers when digital companies do it, as they are perceived as predatory prices or abusive business models for zero-priced products and services.[63]

Popularized by the Neo-Brandeisian Lina Khan, the idea according to which big tech companies charge predatory prices first came to the fore with Amazon’s successful competition in the retail sector.[64] Without economic data and evidence of the essential marginal cost, Khan made the assumption that Amazon charged prices below its marginal costs before recoupments.[65] Unsubstantiated and unevidenced since the publication of the influential article in 2017, this assumption nevertheless generated groupthink such that it is now taken for granted that Amazon and other tech companies charge predatory prices.[66]

Instead of traditional boutiques’ high-expenditure strategy, online businesses’ cost-saving rationale is overlooked whenever claims that low prices are predatory prices. From the perspective of traditional businesses, it may appear that digital retailers engage in predatory pricing. Rather than being an abuse of competition, because of low-cost business models, price competition is the essence of the competitive process.[67] In that regard, one must recall Schumpeter’s description of the “perennial gale of creative destruction” when he wrote,

In analyzing such business strategy ex visu of a given point of time, the investigating economist or government agent sees price policies that seem to him predatory and restrictions of output that seem to him synonymous with loss of opportunities to produce. He does not see that this type’s restrictions are, in the conditions of the perennial gale, incidents, often unavoidable incidents, of a long-run process of expansion that they protect rather than impede. There is no more of paradox in this than there is in saying that motorcars are traveling faster than they otherwise would because they are provided with brakes.[68]

Unfortunately, the DMA implicitly enshrines Khan’s flawed and ultimately ideological assumption. Indeed, it follows her argument that large platforms prioritize growth over revenue, and long-term market capitalization over short-term capital returns:

In addition, a very high market capitalization, a very high ratio of equity value over profit, or a very high turnover derived from end users of a single core platform service can point to the tipping of the market or leveraging potential of such providers. Together with market capitalization, high growth rates, or decelerating growth rates read together with profitability growth, are examples of dynamic parameters that are particularly relevant to identifying such providers of core platform services that are foreseen to become entrenched.[69]

Yet, what is most ironic about this attack is it is precisely the kind of behavior many analysts and critics have argued EU and U.S. companies should engage in. In the 1980s and1990s, Japanese companies were praised in the West because, much like large platforms today, they prioritized growth over revenue, and long-term market capitalization over short-term capital returns. In the last decade, while Western companies have been regularly attacked for focusing on stock market values and short-term capital returns, a set of companies that are not doing that has emerged. They are behaving in ways that maximize long-term shareholder and societal value—yet are being attacked for it.

Higher efficiency, scale economies, more significant innovation, and network effects are all enablers for digital platforms to offer lower prices and thereby enter different lines of business.[70] According to the DMA, these positive parameters illustrate potential leveraging effects that are made possible by predatory pricing. Thus, “low prices” perceived positively offline are, according to the DMA, considered digital abuses of dominance in the form of “predatory prices.”

The same is true regarding zero-priced products. Ad-funded products are attacked for data accumulation abuses and subsequently construed as abuses of dominance. The ad-funded business model has existed since the rise of the advertising industry in the early 1900s, and has traditionally been praised rather than contemned by regulators.[71] However, overlooking again the unparalleled consumer benefits associated with zero-priced products, especially to lower-income consumers, antitrust enforcers shun the pro-competitive benefits of zero-priced products to effectively prosecute and investigate practices that would, in the off-line world, be praised.

Would TV channels be investigated if TV advertisers enabled TV programs to be broadcasted for free while accumulating data on viewers’ attention patterns through such services as Neilson ratings? Would newspapers with free online services be investigated if they showed targeted ads? Would the French inventor J.C. Decaux have been investigated if he has offered free bus stops to communities under the condition that they include advertising spaces?[72] Ad-funded business models are old, and have, unsurprisingly, entered the online space together with the digital disruption. Indeed, Wired magazine has argued that this business model is becoming widespread, even coming up with the term “freeconomics.”[73]

The DMA makes ad-funded zero-priced products suspicious, whereas similar off-line zero-priced products are exempt from such suspicion and heavy obligations.[74] Due to the disruptive effects of digital innovation on traditional businesses as illustrations of “gales of creative destruction,” what are considered conventional commercial business practices become digitally-enabled abuses of non-dominance under the DMA because they are ex ante prohibited.[75]And regulators should remember that if the free model were made more difficult, upper-income Europeans would likely see insignificant impacts, as they could easily afford the resulting subscription fees. Lower-income Europeans would either be digitally excluded or have to make tough choices about where to spend their limited income.

Market Tipping and Market Concentration

One final rationale, only recently introduced in competition debates, for specifically regulating digital markets as opposed to other markets is that digital markets tend to “tip”. Market tipping suggests that one or few companies earn inordinately high profits and obtain market positions that are almost invulnerable to competition. Market concentration in the digital sector reaches unparalleled dimensions, as recital 26 of the DMA encapsulates,

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. The same specific features of core platform services make them prone to tipping: once a service provider has obtained a certain advantage over rivals or potential challengers in terms of scale or intermediation power, its position may become unassailable and the situation may evolve to the point that it is likely to become durable and entrenched in the near future.[76]

Market Tipping as a Superfluous Concept

The DMA’s underlying assumption is that gatekeepers exist only because network effects generate market tipping. But what are the characteristics of market tipping? When do they materialize? Market tipping describes a situation where firms have market dominance, thus describing a situation already covered by current EU competition rules. Market tipping is thus a superfluous concept in light of the traditional notion of market dominance.

The NCAs, in their contributions to the impact assessment of the new competition tool, defined market tipping: “So-called tipping (or ‘winner takes most’) markets are markets where the number of users is a key element for business success: if a firm reaches a critical threshold of customers, it gets a disproportionate advantage in capturing remaining customers.... only one or very few companies will remain on those markets in the long term.”[77]

The DMA allegedly needs to be adopted because, unlike other markets, digital markets are prone to tipping. Otherwise, the DMA would have a broader scope.[78]

The DMA further specifies that “undertakings can try to induce this tipping and emerge as gatekeepers by using some of the unfair conditions and practices regulated in this Regulation.”[79] This statement is confusing on multiple grounds. If market tipping is a consequence of companies’ driving force to gain more market power by competing against rivals, then it seems to deride the very process of competition.

Access to a gatekeeper position should not be a competitive concern. When market dominance is grounded in competition on the merits, it becomes a legitimate reward for effort and innovation. Also, potential entry by new competitors disciplines incumbents. Scaling-up is a legitimate objective as long as the competition takes place on the merits. Thus, market tipping sanctions the process of competition.[80]

The statement also refers to “unfair conditions and practices.” But if the identified practices are anticompetitive, the concept of market tipping becomes pointless to investigate, prosecute, and sanction practices that, no matter what, violate competition rules.

Market Tipping and Dominance

If dominant companies engage in anticompetitive behaviors, then Article 102 Treaty on the Functioning of the EU (TFEU) is already entirely actionable to sanction illegal conduct. Market tipping proves to be a useless concept for dominant companies.

Suppose the nondominant companies engage in anticompetitive behaviors. In that case, competition rules are specially designed not to be concerned with those behaviors because of these practices’ de minimis aspect: They are too minimal to have tangible effects on the market.

Also, market tipping appears fruitless with respect to the concept of dominance. Does this imply that market tipping refers to some “collective dominance”—meaning a dominant position enjoyed by multiple undertakings? Again, the concept of market tipping is superfluous. Article 102 TFEU already provides for the legal basis necessary to investigate competition problems of collective dominance. Indeed, it explicitly refers to the prohibition of abuses “by one or more undertakings” of a dominant position. Collective dominance, or “joint dominant position,” is a well-known and developed concept by the Court of Justice.[81] Indeed, the Court stated that

the expression of “one or more undertakings” in Article [102 TFEU] implies that a dominant position may be held by two or more economic entities legally independent of each other, provided that from an economic point of view they present themselves or act together on a particular market as a collective entity. This is how the expression “collective dominant position,” as used in the remainder of this judgment, should be understood.[82]

Consequently, since collective dominance positions fall within the ambit of Article 102 TFEU, market tipping proves superfluous with the existing notion of “collective dominance.” Market tipping does signify a collective dominance position. Moreover, the concept of collective dominance is itself also caught within the remits of Article 101 TFEU.[83] Therefore, because competition rules already address issues of collective dominance, market tipping as a concept tailor-made for digital markets adds nothing but confusion and redundancy.

Furthermore, the confusion of the above statement of recital 26 of the DMA reaches its climax with the last throng of the sentence: “regulated in this Regulation.” Practices are now considered unfair whenever the proposed DMA identifies them as unfair, even though they were not considered unfair until now. Such retrospective application of the competition laws is legally erratic and hazardous. It is also economically harmful, as it creates considerable risk for legal costs, which may generate anti-innovation risk-averse attitudes.

Market Tipping and Concentration

Does market tipping refer to market concentration? It seems that market tipping refers to an oligopolistic market structure wherein few firms dominate the defined relevant market. With respect to market concentration, market tipping is either superfluous (as opposed to merger control) or obsolete (suggesting a return to a structuralist approach).

But does market tipping refer to market concentration in the sense of merger control? In other words, does market tipping refer to the alleged inefficacy of merger control to slow down acquisitions by large companies? It does seem to be the case, as Article 12 of the DMA lays down an obligation for gatekeepers to notify the Commission of any envisaged acquisitions. Article 12(1) of the DMA reads,

A gatekeeper shall inform the Commission of any intended concentration within the meaning of Article 3 of Regulation (EC) No 139/2004 involving another provider of core platform services or of any other services provided in the digital sector irrespective of whether it is notifiable to a Union competition authority under Regulation (EC) No 139/2004 or to a competent national competition authority under national merger rules.

A gatekeeper shall inform the Commission of such a concentration prior to its implementation and following the conclusion of the agreement, the announcement of the public bid, or the acquisition of a controlling interest.

If market tipping designates market concentration created by mergers (rather than by collective dominance), the notion of “market tipping” in the DMA proves problematic. Indeed, merger control is commonly considered to have been unable to scrutinize previous mergers realized by large platforms. American Neo-Brandeisians and European Ordoliberals concur on the assessment according to which merger control has been too lenient and insufficiently well equipped to deal with the acquisitions made by large digital platforms.[84]

Consequently, large companies often face tremendous competition, and that scale can boost innovation and productivity.[85] As Bighelli et al. demonstrated,

[W]e find positive and significant correlations between rising sector-level concentration and increases in sector-level productivity and allocative efficiency. Increasing market concentration in Europe should not necessarily be seen as a cause for concern related to a weaker competitive environment. This has important consequences for antitrust and industrial policy, which must carefully evaluate the costs and benefits of increasing market concentration.[86]

Increased efficiency through concentration enables companies to cope with the competitive constraints larger players exert. Thus, concentration at national levels will enable firms to compete with not only their direct rivals at both European and international levels, but also with other large players in adjacent markets. In that regard, one can wonder whether the DMA’s underlying assumption that market concentration is detrimental to markets’ functioning is indeed a “careful evaluation” of the state of affairs.

Let us assume these criticisms—namely, that merger controls wrongly approve mergers that should have been rejected—hold some water. Considering that merger controls remain blind on many acquisitions, this argument does not in and of itself justify the DMA. Indeed, altering merger control criteria now would mean the European Union Merger Control Regulation (EUMCR) of 2004 should also have been amended. Instead, Article 12 of the DMA adopts this baroque position to change the rules of concentration control in Europe without amending the primary regulation relevant for such control. Accordingly, Article 12 of the DMA appears bound to refer to the EUMCR but exclude its applicability in that respect.

In summary, the market tipping concept that underpins the DMA appears to be redundant in that it refers to either the notion of “dominance” and current rules and decisional practices provide for sufficient enforcement mechanisms, or the notion of “concentration through merger,” and thus, the relevant EUMCR should have been amended to provide greater clarity, consistency, and coherence. However, the concept of market-tipping redundancy is profoundly harmful to adequately understanding how the markets work.

Market Tipping as a Harmful Concept

The very concept of market tipping is problematic. In defining the concept, Motta and Peitz argued that “because network effects reward firms with large customer base, they are subject to market tipping, a notion which captures the idea that once a firm has obtained a certain advantage over rivals in terms of market share, its position may become unassailable, and the market may tend to a situation of monopoly.”[87]

Each of the elements of this preliminary insight is worth discussing.

Box 3: A Critical Assessment of the Characteristics of Market Tipping

First, to consider that network effects reward firms with a large customer base is right, but only to the extent that the double-edged nature of network externalities is accounted for. Network effects not only reward firms with a large customer base, but also punish firms with large customer losses. Indeed, network externalities are either positive or negative, as the low entry costs associated with network-enabled rewards (i.e., a virtuous circle enjoyed by an innovative firm) are indistinguishable from the low exit costs network-enabled punishments (i.e., a vicious circle experienced by a non-innovative firm).[88] Consequently, network effects cause firms to grow or shrink with a reward/punishment system inherent to the competitive process.[89] The competitive-process dynamics demonstrate that network effects do not represent causes of unassailability unless the fundamental element of temporality is considered. Indeed, a large customer base may rapidly erode due to negative network externalities. In short, network effects emphasize competition for the market instead of competition within the market.[90]

Second, market tipping may refer to instances wherein some firms have obtained a certain advantage over rivals in terms of market share. This is better described by the substitute to market tipping, which is the “winner takes most” phenomenon. After relinquishing the theoretical and misleading notion of “winner takes all,” the literature appears to have reached a consensus on the worrying concept of “winner takes most.”[91] Obviously, “take” refers to taking market shares. [92] However obsolete the notion of “market share” in digital markets may be, the “winner takes most” concern is at odds with the fundamentals of the competitive process.[93] It is precisely the prospect of gaining a comparative advantage over rivals that constitutes the engine as well as the result of competition and innovation. Absent such a hopeful chance, a competitor may no longer compete or innovate, and may be deterred from entering the market altogether. Therefore, designating the detrimental notion of “market tipping” through the description of the positive outcome according to which “winner takes most” is tantamount to refusing the very process of competition on the merits. It would indeed constitute the consecration of competition on the lack of merits. “Winner takes few” or “loser takes most,” both opposite outcomes to market tipping, would destroy the dynamics of the competitive process to protect inefficient and less-innovative firms.

Third, market tipping may refer to some allegedly unassailable market position. The pretense of knowing, on par with the regulator, what is and what is not unassailable is highly debatable. Experience teaches us that what were once deemed unassailable have become passé market actors. Also, current market conditions prove that unassailability is all relative given the extreme rivalry and fascinating innovation portrayed by some market leaders. And finally, unassailability is time relevant, intellectual property (IP)  rights dependent, and subject to tremendous discretionary power. The alleged unassailability of gatekeepers supposes that these firms enjoy a stable entrenched market position and should be regulated as “natural monopolies.”[94] Although highly controversial, especially in digital markets characterized by disruptive innovation, the notion of gatekeepers being inevitably prone to natural monopoly regulations is questionable, and yet perceptible in the DMA.[95] For these reasons, it appears that any unassailable market position is instrumental to the qualification of a market actor as a gatekeeper, irrespective of the error-cost judgments of such a conclusion on the alleged unassailability.

Fourth, once it has been concluded that a market actor enjoys an unassailable market position, however realistic such a conclusion may be, it still remains to be proven that the said market actor tends to monopolize. In parlance reminiscent of Section 2 of the U.S. Sherman Act, market tipping thus suggests a willingness and an ability to monopolize the market. Therefore, a market actor that has taken most of the market shares of a market due to both inherent market characteristics and superior efficiency is expected not to grow further in market shares. Should that company grow further, it may be considered that the market tipped and the company is likely to be treated as a gatekeeper. Such a company, absent evidenced anticompetitive practices, shall be subject to an entire range of regulatory prohibitions because its conduct has become suspicious irrespective of consumer benefits and innovation efforts.

According to NCAs, market tipping leads to i) efficient or innovative firms disappearing; ii) less competition overall; iii) less consumer choice; and iv) higher consumer prices.[96] Despite risks of “stifling of innovation incentives,” NCAs recommend intervening early with ex ante competition tools since “Article 101/102 TFEU are not suitable and sufficiently effective instruments” for addressing market tipping.[97]

NCAs’ argument ignores potential scale and scope benefits from market tipping and assumes that more consumer choice and more competition overall is good. If this were true, would consumers not be better off if there were six major aircraft companies, rather than just two (Airbus and Boeing). After all, there would be more competition and choice. There would also be significantly higher costs and less innovation because companies would be less able to cover their massive fixed costs from what would be dramatically reduced sales. The optimal industry structure, at least in innovation industries, is not Adam Smith-like hypercompetition among many rivals. Another way of saying this is competition is not the goal but rather a means to the end of productivity and innovation.

Moreover, these negative economic consequences are hardly evidenced in digital markets.[98] Rivals exiting from the marketplace may be attributed to the inevitable result of the competitive process, should the remaining firms epitomize superior innovation or efficiency (through lower prices and higher quality). Innovation is beneficial to consumers and society and may irremediably kick out the market’s innovation laggards. Even when innovation is minimal, the displacement of a less-efficient company can be inherent to the competitive process’s normal functioning. This is the case for imitation—legitimately considered not to be a powerful, innovative endeavor. Nevertheless, imitation itself can deliver consumer benefits and spur innovation.

One of the numerous illustrations is provided when Apple freely offers as a new OS feature a digital service initially supplied by a third-party developer at a cost. As this refers to Apple’s decision to integrate into Sherlock (Apple’s early search function in OS) features similar to the third-party software called Watson, the practice is referred to as “Sherlocking.”[99] This is especially the case when such disappearances occur because consumer welfare is increased.[100] In the case of Apple’s Sherlocking of certain products, consumers were given for free a product they once either ignored or had to pay for. Recently, Google was criticized for providing Google Maps for free on smartphones with the complaint that doing so hurt companies that were selling digital mapping services for hefty fees. This kind of complaint represents the ultimate confusion and reversal of what competition should be about. Competition policy should not be about firms’ welfare. If it were, it would be tantamount to significant hindrances to innovation. In short, consumer welfare and innovation are essential to any antitrust enforcement.[101]

Antitrust enforcement may in fact encourage “zombie firms” (low productivity companies that would typically exit in a competitive market) to remain.[102] In contrast, market tipping follows from this aggressive, yet undoubtedly competitive, process of creative destruction. Therefore, it is certainly unclear exactly what tipping means. To be sure, it not only implies that markets have well-accepted market definitions but also reinvigorates an overdue structural analysis to digital markets. On the contrary, market tipping is a highly contestable notion because the market itself is often poorly defined. The structural analysis belittles the analysis of competitive pressure exerted at the firm level.

Market tipping is a justificatory concept for intervention whenever markets are deemed not to be ideally structured (i.e., an atomized vision of markets). Market tipping is the modern version of the rebuffed structuralist perspective of the structure-conduct-performance (SCP) paradigm that discards innovation considerations.

The DMA describes digital markets as prone to market tipping. Yet, other sectors can concentrate, sometimes much more firmly than can the digital sector. Second, there is nothing inherently bad about tipping as long as consumers continue to benefit and innovation is strong. The fact that the major digital firms are also, according to the EU R&D 5000 index, among the leading R&D investors in the world, suggests that such concentration is pro-innovation.

Moreover even assuming that digital markets tip, the network effects at the source of tipping can be equally powerful to un-tip the markets as consumers find value in other networks. As market tipping results from a “winner takes all” feature of the digital economy, the associated “loser loses all” feature cannot be ignored. Low exit barriers and massive crowding-out effects lead consumers to switch rapidly. In that regard, market tipping is fragile, transitory, and subject to considerable disciplining forces from rivals. With the COVID-19 crisis, massive crowding-out effects took place at the expense of platforms such as Slack and Skype to the benefit of platforms such as Teams and Zoom. The fall of powerful incumbents can happen whenever consumers massively exit the market (or multi-home with the choice of using rivals’ products simultaneously). A case in point here is the rapid rise of alternative social media platforms to Facebook. However, Facebook is erroneously portrayed as having tipped the social media platform market (if such a market even exists). The sudden rise of alternative social media platforms such as TikTok, Snapchat, Clubhouse, and many others demonstrates the ability of negative network effects to un-tip the allegedly tipped market.

Market tipping is loosely delineated, and can refer to market concentration, which itself is also depicted more positively as market consolidation.[103] Market consolidation may result from legitimate competition, as companies respond to intense rivalry by consolidating their market position. But sometimes diseconomies of scale and other frictions can threaten them to be soon outcompeted.[104] Indeed, Motta and Peitz recognized this widely accepted phenomenon:

We acknowledge that in markets with scale economies, network effects, or switching costs, there may be fierce competition during the period prior to market consolidation, with profit sacrifices being made on the expectation of future profit recovery after the market has consolidated. Ex post intervention should therefore include considerations of the legitimacy of such dynamic business strategies; i.e., firms should not necessarily be denied the recovery of upfront investments and profit sacrifices.[105]

In addition to patent holders, large companies, blessed with network externalities and disruptive innovations, profit from innovation. The network effects function as returns on investments to recoup the sunk costs from current or previous lost investments. Recoupments can take place on either side of the multisided platform. Indeed, Evans considered that “the economics of platform businesses suggests that certain practices that may appear anti-competitive—recouping losses from ‘low prices’ on one side through ‘high prices’ on the other side—are natural, pro-competitive practices.”[106]

Large sunk costs of innovating and building a network necessitate recoupment strategies—namely, subsequently being profitable with products and services whose marginal costs of producing are very low. Consequently, the platform’s added network externalities and quick growth indicate not anticompetitive conduct or considerable market power, but rather recoupment strategies that are the result of regular post-investment practices deriving from innovations and technological breakthroughs. In other words, bigness via market consolidation may not illustrate anticompetitive practices, especially when network externalities are essential, and that competition through innovation takes place for the market rather than within the market.[107]

In that framework, market consolidation not only results from competitive pressures and represents efficiencies by reaping scale economies, but it also enables subsequent competition to take place. Large firms will fiercely compete against one another at a comparable scale and be able to make the kinds of large investments necessary to mount a competitive challenge.[108] If market consolidation can now be perceived as pro-competitive, then the fundamental assumptions underpinning the DMA crumble.

What the DMA considers to be market tipping is nothing but market concentration, which itself is a market consolidation potentially corresponding to efficient responses (i.e., scale economies) and intense competitive pressures from existing or potential rivals.

Multiple examples illustrate these market mechanisms, from supermarkets merging to cope with online competition exerted by Amazon and others to training companies trying to merge to anticipating Chinese competitors’ entry to Apple entering the car industry in order to compete with another tech company—Tesla.[109] These concentrations are, in fact, market or firm consolidations for immediate innovation capabilities and subsequent competitive tensions.[110] The case for or against market concentration often fails to take into account the enhancement of firms’ capabilities and too often pares down to legal discussions without economic soundness. As Janusz Ordover wrote, “Arguments for and against a merger that turn upon distinctions between broad and narrow market definitions are, to an economic purist, an inadequate substitute for, and a diversion from, sound direct assessment of a merger’s effect.”[111]

Consequently, mergers need to be reassessed from a more dynamic perspective—namely, without focusing on the static, short-term effect on market concentration. Some have nations adopted reforms in that direction or are considering them.[112] However, the DMA would treat acquisitions envisaged by the so-called gatekeepers as suspect. Article 12 of the DMA proceeds from the negatively connoted notion of “market tipping,” which blurs the distinction between desirable and undesirable market concentration.

In conclusion, the notion of “market tipping” erroneously conveys the belief that digital markets tend to tip, as it assumes digital distribution channels are markets separate from other product markets. It also equates market consolidation (due to network effects, efficiency, and innovation) as systematically detrimental to competition and innovation. These assumptions are unsubstantiated and even contradicted by economic history and market realities.

Digital markets should not be treated, for antitrust purposes, separately from other markets, as advised by the NCAs themselves. In that regard, the recent U.S. Competition and Antitrust Law Enforcement Reform Act introduced by Sen. Amy Klobuchar (D-MN) to reinvigorate antitrust at least has the benefit of avoiding the fundamental pit into which the DMA falls.[113]

The Nebulous Concept of Gatekeeper

Large digital platforms are targets of the techlash, at least among the punditry.[114] The DMA reflects that techlash by targeting the digital markets only—and within those markets, only the superstar platforms that would be labeled as gatekeepers. The regulatory playing field is now unlevel, thereby impeding the very notion of “competition on the merits.” But the lynchpin notion of “gatekeeper” is legally vague and economically damaging.

The reason for the EU’s focus on gatekeepers is the belief that some platforms have amassed unparalleled market power, enjoy entrenched market positions, and are likely to prevail in their unassailable positions. Without a doubt, these well-identified firms are just a handful of corporations, mostly American and Chinese.[115]

The DMA’s notion of “gatekeeper” is legally indeterminate and economically harmful. It castigates competitive success and introduces wrong incentives via tremendous threshold effects.

The term “gatekeeper” has changed.[116] First referred to in 2019 as “quasi gatekeepers” in the Platform-to-Business Regulation’s (P2B Regulation) proposal, these large platforms have become full “digital gatekeepers” with the DMA’s proposal.[117]

The DMA is premised on the view that advanced economies never experienced such large and powerful corporations in the past. However, since the rise of the industrial revolution, which enabled firms to take advantage of economies of scale, superstar firms have consistently operated and have frequently raised similar concerns as those expressed in the DMA. Superstar firms (and individuals) often fail, either due to the magnitude of the investments involved, from lethargy, or, most often, from tectonic shifts in technology.[118]

To be sure, some platforms are big.[119] In 2020, with a brand value of $352 billion, Apple was the largest tech company, followed by Microsoft at $326 billion and Google at $324 billion. However, Chinese Tencent and its $151 billion brand valuation show the growing power of the so-called “BATX”—Baidu, Alibaba, Tencent, and Xiaomi, the four biggest tech firms in China.[120]

Market capitalization increases have resulted from both consumer demand during COVID-19 and innovation.[121] In short, these companies are undoubtedly big—but bigness is neither bad nor new.

Box 4: Bigness in Economic History

Contrary to the idea behind antitrust, big is often good. Indeed, big businesses on average deliver many benefits to society compared with smaller ones.[122] This holds true for most indicators, including wages and benefits, worker health and safety, productivity, research and development (R&D) and innovation, and exporting, as large firms mostly outperform small firms.[123] As Robert Atkinson and Michael Lind noted,

In short, on virtually every measure, large businesses perform better than small. This is not meant to denigrate small “Main Street” businesses. Most small business owners take risks, work hard, and contribute to their communities. But we should not let sentiment get in the way of reality. Economic prosperity will be determined principally by large firms, not by small firms, and least of all by the vast majority of small firms whose owners do not intend them to grow beyond a few employees.[124]

Yet, it is popular to disparage large firms, and elected officials can gain voters’ approval to defend nostalgic and romanticized idea of small firms against big bad corporations.[125] Nevertheless, large firms provide superior benefits to society, and it is the ability of start-ups to get big that determines the long-term success of the entrepreneurial spirit.[126]

Superstar firms almost always succeed because they make investments in intangible capital, benefit from network effects, and are more efficient.[127] Higher efficiency, enabled through innovation and capital investments, is a better explanation for the rise of superstar firms than speculations about the alleged diminution of competition and the oft-discussed under-enforcement of antitrust laws.[128]

Not only are big companies generally good, but they are not new. Former member of Franklin Delano Roosevelt’s Brains Trust, Adolph Berle, stated that “in 1962, bigness is familiar. And more of it seems inevitable, as new consolidations make regular, and not unsympathetic, headlines throughout the country.”[129]

The Schumpeterian perspective of the process of creative destruction describes technological innovations replacing incumbents with new entrants that compete disruptively with established large firms. In any respect, bigness is neither new nor surprising, as it emerges from superior efficiency, first-mover advantage, or both, according to innovations.

In addition, we traditionally overstate the problem of monopoly, notably by ignoring both international competition and technological disruption.[130] In that regard, Schumpeter wrote in 1942 about the “imaginary golden age of perfect competition that at some time somehow metamorphosed itself into the monopolistic age, whereas it is quite clear that perfect competition has at no time been more of a reality than it is at present.”[131]

He further noted that “the modern standard of life of the masses evolved during the period of relative unfettered ‘big business.’”[132] He considered that big business could deliver a greater quantity and quality of products precisely because they use their strategic market position:

But when all the facts of the case are taken into consideration, it is no longer correct to say that perfect competition wins out … For through a concern that has to accept and cannot set prices would, in fact, use all of its capacity that can produce at marginal costs covered by the ruling prices, it does not follow that it would ever have the quantity and quality of capacity that big business has created and was able to create precisely because it is in a position to use it “strategically.”[133]

Big firms’ strategic position corresponds to large platforms’ unavoidability assumption as essential criteria for designating gatekeepers—although the unavoidability assumption is both overstated and misconceived.[134] Digital gatekeepers are Schumpeter’s industrial behemoths of today, driving innovation because of intense competitive rivalry. Digital gatekeepers often experience monopolistic competition—or “moligopolies”—which enable them to derive profits for further innovation.[135] Competition for large companies, often with other equally large companies, can be very intense despite the reduced number of firms. Duopolies are, in that respect, instances when competitive rivalry can be fierce.[136] For example, Boeing competes intensely with Airbus, as does Google’s Android OS with Apple’s iOS.

Competition for large companies stems from both entrant companies and foreign companies that enter a market having already accumulated a user base and valuable experience in their home market.[137] Prime examples include Chinese Alibaba and Xiaomi competing with Amazon, and Apple, respectively.

Irrespective of the merits of digital gatekeepers, the DMA intends to intervene and regulate rather heavily the ongoing digital revolution by targeting the so-called digital gatekeepers. But exactly who are those gatekeepers? What are their characteristics?

Large tech companies, designated as gatekeepers under the DMA or as dominant firms under Article 102 TFEU, see their conducts assessed with suspicion, if not outright reprehension. Such suspicion hinders both gatekeepers’ innovation capabilities and their market partners’ innovation capabilities.

The DMA designates gatekeepers using qualitative criteria as well as quantitative criteria. Qualitative criteria are laid down in Article 3(1) of the DMA, while quantitative criteria are presented in Article 3(2). According to the qualitative criteria of Article 3(1), in order for a provider of core platform services to be designated as a gatekeeper, it must:

  1. have a significant impact on the internal market;
  2. operate a core platform service that serves as an important gateway for business users to reach end users; and
  3. enjoy an entrenched and durable position in its operations, or will foreseeably enjoy such a position in the near future.

Criterion (a) refers to the subject designation: The requirement of a “significant impact on the internal market” implies that a large platform is concerned. Criterion (b) refers to the gatekeeping status stricto sensu: The requirement for the large platform to be an “important gateway for business users” implies that the platform has become an “unavoidable intermediary.” Criterion (c) refers to the contestability element: The large platform's requirement to enjoy “an entrenched and durable position” implies that the gatekeeper’s market position has allegedly become uncontestable. The following chart recaps these qualitative criteria. The designated subjects (i.e., large platforms) generate structural risks to competition because they have become an unavoidable market player and enjoy unassailable market position.

Article 3(1) Qualitative Criteria 1

Before delving into the quantitative criteria of Article 3(2) of the DMA that aim to objectivize the gatekeeper designation, the most controversial aspect of such designation pares down to the qualitative criteria outlined in the DMA. The following section reviews each criteria—namely, that gatekeepers are large digital platforms that are both unavoidable and unassailable. It then discusses the quantitative criteria before validating the concept of gatekeeper.

Digital Gatekeepers as Large Digital Platforms

The Commission defines digital gatekeepers as large online platforms.[138] While the DMA implicitly endorses the flawed motto of “big is bad,” it seemingly creates with the notion of “gatekeeper” a different concept than the well-known (and well-practiced) concept of dominance. Large platforms are identified as the winners of the digital competition:

Large platforms have emerged benefitting from characteristics of the sector such as strong network effects, often embedded in their own platform ecosystems, and these platforms represent key structuring elements of today’s digital economy, intermediating the majority of transactions between end users and business users. Many of these undertakings are also comprehensively tracking and profiling end users. A few large platforms increasingly act as gateways or gatekeepers.[139]

Intuitively, the concept of gatekeepers seems to refer to very strong levels of market dominance (i.e., the controversial concept of “super dominance”).[140] Super dominance, also referred to as a “quasi-monopoly” or “overwhelmingly dominant position,” arises whenever a firm controls more than 70 percent of its market share. European competition history has tersely used the concept of super dominance.[141] Although distinctions for antitrust purposes between a super-dominant position and a mere dominant position are yet to be deciphered, the concept of super dominance can nevertheless be said to be a useful, convincing concept. Super dominance may suggest a quasi-monopoly market position.

In the Microsoft case, the European Commission found that the company attained an “overwhelmingly dominant position,” thereby leading Microsoft to bear “particularly” the “special responsibility” which pertains to dominant undertakings.[142] In this case, such an “overwhelmingly dominant position” led Microsoft to being bound to ensure the interoperability of its PC operating system with software manufacturers.

Consequently, under EU competition rules, super-dominant firms can be compelled to grant access to the interface and provide interoperability with its business users. Assuming that these obligations are well founded, as they were in the Microsoft case, it cannot be legitimately argued that EU competition law does not provide the adequate tools to force super-dominant companies to grant access to some of their data or facilities.

Contrary to the notion of “super dominance,” the notion of “gatekeeper” suggests the enjoyment of such strong dominant positions that such a company does not merely enjoy a quasi-monopoly market position subject to it being assailed by rivals, but instead enjoys an unassailable status resembling the “quiet life” of monopolies, as described by John Hicks.[143]

In reality, not only do firms not need to be super dominant to be classified as gatekeepers, but most puzzlingly, they may not even enjoy dominance. The DMA regulates digital gatekeepers because they generate competition risks even if they are nondominant in their relevant markets. This confusing aspect is plainly stated: “It should be noted that a gatekeeper may not necessarily be dominant within the meaning of Article 102 TFEU.”[144]

Under traditional competition laws, the exercise of dominance in the market is not prohibited (only abuses are), but the antitrust authorities need to prove dominance (often by resorting to contestable market definitions). Market definitions are fundamentally ill-suited to analysis in digital markets. To include nondominant companies in the category of gatekeepers is an implicit acknowledgment of the Commission’s flawed analysis of market definitions in significant cases. Indeed, the Commission has consistently depicted the traditionally referred-to gatekeepers (i.e., Google, Facebook, Apple, Amazon) as dominant or super dominant.

Both past and ongoing lawsuits are built around the finding that the targeted companies are super dominant in their respective relevant markets. Consequently, to extend the notion of “gatekeepers” to nondominant firms generates incommensurable legal uncertainties for many small and medium-sized companies that might fit into the notion of “gatekeeper.” It also constitutes an embarrassment for the Commission’s claim that gatekeepers have “significant market power, even in the absence of dominance,” which is economically baseless and conceptually flawed.[145]

Nevertheless, in the face of mounting evidence, the Commission appears forced to admit that its aims lead to the sanctioning or regulating of conduct of companies that are considered nondominant under Article 102 TFEU. In other words, the Commission is willing to have it both ways. It aims at suing tech companies under Article 102 TFEU because they are alleged to be dominant in their respective markets; and if such claims of dominance are rebutted, the Commission looks to prohibit conduct under the DMA that falls outside the remits of Article 102 TFEU.

Colloquially, the Commission wants its cake, and to eat it too. However credible its claims of market dominance are, it can effectively target large tech companies because, dominant or not, these companies will be liable under either Article 102 TFEU or the DMA.[146]

The confusion is total when one delves into the details of whether the Commission has built its belief that digital gatekeepers are powerful notwithstanding them enjoying a dominant position or not. Contradictory statements abound, including within the single document of the “Factual summary of the contributions received in the context of the open public consultation on the New Competition Tool.”[147] Indeed, after having stated that gatekeepers “may not necessarily be dominant,” the Commission acknowledged that, among NCAs, “several of these respondents indicated that gatekeepers should be regarded as having market power and thus a dominant position.”[148] Furthermore, among the structural competition problems identified by the European Commission, it appears that “vertical integration and the acquisitions of competitors by dominant players” stand as prime competition risks generated by gatekeepers.[149] Gatekeepers are thus once again considered to be dominant players.

Finally, among the structural competition problems identified in the document, the Commission considered that “structural competition problems may arise in markets where a (not necessarily dominant) company with market power in a core market applies repeated strategies to extend its market position to those related markets.”[150] This phrasing is the one adopted in the DMA proposal wherein the Commission indeed considers that “a gatekeeper may not necessarily be a dominant player, and its practices may not be captured by Article 102 TFEU if there is no demonstrable effect on competition within clearly defined relevant markets.”[151]

The Commission seldom refers to coherently construed competition law concepts when they are instrumental in achieving its goals, but ignores and undermines these very concepts whenever the economic reality does not correspond to its regulatory agenda.[152]

The baroque evolutions of competition law epitomize the paradigm shift of the DMA. Ex ante regulatory interventions substitute the ex post engagement of antitrust liability. Ex ante precautionary interventions inherent to such paradigm shift constitute a chilling effect on innovation for entrepreneurs. 

To sanction nondominant companies, irrespective of the principles laid down in Article 102 TFEU, is both economically harmful and legally weak. These justifications are twofold. The digital gatekeeper, while nondominant, has become either an unavoidable intermediary or has attained “unassailable market power.” Both legal justifications unreasonably widen the ambit of EU competition rules. And both assumptions regarding gatekeepers are questionable—and now being questioned.

Digital Gatekeepers—the Unavoidability Assumption

A fundamental assumption about the designation of digital gatekeepers is the perception that these companies find themselves to be “unavoidable” for business users that want to reach end users. This fundamental assumption of the DMA is an excessively strong one.

Paving the way for the DMA proposal, the Crémer Report of 2019 introduced the concept of “unavoidable trading partner” in mainstream policy thinking at the European level. Indeed, the report states that “even in an apparently fragmented marketplace, there can be market power. This kind of market power is linked to the concept of ‘unavoidable trading partner’ and has sometimes been called ‘intermediation power’ in the area of platforms.”[153]

The notion of “unavoidable trading partner” fails to prove useful or illuminating, as intermediaries are the multisided platforms that best match supply and demand—namely, service providers and end users—through digital technologies.[154] They consist of an endless number of platforms with countless different business models. Still, notable instances include the intermediaries in the “gig economy” (e.g., Uber, Bookings, Airbnb, and BlaBlaCar), in the advertising sector (e.g., Google and Facebook), and in the device sector (e.g., Apple Store and Google Play Store).

As matchmakers, intermediaries are not specific to the digital economy, with multisided platform actors are present in traditional industries such as advertising, payments, other financial services, and energy.[155] The DMA distinguishes digital intermediaries from traditional intermediaries without a clear rationale. Indeed, although the DMA does not explicitly refer to the unavoidability assumption, it considers digital intermediaries to be of particular concern for antitrust purposes because they capture the majority of transactions:

Large platforms have emerged benefitting from characteristics of the sector such as strong network effects, often embedded in their own platform ecosystems, and these platforms represent key structuring elements of today’s digital economy, intermediating the majority of transactions between end users and business users. Many of these undertakings are also comprehensively tracking and profiling end users.[156]

These large intermediaries are said to enjoy “weak contestability and unfair practices” because they “mostly directly intermediate between business users and end users.”[157]

The intermediary status is fundamental to the DMA. Although it does not use the term “unavoidable,” the Commission Staff Working Document, which underpins the DMA’s rationale, refers to it.[158] Stakeholders’ responses also appear to have influenced the Commission. Many advocated for the recognition of the notion of unavoidable trading partner: “For the definition of gatekeeper platforms, some stakeholders suggest considering the criteria of ‘economic dependence’ on certain platforms which makes them unavoidable trading partners and make it more difficult for consumers to avoid dealing with them.”[159]

More specifically, the Commission makes this notion its own when it specifically states,

The analysis also revealed power imbalances among platforms that are reflected in data sharing arrangements. Google and Facebook have the central position in online marketing and advertising, to the extent that they are unavoidable trading partners, including other platforms from the analysed sectors. This puts them in a position to determine the terms and conditions of data access and data reuse.[160]

Unavoidability here may refer to Google’s and Facebook’s market powers in the advertising markets. The distinction of the digital advertising market as separate from the media advertising market can, first of all, be contested from a marketing perspective. Companies advertise online, or on TV, or radio, or in newspapers, or a combination of these media channels. Even if one considers the narrow digital advertising market as separate from the broader media advertising market, both Google’s and Facebook's unavoidability in that market can still be questioned.

Indeed, with approximately more than 40 percent of market shares in major European digital advertising markets, Google can be said to be in a dominant position. With only about 25 percent of the market share, Facebook can hardly be said to enjoy a dominant position.[161] Instead, other competitors such as Amazon and other social media platforms such as Twitter constitute non-negligible market players.

For instance, a clothing company may want to advertise exclusively on Amazon as a key e-commerce platform. In contrast, a pasta company may want to advertise exclusively on TV during prime time to catch people’s attention during dinner.[162] For the DMA’s Staff Working Document, to revert to the alleged ad duopoly of Facebook and Google to illustrate the “unavoidability” assumption demonstrates how strong (and, thus unsubstantiated) this assumption proves to be. Indeed, nondominant companies are inherently avoidable—thereby excluding Facebook and other digital advertising from the category of being an unavoidable trading partner. But dominant companies are also avoidable by business partners.[163]

Moreover, one of the services detailed as part of the core platform services—a notion essential to the designation of gatekeepers—is the “online intermediation services.” In the DMA, online intermediation services refer to “information society services” of Article 2(2) of the P2B Regulation wherein it is acknowledged that these services “allow business users to offer goods or services to consumers, to facilitate the initiating of the direct transaction between those business users and consumers, irrespective of where those transactions are ultimately concluded.”[164] These intermediaries are essentially transaction cost minimizers.[165] They efficiently match supply and demand in once-inefficient markets such as the taxi industry being disrupted by Uber, Lyft, and other intermediaries.[166] Can we assume that Uber or any other app from the gig economy has become unavoidable such that competition from traditional actors, say taxi drivers, has become nonexistent? The unavoidability assumption of the gatekeepers is problematic because it is unrealistic.

This assumption of intermediaries as unavoidable trading partners is both unsupported economically and legally misleading. It is a concept that is vague and contrary to what certain authors may attempt to advance. Indeed, Alexiadis and de Streel considered that “the implications of dealing with an ‘unavoidable trading partner’ are clear in a digital platform environment where the platform operator has market power.” However, the implications are anything but clear, especially when looking at what they consider to be the key characteristics of an unavoidable trading partner:

Particular manifestations of the doctrine of an unavoidable trading partner lie in the twin ideas that:

  • Unilateral market power might exist in relation to “must have” content where certain TV programmes are critically important commercial inputs in order to be able to attract advertising and subscribers.
  • Conglomerate market power might exist where a merged entity holds strong positions across a number of neighbouring markets, especially where customers feel that one or more of those products enjoy a “must-have” quality.[167]

Consequently, a platform alleged to have must-have qualities—namely, superior quality products or innovative characteristics—may fall within the category of unavoidable trading partner and thereby qualify as digital gatekeepers, subject to quantitative criteria met. Such a hasty conclusion may prevent superior efficiencies, innovative ideas, and consumer preferences from being fully expressed and reaped. Indeed, outcompeted rivals due to inferior quality and innovation will lead the successful platform to fall within the ambit of the stringent gatekeepers’ regulation of the DMA. Must-have qualities can hardly be argued as being a clear, objective standard through which trading partners can be assumed to be unavoidable.

More generally, the notion of “unavoidable trading partner” is relatively recent in European competition literature and practice. We can trace its origins in the notion of an “obligatory trading partner,” which first appeared in the Virgin/British Airways case in 1999. Virgin Atlantic Airways complained against British Airways’ rebate schemes. In its Statements of Objections, the Commission considered that British Airways was dominant in the U.K. market for travel agent services, notably based on “the fact that it is an obligatory trading partner for travel agents wishing to offer a full service to their customers.”[168] Also, in the Deutsche Bahn case, it was considered that the German Railways had a “statutory monopoly” on rail-transport services and thus placed “those seeking the services in a position of economic dependence on the supplier.”[169]

This notion of “obligatory trading partner” is a translation from the French commercial code “partenaire obligatoire,” but is mostly influenced by German competition law.[170] Indeed, Section 20 of the German Act against Restraints of Competition is entitled “Prohibited Conduct of Undertakings with Relative or Superior Market Power.” Section 20(1) particularly introduces the concept of “relative market power,” as German competition also applies to “undertakings and associations of undertakings to the extent that small or medium-sized enterprises as suppliers or purchasers of a certain type of goods or commercial services depend on them in such a way that sufficient and reasonable possibilities of switching to other undertakings do not exist (relative market power).”[171]

In accordance with German Ordoliberalism’s commitment to defend small and medium-sized companies against large companies, German competition law influenced the development of the notion of “unavoidable trading partner” from that of “obligatory trading partner.” In the development of this vague notion, it has commonly been accepted that the concept of an obligatory trading partner was only a means of defining the dominant position. Indeed, the Commission stated in 1986,

When considering dominant positions of a vertical or conglomerate nature, a somewhat artificial definition of the relevant market is sometimes required. Recourse to the concept of obligatory trading partner, which concerns the relationship of dependence which may exist between two economic operators owing to their commercial dealings, may prove to be useful. The Commission is studying the possibility of employing this concept as a supplementary tool of cases of abuse of dominant positions.[172]

However, with the DMA, the notion of an “unavoidable trading partner” becomes disconnected from the notion of dominance. Instead, the notion of an “unavoidable trading partner” is restored to impose obligations despite the lack of a dominant position. In that regard, the shift from the application of the notion of “obligatory trading partner” applied to public utility matters (e.g., train and airlines industries) to the application of the notion of “unavoidable trading partner” to nondominant companies is perplexing.

Thus, the Commission drops the consumer harm requirement since nondominant companies become subject to regulatory obligations absent consumers’ or rivals’ harm. The Commission applies to highly competitive companies and rapidly changing markets the insights that were once developed for public utility markets, such as markets dominated by legal monopolies. In order words, the public utility’s thinking, and its associated essential facilities doctrine, is wrongly applied to a fundamentally different sector—the digital industry.

Box 5: Unavoidable Trading Partners as Situational Monopolies?

The notion of “unavoidable trading partners” relates to the notion of “situational monopolies” described by Michael Trebilcock.[173] He distinguishes between situational monopolies and “structural monopolies.” Any superior bargaining power enjoyed by a company may transitorily place that firm in a position of situational monopoly, as the business user's dependency on the platform enables the latter to charge quasi-monopoly prices. Situational monopolies may result from the “serendipitous circumstances of the interaction between the two parties in question.”[174] On the other hand, antitrust rules address the notion of “structural monopolies.” Should we adopt Trebilcock’s distinction between situational and structural monopolies, this distinction further undermines the relevance of the unavoidability assumption.

Indeed, suppose unavoidable trading partners are situational monopolies in Trebilcock’s dichotomy. According to Trebilcock, competition rules must not be an antitrust concern but rather a contract law concern. Bargaining imbalances, along with fairness in trading negotiations, must remain adjudicated by the courts that review contractual provisions and are subject to either contract law principles or any particular applicable regulatory provision.

This is precisely the case in Europe with the recently adopted and universally accepted P2B Regulation. This 2019 Regulation specifically addresses the contractual provisions platform intermediaries must comply with in dealing with trading partners. Consequently, assuming that unavoidable trading partners are situational monopolies, as hinted by Trebilcock, the European regulatory framework (via contract law principles and specific regulation) already states the numerous contractual principles applicable to platform intermediaries. Consequently, the DMA is both superfluous with respect to the P2B Regulation and ill-suited as it undermines the effectiveness of the contract law principles.

On the other hand, suppose that unavoidable trading partners are structural monopolies in Trebilcock’s dichotomy. Then, one must bring the evidence that the platforms are dominant in their relevant markets and that such a dominant position has led to abusive practices.

Because competition law ought to be concerned with structural monopolies and not situational monopolies, the concept of unavoidable trading partner becomes unnecessary, considering the well-accepted concept of dominance.

To characterize a platform as being “unavoidable” indeed overlooks the relativity of such a market position. A platform may retain many unavoidable elements in the short/medium term only because it competes by innovating. Here is where the concept of unavoidability of the platform proves to be indistinguishable from the concept of unassailability introduced in the DMA.

Either the platform's unavoidability refers to a nondominant position, and current contractual provisions must apply, or unavoidability of the platform refers to a dominant position, and current competition rules of Article 102 TFEU can apply.

Should regulators eventually embrace a dynamic perspective, the allegedly unavoidable platform will have become assailed and thus no longer be unavoidable. Reversely, from a dynamic perspective, the so-called unassailable position of the platform becomes irrelevant once the platform has been avoided because of disruptive innovation. The prospect of assailing the platform becomes pointless if the platform’s services and products have been outcompeted by disruptive innovation. For a platform to be labeled as a digital gatekeeper, it requires some unassailable position—but it implies that what is to be assailed is worth competing over. This is a rebuttable belief.

Digital Gatekeepers—the Unassailability Assumption

Gatekeepers are regulated because they are deemed to be unassailable—a questionable belief that should not, but does, underpin the DMA. The assumption of unassailability relates to the allegedly powerful status ascribed to intermediaries—for platforms considered dominant or nondominant. On the contrary, the DMA’s assumption is not so much related to status but to the alleged unassailable market positions enjoyed by gatekeepers.

The DMA identifies three fundamental aspects for core platform providers to be designated as gatekeepers. According to Article 3(2) of the DMA proposal, a core platform service provider is a gatekeeper if:

  1. it has a significant impact on the internal market;
  2. it operates a core platform service which serves as an important gateway for business users to reach end users; and
  3. it enjoys an entrenched and durable position in its operations or it is foreseeable that it will enjoy such a position in the near future.[175]

The first element refers to the size of the platform, the second element refers to the unavoidability assumption previously discussed, and the third and final element refers to the unassailability assumption.

As opposed to the unavoidability assumption, which was present in the Crémer Report, the unassailability assumption was not explicitly mentioned therein. The Commission created this controversial notion—in the Staff Working Document accompanying the DMA proposal— as a consequence of “market tipping”:

The presence of network effects and the multi-sidedness of certain markets imply that even markets where initially multiple competitors are active are particularly prone to tipping: once a firm has obtained a certain advantage over rivals in terms of market share, its position may become unassailable, and the market may gravitate towards a situation of dominance or (quasi)-monopoly.[176]

More importantly, the DMA assumes that regulation is needed to address unassailable market power issues by digital platforms, positing that some market power can be unassailable. Indeed, the DMA proposal states,

A particular subset of rules should apply to those providers of core platform services that are foreseen to enjoy an entrenched and durable position soon. The same specific features of core platform services make them prone to tipping: once a service provider has obtained a certain advantage over rivals or potential challengers in terms of scale or intermediation power, its position may become unassailable, and the situation may evolve to the point that it is likely to become durable and entrenched in the near future.[177]

Box 6: Unassailable Market Position—a Dynamic Approach?

Metaphorically speaking, the unavoidability assumption is a photographic snapshot of the market that helps determine which firms enjoy a powerful intermediary status.

By opposition, the unassailability assumption is more of a movie clip of the market that reveals which firms successfully resist multiple assaults from rivals without being dethroned. The time dimension of the “durable position” is illustrative of this perspective.

In that respect, the unassailability assumption can embrace a more dynamic approach than the highly static approach associated with the unavoidable trading partner's notion.

However partially true, the dynamic approach inherent to the unassailability assumption is not dynamic in the Schumpeterian sense of the analysis. Instead, the unassailability assumption's dynamic approach is shortsighted and therefore ill-suited to portray market realities. The inadequacy of this other strong assumption is, however, core to the DMA’s raison d’être.

It is the entrenched and durable position of a company that causes the assumption of unassailability to be fulfilled. What does an “entrenched and durable position” mean? And what does the “near future” suggest?

The Unassailability Assumption: The Entrenched and Durable Position

The DMA may suggest a dynamic approach to competition as illustrated with the time-dimension inherent to the unassailability assumption. Indeed, the DMA defines the entrenched and durable position in the following manner:

An entrenched and durable position in its operations or the foreseeability of achieving such a position future occurs notably where the contestability of the position of the provider of the core platform service is limited. This is likely to be the case where that provider has provided a core platform service in at least three Member States to a very high number of business users and end users during at least three years.[178]

Leaving aside the debatable elements of the definition of the entrenched and durable position, the DMA explicitly refers to the time-dimension inherent to the unassailability assumption.

The Commission considers that a market position is entrenched and durable whenever a company provides a core platform service to many (business and end) users for a period of three years. This exceedingly limited timespan to conclude that a company has become unassailable precludes a credible dynamic approach to competition policy.

How could an entrepreneur consider that the company has reached an unassailable market position only because it reaches a high level of users for three years only? The rapidly changing environment of the digital competition must, on the contrary, suggest to antitrust enforcers that market positions are often fragile. This fragility derives from network externalities. If network externalities generate the “winner takes most” phenomenon, they always generate the “loser loses all” phenomenon.

As exit and switching costs are low, platforms' crowding-out effects prove robust market features disciplining incumbents and rivals. Because of the winner-take-all phenomenon applicable to rivals, the incumbent may suddenly face a surge of exit. In the digital economy, the surge of entry comes with a surge of exit, as network externalities work both ways, contrary to the DMA’s view.

Many examples can illustrate this. For instance, the superior efficiency and ease of the video-conferencing company Zoom have crowded-out the incumbent Microsoft’s Skype.[179] Facebook’s slight change of privacy settings about WhatsApp has partially crowded out the messaging app to favor rivals such as Telegram, Signal, and other rivals.[180] Under the current unassailable market position assumption, indeed, Apple’s iOS would have been considered to be enjoying an unassailable market position in 2008 when it launched the iPhone. Two years later, Android OS has become a leading OS for smartphones.[181] The mobile OS of the past (e.g., Nokia, Blackberry, etc.) could have been considered to enjoy unassailable market positions while the Android OS may not have been considered to be constituting a credible threat.[182]

Moreover, the more consumers resort to multi-homing, the less the market positions of digital companies are deemed unassailable. The DMA assumes that gatekeepers’ customers are locked into one’s gatekeeper’s digital ecosystem. Customers cannot easily switch from one digital ecosystem to another. These beliefs and assumptions are hardly corroborated in real life.

Furthermore, gatekeepers can include companies that are not dominant in their relevant markets. Indeed, “a gatekeeper may not necessarily be dominant within the meaning of Article 102 TFEU.” Such prospect of regulating gatekeepers that are nondominant and thereby exempted from antitrust liability under Article 102 TFEU directly contradicts Motta and Peitz’s abovementioned definition, where gatekeepers are said to have an unassailable market position evidenced by “a certain advantage over rivals in terms of market share.” To assert that gatekeepers include dominant companies (thereby already subject to Article 102 TFEU) and companies that are nondominant (thereby expanding the reach of antitrust by circumventing the limits of Article 102 TFEU) demonstrates a fundamental flaw in the reasoning as discussed above.

However, the idea of unassailable market power by digital platforms appears highly questionable in the academic literature. Only Larouche and de Streel, whose work was used by the Commission ahead of the issuance of the DMA proposal, mentioned this idea.[183] This idea appears intrinsically linked with the notion of “market tipping”—because digital platforms have unassailable market power, the market has tipped in favor of (few and large) digital platforms.[184]

Furthermore, the assumption of unassailability proves all the more perplexing and damaging to competition and innovation as it encompasses situations where companies can “foreseeably” become entrenched durably. Indeed, the mere prospect, according to the regulator, that a company may provide its services to a large number of business/end users and may enjoy an entrenched position for three years, the DMA suggests that the Commission can conclude that such company is likely to become unassailable. However problematic this notion of “potential assailability” may be, the DMA includes these situations:

Together with market capitalisation, high growth rates, or decelerating growth rates read together with profitability growth, are examples of dynamic parameters that are particularly relevant to identifying such providers of core platform services that are foreseen to become entrenched […] 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. The same specific features of core platform services make them prone to tipping: once a service provider has obtained a certain advantage over rivals or potential challengers in terms of scale or intermediation power, its position may become unassailable, and the situation may evolve to the point that it is likely to become durable and entrenched in the near future.[185]

If the notion of “unassailable” market position was already nebulous and tersely justified from an innovation perspective, then the mere likelihood of an unassailable market position via competition on the merits can lead the company to be considered a gatekeeper. Such a potential gatekeeper will be subject to the DMA’s obligations. The deterrence to innovate and grow can hardly be more pervasive.

There cannot be a more powerful innovation deterrent effect for entrepreneurs to suggest that their innovation leading to first-mover advantages and leading market positions over the next three years may transform their companies into gatekeepers subject to numerous regulatory obligations.

The “entrenched and durable position” is thus enjoyed by a company for only three years to qualify as enjoying a unassailable market power as condition to be designated as gatekeeper. This notion of “entrenched and durable position” is in itself problematic.

The Entrenched and Durable Position Under Siege

Economic evidence refutes the alleged unassailability of digital players. In technology markets, incumbents are successfully challenged through indirect rather than direct entries into the market. Indeed, having compiled decades of data from platform markets, Bresnahan and Greenstein showed that

a new platform increases its likelihood of attracting enough customers (a “critical mass”) if it serves a completely new, uncontested, segment of demand. After some investment in components, a platform attains sufficient capabilities to attract a larger network of suppliers and … edge closer to contested bodies of demand. The new platform can eventually grow strong enough to move into an old platform’s market.[186]

Instead of direct rivalry, the indirect entry pattern suggests that niche creation is a viable strategy and an effective competitive constraint disciplining the incumbents and possibly out-competing them.[187] This led Nicolas Petit to consider indirect entry as a viable strategy in digital markets, noting that

a process of indirect entry characterizes the digital industry from its early days to the present. Search engines leapfrogged portals and browsers as entry points to the web. Mobile telephony cannibalized desktop computing. Social networking redefined personal and professional communications. In a way, each of today’s dominant consumer-facing platforms emerged by indirect entry.[188]

As an illustration, he also noted that “Microsoft almost systematically failed when they tried to take on platform incumbents by direct competition.”[189]

It follows that incumbents’ network externalities and technological innovation provide them with a short-term advantage, frequently resulting from an innovator’s first-mover advantage. Is such a limited advantage equivalent to an unassailable market position?

Incumbents’ alleged advantage of an unassailable market position depends on the market’s definitions. Although highly controversial, especially in digital industries, defining markets may reveal niche markets wherein (short-term) indirect entry can occur. To illustrate, let us assume that, according to the DMA’s perspective, Facebook has an unassailable market position on the social media platform market. To define the relevant market as the social media platform market is highly contemptuous. Not only the relevant market of Facebook is the advertising (attention) market, but also, assuming that the social networking platform market is the relevant market, such market is made of a wide variety of platforms which exert considerable competitive constraints on Facebook.

In reality, social networking for expressing personal opinions is said to be dominated by Twitter. Snapchat leads photo social networking, while social networking for younger users is the dominion of TikTok. LinkedIn is the go-to professional social networking platform, YouTube rules video social networking, and WeChat is China’s primary social networking platform. Pinterest dominates the content-discovery social media platform.[190]

Market positions of large digital platforms are consistently challenged through severe and credible risks by the one unique feature that ensures effective competition: innovation. Indeed, Joseph Schumpeter wrote that “innovation offers the carrot of spectacular reward or the stick of destitution.”[191] Today, we often see, admire, or equally feel simmering hostility toward the digital gatekeepers’ “spectacular rewards.”

But antitrust enforcers rarely take into account the extent to which digital gatekeepers have faced—and continue to face—phenomenal failures. Nor do antitrust enforcers assess the extent to which these digital gatekeepers have disrupted laggard innovators for the benefit of both social innovation and consumers. The “sticks of destitution” lashed—and continue to lash—out. The digital gatekeepers frequently come under this natural, and commendable, innovation lash. Thus, there is no need to fuel a populist techlash, however politically rewarding it may be.

Unfortunately, when addressing the issue of digital gatekeepers, the DMA’s proposal overlooks the dynamics of innovation, for example, notably disregarding Amara’s law. American scientist and former president of the Institute for the Future, Roy Amara, stated that “we tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.”[192] Applied to digital gatekeepers, Amara’s law suggests that a dynamic competition perspective would lead us to consider that technological innovation by digital start-ups can disrupt digital incumbents irrespective of their innovations.

Indeed, a recent notorious example illustrates this. During the COVID-19 pandemic, the previously largely unnoticed company Zoom suddenly “dominated” the market for videoconferencing. Successfully erupting due to the health crisis, it continues to out-compete the digital incumbents, despite their subsequent innovations. Indeed, Google Meet and Microsoft Teams only partially (and mostly ineffectively) attempted to gain some of Zoom’s user base. Zoom’s success, derived from a small innovation to a technology first created by incumbent Skype, ushered in long-term technological disruption and social upheaval in the way people socially interact with one another online. Zoom dislodged the digital incumbent, Microsoft’s Skype, not with direct entry with imitation but rather through a small yet significant innovative feature—namely, the ability to connect through a weblink instead of complex sign-ins and expensive phone calls.

Does the DMA’s proposal integrate these disruptive and noticeable features of the digital competition? It is dubious that the DMA considers the dynamic competition inherent to the digital economy.[193] Indeed, discarding the essential notion of “indirect entry,” the DMA assumes that entrenched positions are conducive to unassailable market positions. These unassailable market positions allegedly call for remedying the so-called “structural risks for competition.”[194] In a long-overdue SCP, the Commission thus denies rivals’ innovations any credible ability to effectively compete over the incumbents’ market positions.

The pitfalls of the notion of market positions being “unassailable” pare down to the very implications this notion suggests: Market positions are alleged to be unassailable because they epitomize bottlenecks of market power. This controversial notion itself refers to the idea that the market actors hold essential assets that can be neither replicated nor obtained through rivals’ legitimate efforts. These indispensable and yet unobtainable essential assets require the application of the essential facilities doctrine.

Therefore, to assert that Amazon, Google, Apple, or Facebook, like many others (such as Microsoft,, Uber, and Netflix), have unassailable market powers is equivalent to a call for applying the inappropriate essential facilities doctrine to the digital sector.[195] Because these market positions are allegedly unassailable, structural remedies are inevitable. Either a breakup of companies or a public utility-style structural regulation can be the only solution to an unidentified problem. These regulatory calls prove incredibly popular in Europe, where politicians resort to “digital sovereignty” as a convenient means to endearing themselves in the eyes of resentful voters regarding the U.S. and Chinese tech behemoths.[196]

The allegedly unassailable market power of gatekeepers involves their control of essential assets, which is anticipated to be sustained “at least in the medium term” the Commission argued:

Gatekeepers control access to a number of customers and/or to a given input/service such as data, which—at least in the medium term—cannot be reached otherwise. Typically, customers of gatekeepers cannot switch easily and therefore only use the gatekeeper’s offering (‘single-homing’).[197]

The European Commission’s perception of the digital gatekeeper implies that control over essential data is expected to be exerted over the medium term. It discards not only short-term analysis (obviously flawed), but also long-term analysis (albeit necessary). But exactly what does “the medium term” refer to? Medium term means “three years,” as referred to in the section of the DMA that defines the entrenched and durable position. This relatively short period of time is notoriously the same time span the Commission accepts for forecasting market dynamics.[198] The European Commission adopts a relatively static approach consisting of an unassailable market position that cannot be effectively challenged in the medium term.

To illustrate, let us assume that Amazon enjoys more than 30 percent of e-commerce market and has amassed essential data. The company is considered to be a gatekeeper because no rival could, within three years, dislodge the online platform and enjoy its 30 percent market share. Irrespective of analyses of the competition on the merits and of the incumbent’s innovativeness, an artificial drive to displace Amazon with either another company or many others is equivalent to artificial selection by enforcers at the expense of consumer choice. Indeed, because antitrust enforcers may come to think that the market would not replace Amazon with one or multiple rivals enjoying that 30 percent of e-commerce market share within three years, they may intervene under the DMA, imposing public utility-style regulations. Structural separations may ultimately be imposed, regardless of innovation deterrence or consumer harm. Should such a static approach, which justifies intervention whenever an online platform is expected to enjoy unassailable market power within the next three years, be commended?

The approach would sanction competition on the merits based on relatively short-term analysis that existing or potential rivals could not dislodge an incumbent and similarly enjoy its market position. Whereas the expected benefits of such regulatory reshuffling of market actors are most questionable, the consumer costs and innovation deterrence are indubitable.

Thus, it appears that both the unavoidability and unassailability assumptions of the DMA’s Article 3(1) lack the legal clarity and economic robustness for these assumptions to legitimately underpin the DMA.

Digital Gatekeepers—the Quantitative Criteria

Given the legal vagueness and the economic weaknesses of the qualitative criteria, the Commission allegedly noted these pitfalls and attempted to “objectivize” the subjective qualitative criteria with criteria that is more quantitative. Article 3(2) of the DMA provides for quantitative criteria to designate companies as gatekeepers. However, these quantitative criteria fail to provide the necessary legal clarity and economic soundness the qualitative criteria so desperately need. Moreover, the quantitative criteria reinforce the precautionary approach inherent to the DMA with an unequal level playing field depending on the presumed categorization of the companies.

The Failed Objectivization of the Designation of Gatekeepers

The quantitative criteria used to designate gatekeepers are listed in Article 3(2) of the DMA. In a European Commission Impact Assessment Report, most respondents advocated for a mixture of both qualitative and quantitative criteria—although several respondents advocated for qualitative criteria only.[199] Respondents justified quantitative criteria based on the need for greater clarification with respect to qualitative criteria in designating gatekeepers. Unfortunately, however, the quantitative criteria do not provide the expected clarification. Per Article 3(2) of the DMA:

A provider of core platform services shall be presumed to satisfy:

  1. the requirement in paragraph 1 point (a) where the undertaking to which it belongs achieves an annual EEA [European Economic Area] turnover equal to or above EUR 6.5 billion in the last three financial years, or where the average market capitalisation or the equivalent fair market value of the undertaking to which it belongs amounted to at least EUR 65 billion in the last financial year, and it provides a core platform service in at least three Member States;
  2. the requirement in paragraph 1 point (b) where it provides a core platform service that has more than 45 million monthly active end users established or located in the Union and more than 10,000 yearly active business users established in the Union in the last financial year;
  3. for the purpose of the first subparagraph, monthly active end users shall refer to the average number of monthly active end users throughout the largest part of the last financial year;
  4. the requirement in paragraph 1 point (c) where the thresholds in point (b) were met in each of the last three financial years.

The three quantitative criteria correspond to the three qualitative criteria of Article 3(1), respectively. Here is what a consolidated version might look like:

  • A gatekeeper is said to have “a significant impact on the internal market” whenever “the undertaking to which it belongs achieves an annual EEA turnover equal to or above EUR 6.5 billion in the last three financial years, or where the average market capitalization or the equivalent fair market value of the undertaking to which it belongs amounted to at least EUR 65 billion in the last financial year, and it provides a core platform service in at least three Member States.”
  • A gatekeeper is said to be “an important gateway” in the sense that it becomes an unavoidable intermediary whenever the company “has more than 45 million monthly active end users established or located in the Union and more than 10,000 yearly active business users established in the Union in the last financial year.”
  • A gatekeeper is said to have become unassailable once it “enjoys an entrenched and durable position … in each of the last three financial years.”

Static Criteria at the Expense of a More Dynamic Approach

The Impact Assessment of the DMA considers that quantitative thresholds are to be “constructed from indicators for size … and for economic dependency” as well as by “measures of persistence.”[200] Each of these indicators corresponds to the thresholds listed in Article 3(2), respectively.

The Impact Assessment identified three policy options available to the European Commission:

Option 1: a “non-dynamic option” wherein gatekeepers are designated via quantitative criteria only

Option 2: a “semi-flexible option” that designates gatekeepers using a combination of qualitative and quantitative criteria

Option 3: a “fully flexible option” that allows gatekeepers to be designated “following a market investigation” on qualitative criteria only.[201]

The Commission ultimately chose Option 2, thereby explicitly discarding the dynamic approach. In this intermediate option, the Commission has again cut down the middle in terms of thresholds.

Box 7: The DMA’s Intermediate Threshold

Interestingly, the Impact Assessment considers that either of the “two meaningful thresholds” would be a “low threshold” or a “high threshold”:

Low threshold: Fixing the EEA annual group turnover threshold at the level of EUR 6.5–7.5 billion and the required number of core platform services showing dependencies at 30–45 million end users and 10,000 business users during several years at a single one. This threshold would result in 10 to 15 providers of core platform services;

High threshold: Fixing the threshold at an EEA annual group turnover of EUR 5–6 billion but also including a minimum of two core platform services, with at least one showing dependencies at 30–45 million end users and 10,000 business users would reduce the group of providers captured to an estimated number of five to seven companies.”[202]

It thus appears that Article 3(2) of the DMA chooses none of these “two meaningful thresholds” and instead opts for a middle way. It does not require “a minimum of two core platform services” of the “high threshold” (only one would thus suffice). Still, it applies figures from the “low threshold” for turnover (i.e., EUR 6.5 billion annual turnovers).

What does this suggest? It may mean that the Commission wants to target, with the DMA, between 7 and 10 companies designated as gatekeepers.

Again, the Commission tries to address the pitfalls of over-inclusiveness and under-inclusiveness by choosing a middle ground that avoids the critique that these thresholds are both too broad, because some gatekeepers may not be considered to be enjoying an unassailable market position and it may deter medium-sized companies from scaling up in order to circumvent the stringent regulation of the DMA, and too narrow, as it identifies exclusively non-European companies and deteriorates their market conditions in favor of their rivals. Gatekeepers’ rivals may free ride on the DMA’s newly created opportunities for litigation and gain mandatory access to a range of assets and data.

The Commission chose the thresholds that overlook a dynamic approach to the designation of gatekeepers following market investigations and fact-finding endeavors rather than by some ex ante (qualitative, quantitative, or both) criteria. The neglect of such a dynamic approach is hardly justifiable. Indeed, commenting on the European Commission’s Impact Assessment of the DMA, the European Scrutiny Board delivered an opinion that cast doubt on the relevance of the quantitative criteria specifically:

The report should better define and justify the measures covered under the options. It should demonstrate why the proposed set of cumulative quantitative thresholds (under the ‘non-dynamic’ and ‘semi-flexible’ options) can be considered as a robust and reliable trigger across all selected core platform services for the (quasi-automatic) designation of gatekeepers and the imposition of obligations. It should better explain why a market investigation is not deemed necessary or proportionate in these situations.[203]

It thus appears that the designation of gatekeepers, albeit an error-prone exercise, would be better achieved using economic, evidence-based analysis inherent to market investigations rather than a static, non-flexible approach involving ex ante contestable criteria.

Quantitative Presumptions, Not Criteria

The benefits of quantitative criteria pertain to the alleged objectivization such criteria were deemed to achieve. This objectivization has failed in light of the DMA proposal.

One would think that the quantitative criteria aimed at clarifying the qualitative criteria would be clear-cut and help companies, entrepreneurs, and enforcers designate gatekeepers. On careful analysis of the DMA proposal, it appears that these quantitative criteria are, in fact, not “criteria” at all, but rather “presumptions.” Indeed, Article 3(2) states, “A provider of core platform services shall be presumed to satisfy” before unfolding the quantitative criteria.

Article 3(2) lays down “indicators” aimed at circumventing the qualitative criteria of Article 3(1).[204] According to the DMA, the companies that fulfill these indicators are presumed to be gatekeepers. Therefore, contrary to the announcement suggesting a mixture of qualitative and quantitative criteria, the quantitative presumptions are outlined so as to not require qualitative criteria.

Article 3 does not lay down two series of criteria (qualitative and quantitative), but rather one series of qualitative criteria and a set of quantitative presumptions.

After a company fulfills the criteria of Article 3(2), it must notify the European Commission. But even without such notification, the Commission can systematically designate any company that meets the quantitative indicators.[205] The importance of the notification system encapsulated in the DMA is thus limited. Regardless of whether or not the company notifies the European Commission, the Commission’s self-granted powers will enable it to designate the company as a gatekeeper.

The presumptions in Article 3(2) with the qualitative indicators are extremely strong and highly questionable. The first indicator presumes that a company has “a significant impact on the internal market” whenever “the undertaking to which it belongs achieves an annual EEA turnover equal to or above EUR 6.5 billion in the last three financial years, or where the average market capitalization or the equivalent fair market value of the undertaking to which it belongs amounted to at least EUR 6.5 billion in the last financial year, and it provides a core platform service in at least three Member States.” Size as the parameter is the embodiment of the “big is bad” motto inherent to the populist techlash expressed recently against disruptive and innovation companies. More specifically, using the turnover to assess the “bigness” appears fatally unwise.

Indeed, the turnover refers to the amount of business done by an enterprise in a given year. It shows the speeds at which both payments are received from debtors and inventory is sold (inventory turnover). As the Impact Assessment itself recognizes, “Turnover is an indicator of the number of transactions intermediated by a given platform.”[206] Therefore, to place a company under a gatekeeper status based on its volume, irrespective of the profitability derived from these transactions, amounts to a dramatic increase in transaction costs despite no evidence of anticompetitive conduct.

Turnover reflects efficiency. The more a company replaces its assets with new ones (due to innovation, change of strategies, adaptation to a competitive environment), the more the turnover increases. Therefore, turnover is the exact opposite of a lethargic company resting on its laurels. By opposition, revenue as a measurement of size reflects the profitability (i.e., market power) of a company instead of turnover reflecting internal efficiency. If the Commission were genuinely concerned with the profitability of firms, it would have chosen revenue rather than turnover as measurement criteria of size.

To categorize companies according to turnover is tantamount to sanctioning their efficiencies, with the most agile and disruptive companies falling under the DMA. The European Commission thus undermines the once-influential “as efficient competitor” test within EU competition law.

Turnover is a fundamentally inappropriate criterion to assess market power. And yet, the Commission relies on a calculus of turnover.[207] Indeed, a 2018 Impact Assessment for a accompanying an earlier regulatory proposal notes,

It is assumed that the negative impacts of ranking lead to a loss in yearly turnover of between 1% and 2%, most of which is permanent due to the difficulty in redirecting sales to other channels. These assumptions have been applied to the total turnover in the different sectors considered but exclude the issue of ranking in online general search engines for which insufficient evidence is available at present to allow a robust quantification of any systematic negative impacts. On that basis, the reduction of sales through platforms for EU business users caused by the practices at stake can be estimated to amount to between €1.27 and €2.35 billion per year.[208]

To infer from inferior ranking an anticompetitive conduct without properly assessing the reasons underpinning such an inferior ranking (e.g., quality, innovation, consumer preferences, etc.) constitutes a flawed conclusion. Not only is a fall in turnover due to a negative ranking possibly fully justified (e.g., Are poorly ranked restaurants complaining against well-ranked restaurants?), but the demotion of some companies, by definition, also suggests the promotion of other companies in these rankings.

Therefore, in a worst-case scenario, the total net welfare would be equal to before. In the most probable of the best-case scenarios, it increases due to the selection of products based on the innovation of consumer preferences. The turnover criterion is flawed, and used in a defective manner under the DMA proposal.

Would the many European companies with tens of billions of euros of annual turnover together be categorized as a “large platform”? For instance, would the French supermarket Auchan and its $61 billion annual turnover qualify as a gatekeeper for equal footing with its competitor Amazon? Would Finnish Nokia or Swedish Ericsson, with their annual turnovers of $53 billion and $34 billion, respectively, be designated as gatekeepers so as to be placed under an equivalent regulatory framework as their American or Chinese rivals?

European companies started to realize the breadth of the DMA’s scope and thus express concerns about the vagueness of the notion of “gatekeeper” and the intrusiveness of the obligations applicable to gatekeepers.

As one of the many illustrations of expressed concerns, the French association of large companies, responding to the DMA public consultation, stated,

The options considered by the Commission are worrying because they target non-dominant companies holding power in a market without this concept being defined or in related markets without any abuse, such as oligopolistic market structures which are neither new nor objectionable per se; and do not clearly define the target market as digital or non-digital, or European or global. Potentially, any economic actor in a non-abusive dominant position could be within the scope of this tool, based on the analysis of market structures. Its intrusive nature, based on an analysis which requires that many complex interactions are taken into account, is likely to call into question the building of real industrial strategies.[209]

Not only is size an inappropriate criterion to assess the competitive constraints in a market and the possible anticompetitive practices, but turnover is a misguided standard that would encompass too many companies.

The other indicators of the first element of Article 3(2) are equally dubious. Indeed, the market capitalization of €65 billion in the last fiscal year may affect such companies as the big European pharmaceutical firms, the big European energy companies, the big European car manufacturers, etc. Of course, none of them provide core platform services and thus would be exempt from the DMA’s obligations.

Finally, the criterion that gatekeepers must meet their turnover threshold in at least three member states carries enormous unintended consequences, as it would induce companies to remain within their own national markets and avoid entering into a third member state’s market when non-European companies can be entered without being subject to the DMA.

For example, suppose a large core platform service provider dominates the German and French markets—already a generous portion of the whole European market. The company decides not to enter a third member state’s market, say, Luxembourg’s, and instead enters the U.S. or Turkish market, thereby avoiding the obligations of the DMA. Of course, the “circumvention risks” of the DMA’s obligations by potential gatekeepers are essential.[210] The Commission attempted to anticipate such risks with Article 3(6) of the DMA. In order to avoid circumvention, Article 3(2) indicators are irrelevant, as the Commission would not use these presumptions in order to designate a broader range of companies as gatekeepers. This reveals the looming arbitrariness of the quantitative criteria.

The Arbitrariness of the Designation of Gatekeepers

The quantitative indicators do not clarify the qualitative indicators, but rather merely indicate that the qualitative ones are met whenever the quantitative indicators are. Still, the assumption involving quantitative indicators does not prove to be of real importance.

Indeed, a presumption generally aims at speeding up the reasoning process with shortcuts. A legal presumption thus provides quicker ways to reach a legal conclusion. The presumptions of Article 3(2) should enable regulators to conclude the designation of gatekeepers more quickly. However, the goal of Article 3(2)’s presumption is not achieved.

Article 3(6) undermines the relevance of Article 3(2) because of the European Commission’s fear of under-inclusiveness. A company that exhibits the characteristics of a gatekeeper under the qualitative indicators of Article 3(1), but does not fulfill the presumption of the Article 3(2)’s quantitative indicators, may nevertheless be considered a gatekeeper at the discretion of the European Commission. Article 3(6) reads:

The Commission may identify as a gatekeeper, in accordance with the procedure laid down in Article 15, any provider of core platform services that meets each of the requirements of paragraph 1 but does not satisfy each of the thresholds of paragraph 2 or has presented sufficiently substantiated arguments in accordance with paragraph 4.[211]

Consequently, reverting to qualitative criteria only, the Commission will be able to assess whether a large platform has an unavoidable intermediary power and enjoys an unassailable market position. Without further market investigation suggested in policy Option 3 in the Impact Assessment, the ex ante designation of gatekeepers based on qualitative criteria only corresponds to none of the policy options identified by the Impact Assessment. Designating gatekeepers exclusively on qualitative criteria is indeed not advised in the Impact Assessment Report precisely because the discretionary power inherent to such a subjective analysis creates a scope for arbitrariness.

The quantitative presumptions of Article 3(2) are pointless given the European Commission’s ability to ignore them under Article 3(6). The quantitative presumptions do not provide the legal clarity they were intended to provide.

The qualitative criteria of Article 3(6) of the DMA further reinforce the risks of arbitrariness:

For that purpose, the Commission shall take into account the following elements:

  1. the size, including turnover and market capitalisation, operations and position of the provider of core platform services;
  2. the number of business users depending on the core platform service to reach end users and the number of end users;
  3. entry barriers derived from network effects and data driven advantages, in particular in relation to the provider’s access to and collection of personal and non-personal data or analytics capabilities;
  4. scale and scope effects the provider benefits from, including with regard to data;
  5. business user or end-user lock-in;
  6. other structural market characteristics.

In conducting its assessment, the Commission shall take into account foreseeable developments of these elements.[212]

Besides the first two elements, which broadly repeat Article 3(2)’s quantitative indicators, the elements from (c) to (f) enumerate fundamental features of the two-sided platforms. These elements may not exemplify market power, but rather, illustrate the natural phenomenon of “winner takes most” inherent to network externalities of the digital economy. Point (f) (“other structural market characteristics”) explicitly refers to a more structuralist approach wherein the preservation of the structure of the markets (i.e., a sufficient number of players in the market) shapes the European Commission’s analysis of the competitive environment in digital industries. The structuralist approach epitomizes the precautionary approach toward innovation discussed ahead.[213]

Despite a company not qualifying as a gatekeeper, the mere “foreseeability” that it may eventually become a gatekeeper can lead to its designation as a gatekeeper, according to Article 15(4) of the DMA. This amounts to a prediction of commercial success for medium-sized companies.

The unpredictability of the designation of the gatekeepers reaches its climax with the last paragraph of Article 3(6) of the DMA. In case a company does not satisfy the quantitative indicators of Article 3(2) and fails to comply with the Commission’s measures, the Commission may unilaterally decide to designate that company as a gatekeeper:

Where the provider of a core platform service that does not satisfy the quantitative thresholds of paragraph 2 fails to comply with the investigative measures ordered by the Commission in a significant manner and the failure persists after the provider has been invited to comply within a reasonable time-limit and to submit observations, the Commission shall be entitled to designate that provider as a gatekeeper based on facts available.[214]

Failure to comply with the Commission’s investigative measures should lead to penalties and fines, as it is commonly accepted and applied under EU law not to reclassify a company as gatekeeper without full-fledged analysis.[215]

Under the DMA, the designation of gatekeepers resembles a discretionary power that may evolve toward a Commission’s arbitrary decision. Indeed, the Commission can ignore the qualitative indicators of Article 3(1) whenever the presumptive indicators of Article 3(2) are met. Even if they are not met, a company can still be designated as a gatekeeper, according to Article 3(6), based on some elements of the markets or on a failure to comply with the Commission’s investigative measures.

In other words, a company suspected by the Commission to qualify as gatekeeper can hardly escape the wrath of the Commission and its ability, due to self-granted regulatory powers, to designate the company accordingly.

Digital Gatekeepers—an Assessment

The creation of a category of gatekeeper runs the risks of creating both threshold effects (e.g., uneven level playing field for competition) and entrenchment effects (e.g., gatekeepers’ market positions reinforced by their ability to cope up with obligations). The category, it has been demonstrated, is flawed from both a legal and an economic viewpoint.

Nevertheless, can the questionable gatekeeper status, once ascribed to a company, be removed? Not really, according to Article 4 of the DMA, which suggests that the Commission reviews gatekeepers’ status every two years, and may consider amending or repealing a gatekeeper’s status only if there has been “a substantial change” of the situation or if the initial decision was incorrect or incomplete.[216] The probability of such an advent materializing is relatively low. Not only would the Commission hardly consider its previous decision as being erroneous, but the substantial change required may also prove to be too high a threshold for companies to bring a convincing case.

More generally, the concept of gatekeeper proves unfathomable.

Box 8: From Digital Gatekeepers to Digital Concierges?

The Commission defines large platforms as digital gatekeepers.[217] The idea of digital gatekeeping suggests that gatekeepers are in essence the “concierges of the Internet.”[218] They observe traffic on the Internet and amass information to distribute and accompany users through the relevant places.

Because the concierge “is flexible, disciplined and persistent and is kind and courteous to all customers and staff alike ... Any request will be addressed,” so the use of the metaphor of digital concierges sheds light on the range of services conducted by concierges that go well beyond the gatekeeping service.[219] Indeed, concierge services suggest that digital platforms do not only keep a digital gate but more fundamentally perform services that are designed for customer care. The digital concierge operates in a designated digital ecosystem to deliver relevant services to customer queries, akin to concierge doctors or hotel concierges.[220] Ad-funded digital concierge services are triggered whenever consumers place a query.

As the metaphor of digital gatekeepers as digital concierges implies, while they do provide some services, they do not have unassailable market power.[221] Akin to concierges, digital gatekeepers are considered disposable intermediaries. Despite the assumption that Facebook, Apple, and Google are the gatekeepers of online speech, smartphones, and search engines, respectively, they could each be replaced by Twitter, Google’s Android OS, and other search engines/Internet entry portals.

“The gatekeepers are vulnerable though they pretend to stay in control,” Henry Stevens has argued in the area of health care.[222] In the digital sector, gatekeepers pretend to stay in control while actually being vulnerable—from technical vulnerabilities to competitive vulnerabilities to reputational vulnerabilities.[223]

Can digital gatekeepers be said to “act as private regulators setting the rules of the game on the markets they control,” as the Commission considers?[224] The very idea of private companies acting as private regulators is disingenuous since every company’s terms and conditions, internal processes, and corporate rules adopted to conduct business with trading partners can each be said to be its own “private regulation.”

A company’s corporate governance rules, business strategy, and conditions for dealing with third parties are equivalent to privacy regulations, according to the Commission’s language. In that regard, is Google Android a digital gatekeeper over its open-sourced OS when it deals with third-party app developers? Is Apple a private regulator of its App Store?

Yes, they are both digital gatekeepers, if we infer the gatekeeping role is based on a their ability to define the rules concerning third-party use of their core products.[225] This alarmingly negative portrayal may very well swift to a more balanced assessment that Google Android’s inherent forks enable third-party app developers to freely make the best use of the OS, while Apple’s closed iOS ensures the highest quality and reliability of its digital ecosystem, wherein Apple selects the app developers.[226]

In both cases, the digital concierge services inherently provide different yet valuable information in guiding third-party developers toward optimal use of the digital ecosystems. For Android, the freely accessible OS is conditioned to contractual restrictions (i.e., forks) designed to ensure the services’ economic viability and the compatibility of third-party services therein. For Apple iOS, the gatekeeping services resemble porter services in that entry is made on a selective basis. It both prevents unwarranted entry and incentivizes quality ameliorations for the sake of more excellent safety and system integrity.

Concierge services are provided by both platforms, although Google Android OS mimics full concierge services by contractual restrictions. To further the metaphor, if Google Android OS is a hotel concierge standing at the front desk, Apple iOS is a night-club bouncer standing at the door.

To blur the distinction between the different ranges of concierge services and their fundamental impacts on the overall digital ecosystem thus proves to be a fundamental pitfall of the notion of digital gatekeeper—let alone the contestable qualitative and quantitative criteria this concept rests upon.

To conclude, the notion of “gatekeeper” is legally vague, is economically detrimental to competition and innovation, and misses its essential objectives—namely, to quicken regulatory compliance and avoid apparently useless discussions between companies and the regulator.

The notion of “digital gatekeepers” is not only detrimental to the economy because it discourages large companies from both innovating and competing fairly with rivals, but it also sets incredibly powerful threshold effects that will deter medium-sized companies to scale up.

Companies will ask judges to nullify individual decisions that designate them as gatekeepers. Companies will challenge the obligations imposed to them on the basis that the gatekeeper status they carry is unsubstantiated. Companies will regularly seek to exit the category of gatekeepers and will challenge any decision not to do so.

Consequently, against its regulatory rationale of avoiding lawsuits and intervening by regulation only, the DMA will likely trigger a wealth of lawsuits that may last years before finally decided. Advocates of the DMA disparage the role of the courts and the fundamental function of judges in the adjudication of competition law. Thus, because it so highly detrimental, the gatekeeper status represents one of the most fundamental challenges to a thriving, competitive, and dynamic European innovation economy.

The Gatekeepers’ Core Obligations—Article 5

The DMA identifies several ex ante prohibitions for gatekeepers. Article 5 identifies seven practices that are preemptively considered anticompetitive. This section reviews these practices and the extent to which they are prohibited, both at the expense of innovation incentives and despite their benefits to consumers.

According to the Commission, “[T]he 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.”[227] Each of these three elements raises considerable doubts.

First, the identified obligations may not redress unfairness and harm, but may exacerbate unfair competition and consumer harm. Second, the core obligations of Article 5 (and the ancillary obligations of Article 6) do not provide unambiguous clarity, as scholars and market participants have widely acknowledged. Finally, to assert that these obligations have been elaborated based on sufficient experience raises considerable concerns about the legitimacy of such experience.

Indeed, Margrethe Vestager rebuffed allegations regarding a potential conflict of interest between her mandates as vice president in charge of the “Europe Fit for the Digital Age” agenda and as competition commissioner.[228] It was, however, already clear that digital regulations would be designed from past antitrust cases.[229] Vestager left the members of the European Parliament “unimpressed by her attempts to reassure about a potential conflict of interest.”[230] The proposed reforms of the Digital Single Market were not to be based on the work of DG-Comp or past or current antitrust cases.

Vestager once promised to build “Chinese walls” between competition (past and current) enforcement and digital regulation, promising that “it will not be my pen that will draw” the digital regulations.[231] These walls were crushed down for the DMA included in the Digital Services Package.[232]

Indeed, Vestager officially presented the DMA as being construed from sufficient experience from past and ongoing antitrust cases, contrary to the promises she made during parliamentary hearings.

Past antitrust cases such as Google’s, Facebook’s, and ongoing investigations such as Amazon’s were unfairly used to draw Article 5’s obligations. The once-derided conflict of interest between competition cases and experience and digital regulations now appears legitimate. In other words, the conflict of interest has not been avoided, but rather, the direct use of antitrust information from past and current antitrust cases helped elaborate the DMA.

As none of the few cases relevant to the DMA have yet to produce even a single a judicial ruling, Article 5’s obligations were thus inferred from insufficient experience and undue information.

Article 5’s obligations increase transaction costs in digital markets despite calls for an innovation economy, deter innovation despite Europe’s weakness on digital innovation leaders, and neatly embody precautionary over innovation-based antitrust.

Chapter III is the core part of the DMA. It outlines both the obligations imposed on gatekeepers and the relevant procedures to ensure proper compliance with those obligations, with Article 5 enshrining the fundamental obligations created by the DMA to gatekeepers. The seven prohibited practices are far-reaching and widely defined, paving the way for appreciable discretionary power by the European Commission. This section critically assesses each of these prohibitions successively.

Leveraging, Envelopment, and Bundling

A gatekeeper shall

refrain from combining personal data sourced from these core platform services with personal data from any other services offered by the gatekeeper or with personal data from third-party services, and from signing in end users to other services of the gatekeeper to combine personal data, unless the end user has been presented with the specific choice and provided consent in the sense of Regulation (E.U.) 2016/679.[233]

The first category of prohibited practices consists of combining data of one service with that of another service (be it provided by the gatekeeper or a third party). Referred to as “platform envelopment,” it’s when a gatekeeper (the enveloper) enters an adjacent market by using the data accumulated in the primary market.[234] Enveloping is one of the leveraging strategies that enable market actors to enter markets based on their experience and reputation. Envelopment strategies, developed in a conglomerate fashion, contribute to lower entry barriers wherever network effects and switching costs may insulate a dominant firm in a market from being outcompeted.

Examples include Google Chrome outcompeting Firefox Explorer thanks to Google search engine’s user experience, and Facebook Marketplace entering secondhand platform markets via its Facebook user experience. Due to incumbents’ ability to successfully leverage their dominance from one market to another, the effects of envelopment strategies are assimilated to foreclosure effects.[235]

The DMA appears to disregard the many pro-competitive effects of envelopment strategies that counterbalance these alleged anticompetitive effects in order to command gatekeepers to refrain from engaging in envelopment strategies at the expense of both new entries into some markets and the lowering of prices. If the DMA prohibits platform envelopment strategies, then business developments that brought about disruptive innovations may no longer be possible. Instances of platform envelopment include Uber Eats by Uber, Google Chrome by Google, Amazon Prime by Amazon Fire OS, and LinkedIn Job Listings by LinkedIn.[236]

Furthermore, the DMA prohibits the envelopment strategy not only for gatekeepers’ core services but also for third-party services.[237] This reflects the so-called “conflict of interest” that has recently appeared as an antitrust concern. When a platform offers third-party sellers the ability to sell on that platform, but then subsequently sells third parties’ related products directly to end users, the platform acts as both a platform and a rival retailer. Pro-competitive benefits such as lower prices, increased rivalry, higher quality, or loyalty services (e.g., service delivery, feedback, etc.) abound for these practices.

Supermarkets historically resorted to these practices whenever private-label (or store-brand) products competed with brand-name ones.[238] While their overall quality has also increased, private-label products are popular mainly because they are cheaper, thus making them particularly sought-after in times of economic turmoil.[239]

Nevertheless, the additional role of retailer places the platform under scrutiny, with regulators rejecting private-label products, notwithstanding the consumer benefits they generate in terms of purchasing power and increased competitiveness. When offered by digital gatekeepers, private-label products allegedly result from the unfair dual role played out by the platform wherein data is amassed in order to be used as a retailer. At the launch of its investigation of Amazon and its dual role as platform and retailer, the Commission claimed that this practice was a violation of Article 102 TFEU.[240] It detailed that “when providing a marketplace for independent sellers, Amazon continuously collects data about the activity on its platform. Based on the Commission’s preliminary fact-finding, Amazon appears to use competitively sensitive information—about marketplace sellers, their products and transactions on the marketplace.”[241]

A Statement of Objections on November 10, 2020, confirmed the opening of this investigation. Commissioner Vestager argued that “we must ensure that dual role platforms with market power, such as Amazon, do not distort competition. The date on the activity of third-party sellers should not be used to the benefit of Amazon when it acts as a competitor to these sellers.”[242] Similar to a supermarket manufacturing its own products to compete with brand-name products, Amazon’s platform sometimes offers its own private-label products when those products are popular, which gives the company a competitive edge (such as lower prices, higher quality, distinctive features, etc.).

These practices amount to anticompetitive conduct only if they are deemed to be anticompetitive under Article 102 TFEU—and the ongoing investigation may lead to these conducts being stopped. The DMA has no usefulness in that respect, since Article 102 TFEU can achieve what the DMA seeks to achieve.

The last conduct prohibited in this first category relates to “signing in end users into other services of the gatekeeper to combine personal data.” Single sign-on (SSO) has been “widely adopted throughout the years, in especial to solve the complex problem of credentials management.”[243] Sometimes referred to as sign-ins or sign-ups, SSOs are user- or session-authentication processes that enable a user to enter a name and password to access multiple applications.[244] Users do not have to retype their login details and passwords when, for example, using Facebook Messenger from the Facebook platform, Gmail from the Google search engine, or iTunes from Apple. More generally, SSOs are authentication processes that “can be conducted using several distinct types of credentials, such as something that the subject knows (e.g., a password), possess (e.g., a smart cart), is (e.g., static biometrics), does (e.g., dynamic biometrics), or some other verifiable property (e.g., the subject’s location).”[245] The obvious reason behind such SSOs is users lose or forget their passwords, hence SSOs correspond to consumer demand and preferences.

Giving users the ability to discover additional (free) services provides platforms with an innovation incentive to develop a digital ecosystem wherein applications, services, and software are interoperable and easily accessible. Some digital platforms have pushed the innovation behind SSOs to create an ad hoc app dedicated to easy and secure authentications.[246] Irrespective of the obvious time-saving reasons underpinning these features for users, and the legitimate data interoperability across services these practices can yield, the DMA bans these practices.

These practices identified in the first category are all prohibited “unless the end user has been present with the specific choice and provided consent in the sense of Regulation (E.U.) 2016/679”—namely, the GDPR. In other words, practices become prohibited unless the users have consented to them in compliance with the GDPR. Password-less authentication prohibition would increase transaction costs for consumers and deter digital innovation because of reduced expected interoperability. Compliance with the GDPR suggests that the present requirement for consent appears superfluous. For instance, a Gmail user may no longer be automatically signed in to Google Maps without prior consent. Because users will presumably consent to it, such an obligation becomes trivial, and is otherwise an unnecessary increase in transaction costs.[247]

Does the requirement for user consent required by the GDPR apply only to the last practice identified (“signing in end users to other services of the gatekeeper to combine personal data”) and not to all the practices identified in the first set of prohibitions? If so, the inconsistent treatment across different data combinations by gatekeepers suggests inadequate legal refinement that can only generate legal indeterminacy. Indeed, data via either adjacent services/apps or a combination of data via SSOs should be treated under the same data protection regime—namely, the GDPR—and not under some discrete competition rules, as the DMA suggests.[248]

Suppose the qualification applied to all conducts referred to in Article 5(a). In that case, it would be dubious the extent to which these prohibitions might ever be effective. Digital gatekeepers invariably comply with the GDPR by guaranteeing user consent through ticked boxes and other approval methods compatible with Article 7 of the GDPR.[249] Thus, beyond the mere detrimental competitive effects of the GDPR and its de facto support for large digital platforms able to cope with the regulatory costs generated, it appears that the GDPR already requires all digital platform to use data in a consent-based manner.[250]

Consequently, Article 5(a) prohibitions certainly provide either a detrimental (if effective) ban on some conducts or a superfluous (if repetitive to the GDPR) ban on already illegal behaviors in the European Union. Nevertheless, digital gatekeepers will be subject to two different legal texts that are aimed at the same objective but prone to interpretative discrepancies across these texts. Legal uncertainty and innovation deterrence will inevitably increase for a handful of digital platforms. Absent any added value or rationale regarding Article 5(a), this ban wields no benefits to current practice, with only legal uncertainty and over-deterrence effects.

Prohibitions of Most Favored Customer Clauses

The second practice represents a setback for both digital platforms’ competitive environment and end consumers’ ability to have the cheapest and most qualitative products and services. Indeed, this prohibition bans most favored customer (MFC) clauses outright despite their pro-competitive benefits:

[A gatekeeper shall] allow business users to offer the same products or services to end users through third party online intermediation services at prices or conditions that are different from those offered through the online intermediation services of the gatekeeper.[251]

Such conduct entitles business users (namely, customers, app developers, third-party retailers, etc.) to prevent digital gatekeepers from including so-called MFC clauses in their contractual relationships. Derived from the most favored nation (MFN) principle in international trade, MFC clauses enable, say, a platform to ensure that third-party retailers provide, in its proprietary platform, the best offers (e.g., in terms of price, quality, contractual terms) relative to what third-party retailers offer in other online distribution channels (e.g., third-party platforms, retailers’ websites).

In other words, the platform requires third-party retailers to make available their best offers on that platform, akin to overall marketing strategies such as refund price differences.[252] MFC clauses are equivalent to “price parity” clauses (or “best price” clauses) wherein the competition is primarily based on price. The prohibition of MFC clauses derives from the Commission’s investigation of Amazon’s MFC clauses on its e-books, when, in 2017, it accepted Amazon’s commitment to no longer require publishers to offer Amazon similar (or better) terms and conditions as those offered to its competitors.[253]

Antitrust implications are ambiguous given both the regulators’ sensitivity regarding contractual restraints and pro-competitive benefits generated by the price-competition effect of MFC clauses.[254] The academic literature emphasizes the pro-competitive benefits of MFC clauses.[255] Indeed, the pros of MFC clauses predominantly outweigh their hypothetical anticompetitive effects.[256] This is easily understandable, as MFC clauses are all about lowering prices, increasing the competitive advantage of a platform (be it a supermarket or a digital player) against its rivals. Therefore, MFC clauses foster consumer welfare while saving valuable time and resources for the platform to search and monitor rivals’ lower prices. Furthermore, since the platform does not have to search for and watch its competitors’ price catalogs, they prevent collusive practices and parallel prices.[257]

Moreover, MFC clauses incentivize innovation because contractual arrangements foster long-term investments.[258] Both contracting parties can benefit from a long-term relationship.[259] It thus becomes difficult to see any outstanding anticompetitive effects of MFC clauses whenever consumer welfare, the intensity of rivalry, and prevention of collusions are fully considered. The prohibition of MFC clauses pushes the market further “from a less anticompetitive equilibrium to a more anticompetitive equilibrium.”[260] Therefore, the DMA overlooks the pro-competitive effects of MFC clauses, as acknowledged in the European decisional practice. This neglect does not tackle anticompetitive practices,[261] but rather hampers consumer welfare.[262]

The prohibition makes it more difficult for gatekeepers to ensure end users receive the cheapest and best quality products or services whenever such conditions exist on third-party platforms.[263] Indeed, consumers may continue to use gatekeeper services with higher prices and lower quality than those offered on third-party platforms because of the gatekeeper’s overall attractiveness of service portfolios.

The prohibition will harm consumers, as innovation by gatekeepers will lower due to a lack of both incentives and ability to meet the best offers. For instance, may no longer ensure that its best hotel rates are listed on its platform (thus adversely affecting its whole competitive viability), and Amazon may be precluded from requiring third-party sellers to propose their best offers on the Amazon platform. These platforms may be forced to accept third-party sellers proposing less-competitive offers on their websites as opposed to the third-party sellers’ original websites. Would consumers gain from these prohibitions? Consumers would likely either spend more time searching and comparing offers or pay more for products and services than they would if MFC clauses were enforced.

An MFC clause “engages a seller to apply to a buyer the same conditions offered (by the same seller) to other buyers.”[264] A buyer is offered a price reduction when either other buyers are paying lower prices for related products (ex post MFC clauses) or the seller contractually commits to offer the buyer those same lower prices. In this traditional model, the buyer enjoys bargaining power against the seller, and the seller tries to attract buyers with a commitment not to discriminate. But, in the framework of MFC clauses applied to digital platforms, the platform is mere intermediary and does not sell the end-products (e.g., hotel rooms in the case of

Therefore, MFC clauses do not require buyers to extract rents from the sellers, but rather require sellers (e.g., the digital platforms) to lower prices for buyers (that sell the core products). They help the digital platform to cope up with the intense rivalry. This contributes to the competitive process and benefits end consumers. Because the economic relationship (namely, the enjoyment of bargaining power) is utterly reversed and the buyers are actually sellers to end users, the conclusions derived from the MFC clauses are not accurate. The current competitive relationship between digital platforms and conventional market actors (e.g., hotels) is at odds with traditional actors.

However odd this extrapolation may prove to be from an economic perspective, Article 5(b) embodies such extrapolation with a strict prohibition. This prohibition of MFC clauses is fundamentally flawed and unfair. While companies “cannot offer secret price discounts to buyers”[265] and would be “heavily fined if they were to offer different terms of supply to different buyers,” digital gatekeepers would also be fined if they were to offer similar terms to different customers.[266] Consequently, the prohibition suffers from fundamental pitfalls.

Data Sharing Through Free Riding

The third practice enshrines entitlements for business users, thereby leading to the creation of a free-rider problem:

{A gatekeeper shall] allow business users to promote offers to end users acquired via the core platform service, and to conclude contracts with these end users regardless of whether for that purpose they use the core platform services of the gatekeeper or not, and allow end users to access and use, through the core platform services of the gatekeeper, content, subscriptions, features or other items by using the software application of a business user, where these items have been acquired by the end users from the relevant business user without using the core platform services of the gatekeeper.[267]

This practice also entitles digital companies running on gatekeeper’s platforms (e.g., app developers on app stores; apps on operating systems, etc.) access to end users without using “the core platform services of the gatekeeper.” These core platform services are defined in the explanatory part of the DMA as including:

(i) online intermediation services (incl. for example marketplaces, app stores and online intermediation services in other sectors like mobility, transport or energy) (ii) online search engines, (iii) social networking (iv) video-sharing platform services, (v) number-independent interpersonal electronic communication services, vi) operating systems, (vii) cloud services and (viii) advertising services.[268]

This obligation pertains to a digital duty to deal whereby data interoperability and an improper essential facilities doctrine apply against gatekeepers. Such an obligation applies irrespective of their investments or their proprietary rights over their digital ecosystems.[269] It is modeled after the Revised Payment Services Directive of 2015 (PSD2), which states that, absent market failures or consumer harm, real-time data access enables service interoperability.[270]

This obligation for the gatekeeper to share data with business users to help them reach end users without the need for business users to use the gatekeeper’s core platform services constitutes a setback. For instance, the obligation may detrimentally affect, say, Apple’s App Store without clear benefits.[271] Indeed, app stores are targeted with such an obligation to share data about end users with business customers, subject to no conditions. It jeopardizes these platforms’ business models because adjacent proprietary services are effective ways for these platforms to ensure the economic viability of the core platforms’ services. Under the veil of the prohibition of discrimination across both businesses that use the core platforms services and those that do not, the present obligation inherently generates a free-ridership problem.

Indeed, IP rights protect platforms’ proprietary digital ecosystem, ensuring the internalization of potentially uncompensated externalities and optimizing innovation incentives. The present obligation seems to both disregard the presence of IP rights and overlook innovation incentives altogether.[272] Business users’ enjoyment of accessing the gatekeeper’s consumer database without being subject to contractual conditions generates a free-rider problem by both forcing disclosure of proprietary consumer data and creating unconditional entitlements for business customers to any gatekeeper’s adjacent services. The gatekeeper’s inability to discriminate (let alone exclude) business users based on whether they use their core platform services represents an uncompensated externality. This is the essence of the free-riding problem that stifles innovation.

Assuming, in the vein of the Crémer Report, that data access “may need to be imposed” in order “to serve complementary markets or aftermarkets”[273], such legally enshrined free riding not only deters innovation but also increases prices to end consumers. Although tying and bundling sometimes lead to win-win agreements, the DMA prohibits it whenever it involves a gatekeeper’s services, irrespective of the practical consequences of such a detrimental ban on consumers.

As a metaphor for the off-line world, this obligation would enable tied houses (e.g., bars with distribution agreements with breweries) to be provided with financial and material investments in the pub without the bar owner being required to sell the brewer’s beers. Such free-riding entitlement would deter initial and subsequent investments, increase transaction costs (due to hold-up problems), and limit product innovation. A similar outcome may unfold in the digital markets, wherein the gatekeeper (i.e., provider of the digital ecosystem) could become obligated to accept business users that can reach the gatekeeper’s end users without using the gatekeeper’s core platform services. Additionally, such digital duty to deal disregards cybersecurity threats and the safety of the digital ecosystem. Foreseeably, the gatekeeper may limit its investments based on business customers’ ability to use and exploit the gatekeeper’s digital ecosystem.

As a result, the provision discourages innovation. Additionally, it may result in less consumer choice, lower innovation, increased transaction costs, and overall consumer harm. The law should prevent free-rider problems from arising, not generate them, as this third practice unfortunately does.[274]

Finally, this practice also encompasses the entitlement for end users to “access and use, through the core platform services of the gatekeeper, content, subscriptions, features or other items by using the software application of a business user, where the end users have acquired these items from the relevant business user without using the core platform services of the gatekeeper.”[275] This entitlement to access business users’ apps without using the gatekeeper’s core platform services rewards business users’ abovementioned free ridership as well as encourages end users to weaken the overall economic viability of the digital ecosystem built by the gatekeeper.

As an illustration, an iPhone user may thus become entitled to download, say, a calendar app from somewhere other than Apple’s App Store. Referred to as “sideloading,” this practice creates tremendous cybersecurity risks and privacy and safety concerns, which are blatantly ignored by the European regulator. Even further, assuming  iOS is a software application, could iPhone users be entitled to use iOS’s rivals’ operating systems in an iPhone? The sheer proprietary aspect of the digital ecosystem, let alone associated risks for privacy and cybersecurity, is put into question.

Following such an obligation, one can reasonably predict either an increase in prices for end users (since once-vertically integrated features may be prone to a double-marginalization problem), and a lower level of investments from the gatekeepers in its core platform services (since free ridership from unrewarding customers occurs).[276] In conclusion, while ignoring that these data access may be in “obvious tension with the GDPR,” this third practice enshrines free riding from both the business users and end users at the expense of innovation and economic viability of the gatekeeper’s digital ecosystem.[277]

Out-of-Court Settlements for Antitrust Claims Made Out of Question

The fourth practice prohibited is undoubtedly the most puzzling and surprising conduct included in the list in Article 5. It states that gatekeepers shall “refrain from preventing or restricting business users from raising issues with any relevant public authority relating to any practice of gatekeepers.”[278]

This prohibition refers to the contractual right of one contractual party in regard to the other contacting party. Contracting parties regularly decide that any issues arising out of contract performance may first be addressed through out-of-court settlements. These time-saving, cost-minimizing contractual clauses are ubiquitous as being welfare-enhancing. Should these attempts at out-of-court settlements fail, any contractual party’s fundamental right to seek a remedy before administrative or judicial bodies can never be suppressed. Consequently, this prohibition, which refers to the natural and legal person’s fundamental right to access the justice system (be it initially through administrative appeal or subsequently via the court system), is not only useless given the fundamental right to adequate judicial protection ascribed to any contractual party, but unneeded, as fundamental rights in the EU and at the national and international levels override such a specific and poorly defined entitlement.

This prohibition assumes that gatekeepers only can prevent or restrict market actors from making antitrust claims. More importantly, it is the very existence of such conduct by gatekeepers that is questionable. Such prohibition inherently yields detrimental results concerning the judicial system’s quality and incentivizes opportunistic behaviors.

First, the prohibited conduct consists of the unsupported claim according to which gatekeepers will retaliate against businesses if they raise issues to relevant public authorities. Given the numerous lawsuits and complaints filed by companies against gatekeepers, it is dubious that this prohibited conduct may significantly alter the current reality. This prohibition is either useless or harmful.

This superfluous prohibition nevertheless represents a shift of bargaining power in favor of business users. This change goes so far as to grant these business users an incentive to opportunistically extract rents from gatekeepers. Indeed, while the contract may falsely make the gatekeeper believe that out-of-court settlements are optimal solutions for contract-performance conflicts, the business user may opportunistically raise the stakes by going directly to the public authority or the court system, however small the initial stakes may be. Entitled to do so, the business user may thus accede out-of-court settlements and withdraw its complaint before the public authority.

Would these opportunistic behaviors benefit innovation and consumers? Legal-risk costs could eventually increase transaction costs and final prices for end users. Consequently, such prohibition constitutes a formidable incentive for business users to ignore contractual commitments and extract economic rents irrespective of fair competition. It also, unfortunately, increases distrust in the digital economy, as legal relationships among digital market players inevitably deteriorate because of increased legal risk aversion.

Second, such prohibition may constitute a green light for rivals to embark on (administrative or judicial) rent-seeking behaviors. Such detrimental opportunism at the expense of both gatekeepers and business users’ ability to compete on the merits may inevitably follow suit. Indeed, this banned conduct reveals a positive bias toward gatekeepers’ rivals in regulatory claims.

For instance, an app developer present in the Apple App Store and subject to Apple’s 30 percent commission fee may want to complain about that fee while reaping the App Store platform’s benefits. Such an obligation would make current antitrust complaints both more frequent (as the regulation incentivizes the complaints) and more useless (as the defendant would automatically lose irrespective of its merits). Thus, the current antitrust complaint by Epic Games against Apple is only the tip of the iceberg of the number of complaints that could unfold once such an obligation becomes enforceable.[279] In this regard, against its stated objectives, the DMA does not minimize the number of lawsuits, but rather dangerously incentivizes judicial rent-seeking behaviors. An app developer could indeed circumvent a contractual clause according to which out-of-court settlements may first be sought. Therefore, the gatekeeper may be unable to refrain the app developer from raising the issue with the relevant public authority before any amicable agreement is sought. This prohibition may generate numerous strategic behaviors detrimental to both the overall welfare of and the necessary confidence in economic relationships.

Prohibition of Data Interoperability

The fifth prohibition makes it impossible for a gatekeeper to favor data interoperability across both its services and the services of a third party: “In respect of each of its core platform services identified pursuant to Article 3(7), a gatekeeper shall … refrain from requiring business users to use, offer or interoperate with an identification service of the gatekeeper in the context of services offered by the business users using the core platform services of that gatekeeper.”[280]

This widespread prohibition relates to the ability of gatekeepers to force business users to adopt the platform’s authentication system (ID system or sign-ins). This suggests app developers are forced to use Google’s or Apple’s ID system in their apps. This is problematic because it ignores both consumer benefits and the competitive process.

Despite obvious consumer benefits related to data interoperability, this prohibition does not justify why data interoperability is commonly praised for digital platforms but despised when it involves digital gatekeepers. Consumers may not discriminate between data interoperability involving gatekeepers and not involving gatekeepers. Consumers prefer to sign in to the apps they use through notable gateways. They save time, do not have to remember passwords,  and benefit from data interoperability between the app and the sign-in provider. ID systems offered for free by the platforms are password-management tools that allow consumers to manage their passwords without remembering them or having to pay for chargeable password-management features. Indeed, the password-management market is exceptionally competitive, and often these services are provided at a price.[281]

Prohibiting the so-called gatekeepers from offering password-management services for free to consumers would likely increase consumer search costs, be time-consuming, and prevent consumers from enjoying the benefits of interoperating their data between the platform and the app. Also, such a prohibition may ignore the fundamentals of consumer preferences in digital markets. Instances of forced SSO are scarce, with no evidence of actual harm. Moreover, requiring gatekeepers to use third-party identification services may expose them, and their users, to security vulnerabilities they may not be able to mitigate but be ultimately liable for. These are highly questionable legal premises with fundamentally negative economic implications.

Second, this blank prohibition overlooks the very process of competition, forcing consumers to go against their general preference for sign-on options and choose password-management service providers. Indeed, Google and Facebook benefited from the first-mover advantage by offering widespread sign-in options alongside other sign-up options. In response to Google’s and Facebook's pioneering advances, in 2019, Apple began offering its own sign-in options. Regardless, each of these platforms has continuously offered consumers the choice to sign up with a more traditional option—say, an email and a password.

Nowadays, app developers complain about Apple’s sign-in option.[282] Doing so overlooks the fact that such an option was introduced precisely to strengthen competition against Google’s and Facebook’s sign-in options. Rather than exerting an alleged monopoly power and abusing its position, Apple seems to constrain its rivals and their sign-in options. Consumers are offered a greater choice of options—namely, Google’s, Facebook’s, Apple’s, or one of their own—than ever before. Consequently, competition and choices have both strengthened and increased.

Forced Interoperability of Ancillary Services (or the Implicit Prohibition of Bundling)

The sixth banned conduct disconnects services within a digital ecosystem. “In respect of each of its core platform services identified pursuant to Article 3(7), a gatekeeper shall … refrain from requiring business users or end users to subscribe to or register with any other core platform services identified pursuant to Article 3 or which meets the thresholds in Article 3(2)(b) as a condition to access, sign up or register to any of their core platform services identified pursuant to that Article.”[283]

Article 5(f) makes cross-tying illegal, prohibiting cross-viability of the digital ecosystem integrity. A platform may provide a range of services for free under the condition that other services are tied up. This reverts to the basics of the multisided platform model wherein one side of the platform subsidizes the other side of the platform. Article 5(f) makes the fundamental business model of multisided platforms less viable, thereby deterring investments in the platforms and hindering subsequent innovations.

Large platforms may be prevented from making access to ancillary services subject to subscription or registration to primary digital services by business users or end users. In other words, not only does such an obligation virtually ban tying and bundling for large platforms but it also lays down the obligation for interoperability of the gatekeeper’s ancillary services with business users’ services.

The DMA distinguishes, without real clarity, between the gatekeepers’ core platform services and its ancillary services.[284] As a recurring theme in the DMA, the main concern for the Commission is the alleged dual role of gatekeepers. Here, the dual role pertains to gatekeepers being both developers of operating systems and device manufacturers. The assumption here is that the gatekeeper can technically restrict access to third-party service providers' services in order to promote and favor their own services.

To illustrate, the fact that Google Pay and Apple Pay were designed on Android OS and iOS, respectively, provides an incentive for these OS developers to discriminate against, say, MasterCard, banks’ payment solutions, platforms payments solutions, etc. On June 16, 2020, the Commission, alleging that some of these practices are discriminatory and occur through non-interoperable services, opened an investigation into Apple’s practices regarding Apple Pay because “the Commission has concerns that Apple's terms, conditions, and other measures related to the integration of Apple Pay for the purchase of goods and services on merchant apps and websites on iOS/iPad devices may distort competition and reduce choice and innovation.”[285] Margrethe Vestager stated,:

Mobile payment solutions are rapidly gaining acceptance among users of mobile devices, facilitating payments both online and in physical stores. This growth is accelerated by the coronavirus crisis, with increasing online payments and contactless payments in stores. It appears that Apple sets the conditions on how Apple Pay should be used in merchants' apps and websites. It also reserves the “tap and go” functionality of iPhones to Apple Pay. It is important that Apple's measures do not deny consumers the benefits of new payment technologies, including better choice, quality, innovation and competitive prices. I have therefore decided to take a close look at Apple's practices regarding Apple Pay and their impact on competition.[286]

More generally, the DMA preemptively addresses this concern with Article 5(f), which lays down such an obligation of interoperability of the gatekeepers’ ancillary services because the Commission considers that:

Gatekeepers may also have a dual role as developers of operating systems and device manufacturers, including any technical functionality that such a device may have.… If such a dual role is used in a manner that prevents alternative providers of ancillary services or of software applications to have access under equal conditions to the same operating system, hardware or software features that are available or used in the provision by the gatekeeper of any ancillary services, this could significantly undermine innovation by providers of such ancillary services as well as choice for end users of such ancillary services. The gatekeepers should therefore be obliged to ensure access under equal conditions to, and interoperability with, the same operating system, hardware or software features that are available or used in the provision of any ancillary services by the gatekeeper.[287]

This obligation of nondiscrimination and forced interoperability of ancillary services is highly problematic for several reasons.

First, the regulatory obligation clashes with ongoing investigations. This exacerbates the concerns of a conflict of interest raised during Vestager’s nomination hearing. Also, it demonstrates that current competition rules can adequately address all sorts of competition concerns because the current investigations take place under current competition rules.

Second, the forced interoperability of ancillary services disregards fundamental issues of cybersecurity and digital protection. It is noticeable that the Commission has opened an investigation into Apple’s iOS, which is the OS that most prioritizes security, closedness, and safety of consumer transactions. Indeed, the forced interoperability of ancillary services—mainly payment services—exacerbates the security risks of degrading the quality of the entire operating system. Contractual restrictions thus need to be assessed in light of these risks to the digital ecosystem. Indeed, a report from the Dutch Competition Authority acknowledges these risks concerning in-app purchases (IAPs):

Consumers benefit the IAP system of Apple and Google. This increases the convenience for consumers: a consumer only has to enter their payment details once, and can thereafter pay with just one simple click, and it prevents sensitive data from going to third-party app providers that might not treat this data with care. On the other hand, the requirement to use IAP for certain apps may limit consumer choice: consumers are restricted to the payment systems chosen by Apple and Google.[288]

More generally, this obligation will lead gatekeepers to make their platforms interoperable with third-party service providers on the same terms as their own services. Such an obligation precludes gatekeepers from treating alternative providers of ancillary services or software applications differently from their services to access the operating system, hardware, or software features. Although such a feature may a priori seem desirable, imposing a nondiscrimination principle regardless of the security risks and the data-sharing implications of such interoperability will inevitably lower the quality and reliability of the whole digital ecosystem in order to artificially promote third-party ancillary services providers.[289] This intervention may provide short-term minimal gains for a handful of rivals, but, simultaneously, will provide considerable costs and risks in terms of the viability and reliability of the digital ecosystem’s safety.

Third, it can be argued that this obligation is not, and should not be, a competition concern. At most, it should be a matter of standardization. Indeed, standardization can set technical interoperability requirements that may increase the safety, security, and reliability of the digital ecosystem and maximize competition and innovation. On the other hand, standardization requirements such as the European Telecommunications Standards Institute (ETSI), the World Wide Web Consortium (W3C), the Internet Engineering Task Force (IETF), and the European Multi-Stakeholder Platform on ICT Standardisation can impose interoperable add-ons.[290] Moreover, the EU legal framework for technical standards is rich, diverse, and entirely satisfactory to address the concerns embodied in this obligation:

  1. Directive 2002/21/EC of the European Parliament and of the Council of 7 March 2002 on a common regulatory framework for electronic communications networks and services (Framework Directive);
  2. Directive (EU) 2018/1972 of the European Parliament and of the Council of 11 December 2018 establishing the European Electronic Communications Code (Recast);
  3. Directive 2002/19/EC of the European Parliament and of the Council of 7 March 2002 on access to, and interconnection of, electronic communications networks and associated facilities (Access Directive);
  4. Regulation (EU) No 1025/2012 of the European Parliament and of the Council of 25 October 2012 on European standardization, amending Council Directives 89/686/EEC and 93/15/EEC and Directives 94/9/EC, 94/25/EC, 95/16/EC, 97/23/EC, 98/34/EC, 2004/22/EC, 2007/23/EC, 2009/23/EC, and 2009/105/EC of the European Parliament and of the Council and repealing Council Decision 87/95/EEC and Decision No 1673/2006/EC of the European Parliament and of the Council; and
  5. Directive (EU) 2015/1535 of the European Parliament and of the Council of 9 September 2015 laying down a procedure for the provision of information in the field of technical regulations and of rules on Information Society services.[291]

In other words, it cannot legitimately be argued that technical operability is unachievable with the current EU regulatory framework.

The obligation of technical interoperability is misguided both in its substance, as it would increase consumer cost and degrade quality, and in its procedure, as specific regulatory instruments are better suited to achieve these objectives.

More generally, Application Programming Interfaces (APIs) and communication protocols can achieve technical interoperability between systems and software. Technical interoperability requires both syntactic (i.e., systems that can communicate among them) and semantic (i.e., information exchanged is understandable in both operable systems) systems. It is unclear whether the DMA encompasses both aspects, and thus proves incomplete or unclear.

Finally, this prohibition virtually targets all tying and bundling of services offered by gatekeepers to their app developers.[292] More specifically, contractual tying of primary services with ancillary services becomes prohibited for gatekeepers only.[293] The ban is enforceable irrespective of the pro-competitive, pro-efficiency rationale of such practice.[294] Such prohibition may weaken the overall economic viability of both the digital services and the competitiveness of the prices offered. Gatekeepers will no longer be able to offer adjacent services to consumers irrespective of these services’ complementarities.[295] The disproportionality of such prohibition creates legal questions regarding the validity of the practice in light of the EU general principles of law.

For instance, the DMA may lead consumers to enjoy Facebook Marketplace without a Facebook account, or to enjoy Amazon Video without an Amazon Prime account. These direct attacks on the gatekeepers’ cross-viability of services represent a blatant opportunity for customers and consumers to free ride with à la carte digital services at the expense of the digital ecosystem’s viability. The DMA requires unbundled access to digital services irrespective of the business models inherent to these services.[296]

With Article 5(f)’s prohibition, cross-subsidization becomes suspicious, albeit inherent to the multisided nature of digital platforms. According to Jean Tirole, one the subsidizing segment of the platform (i.e., here the advertising services) allows for the subsidized segment to thrive and flourish (e.g., search engine, social media platforms, ancillary services, disruptive innovations, etc.)[297] A cross-subsidization pricing strategy is core to the free services provided by digital platforms.[298] Regrettably, the prohibition tackles cross-subsidization and wishes to lower prices at the expense of innovation, platform viability, and, ultimately, consumer benefits.

This prohibition is misguided because it unduly prohibits cross-tying concerning ancillary services for gatekeepers only. It addresses an issue of technical interoperability that standardization requirements are best suited to address.

Communication of Advertising Prices and Publishers’ Remunerations

The final obligation aims at fostering price transparency by gatekeepers to publishers and advertisers. “In respect of each of its core platform services identified pursuant to Article 3(7), a gatekeeper shall … provide advertisers and publishers to which it supplies advertising services, upon their request, with information concerning the price paid by the advertiser and publisher, as well as the amount of remuneration paid to the publisher, for the publishing of a given ad and for each of the relevant advertising services provided by the gatekeeper.”[299]

In a quest to advance media diversity, the obligation entails the inability of gatekeepers to retain information on publishers’ remuneration and advertising services’ prices whenever advertisers or publishers request such information. The goal such an obligation seeks to achieve is unclear. It does not relate to the revenue-sharing mandatory schemes currently envisaged in Australia, Europe, and the United States.[300]

The pursued shift of bargaining power between two news aggregators Facebook and Google and publishers is highly questionable, as they do not represent the reciprocal nature of the relationship news aggregators have with news publishers—the former bring web traffic and ad revenue to the latter without the latter having to pay or do anything but generate popular articles.[301] However, regulators and legislators worldwide want to introduce some snippet taxes against these two specific news aggregators only. These legislative proposals aim to protect traditional news publishers from the gales of the digital revolution, which will ultimately harm consumers and deter digital innovation in journalism. This obligation does not address the ill-fated concern of revenue sharing.

It instead lays down transparency rules. The obligation also requires gatekeepers to communicate pricing information for advertisers and publishers—albeit “upon request.” But why? And for whose benefit? Purportedly violating potential confidential contractual clauses, and more generally the privity of contractual obligations, this obligation forces the disclosure of sensitive information for no tangible benefits. It will further demonize gatekeepers at the benefit of their direct rivals—namely, news aggregators not designated as gatekeepers.[302]

There is no general right to access this information, and a relative right to access the information is given to the requester only. Nevertheless, the sharing of pricing information implicitly assumes that advertising prices are opaque and discriminatory, and that publishers' prices are non-competitive. These assumptions are stated relatively explicitly in Recital 42 of the DMA:

The conditions under which gatekeepers provide online advertising services to business users including both advertisers and publishers are often non-transparent and opaque. This opacity is partly linked to the practices of a few platforms, but is also due to the sheer complexity of modern day programmatic advertising. The sector is considered to have become more non-transparent after the introduction of new privacy legislation, and is expected to become even more opaque with the announced removal of third-party cookies. This often leads to a lack of information and knowledge for advertisers and publishers about the conditions of the advertising services they purchased and undermines their ability to switch to alternative providers of online advertising services. Furthermore, the costs of online advertising are likely to be higher than they would be in a fairer, more transparent and contestable platform environment. These higher costs are likely to be reflected in the prices that end users pay for many daily products and services relying on the use of online advertising. Transparency obligations should therefore require gatekeepers to provide advertisers and publishers to whom they supply online advertising services, when requested and to the extent possible, with information that allows both sides to understand the price paid for each of the different advertising services provided as part of the relevant advertising value chain.[303]

These allegations underpinning the obligation of Article 5(g) reflect the “nirvana fallacy.”[304] Absent counterfactual claims, these claims are hardly contestable—and yet, they remain unsubstantiated.[305]

Despite lacking any evidence, the Commission alleges that online advertising prices are non-competitive and too high, and an otherwise structured online advertising market would yield more significant benefits for lower prices. Lacking counterfactuals, this claim can neither be proven and nor debunked.

However, contrary to the European Commission’s presumptions of unreasonable prices for online advertising markets, it has been well substantiated that online advertising prices are extremely competitive, as the competition is both aggressive (for the benefit of end consumers) and highly innovative.

This obligation overlooks the fact that the economic viability of these platforms’ advertising services relates to their ability to provide free services. In other words, the profitability of the advertising side of the platform is core to the free provision of services on the other side of the platform.

Specifiable Obligations—Article 6

The obligations Article 5 lays down for all gatekeepers to comply with are not the only obligations within the DMA. Indeed, Article 6 of the DMA compiles another set of them for gatekeepers whose business models and standard features are severely put into question, or “obligations for gatekeepers susceptible of being further specified.”[306]

Contrary to initial plans by the European Commission, Article 6 is not a “greylist” aside to Article 5’s “blacklist” as Articles 5 and 6 are both blacklists. Every gatekeeper must comply with all obligations of Articles 5 and 6. This is a daunting increase in the asymmetrical regulatory burden gatekeepers must cope with in sheer opposition to their rivals’ regulatory exemptions.

What is the rationale for creating a distinct set of obligations with Article 6 as separate from Article 5? Article 6 obligations are so vague, general, and wide reaching that the Commission felt the need to provide subsequent clarifications for them. Indeed, “susceptible to be specified” must be read as “in need of clarification.” Nevertheless, one may speculate about the precise content of Article 6 obligations.

Before delving into these “specifiable” obligations—as if Article 5 obligations’ vagueness did not require further clarification—one feature distinguishes Article 5’s from 6’s obligations. According to Article 15(4):

When the Commission pursuant to Article 3(6) designates as a gatekeeper a provider of core platform services that does not yet enjoy an entrenched and durable position in its operations, but it is foreseeable that it will enjoy such a position in the near future, it shall declare applicable to that gatekeeper only obligations laid down in Article 5(b) and Article 6(1) points (e), (f), (h) and (i) as specified in the designation decision. The Commission shall only declare applicable those obligations that are appropriate and necessary to prevent that the gatekeeper concerned achieves by unfair means an entrenched and durable position in its operations.[307]

In addition to “foreseeable” gatekeepers potentially harming European SMEs—a highly contestable notion—certain Article 6 obligations will apply to them alongside others from Article 5.

More oddly, Article 7(2) only refers to the obligations laid down in Article 6, while Article 7(1) requires gatekeepers to implement their obligations stated in Articles 5 and 6 of the DMA. But Article 7(2) states,

Where the Commission finds that the measures that the gatekeeper intends to implement pursuant to paragraph 1, or has implemented, do not ensure effective compliance with the relevant obligations laid down in Article 6, it may by decision specify the measures that the gatekeeper concerned shall implement. The Commission shall adopt such a decision within six months from the opening of proceedings pursuant to Article 18.[308]

Article 7(2) implies that gatekeepers must implement Article 5 obligations regardless of the wording, spirit, and objectives of Article 6 obligations, which themselves require further specifications—namely, clarification.

The DMA purportedly organizes legal vagueness with Article 6. This makes Article 6 obligations particularly prone to countless lawsuits and legal and economic arguments. These unintended consequences contradict the DMA’s objective to avoid legal disputes and favor regulatory compliance. Not only will there not be regulatory compliance before legal disputes have clarified the enigmatic meaning of Article 5 and 6 obligations, but most unfortunately, the legal uncertainty generated via Articles 5, 6, and 7 may deter investments and innovation in the digital ecosystems.

Consequently, Article 6 obligations are vague but applicable and invokable directly against gatekeepers that, presumably, may seek judicial clarification before regulatory compliance. This regulatory game may prove to be a lose-lose situation. As the Commission generates endless legal disputes it actually aims to avoid, and thus cannot force compliance, gatekeepers will evolve amid legal vagueness, and business users and consumers will suffer from stifled innovation.

The Commission chose with the DMA to have a “partially flexible framework of designation and updating of obligations, including a regulatory dialogue for the implementation of some.”[309] The regulatory “dialogue” suggested with Article 6 has its own pitfalls and takes place on unequal terms, since the Commission has the final say. Should the dialogue fail to reach a constructive conclusion, the Commission will always have the dual pathways to enforce regulatory obligations, sue the company under traditional competition rules, or both. These unequal bargaining positions hardly pave the way for a fruitful dialogue.

Gatekeepers may be reluctant to enter such a dialogue, which resembles some foolish game wherein all the information, data, and justifications provided by the gatekeepers may be used against them in a looming legal dispute. The DMA reveals that the dialogue will not consider the relevance of applying obligations to gatekeepers, but rather will ensure that these obligations are best complied with as to minimize circumvention risks:

However, it may in certain cases be appropriate for the Commission, following a dialogue with the gatekeeper concerned, to further specify some of the measures that the gatekeeper concerned should adopt in order to effectively comply with those obligations that are susceptible of being further specified. This possibility of a regulatory dialogue should facilitate compliance by gatekeepers and expedite the correct implementation of the Regulation.[310]

From a strategic viewpoint, gatekeepers may be well advised to remain silent, as they will be placed in a situation of being under investigation or even facing prosecution. These dialogues are not genuine. They are investigations without names—an undercover prosecutorial exercise disguised as data compilation. Regulatory dialogues in the DMA will fail from the start.

Should regulatory dialogue start for any reason, they will generate regulatory capture detrimental to society. Indeed, the vicinity between gatekeepers and regulators will ultimately lead digital gatekeepers to be treated differently from one another—and from smaller companies. Finally, domestic gatekeepers will be treated differently from foreign companies since the explicit goal of the DMA is to boost European tech companies in light of American and Chinese tech companies. Regulatory dialogue cannot avoid these endless regulatory captures.

Thus, the discriminatory and discretionary procedures these regulatory dialogues imply will leave such considerations as efficiency, welfare, innovation, and the common good far away from practical, short-term, political considerations. Regulations of telecom companies and data regulation already illustrate such protectionist bias.[311] A similar tendency will inevitably emerge for foreign gatekeepers when they engage in dialogues with the European Commission.

The European Commission's regulatory capture by powerful and influential large European SMEs is a risk that is both real and damaging to the economy. The Commission will therefore specify Article 6 obligations subject to interest group pressures.

Article 6 obligations mostly deal with the prohibition of self-preferencing and data-sharing obligations. They lay down nondiscriminatory principles in the digital ecosystems irrespective of the risks of free riding and overlooking proprietary assets.

Prohibition of Using Third-Party Providers’ Data

Article 6(1)(a) forces gatekeepers to “refrain from using, in competition with business users, any data not publicly available, which is generated through activities by those business users, including by the end users of these business users, of its core platform services or provided by those business users of its core platform services or by the end users of these business users.”[312]

Reflecting the so-called “conflict of interest” between a gatekeeper's dual role as a platform and retailer/distributor, these obligations correspond to the ongoing antitrust investigation of Amazon.[313] Commissioner Vestager said in the opening of this investigation,

We must ensure that dual role platforms with market power, such as Amazon, do not distort competition. Data on the activity of third-party sellers should not be used to the benefit of Amazon when it acts as a competitor to these sellers. The conditions of competition on the Amazon platform must also be fair. Its rules should not artificially favour Amazon's own retail offers or advantage the offers of retailers using Amazon's planning and delivery services. With e-commerce booming, and Amazon being the leading e-commerce platform, a fair and undistorted access to consumers online is important for all sellers.[314]

There are no other current or past antitrust cases in the European Union about such an alleged practice. Indeed, this practice—and the associated regulatory obligation in the DMA—is entirely new. Consequently, it cannot be legitimately argued, again, that the DMA is based on well-known practices and extensive experience when the only case that grounds the present obligation follows the mere opening of an investigation.

Furthermore, these leaks, from the ongoing antitrust investigations to regulatory proposals, raise doubts about the relevance of the ongoing investigations. Should they be stopped so that the regulatory obligation is enforced now rather than in years through judicial decision? Would regulatory duties complement or substitute judicial findings? The Commission conducts contestable interlinkages between investigation teams and teams at the heart of the reforms. However, it is less clear whether the Commission intends to sue and regulate or favor one route. Choosing the cumulative routes (i.e., to sue and regulate) could provide excessive enforcement actions and an unfair focus on one company at the expense of other companies engaged in similar practices.

Amazon has repeatedly denied having engaged in such practice. Thus, it is up to judicial discussion, evidence-based arguments, and applying a case-by-case approach to determine whether a violation of competition rules has taken place. Regardless, the current investigation demonstrates that current rules are mostly sufficient to investigate and sanction potential abuses of dominant positions, contrary to the claims that the DMA is needed.

Also, data protection rules such as the GDPR and its enforcement may prove to be better suited to ensure that data is used appropriately. Despite the appropriateness of current competition rules and the GDPR, Article 6(1)(a) adds on a regulatory obligation without apparent benefit to the existing regulatory framework.

Uninstallation Requirements

Default settings and preinstalled software applications are core to the cross-subsidization element of multisided platforms, as discussed. Platforms may gain no revenue from a range of services and become profitable only when consumers use another range of services. Thus, preinstallation settings are ubiquitous in the digital industry. Nevertheless, Article 6(1)(b) states that:

[A gatekeeper shall] allow end users to un-install any pre-installed software applications on its core platform service without prejudice to the possibility for a gatekeeper to restrict such un-installation in relation to software applications that are essential for the functioning of the operating system or of the device and which cannot technically be offered on a standalone basis by third-parties.[315]"

More precisely, the obligation considers those software applications as “essential” to the digital ecosystem. The DMA is sensible in that it limits the duty to allow uninstallation when disproportionate. One clear example here would be to force Google Android to offer free Android OS without having the right to restrict uninstallation of, say, Google Chrome or Google Play Store in the devices. In contrast, one can arguably understand that both revenue-generating software applications are essential to Google’s ability to provide Android OS for free as a rival to Apple’s iOS. The suitable qualification of Article 6(1)(b) nevertheless contradicts the blames formulated in the Google Android decision of 2018.

In this decision, the Commission expressed concerns about the inability to uninstall the Google Search app and Google Chrome on devices. In reply, Google argued that such impossibility to uninstall apps may be irrelevant since the installation of competing apps is always possible. No clear evidence has yet been found that the inability to uninstall apps powerfully influences consumer choices and, most importantly, prevents consumers from downloading and using competing apps. For instance, the preinstalled Bing search engine on all PCs—wherein Microsoft remains dominant—has not created leverage for Bing to become a more popular search engine. Preinstallations rarely prevent consumers from switching—and only at the margin influence consumer choices.

Nevertheless, Article 6(1)(b) of the DMA assumes without evidence that consumer choice is hampered by such an inability to uninstall apps, although leaving open for discussions (and lengthy litigations) over the extent to which the concerned apps are deemed essential to the platform’s ecosystem.

Mandatory Access to Third-Party App Stores and Side-Loading Apps

Article 6(1)(c) is peculiar, and could have been written explicitly for Apple’s App Store and Google’s Play Store:

[A gatekeeper shall] allow the installation and effective use of third-party software applications or software application stores using, or interoperating with, operating systems of that gatekeeper and allow these software applications or software application stores to be accessed by means other than the core platform services of that gatekeeper. The gatekeeper shall not be prevented from taking proportionate measures to ensure that third-party software applications or software application stores do not endanger the integrity of the hardware or operating system provided by the gatekeeper.[316]

With such an obligation, operating system providers will have to allow competing app stores to be present on Android OS devices. However, this overlooks the fact that many app stores already exist and are available to business users and end users.[317] But it is commonly understood that most app developers “will opt for the $25 route of Google Play as it has one of the most guaranteed audiences.”[318] This reveals a fundamental feature of the app store market—should such a relevant market ever exist.

It also underlies the tremendous network effects inherent to app stores. App developers legitimately want their apps to reach the widest audiences possible, which leads to a natural concentration of the market, as app developers do not want to be present in niche app stores in which the viewability is limited. Therefore, irrespective of the number of app stores the Commission wants to introduce on the smartphones of EU citizens, there is a strong probability that it is the app developers’ choice—and consumer preference—to focus on and be present in app stores with the largest audiences.

It can also be argued that Apple does not allow for alternative app stores on iPhones, with the present obligation clashing with Apple’s restrictive requirements. But it cannot be ignored that these restrictions are core to Apple’s closed and integrated operating system, which maximizes integrity over choice, and safety over openness. The obligation laid down does nevertheless encapsulate the crucial notion of “integrity of proprietary digital ecosystems.” It remains to be seen to what extent the Commission can kink Apple’s proprietary business model. IP rights and trade secrets may constitute legitimate reasons to bar competition rules or regulatory obligations from infringing on both proprietary and considerable safety considerations.

Another argument in favor of a more substantial number of app stores on smartphones relates presumably to gatekeepers' alleged soaring prices. Indeed, the Commission has opened an investigation into Apple’s App Store rules wherein a 30 percent commission applies to all subscription fees.[319] In a spectacular move, Epic Games sued Apple in the United States for its 30 percent commission fee on App Store.[320]

To argue that a 30 percent fee constitutes exploitative prices tantamount to abuse of dominant position under Article 102 TFEU is a strong claim yet to be substantiated. Franchise fees are often close to 30 percent. Also, telecom fees before the advent of the iPhone in 2008 charged 70 percent, leaving a considerable financial relief of 30 percent of revenue to the software/app developers. Consequently, more competition in app stores may not bring about a fall prices as the management of the app stores brings with it non-negligible costs, and fees are already set at a competitive price compared with previous fees charged by telecom companies.

The mandatory access obligation runs the risk of undermining both proprietary and safety aspects of app stores based on the unfounded claim that prices are too high for a service that has enabled apps to develop at an unprecedented scale. Again, this prohibition disturbs a flourishing app market in order to achieve unrealistic objectives at the expense of both app developers and consumers.

Prohibition of Self-Preferencing

Article 6(1)(d) is a highly problematic obligation that may significantly impact innovation in digital sectors and fairness of competition with off-line/more-traditional companies. This Article prohibits self-preferencing—namely, a platform's ability to promote its services and products in its rankings, stating that “a gatekeeper shall … refrain from treating more favourably in ranking services and products offered by the gatekeeper itself or by any third party belonging to the same undertaking compared to similar services or products of third party and apply fair and nondiscriminatory conditions to such ranking.”[321]

Algorithm-driven rankings reveal consumer preferences. In this respect, self-preferencing is frequently consumer preferencing.[322] Algorithmic biases are either rare or hard to evidence. Unless the regulator wants to dive deep into the complex mathematical formulas of algorithms, the fact that the platform’s own products and services are better ranked than those of downstream rivals may merely reflect consumer preferences.

For platforms to systematically, and against consumer preferences, promote less efficient products and services only because they own them may degrade the quality of the platform. Consumers may thus resort to platform competition and leave the platform altogether. If Google Shopping ranks results in an unsatisfactory way, consumers will shop on Amazon. And if Amazon ranks results in an equally non-consumer-oriented manner, consumers will shop at Shopify, Etsy, or tens of other competitors which are only one click away. Should self-preferencing be evidenced, it can still be argued that this leads to pro-efficiency outcomes.[323] Prohibition of self-preferencing may also clash with a company's legitimate pursuit of commercial interests as it partakes in competition on the merits.[324]

A ban on self-preferencing will harm both consumers and innovation, as product offerings will be reduced. The platform will refrain from offering competitive products simply because it is the platform, not because it cannot deliver efficient products. Such a ban will result in foregone consumer benefits.

Article 6(1)(d) constitutes a direct blow to the digital platforms’ business models. But as the DMA sits on an unequal stance, self-preferencing will remain essential to the off-line world—from supermarkets to banking to commodity markets. And again, should self-preferencing be limited, let alone prohibited, current competition rules fully provide the necessary actionable legal basis for so doing, as evidenced by the Google Shopping decision currently under judicial review.[325]

Instead, the prohibition of self-preferencing may induce considerable unintended consequences, such as de facto treatment of the platform under the essential facilities doctrine. Entrenchment effects will increase rather than decrease. Nondiscrimination principles entice the implementation of an essential facilities doctrine in platforms that nevertheless are not indispensable and whose stronger market position does not need to be entrenched. Unfortunately, the DMA overlooks the well-known and widely discussed unintended consequences of such a ban—and the arguments both against it and for self-preferencing .[326]

Prohibition of Lock-Ins

Article 6(1)(e) states that “a gatekeeper shall … refrain from technically restricting the ability of end users to switch between and subscribe to different software applications and services to be accessed using the operating system of the gatekeeper, including as regards the choice of Internet access provider for end users.”[327]

This obligation prohibits “lock-ins” of consumers, as gatekeepers may not impose technical restrictions on switching end users. End users will be entitled to delete apps and switch away from default apps without technical restriction:

Gatekeepers should therefore ensure a free choice irrespective of whether they are the manufacturer of any hardware by means of which such software applications or services are accessed and shall not raise artificial technical barriers so as to make switching impossible or ineffective. The mere offering of a given product or service to end users, including by means of pre-installation, as well as the improvement of end user offering, such as better prices or increased quality, would not in itself constitute a barrier to switching.[328]

The idea that switching costs between operating systems are high for end users is prevalent in the Google Android decision. It was wrongly believed that switching costs for end users between Android OS and Apple’s iOS are so significant that Google Android and Apple iOS are not direct competitors. In its lightly substantiated decision, the Commission considered that:

Users of Google Android devices would face substantial costs when switching to iOS devices…. These include the need to download and purchase existing apps for the new smart mobile OS, the need to learn and become familiar with a new interface and the need to transfer a large amount of data through often inconvenient and imperfect mechanisms…. The existence of substantial switching costs has been confirmed …The existence of substantial switching costs is also confirmed by Apple's launch in September 2015 of a “Move to iOS” app as part of its iOS 9 release, as an attempt to make switching easier.[329]

To claim that switching from an iPhone to an Android smartphone, and vice versa, generates “substantial costs” is obviously an exaggeration. The Commission itself acknowledges that the switch is eased by tech companies themselves, such as the launch by Apple of the Move to iOS app, with which adaptation from one device to another takes a few hours, if not only minutes. If we are to take consumer behaviors seriously, the competitive constraints exerted by one operating system over another cannot be discarded as abruptly as the Commission did in the Google Android decision.

Actually, in practical terms, as by requiring only seven steps within the app, Apple ensures that switching from Android OS to Apple iPhone is as easy as it can get.[330] At the same time, switching to Android OS from Apple’s iOS seems equally easy.[331]

Either way, the process seems to be as easy (if not easier) as switching from one computer to another, one car to another, one TV to another, etc. But would we argue that these products are not in competition with one another only because of the marginal (and insubstantial) switching costs? Switching from one operating system to another is easy, and well documented.[332] In fact, the lock-in effect of phone contracts seems much more powerful than any technical lock-in effect from operating systems.

Technical lock-in referred to in Article 6(1)(e) may relate to the Google Android decision wherein the lock-in effects were not convincing for most commentators. Lock-in effects of this obligation may also refer to the prohibition of gatekeepers to lock users into a particular Internet service provider. Indeed, Recital 51 states this obligation more clearly when it justifies the obligation:

Gatekeepers can hamper the ability of end users to access online content and services including software applications. Therefore, rules should be established to ensure that the rights of end users to access an open internet are not compromised by the conduct of gatekeepers. Gatekeepers can also technically limit the ability of end users to effectively switch between different Internet access service providers, in particular through their control over operating systems or hardware. This distorts the level playing field for Internet access services and ultimately harms end users. It should therefore be ensured that gatekeepers do not unduly restrict end users in choosing their Internet access service provider.[333]

Such prohibition seems less detrimental to consumers and innovation than the prohibition of technical lock-ins whose effects are exaggerated and can be legitimately justified by different proprietary standards. It is unclear as to what this prohibition adds compared with the privately initiated lowering of barriers to switching.

Mandatory Interoperable Add-Ons

Article 6(1)(f) will force gatekeepers to allow other “ancillary services providers” (such as payment service providers, cloud service providers, sign-in service providers, etc.) to be interoperable with the gatekeeper’s core platform services. Indeed, the obligation will entitle “business users and providers of ancillary services access to and interoperability with the same operating system, hardware or software features that are available or used in the provision by the gatekeeper of any ancillary services.”[334]

This entitlement is equivalent to regulatory-embedded free riding. Not only will a gatekeeper no longer be able to favor their ancillary services, but rivals’ ancillary services will be entitled to have the same treatment as the gatekeeper’s ancillary services, irrespective of both the gatekeeper’s maintenance responsibilities inherent to the digital ecosystem and the innovation incentives for the gatekeepers to invest in downstream services.

In other words, the obligation strongly incentivizes the gatekeeper to refrain from providing ancillary services because such a provision would not generate a specific advantage for the platform due to the nondiscriminatory access entailed by Article 6(1)(f)’s obligation. The alleged dual role of the gatekeeper as provider of both core platform and ancillary services again focuses the primary source of criticism:

If such a dual role is used in a manner that prevents alternative providers of ancillary services or software applications to have access under equal conditions to the same operating system, hardware or software features that are available or used in the provision by the gatekeeper of any ancillary services, this could significantly undermine innovation by providers of such ancillary services as well as choice for end users of such ancillary services. The gatekeepers should therefore be obliged to ensure access under equal conditions to, and interoperability with, the same operating system, hardware or software features that are available or used in the provision of any ancillary services by the gatekeeper.[335]

This obligation is extraordinarily far-reaching and may entirely disrupt the gatekeeper’s business model, if not its viability. For instance, does this obligation infer that Google News may not be portrayed in a prominent manner? Should competitors’ news aggregators be favored in Google Search irrespective of the rights and entitlement of Google on its own search engine platform? Does this obligation infer that Microsoft should uninstall Bing from Windows PCs and let other ancillary service providers such as Google, DuckDuckGo, etc. be treated equally irrespective of Microsoft’s rights on Windows PCs? Does this obligation suggest that Uber cannot promote Uber Eats any longer but should treat Deliveroo on the same terms as Uber’s own transportation services app?

The implications of such obligations are countless—as will be the legal disputes, presumably. Cross-subsidization and the viability of the whole digital ecosystem are undoubtedly under threat with these obligations. Ultimately, with decreased profitability and increased economic duress, consumers will pay a higher price, and innovation will deplete.

Mandatory Tools for Advertisers and Publishers

In an additional attempt to further heat up the debate and tensions between news aggregators and news publishers, Article 6(1)(g) lays down an original obligation consisting of forcing gatekeepers to “provide advertisers and publishers, upon their request and free of charge, with access to the performance measuring tools of the gatekeeper and the information necessary for advertisers and publishers to carry out their own independent verification of the ad inventory.”[336]

Echoing Article 5(g), Article 6(1)(g) states,

To further enhance fairness, transparency, and contestability of online advertising services designated under this Regulation as well as those that are fully integrated with other core platform services of the same provider, the designated gatekeepers should therefore provide advertisers and publishers, when requested, with free of charge access to the performance measuring tools of the gatekeeper and the information necessary for advertisers, advertising agencies acting on behalf of a company placing advertising, as well as for publishers to carry out their own independent verification of the provision of the relevant online advertising services.[337]

The assumption holds that the condition pertaining to advertising services is “non-transparent.” This assumption alleges that such opacity results from the gatekeeper’s unilateral choices. But confidential contracts between news publishers and news aggregators may prevent the publication and disclosure of confidential clauses. This means the present obligations may also disregard the news publishers’ willingness to retain the confidentiality of the clauses of the contracts signed with news aggregators. For instance, the agreement between Google and French news publishers was said to have “infuriated many other French outlets, which deemed it unfair and opaque.”[338] News publishers wanted to keep the agreement confidential for obvious financial reasons.

This mandatory transparency may not only run counter to contractual confidentiality clauses but also, from a competitive viewpoint, make public information to such an extent that collusive practices and parallel pricing become easier to maintain. Indeed, advertising prices will be compared, shared, and equalized across news aggregators so that the advertising prices may very well be less competitive and more cartelized. Competition will halt inasmuch as innovation will deplete. Transparency will increase, but at a great cost.

Data Portability and Data Instant Accessibility

Article 6(1)(h) creates a new obligation for gatekeepers concerning data management: “[A] gatekeepers shall … provide effective portability of data generated through the activity of a business user or end user and shall, in particular, provide tools for end users to facilitate the exercise of data portability, in line with Regulation EU 2016/679, including by the provision of continuous and real-time access.[339]

This obligation is a transposition of the spirit of the GDPR to business users and end users concerning their data held by gatekeepers, granting a right to access, retrieve, and collect their data at any time from gatekeepers. Recital 54 of the DMA outlines that “business users and end users should be granted effective and immediate access to the data they provided or generated in the context of their use of the relevant core platform services of the gatekeeper, in a structured, commonly used and machine-readable format.”[340] The data portability and associated instant accessibility rights enshrined in this obligation seem unsurprising when read in conjunction with the GDPR.

Nevertheless, this obligation is incredibly detailed and demanding, with far-reaching consequences for gatekeepers, and immediate access is both technically demanding and legally contestable—how “immediate” ought the access be? Data portability must be instantaneously effective and performed through high-quality application programming interfaces. Recital 54  states that “facilitating switching or multi-homing should lead, it is assumed, to an increased choice for business users and end users. It would create an incentive for gatekeepers and business users to innovate.”[341]

These detailed obligations are daunting, and prohibitively demanding for gatekeepers that have to handle millions of business users’ data and hundreds of millions of end users’ data. To make them immediately accessible requires massive computational capabilities, which may be challenging even for gatekeepers. To overlook the capability constraints of gatekeepers may discard real barriers to delivering immediate access.

More insidiously, this obligation is illustrative of the kind of obligation a foreseeable gatekeeper, referred to in Article 15(4), will never be able to cope with—and may deter SMEs from becoming a gatekeeper, or even something close to a gatekeeper such as a foreseeable one, themselves. Indeed, such a prospect would immediately run the risks of being subject to incommensurable data-sharing obligations such as those of Article 6(1)(h), on top of the already enforceable GDPR. In other words, this obligation would constitute a formidable barrier to expansion—thereby reducing, rather than intensifying, the competition of SMEs concerning established gatekeepers—and signify that current gatekeepers will have to comply with complex data obligations. In contrast, smaller rivals may think twice before reaching the quantitative indicators of Article 3 or enjoying an “entrenched” market position. Otherwise, these smaller rivals will bear the costs associated with Article 6(1)(h) and similar obligations.

Data Access For Business Users

Remarkably similar to the previous obligation, per Article 6(1)(i) of the DMA,

[A gatekeeper shall} … provide business users, or third parties authorized by a business user, free of charge, with effective, high-quality, continuous and real-time access and use of aggregated or non-aggregated data, that is provided for or generated in the context of the use of the relevant core platform services by those business users and the end users engaging with the products or services provided by those business users; for personal data, provide access and use only where directly connected with the use effectuated by the end user in respect of the products or services offered by the relevant business user through the relevant core platform service, and when the end user opts in to such sharing with a consent in the sense of the Regulation (EU) 2016/679.[342]

This obligation implies that “a gatekeeper should not use any contractual or other restrictions to prevent business users from accessing relevant data and should enable business users to obtain consent of their end users for such data access and retrieval, where such consent is required under Regulation (EU) 2016/679 and Directive 2002/58/EC.”[343]

The obligation follows the previous obligation but may additionally raise doubt as per its relevance. Since Regulation 2016/679 and Directive 2002/58 already provide such access to data, it is unclear how much this obligation provides valuable information and useful mandatory implications for gatekeepers. Finally, this obligation is so repetitive that it legitimately may be perceived as redundant to the previous one.

FRAND Access to Ranking and View Data for Search

Fair, Reasonable and Non-Discriminatory (FRAND) access to data, although implicit throughout the DMA, is explicitly enshrined in Article 6(1)(j), which states that gatekeepers shall:

provide to any third party providers of online search engines, upon their request, with access on fair, reasonable and non-discriminatory terms to ranking, query, click and view data in relation to free and paid search generated by end users on online search engines of the gatekeeper, subject to anonymisation for the query, click and view data that constitutes personal data.[344]

In a clear reiteration of the Google Shopping decision rationale, this obligation provides FRAND access to business users of aggregated datasets containing information about their search history and behaviors.[345] Gatekeepers allegedly have unassailable market positions thanks to their control of these datasets and search engines:

Gatekeepers should therefore be obliged to provide access, on fair, reasonable and non-discriminatory terms, to these ranking, query, click and view data in relation to free and paid search generated by consumers on online search engine services to other providers of such services, so that these third-party providers can optimise their services and contest the relevant core platform services. Such access should also be given to third parties contracted by a search engine provider, who are acting as processors of this data for that search engine.[346]

Business users will possibly free ride on the gatekeepers’ core platform services to provide ancillary services because the FRAND access vague and prone to legal disputes.

Irrespective of the gatekeepers’ proprietary rights and competition on the merits, Article 6(1)(j) grants FRAND access to business users thereby entitled to free ride the gatekeepers' core platform services developed over the years.

A report from the Panel of Economic Experts following the DMA acknowledges the indeterminacy of the notion of “fairness” in the DMA, as “the precise definitions of fairness and the means to measure it have not been provided.”[347] The report is critical of this unfortunate imprecision, as there are no details on what FRAND means, except for a somewhat vague reference in Recital 57 of the DMA: "Pricing or other general access conditions should be considered unfair if they lead to an imbalance of rights and obligations imposed on business users or confer an advantage on the gatekeeper which is disproportionate to the service provided by the gatekeeper to business users or lead to a disadvantage for business users in providing the same or similar services as the gatekeeper.”[348]

Although supportive of the FRAND access requirements, the report understands the need to for precision in the requirements following FRAND access to online search engines as required in Article 6(1)(j), but also with the FRAND access requirements outlined in the next article concerning app stores (Article 6(1)(k)). Indeed, the “fairness” requirement may suggest “fairness as efficiency” justice or “fairness as equality” justice.

Suppose fairness means efficient conduct. Efficient conduct may harm less efficient rivals and benefit society in general, as it is the essence of the competitive process.[349] Suppose fairness means equality. Inefficient conduct may be protected by regulation against the competitive process and innovation. Indeed, in the example of search engines, there cannot be equality since, per definition, rankings imply an ordering of search results that may benefit some and harm others—which is inherent to rankings, unless we revert to Yellow Pages-type ordering. Inequality of treatment is justified whenever such treatment distinguishes between the efficiencies of rivals and follows consumer preferences.

Consequently, to require “fair” access as a form of equality of treatment irrespective of consumer preferences and the efficiency of the business users may decrease social welfare, promote inefficient business users at the expense of efficiency, and deter innovation, for increased efficiency by business users violates consumer preferences.[350]

FRAND access to search engine services will lead to detrimental neglect of the overall business model of search engines—wherein cross-subsidization remains essential—and involve a de facto application of the essential facilities doctrine and legal and technical separation of services.[351] The essential facilities doctrine entrenches, rather than displaces, market positions.

The DMA unfolds considerable unintended consequences for the sake of fair access to certain core platform services efficiently provided by gatekeepers that enjoy not only a first-mover advantage but also efficiency superiority through constant innovation.[352]

For the DMA to apply FRAND access to search engine services would entail the digital platform being assumed to hold standard-essential patents (SEPs) justifying the application of the essential facilities doctrine.[353] IP rights limit the extent to which FRAND access can be imposed.[354] SEPs are core to the legal basis for FRAND access.[355] Indeed, the U.S. Supreme Court has notably considered that “undertaking to grant licenses on FRAND terms creates legitimate expectations on the part of third parties that the proprietor of the SEP will, in fact, grant licenses on such terms.”[356]

Search engines may well argue that they do not hold SEPs, given the competition among search engines, irrespective of the market shares. Search engine patents are not essential because substitutes exist. As SEPs “do not allow for inventing or patenting around, thereby posing a barrier to entry for new entrants,” FRAND commitments apply to SEPs.[357] But can it be legitimately argued that search engines’ algorithms are SEPs? The presence of alternative search engines demonstrates that they are not.

Thus, FRAND commitments are inappropriate in that competitive environment. According to EU law, it cannot be convincingly argued that the algorithms of search engines are SEPs because SEPs ought to be “indispensable to all competitors which envisage manufacturing products that comply with the standard to which it is linked” as required by the Court.[358] For instance, even Google can be considered not to be “indispensable” for Internet search since Google evolves in a highly competitive environment.[359]

The DMA imposes FRAND access without exact consistency or relevance with the existing decisional practice of the EU competition law. Consequently, a two-tier regime may unfold—one under the DMA obligations, and the other under the traditional competition rules. It is unclear as to how this creates the “fair” competition the DMA aims to pursue.

FRAND Access to App Stores

The last obligation of Article 6 underlies the significant concerns the Commission has with the functioning of the app stores. Aside from Article 6(1)(c), the obligation laid down in Article 6(1)(k) imposes FRAND access for business users to app stores. The gatekeeper is indeed required to “apply fair and non-discriminatory general conditions of access for business users to its software application store designated pursuant to Article 3 of this Regulation.”[360]

FRAND access applied to digital platforms raises many questions. It incentivizes free riding while minimizing the previous and ongoing costs incurred to have the core platform services running for business users.

As discussed, FRAND access may be inappropriate to most digital services since the crucial criterion of indispensability is missing. Irrespective of these considerations, the DMA concludes stakeholders that complained about “high commission fees, unreasonable transfers of liability to the app developer with mutual liability being accepted by the platform operator, and the lack of notice given for technical changes in the app stores, which then requires the app to be amended in some cases resulting in lack of functionality.”[361]

These app developers' claims need to be rebalanced in light of the previous fees of 70 percent charged by telecom companies regarding the platform’s integrity and safety concerns. Despite the relatively minor complaints raised by app developers, the DMA imposes widespread FRAND access to app stores, with “fairness” assessed in light of the following characteristics:

[P]rices charged or conditions imposed for the same or similar services by other providers of software application stores; prices charged or conditions imposed by the provider of the software application store for different related or similar services or to different types of end users; prices charged or conditions imposed by the provider of the software application store for the same service in different geographic regions; prices charged or conditions imposed by the provider of the software application store for the same service the gatekeeper offers to itself. This obligation should not establish an access right and it should be without prejudice to the ability of providers of software application stores to take the required responsibility in the fight against illegal and unwanted content as set out in Regulation [Digital Services Act].[362]

According to the DMA, Apple’s App Store and Google’s Play Store have contestable market positions that are currently being contested. For instance, Indian entrepreneurs have coalesced to build “a large-scale platform that will host local apps and break the duopoly of Google's Play Store and Apple's App Store.”[363]

It is surprising that none of the European tech companies, be it alone or in coalition, are able to start up an alternative app store, whereas Indian tech entrepreneurs can. Thus, the market for app stores (should such a thing be a relevant product market) is contestable but uncontested in Europe by lack of entrepreneurship and ambition.

The current app stores are neither indispensable nor uncontestable, but at the moment constitute a duopoly, which may not prevent the environment from being competitive and contestable. Duopolies can exhibit considerable competitive constraints.

According to the different business models, app stores compete to attract apps according to Apple’s closed iOS and Google’s open Android OS. Furthermore, the market environment is highly competitive in terms of potential competition. App stores can be created as standalone services (e.g., the Indian initiative), but the app stores can be outcompeted by indirect entry. Indeed, the current duopoly emerged only because of the Windows Phone, BlackBerry, and other device manufacturers' failures.[364]

This reveals not so much that it is the app store “market” that can be competed against directly, but rather that smartphone devices may lead to contestability of the app store's duopoly. In other words, should Windows design a better smartphone, or should BlackBerry be successful again with its new products, these alternative ecosystems may easily design their own app stores, thereby throttling the current app stores' duopoly.[365] The DMA overlooks the fundamental nature of the potential indirect entry. From a static viewpoint, the DMA looks at the narrow app store market and wrongly deduces it to be uncontestable.

Also, there is no such thing as an app store market. The market remains the market for smartphones and mobile OSs. Should competitors design alternative smartphones with alternative mobile OSs that correspond to consumer preferences, consumers will choose these alternatives and new app stores may emerge. This trend is currently emerging.[366] However, the backward-looking perspective of the DMA precludes any forward-looking analysis of the reality of both the competitive constraints exerted and the contestability of app stores.

As a conclusion regarding the obligations of DMA’s Articles 5 and 6, it appears that the DMA’s additional regulatory burden represents a prevalence of precaution through regulation over innovation through disruption. Ex ante competition prohibitions prevent a rational, innovation-based analysis of the allegedly harmful conducts’ pro- and anticompetitive effects. The DMA takes unreasonable aims at weakening gatekeepers’ innovativeness of creating and entering markets.

The DMA may nevertheless prove to be most effective in removing digital gatekeepers and replacing them with regulatory walls wherever consumer prices increase, consumer quality decreases, and entrenched market positions' overall contestability diminishes rather than increases. The DMA may potentially harm gatekeepers—and certainly will replace them with walls for consumers and innovation.

The DMA imposes obligations to digital gatekeepers as if the Internet were the digital tabernacle—a sacred place to be removed from humanly exerted market power for the sake of a multitude of choices of operators, an absence of natural network effects, and a lack of implications of regulatory burdens on innovation incentives. The DMA, despite a blatant inability to have predicted their more-recent evolution, portrays a pretense of knowledge on how the digital landscape and practices will evolve over the next few years. These obligations thwart the European innovation economy, whereas it is in dire need of a jump-start.

Precautionary Antitrust in the DMA

The DMA embodies a transformational shift from ex post antitrust enforcement to ex ante regulatory compliance, albeit for a narrowly selected set of companies. Regulatory standards would replace evidence-based antitrust laws—a shift that departs from the traditional error-cost framework that dominated antitrust laws over the last decades. It downplays the analysis of balancing the cost of the intervention (i.e., false positives) against the cost of nonintervention (i.e., false negatives). And it avoids the balancing exercise inherent to the error-cost framework of antitrust enforcement.

The risk-averse atmosphere enticed by the DMA is perhaps what the European innovation economy needed the least. Following a techlash spurred by American Neo-Brandeisians and some European Ordoliberals, the DMA embodies a detrimental precautionary antitrust framework for many years to come.

Characteristics of Precautionary Antitrust in the DMA

The DMA illustrates the Commission's precautionary approach to competition and innovation. Indeed, after years of incremental decisions wherein an increasingly risk-averse stance permeated European competition enforcement, the DMA implicitly applies the core elements of the precautionary principle to EU competition rules.

The precautionary principle is a regulatory principle that runs counter to innovation. European institutions themselves have acknowledged the costs and innovation deterrence effects of this principle. Indeed, they have advocated for an “innovation principle” to limit the detrimental effects of the precautionary principle on introducing new products, processes, and business models—in short, in disrupting an economy in need of disruption, particularly in Europe.[367]

However, the EU antitrust practice continuously reinforces a risk-averse, non-innovation-based approach to competition matters. The DMA only magnifies the precautionary principle. The precautionary principle is costly as it exacerbates transaction costs, deters innovation, and puts the burden on companies to demonstrate the long-term effects of innovative conduct and products. Unfortunately, the DMA embraces the precautionary approach to regulating innovative and dynamic firms that engage in disruptive, unpredictable technologies. Indeed, each of the elements of the precautionary principle pervades the DMA.

Regulating Amid Uncertainties

The regulation of competition law in the digital sector occurs amid considerable uncertainties for regulators regarding how the regulatory obligations may affect the evolution of these fast-moving markets.

The DMA acknowledges the uncertainties regarding dynamics of the fast-moving innovation markets.[368] For instance, it acknowledges the “fast-moving and dynamic nature of digital markets” and the “dynamic nature of the platform economy.”[369] Also, the Inception Impact Assessment requires that “any solution should be future-proof, thus allowing the Commission to address novel issues is constantly evolving markets without introducing uncertainty in terms of its scope of application.”[370] Richard Gilbert neatly summed up such difficulty when he wrote that “creative destruction complicates predictions of market outcomes, but it does not make antitrust enforcement irrelevant or unnecessary.”[371]

The Commission elaborated speculations of risks in the digital economy amid considerable uncertainties (i.e., unknown unknowns).[372] The Commission was strongly advised to engage in cost-benefit analysis inherent to a case-by-case approach to interventions in the digital sectors of the economy.[373] The Commission now engages in the task of predicting consequential effects of interventions and consequential effects of noninterventions.[374] However, the Commission discounts market uncertainties in order to justify “timely intervention” instead of traditional antitrust enforcement.[375] These uncertainties constitute risks according to the European Commission’s precautionary approach. Market uncertainties allegedly carry two specific risks:

  • 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 anti-competitive 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).[376]

Against the rationale of innovation dynamics, which may justify consolidation and scale economies, the mere concentration of the market, as well as the growth of existing market players, thus constitutes risks that, from a precautionary perspective, needs to be addressed as early and as powerfully as possible.[377] The mere exhibition of some risks justifies interventions against designated gatekeepers.[378]

The DMA’s precautionary approach overlooks a cost-benefit analysis wherein risks, costs, and benefits are weighed in a rational, casuistic approach.[379] But the risks referred to in the precautionary approach differ from those risks used in cost-benefit analysis: They are not expected risks, but rather merely hypothetical risks and hypothetical damages.[380] The paradigm shift takes place amid uncertainties.[381]

As Commissioner Vestager has noted: “We can do investigations if it need be, so that we can see if gatekeepers are emerging and we can impose on them obligations. So, we can make sure that our intervention comes at the right place at the right time, before the market tips and a new gatekeeper emerges.”[382] Despite uncertainties, the precautionary approach ensures that regulation trumps innovation at the expense of the markets' dynamism and reasonable and non-negligible costs for consumers.

Consumer Choice as the New Precautionary Theory of Harm

Traditionally, under competition rules, intervention should occur only when there is (or likely to be) consumer harm.[383] However, the historical requirement of consumer harm for competition law to justify interventions has gradually but inexorably faded away in favor of another standard for intervention: consumer choice.[384]

Consumer choice is a dubious goal. Imagine two markets: one that provides consumers with a multitude of higher prices and lower quality products, and one that provides them with a small number of lower cost, higher quality products. The EU’s new standard would prefer the former, while virtually all consumers would prefer the latter. This is even more true given social science research showing that too much choice lowers consumer welfare.

A consumer choice standard is even more questionable given the considerable risk that it will be used to protect inefficient rivals that will claim they are simply providing a more choice.[385] Yet, under the DMA, conduct that merely reduces consumer choice may be found to be anticompetitive, irrespective of the possible inefficiencies of rivals or the benefits to consumers.[386] As Nazzini affirmed, “[W]hen consumer choice is seen as an objective in its own right, it may become a disguised form of competitor protection: a competitor deserves to be protected solely on the basis that it offers a differentiated product.”[387]

In this sense, the Commission’s goal to achieve “free choice for users in the digital sector” is likely to harm both the digital economy and consumer welfare. Just as there is limited consumer choice in many markets with significant fixed costs and scale economies (e.g., aerospace) how could there be “free choice” when the breadth of digital services is restricted by the massive capital investments companies must incur, and that such investments require commercial success? For instance, how could there be a free choice between app stores when consumers have consistently preferred and bought, over the last few years, smartphones with Android OS or iOS, thereby kicking out of the market’s less attractive phones such as from Nokia and BlackBerry? How free are consumers if they have the choice between, say, 10 cloud services providers instead of 8? Under the consumer choice standard, every conduct that may increase efficiency but reduce customer/consumer choice appears anticompetitive.[388]

Free choice for consumers is a chimera in markets. It proves incredibly misguided as a digital industry goal. Once essential to the competition law framework, efficiency becomes secondary to the nebulous concept of free choice.

Indeed, the DMA imposes obligations to gatekeepers no matter what their efficiencies.[389] Per the DMA, “Some of these providers exercise control over whole platform ecosystems in the digital economy and are structurally extremely difficult to challenge or contest by existing, or new market operators, irrespective of how innovative and efficient these may be.”.[390]

Smaller rivals can be less efficient in the digital economy since scale economies enticed by network effects are crucial to it. Indeed, would a search engine with fewer webpages crawled be considered as efficient as a comprehensive web crawler? Would advertisers consider a social media platform with fewer members to be as efficient as a smaller social media platform with a narrower audience? Size matters in the digital economy. Although it also may not, there is a strong probability that size may automatically generate efficiencies. Consequently, smaller rivals may be less efficient than larger incumbents. However, the DMA embraces a precautionary approach to a competition wherein smaller rivals become entitled to compete and access proprietary assets irrespective of their merits.

Thus, consumer choice as a new standard of competition law is, both implicitly and explicitly, present throughout the DMA. FRAND access best illustrates this regulatory trend wherein small rivals are granted strategic access to proprietary assets of the gatekeepers in order to give consumers a more extensive choice. FRAND access will have significant consequences in innovation deterrence and investment decline as the DMA legalizes free riding.[391]

The Commission appears to believe, wrongly, that increased consumer choice is associated with increased innovation, reporting that “intervention tackling … would both create the right innovation incentives and contribute to increased consumer choice paving the way for new platforms and innovative and privacy-friendly services.”[392] Nevertheless, the DMA overlooks the degradation of the innovation incentives for the gatekeepers. Innovation incentives for business users to access strategic assets by the gatekeepers may be equivalent to free riding.[393]

The lack of (consumer) harm reveals the precautionary nature of the shift from consumer welfare to choice. Indeed, the precautionary principle rests on the very remote, hypothetical possibility of harm for justifying interventions and regulations.[394] The shift from the need to evidence harm to the mere reduction of consumer/customer choice (irrespective of the efficiency and merits of such reduction) is akin to a precautionary approach to competition matters.

When the DMA talks about increasing consumer choice, what it is really saying is that it wants to increase producer diversity (i.e., the number of European producers).

An increase in consumer choice means an increase in the number of business participants irrespective of consumer preferences—and the Commission hopes and expects those business participants to be European. So in realty, when the DMA talks about increasing consumer choice, what it is really saying is that it wants to increase the diversity of producers—namely, with European companies.

In that regard, the DMA pursues economic deconcentration, regardless of its desire for innovation or consumer welfare. Economic deconcentration is consistent with Neo-Brandeisian and Ordoliberal theories, but is fundamentally at odds with the dynamics of digital markets.

Reversed Burden of Proof—Harmful Unless Proven Otherwise

Core to the precautionary principle is the belief that it is no longer necessary for regulators to justify their interventions. Instead, it is up to the market participants to justify not needing regulators to intervene.[395] This reversal of the burden of proof has long been envisaged by Commissioner Vestager for the regulation of competition in digital markets.[396]

The prospect of reversing the burden of proof against tech companies spurred a general outcry among commentators who legitimately considered such a proposal as a blow to innovation incentives and a violation of fundamental legal principles.[397] Nevertheless, the Crémer Report, instrumental in paving the way for the DMA, alluded to such reversal when it concluded that “some modifications of the established tests, including the allocation of the burden of proof and the definition of the standard of proof, may be called for.”[398]

In a different policy context, former IMF chief economist Olivier Blanchard, together with Leandro and Zettelmeyer, considered, “Simplicity is attractive, but not feasible. And even a complex rule is very unlikely to adequately capture the relevant contingencies, in part because many are impossible to predict ex ante.”[399] Jacques Crémer recently captured the essence of the precautionary approach inherent to the paradigm shift when he stated,

One of the problems with competition policy is that proving anti-competitiveness is extremely difficult. The opacity of platforms means that no one knows exactly how algorithms determine Amazon’s choice or Booking’s rankings. This problem can be eased by changing the burden of proof, as Yves-Alexandre de Montjoye, Heike Schweitzer, and I argued in our report for the European Commission. You could say that multihoming is illegal, for instance, unless the platform can show it’s pro-competitive.[400]

It is not only multi-homing that is presumed to be wrong irrespective of its consumer and innovation benefits, but many other practices would likely be included, such as the sharing or non-sharing of data, the raising or lowering of prices, changes in quality of service, etc.

There are many instances in which a company is unable to demonstrate the pro-competitive effects of its practices. It will likely follow that these practices, although innovative and disruptive, will ultimately be presumed to be illegal, even with tangible innovation benefits.

The regulatory approach to innovation with a dangerous shift of the burden of proof to the innovator epitomizes the precautionary principle. Both reluctance to endorse changes and the unmatchable requirement to demonstrate benefits of potential innovations will deter innovation.

Although the European institutions long acknowledged that “regulatory burdens are often perceived as a major obstacle to innovation,” the reversal of the burden of proof constitutes an additional regulatory burden.[401] Market participants must now show the lack of damage to competitors generated by their innovations, both in the short and long term. Would the microwave oven, for example, have ever been commercialized under such impossibly high evidentiary burdens? The deterrent effects are maximal, and gatekeepers will experience them.

The DMA involves a reversal of the burden of proof in several ways. First, it becomes an ex ante regulatory rulebook, implying that the burden of proof is reversed. Indeed, gatekeepers need to comply with the obligations of Articles 5 and 6 without the Commission having to engage in any prior endeavor.

Second, according to the reversed burden of proof, Article 8 may allow gatekeepers to seek suspension of these obligations, provided that they “demonstrate that compliance with that specific obligation would endanger, due to exceptional circumstances beyond the control of the gatekeeper, the economic viability of the operation of the gatekeeper in the Union, and only to the extent necessary to address such threat to its viability.”[402]

In line with the precautionary principle, the gatekeeper must demonstrate the need for the Commission not to enforce regulatory obligations and exempt it from such obligations.

Third, Article 9 allows gatekeepers to request to be exempted from the application of Articles 5 and 6 for “overriding reasons of public interest,” which include “public morality,” “public health,” and “public security.”[403] If gatekeepers can convincingly demonstrate that there are overriding reasons of public interest for obligations not to apply to them, the Commission may suspend the application of the relevant obligations. Again, the market participant bears the burden of proving public interest.

Moreover, companies that meet the quantitative indicators of Article 3(2) of the DMA are presumed to be gatekeepers.[404] The company presumed to be a gatekeeper must rebut it:

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.[405]

Article 3(4)2 states, “Where the gatekeeper presents such sufficiently substantiated arguments to demonstrate that it does not satisfy the requirements of paragraph 1, the Commission shall apply paragraph 6 to assess whether the criteria in paragraph 1 are met.”[406] Article 3(6) only lists qualitative indicators that may lead the Commission to designate the company as a gatekeeper.

Consequently, the targeted companies are designated as gatekeepers (through qualitative and/or quantitative indicators) unless they provide “sufficiently substantiated arguments” to prove differently. The threshold seems to be relatively high, and the probability that the company ends up being designated as a gatekeeper remains equally high. Consequently, the burden of proof is reversed at the expense of the innovation incentives and against law principles.

Interim Measures and the Assumption of Irreversibility

The precautionary principle is about regulatory interventions that take place in an early and timely fashion. The arguments invoked for such early regulatory intervention for competition, as opposed to the traditional administrative and judicial investigations, are twofold: i) interventions must quickly occur, otherwise the damage may become irreversible and irreparable in a language directly borrowed from the precautionary principle; and ii) traditional law enforcement processes are too time consuming and do not correspond to the market’s rapidly changing environment. These two arguments are unsubstantiated and contradictory.

Interim measures of Article 22 of the DMA have been explicitly referred to as being “precautionary measures.” Indeed, interim measures are explicitly “precautionary” in their nature. Commissioner Vestager made this explicit in her answer to a member of the European Parliament (MEP) on July 5, 2017. Spanish MEP Ramon Luis Valcarcel Siso, on the heels of the 2017 Google Shopping decision, asked, 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.[407] The answer soon materialized with interim measures being adopted (in the Broadcom case) for the first time in 20 years of the European Commission’s decisional practice.[408]

Also, the Motta Report written ahead of the DMA proposal, and requested by the Commission as part of the public consultation on the market investigation rules envisaged, explicitly recognized the inspiration of the precautionary principle to antitrust matters: “Conceivably, an NCT investigation might allow intervention even without proving that the conduct is abusive: quite simply, if it is thought that the adverse (dynamic) effect on competition is sufficiently high, then by applying a sort of precautionary principle the conduct could be discontinued.” (emphasis in original)[409]

Therefore, interim measures inherently embody the precautionary principle. Interim measures have been dormant instruments of EU competition policy for many years.[410] Article 8 of the Council Regulation (EC) No. 1/2003 of December 16, 2002, provides for interim measures. It states that "in cases of urgency due to the risk of serious and irreparable damage to competition, the Commission, acting on its initiative, may be a decision, on the basis of a prima facie finding of infringement, order interim measures" (Article 8(1)). The interim measures decision can be renewed if deemed necessary and appropriate (Article 8(2)). The Regulation also acknowledges the member states' competition authorities to order interim measures when applying EU competition rules (Article 5).

Precautionary measures adopted under interim rules are justified on the basis of urgency of an irreversible harm: “In case of urgency due to the risk of serious and irreparable 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.”[411]

The notion of a “risk of serious and irreparable damage for business users or end users of gatekeepers” borrows from precautionary rhetoric and raises many questions as per the probability that innovation will be stifled and consumers harmed.

It is not the materiality of irreparable damage but the mere risk of it that justifies urgent intervention. What are “risks” deemed to potentially lead to irreparable damage to companies? Aggressive competition may harm less efficient rivals and innovation laggards, while unresponsive competitors seem to generate such risks. Disruptive innovations inevitably create risks for non-innovative rivals of being displaced by competition.

With precautionary reasoning, innovation is irreversibly damaged and consumers are irreparably harmed. This precautionary approach aims at preserving the status quo and protecting competitors—not competition.

In other words, the mere risk of irreversible harm constitutes a formidable legal basis for less efficient rivals and less innovative competitors to delay, if not halt, the introduction of new products and services by gatekeepers. The risk of serious and irreparable damage represents the best venue for rent-seeking behaviors by sluggish competitors at the expense of digital technologies' dynamism and disruptive nature.

Second, the notion of “serious and irreparable damage” not only suggests that some damage from competitive rivalry may be reparable or reversible—an illusory consideration—but also suggest that whenever serious and irreparable damage is expected to occur, the regulator’s intervention in the market becomes necessary and desirable. This notion of “damage” is all the more contestable since it may not need to materialize in order to justify intervention. The mere fact that the market may irreversibly tip justifies intervention.[412] Indeed, Recital 26 of the DMA makes clear that the mere possibility of tipping justifies quick interventions.[413]

Thus, some companies' success leading to an oligopolistic structure of the market shall be a sufficient basis for intervention, absent any damage. The mere fact that an alleged dominant firm integrates databases from two different markets is said to have “long-term irreparable effects to the competition.”[414] In other words, the reduction in the number of firms, irrespective of the functioning of the competitive process, may constitute grounds for claims according to which a market is under the risk of tipping, therefore threatening to generate irreversible harm to competitors and justifying interventions.

The European Commission’s vision of atomized markets, regardless of the network effects and the nature of the digital competition, entails re-designing oligopolistic markets. This Ordoliberal vision of atomized markets is both passé and inappropriate for digital markets.

Third and finally, the whole rationale for “early and timely” intervention by regulation rather than traditional administrative and judicial processes misses a fundamental point. The time necessary for investigative measures and judicial review is a core element of the rule-of-law principle and helps regulators and judges make better decisions, enlightened by time and retrospect.

Instead of the calm, evidenced-based, retrospective nature of traditional competition law enforcement, interim measures (and, to a lesser extent, market investigation rules) freeze a market situation with little evidence and considerable impact on innovation—it is the status quo preserved. Equally, the DMA's market investigation rules include a precautionary dimension—although measures may not be automatically adopted following market investigations.[415] Proponents of interim measures acknowledge their swiftness and expedited nature.[416]

There is a fundamental contradiction between, on the one hand, the recognition of digital sectors being fast moving, highly uncertain, and unpredictable, and on the other hand, the burning desire to intervene as early and as powerfully as possible to prevent unforeseen consequences and damage from materializing in a quest for a uncertain benefits.

In conclusion, the DMA portrays the fundamental characteristics of the precautionary principle. It entails ex ante intervention amid uncertainties, reverses the burden of proof so that companies have to justify why they do not qualify for the regulatory obligations, and preserves the status quo against irreversible changes inherent to disruptive and innovative practices. The DMA favors precaution over innovation, engrains a static perspective to a highly dynamic competition process, and, finally, deters disruptive innovation at the expense of consumer benefits. The DMA fossilizes, rather than jump-starts, digital competition, despite a much-awaited thriving and dazzling European innovation economy. The DMA embodies precautionary antitrust, although it should have propelled a dynamic approach to antitrust concerns—or “dynamic antitrust.”

Innovation Concerns Are Absent

A dynamic approach to antitrust is the opposite of a precautionary approach. In other words, dynamic antitrust fosters innovation and long-term analysis of market dynamics. It is size-neutral in the sense that big companies are not necessarily bad. Antitrust analysis in a dynamic approach derives from desirable objectives, such as scale economies, global competitiveness for industrial champions, robustness to experience market shocks, greater innovativeness, and greater social and economic protection for stakeholders.

Also, a dynamic approach to antitrust would minimize the threshold effects of regulations—namely, categorizing companies according to some discrete parameters (such as size) to incentivize growth and minimize deterring barriers to expansion. The essential intellectual perspective of dynamic antitrust lies in the fact that, beyond competition as a source of innovation, it most importantly and dramatically is innovation that constitutes a source of competition. Companies compete through innovation, particularly concerning digital technologies.

Unfortunately, the DMA seems ignorant of the considerable amount of literature and knowledge of the need to incentivize innovation as a source of competition rather than merely designing competitive market structures that are expected to lead to innovation outcomes. Companies escape competition through innovation, and enter into competition with other companies through innovation.[417]

Finally, a dynamic approach to antitrust issues would better consider the need to preserve and incentivize the building of “dynamic capabilities” by companies.[418] The multiple IP rights present in digital technologies inevitably amount to some exclusion only for the benefit of future competition and rapid innovation. Also, the sensing of opportunities and the improvement of complementarities may justify leveraging practices that disrupt adjacent markets, thereby providing stronger competition. A dynamic approach to antirust would ensure that firms’ dynamic capabilities allow for stronger competition and the necessary asset orchestration for innovation to emerge. A dynamic approach to antirust aims to strengthen firms’ dynamic capabilities by enabling firms to compete through innovation. It avoids weakening these dynamic capabilities for market structure considerations.

Unfortunately, these fundamental features of dynamic antitrust are notoriously absent in the DMA.

The Assault on Scale Economies

The populist techlash generated a resurgence of the “big is bad” motto, according to which large companies are inherently detrimental to economic growth and consumers.[419] Unfortunately, the DMA falls prey to this popular yet unsubstantiated claim that small companies can achieve more desirable outcomes than bigger ones. It is thus believed that larger companies’ conduct is motivated mainly by monopolization rather than by innovation.

The first paragraph of the DMA reveals the fundamental bias of the proposal against large tech companies and the associated positive bias toward smaller companies. It states that “whereas over 10,000 online platforms operate in Europe’s digital economy, most of which are SMEs, a small number of large online platforms capture the biggest share of the overall value generated.”[420] The Commission does play small platforms against large platforms with the “gatekeeper instrument” the DMA represents.[421]

An overdue structuralist approach at the expense of a behavioral, case-by-case approach prevails in the DMA. The structuralist approach enticing a concern for “large platforms” appears in the next paragraph:

Large platforms have emerged benefitting from characteristics of the sector such as strong network effects, often embedded in their own platform ecosystems, and these platforms represent key structuring elements of today’s digital economy, intermediating the majority of transactions between end users and business users. Many of these undertakings are also comprehensively tracking and profiling end users. A few large platforms increasingly act as gateways or gatekeepers between business users and end users and enjoy an entrenched and durable position, often as a result of the creation of conglomerate ecosystems around their core platform services, which reinforces existing entry barriers.[422]

The DMA overlooks entirely the fact that the greatest hindrance to the completion of the Digital Single Market are national regulatory barriers that impede firms’ scalability and deter market contestability.[423] Instead of identifying the real causes of the Digital Single Market's barriers, the DMA represents an assault on large tech companies designated as gatekeepers, mostly based on size indicators.

Box 9: Breakups Under the DMA—Article 16

The DMA explicitly envisages structural remedies (i.e., break-ups) against large companies. According to the European Commission, this would take place whenever both a gatekeeper engages in “systematic noncompliance” with DMA obligations and the gatekeeper’s size increases too much.

The DMA explicitly targets gatekeepers’ growth and potential noncompliance with DMA obligations as sufficient grounds for breaking tech companies up into pieces according to a yet clarified method.

Recital 64 of the DMA specifies this threat of structural remedies plainly:

The Commission should investigate and assess whether additional behavioural, or, where appropriate, structural remedies are justified, in order to ensure that the gatekeeper cannot frustrate the objectives of this Regulation by systematic noncompliance with one or several of the obligations laid down in this Regulation, which has further strengthened its gatekeeper position. This would be the case if the gatekeeper’s size in the internal market has further increased, economic dependency of business users and end users on the gatekeeper’s core platform services has further strengthened as their number has further increased and the gatekeeper benefits from increased entrenchment of its position.[424]

This worrying objective enshrined in Article 16 of the DMA states that “systemic noncompliance” with DMA obligations—defined as three noncompliance decisions over a five-year span—will lead to the breakup of the company if it has increased in size.[425] Thus, systemic noncompliance without strengthening and expanding the gatekeeper’s market position cannot lead to breakups. Consequently, one can legitimately induce that the growth in the size of the gatekeeper, following systemic noncompliance is the cause of disagreement, thereby triggering structural remedies against the platform.

In other words, more than systemic noncompliance, breakups of tech companies will be possible based on the size increase of the gatekeeper. Indeed, Article 16(4) of the DMA clearly expresses the assault against further expansion of the gatekeeper’s size in the following terms:

A gatekeeper shall be deemed to have further strengthened or extended its gatekeeper position in relation to the characteristics under Article 3(1), where its impact on the internal market has further increased, its importance as a gateway for business users to reach end users has further increased, or the gatekeeper enjoys a further entrenched and durable position in its operations.[426]

The assault on companies’ size pervades the entire DMA. Indeed, the sheer bigness of undertakings becomes suspicious, regardless of the merits accrued from such bigness.

The DMA entails a lose-lose situation. In contrast, clear pathways to complete the Digital Single Markets (e.g., removal of national regulatory barriers, enhancing European SMEs’ dynamic capabilities, fostering the European knowledge economy, etc.) represent win-win solutions.

Overlooking the economic evidence that favors large platforms, the DMA reveals a bias toward European SMEs. The DMA takes aims at large platforms, enhancing the small companies by targeting the larger ones, which may prove to be a lose-lose situation. The SMEs will not experience substantial benefits since their inherent disadvantages (i.e., limited scale) will not fade away—and larger platforms may be hurt considerably without tangible benefits for consumers.[427]

Contrary to the DMA’s precautionary approach, a dynamic approach to antitrust matters would be size neutral and focused on firms’ conduct rather than on market structures and firms’ designations. Unfortunately, the DMA explicitly rejects such a desirable dynamic approach.

Static Approach for a Dynamic World

The dynamic approach to antitrust matters diverges from the DMA’s inherently static approach in its appraisal of markets' functioning, the working of the competitive process, and its relationship with innovation.

One may argue that the DMA encapsulates dynamic elements. For instance, Article 4 indeed allows for the review of the status of gatekeepers—which is necessary given the evolutions of the market positions.[428] Article 4(2) imposes the Commission to review the status of gatekeepers “at least every 2 years.”[429] Similarly, Articles 8 and 9 of the DMA provide for the necessary suspension and exemption of the DMA’s obligations subject to convincing arguments.

Article 10 allows the Commission to update the obligations for gatekeepers should “the need for new obligations addressing practices” emerge.[430] Interestingly, the DMA considers that such “introduction of the dynamic updating of unfair practices would be subject to ensuring a full respect for the fundamental rights to fair proceedings and good administration as enshrined in the [European Convention on Human Rights], which are binding on the EU institutions.”[431] Market investigations may also add new services to the range of core platform services subject to the DMA—new practices deemed to be “unfair.”[432] The DMA describes this possibility as “a dynamic mechanism allowing to update the list of obligations in case new practices are deemed unfair after a market investigation.”[433]

However, these marginal elements of the “dynamic” assessment of the competition in the digital sector may not be conflated with a dynamic approach to antitrust concerns. They are the basis for updates (or, often, expansion) of the DMA’s obligations to address the evolution of markets. It does not necessarily mean that the analysis will not be a static analysis that almost exclusively assesses competitive constraints as those exerted within the current market players. The DMA mostly discards competitive constraints exerted by potential competition (i.e., potential entrants, substitutable supply, or both).

Indeed, the DMA makes no mention of the fundamental notion of “potential competition.” Potential entry constitutes competitive constraints by plausible rivals. These constraints are fully ignored in the DMA.[434] In fact, one of the few references to potential competitors is made concerning the technical definition of core platform services.[435] Another reference, equally made for describing market tipping, is that “once a service provider has obtained a certain advantage over rivals or potential challengers in terms of scale or intermediation power, its position may become unassailable and the situation may evolve to the point that it is likely to become durable and entrenched in the near future.”[436]

More precisely, mere references of “potential entrants” are made regarding the right of these entrants to access the gatekeepers’ core platform services.[437] In other words, no dynamic competition consideration in terms of innovation incentives is palatable in the DMA. This state of affairs contradicts the needs identified by both the Commission itself and the economic literature.[438]

Many examples illustrate the lack of dynamic competition perspective—or dynamic antitrust. For instance, the multiple definitions, categorizations, and discriminatory rules enshrined in the DMA preclude a dynamic, holistic approach according to which the anticompetitive effects of conduct are assessed irrespective of the designation’s definitions of the subject or the practice.

Not only does the DMA questionably define the digital market (as opposed to traditional markets) and discriminate between gatekeepers and other rivals, but it also lists the digital services subject to the DMA. Pursuant to Article 2(2), such a listing appears extremely narrow, is contingent on today’s realities, and, presumably, will soon be outdated with new services that may overlap multiple categories or not fit into any of these categories.

Although new services can be added to the list according to Article 17, such a listing precludes a dynamic approach that’s focused on effects rather than legal categorizations. This listing is prone to error and has little relevance to an appropriate antitrust enforcement that ensures fair competition, as the DMA claims to pursue.

One prominent example of such inappropriateness is the distinction between “online search engines” (in passing, what are “off-line search engines”?) and advertising services. Search engines live, develop, and flourish only because they are mostly ad funded. One cannot imagine an online search engine with a pay-per-query business model. Consequently, there is no market for search engines not tied up with a broader advertising market. The exaggeratedly subtle distinctions in the DMA illustrate the lack of dynamic perspective aimed at fully considering the breadth and complexity of the competitive constraints exerted on tech companies.

Also, it is commonly understood that a dynamic approach to competition would shift the focus from being overly price-centric to becoming more innovation-centric, in the words of Prof. Richard Gilbert:

Antitrust enforcement should evolve from being price-centric to innovation-centric. Price-centric antitrust enforcement prevents mergers that are likely to raise prices and prevents firm conduct that excludes competition for existing products and services. Innovation-centric antitrust enforcement does not abandon these concerns, but it augments them by challenging merger and firm conduct that are likely to harm innovation and competition for products that do not presently exist. Innovation-centric competition policy will achieve goals that price-centric enforcement neglects, such as ensuring opportunities for entrepreneurs to compete and thrive.[439]

Contrary to such an innovation-centric approach, the DMA focuses on prices, assuming that higher prices represent monopoly power, without considering the possibility of increased quality or product innovation: “Unfair practices and lack of contestability lead to inefficient outcomes in the digital sector in terms of higher prices, lower quality, as well as less choice and innovation to the detriment of European consumers.”[440]

The DMA allegedly “helps businesses overcome the barriers stemming from market failures or gatekeepers’ unfair business practices. This will help to foster the emergence of alternative platforms, which could deliver high-quality, innovative products and services at affordable prices.”[441] Contrary to economic literature, which sees pro-competitive advantages in price discrimination, the DMA blatantly prohibits price discrimination for gatekeepers:

To ensure that business users of online intermediation services of gatekeepers can freely choose alternative online intermediation services and differentiate the conditions under which they offer their products or services to their end users, it should not be accepted that gatekeepers limit business users from choosing to differentiate commercial conditions, including price.[442]

The DMA assumes that “higher costs [of online advertising] are likely to be reflected in the prices that end users pay for many daily products and services relying on the use of online advertising.”[443] The overall goal of the DMA is mainly to decrease consumer prices and increase consumer choice. Indeed, the expected impact of the DMA is outlined in the DMA’s legislative financial statement: “Interventions aiming at increasing the contestability of the digital sector would have a significant positive and growing contribution to achieve all of the potential benefits of a Digital Single Market, also resulting in lower prices and greater consumer choice, productivity gains and innovation.”[444]

It derives from the DMA that prices for products and services in digital industries are assumed to be high and that the DMA may lower them. This contradicts the ad-funded business model of a wide range of platforms, and the lower prices offered by digital platforms as opposed to traditional businesses.

Moreover, the DMA’s assumption concerning prices overlooks an increase in prices due to increased quality (i.e., increased costs) or increased innovation (i.e., recoupment costs after sunk costs). However, both justifications for price increases are at the center of the innovation-centric antitrust enforcement Prof. Gilbert called for.

Price increases for innovation purposes are the source of subsequent effective competition in the market. The European Commission’s DMA ignores the foremost important element of dynamic competition, which is not to assume that price increases are anticompetitive behaviors in highly innovative industries.

Dynamic Capabilities Discarded

To ensure that firms innovate as a source of effective competition, antitrust regulators must support, rather than reject, innovation incentives. Firms’ capabilities that constitute the engine for future disruptive innovation are commonly referred to as “dynamic capabilities.”[445] Dynamic capabilities differ from ordinary capabilities concerning innovation:

If an enterprise possesses resource/competences but lacks dynamic capabilities, it has a chance to make a competitive return for a short period, but superior returns cannot be sustained. It may earn Ricardian (quasi-)rents, but such quasi-rents will be competed away, often rather quickly. It cannot earn Schumpeterian rents because it hasn’t built the capacity to be continually innovative. Nor is it likely to be able to earn monopoly (Porterian) rents since these require exclusive behavior or strategic manipulation.[446]

Embedded in the Schumpeterian perspective, dynamic capabilities are the knowledge capacity derived from scarce resource accumulation.[447] Consequently, for firms to innovate and compete with rivals, they need to constantly build dynamic capabilities. Therefore, contrary to the prevailing paradigm embraced by Ordoliberals and Neo-Brandeisians implicit in the DMA, it is not perfect competition that drives innovation; rather, it is imperfect competition. And innovation stemming from imperfect competition can be source of competition, a concept that regrettably appears absent in the DMA.

Indeed, what is Article A5(4)’s notion of “foreseeable” gatekeeper if not a company presumptively building dynamic capabilities through innovation? Article 15(4) of the DMA assumes a gatekeeper is a company enjoying monopoly rents, rather than a company aggressively competing to innovate in order to put itself into a respectable market position. The distinction between gatekeepers and foreseeable gatekeepers is unclear—and runs the risk of regulating and prohibiting conduct by firms developing their dynamic capabilities through strategic resources.

These strategic resources are subject to the DMA’s regulations and prohibitions. The notion of “unavoidable trading partner” illustrates the disregard of the DMA for dynamic capabilities. Suppose tech companies failed to compete with incumbent firms. They may try to invest heavily in acquiring resources key to digital industries' operation to leverage some dynamic capabilities essential to compete against incumbents in adjacent markets from this niche market position.

This niche market position may lead to an intermediation market position conducive to the designation of the company as gatekeeper. This is explicitly envisaged in the DMA:

[O]nce a service provider has obtained a certain advantage over rivals or potential challengers in terms of scale or intermediation power, its position may become unassailable and the situation may evolve to the point that it is likely to become durable and entrenched in the near future. Undertakings can try to induce this tipping and emerge as a 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.[448]

Suppose a tech company wants to compete with a search engine. It may very well start with an indirect entry into the market for cloud services and develop traditional computer processing and data-center capabilities. Following the access to a large amount of data, it may learn and invest in human capital to use these cloud computing services to crawl the web with algorithms leveraged from its cloud services toward web-crawling activities.

Before tech companies are able to reach the necessary scalability and build the necessary learning processes specific to search engines, the DMA may come to the fore and put the tech companies' cloud services under the DMA obligations for third-party access, nondiscrimination, the prohibition of leveraging, etc. Indeed, Article 3 provides for the designation of cloud services companies. For example, a company will not be able to exert effective competitive constraints onto the search engine, as it would implicitly have to focus its business on cloud services. The company would also be more concerned with regulatory compliance of its cloud services concerning its business users' rights under the DMA, rather than expanding.

This sort of misconceived competitive assessment will keep repeating as long as dynamic capabilities are absent from the regulator’s concerns for building up companies' innovation capacities to compete.

The appropriability problem is crucial for firms’ ability to capture value from innovation—in other words, to profit from innovation.[449] IP rights solve appropriability problems (e.g., patents, copyrights, trade secrets, trademarks, etc.). But innovation in the marketplace, even in technology markets wherein IP rights are ubiquitous, does not always allow for full appropriability—or for a full capture of the innovation.

The discrepancy between innovation investments and returns from these investments stifles innovation incentives and creates the free-riding problem. Property rights help avoid or minimize the cost of free riding by rivals in their competitors’ assets and innovation. The DMA, however, enables free riding.

FRAND accesses of Article 6 distort appropriability and the ability of firms to capture value from innovation and investments made in the allegedly “essential” assets. Why would business users invest in creating a competing app store if the DMA entitles them to free ride and be treated equally with the app store’s owner? In reverse, why would the owner keep investing and innovating in an app store that may yield considerable benefits for its downstream rivals at the expense of its appropriability capacity? The app store’s owner may seek alternative channels to reach end users and stop investing in the app store, at the expense of business users.

Why would a platform keep on providing free news aggregation services to derive no profits if they cannot appropriate value from these services, such as consumer satisfaction for the overall platform? The news aggregation services may be discontinued in order for the platform to focus on its most profitable services—thereby distorting the incentives for the company to develop capabilities irrespective of their capacity to generate extra profitability.

Also, the prohibition on default setting represents a considerable weakening of companies’ ability to develop opportunities using a cross-subsidizing services that enable free/low-cost services to be provided in exchange for more profitable services to be made readily available to end users. The prohibition on default setting is a formidable blow to the ability of firms to develop synergies so that they may indirectly enter and compete with rivals. For instance, the prohibition of default-setting cannot exclude the prohibition on Microsoft to have Bing as the default search engine on every Windows PCs—thereby weakening Microsoft’s dynamic ability to learn and sense opportunities in order to compete with Google’s search engine.

Equally, the prohibition on default-setting cannot exclude the prohibition of to have a selected range of rental car companies listed for end users to pick from when they book their reservations. Indeed, why would another rental car company be prevented from being listed on the app? The prohibition of default-setting may be far-reaching but will undoubtedly overlook firms’ incremental building of dynamic capabilities. In the case of, selecting a small set of partners for rental cars may enable it to improve its knowledge and learning process concerning customer satisfaction and work on necessary innovations. Regardless of these considerations, rival rental car companies may successfully claim their rights of access as business users under the DMA.

The DMA’s pursuit for free choice, for business users’ entitlement to have nondiscriminatory access to core platform services irrespective of their customer patterns, and the DMA’s tendency to overlook cross-subsidization across platforms’ services to offer free/low prices, play a part in the overall neglect of dynamic capabilities in the European Commission’s proposal.

Such neglect will inevitably distort competition by undermining innovation incentives and increasing costs under the DMA’s overall static approach to the dynamics of digital competition.


This report discussed the most problematic aspects of the DMA. Some other aspects that have not been covered represent either no real concern or concerns of a lesser extent. For instance, Article 12 of the DMA imposes an obligation for gatekeepers to inform the Commission about concentration. Contrary to a direct prohibition of mergers, this proposal appears quite sensible and proportionate. Also, the market investigation rules in Chapter IV of the DMA are justified as long as they help the Commission gather evidence and engage in fact-finding exercises.

However, the Digital Markets Advisory Committee, created with Article 32 of the DMA and composed of representatives of EU member states, constitutes a worrying politicization of the competition enforcement process. How could the bigger member states not have a more influential voice in such a committee? How can such a committee avoid politically-driven competition decisions with a real risk of protectionism? This committee reintroduces politicization over competition law issues at a time when antitrust populism is least needed.

Overall, the DMA materializes a paradigm shift from ex post antitrust enforcement toward ex ante regulatory compliance. In achieving such a shift, the DMA embodies the elements of the precautionary principle, thereby making the DMA a prime instance of precautionary antitrust.

Precautionary antitrust clashes with innovation-based antitrust wherever evidence-based, innovation-centric analyses are made on a case-by-case basis—in other words, “dynamic antitrust.”

The Commission could improve the proposal in the following ways:

  • Restore a fair level playing field with a reform of competition law applicable to all firms, not only those operating in digital markets.
  • Eliminate the convoluted classification of “gatekeepers,” which creates threshold and entrenchment effects, and will inevitably lead to endless legal disputes against the DMA’s stated objectives.
  • Develop market investigation rules with capacity-building by the European Commission, with staff resources expanded in order to take evidence-based, fact-finding exercises seriously.
  • Ensure that the DG-Comp in charge of market investigation rules is not in charge of antitrust enforcement in order to avoid a conflict of interest and confirmation bias. A new team insulated from the DG-Comp needs to be created.
  • Recognize the need to analyze competition issues dynamically with an explicit focus on longer-term analysis and providing firms with the ability to justify their conduct thanks to a generalized rule of reason.


[1]Margrethe Vestager, “Building trust in technology,” Speech at EPC Webinar, Digital Clearinghouse, October 29, 2020, wherein Vice President Vestager argued that “the Digital services Act, which will update the E-Commerce Directive, and require digital services to take more responsibility for dealing with illegal content and dangerous products.”

[2]Margrethe Vestager, “A Europe of Rights and Values in the Digital Decade,” Speech at the EU Council’s High-level Presidency Conference, December 8, 2020 (noting that “the traffic light was the centerpiece in a new system of rules and devices which brought order to the chaos ... But we have also seen that as the digital traffic increases, things can become chaotic, just like on busy city streets”). This analogy is irrelevant because car crashes and pedestrians being killed are real in road traffic, whereas in digital traffic the only risk is the markets being tipped with light; and also because red lights become green and vice versa, while regulatory prohibitions may never become regulatory green lights. Finally, road traffic lights are applicable to all cars and drivers irrespective of their size or wealth, unlike digital traffic lights which are applicable only to a well-identified set of digital players.

[3]See Article 3 of the Regulation 2016/679 of the European Parliament and of the Council of April 27, 2016, on the protection of natural persons about the processing of personal data and the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation). More specifically, para.101 of the GDPR’s Preamble states that “when personal data are transferred from the Union to controllers, processors or other recipients in third countries or to international organizations, the level of protection of natural persons ensured in the Union by this Regulation should not be undermined, including in cases of onward transfers of personal data from the third country or international organization to controllers, processors in the same or another third country or international organization. In any event, transfers to third countries and international organizations may only be conducted in full compliance with this Regulation. A transfer could take place only if, subject to the other provisions of this Regulation, the conditions laid down in the provisions of this Regulation relating to the transfer of personal data to third countries or international organizations are complied with by the controller or processor.” The wider scope of the GDPR was explicitly detailed by the Court of Justice in the Google LLC v CNIL case wherein the Court had to assess the extraterritorial effects of the Directive applicable at the time of the ruling. See C-507/17, Google LLC v Commission nationale de l’informatique et des libertés (CNIL), ECLI: EU :C :2019 :772, para.61 (“while the EU legislature has, in Article 17(3)(a) of Regulation 2016/679, struck a balance between that right and that freedom so far as the Union is concerned ... it must be found that, by contrast, it has not, to date, struck such a balance as regards the scope of a de-referencing outside the Union”). On the extraterritorial effects of the GDPR, see Gert Vermeulen, Eva Lievens (Eds.) “Data Protection and Privacy under Pressure. Transatlantic tensions, EU surveillance, and big data,” Antwerp: Maklu (2017); Paul de Hert, Michal Czerniawski, "Expanding the European data protection scope beyond territory: Article 3 of the General Data Protection Regulation in its wider context, 6 International Data Privacy Law 3, 230–243 (2016),  (considering that the GDPR has wider extraterritorial effects than any previous EU data protection rules); Benjamin Greze, “The extra-territorial enforcement of the GDPR: a genuine issue and the quest for alternatives,” 9 International Data Privacy Law 2, 109–128  (discussing extra-territorial voluntary compliance or enforcement actions of the GDPR rules); Adele Azzi, “The Challenges Faced by the Extraterritorial Scope of the General Data Protection Regulation,” 9 Journal of Intellectual Property Information Technology & Electronic Communications Law, 126–137 (which considers that the GDPR “amounts to a unilateral expansion of the application of European law to non-EU businesses”).

[4]Margrethe Vestager, “Building trust in technology.”

[5]Damien Gerardin, “The EU Digital Markets Act in ten points,” The Platform Law Blog, December 16, 2020  (noting that the “proposed obligations could have far reaching consequences”).

[6]Government of Ireland, The Irish Position on the EU Commission’s proposed Digital Services Act package—Submission to the Public Consultation (Dublin: September, 2020), See European Commission, Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20, June 2019, on promoting fairness and transparency for business users of online intermediation services, L.186/57, July 11, 2019.

[7]Ibid, 6, “The many Member States, including Ireland, blocked attempts to include ‘blacklisted’ practices without merit and evidence of economic harm during the negotiations on the Platforms to Business Regulation. The evidence base required to prohibit, per se, certain practices did not exist in Q1 2019, and Ireland remains to be convinced that this evidence base exists now.”

[8]Digital Markets Act, recital 79 (Brussels: European Commission, December 15, 2020),

[9]Margrethe Vestager, “Building trust in technology.”


[11]The “new competition tool” was a European Commission’s initiative, now defunct, which aimed at providing the European Commission with antitrust tools to leave ex post antitrust liability engagement in order to intervene ex ante. On June 2, 2020, the European Commission indeed published its initial impact assessment and a public consultation on “new competition tools.” The underlying idea is that Article 101 and 102 TFEU do not properly work to address “structural problems with competition, when we can’t do that with the powers that we already have—or when those existing powers would be much less effective.” See Margrethe Vestager, Competition in a Digital Age: Changing Environment for Changing Times, ASCOLA Annual Conference, June 26, 2020 (the purpose of which is “not by fining companies or finding them guilty of breaking the rules—but by imposing obligations that would protect competition. That could involve duties to behave in a certain way—to make data available to others, for instance. As a last resort, it could even mean breaking up companies, to protect competition”). These alleged structural problems are of two sorts: “structural risks for competition” wherein gatekeepers have tipped markets so that it is assumed that new entrants cannot emerge, and “structural lack of competition” wherein markets fail structurally due to high concentration and/or risks of tacit collusions. See Margrethe Vestager, Statement to the US Committee on the Judiciary, Subcommittee on Antitrust, Commercial and Administrative Law, United States House of Representatives, July 30, 2020. See also Samuel Stolton, “Commission charts new competition ‘fit for the digital age’,” Euractiv, June 12, 2020  (quoting the inception impact assessment which expects this tool to facilitate “timely and effective intervention against structural competition problems across markets”); Marie Cartapanis and Frederic Marty, “Towards New Tools in Competition Law: Some legal and economic considerations,” Competition Forum: Law & Economics  (November, 2020),; Council of Bars and Law Societies of Europe, CCBE comments for a ‘New Competition Tool’ as part of a future legislation to be proposed by the European Commission,” October 14, 2020  (considering that such tool “would drastically change the nature of competition law instruments and equilibrium of powers, while lowering the standard of proof and putting at risk in a disproportionate manner parties’ rights of defense. The CCBE does not believe the Commission requires a tool which goes beyond its existing ability to perform sector inquiries”); Motta and Peitz, “Intervention triggers and underlying theories of harm” (considering such a tool may address market malfunctions and also address “markets which are not currently experiencing problems, but which for different reasons ... may be at risk soon. In such cases, it is conceivable that the NCT might provide a preventive tool of intervention that is currently not available under EU competition law.” The authors might have more accurately referred to “precautionary” tool (as per the precautionary principle)).

Digital Markets Act, Article 15. Although no longer referred to as “new competition tool,” the obligations and prohibitions (“do’s and don’ts”) resemble such competition tool. On the extent to which the DMA’s obligations and prohibitions are tantamount to the once imagined “new competition tool,” we certainly need more time and a final text to appraise more clearly.

[12]Billy Perrigo, “How the E. U’s Sweeping New Regulations Against Big Tech Could Have an Impact Beyond Europe,” Time, December 15, 2020 (considering that “along with fines of up to 10% of global turnover for violations of antitrust law, the European Commission threatens that it could break up the businesses of repeat offenders—a step further than anything the incoming Biden Administration has pledged to do”).

[13]Aurelien Portuese, “European Competition Enforcement and the Digital Economy: The Birthplace of Precautionary Antitrust,” in Global Antitrust Institute (Ed.), Report on the Digital Economy (Arlington, VA: Global Antitrust Institute, 2020), 597–651, 651 (“the new competition tools proposed by the European Commission strongly advocate for a shift from ex-post antitrust liability balancing pro and anti-competitive effects, towards ex-ante regulatory obligations akin to sectoral regulations. Should these types of initiatives materialize in Europe and abroad, the precautionary shift will radically change competition enforcement and the antitrust discipline as we understand it.”).

[14]Digital Markets Act, 8.

[15]European Commission, Shaping Europe’s Digital Future (Luxembourg: Publications Office of the European Union, February 2020), 5,


[17]Digital Markets Act, Article 1(1).

[18]Gregory S. Crawford, Patrick Rey, and Monika Schnitzer, “An Economic Evaluation of the EC’s Proposed ‘New Competition Tool’,” EAGCP Report (Luxembourg: Publications Office of the European Union, October 2020), 14,

[19]Motta and Peitz, “Intervention triggers and underlying theories of harm,” 4–5 (distinguishing with respect to the “New Competition Tool” between horizontal scope and limited scope and advising not to choose a tool with limited (digital only) scope).

[20]European Commission, “Factual summary of the contributions received in the context of the open public consultation on the New Competition Tool” (European Commission, October 2020),

[21]Competition and Antitrust Law Enforcement Reform Act of 2021, S. SIL21191 6C1, 117th Congress (2021), The “CALERA” Bill introduced by Sen. Klobuchar (D-MIN) does not refer to digital markets or the digital sector. The only slight reference one can find is in Sec. 26A(e)(7).

[22]Digital Markets Act, 1.


[24]CERRE Report, 31, which refers to the “digital route to market.”

[25]Nitin Joglekar, Geoffrey Parker, and Jagjit Singh Srai, “Winning the Race for Survival: How Advanced Manufacturing Technologies Are Driving Business-Model Innovation,” White Paper (WEF, May 2020), 6, (noting that “manufacturing companies having noticed megatrends—from digitization to the imperative of environmental sustainability and globalization—and have begun investing in projects that exploit data”).

[26]Digital Markets Act, Article 2(2)(4).

[27]Digital Markets Act, Article 2(2)(3).

[28]Article 1(b) of the Directive (EU) 2015/1535 of the European Parliament and of the Council of September 9, 2015, laying down a procedure for the provision of information in the field of technical regulations and of rules on Information Society services (codification), L.241/1, September 17, 2015.

[29]Ibid, Article 1.1.b(i).

[30]Ibid, Article 1.1.b(ii). It is remarkable that, although the DMA explicitly refers to the 2015 Directive for determining its scope, this Directive contains the word “digital” only at that occurrence—namely, cursively, under brackets, to designate “digital compression.” It thus appears that the digital sector, let alone digital markets, is obviously not the primary concern of the 2015 Directive, although the DMA crucially relies upon its premise. This paradox further undermines the clarity of the DMA’s scope.

[31]Ibid, Article 1.1.b(iii).

[32]Wolter Lemstra, “An Integrated Regulatory Framework for Digital Networks and Services. Appendix 2: Imagine 2025” (CERRE Policy Report, January 2016), 15,

[33]Jacques Crémer, Roundtable “Regulation and Antitrust of Digital Platforms” at the Conference on Digital Platforms: Opportunities and Challenges? (Online, October 27 & 28, 2020), 8,

[34]Yougjin Yoo et al., “Organizing for innovation in the digitized world,” Organization Science 23, No. 5 (2012) 1398–1408, (noting on 1398 that “everyday products such as TVs, watches, and cars now have embedded software-based digital capabilities, and organizations are routinely creating management systems composed of intelligent machines with digital sensors, networks, and processors” and consider that “one of the most important traits of innovation processes and outcomes with pervasive digital technology is the emergence of a platform as the central focus of the innovation.”; Erik Brynjolfsson, Yu Hu, and Michael D. Smith, “Long Tails vs Superstars: The Effect of Information Technology on Product Variety and Sales Concentration Patterns,” Information System Research 21, No.4 (December 2010) 736–747, DOI:10.1287/isre.1100.0325.

[35]Joglekar, Parker, and Srai, “Winning the Race for Survival,” 6,  noting that, amidst the COVID-19 crisis, “Amazon Marketplace is able to more fluidly connect supply and demand” and that “the firms best-positioned to adapt quickly to disruption are those that can digitally link their design and operations all the way to the manufacturing floor to create seamless customers experience.”

[36]Brynjolfsson, Hu, and Smith, “Long Tails vs Superstars,” 736-747.

[37]Autorite de la Concurrence, FNAC/Darty, 16-DCC-111, July 27, 2016,; Decision 18-D-16, July 27, 2018; Autorité de la Concurrence, FNAC/Darty, 18-D-16, July 27, 2018, (fining Fnac Darty for failure to comply with divesture obligations).

[38]For a discussion of the decision in English language, see Orrick, “Fnac-Darty: A Landmark Merger Decision in France,” Antitrust Watch, August 5, 2016, (the merger decision constitutes “a significant step in Europe for analysis of relevant markets that include both physical and e-commerce distribution channels.”); Willkie Farr & Gallagher, “The French Competition Authority authorizes Fnac group to acquire Darty,” July 18, 2016, (considering the decision as “remarkable” because “retail sector included both in-store and online retail channels.”). For a broader discussion, see Simon Genevaz and Jerome Vidal, “Going Digital: How Online Competition Changed Market Definition and Swayed Competition Analysis in Fnac/Darty,” Journal of European Competition Law & Practice 8, No.1 (2017), 30–35, (concluding that “agencies will therefore continue to conduct detailed local analysis to assess whether customers have sufficient retail alternatives, may they be physical or digital, to shop in a competitive environment.”).

[39]Adeline Ochs, Jean-Francois Lemoine, an Olivier Badot, “The Phygital Shopping Experience: An Attempt at Conceptualization and Empirical Investigation,”; Patricia Rossi and Nina Krey (Eds.) “Marketing Transformation: Marketing Practice in an Ever Changing World,” AMSWMC 2017, Developments in Marketing Science: Proceedings of the Academy of Marketing Science (Cham, Switzerland: Springer, 2017) 61–74, (“one of the solutions suggested by companies to accompany consumers’ numerous physico-digital hybridizations is to proceed with hybridization themselves in their points of sale.”).

[40]Erik Brynjolksson, Yu Jeffrey, and Mohammad S. Rahman, “Competing in the Age of Omnichannel Retailing,” MIT Sloan Management Review, May 21, 2013, (“as the retailing industry evolves toward a seamless ‘omnichannel retailing’ experience, the distinctions between physical and online will vanish, turning the world into a showroom without walls.”); Deloitte, “The omnichannel opportunity. Unlocking the power of the connected consumer,” A Report for eBay, February 2014, (“by integrating and aligning channels, omnichannel retailing provides a flexible and seamless shopping experience to consumers, regardless of whether the customer walks into a store, browses on a website or orders via a mobile phone.”); Marco Savastano, Riccardo Barnabei, and Francesco Ricotta, “Going Online While Purchasing Offline: an Explorative Analysis of Omnichannel Shopping Behaviour in Retail Settings,” January 2016,; Sara Rosengren et al., “Catering to the Digital Consumer: From Multichannel to Omnichannel Retailing,” Chapter 4, in Per Andersson et al. (Eds.), “Managing Digital Transformation” (Stockholm: Stockholm School of Economics Institute for Research, 2018) 97–114,

[41]Consumers glean information in-store but shop online.

[42]See Annie Palmer, “Amazon just opened a cashierless supermarket—here are all the ways it’s trying to upend the grocery industry,” CNBC, February 25, 2020,; See also Rupinder P. Jindal et al., “Omnichannel battle between Amazon and Walmart: Is the focus on delivery the best strategy?” Journal of Business Research 122 (2021) 2700-280; Griffin Kao et al., “Online to Offline (O2O),” in Turning Silicon Into Gold (Berkeley, CA: Apress, 2020) 75–80. Also, in the merger decision between Picwic and Toys“R”Us, the French Competition Authority also considered the extent to which in-store retailers developed digital tools to enable consumers to see the products’ availabilities and to shop online. See Autorite de la Concurrence, Decision 19-DCC-65, April 17, 2019,

[43]Nick Kostov and Sam Schechner, “GDPR Has Been a Boon for Google and Facebook,” The Wall Street Journal, June 17, 2019, (“the rules have also made it harder for third parties to collect lucrative personal information like location data in Europe to target ads. This gives the tech giants another advantage: they have direct relationships with consumers that use their products, allowing them to ask for consent directly from a much large pool of individuals”); Mark Scott, Laurens Cerulus, and Steven Overly, “How Silicon Valley gamed Europe’s privacy rules,” Politico, May 22, 2019, (noting “small firms—whose fortunes were of special concern to the framers of the region’s privacy revamp—also have suffered from the relatively high compliance costs and the perception, at least among some investors, that they can’t compete with Silicon Valley’s biggest names.”); Nicholas Vinocur, “’We have a huge problem’: European tech regulator despairs over lack of enforcement,” Politico, December 27, 2019,; Jedidiah Yueh, “GDPR Will Make Big Tech Even Bigger,” Forbes, June 26, 2018,; Alex Moazed, “How the GDPR is Helping Big Tech and Hurting the Competition,”,,GDPR%20is%20missing%20its%20targets.&text=GDPR%20is%20helping%20them.,in%20some%20cases%20beyond%20repair (concluding that “GDPR is knocking off small-to-midsize competitors, and reinforcing the economic moat the tech monopolies already enjoy … Many European adtech companies are folding in response.” Most ironically, among the unintended consequences of the GDPR, the Regulation eased tech companies to bypass user consent.) See Noyb, “Vienna Superior Court: Facebook can ‘bypass’ GDPR consent, but must give access to data,”, December 29, 2020, (noting that “the GDPR allows different bases for the processing of personal data: for example, consent or a contract. Civil law contracts do not need to fulfill the strict requirements of ‘consent’ under the GDPR. This would mean that the company does not have to give users a free choice and obtain separated and unambiguous consent.”).

[44]Such a phenomenon is notoriously illustrated with, for example, the techlash against Uber wherein regulation takes place in order to insulate taxi drivers from the digital disruption brought about by the car-hailing platform. See Michael B. Horn, “Uber, Disruptive Innovation and Regulated Markets,” Forbes, June 20, 2016, (noting however that “the history of disruptive innovation suggests that these regulations that preserve the status quo will likely fade away over time.”); Feng Li, “Outdated regulations halt Uber in its tracks, but innovation must prevail,” The Conversation, October 25, 2017,; Veena B. Dubal, Ruth Berins Collier, and Christopher L. Carter, “Disrupting Regulation, Regulating Disruption: The Politics of Uber in the United States,” Perspectives on Politics 16 (2018), 919,; A similar argument can developed with the “Red Flag” regulation of the 1865 Locomotive Act in the United Kingdom. See Rhodri Marsden, “Lord Winchelsea’s Red Flag,” The Independent, July 3, 2015,

[45]Philippe Aghion et al., “Competition, Imitation and Growth with Step-by-Step Innovation,” Review of Economic Studies, LXVIII (2001), 467–492, (finding on 470 that “with respect to imitation, we find that a lot of it is always bad for growth. That is, as the ease of imitation goes to infinity the grow rate always falls to zero.”); Philippe Aghion et al., “Competition and Innovation: An Inverted-U Relationship,” Quarterly Journal of Economics (May 2005), 701–728,

[46]Philippe Aghion et al., “Competition, Imitation and Growth with Step-by-Step Innovation,” 467–492,

[47]Digital Markets Act, para.52.

[48]Thomas P. Lyon and Haizou Huang, “Asymmetric Regulation and Incentives for Innovation,” Industrial and Corporate Change 4, Issue 4 (1995), 769–776, (wherein the authors find, “We think of regulation as simply imposing a ‘tax’ on the regulated firm’s profits ... A higher regulatory tax may slow the pace of change because the regulated firms is no longer willing to innovate, and the industry shifts from a world in which both firms [the regulated innovator and the rival unregulated] innovate to a world where only the unregulated firm innovates.”).

[49]Ibid, 470.

[50]Digital Markets Act, 11.

[51]Joglekar, Parker, and Srai, “Winning the Race for Survival,” 6 (noting that “The firms best-positioned to adapt quickly to disruption are those that can digitally link their design and operations all the way to the manufacturing floor to create seamless customers experience.”).

[52]Joe Kennedy, “Monopoly Myths: Are Superstar Firms Stifling Competition or Just Beating It?” (ITIF, January 2021), /publications/2021/01/11/monopoly-myths-are-superstar-firms-stifling-competition-or-just-beating-it.

[53]Alexandre de Streel, “Conduct and Theories of Harm,” Issue Paper + Discussion Summary, in Alexandre de Streel (Ed.) Digital Markets Act. Making Economic Regulation of Platforms Fit for the Digital Age (Brussels, CERRE Report, December 2020), 12 (describing these practices as illustration of switching costs increases). On the Google Android decision referred to, see European Commission, Google Android, Case AT.40099, July 18, 2018, (stating controversially at 170 that “users are unlikely to look for, download, and use alternative apps, at least when the app that is pre-installed, premium placed and/or set as default already delivers the required functionality to a satisfactory level.”) On the reasons why consumer bias cannot be alleged in the Google cases in a manner similar to that of the Microsoft cases, see Aaron S. Edlin and Robert G. Harris, “The Role of Switching Costs in Antitrust Analysis: A Comparison of Microsoft and Google,” Yale Law Journal & Technology 15 (2013), 169–213, (concluding that “any monopolization case against Google Search would have to be very different from the Microsoft browser case, because the cost for a user switching from Google Search is much lower than was the cost in the 1990s (or today) of switching away from the Microsoft operating system.”).

[54]Default “bias” does not exist whenever the rational loss aversion incited by the status quo is part of the analysis. Indeed, “one implication of loss aversion is that individuals have a strong tendency to remain at the status quo, because the disadvantages of leaving it loom larger than advantages” argued Daniel Kahneman, Jack L. Knetsch, and Richard H. Thaler, “Anomalies. The Endowment Effect, Loss Aversion, and Status Quo Bias,” Journal of Economic Perspectives 5, No.1 (1991) 193–206, (concluding from experiments that “the advantage of the status quo increases with the number of alternatives.”. Invoking the status quo bias of consumer therefore means that consumers have too many choices—thus, status quo may reveal excessive competition rather than lack of competition contrary to antitrust assumptions.)

[55]Aurelien Portuese, “The Antitrust Prohibition of Favoritism, or the Imposition of Corporate Selflessness, Truth on the Market,” December 16, 2020, (concluding that “as a free webware, Google search’s default setting cannot be compared to default installation in the Microsoft case since minimal consumer stickiness entails (almost) no switching costs.”).

[56]Digital Markets Act, para.41.

[57]Ibid, para.46.

[58]Henry Samuel, “France to make ‘Gallic Google’ Quant government’s default search engine,” The Telegraph, January 9, 2020,; See also Mark Scott, “Qwant Wants to Be Alternative to Google,” December 31, 2014,

[59]Cass. R. Sunstein, “Nudging: A Very Short Guide,” Journal of Consumer Policy 37, 583–588, DOI 10.1007/s10603-014-9273-1 (noting at 585 that “some kind of default rule is essentially inevitable, and hence it is a mistake to object to default rules as such ... In many contexts, default rules are indispensable because it is too burdensome and time-consuming to require people to choose.”).

[60]Richard H. Thaler, Cass R. Sunstein, and John P. Balz, “Choice Architecture,” in Eldar Shafir (Ed.), The Behavioral Foundations of Public Policy, Ch.25, 428–439, (noting at 430 that “defaults are ubiquitous and powerful. They are also unavoidable in the sense that for any node of a choice architecture system, there must be an associated rule that determines what happens to the decision make if she does nothing.”)

[61]Daniel G. Goldstein et al., “Nudge Your Customers Toward Better Choices,” Harvard Business Review, December 2008,

[62]European Commission, Google Android, Case AT.40099, July 18, 2018, C(2018) 4761 final, paras.481–482,

[63]Joe Kennedy, “Monopoly Myths.”

[64]Lina Khan, “Amazon Antitrust’s Paradox,” Yale Law Journal 126 (2017), 564–907, (alleging that “Amazon’s strategy has enabled it to use predatory pricing tactics without triggering the scrutiny of predatory pricing laws.”). See also Robert D. Atkinson, “How Progressives Have Spun Dubious Theories and Faulty Research Into a Harmful New Antitrust Doctrine” (ITIF, March 2021), /publications/2021/03/10/how-progressives-have-spun-dubious-theories-and-faulty-research-harmful-new.

[65]Lina Khan concluded that Amazon has predatorily priced its products despite no evidence of any product being priced below marginal cost and then recouping the costs. Indeed, they are allegations rather than evidence. See Ibid, stating “Although Amazon may be recouping its initial losses in e-books through markups on physical books.”).

[66]Kevin Carty, Leah Douglas, and Lina Khan, “6 Ideas to Rein in Silicon Valley, Open Up the Internet, and Make Tech Work for Everyone,” New York Mag, December 11, 2017, (alleging that “though technically illegal, predatory pricing is a routine way for Silicon Valley firms to squash nascent competitors.”); Simon Edelsten, “Taking a chance with monopoly?” Financial Times, September 8, 2020,

[67]Peter Weill and Stephanie L. Woerner, “Thriving in an Increasingly Digital Ecosystem,” MIT Sloan Management Review 56, No.4 (2015) 27–34,

[68]Joseph A. Schumpeter, Capitalism, Socialism, and Democracy, 3rd Edition (New York: HarperCollins Publishers, 2008) [1942], 88.

[69]Digital Markets Act, para.25. In passing, one can note the blatant contradiction wherein both “high growth rates” and “decelerating growth rates” are evidence that the “gatekeeper” is about to enjoy an entrenched market position.

[70]For the notable example of Amazon, see Alberto Cavallo, “More Amazon Effects: Online Competition and Pricing Behaviors,” NBER Working Paper Series, October 2018,; Richard J. Gilbert, “E-books: A Tale of Digital Disruption,” Journal of Economic Perspectives 29 (2015) 165–184, Despite predatory pricing, Amazon does not seem to enjoy a monopoly life, see Peter Cohan, “Why Amazon Can’t Beat Shopify,” Forbes, December 23, 2020, For a broader discussion, see Nicolas Petit, Big Tech & The Digital Economy. The Moligopoly Scenario (Oxford: Oxford University Press, 2020), 97–99.

[71]Joe Kennedy, “Monopoly Myths.”

[72]JCDecaux, “The invention of a new business model,”, February 17, 2014,

[73]Chris Anderson, “Free! Why $0.00 Is the Future of Business,” Wired, February, 25, 2020,

[74]Interestingly, following the EU Google Android’s fining decision, the zero-priced business model has been questioned for European consumers so much so that Android may be no longer offered for free to business users. See Foo Yun Chee, “Google to charge Android partners up to $40 per device for apps,” Reuters, October 19, 2018, (“Google will charge hardware firms up to $40 per device to use its apps under a new licensing system to replace one that the European Union this year deemed anti-competitive.”); Ron Amadeo, “After $5 billion EU antitrust fine, Google will start charging for Android apps,” Ars Technica, October 17, 2019,

[75]Joseph A. Schumpeter, Capitalism, Socialism and Democracy, 81–86.

[76]Digital Markets Act, para.26.

[77]Summary of the contributions of the National Competition Authorities to the impact assessment of the new competition tool, Europa website (2020), 6, available at:

[78]See Motta and Peitz, “Intervention triggers and underlying theories of harm,” 4–5 (“choosing a competition tool with a horizontal scope ... is in our view the best choice.”); Competition and Antitrust Law Enforcement Reform Act of 2021, S. 21191, 117th Congress (2021).

[79]Digital Markets Act, para.26.

[80]Joe Kennedy, “Monopoly Myths.”

[81]European Commission, Flat Glass (Italy), Case COMP IV/31.906, OJ L 33, para.79 noting “the collective dominant position of FP, SIV and VP derives from the following factors. The joint market shares of some 79% for non-automotive glass and some 95% for automotive glass are sufficient in themselves to give FP, SIV and VP a dominant position on the Italian market in flat glass” before concluding at para.80 that “FP, SIV, and VP have abused their collective dominant position on the Italian market in flat glass, which constitutes a substantial part of the common market.”

[82]Joined cases T-24-6 & 29/93, Compagnie maritime belge transports SA and Compagnie maritime belge SA, Dafra-Lines A/S, Deutsche Afrika-Linien GmbH & Co and Nedlloyd Lijnen BV v Commission (1996) ECLI:EU:T:1996:139/.

[83]Should past communication among alleged firms be established, the collective dominance shall be treated as collusive practices and will be prosecuted under Article 101 TFEU. Should past communication among alleged firms be inexistent, the collective dominance shall be treated as joint market dominance and might, subject to anticompetitive conducts, be prosecuted under Article 102 TFEU. See, for a general discussion, Ioanis Lianos, Valentine Korah, and Paolo Siciliani, Competition Law. Analysis, Cases, & Materials (New York: Oxford University Press, 2019), 1418.

[84]Joe Kennedy, “Monopoly Myths.”

[85]Tommaso Bighelli et al., “Increasing market concentration in Europe is more likely to be a sign of strength than a cause for concern,” Vox, October 13, 2020,


[87]Motta and Peitz, “Intervention triggers and underlying theories of harm,” 12.

[88]Let alone the ability for consumers to multi-home and thus to mitigate network effects of platforms. See Shengli Li, Yipeng Liu, and Subhajyoti Banyopadyopadhyay, “Network effects in online two-sided market platforms: A research note,” 49 Decision Support Systems, 245–249 (2010) (“considering multi-homing among agents also has potential to counteract the tendency towards tipping and the lock-in effects in platforms with strong cross-group network externalities ”); Motta and Peitz, “Intervention triggers and underlying theories of harm,” 12 (noting that “tipping does not necessarily occur, since there are situations in which consumer multi-home, and others where network effects may co-exist with different tastes, so that more than one network goo may well continue to receive a significant share of customers even beyond the short-run”).

[89]Sangeet Paul Choudary, “Reverse Network Effects: Why Today’s Social Networks Can Fail As They Grow Larger,” Wired, March 2014 (noting that “reverse network effects often cause a large and thriving network to implode.”).

[90]Competition for the market generated by network effects is worsened by consumers’ switching costs: competition for the market involve investments and innovation to gain necessary scalability, but these investments require legitimate recoupments. See Motta and Peitz, “Intervention triggers and underlying theories of harm,” 17 (which “acknowledge[s] that in markets with scale economies, network effects, or switching costs there may be fierce competition during the period prior to market consolidation, with profit sacrifices being made on the expectation of future profit recovery after the market has consolidated. Ex post [antitrust] intervention should therefore include considerations of the legitimacy of such dynamic business strategies, i.e., firms should not necessarily be denied the recovery of upfront investments and profit sacrifices.”).

[91]On the confusion between “winner takes all” (possibly detrimental as being monopolization) and “winner takes most” (certainly beneficial as being competitive outcome), see the European Commission’s own questionnaire for the Digital Services Acts packages wherein ex ante competition tools were envisaged and consulted, European Commission, Consultation on the Digital Services Act packages, Public Consultation, June 2 to September 8, 2020,, at 33 wherein the Commission writes “the concentration of economic power in then platform economy creates a small number of “winner takes all/most” online platforms. The winner online platforms can also readily take over (potential) competitors and it is exceedingly difficult for an existing competitor or potential new entrant to overcome the winner’s competitive edge.” The “winner takes all” idiom has progressively given way to the “winner takes most” since it has been commonly accepted that even super dominant firms always face rivals in the same relevant product markets, thereby capturing most (and not all) of market shares if this criterion ought to remain relevant. Compare David S. Evans and Richard Schmalensee, “Some Economic Aspects of Antitrust Analysis in Dynamically Competitive Industries,” in Adame B. Jaff, Josh Lerner, and Scott Stern (Eds.) Innovation Policy and the Economy. Vol.2, Cambridge MA: MIT Press, 1–50 (stating that “in most of this literature, any given industry is assumed for simplicity to experience one and only one race, after which the winner enjoys a monopoly position forever”); Vincenzo Denicolo and Luigi Franzoni, “On the Winner-Take-All Principle in Innovation Races,” 8 Journal of European Economic Association, 1133–1158 (2010); Alden Abbott, “Antitrust and the Winner-Take-All Economy,” The Heritage Foundation, January 23, 2018; Robert H. Frank and Philip J. Cook, “Winner-Take-All Markets,” 1 Studies in Microeconomics 2, 131–154; Herbert Hovenkamp, “Antitrust and Platform Monopoly,” 130 Yale Law Journal (2021) (wherein it is summed up that “a ‘winner-take-all’ market is one in which the equilibrium number of sellers at any time is one ... Notwithstanding overwhelming evidence to the contrary, the market for digital platforms is often said to be winner-take-all. This is rarely true.”).

“Winner takes most,” obviously inherent to the positive outcome of the competitive process, appears synonymous to market tipping, which is aimed at being avoided by the DMA. Indeed, for instance, the Report by the Economic Advisory Group on Competition Policy (EAGCP) advising the European Commission on the possible New Competition Tool considered that market tipping is “when the economic fundamentals favor winner-take-most outcomes,” in Gregory S. Crawford, Patrick Rey, and Monika Schnitzer, “An Economic Evaluation of the EC’s Proposed ‘New Competition Tool’,” Report by the Economic Advisory Group on Competition Policy, Luxembourg: Office of the European Union.

[92]European Commission, Evaluation of the Commission Notice on the definition of relevant market for the purposes of Community competition law, Public Consultation, June 26 to October 9, 2020, (which resulted in many respondents considered that the Commission’s Notice “should not presume that market shares are the primary source for identifying a dominant position.” (p11) and “around one-quarter of the respondents who expressed a view indicated that the Notice provides correct, comprehensive and clear guidance as regards market share calculation, while the remaining respondents disagreed” given the presence of “zero price services and multi-sided platforms” (pp.19–20). Finally, “several respondents also indicated that static market shares are not a good indicated in dynamic markets, where innovation has a strong role” (p.20). See European Commission, Summary of the stakeholder consultation to the Evaluation of the Market Definition Notice, Ref. Ares (2020)7730543, December 18, 2020.

[93]Ibid. “winner takes most” outcome appears intertwined with the assumption that markets have become no longer contestable. For, market contestability implicitly calls for against “winner takes most” outcome, however beneficial incentives such an outcome may produce. Indeed, the EACGP has written “[W]e adopt as an organizing principle the goal of ensuring that such markets are contestable, not only for existing competitors currently operating in the market but also, for future competitors who could displace whoever is the winner, particularly if the economics of the existing market suggest that there will necessarily be a ‘winner-take-most’ outcome in the long run” (emphasis in original), in ibid at 15.

[94]Gregory S. Crawford, Patrick Rey, and Monika Schnitzer, “An Economic Evaluation of the EC’s Proposed ‘New Competition Tool,’” Report by the Economic Advisory Group on Competition Policy, Luxembourg: Office of the European Union, 17 (while discussing the proposed gatekeepers’ regulation, the EAGCP “see regulation as being reserved for ‘natural monopoly’ environments where there is high and durable market concentration”) (emphasis in original).

[95]Notwithstanding the terse evidence available to convincingly argue that digital platforms are natural monopolies akin to railroads or harbors, the antitrust implications of natural monopolies are even more contradictory, and one can hardly suggest a clear antitrust rule out of the belief that digital platforms are natural monopolies. Indeed, as Hovenkamp eloquently sums up this contradiction, natural monopolies can justify utility-style regulations, thereby justifying a single firm with rivals removed from the market since utility-style regulation resembles public services rather than competitive rivalry. Therefore, the natural monopoly defense has justified exclusionary practices against rivals in the US. Antitrust laws become irrelevant and are substituted with utility regulations. See Greenville Publ’g Co. v. Daily Reflector, Inc., 496 F.2d 391, 397 (4th Cir. 1974); Union Leader Corp. v. Newspapers of New England, Inc., 284 F.2d 582, 584 (1st Cir. 1960); City of Cleveland v. Cleveland Elec. Illuminating Co., 538 F. Supp. 1306, 1314–15 (N.D. Ohio 1980). On the other hand, to designate a digital actor as natural monopoly may bolster more aggressive antitrust enforcement since a super dominant platform would come under antitrust scrutiny. See This antitrust implication is prevalent in the confusing “common carrier antitrust” concept wherein both antitrust enforcement as well as utility-style regulation are applicable to a single market actor. Neo-Brandeisians advocate for common carrier obligations for digital platforms? See Lina Khan, “Amazon’s Antitrust Paradox,” 126 Yale Law Journal (2017), 710–805, (common carrier obligations are defined on 799 as “requiring platforms to ensure open and fair access to other business” and such obligations command for both net neutrality and search neutrality); Lina Khan, “The Separation of Platforms and Commerce,” 119 Columbia Law Review, 973–1098 (2018); Tim Wu, “Why Have a Telecommunications Law? Anti-Discrimination Norms in Communications,” 5, Journal on Telecommunications and High Technology Law 15, 31 (2006) (arguing that “it is the role of the carrier plays in the economy that necessitates duties of common carriage, not necessarily the potential for abuse of market power”). For a critical discussion from a European perspective, see Pablo Ibanez Colomo, “The Commission sends Amazon an SO: the rise of common carrier antitrust, Chillin’ Competition,” November 10, 2020 (concluding that “common carrier antitrust tends to lead to a relaxation of [antitrust] assessment. Under this sui generis approach to Article 102 TFEU, any distortions of competition that come from self-preferencing are considered problematic, even when they intensify rivalry”).

[96]Summary of the contributions of the National Competition Authorities to the impact assessment of the new competition tool, Europa website (2020), 6, available at:


[98]Joe Kennedy, “Monopoly Myths.”

[99]Apple’s built-in search feature was once named “Sherlock.” Watson, the Karelia company’s companion to Sherlock, enabled users for $30 to also search the web. Subsequent versions of Apple’s Sherlock also enabled users to search the web. As Sherlock was offered for free, Karelia’s Watson became irrelevant and unattractive. Undoubtedly, Watson was once innovative. However, it became inefficient after Sherlock’s free subsequent versions and given that Watson neither offered higher quality nor a price worth paying. Consequently, to hastily conclude that the disappearance of Watson is tantamount to the disappearance of an efficient and innovative company disregards the consumer welfare increase associated with such a disappearance. Consumers enjoyed a product they either never used or used at a cost.

[100]William Gallagher, “Developers talk about being ‘Sherlocked’ as Apple uses them ‘for market research,’” Apple Insider, June 6, 2019; Justin Pot, “What Does It Mean When Apple ‘Sherlocks’ an App?” How-To Geek, March 14, 2017.

[101]Motta and Peitz, “Intervention triggers and underlying theories of harm,” 31 (concluding as per “Sherlocking” that “it is not clear to us to what extent these business models or products or ideas are worth protecting and/or cannot be protected, and whether such ‘cloning’ is really harmful to consumers”).

[102]Dan Andrews, Müge Adalet McGowan, and Valentine Millot, “Confronting The Zombies: Policies for Productivity Revival,” Economic Policy Paper, No 21, OECD Economic Policy, December 2017,

[103]On the notion of market consolidation as necessary scale-up pathway to entrepreneurial success, see Graeme K. Deans, Fritz Kroeger, and Stefan Zeisel, “The Consolidation Curve,” Harvard Business Review, December 2002, (stating “everyone knows that most new industries are fragmented and consolidate as they mature”); Costas Markides and Paul Geroski, “The Innovator’s Prescription: The Art of Scale,” Summer 2004, Issue 35, June 1, 2004,

[104]Joe Kennedy, “Monopoly Myths.” 

[105]Motta and Peitz, “Intervention triggers and underlying theories of harm,” 17.

[106]David S. Evans, “The Antitrust Economics of Multi-Sided Platform Markets,” Yale Journal on Regulation 20, 325–381, 356,

[107]Lianos, Korah, and Siciliani, Competition Law. Analysis, (noting on 229 that in digital industries, “competition ‘for’ (rather than ‘in’) in the market means that consumers benefit greatly from lower prices, more convenient mode of consumption and strong innovation.”).

[108]Joe Kennedy, “Monopoly Myths.” 

[109]See, for instance, Kim Souza, “The Supply Side: Grocery consolidation likely to continue in challenging market,” TalkBusiness, April 14, 2019,; Richard Lough, “Explainer: Why Siemens-Alstom rail merger is creating European tensions,” Reuters, January 17, 2019,; Michael Wayland, “An Apple car would increase pressure on Tesla and other automakers,” CNBC, February 6, 2021.

[110]In the tech sector, this subsequent larger scale competition is evidenced with the “moligopoly scenario,” see Nicolas Petit, Big Tech & The Digital Economy. The Moligopoly Scenario (Oxford: Oxford University Press, 2020). More generally, see David J. Teece, Dynamic Capabilities & Strategic Management. Organizing for Innovation and Growth (Cambridge, MA: Oxford University Press, 2009), (concluding on 249 that “market concentration is likely to be an outcome of market selection, which in turn depends on the uneven exploitation of learning opportunities; that is, concentration has little to do with market power.”).

[111]Janusz Ordover, cited in Teece, Dynamic Capabilities, 250.

[112]See, for instance, the reform of the U.S. Merger Guidelines in 2010 and the ongoing reform of the EU rules on market definitions: U.S. Department of Justice and the Federal Trade Commission, Horizontal Merger Guidelines, August 19, 2010,; European Commission, Evaluation of the Commission Notice on the definition of relevant market for the purposes of Community competition law (Brussels, June 26, 2020-October 9, 2020),

[113]Competition and Antitrust Law Enforcement Reform Act of 2021, SIL21191, 117th Cong. (2021).

[114]Robert D. Atkinson et al., “A Policymaker’s Guide to the ‘Techlash’—What It is and Why It’s a Threat to Growth and Progress” (ITIF, 2020), (arguing for policymakers to “resist techlash and embrace pragmatic ‘tech realism’—recognizing technology is a fundamental force of human progress that also can pose real challenges, which deserve smart, thoroughly considered, and effective responses.”).

[115]The digital protectionism of the DMA can be inferred from “Report from the Panel of Economic Experts on the DMA” which found that, with respect to the designation of gatekeepers, “these thresholds are designed to capture the largest online platforms, where potential harm is the greatest. Effectively, it comes down to the GAFAM tech giants (Google, Apple, Facebook, Amazon, and Microsoft), possibly a few more. Large Chinese platforms still have insufficient foothold in the EU market to match the criteria.” See in Luis Cabral et al., “The EU Digital Markets Act. A Report from a Panel of Economic Experts” (Luxembourg: Publications Office of the European Union, 2021), 9; Thomas Husson, “EU Digital Markets And Services Acts: A Bold Move,” Forrester, December 15, 2020, (reporting that French President Emmanuel Macron identified the U.S. GAFA and the Chinese BATX as reasons for EU regulations).`

[116]The term “gatekeeper” seems to have the preference of the Commission under the DMA’s proposal, albeit other concepts have been used in the literature and in agency reports. For instance, large platforms were referred to as holding “strategic market status” in the Furman Report, as enjoying “bottleneck power” in the Stigler Report, or being “structuring digital platforms,” but plainly as “gatekeepers” in the Commission’s commissioned Crémer Report. See, respectively, Jason Furman, Diane Coyle, Amelia Fletcher, Philip Marsden, Derek McAuley, “Unlocking Digital Competition. Report of the Digital Competition Expert Panel” (London: UK Government, March 2019), 10, (wherein a digital company is said to hold a “strategic market status” whenever it is “enduring market power over a strategic bottleneck market”); Luigi Zingales, Guy Rolnik, and Filippo Maria Lancieri (Eds.), Stigler Committee on Digital Platforms. Final Report (Chicago, September 2019), 32, (wherein “bottleneck power” is described as “a situation where consumer primarily single-home and rely upon a single service provider, which makes obtaining access to those consumers for the relevant activity by other service providers prohibitively costly.”); Jacques Crémer, Yves-Alexandre de Montjoye, and Heike Schweitzer, “Competition policy for the digital era” (Brussels: European Commission, January 2019), 48, (“control over the devices allows a platform to become a gatekeeper in terms of access to consumer data and capacity to deliver content and services.”).

[117]European Commission, Proposal for a Regulation of the European Parliament and of the Council on promoting fairness and transparency for business users of online intermediation services, COM(2018) 238 final, April 26, 2018, 1, (“This growing intermediation of transactions through online platforms, combined with strong indirect network effects that can be fueled by data-driven advantages by the online platforms, lead to an increased dependency of businesses on online platforms as quasi ‘gatekeepers’ to markets and consumers.”)

This radical and rapid change in adopting a nascent antitrust concept was presumably boosted by the European Commission’s own reinforcing contract bid. See Samuel Stolton, “Commission to contract €600,000 study on gatekeeping power of digital platforms,”, May 11, 2020, (“a call for tenders document, detailing an upper limit of €600,000 for the study, states that the report should include ‘robust data and insights as regards issues linked with significant network effects [and] gatekeeping power.’”).

[118]Sherwin Rose, “The Economics of Superstars,” The American Economic Review 71, No.5 (December 1981) 845–858,, 845 (noting that “the phenomenon of Superstars ... seems to be increasingly important in the modern world.”); David Autor et al., “The Fall of The Labor Share and The Rise of Superstar Firms,” The Quarterly Journal of Economics 135, Issue 2 (2020) 645–709,, (which demonstrates on 650 that the fall in the labor share is largely due to reallocation of sales and value between firms as a result of growth in productivity and innovation).

[119]Nick Bastone, “The 14 biggest product flops in Google history,” Business Insider, November 26, 2018, (“when a company tried as many crazy new ideas as Google, there are bound to be a few epic fails along the way.”); Clifford Colby, “Apple’s worst failures of all time, from AirPower to Bendgate to the Newton,” CNET, April 2, 2019, (“from unforeseen glitches and lower-than-expected sales to design fails—it doesn’t always go right with Apple.”); Irina Ivanova, “Facebook’s biggest failures,” CBS News, March 22, 2018, (reviewing “some of the past controversies that have dogged the social networking giant”); John Rossman, “How Amazon succeeds by learning from failure,”, May 2, 2019, (“Amazon Destinations, Amazon Wallet, Amazon Music Importer, Amazon Webstore—the list of Amazon failures is long.”); David Hayward, “Microsoft’s Greatest Failures,”, April 8, 2020, (“the company has, along the way, come up with a few stinkers ... These new elements haven’t always worked, but at least they’ve tried.”); David Trainer, “Netflix’s Original Content Strategy is Failing,” Forbes, July 19, 2019, (“these disappointing subscriber numbers show that Netflix’s investment in original content has failed to deliver the sustainable competitive advantage required to justify its valuation. The loss of licensed content, increased competition, and higher prices in the future mean investors should expect more disappointing subscriber numbers going forward.”).

[120]An acronym that stands for the Chinese companies Baidu, Alibaba, Tencent, and Xiaomi, with cumulative market capitalization of approximatively $950 billion in 2020.

[121]Robin Wigglesworth, “How Big Tech got even bigger in the Covid-19 era,” Financial Times, May 1, 2020,; Kirk Miller, “A Look at the Tech Companies That Have Capitalized Off of Coronavirus,” InsideHook, February 10, 2021,; Richard Waters et al., “Big Tech shows its resilience to pandemic and politics,” Financial Times, October 30, 2020,

[122]Robert Atkinson, Michael Lind, Big Is Beautiful. Debunking The Myth of Small Business (Cambridge, MA: MIT Press, 2018).

[123]Ibid, 97–103.

[124]Ibid, 97.

[125]From politicians to policy advocates, big tech companies are regularly portrayed as the cause of the economic difficulties of traditional businesses disrupted by the digital innovations of these companies. See, for instance, Jonathan Taplin, “It is Time to Break Up Google!” The New York Times, April 22, 2017; Elizabeth Warren, “Here’s how we can break up Big Tech,” Medium, March 8, 2019,; Lina Khan, “The Separation of Platforms and Commerce,” Columbia Law Review 119, 973–984,; Zephyr Teachout, Break’ Em Up: Recovering Our Freedom from Big Ag, Big Tech, and Big Money (New York: All Points Books, 2020); Emily Chang, Brotopia: Breaking Up the Boys’ Club of Silicon Valley (New York: Penguin, 2018); Scott Galloway, The Four. The Hidden DNA of Amazon, Apple, Facebook, and Google (New York: Penguin, 2017); Matt Stoller, Goliath: The 100-Year War Between Monopoly Power and Democracy (New York: Simon & Schuster, 2019); Tim Wu, The Curse of Bigness: Antitrust in the New Gilded Age (Columbia: Columbia Global Reports, 2018); Rana Foroohar, Don’t Be Evil: The Case Against Big Tech (New York: Penguin, 2019); Rob Larson, Bit Tyrants: The Political Economy of Silicon Valley (Chicago: Haymarket Books, 2020).

[126]Subsidization and tax breaks for companies competing with Amazon is, for instance, advocated by the Institute of Self-Reliance. See Stacy Mitchell and Olivia LaVecchia, “Amazon’s Stranglehold: How the Company’s Tightening Grip Is Stifling Competition, Eroding Jobs, and Threatening Communities” (Institute for Local Self-Reliance, November 2016), 70–71, (advocating for officials to look “to loan funds that help small business access credit ... purchasing policies that use public dollars to strengthen the local economy, among other strategies.”).

[127]Ibid, 650–651.

[128]Ibid, 651 debunking the ideas according to which “the rise of superstar firms ... reflect a diminution of competition, due to weaker U.S. antitrust enforcement [whereas] concentrating sectors appear to be growing more productive and innovative.”

[129]A.A. Berle, “Bigness: Curse or Opportunity?” The New York Times, February 18, 1962,

[130]Solomon Fabricant, “Is Monopoly Increasing?” The Journal of Economic History 13, No.1, 1953, (noting particularly at 91 that “for example, most industries are considered to be national in scope: concentration within local markets is taken into account only when known to be significant—which probably understates monopoly; and international competition is largely ignored—which overstates monopoly.”).

[131]Joseph A. Schumpeter, Capitalism, Socialism, and Democracy, 3rd Edition (New York: HarperCollins Publishers, 2008) [1942], 81.


[133]Ibid, 105–106.

[134]See III.3.

[135]Nicolas Petit, Big Tech & The Digital Economy. The Moligopoly Scenario (Oxford: Oxford University Press, 2020), 62–63 (stating that “while there is an undisputable trend toward industry concentration in the digital economy, there is also a competitive force behind it ... our analysis allows us to entertain doubts that big tech firms deserve to be considered as monopolies.”).

[136]Luciano Fanti et al., “Local and global dynamics in a duopoly with price competition and market share delegation,” Charo, Solitons & Fractals 69, 253–270,; Matt Oguz, “A beautiful duopoly,” TechCrunch, May 8, 2019,

[137]Joe Kennedy, “Monopoly Myths.”  

[138]The DMA does not make the useful distinction between platforms and aggregators. However, such a distinction can be of interest. See Thibault Schrepel, “Platforms or Aggregators: Implications for Digital Antitrust Law,” Journal of European Competition Law & Practice 12, No.1 (2021), 1–3 (suggesting that the per se prohibitions of consumer choice reduction by aggregators amount to “holding aggregators’ business models as anticompetitive.”).

[139]Digital Markets Act, 1.

[140]Lianos, Korah, and Siciliani, Competition Law. Analysis, 229 (considering that “it is often argued that hyper-competitive rivalry is the norm in digital industries subject to ‘winners-take-all’ competitive dynamics, where a position of super-dominance is the market outcome of strategies based on very aggressive pricing and/or relentless product and process innovation.”).

[141]The concept of “super dominance” was first referred to by Advocate General Fennelly in the case of Compagnie Maritime Belge, wherein he stated that “To my mind, Article 86 [now Article 82] cannot be interpreted as permitting monopolists or quasi-monopolists to exploit the very significant market power which their superdominance confers so as to preclude the emergence either of a new or additional competitor. Where an undertaking, or group of undertakings ... enjoys a position of such overwhelming dominance verging on monopoly, comparable to that which existed in the present case at the moment when G & C entered the relevant market, it would not be consonant with the particularly onerous special obligation affecting such a dominant undertaking not to impair further the structure of the feeble existing competition for them to react, even to aggressive price competition from a new entrant, with a policy of targeted, selective price cuts designed to eliminate that competitor,” Opinion of Advocate General Fennelly delivered on October 29, 1998, in Compagnie maritime belge transports SA (C-395/96 P), Compagnie maritime belge SA (C-395/96 P) and Dafra-Lines A/S (C-396/96 P) v Commission, Joined cases C-395/96 P and C-396/96 P, [2000] E.C.R. I-1365 at para. 137.

[142]European Commission, Microsoft Corp. Case COMP/C-3/27.792 Microsoft, November 10, 2005, (at para.8 stating, “Any interest claimed by Microsoft as requiring protection as well as any justification brought forward by Microsoft as to the necessity and proportionality of a condition in question must be considered in the light of Microsoft’s special responsibility as a dominant undertaking. This is a responsibility that weighs particularly on Microsoft since that undertaking enjoys an overwhelmingly dominant position.”).

[143]“The best of all monopoly profits is a quiet life,” once wrote John Hicks. See John Hicks, “Annual survey of economic theory: The theory of monopoly,” Econometrica 3 (1935), 1–20. See also, Stephen A. Rhoades and Roger D. Rutz, “Market Power and Firm Risk. A Test of the ‘Quiet Life’ Hypothesis,” Journal of Monetary Economics 9 (1982) 73–85,

[144]European Commission, “Factual summary of the contributions received in the context of the open public consultation on the New Competition Tool” (European Commission, October 2020), 15, The confusion goes further by first considering that digital gatekeepers can be nondominant companies, and then by considering that the DMA should apply to nondominant companies as well as to gatekeepers—thereby suggesting that gatekeepers are dominant companies. See ibid, 1, “As regards the intervention trigger for the NCT, the majority of respondents that expressed a view in this regard considered that such a tool should focus on structural competition problems, thus being applicable to all companies in a market, rather than only to dominant companies or gatekeepers or digital platforms.”

[145]Summary of the contributions of the National Competition Authorities to the impact assessment of the new competition tool, Europa website (2020), 7, available at: (reporting that NCAs consider that “gatekeepers may escape the dominance test under Article 102 TFEU. Furthermore, Article 102 TFEU is not a tool that is designed to preserve or enhance competition, but rather to restore competition once there is distortion ... At the same time, some NCAs specify that Article 102 TFEU is appropriate to deal with certain anticompetitive conducts in gatekeeper markets,”).

[146]NCAs support such endeavors. For, “the general view among NCAs with relevant experience is [that gatekeeper’s ex ante regulation] should leave from the tradition dominance concept and thus apply to all undertakings in a market, including dominant but also non-dominant companies. This is because a dominance-based tool would fail to cover many important structural competition problems that are not caused by dominant companies. At the same time, as some NCAs point out, using the dominance threshold would have the advantage of relying on well-tested legal concepts,” in Summary of the contributions of the National Competition Authorities to the impact assessment of the new competition tool, Europa website (2020), 8–9, available at:

[147]Summary of the contributions of the National Competition Authorities to the impact assessment of the new competition tool, Europa website (2020), 6, available at:

[148]Ibid, 15.

[149]Ibid, 16.

[150]Ibid, 10.

[151]Digital Markets Act, 8. This phrasing is repeated once in the DMA. See, Digital Markets Act, 15, “existing Union law does not address, or does not address effectively, the identified challenges to the well-functioning of the internal market posed by the conduct of gatekeepers, which are not necessarily dominant in competition-law terms.”

[152]On the many contradictions antitrust regulators generate with respect to tech companies, see Mark A. Lemley, “The Contradictions of Platform Regulation,” Working Paper (February 2021),

[153]Jacques Crémer, Yves-Alexandre de Montjoye, and Heike Schweitzer, “Competition Policy for the Digital Era” (Brussels: European Commission, 2019), 4, For comparison, the Furman Report refers to “strategic market status” which both hints at the narrow notion of unavoidability and the broader notion of “gatekeeper.” See Jason Furman et al., “Unlocking digital competition.” For a general discussion on the notion of unavoidability in these reports, see Aurelien Portuese, “European Competition Enforcement and the Digital Economy: The Birthplace of Precautionary Antitrust,” Report on the Digital Economy (Arlington, VA: Global Antitrust Institute, 2020), 597–651,

[154]See Paul A. Johnson, “Suggestions for competition authorities when assessing vertical restraints in multi-sided platforms,” in OECD, “Rethinking Antitrust Tools for Multi-Sided Platforms” (Paris: OECD, 2018), 201–211, (“platforms, equivalently termed ‘two-sided markets’ or ‘multi-sided markets’ are intermediaries.”)

[155]Bernard Caillaud and Bruno Jullien, “Chicken & Egg: Competition among Intermediation Service Providers,” The RAND Journal of Economics 34, No.2 (2003), 309–328,

[156]Digital Markets Act, 1.

[157]Ibid, 2.

[158]The Staff Working Document refers to the stakeholders’ responses noting that “certain platforms and their ecosystems have become unavoidable to access a large variety of contents and services on the internet. Those structuring platforms have become gatekeepers not only within their services, but for the internet at large,” and cites the European Data Protection Board’s Guidelines 08/2020 on the targeting of social media users wherein it is explained that “the unrivalled insight capabilities provided by the platform may make it an 'unavoidable trading partner' for online marketers.” See European Commission, Commission Staff Working Document, Impact Assessment Report Accompanying the document for a Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), SWD(2020) 363 final, December 15, 2020, Part 1/2, Also, this unavoidability assumption of intermediaries was present in the PSB Regulation’s Staff Working Document. See European Commission, Commission Staff Working Document, Impact Assessment, accompanying the document, Proposal for a Regulation of the European Parliament and of the Council on promoting fairness and transparency for business users of online intermediation services, COM(2018) 238 final, April 26, 2018, (wherein it was also stated that “The European Economic and Social Committee noted that online platforms benefit from a strong first-mover advantage, and that those exploiting network effects can become an unavoidable trading partner for businesses.”)

[159]European Commission, Commission Staff Working Document, Impact Assessment Report Accompanying the document for a Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), SWD(2020) 363 final, December 15, 2020, Part 2/2, 34,

[160]European Commission, Commission Staff Working Document, Impact Assessment Report Accompanying the document for a Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), SWD(2020) 363 final, December 15, 2020, Part 2/2, 91,

[161]Andreas Schwabe et al., “Getting the Most From Europe’s Marketing Ecosystem,” Boston Consulting Group, May 22, 2020,

[162]Lara O’Reilly and Laura Stevens, “Amazon, With Little Fanfare, Emerges as an Advertising Giant,” The Wall Street Journal, November 27, 2018, (noting that “it push challenges the big ad sellers, including Facebook, Google and TV networks”); Matthew Capala, “Google Ads vs Amazon Advertising: Where to Put Your Advertising Dollars,” Alphametic, March 19, 2019,; Pymnts, “Google Offers Free Ads To Lure Merchant From Amazon,”, April 21, 2020,

[163]For the case of digital advertising market, Google seems to be a highly avoidable trading partner. See Zaid Ammari, “How to Market Online Without Relying on Google AdWords,” Forbes, February 27, 2019,; PPC Hero, “10 Alternatives To Google and Facebook Ads,” PPC Hero, February 8, 2021,; Matteo Duo, “21 Best AdSense Alternatives to Consider for Your Website in 2021,” Kinsta, December 30, 2020,

[164]Article 2(2)(b) of the Regulation (EU) 2019/1150 of the European Parliament and of the Council of June 20, 2019 on promoting fairness and transparency for business users of online intermediation services, July 11, 2019, L.186/57, (referred to as the “Platform-to-Business” Regulation, or “PSB Regulation”).

[165]Jean-Claud Rochet and Jean Tirole, “Platform competition in two-sided markets,” Journal of European Economic Association 1, No.4 (2003), 990–1029,; OECD, “Rethinking Antitrust Tools for Multi-Sided Platforms” (Paris: OECD, 2018),

[166]Judd Cramer and Alan B. Krueger, “Disruptive Change in the Taxi Business: The Case of Uber,” American Economic Review 106, No.5 (2016), 177–182, DOI: 10.1257/aer.p20161002; Carla Sanchez Armas, “Uber vs. Taxi,” Harvard Business School, March 31, 2020,

[167]Peter Alexiadis and Alexandre de Streel, “Designing an EU Intervention Standard for Digital Platforms,” RSCAS 2020/14, EUI Working Papers (2020), 6–7,

[168]European Commission, Virgin/British Airways, IV/D-2/34.780, February 4, 2000, 2000/74/EC, (noting in para.92 that “BA's position on the markets for air transport make it an obligatory business partner for travel agents. As BA has pointed out, the IATA passenger agency programme may oblige it to deal with all travel agents who meet certain objective criteria.”).

[169]European Court of Justice, Deutsche Bahn (German Railways), October 21, 1997, ECR II-1689, para.57.

[170]Article 8 of the Ordonnance No.86-1243 on Freedom in Pricing and Competition, succeeded by Article L.420-2 of the “Code de Commerce.”

[171]Section 20(1) of the Act Against Restraints of Competition in the version published on June 26, 2013 (Bundesgesetzblatt (Federal Law Gazette), I (2013), 1750, 3245 as last amended by Article 10 of the Act of July 12, 2018 (Federal Law Gazette I), 1151,

[172]European Commission, Sixteenth Report on Competition Policy (Brussels: Commission of the European Communities, 1987), 234,

[173]Michael J. Trebilcock, The Limits of Freedom of Contract (Cambridge, MA: Harvard University Press, 1997); Michael J. Trebilcock, “Critiques of the Limits of Freedom of Contract: A Rejoinder,” Symposium: Michael J. Trebilcock’s The Limits of Freedom of Contact, Osgoode Hall Law Journal 33 (1995),

[174]Michael J. Trebilcock, “Critiques of the Limits of Freedom of Contract,” 355.

[175]Article 3(2) of the Digital Markets Act, 36. See also Digital Markets Act, 2.

[176]European Commission, Executive Summary of the Impact Assessment Report accompanying the document Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), SWD (2020) 364 final (Brussels: European Commission, December 2020), 21.

[177]Digital Markets Act, para.26.

[178]Digital Markets Act, 19.

[179]Brandon Vigliarolo, “Zoom, Teams, Skype, Webex: Business collaboration software becoming a battleground, Aternity says,” TechRepublic, February 11, 2021,

[180]Rae Hodge, “Signal, WhatsApp and Telegram: All the major security differences between messaging apps,” CNET, February 19, 2021,

[181]Statista, “Global market share held by the leading smartphone operating systems in sales to end users from 1st quarter 2009 to 2nd quarter 2018,” Statista, November 25, 2020,

[182]The “assailable” market position of Apple was however predictable for some. See Henry Blodget, “Hey, Apple, Wake Up—It’s Happening Again,” Business Insider, January 5, 2010,

[183]Pierre Larouche and Alexandre de Streel, “Interplay between the New Competition Tool and Sector-Specific Regulation in the EU,” Expert Study (Brussels: European Commission, October 2020), 9, (wherein the authors refer to “structural risks for competition” illustrated by tipping markets wherein “powerful market players with an entrenched market and/or gatekeeper position” are believed to represent a “threat for competition.”). See also Alexandre de Streel, “Digital markets act: making economic regulation of platforms fit for the digital age” (Brussels: CERRE, 2020), 62 , (which considers that data sharing obligations is “viewed as a remedy to concerns that digital platforms performing a gatekeeper role may have acquired an unassailable position within their core market by virtue of the scale and scope of the data which they hold, given network and feedback effects.”).

[184]Motta and Peitz, “Intervention triggers and underlying theories of harm,” 12 (noting that market tipping refers to the idea that “once a firm has obtained a certain advantage over rivals in terms of market share, its position may become unassailable and the market may tend to a situation of monopoly.”).

[185]Digital Markets Act, 20.

[186]Timothy F. Bresnahan and Shane Greenstein, “Technological Competition and the Structure of the Computer Industry,” The Journal of Industrial Economics 47, No.1 (1999) 1–40,

[187]David J. Teece, “Profiting from Technological Innovation: Implications for Integration, Collaboration and Public Policy,” Research Policy 15 (1986), 285–305, See, more recently, David J. Teece, Dynamic Capabilities & Strategic Management. Organizing for Innovation and Growth (Cambridge, MA: Oxford University Press, 2009), 242 (summing up that “the advent of technological ensembles or paradigms is usually marked by a wave of new competitors entering an industry to sustain success. Incumbents must master discontinuities as well as incremental change and improvement.”).

[188]Nicolas Petit, “Big Tech Platforms and Schumpeter’s Creative Destruction,” ProMarket, November 19, 2020, (adding that “firms do not enter by imitation, replication, or substitution of incumbent firms’ capabilities.”).

[189]Ibid. More generally, see Nicolas Petit, “Big Tech & The Digital Economy. The Moligopoly Scenario” (Oxford: Oxford University Press, 2020), 257 (who concludes his analysis of big tech’s rivalry under the “moligopoly framework” by stating that “firms can take competition from several sources, in carious dimensions. Zero sum game rivalry for market share is just one of them. Technology is another. Digital industries exhibit a variety of intrinsic properties that work together to impose on firms a pressure equivalent to oligopoly competition … Tech firms compete with others by a process of indirect entry and reconfigure existing channels of competition.”).

[190]Billy Cheung, “Who are Facebook’s Main Competitors?” Investopedia, March 17, 2020,; Prinona Das, “The 10 Biggest Competitors of Facebook,” Feedough, October 21, 2019, (noting that Facebook, “with increasing popularity, it also gained strong competitors.”). This competitive portrayal of the social networking market, if there is one, as opposed to the advertising market, is a stark contradiction with the popular depiction of Facebook’s alleged “monopoly power.” See Dina Srinivasan, The Antitrust Case Against Facebook: A Monopolist’s Journey Towards Pervasive Surveillance in Spite of Consumers’ Preference for Privacy,” Berkeley Business Law Journal 16, Issue 1 (2019) 39–101, (noting on 43 “colloquially, and in the press, Facebook is a monopoly.”).

[191]Quoted in Matt Ridley, How Innovation Works (London: HarperCollins Publishers, 2002) (noting that “innovation often disappoint in its early years, only to exceed expectations once it gets going.”).

[192]Roy Amara, “Roy Amara 1925–2007, American futurologist,” in Oxford Essential Quotations, edited by Susan Ratcliffe (Oxford: Oxford University Press, 2016),

[193]Larouche and de Streel, “Interplay between the New Competition Tool and Sector-Specific Regulation,” (which on 9, in a reference reminiscent of the Structure-Conduct-Performance defund paradigm, refers to the risks created by “certain Structural Competition Problems (SCP) due to problematic market features.”).

[194]Digital Markets Act, 21 (targeting market tipping: “once a firm has obtained a certain advantage over rivals in terms of market share, its position may become unassailable and the market may gravitate towards a situation of dominance or (quasi)-monopoly.”). It is noticeable that never does the Digital Markets Act consider the potential for indirect entry by rivals into incumbent market positions. Indirect network effects reinforcing gatekeepers’ market power are the only indirect effects to be considered by the Digital Markets Act. This partial assessment undermines the DMA’s accuracy of the competitive constraints exerted by rivals and faced by gatekeepers.

[195]See Nikolas Guggenberger, “Essential Platforms,” Yale Law & Economics Research Paper (2020), 8–9,“I suggest a two-tiered remedy for digital bottlenecks: At its first level, regulators and courts must bar discrimination and self-preferencing by platforms. At its second level, after an appropriate amortization period, beginning with the tipping of the market, antitrust enforcers must upend platform-monopolies entirely.”). See also K. Sabeel Rahman, “Regulating Informational Infrastructure: Internet Platforms as the New Public Utilities,” Georgetown Law Technology Review 2 (2018), 234–251,–51.pdf (recommending “public utility regulation” to digital gatekeepers wherein “structuralist” regulations are advised); Lina Khan, “The Separation of Platforms and Commerce,” Columbia Law Review 119, No.4 (2019), 973–1098 (suggesting structural separation of platforms and using India’s populist structural separation of Amazon as an inspiration). These misguided calls for structural separations of platform aim at applying, to a competitive digital sector, solutions applied to sectors of the past wherein natural monopolies were ubiquitous. Absent the criteria for the essential facilities doctrine to be applicable to the digital sector, these unreasonable calls should be received with caution at minimal.

[196]Luciano Floridi, “The Fight for Digital Sovereignty: What It Is, and Why It Matters, Especially for the EU,” Philosophy & Technology 33, 369–378, (acknowledging on 374 “the risk, when supporting national digital sovereignty, is to end up supporting digital sovereignism or digital statism.”); Eanna Kelly, “Decoding Europe’s new fascination with ‘tech sovereignty’,” Science | Business, September 3, 2020,; These calls for digital sovereignty overlook the fundamental difference between a Western democracy and a Asian dictatorship rival. See Charlene Barshefsky, “EU digital protectionism risks damaging ties with the US,” Financial Times, August 2, 2020, Instead, we urgently need a transatlantic tech partnership, see Tom Wheeler, “Time for a U.S.-EU Digital Alliance,” January 21, 2021, (“a transatlantic digital alliance can begin to be built around common analytical systems to assess the digital marketplace.”).

[197]European Commission, “Factual summary of the contributions received in the context of the open public consultation on the New Competition Tool” (European Commission, October 2020), 14,

[198]Beyond two years, market entry and market dynamics are not considered to be “timely.” For instance, see European Commission, Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, February 5, 2004, para.74 (“The Commission examines whether entry would be sufficiently swift and sustained to deter or defeat the exercise of market power. What constitutes an appropriate time period depends on the characteristics and dynamics of the market, as well as on the specific capabilities of potential entrants (100). However, entry is normally only considered timely if it occurs within two years.”).

[199]European Commission, Commission Staff Working Document, Impact Assessment Report Accompanying the document for a Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), SWD(2020) 363 final, December 15, 2020, Part 1/2, 46,


[201]Ibid, 64.

[202]European Commission, Commission Staff Working Document, Impact Assessment Report Accompanying the document for a Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), SWD(2020) 363 final, 48.

[203]European Commission, Regulatory Scrutiny Board Opinion, Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), SEC(2020) 437, para.3, December 10, 2020,

[204]Ibid, 46 referring to quantitative thresholds as “indicators.”

[205]Article 3(3) of the Digital Markets Act.

[206]European Commission, Commission Staff Working Document, Impact Assessment Report Accompanying the document for a Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), SWD(2020) 363 final, December 15, 2020, Part 1/2, 81,

[207]For instance, the small turnovers of the companies having responded to the public consultation prior the DMA are noticeable. Indeed, it is noted by the European Commission that “In terms of annual turnover, more than half of the participating companies/business organizations indicated a turnover of over EUR 50 million per year. 13.3% make an annual turnover of smaller than or equal to EUR 2 million, 3.8% of the respondent revealed an annual turnover of smaller than or equal to EUR 10 Mio, whereas 6.2% specified an annual turnover of smaller than or equal to EUR 50 Mio.” See European Commission, Commission Staff Working Document, Impact Assessment Report Accompanying the document for a Proposal for a Regulation of the European Parliament and of the Council on contestable and fair markets in the digital sector (Digital Markets Act), SWD(2020) 363 final, December 15, 2020, Part 2/2, 26,

[208]European Commission, Commission Staff Working Document, Impact Assessment, Report Accompanying the document for a Proposal for a Regulation of the  European Parliament and of the Council on promoting fairness and transparency for business users of online intermediation services, SWD(2018) 138 final, April 26, 2018, Part 1/2, 27, .

[209]AFEP, “European Commission Consultation on a New Competition Tool,” September 2020,

[210]Interestingly, the DMA explicitly refers these “circumvention risks,” hinting to the fact that they loom large. See Digital Markets Act, para.29 and Article 11.

[211]Digital Markets Act, Article 3(6).

[212]Article 3(6) of the Digital Markets Act.

[213]See, for a general discussion about how the DMA sanctions precautionary antitrust, below Part VIII.


[215]Directive (EU) 2019/1 of the European Parliament and of the Council of December 11, 2018, to empower the competition authorities of the member states to be more effective enforcers and to ensure the proper functioning of the internal market, January 14, 2019,

[216]Article 4 of the Digital Markets Act.

[217]The DMA’s explanatory memorandum indeed states that “a few large platforms increasingly act as gateways or gatekeepers between business users and end users,” that “these gatekeepers have a major impact on, have substantial control over the access to, and are entrenched in digital markets,” “gatekeeper power of these large digital platforms is often misused by means of unfair behaviour” in the Digital Markets Act, 1–2.

[218]One concierge used interchangeably with gatekeeper, see Keith Rosen, The Complete Idiot’s Guide to Cold Calling (New York: Alpha Books, 2004), 208–209 (which intends to “change the nickname of ‘gatekeeper’ to ‘concierge’. It’s assumed that this gatekeeper is an evil ogre before you even get a change to meet them. Maybe the lesson here is to recognize that this negative perception has tainted your approach and how you handle the gatekeeper ... Think about your reaction to the word ‘gatekeeper’. What thoughts does it conjure up for you? Now, think about the word ‘concierge’ ... The concierge secretly wants to help you. Treat the gatekeeper more like a concierge.”). See also Chris Dawber, “Gatekeeper Versus Concierge: Reworking the Complexities of Acute Mental Health Care Through Metaphor,” Social Alternatives 33, No.3 (2014) 53–59, (calling on 55 for the metaphor to be used in the health care field because “the language of the gatekeeper is one of exclusion [whereas] the metaphor of the concierge is provided as a way of envisaging an approach ... that epitomizes a focus on consumers and an inclusive, responsive approach to service provision. A comparison of the gatekeeper and the concierge provides an opportunity to view a ... role that is less defensive, more consumer focused.”).

[219]Chris Dawber, “Gatekeeper Versus Concierge,” 56.

[220]Concierge doctors offer personalized medical services to patients in exchange for an up-front fee. Hotel concierges assist customers with any need they may have, acting as a personal assistant to guests.

[221]See below for a discussion on the alleged unassailability of the gatekeeper’s market power.

[222]Quoted in Chris Dawber, “Gatekeeper Versus Concierge.”

[223]Technical vulnerabilities take the form of outages. Competitive vulnerabilities are illustrated by disruptive innovators disciplining market incumbents. Reputational vulnerabilities take place when incumbents face public scandals which jettison their market positions, such as Facebook’s Cambridge Analytica.

[224]Samuel Stolton, “Commission to contract €600,000 study on gatekeeping power of digital platforms,”, May 11, 2020,

[225]See Damien Gerardin, “What is a digital gatekeeper?” The Platform Law Blog, October 5, 2020, (which argues that “the App Store is a gatekeeper due to the presence of an insurmountable barrier to entry, which that Apple does not allow other app stores on iOS devices.” Whenever a large platform is said to “’control access to critical online services which allow to reach a large category of users,” it is deemed to be a gatekeeper. In that regard, “it may be hard, if at all possible, for many business to compete without the search traffic generated by Google. Similarly, for many resellers access to the Amazon platform is critical to earn a living.” Many instances undermine these beliefs—the success of Etsy being one among many.) See John Ballard, “Why Etsy Stock Surged 301% in 2020,” The Motley Fool, January 7, 2021, (noting that “the online platform is gaining significant market share in e-commerce”). See also Colin Darretta, “How to Win at E-Commerce Without Facebook or Google,” Forbes, August 22, 2019, (reporting experiences and concluding that “we’ve been able to do it all while raising a very modest amount of capital relative to the size of the business we’ve built. This enables you to own more of the business you worked so hard to create and, perhaps more importantly, build a profitable enterprise along the way that will not rely on outside capital or a third-party platform like Facebook or Google to stay alive.”). On the competition in app stores, while Apple’s App Store made the choice to favor monetizing its mobile content over quantity and open access to its App store, it can hardly be said that there is no competition on the very market of app stores. See Statista Research Development, “Number of apps available in leading app stores 2020,” Statista, February 4, 2021, (“whereas Apple and Google dominate the conversation regarding downloads and spending, there are as well as few other players with various degrees of market success and relevance.”).

[226]See for comparison of OS business models, Roland M. Müller, Björn Kjil, and Josef K.J. Martens, “A Comparison of Inter-Organizational Business Models of Mobil App Stores: There is more than Open vs Closed,” Journal of Theoretical and Applied Electronic Commerce Research 6, Issue 2 (2011), 63–76,; Salman Q. Mian, Jose Teixeira, and Eija Koskivaara, “Open-source software implications in the competitive mobile platforms market”; Tomas Skersys et al., (Eds.), Building the e-World Ecosystem, I3E (Berlin: Springer, 2011), 110–128, (“The technological developments in various mobile device platforms has eventually introduced tough competition, with eventually consumer winning in the end.”).

[227]Digital Markets Act, 6.

[228]Aoife White, “EU’s Vestager Defends Dual Role as Antitrust and Tech Chief,” Bloomberg Tax, October 16, 2019,; European Parliament, Hearing of Executive Vice President-designate Margrethe Vestager, Press Releases, October 8, 2019, (reporting that “MEPs also brought up the potential conflict of interest that Ms Vestager could face while executing her wide portfolio, especially on ensuring competition and encouraging the development of digital companies.”) See also European Parliament, “Hearing of Margrethe Vestager, Executive Vice President-designate, Europe fit for the Digital Age: Opening statement by Margrethe Vestager,” October 8, 2019,


[230]Dave Anderson, “Confirmed: The Von der Leyen Commission featuring a more powerful Commissioner Vestager,” Lexology, December 2, 2019, (reporting that “many individuals believe the dual role creates a conflict of interest. Wearing her first hat, Vestager is the champion of Europe’s digital economy, while her other hat deems her the Union’s competition law enforcer. In respect of the latter hat, it is no secret that in targeting the big tech companies Vestager has collected a few political enemies along the way. Trying to reconcile the new Commission’s clear move toward an industrial policy push (including the drive for EU “Digital Sovereignty”), with the competition policy side of her job, could prove to be a tricky tightrope act for Vestager.”).

[231]Javier Espinoza, “Vestager wans Big Tech she will move beyond competition fines,” Financial Times, October 8, 2019,

[232]European Commission, “European fit for the Digital Age: Commission proposes new rules for digital platforms,” Press Release, December 15, 2020, wherein Vestager publicly stated, “The two proposals serve one purpose: to make sure that we, as users, have access to a wide choice of safe products and services online. And that businesses operating in Europe can freely and fairly compete online just as they do offline.”

[233]Article 5(a), Digital Markets Act.

[234]Thomas Eisenmann, Geoffrey Parker, and Marshall Van Alstyne, “Platform Envelopment,” 32 Strategic Management Journal 12 (2011) (in which the term “envelopment” was “entry by one platform provider into another’s market by bundling its own platform’s functionality with that of the target’s so as to leverage shared user relationships and common components”); Daniele Condorelli and Jorge Padilla, “Harnessing Platform Envelopment in the Digital World,” Journal of Competition Law & Economics, 1–45 (2020).

[235]Sebastian Hermes et al., “A Taxonomy of Platform Envelopment: Revealing Patterns and Particularities,” AMCIS 2020 Proceedings, 17,

[236]For instances, see ibid at 4.

[237]This is mentioned in the “or with personal data from third-party services” part of the prohibited practice.

[238]John Quelch and David Hardin, “Brand Versus Private Labels: Fighting to Win,” Harvard Business Review, January–February 1996 (noting that “in European supermarkets, higher private-label sales result in higher average pretax profits. U.S. supermarkets average only 15% of sales from private labels; they average 2% pretax profits from all sales. By contrast, European grocery stores such as Sainsbury’s, with 54% of its sales coming from private labels, and Tesco, with 41%, average 7% pretax profits”); Stephen J. Hoch and Shumeet Banerji, “When Do Private Labels Succeed?” MIT Sloan Management Magazine, July 15, 1993 (explaining private-label products’ success mostly due to high-quality and low-variability levels). See also Michael S. Pepe, Russell Abratt, and Paul Dion, “Competition advantage, private-label brands, and category profitability,” 28 Journal of Marketing Management (2012), 1–2, 154–172; Sandra Maria Correia Loureiro, “Exploring the attractiveness of manufacturer brands and retailer own-brands in supermarket context,” 45 International Journal of Retail & Distribution Management 10 (2019), 1095–1113 (which discusses the brand image as the most effective feature for brand reputation); Rasha H.A. Mostafa and Reham I. Elseidi, “Factors affecting consumers’ willingness to buy private label brands (PLBs): Applied study on hypermarkets,” 22 Spanish Journal of Marketing 3 (2018), 341–361 (demonstrating that store image is crucial in having consumers purchasing private-label products).

[239]Quelch and Hardin, “Brand Versus Private Labels” (noting “private-label market share generally goes up when the economy is suffering and down in stronger economic periods”); Louis Biscotti, “Private Label Soars During Pandemic, Battling Brands,” Forbes, November 23, 2020, (noting “private labels was doing well pre-pandemic. Thanks to COVID-19, it’s accelerating, although a rush to comfort has sent many consumers back to big brands”).

[240]European Commission, “Antitrust: Commission opens investigations into possible anticompetitive conduct of Amazon,” Press Release, July 17, 2019, wherein Commissioner Vestager argued that “I have therefore decided to take an awfully close look at Amazon’s business practices and its dual role as marketplace and retailer, to assess its compliance with EU competition rules.” See Pablo Ibanez Colomo, “The Commission sends Amazon an SO: the rise of common carrier antitrust,” Chillin’ Competition, November 10, 2020 (which notes that these investigations “signal the rise of common carrier antitrust,” a questionable concept in antitrust, as discussed in Part IV).

[241]European Commission, “Antitrust: Commission opens investigations into possible anticompetitive conduct of Amazon,” Press Release, July 17, 2019. See also Simon Van Dorpe, “EU antitrust probe targets Amazon’s ‘dual role,’” Politico, July 17, 2019.

[242]European Commission, “Antitrust: Commission sends Statement of Objections to Amazon for the use of non-public independent seller data and opens second investigation into its e-commerce business practice,” Press Release, November 10, 2020 (along with the issuance of a Statement of Objections on Amazon’s use of marketplace seller data, this press release informs the opening of investigations into Amazon practices about its ‘Buy Box’ and Prime label).

[243]Glauber C. Batista et al., “A Taxonomy Model for Single sign-on Oriented towards Cloud Computing, Proceedings of the 8th International Conference on Cloud Computing and Services Science” (2019), 573–581,

[244]Ibid at 580 (noting that “the core characteristic of SSO solutions is to provide a unique, system-wide identifier to each user. Several SSO-enabled services can then rely on this identifier, providing pervasive authentication and authorization for its users”).

[245]Ibid at 574.

[246]Steve Dent, “Microsoft Authenticator can now store and autofill mobile device passwords,” Engadget, December 16, 2020; Microsoft, “Enable password less sign-in with the Microsoft Authenticator app,” November 11, 2020. For a general discussion, see Joseph Bonneau et al., “The Quest to Replace Passwords: A Framework for Comparative Evaluation of Web Authentication Schemees,” IEEEE Symposium on Security and Privacy (2012), 553–567,; Jan Krämer, Daniel Schnurr, and Michael Wohlfarth, “Winners, Losers, and Facebook: The Role of Social Logins in the Online Advertising Ecosystem,” 65 Management Science 4 (2019), 1678–1699; Furkan Alaca and Paul C. van Oorschot, “Comparative Analysis and Framework Evaluating Web Single Sign-on Systems,” 53 ACM Computing Surveys 5, Article 2 (2020), 1–34. Ironically, SSO can be integrated with a company called … Gatekeeper. See Gatekeeper, “Managing SSO Integration,” (2020).

[247]Built on privacy concerns, Apple launched its own SSO button and made it mandatory for app developer to integrate such a button whenever similar buttons from Google or Facebook are present. This obligation is necessary given the app developer’s reluctance not to collect data derived from SSOs. The announcement was made the same day the U.S. Justice Department was authorized to investigate Apple for potential violations of antitrust laws. See Jason Abbruzzese and Allan Smith, “Apple’s enormous power play comes in a small privacy feature,” NBC News, June 4, 2019; Casey Newton, “With an antitrust case looming, Apple’s new login tool is tempting fate,” The Verge, June 5, 2019; “U.S. Justice Dept considering Apple probe,” Reuters, June 3, 2019. Nevertheless, Apple’s SSO may not fall within the remit of the prohibited practice because it is built on the objective not to “combine personal data.”

[248]Article 5.1(f) of the GDPR requires personal data to be “processed in a manner that ensures appropriate security of the personal data, including protection against unauthorized or unlawful processing and against accidental loss, destruction or damage, using appropriate technical or organizational measures (“integrity and confidentiality”),” whereas Article 6.1(a) conditions lawfulness of data processing to the fact that “the data subject has given consent to the processing of his or her personal data for one or more specific purposes.” Therefore, SSOs are always GDPR-compliant whenever they both ensure security and are consent based.

[249]Article 7 of the GDPR lays down the condition for free consent to data combination, namely, that the user’s consent is presented “clearly distinguishable from other matters,” “intelligible and easily accessible form,” and consent can be withdrawn “at any time.” More generally, consent-based use of personal data is derived from Article 8.2 of the EU Charter of Fundamental Rights which reads: “[personal] data must be processed fairly for specified purposes and based on the consent of the person concerned or some other legitimate basis laid down by law. Everyone has the right of access to data which has been collected concerning him or her, and the right to have it rectified.” Equally, Article 16(1) of the TFEU enshrines the “right to the protection of personal data .” See, more generally, Eline Chivot and Daniel Castro, “What the Evidence Shows About the Impact of the GDPR After One Year,” Center for Data Innovation, June 17, 2019; Written Statement For The Record David Hoffman, Associate General Counsel and Global Privacy Officer, Intel Corporation, United States Senate Committee on the Judiciary Hearing on “GDPR & CCPA: Opt-ins, Consumer Control, and the Impact on Competition and Innovation” (March 12, 2019) (“The notice-and-consent model is fatally flawed; it must be replaced. People do not have time to read privacy policies for every interaction where their personal data will be collected and used. Even if they did read these policies, it is unlikely they would be able to understand how this data will be used”).

[250]Michal S. Gal and Oshrit Aviv, “The Competitive Effects of the GDPR,” 16 Journal of Competition Law & Economics 3 (2020), 349–391, (at 353 “[I]t is often costly, and sometimes impossible, to obtain informed consent from data subjects to have their data shared with the data receiver, as may be required by the GDPR. This effect is strengthened in a multiproduct and/or multiservice environment, in which consent is required for each different use of the data.”); James David Campbell, Avi Goldfarb, and Catherine E. Tucker, "Privacy Regulation and Market Structure, 24(1) Journal of Economics & Management Strategy 47 (2015) (on 48–49, “[R]ather than increasing competition, the nature of transaction costs implied by [opt-in] privacy regulation suggests that privacy regulation may be anti-competitive ... [I]n some cases where entry had been profitable without regulation, [some firms] will choose not to enter.”). See also Gal and Aviv, “The Competitive Effects of the GDPR,” 370 (“the magnitude of these [GDPR] obligations increase only up to the point where the maximal requirements are applied (due to the dataset’s size and/or sensitivity). Beyond that point, the marginal costs of meeting legal requirements might fall due to economies of scale and scope. This, in turn, may benefit large firms relative to medium-sized ones which must still comply with such requirements. Put differently, the larger the firm, the lower its per-datum compliance costs, relative to smaller firms which must also comply with similar requirements.”); Campbell, Goldfarb, and Tucker, "Privacy Regulation and Market Structure.”

[251]Article 5(b), Digital Markets Act.

[252]Refunding the price difference can be said to be an ex post price adjustment strategy while most favored customer clause can be said to be tantamount to an ex ante price adjustment strategy in that consumers are offered the lowest price, contractually (MFC) or financially (refund). These marketing strategies are so widespread that some supermarkets introduced the “double the difference” offer whereby they refund consumers twice the price difference to both assert their lowest prices and deter consumers from shopping at rivals. See Russel Parsons, “Tesco pulls ‘double the difference’ offer,” Marketing Week, April 28, 2011 (alleging that the offer ended because some consumers abused the system). Massimo Motta, Competition Policy: Theory and Practice, Cambridge: Cambridge University Press (2004), 157, wherein Motta distinguishes between “retroactive” MFC (ex post) and “contemporaneous” MFC (ex ante).

[253]European Commission, “Antitrust: Commission accepts commitments from Amazon on e-books,” Press Release, IP/17/1223, May 4, 2017 (the commitments were lauded by Commissioner Vestager as increasing consumer choice and increased competition).

[254]MFC clauses are exempted from antitrust scrutiny in Europe whenever the Vertical Agreement Block Exemption (“VABE”) applies—meaning when market shares of each of the parties to the vertical agreement are below 30% and when there is no “hardcore” restriction in the agreement such as resale price maintenance. See Commission Regulation, 330/2010 of April 20, 2010, on the application of Article 1010(3) of the Treaty on the Functioning of the European Union to categories of vertical agreements and concerted practices, L102/1 (2010). European NCAs have investigated MFC for online hotel bookings, insurance price comparison websites, Amazon marketplace, Apple e-books, Amazon e-books. See Philippe Chappate and Keery O’Connell, “European Union—E-commerce: Most Favoured Nation Clauses,” Global Competition Review, October 15, 2019.

[255]Jonathan B. Baker, “Vertical Restraints with Horizontal Consequences: Competitive Effects of ‘Most Favored-Customer’ Clauses,” 64 Antitrust Law Journal 517 (1996); Pinar Akman and Morten Hviid, “A Most-Favoured-Customer Clause with a Twist,” 2 European Competition Journal 57 (2006); Jonathan B. Baker and Judith A. Chevalier, “The Competitive Consequences of Most-Favored-Nation Provisions,” 27 Antitrust Magazine, 20 (2013); Steven C. Salop and Fiona Scott Morton, “Developing an Administrable MFN Enforcement Policy,” 27 Antitrust Magazine 15 (2013).

[256]The hypothetical anticompetitive effects of MFC clauses are not generally convincing. These comprise four types of competitive risks. First, MFC clauses may ease coordination: MFC clauses may help price monitoring and price fixing. The opposite can be argued: Absent an incentive for the platform to monitor and surveil third-party retailers’ prices, its focus will shift toward platform quality and network rather than remain on price surveillance. See Motta, Competition Policy, 157, (“[I]t is not clear that a [MFC] clause would facilitate (tacit or overt) collusion,” and on ft 53, “in other words, I know of no model where [MFC clause] is found to increase sustainability of collusion in an infinite horizon game.”) Second, MFC clauses may dampen competition with parallel accommodating conduct whereby third-party retailers are incentivized to compete less aggressively on prices and anticipate rivals to do similarly. Overall, a less-intense rivalry emerges, the theory goes. However, absent cartelized behavior, it is dubious that no single rival will not take advantage of the economic prospects of stealing the entire business. Third, MFC clauses may raise rivals’ and entrants’ costs since the platform has become too competitive so that no rival or entrant can offer similar services since best prices are already guaranteed by the platform. This disregards the fact that a platform’s attractiveness for business customers does not revolve solely on end users’ prices but on customers’ prices (e.g., fees charged by to hotels) and quality features (e.g., customer service, design, and applications, etc.). Thus, a rival to Airbnb or slashing the franchise fee these platforms charge their customers may attract, ceteris paribus, new customers and, thanks to positive network externalities, end users as well. The fourth and final argument pertains to the increase of the seller’s bargaining power: The story says that a monopolist platform can charge monopoly prices and maintain them above competitive prices through MFC clauses thanks to its bargaining power, which is thus self-reinforced. This argument assumes both a monopoly at the upstream level (platform) and a collective willingness from third-party retailers not to compete against one another (downstream level). Both assumptions make this argument highly unrealistic. These arguments are summed up in Baker and Chevalier, “The Competitive Consequences of Most-Favored-Nation Provisions,” 23–24.

[257]On the inadequacy of the prohibition of MFC clauses with respect to the prohibition of collusive practice under Article 101 TFEU, see Pinar Akman, “A Competition Law Assessment of Platform Most-Favored-Customer Clauses,” 12 Journal of Competition Law & Economics 4 (2016), 781–833, (noting on 784: “[T]he way in which some of the authorities have handled these [MFC] clauses opens up the possibility of creating a legal anomaly in the laws concerning the prohibition of anticompetitive agreements ... This is because the investigations of the NCA into anticompetitive ‘agreement’ are directed against only one party to the agreement, and the decisions taken are addressed to only one of the parties ... [This] creates a mismatch between the authorities’ theory of harm and their action, and thereby raises the question as to whether they are operating on the basis of the correct theory of harm.” In other words, when under antitrust scrutiny, MFC clauses are investigated as collective collusive practices but are prosecuted against one individual platform only. This incoherence weakens the very legal basis underpinning the antitrust enforcement actions.).

[258]On how contractual arrangements can overcome the commitment problems, see Motta, Competition Policy, 344–347.

[259]Baker and Chevalier, “The Competitive Consequences of Most-Favored-Nation Provisions” (notably, it is argued that MFC clauses control for both the “hold up” problems related to investment-specific costs and for the delays in transactions with the “hold out” problems by a transaction cost minimization); Jan Peter van der Veer, “Antitrust Scrutiny of Most-Favoured-Customer Clauses: An Economic Analysis,” 4 Journal of European Competition Law & Practice (2013), 501; Akman, “A Competition Law Assessment” (2016), 781–833.

[260]Ibid (noting at 786 the lack of understanding by regulators of the diverse types of MFC clauses and their respective antitrust implications).

[261]Commission prevented the hotel comparison website to make use of their MFC clauses on its hotel partners but preserved its ability to maintain its price-matching guarantees (PMGs) involving promises by a seller to match the prices of competitors for the same customer. But PMGs generate much larger anticompetitive effects as opposed to MFC clauses. See Press Release, Italian Competition Authority, “Commitments Offered by Closed the Investigation in Italy, France and Sweden” (Apr. 21, 2015),; See also Press Release, Bundeskartellamt, “Bundeskartellamt Issues Statement of Objections Regarding’s ‘Best Price’ Clauses” (Apr. 2, 2015), http://www. html; Press Release, Bundeskartellamt, “Narrow ‘Best Price’ Clauses of Booking Also Anticompetitive” (Dec. 23, 2015), (stating that “best price clauses,” or MFC clauses, are illegal even after reduction of these clauses’ scope, and after the Düsseldorf Higher Regional Court confirmed in January 2015 the Bundeskartellamt’s decision that these clauses by Booking’s competitor HRS were illegal). See also ibid wherein it is argued at 786 that “it is likely that competition law intervention in this market has banned the potentially less anticompetitive clause (MFC clause) thereby pushing the hotels to make more use of their potentially more anticompetitive clause (PMG).”

[262]Consumers may continue to shop on the platforms they know, unaware that the prices they see are no longer the lowest prices available on the Internet or offline.

[263]Against the decisional background, European companies such as the Dutch company may qualify as gatekeepers despite their reluctance to be so considered. See Glenn Fogel, Booking Holdings, Inc., Q3 2020 Results—Earnings Call Transcript, Seeking Alpha, November 5, 2020 (arguing that “there has been questions and speculations that maybe one of the designated gatekeepers, which we firmly believe would be incorrect for a number of reasons. The principal reasons are that the accommodations market in Europe is incredibly open and incredibly competitive ... booked about 7% of all potentially bookable accommodation room nights across all properties on our platform globally in 2019. Across Europe, that number was about 11%.”); Hannah Boland, “ boss on why Brussels is hampering Europe’s tech ambitions,” Telegraph, November 22, 2020; Javier Espinoza, “EU vs Big Tech: Brussels’ bid to weaken the digital gatekeepers,” Financial Times, December 8, 2020 (which reports that companies such as Airbnb and may qualify as gatekeepers).

[264]Massimo Motta, Competition Policy: Theory and Practice, Cambridge: Cambridge University Press (2004), 157.

[265]Motta, Competition Policy, 343 (considering the prohibition of price discrimination is “equivalent to enforcement of the [MFC] clause. This is precisely what happens under the current EU competition law”). It is thus clear that EU competition prohibits one thing (price discrimination) and now its opposite (price transparency).

[266]Ibid, citing the Michelin case as illustration of the antidiscrimination principle in EU competition law. See Case 322/81 NV Nederlandsche Banden Industrie Michelin v Commission, ECLI: EU: C :1983 :313; Case T-203/01, Manufacture française des pneumatiques Michelin v Commission, ECLI: EU: T :2003 :250.

[267]Article 5(c), Digital Markets Act.

[268]European Commission, Proposal for a Regulation of the European Parliament and of the Council on the contestable and fair markets in the digital sector (Digital Markets Act), COM (2020) 842, 2.

[269]For a general discussion, see Gus Hurwitz, “Digital Duty to Deal, Data Portability, and Interoperability,” The Global Antitrust Institute—Report on the Digital Economy (2020), 1024–1059.

[270]Directive (EU) 2015/2366 of the European Parliament and of the Council of November 25, 2015, on payment services in the internal market, amending Directive 2002/65/EC, 2009/110/EC and 2013/36/EU and Regulation (EU) No 1093/2010, and repealing Directive 2007/64/EC, OJL 337, December 23, 2015.

[271]The European Commission has found three broad models of data-sharing: i) an Open Data approach; ii) data monetization on a data marketplace; and iii) data exchange in a closed platform. See European Commission, “Towards a Common European Data Space,” COM (2018), 232 final. More generally, see Heike Schweitzer and Robert Welker, “A Legal Framework for Access to Data—A Competition Policy Perspective” (September 15, 2020). Josef Drexl (ed.), “Data access, consumer interests and public interest,” Forthcoming, available at

[272]Particularly, see Case C-170/13 Huawei [2015] ECLI:EU:C:2015:477 for data access despite standard essential patents.

[273]Crémer, de Montjoye, and Schweitzer, “Competition Policy for the Digital Era,” 9.

[274]Ibid wherein at 10 the data controller who is under data access obligations may be subject to both sector-specific regulations as well as updated competition tools. Such expansion of the traditional antitrust analysis, as well as the disregard for proprietary aspect associated with data accumulation, is justified based on a belief that gatekeepers have “quasi-infrastructural resources, like public sector data” as outlined in Schweitzer and Welker, “A Legal Framework for Access to Data;” Josef Drexl (ed.), “Data access, consumer interests and public interest,” Forthcoming, 27, available at

[275]Digital Markets Act, Article 5(c).

[276]Double marginalization problem arises whenever upstream and downstream market players both exhibit some degree of market power, thereby generating markup above marginal cost. Vertical integration solves the double marginalization problem. See Preston MacAfee and Marius Schwartz, “Opportunism in Multilateral Vertical Contracting: Nondiscrimination, Exclusivity, and Uniformity,” 84 American Economic Review (1994), 210–230 ; Esther Gal-Or, “Duopolistic Vertical Restraints,” 35 European Economic Review (1991), 1237–1253; Oliver Hart and Jean Tirole, “Vertical Integration and Market Foreclosure, Brookings Papers on Economic Activity: Microeconomics” (1990), 205–286; Francis Lafontaine and Margaret Slade, “Vertical Integration and Firm Boundaries: The Evidence,” 44 Journal of Economic Literature (2007), 629. One of the recent cases wherein the elimination of double marginalization problem was at the core of the antitrust debate was the approved merger between AT&T and Werner wherein Judge Leon seminally defined double marginalization as “the situation in which two different firms in the same industry, but at different levels in the supply chain, each apply their own markups (reflecting their own margins) in pricing their products. Those ‘stacked’ margins are both incorporated into the closing price that consumers must pay for the product. By vertically integrating two such firms into one, the merged company can ‘shrink that total margin so there's one instead of two,’ leading to lower prices for consumers,” in United States v. AT&T, Inc., No. 17-cv-2511 (RJL), 2018 WL 2930849 (D.D.C. June 12, 2018).

[277]Schweitzer and Welker, “A Legal Framework for Access to Data;” Josef Drexl (ed.), “Data access, consumer interests and public interest,” Forthcoming, available at at 28.

[278]Article 5(d), Digital Markets Act.

[279]See Jon Porter, “Epic Games brings Apple fight to the EU with new antitrust complaint,” The Verge, February 17, 2021, In the United States, a similar complaint has been filed, see Epic Games Inc. v Apple Inc., 422 F. See also Kyle Orland, “Valve gets dragged into Apple and Epic’s legal fight over Fortnite,” Ars Technica, February 19, 2021,

[280]Article 5(e), Digital Markets Acts.

[281]Neil J. Rubenking and Ben Moore, “The Best Password Managers for 2021,”, February 17, 2021,

[282]Josh Sisco, “Apple’s App Sign-in Button Becomes Hot-Button Issue in U.S. Antitrust Probe,” The Information, February 23, 2021,

[283]Article 3(f), Digital Markets Act.

[284]Recital 52, Digital Markets Act.

[285]European Commission, Antitrust: Commission opens investigations into Apple practices regarding Apple Pay, June 16, 2020,


[287]Recital 52, Digital Markets Act.

[288]Autoriteit Consument & Markt, Market study into mobile app stores. Report, April 11, 2019, 7,

[289]Ibid, (noting on 6 that [The Dutch Competition Authority] also spoke to Apple and Google and asked them about their views on certain topics, then used terms and conditions and several processes. Apple and Google point top such matters as integrity, safety, and the quality of the app stores and the ecosystems, the investments they made to develop the app stores, and the opportunities the app stores give to app providers. According to Apple, favoring their own apps over third-party apps would not be rational. Apple wants to offer the best services possible to its users and therefore has no incentive to refuse a third party that offers a higher quality app.)

[290]ETSI website, History,; W3C website, Standards, (where it is stated that “W3C develops these technical specifications and guidelines through a process designed to maximize consensus about the content of a technical report, to ensure high technical and editorial quality, and to earn endorsement by W3C and the broader community”); Internet Engineering Task Force, Internet Standards,; European Commission, “European Multi Stakeholder Platform on ICT Standardisation,” Shaping Europe’s Digital Future, October 29, 2020,

[291]See, more generally, Ian Brown, “The Technical Components of Interoperability as a Tool for Competition Regulation,” Open Forum Academy, November 2020,, 11 et seq.

[292]On the definition of tying and bundling, see European Commission, “Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct by dominant undertakings,” February 24, 2009, 2009/C, para.48, (wherein it is noted that “‘Tying’ usually refers to situations where customers that purchase one product (the tying product) are required also to purchase another product from the dominant undertaking (the tied product). Tying can take place on a technical or contractual basis. ‘Bundling’ usually refers to the way products are offered and priced by the dominant undertaking. In the case of pure bundling the products are only sold jointly in fixed proportions. In the case of mixed bundling, often referred to as a multi-product rebate, the products are also made available separately, but the sum of the prices when sold separately is higher than the bundled price.”).

[293]For a seminal case in the EU decisional practice, see Eurofix-Bauco v Haiti (1988) OJ L65/19, para.4.

[294]See Michael D. Whinston, “Tying, Foreclosure, and Exclusion,” American Economic Association 80 (1990) 837–589,

[295]This is illustrated with Google tying Google Chrome (delivered for free) with Google Search (also delivered for free but generating advertising revenue). Such cross-tying ensuring cross-viability of the platform was nevertheless central to the Google Android fining decision in 2018 currently under judicial review. See European Commission, Google Android, Case AT.40099, July 18, 2018,

[296]See Gregory S. Crawford, Patrick Rey, and Monika Schnitzer, “An Economic Evaluation of the EC’s Proposed ‘New Competition Tool,’” Report by the Economic Advisory Group on Competition Policy, Luxembourg: Office of the European Union, 18 (wherein unbundled access is classified as the most invasive and most costly regulatory interventions).

[297]Jean-Charles Rochet and Jean Tirole, “Platform Competition in Two-Sided Markets,” Journal of European Economic Association 1 (2003), 990–1029,

[298]Mahdi Tavalaei, Juan Santalo, Annabelle Gawer, “What is the Impact of Within-Platform Competition in Two-Sided Markets?” (2017), 992,

[299]Article 5(g), Digital Markets Act.

[300]Natasha Lomas, “After Facebook’s news flex, Australia passes bargaining code for platforms and publishers,” TechCrunch, February 25, 2021,; Mathieu Rosemain, “French antitrust investigators say Google breached its orders on talks with news publishers,” Reuters, February 23, 2021,; French Competition Authority, “Related rights: the Autorité has granted requests for urgent interim measures presented by press publishers and the news agency AGP” (Agence France Presse), April 9, 2020, The European regulatory framework tackles news aggregators under the regulation of protection “neighbouring rights,” see Directive (EU) 2019/790 of the European Parliament and of the Council of April 17, 2019, on copyright and related rights in the Digital Single Market and amending Directives 96/9/EC and 2001/29/EC, OJ L 130/92, May 17, 2019. See also Bernt Hugenholtz, “Neighbouring Rights are Obsolete,” International Review of Intellectual Property and Competition Law 50 (2019), 1006–1011 (concluding that “All this makes the new press publisher’s right an inherently unbalanced right in the bad-old tradition of no-threshold neighbouring rights. This concept, I argue, is passe´.”).

[301]Aurelien Portuese, “The Digital Coase Theorem and the News,” Competition Policy International, March 12, 2021,

[302]Stefanie Fogel, “The 10 Best News Aggregators of 2021,” Lifewire, January 3, 2020,

[303]Recital 42, Digital Markets Act.

[304]The “nirvana fallacy” inherent to the antitrust counterfactual claims was first advanced by Harold Demsetz, “Information and Efficiency: Another Viewpoint,” The Journal of Law & Economics 12 (1969), 1–22,

[305]These claims reflect those formulated in the UK Competition Authority Report. See Competition & Markets Authority, “Online Platforms and digital advertising. Market study final report,” July 1, 2020, (wherein it is claimed that “Google and Bing’s search prices suggest that Google’s prices are 30–40% higher on desktop and mobile when comparing like-for-like search terms” while ignoring the difference in viewability of Google and Bing and ignoring the associated extra costs inherent to the management of a larger search engine and a larger range of (often free) services.).

[306]Article 6, Digital Markets Act.

[307]Article 15(4), Digital Markets Act.

[308]Article 7(2), Digital Markets Act.

[309]This is referred to as “Option 2 - Partially flexible framework of designation and updating of obligations, including regulatory dialogue for the implementation of some” in Digital Markets Act, 9.

[310]Recital 58, Digital Markets Act.

[311]Scott Bicheno, “Europe dabbles with big data protectionism,” Ars Technica, November 26, 2020,

[312]Article 6(a), Digital Markets Acts.

[313]European Commission, “Antitrust: Commission sends Statement of Objections to Amazon for the use of non-public independent seller data and opens second investigation into its e-commerce business practices,” Press Release, November 10, 2020,


[315]Article 6(1)(b), Digital Markets Act.

[316]Article 6(1)(c), Digital Markets Act.

[317]TheAppSolutions, 9 Alternative Android App Stores,


[319]European Commission, “Antitrust: Commission opens investigations into Apple’s App Store rules,” Press Release, June 16, 2020,

[320]Friso Bostoen, “Epic v Apple (3): two perspective on app stores’ 30% commission fee,” Lexxion, November 3, 2020, A similar lawsuit was launched in 2019, see Class action complaint of Donald R. Cameron and Pure Sweat Basketball, Inc., against Apple Inc., before the US District Court for the Northern District of California, available at

[321]Article 6(1)(d), Digital Markets Act.

[322]Aurelien Portuese, “Two Pitfalls Lawmakers Must Avoid in House Antitrust Subcommittee Hearings on ‘Reviving Competition’,” Innovation Files, February 24, 2021, /publications/2021/02/24/two-pitfalls-lawmakers-must-avoid-house-antitrust-subcommittee-hearings.

[323]Pedro Caro de Sousa, “What Shall We Do About Self-Preferencing?” Competition Policy International, June 2020, (noting that “it is normal for companies to promote their own products over those of their competitors, and, in many cases, this can lead to efficiencies.”).

[324]Portuese, “The Antitrust Prohibition of Favoritism.”

[325]European Commission, Google Search (Shopping), AT.39740, June 27, 2017,

[326]See, for instance, Michael A. Salinger, “Self-Preferencing,” Report on the Digital Economy (Arlington, VA: Global Antitrust Institute, November 2020), 329–368,; Pablo Ibanez Colomo, “Self-Preferencing: Yet Another Epithet in Need of Limiting Principles” (2020), 43, 417–446,

[327]Article 6(1)(e), Digital Markets Act.

[328]Recital 50, Digital Markets Act.

[329]European Commission, Google Android, AT.40099, July 18, 2018,

[330]Apple, “Switching to iPhone is easy,”

[332]JR Raphael, “iPhone to Android: The ultimate switching guide,” Computerworld, February 7, 2020,; Jan Schroder, “"Apple to Android: Breaking Up’s Not So Hard To Do,”, December 25, 2020,; Ryan Whitwam, “Switching from Android to iPhone: Here’s where Apple’s phone wins and fails,” PCWorld, March 27, 2020,

[333]Recital 51, Digital Markets Act.

[334]Article 6(1)(f), Digital Markets Act.

[335]Recital 52, Digital Markets Act.

[336]Article 6(1)(g), Digital Markets Act.

[337]Recital 52, Digital Markets Act.

[338]Mathieu Rosemain, “Google’s $76 million deal with French publishers leaves many outlets infuriated,” Reuters, February 12, 2021, See also, Siladitya Ray, “Google Nears Deals With Major Australian News Publishers as Lawmakers Debate Plans to Make Firms Pay For News,” Forbes, February 15, 2021,

[339]Article 6(1)(h), Digital Markets Act.

[340]Recital 54, Digital Markets Act.


[342]Digital Markets Act, Article 6(1)(i).

[343]Recital 55, Digital Markets Act.

[344]Digital Markets Act, Article 6(1)(j).

[345]For a general discussion about the relevance of the applicability of the FRAND regime to digital platforms under EU competition law, see Mathew Heim and Igor Nikolic, “A FRAND Regime for Dominant Digital Platforms,” Journal of Intellectual Property, Information Technology and Electronic Commerce Law 10 (2019), 38–55, (which acknowledges that “in addition to more ‘classic’ competition concerns, new issues, not traditionally within the competition policy space, are increasingly being voiced.”); See also, Oscar Borgogno and Giuseppe Colangelo, “Disentangling the FRAND Conundrum,” DEEP-IN Research Paper (2019),

[346]Recital 56, Digital Markets Act.

[347]Luís Cabral et al., “The EU Digital Markets Act A Report from a Panel of Economic Experts” (Luxembourg: Publications Office of the European Union, 2021),

[348]Ibid, 29.

[349]Ibid, 29 (noting “unfair = harmful. This raises several questions: Harmful compared to what standard? Harmful to whom? ... Network externalities generate welfare gains for society but access conditions to the network may penalise some users.”).

[350]Ibid, 29 (noting “unfair = unequal. Again the questions come up: compared to what standard of equality? How do we operationalise this? For example, is a 70/30 sharing rule for ad revenue between publishers and intermediaries, or between app developers and app store owners, a fair deal? Should consumers, whose time is consumed, participate in the equation for sharing advertising revenue?”).

[351]Ibid, 30 (noting that “to operationalize FRAND terms, gatekeepers might legally and technically separate vertical services on top of the platform from the infrastructure that constitute the platform.”).

[352]“History of Google Algorithm Updates,”,; Mike Shaw, “Why Google is the Best Search Engine (and Why Businesses Should Care),” Towermarketing, June 15, 2020,

[353]See, for instance, European Commission, Motorola—Enforcement GPRS standard essential patents, AT.39985, April, 29, 2014,; European Commission, Samsung—Enforcement of UMTS standard essential patents, COMP/C-39939, April 29, 2014,

[354]Case 170/13, Huawei Technologies Co Ltd v ZTE Corp and ZTE Deutschland GmbH, ECLI:EU:C:2015:477.

[355]European Commission, Communication from the Commission to the European Parliament, the Council and the European Economic and Social Committee, COM(2017) 712 final, November 29, 2017, (which lays down “general principles for FRAND licensing terms for SEPs”).

[356]Case 170/13, Huawei Technologies Co Ltd v ZTE Corp and ZTE Deutschland GmbH, ECLI:EU:C:2015:477, para.53.

[357]Oscar Borgogno, Giuseppe Colangelo, “Disentangling the FRAND Conundrum,” DEEP-IN Research Paper (2019), 3,

[358]Case 170/13, Huawei Technologies Co Ltd v ZTE Corp and ZTE Deutschland GmbH, ECLI:EU:C:2015:477, para.49.

[359], “French gvt offices ditch Google for local search engine Qwant in another possible anti-US move,”, November 22, 2018,; Amanda Meade, “Microsoft’s Bing ready to step in if Google pulls search from Australia, minister says,” The Guardian, February 1, 2021,; Matt Southern, “DuckDuckGo Hits New Record: 100 Million Searches Per Day,” Searchenginejournal, January 17, 2021,

[360]Article 6(1)(k), Digital Markets Act.

[361]Impact Assessment Report, Part 2/2, Digital Markets Act, 31.

[362]Recital 57, Digital Markets Act.

[363]Surabhi Agarwal, “Startup founders bat for an Indian app store,” The Economic Times, October 1, 2020, See also, Samreen Ahmad and Peerzada Abrar, “As demands grow for a local app store, analysts see long road ahead,” The Economic Times, October 3, 2020,

[364]The history reveals the fall of alternative mobile OS due to less efficient and less popular smartphone devices, see Charticulator, “Market share of mobile operating systems,”; Scott D. Anthony, “The Real Lessons From the Fates of BlackBerry and Nokia,” Bloomberg, October 8, 2013,; Travis Hessman, “The Road to Failure: Nokia, Blackberry and…Apple?” Industry Week, September 6, 2013, (considering that “they fall because they didn’t stray ... To stay alive, to stay relevant, top companies have to continue to invest in groundbreaking disruptive technologies.”).

[365]For a general comparative study, see Aadityia Jain, Samridha Raj, and Bala Buksh, “A Comparative Study of Mobile Operating Systems with Special Emphasis on Android OS,” International Journal of Computer & Mathematical Sciences 5 (2016) 5–14,

[366]Melanie Schaffer, “Is BlackBerry Back? With New Partners The Company Is Ready For a Closer Look,” February 14, 2021,; Brien Posey, “Why Windows Phone Is Dead, But Not Completely Gone,” Redmond Magazine, July 7, 2020,; Brad Moon, “Nokia is Preparing to Come Roaring Back in the New Decade,” Investor Place, December 31, 2019,; Omar Zahran, “Life After Death: The Resurgence of Nokia,” Medium, April 11, 2020,

[367]European Commission, “What is the Innovation Principle?” November 25, 2019,; Council of the European Union, “Outcome of Proceedings,” 126383/1/20 REV 1, November 16, 2020, See also Aurelien Portuese and Julien Pillot, “The Case for an Innovation Principle.” 214–237,

[368]See, as illustrations, Crémer, de Montjoye, and Schweitzer, “Competition Policy for the Digital Era,” (noting on 42 that “In a digital world, where the future is more uncertain and less understood, there will be underenforcement if we insist that the harm be identified with a high degree of probability. In some cases, one may be able to use the error-cost framework that we discuss below to compute what economists would call the “expected” consumer welfare, but in many cases this will be too complicated” and noting at 50 that “The digital economy is characterised by a high degree of innovation and rapid changes in the markets. Furthermore, its economics are new and very different from those of “standard” industries. As a consequence, it is clear that there will be uncertainty about the consequences of any competition policy intervention or non-intervention”); Autorité de la Concurrence/Bundeskartellamt, “Algorithms and Competition” (Paris: Autorité, November 2019), 44, (that “so far little is known about the actual real-world use of advanced techniques for pricing purposes. In particular, it remains to be seen if and how pricing algorithms can arrive at some kind of communication. This uncertainty is partly caused by the fact that the exact nature of potential ‘algorithmic communication’ cannot be anticipated”); Stigler Committee on Digital Platforms (2019) Final Report, Stigler Center for the Study of the Economy and the State. University of Chicago, available at: (noting on 32, “Because the cost of false negatives is high and there is uncertainty, the public interest requires the [Digital Authority] to take a more interventionist approach in these settings”); UNCTAD, “The ‘New’ Digital Economy and Development,” UNCTAD Technical Notes on ICT for Development N°8 (Geneva: UNCTAD, 2019), 2,; BRICS Report, “Digital Era Competition: A BRICS View” (Moscow, Russia: BRICS Competition Law and Policy Centre, 2019), 166, (noting that “an extra uncertainty appears in case of [multi-sided platforms] due to the network effect when the number of users becomes an important determinant of the platform efficiency (whatever it is measured).”).

[369]Digital Markets Act, 12 and 21, respectively.

[370]European Commission, Inception Impact Assessment, New Competition Tool, Ares(2020)2877634, June 4, 2020,

[371]Richard J. Gilbert, Innovation Matters. Competition Policy for the High-Technology Economy (Cambridge, MA: MIT Press, 2020), 15–16.

[372]For an illuminating discussion, see Bruce D. Hoffman and Joseph J. Simons, “Known unknowns: Uncertainty and its implications for antitrust policy and enforcement in the standard-setting content,” The Antitrust Bulletin 57 (2012), 89–116, For a general discussion, see Ray Pawson, Geoff Wong, and Lesley Owen, “Known Knows, Know Unknowns, Unknown Unknowns: The Predicament of Evidenced-Based Policy,” American Journal of Evaluation 32, DOI: 10.1177/1098214011403831 citing Donald Rumsfeld having stated, “There are known knowns. These are things we know that we know. There are known unknowns. That is to say, there are things that we now know we don’t know. But there are also unknown unknowns. These are things we do not know we don’t know.”

[373]See for instance the Motta and Peitz, “Intervention triggers and underlying theories of harm,” 4 (advising that “it should be kept in mind that markets are complex, and that different market features and firms’ conducts interact to determine market outcomes in ways that are not always is to foresee. In other words, uncertainty may exist about the impact of a concrete intervention. We submit that the [European Commission] should carry out a cost-benefit analysis that allows for probabilistic assessments.”).

[374]Ibid (noting at 48 that “In the cases where competition is at risk (rather than already seriously affected) so that consumer harm is likely to arise in the future, the additional difficulty is that not only the counterfactual that would arise following an intervention but also the outcome absent intervention has to be predicted.”).

[375]Digital Markets Act, 10.

[376]European Commission, Inception Impact Assessment, New Competition Tool, Ares(2020)2877634, June 4, 2020,

[377]The fundamental assumptions underpinning the DMA according to which market concentration is detrimental and large companies must be avoided appear to be unsubstantiated and embedded in the widespread populist techlash. See, Bighelli et al., “Increasing market concentration in Europe,”,Increasing%20market%20concentration%20in%20Europe%20is%20more%20likely%20to%20be,than%20a%20cause%20for%20concern&text=Aggregate%20firm%20concentration%20has%20increased,productivity%20at%20the%20sector%20level (stating, “Increasing market concentration in Europe should not necessarily be seen as a cause for concern related to a weaker competitive environment. This has important consequences for antitrust and industrial policy, which must carefully evaluate the costs and benefits of increasing market concentration.”); Robert D. Atkinson and Michael Lind, “Big Is Beautiful. Debunking the Myth of Small Business” (Cambridge, MA: MIT Press, 2018); Aurelien Portuese and Joshua D. Wright, “Antitrust Populism: Towards a Taxonomy,” Stanford Journal of Law, Business & Finance 21 (2020)

[378]Digital Markets Act, 6 (noting that the Commission can designate gatekeepers “the providers of core platform services that exhibit the same or similar risks for fairness and contestability of the market and at the same time guarantees that the obligations apply to the relevant providers of core platform services only.”).

[379]Recital 13, Digital Markets Act.

[380]Aurelien Portuese and Julien Pillot, “The Case for an Innovation Principle: A Comparative Law & Economics Analysis,” Manchester Journal of International Economic Law 15 (2018) 214–237,

[381]Evidence of this risk-based approach inherent to the DMA as opposed to the traditional evidenced-based approach of competition law is present in Recital 32 wherein it is stated that “rules are needed to address the risk of harmful effects of unfair practices imposed by gatekeepers”; Recital 64, wherein the Commission considers, “Changes to the structure of an undertaking as it existed before the systematic non-compliance was established would only be proportionate where there is a substantial risk that this systematic noncompliance results from the very structure of the undertaking concerned”; Article 22, wherein the Commission may intervene, “In case of urgency due to the risk of serious and irreparable damage for business users or end users of gatekeepers.”; and Recital 27, wherein it is sought “to avoid the qualified risk of unfair conditions and practices.”

[382]Margrethe Vestager, “Statement by Executive Vice-President Vestager on the Commission proposal on new rules for digital platforms,” December 15, 2020,

[383]See, for a general discussion of the consumer welfare standard in the EU competition, Renato Nazzini, The Foundations of European Union Competition Law: The Objectives and Principle of Article 102 (Oxford: Oxford University Press, 2011), 44–45.

[384]The European Ordoliberalism which promotes consumer choice by protecting companies’ right to economic freedom has lost preeminence with the rise of the Chicago School. With the more recent demise of the Chicago School, Ordoliberalism and its inherent protection of rivals at the expense of the efficiency principle gains momentum again. See Peter Behrens, “The ‘Consumer Choice’ Paradigm in German Ordoliberalism and its Impact Upon EU Competition Law,” Europa-Kolleg Hamburg, Discussion Paper No.1/14 (2015),; Kevin Coates, Competition Law and Regulation in Technology Markets (Oxford: Oxford University Press, 2011), Chap. 2. See also Neil W. Averitt and Robert H. Lande, “Using the ‘Consumer Choice’ Approach to Antitrust Law,” Antitrust Law Journal 74 (2007), 175–264,

[385]For a discussion, see Thomas J. Rosch, “Can consumer choice promote trans-Atlantic convergence of competition law and policy?” in Paul Nihoul, Nicolas Charbit, and Elisa Ramundo (Eds.), Choice — A New Standard for Competition Law Analysis? (Paris: Concurrences, 2016), 265–282.

[386]Indeed, the “as-efficient” test is notoriously discarded as illustrated by the lack of any reference to the test in the DMA.

[387]Nazzini, The Foundations of European Union Competition Law, 32. See also Joshua D. Wright and Douglas H. Ginsburg, “The Goals of Antitrust: Welfare Trumps Choice,” 81 Fordham Law Review 81 (2013), 2405–2423.

[388]Anticompetitive effects of some conducts are implicitly stated, although the pro-efficiency rationale of these conducts is explicitly discarded. For instance, “default settings ... adversely impact customer choice” is one of these claims which ignores the pro-competitive effects of default settings. See European Commission, “Impact Assessment Report,” Part 2/2, SWD(2020) 363 final, December 15, 2020, 18,

[389]This aspect of competition appears overlooked in the DMA as it assumes that free choice needs to be granted irrespective of the technical barriers and the risks inherent to full openness of operating systems and software applications. Indeed, Recital 41 of the DMA states, “Gatekeepers should not restrict the free choice of end users by technically preventing switching between or subscription to different software applications and services. Gatekeepers should therefore ensure a free choice irrespective of whether they are the manufacturer of any hardware by means of which such software applications or services are accessed and should not raise artificial technical barriers so as to make switching impossible or ineffective.”

[390]Recital 3, Digital Markets Act.

[391]Article 6(1)(j) and Article 6(1)(k), Digital Markets Act.

[392]Digital Markets Act, 8.

[393]Article 5 (c) (e), and (f), respectively, Digital Markets Act.

[394]See Stephen Charest, “Bayesian Approaches to the Precautionary Principle,” Duke Environmental Law & Policy Forum 12 (2002), 265–291,

[395]See, for a general discussion about the reversed burden of proof, Cass Sunstein, “Beyond the Precautionary Principle,” University of Pennsylvania Law Review, 151 (2003), 1004–5,

[396]Javier Espinoza and Sam Fleming, “Margrethe Vestager eyes toughening ‘burden of proof’ for Big Tech,” Financial Times, October 30, 2019, (using an example to illustrate that reversal, it was reported that she oddly argued that “Say for instance, Uber started offering higher rates for those drivers who used its platform more often,” said this person. “This would put competitors at a disadvantage because drivers would start favouring Uber to carry out their trips over competing apps. Under the proposed change it would be Uber who would need to show its behaviour is causing no harm to competition rather than the commission having to prove it.”).

[398]Crémer, de Montjoye, and Schweitzer, “Competition Policy for the Digital Era,” 51.

[399]Olivier Blanchard, Alvaro Leandro, and Jeromin Zettelmeyer, “21-1 Redesigning EU Fiscal Rules: From Rules to Standards,” PIIE Working Papers, February 2021, 3,

[400]Jacques Crémer, Roundtable “Conference on Digital Platforms: Opportunities and Challenges?” Online, October 27 and 28, 2020,, 8.

[401]European Political Strategy Centre, “Towards an Innovation Principle Endorsed by Better Regulation,” EPSC Strategic Notes. Issue 14, 30 June 2016, 4.

[402]Article 8, Digital Markets Act.

[403]Article 9, Digital Markets Act.

[404]Article 3, Digital Markets Act.

[405]Recital 23, Digital Markets Act.

[406]Digital Markets Act, Article 3(4)2.

[407]Valcarcel Siso (2017) Applying precautionary measures in antitrust cases. Question for written answer E-004559/2017. European Parliament. Rule 130, PE 607.713.

[408]European Commission, “Antitrust: Commission imposes interim measures on Broadcom in TV and modem chipset markets,” Press Release, October 16, 2020, (wherein Commissioner Vestager argued that “Broadcom's behaviour is likely, in the absence of intervention, to create serious and irreversible harm to competition. We cannot let this happen, or else European customers and consumers would face higher prices and less choice and innovation. We therefore ordered Broadcom to immediately stop its conduct.”); See also European Commission, Broadcom, AT.40608, October 7, 2020, para.10, p (referring to “the need to impose interim measures due to the risk of serious and irreparable harm brought about by Broadcom’s conduct.”).

[409]Motta and Peitz, “Intervention triggers and underlying theories of harm,” 32 (wherein at 12 the authors noted that “We recognize that in practice identifying what is the number of firms that the market would admit is inherently difficult, but the market investigation may have a ‘precautionary’ or ‘preventive’ role, and point to tools that an incumbent may use and that would lead to having only one network being active.”).

[410]Commission’s interim measures have not been numerous over the years. Before Broadcom decision, the Commission ordered interim measures in Commission Decision of August 18, 1982 (IV/30.6969 - Distribution system of Ford Werke AG - interim measures, OJ 1092 L256; Commission Decision of 29 July 1983 (IV/30.698 - ECS/AKZO: interim measures), OJ 1983 L252; Commission Decision of 29 July 1987 (IV/32279 - BBI/Boosey & Hawkes: interim measures) OJ 1987 L286; Commission Decision of 26 March 1990 adopting (IV/33.157 Ecosystem / Peugeot - Provisional measures); Commission Decision of 25 March 1992 (IV/34.072 - Mars/Langnese and Schoeller - interim measures); Commission Decision of June 11, 1992 (IV/34.174 - Sealink/B&J - Halyhead: interim measures); Commission Decision of July 3, 2001 (Case COMP D3/38.044 - NDC Health/IMS Health: interim measures, OJ 2002 L59).

[411]Article 22, Digital Markets Act.

[412]European Commission, “Impact Assessment Report,” Part 2/2, SWD(2020) 363 final, December 15, 2020, 36,

[413]Recital 26, Digital Markets Act.

[414]Motta and Peitz, “Intervention triggers and underlying theories of harm,” 32.

[415]Chapter IV, Digital Markets Act. Market investigation rules appear to be the most sensible, the most reasonable provisions of the DMA and therefore do not necessitate comprehensive commentary at this stage.

[416]Summary of the contributions of the National Competition Authorities to the impact assessment of the new competition tool, Europa website (2020), 7, 24, available at:

[417]For a seminal treatment of this causal relationship, see David J. Teece, Competing Through Innovation. Technology Strategy and Antitrust Policies (Cheltenham, UK: Edward Elgar Publishing, 2013).

[418]David J. Teece, Dynamic Capabilities & Strategic Management. Organizing for Innovation and Growth (Oxford: Oxford University Press, 2009).

[419]For a general discussion of this misguided claim, see Robert Atkinson and Michael Lind, Big Is Beautiful. Debunking The Myth of Small Business (Cambridge, MA: MIT Press, 2018); Aurelien Portuese, “Beyond Antitrust Populism: Towards Robust Antitrust,” Economic Affairs 40 (2020), 237–258; Jonathan M. Barnett, “Big is not necessarily bad,” The Hill, July 30, 2020,; Christopher Marchese, “Debunking the ‘Big is Bad’ Bogeyman: How Facebook Benefits Consumers,” George Mason Law Review 28 (2020) 1–47.

[420]Digital Markets Act, 1.

[421]Indeed, this is stated clearly in the Digital Markets Act, 8, wherein the European Commission noted that “online platforms were split on the issue, with the majority of large online platforms and their representative associations questioning the need for a new gatekeeper instrument. On the other side, many small and medium sized platforms, in particular those that are business users of large online platforms, expressed their support for a new gatekeeper instrument.”


[423]The DMA approaches the issues of “fragmentation” of the Digital Single Market only through the angle of member states being precluded from adopting subsequent national regulations similar to the DMA—not from the fundamental angle that current regulatory discrepancies and multiplications of norms hamper the completion of the Digital Single Market. The DMA indeed states, “Regulatory initiatives by Member States cannot fully address these effects; without action at EU level, they could lead to a fragmentation of the Internal Market,” 1. However, even this objective of precluding subsequent national regulations conducive of fragmentation of the Digital Single Market seems to be failing in light of the recent adoption of the German competition law amendments.

[424]Recital 64, Digital Markets Act.

[425]Article 16, Digital Markets Act.

[426]Article 16(4), Digital Markets Act.

[427]Not only is scale crucial for success, but it is so important that it defines tech companies. For instance, large and competitive cloud services companies are called “hyperscalers.” See Greg Jones, “Why hyperscalers count on colocation providers to accommodate high demand for data center capacity and services,” DataCenterDynamics, March 10, 2020,

[428]Article 4 Digital Markets Act, which states that “Commission may upon request or its own initiative reconsider, amend or repeal at any moment a decision” of designating a company as a gatekeeper.

[429]Article 4(2), Digital Markets Act.

[430]Article 10, Digital Markets Act.

[431]Digital Markets Act, 11.

[432]Article 17, Digital Markets Act.

[433]Digital Markets Act, 65.

[434]This is particularly telling given the fact that the DMA proposal takes place during the Public Consultation on Market Definition rules wherein, following the Alstom-Siemens blocked merger debacle, the European Commission acknowledged the need for competition rules to better take into consideration “potential competition” and to have a more dynamic approach. See European Commission, “EU Competition law—market definition notice (evaluation), June 26, 2020,; European Commission, Summary of the stakeholder consultation to the Evaluation of the Market Definition Notice, Ares(2020)7730543, December 18, 2020, (noting particularly at 14 that “Several respondents criticise that potential competition is only touched upon briefly in the Notice while more detailed guidance is contained in the Commission’s guidance on the competitive assessment. They claim that potential competition should be discussed more in detail in a revised Notice given the rapidly changing market realities, in particular because of technological developments. Furthermore, some respondents point to the specificities of digital markets and submit that the constraints imposed by successful digital platforms in adjacent markets should be taken into account and that the boundaries between supply-side substitutability and potential competition may be more blurred in digital contexts.”) The need to better integration potential competition (or dynamic competition) into the Commission’s analysis is reported at 16 to be widely shared: “The Commission should focus more on potential competition and external constraints (e.g., from traditional sectors on digital markets and vice versa),” see also, at 20, “Going beyond the technicalities of market share calculations and therefore the scope of the market definition notice, some businesses and business associations also explained that market share calculation does not take into account competitive constraints from outside the market and from potential competition, and thus do not adequately reflect market power.” On the blocked merger between Alstom and Siemens, which paved the way for a reflection on market definition rules, see Sam Fleming, Jim Brunsden, and Javier Espinoza, “Margrethe Vestager: the star commissioner brought down to earth,” Financial Times, July 17, 2020,

[435]Digital Markets Act, 5-6 (noting that “the core platform services in scope are only those where there is strong evidence of (i) high concentration, where usually one or very few large online platforms set the commercial conditions with considerable autonomy from their (potential) challengers, customers or consumers.”).

[436]Recital 26, Digital Markets Act.

[437]See, for instance, Recital 28, Digital Markets Act, “This should allow the Commission to intervene in time and effectively, while fully respecting the proportionality of the considered measures. It should also reassure actual or potential market participants about the fairness and contestability of the services concerned.” See also Recital 61 wherein it is noted that “Ensuring an adequate level of transparency of profiling practices employed by gatekeepers facilitates contestability of core platform services, by putting external pressure on gatekeepers to prevent making deep consumer profiling the industry standard, given that potential entrants or start-up providers cannot access data to the same extent and depth, and at a similar scale.”

[438]The European Commission has recognized such a need in the reform of the Market Definition rules. With respect to the literature, see the seminal article, Gregory Sidak and David J. Teece, “Dynamic Competition in Antitrust Law,” Journal of Competition Law & Economics 5 (2009), 581–631, doi:10.1093/joclec/nhp024.

[439]Richard J. Gilbert, Innovation Matters. Competition Policy for the High-Technology Economy (Cambridge, MA: MIT Press, 2020), 235.

[440]Digital Market Act, 1.

[441]Digital Markets Act, 10.

[442]Recital 37, Digital Markets Act.

[443]Recital 42, Digital Markets Act.

[444]Digital Markets Act, 59.

[445]David J. Teece, Dynamic Capabilities & Strategic Management. Organizing for Innovation and Growth (Cambridge, MA: Oxford University Press, 2009).

[446]Ibid, 87–88.

[447]Ibid, 89.

[448]Recital 26, Digital Markets Act.

[449]David J. Teece, Dynamic Capabilities & Strategic Management. Organizing for Innovation and Growth (Cambridge, MA: Oxford University Press, 2009), 236.

Back to Top