The European Union’s Digital Markets Act Seeks to Regulate Competition with Little Regard to Impact on Consumers

Photo Credit: Getty

Henrique Schneider is the deputy CEO of the Swiss Federation of Small and Medium Enterprises

Coming full circle: For much of the 19th century, “trusts” were a major fixture of the U.S. economy. For example, railroads, oil, steel, and sugar were organized in trusts. Two of the most famous companies that were deemed as trusts were U.S. Steel and Standard Oil. More than trusts, they were near-monopolies that controlled the supply of their products, as well as the price. The problem of one company controlling one sector of the economy is not so much that such control imposes restraints on competition, but that it creates barriers to market entry.

Barring competitors from entering a market can lead to monopolization, and with it, decline in quality of the dominant company’s product.[1] That is why the Sherman Act bars the “monopolization” of markets, not the restraint to competition resulting from certain cooperative arrangements. The original text is clear on that topic:

Section 1: Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal.

Section 2: Every person who shall monopolize, or attempt to monopolize, or combine or conspire with any other person or persons, to monopolize any part of the trade or commerce among the several States, or with foreign nations, shall be deemed guilty of a felony.[2]

President Theodore Roosevelt, despite his reputation as a trustbuster, did not use the new law at will, but rather targeted those trusts that he perceived had especially damaging effects on the economy. He believed that these enterprises created market distortions that enabled dominant firms to engage in rent-seeking and helped them deter new market entrants from challenging their monopolies.[3] Both the Sherman Act and the Theodore Roosevelt administration went after the effects of restraints on trade. Roosevelt opposed any economic agent having the power to systematically suppress the ability of another to enter competition. According to some historians, Roosevelt himself believed that Section 1 of the Sherman Act, addressing state agents, was even more important than Section 2, addressing private attempts in monopolization.[4]

John D. Rockefeller, the head of Standard Oil, the nation’s largest trust, was distrustful of competition. “Competition is a sin,” he famously pronounced.[5] He voiced moral and economic arguments for organizing an economy or its sectors into trusts. Rockefeller’s moral argument is that trusts are a form of cooperation that helps make economic advancement accessible to everyone, not just to the profitable enterprises in a given industry. His economic argument was based on the effects of scale generated by trusts. The larger a company, the cheaper a good it can produce, to the benefit of all its customers.[6]

Fast forward to the 2020s: A unique type of entity appears to have taken control of the Internet. Known as “gatekeepers,” they manage access and both upstream and downstream business practices, much like trusts did in their day. And like trusts, they might be able to bar any non-compliant actors from accessing the Internet for various economic activities, stifling competition and economic freedom. If that view is correct, the European Union’s (EU) Digital Market Act (DMA) could play the same role that the Sherman Act once played offline—one could say that the European Commission intends to become the Teddy Roosevelt of the digital age.

The problem is not so much that the analogy fails, but that the economic reasoning behind the DMA fails, as well. This paper explores the DMA’s provisions, assesses the economic reasoning behind it, and discusses its implications for global innovation.

Contents of the DMA. As of this writing, the Digital Markets Act, a proposal by the European Commission (EC), is still in the consultation phase. That means that the European Parliament can still debate its contents and propose changes. However, in the EU, such proposals are often passed without much change.[7] The DMA is modeled on existing and recent antitrust—or, in EU terminology, competition—investigations driven by complaints from competitors seeking specific remedies in specific market contexts.

Intended to serve as a new tool for competition policy enforcement, it introduces new policy objectives of “fairness and contestability.”[8] By doing so, the DMA changes EU competition law in a substantive way, the most important being the introduction of ex ante regulation and an inversion of the burden of proof in antitrust investigation.[9] These changes will be discussed later in this paper, together with an overall assessment of the DMA.

The main motivation for the DMA is explained in the European Commission’s “Explanatory Memorandum”:

Weak contestability and unfair practices in the digital sector are more frequent and pronounced in certain digital services than others. This is the case in particular for widespread and commonly used digital services and infrastructures that mostly directly intermediate between business users and end users. The enforcement experience under EU competition rules, numerous expert reports and studies and the results of the OPC [open public consultation] show that there are a number of digital services that have the following features: (i) highly concentrated multi-sided platform services, where usually one or very few large digital platforms set the commercial conditions with considerable autonomy; (ii) a few large digital platforms act as gateways for business users to reach their customers and vice-versa; and (iii) gatekeeper power of these large digital platforms is often misused by means of unfair behaviour vis-à-vis economically dependent business users and customers. The proposal is therefore further limited to a number of “core platform services” where the identified problems are most evident and prominent and where the presence of a limited number of large online platforms that serve as gateways for business users and end users has led or is likely to lead to weak contestability of these services and of the markets in which these intervene. These core platform services include: (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, including advertising networks, advertising exchanges and any other advertising intermediation services, where these advertising services are being related to one or more of the other core platform services mentioned above.[10]

The explanatory memorandum also recognizes that these phenomena are happening outside the digital sector as well, but it notes that they are more relevant in the digital one since they European Commission considers them the “most pressing.”[11] It also states that an EU-wide regulation of the digital sector can better achieve the aims of integration and regulatory stability than an array of “divergent national rules to address the problems arising from the significant degree of dependency of business users on core platform services provided by gatekeepers.”[12]

Gatekeeper Definition. The DMA introduces the notions of “digital sector” and “gatekeeper.”[13] The idea of a “digital sector” is already problematic. It presupposes a clear separation between digital and non-digital economic activities. This separation makes no empirical sense. Amazon and Uber, for example, rely on the integration of both digital and non-digital resources, while a local bakery can use online services to sell its products or procure ingredients. Rather, digital and non-digital are better understood as complementary channels and not as sectors, as will be argued further in this text.

The introduction of the “gatekeeper” concept is even more problematic. It requires a lengthy explanation in the proposed legislation and its interpreting materials. The length of the definition is an indication that the European Commission is aware that it is introducing a completely new concept in antitrust.

The DMA describes gatekeepers as “structuring elements”[14] of the digital economy that enjoy an entrenched position as intermediator—someone who brings other agents together in order for them to interact or transact. Intermediators facilitate the exchange between agents. Gatekeepers’ business models rely on the intermediation of large groups of agents creating network effects. While these increase the overall value of the transacting group and the utility of the network’s organizer, according to the EC’s reasoning, they also make businesses and consumers alike increasingly dependent on these gatekeepers.[15] That dependence may allow gatekeepers to engage in unfair practices that ultimately harm consumer welfare.[16] Therefore, the Commission concludes, regardless of harmful effects, gatekeepers have a major impact on digital markets and require oversight.[17]

Article 3(1) of the DMA makes the definition operational. It begins:

A provider of core platform services shall be designated as gatekeeper if: (a) it has a significant impact on the internal market; (b) it operates a core platform service which serves as an important gateway for business users to reach end users; and (c) 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.[18]

It then specifies criteria for gatekeepers. Providers of “core platform services” may be designated as gatekeepers. These include search engines, social networks, operating systems, intermediating platforms (connecting sellers and buyers), video sharing sites, communication services, cloud computing, and advertising services.

Table 1: Qualitative Criteria and Thresholds under DMA Article 3

CriteriaPresumption Threshold
The provider has a significant impact on the internal market.If the entity that owns the provider of the core platform service: Achieves an annual turnover within the European Economic Area of at least €6.5 billion ($7.1 billion) in the last three fiscal years. Or Its average market capitalization or equivalent fair market value amounted to at least €65 billion in the last fiscal year, and It provides a core platform service in at least three EU member states.
The service is an important gateway for businesses to reach end users.If the core platform service, in the last fiscal year, has: More than 45 million monthly active end users established or located in the EU, and More than 10,000 yearly active business users in the EU.
The provider’s position is entrenched and durable, or is likely to be so in the future.If it meets the important gateway presumption thresholds in each of the last three fiscal years.

Once a provider is deemed to be a core platform service provider, it may be identified as a gatekeeper based on the qualitative criteria set out by DMA Article 3(2)-(8). The proposal specifies thresholds for when each of the qualitative criteria is presumed to be met. However, even if the presumption thresholds are not met, the Commission may still identify an entity as a gatekeeper.[19] The qualitative criteria and presumption thresholds are listed in Table 1.

The purpose of this framework is to provide flexibility for enforcement. The qualitative criteria allow the European Commission to designate as gatekeepers any core platform service providers that exhibit similar “risks for fairness and contestability” to companies that meet the presumption thresholds.[20] Providers that meet these criteria may dispute the designation before the Commission by providing evidence that they do not enjoy a gatekeeper position.[21] In both of these cases, the Commission has to assess the size, operations, and position of the provider, the number of businesses that depend on the platform, the number of end users, the entry barriers generated by the advantages inherent to the provider’s platform and capabilities (network effect, data), the network effects benefiting the provider, the lock-in effect for users, and other market characteristics.[22]

Ex Ante Regulation in Articles 5 and 6. The DMA creates a new legal economic entity, the “gatekeeper.” Even if Article 3 of the DMA tries to give a classificatory overview over what a “gatekeeper” might be, its specific meaning only becomes apparent when Articles 5 and 6 are considered. Articles 5 and 6 of the DMA lay out rules for gatekeepers and their obligations to end users and third parties. Both articles’ provisions are applicable to all gatekeepers independently of any harm or benefit that their practices might cause, and are applicable ex ante. As such, they regulate gatekeepers, rather than attempt to remedy harmful market behavior by gatekeepers.

Article 5 consists of seven obligations for gatekeepers:

  • Article 5(a) prevents the combination of personal data from other services offered by the same platform.
  • Article 5(b) prohibits gatekeepers from providing services at different costs and conditions across various platforms via third-party intermediaries and builds in protections to allow businesses to interact with consumers outside of the core platform service.
  • Article 5(c) stipulates that gatekeepers must allow business users to promote offers to, and conclude contracts with, end users acquired via the core platform service, regardless of whether the users also used other services provided by the gatekeeper.
  • Article 5(d) prevents gatekeepers from restricting the ability of business owners to raise “issues with any relevant public authority.”
  • Article 5(e) stipulates that gatekeepers must “refrain from requiring business users to use, offer, or interoperate with an identification service of the gatekeeper.”
  • Article 5(f) precludes the ability of gatekeepers to force businesses or end users to subscribe to the core platform service as a condition of accessing those services.
  • Article 5(g) establishes transparency guidelines in advertising prices, stipulating that advertisers and publishers reserve the right to request data on ad relevance and revenue. [23]

Article 6 primarily deals with self-preferencing, discriminatory ranking, and data-sharing obligations similar to those outlined in Article 5(g):

  • Article 6(1)(a) prohibits the use of non-public data by gatekeepers. This rule affects gatekeepers’ ability to gather data generated by both end users and business users.
  • Article 6(1)(b) mandates that end users must be able to uninstall preinstalled applications.
  • Article 6(1)(c) requires gatekeepers to “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.”
  • Article 6(1)(d) prohibits self-preferencing, particularly by search engines mandating nondiscriminatory ranking for third-party providers of online search engines.
  • Article 6(1)(e) establishes guardrails to prevent discrimination within application software stores.
  • Article 6(1)(f) requires gatekeepers to ensure interoperability between third-party software applications.
  • Article 6(1)(g) mandates gatekeepers to provide advertisers and publishers access to “performance measuring tools,” which, in compliance with the European Union’s General Data Protection Regulation (GDPR), may include data on ad revenue. This data must be provided free of charge.
  • Article 6(1)(h) creates rules for gatekeepers to facilitate data portability to prevent siloed storage of user data that would otherwise lead to users to being locked into certain platforms.
  • Article 6(1)(i) requires gatekeepers to provide business users and authorized third parties with “high-quality, continuous, and real-time access and use of aggregated or non-aggregated data.”
  • Articles 6(1)(j) and 6(1)(k) require gatekeepers to “apply fair and non-discriminatory general conditions of access for business users to its software application store.”[24]
  • Article 6(2) attempts to define the relevant data, stating: “For the purposes of point (a) of paragraph 1 data that is not publicly available shall include any aggregated and non-aggregated data generated by business users that can be inferred from, or collected through, the commercial activities of business users or their customers on the core platform service of the gatekeeper.”[25]


Companies Subject to the DMA. By stipulating ex ante obligations for “gatekeepers,” the Commission appears to be attempting to circumvent practical problems of enforcement by legislating its aims into the DMA. The companies most likely to be covered by the DMA are those currently being investigated by the Commission under EU competition law, based on previous and ongoing EU competition cases. Table 2 shows how the DMA could be applied to ongoing investigations.

Table 2: Companies Likely Subject to the DMA and Examples of Regulations[26]

RuleCompanyRegulatory Impact
5(a)Facebook, GooglePrevents Facebook from harvesting and importing personal data from Instagram that it could use to target advertising to users.
5(b)Amazon, Online Travel AgenciesPrevents online travel sites from using most favored nation clauses[27] guaranteeing the lowest bid-price of a hotel to be offered via the travel site. For example, an online booking platform could not require hotels using that platform to offer the lowest price for a room on the platform; the hotel would still be free to offer a lower price on its own.
5(c)AppleRequires Apple to allow apps to be offered outside of its App Store.
5(d)GoogleProhibits Google from banning an individual business in response to that business raising an issue regarding Google’s compliance with the DMA.
5(e)Facebook, GoogleFacebook and Google have their own sets of applied technologies (tech stacks) for identifying users, or customers. The rule requires gatekeepers to share some aspects of their user identification solutions with smaller advertisers that use the gatekeeper’s services and rely on its data to reach consumers.
5(f)Facebook, GoogleThis could refer to various advertising ties that gatekeepers have established between their various businesses—for example, Alphabet might establish links between Google Ads and YouTube. According to the Commission, these links may impair competitors’ ability to advertise. The rule might require Facebook and Google, and possibly Amazon, to sever those links or allow competitors similar linking.
5(g)GoogleThis rule might be similar to the one in 5(f), but targeting Google’s link with AdX in the search engine.
6(1)(a)Amazon, GoogleRequire Amazon to disclose all seller data.
6(1)(b)Google, Apple, MicrosoftCurbs preinstallation of apps.
6(1)(c)Apple, GoogleThis rule, which complements 5(c), requires Apple to allow users to download apps without using the App Store. While 5(c) targets the tech stack, this targets the interface.
6(1)(d)Google, Amazon, AppleProhibits Google Shopping from self-preferencing its own offerings.
6(1)(e)AppleRequires Google to allow users to access some services, such as Google Hangout, without subscribing.
6(1)(f)Google, Facebook, AppleRequires Facebook to open its proprietary application programming interface to allow app developers to access data or functionalities on Facebook’s platform, photo-sharing site Instagram, and software components, in order to allow them to better interact with Facebook.
6(1)(g)Google, FacebookBans self-advertising across channels.
6(1)(h)All CompaniesThese are established policy goals stated in the GDPR. They are largely restated here.
6(1)(i)All Companies
6(1)(j)All Companies
6(1)(k)Apple, GoogleRequires Apple to allow any app developer to distribute its app through the App Store, even when it does not accept Apple’s general terms and technical standards.

Assessing the DMA. In many ways, the DMA is a regulatory innovation. For the first time, the “digital sector” and its “gatekeepers” are subject to a sui generis law. Is this novelty beneficial or harmful to competition, and to the economy in general? There are three different aspects to this question.

  • First, is the DMA compatible with a free society and economy, which the EU claims to advocate?
  • Second, is the model on which the DMA bases its rules sound? Does a “digital sector” exist, and is it necessarily dependent on “gatekeepers”? Does antitrust policy properly address potential harm to competition process in the digital realm?
  • Third, are the novel obligations of gatekeepers pro- or anti-competitive?

The first question involves a general discussion about whether the DMA improves competition, its compatibility with antitrust policy, and its economic effects.

Digital Sector Definition.The DMA defines the digital sector as a category of products and services provided through “information society services,”[28] defined in Article 1(1) point (b) of Directive (EU) 2015/1535[29] as “any service normally provided for remuneration, at a distance, by electronic means and at the individual request of a recipient of services.”[30] This definition breaks down as follows:

  • “[A]t a distance” means that the service is provided without the parties being simultaneously present.
  • “[B]y electronic means” means that the service is sent 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, radio, or other electromagnetic media.
  • “[A]t the individual request of a recipient of services” specifies that the service is provided through the transmission of data on individual request.[31]

The problem with these legal definitions lies in the DMA’s underlying premise—that digital markets can be separated from non-digital markets for purposes of applying ex ante regulatory requirements. Many industries, not just online platforms, are increasingly being transformed by digital technologies.  Brick-and-mortar businesses, such as supermarkets, also try to attract customers with tailored advertisements. They also collect customer data and often sell it—anonymized—to digital marketing firms.

Firms in a wide array of industries now compete against one another to reach consumers through multiple business channels, including digital ones. The practices the DMA seeks to regulate are not specific to the online realm, and are often deployed by both offline businesses models and firms using multiple marketing channels. Furthermore, businesses of all types are employing more digital tools to extract value from data, using digital outlets to branch out into new markets and update their marketing strategies. All these businesses, including their online activities, are already regulated by EU antitrust and other statutes.

If “digital transformation” is indeed happening, then everything is digital.[32] But if everything is digital, what is the point of regulating the digital realm separately? One could argue that the DMA is an attempt by the EU to create a harmonized and level playing field as a basis for future digitization. That goal is based on a faulty premise. The digital is a technology stack—a combination of different elements, some technical, some managerial, and some skill-based. It is not possible to separate out  all of these elements. But even if they could be, why is the EU attemting to regulate it differently than it regulates other sectors? The so-called digital sector is being regulated ex ante while the so-called non-digital sector is being regulated ex post.

Separating the digital from the non-digital is conceptually nonsensical and practically impossible. Even if the DMA were to focus on a narrow view of digital services, those almost always rely on non-digital services, tangible goods, or infrastructure. For example, while Uber is an app, it relies on thousands of real-world drivers, cars, and streets, combined with other digital services, such as payment and routing, which are themselves dependent on several non-digital elements.[33]

Within a given industry, digital and analog actors compete to reach end users through various means, each with specific costs and benefits. For example, when Amazon first started, it focused on selling books and music online, and promptly became the leading online retailer of books and music. Does that make Amazon a monopoly in the digital market for books and music? No, because there is no online bookshop market distinct from the brick-and-mortar one. The same is true for music and any other product category. The only relevant market for antitrust purposes is the product market, encompassing both online and offline distribution channels.

The French National Competition Authority (Autorité de la Concurrence) adopted this position in the case of the 2016 merger of French retailers Fnac and Darty.[34] In its competition assessment, the Competition Authority determined that brick-and-mortar and online markets were part of the same relevant market for antitrust purposes. Online sellers and traditional stores compete with one another. Therefore, competition policy applies equally to all products, regardless of their distribution channels, whether 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 existing markets. Competition takes place in that broader product market, not on the digital channel.[35]

So, according to the DMA, what is the ratio of physical versus digital sales necessary for a company to be considered digital in nature? For example, if Amazon were to increase its offline sales, would it be considered to be a non-digital company under the DMA? The DMA cannot answer that, since it does not consider the integration of business models and markets, let alone their inner dynamics. Regulating companies according to fixed, rigid categories risks creating misunderstandings of the business realities inherent to highly dynamic markets.

Assumptions. Independent from the conceptual problems linked to the notion of a “digital sector,” the model upon which the DMA is based shows serious biases and shortcomings. The DMA Impact Assessment (IA), which serves here as a proxy for assessing the DMA’s economic model, is riddled with unsubstantiated assumptions. For example:[36]

  • The DMA will cause research and development (R&D) spending in the information and communication technology sectors of each member state to double, with related increases in employment.
  • Because of the DMA, a significant amount of spending on merger and acquisition activity by large technology companies will be replaced by R&D spending.
  • These increases in R&D spending will augment value to users.
  • The DMA’s obligations are narrowly targeted so that they will have no negative effect for users regarding costs, quality, or functionality of core platform services, either today or in the future.
  • Regulatory interventions in the digital sector will only involve minimal compliance costs and no harmful side effects on innovation or competitive incentives.
  • The DMA will reduce regulatory fragmentation caused by diverging national approaches, but in its absence a substantial amount of cross-border trade intermediated by online marketplaces would be lost by 2025.
  • All these benefits can be achieved with only 30 to 90 European Commission employees, and no other enforcement costs at EU institutional level.

The Impact Assessment does not evaluate the economic impact of the specific prohibitions or obligations of the DMA.

Economic Effects of the DMA. Based on the dubious notion of a digital sector and the lack of critical evaluation—and despite the Impact Assessment’s overly optimistic assumptions—the misunderstanding of the economics of digital transformation and of the dynamics of innovative markets displayed in the DMA could lead to many adverse effects, as explained below.[37]

The DMA could hinder the widespread adoption of digital technologies across several industries. It would impose additional regulatory costs on companies that have already established an online presence, which could undermine those companies’ competitiveness. Akin to the GDPR’s effect of protecting big tech companies from new competition by startups by raising the costs of complying with the regulation and thereby creating a barrier to market entry by new competitors, the DMA imposes new regulatory costs that both raise existing barriers to entry and create new ones, which could slow down digital innovation.[38]

The regulation makes imitation more attractive than innovation. The risk of free riding is present in the DMA in multiple instances. The regulatory obligation to share data, grant access, and encourage innovation among rivals makes it less costly—if not free—for firms to copy market leaders’ innovations. Thus, innovation laggards will benefit from the regulation at the expense of innovation leaders, who are thus deterred from innovating in order to avoid regulation-driven free-ridership problems in the future.

Note that this argument is not based on intellectual property considerations, but on the cost of developing and sharing operative capabilities. For example, if a gatekeeper develops a novel means of identifying users, it could be forced to share it with its competitors. In such an instance, there is little first-mover benefit, since the first mover can be required to share its capabilities even before deploying them.

Regulation-created rivalry artificially generates competition at the expense of innovative market leaders, who are required to allow access to rivals—who gain a strategic advantage through regulation, at no cost. For example, the DMA states that gatekeepers should 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. That means that if Microsoft links its Outlook email application with LinkedIn (both Microsoft services), it could be required to offer similar links with competitors’ business networks or to sever the original tie-in. This obligation, laid down in DMA Articles 5 and 6, overlooks the innovation dynamics resulting from the initial creation and subsequent innovation of designing an ecosystem in which several services are interlinked. [39] Ecosystems may be deployed by a gatekeeper favoring its ancillary services at first, but if the ecosystem is successful, it will be driven to expand and incorporate services offered by third parties.[40]

As a practical matter, what would be the innovation incentives for Apple if it were prevented from favoring its proprietary apps—such as iMessage, Maps, and Safari—by either preinstalling them or placing them prominently in the App Store? The DMA’s equal access requirement would prevent self-preferencing and undermine the company’s proprietary assets and services, and thereby discourage innovation, both upstream—updates on operating systems, as Apple’s ability to appropriate its innovations would decrease—and downstream—updates and development of apps would be hindered, since the expected benefits derived from these investments would decrease. Similarly, granting all app developers equal access to Android OS—to the whole program code, not just to the tech stack necessary to develop apps—without Google being able to self-preference its apps, would put Android OS’ openly licensable characteristics at risk. Google might then opt to recoup its investments and innovations through a more traditional, fee-based business model. In both instances, innovation laggards would benefit, while innovation leaders would cut back on investment. The overall impact on innovation and competition would likely not be positive.[41]

Furthermore, the DMA skews the regulatory playing field for online companies but not for their offline competitors. This is especially a problem for companies designated as “gatekeepers.” The DMA’s focus on increasing the contestability of core platform services suggests that the DMA is mainly designed to uproot the digital gatekeepers’ dominant market positions in favor of other companies. Yet, firms’ dominant positions are never permanent. Such a goal would amount to regulators picking winners and losers.

It also does not yield clear benefits regarding consumer welfare and innovation, especially if the newly dominant firms’ services correspond less to consumer preferences.

Finally, it incentivizes rent-seeking and free riding at the expense of innovation incentives. For example, does making Google’s dominant search engine subject to greater contestability mean promoting Microsoft’s Bing or the French search engine Qwant? That would ignore consumer preferences and likely deter investment by Google due to the potential of other firms piggybacking on its innovations.

The DMA and Antitrust. Under the consumer welfare standard, antitrust law allows for market leadership and large firms. Even potentially harmful conduct is permitted as long as it creates, on balance, improved efficiencies and benefits for consumers, including lower prices. In fact, antitrust enforcement in most developed countries has trended toward an objective, economic, effects-based approach focused on establishing anticompetitive conduct through due process. These investigations put greater weight on efficiency objectives and consumer welfare rather than on a certain level of market concentration.[42] On the other hand, antitrust enforcement officials in Europe seem to favor protecting potential competitors, even if market leaders outperformed competitors, gaining customer loyalty through innovation and acquisition-strategies.[43]

The European Commission has largely assumed that the DMA will yield good outcomes, but it is difficult to assess the real economic effects of its many prohibitions and regulations.[44] The DMA constitutes a dramatic turn toward the use of an ex-ante regulatory mechanism intended to overlap and operate in parallel with traditional European antitrust methods. As a sweeping overhaul of European competition policy, the DMA could threaten certain operations of U.S. firms in Europe, discriminate in favor of European companies, and provide opportunities for global rivals—including subsidized Chinese firms—to achieve long-term non-economic goals, such as boosting their respective governments’ geostrategic position.

By introducing ex ante remedies, the DMA pushes antitrust policy away from its current adjudicatory stance toward a priori regulatory compliance based on set requirements. Conceptually speaking, antitrust law is supposed to support market processes, which are expected to be generally self-regulating, with regulators only stepping in when markets produce outcomes that are detrimental to competition. The burden of proof is on the competition authorities to establish the existence of anticompetitive behavior. By contrast, the DMA assumes that digital gatekeepers will act in anticompetitive fashion if left unchecked, and therefore need to be directed by regulators. The burden of proof is on the digital gatekeepers, who have to conform with the EU competition authority’s stipulations.

Shifting the focus of EU competition regulation from market efficiency to specific policy and economic objectives would push competition law toward a “big is bad” approach that favors smaller European competitors.[45] In other words, a company’s size will determine whether the new set of ex ante competition rules apply to it. U.S. platforms are particularly concerned with the prohibitions and obligations that would limit their ability to engage in conduct that is pro-competitive, efficient, and welfare-enhancing. That approach would ignore the dynamic competition that gatekeepers bring to the market, consumer welfare, and the innovation and investment incentives necessary for future technological breakthroughs.

Prioritizing the interests of competitors, suppliers, and business users of digital platforms, the DMA regulates ex ante the activities and prices of companies designated as digital gatekeepers. Worse, it does not allow the European Commission to modify the imposed obligations if they prove counterproductive or harmful. This lack of a safety valve is particularly troubling, given that much of what the DMA will prohibit creates value for consumers and business users alike.

The DMA does not allow for any of the safeguards that currently exist for competition policy enforcement. These safeguards include the competition authority’s assessment of the likely effect of conduct, both beneficial and detrimental to competition. It considers a counterfactual, assessing what the world would look like in the absence of allegedly anticompetitive conduct. These safeguards help keep overly aggressive and purely formal enforcement in check, ensuring that competition is maintained while allowing first movers to earn a profit from their innovative products and services.[46]

Given the size and scale of the companies likely to be designated as gatekeepers under the DMA, one would hope for more safeguards to ensure that this regulation does not inadvertently do more harm than good. That is why courts have imposed evidentiary requirements, under competition law, for imposing the far-reaching remedies proposed under the DMA. These safeguards are necessary to protect fundamental principles of an open market economy, including property rights, freedom to contract, and the freedom to run a business.

Breaking up companies, digital tools, or services requires proof of harm that government purports to remedy via mandated product design changes. The DMA would do away with that as well. Without these checks and balances, the DMA could soon become an attractive alternative for overzealous enforcers seeking to avoid the constraints of competition law. This is particularly true if, fragmented national enforcement of the DMA were to be allowed—for example Germany’s overly strict approach vs. Dutch competition authorities’ more pragmatic stance.[47]

Conclusion. The DMA would replace effects-driven antitrust enforcement with ex ante regulation based on faulty economic reasoning and dubious modelling. The DMA could slow the adoption of digital technologies by European industry for three main reasons:

  • Increasing regulatory costs will drive up barriers to entry, thereby reinforcing rigidities in the economy and cementing the status of dominant companies. Incentives for smaller and medium-sized platforms to innovate and scale up will be suppressed, as growth and success are likely to be met with increased regulatory scrutiny, potential legal liability, and an inability to claim the earned monetary rewards of innovation in Europe.
  • The DMA’s ex ante regulatory approach will reduce incentives for large online platforms to provide new, innovative products and services to European businesses and consumers. Restrictions on gatekeepers, including prohibitions on bundling and market entry, would discourage platforms from investing in the development of new products and services. Specifically, DMA Articles 5 and 6 overlook the innovation dynamics resulting from the initial creation and subsequent innovations of a service. Companies closing in on the threshold for meeting gatekeeper status may be disincentivized from creating a new service that would bring in additional users. For example, this concern could persuade a large platform provider against investing in a new telehealth service for fear of outgrowing its established regulatory category. DMA restrictions will hamper existing gatekeepers from competitively constraining one another, particularly outside of their own “lane,” such as for example, Apple with search or Microsoft with digital advertising.
  • As large online platforms are constrained, opportunities for natural affiliations between relatively less digitized European firms and highly digitized U.S. firms will be reduced.

Is there a way out?

Repealing the DMA would solve the problems it creates and allow competition to unfold, while all actions of companies with digital operations would remain subject to Articles 101 and 102 of the Treaty on the Functioning of the European Union, the foundation of EU antitrust regulation.

As a second-best option, the DMA could be reformed. Instead of introducing ex ante regulation and the gatekeeper concept, it could specify substantiality criteria. Firms with activities on digital channels as specified by the Articles 5 and 6 of the DMA would need to explain, upon inquiry by the European Commission, how their activities do not infringe Article 101 and 102 TFEU.

A third-best option is to allow businesses deemed as gatekeepers to challenge that categorization before the European Court of Justice. Such a challenge should be independent from any specific case put forward by the European Commission, and the Court should be allowed to assess material, economic criteria, and weigh them against the formal criteria set out in the DMA.


[*] Henrique Schneider is the deputy CEO of the Swiss Federation of Small and Medium Enterprises.

[1] Israel Kirzner, “The driving force of the market: The idea of competition in contemporary economic theory and in the Austrian theory of the market process,” in David L. Prychitko, ed., Why Economists Disagree: An Introduction to the Alternative Schools of Thought (Albany, NY: State University of New York Press, 1998), pp, 27-52.

[2] 15 U.S.C., § 1 and 2.

[3] For complete biography of Theodore Roosevelt, refer to the three-part oeuvre by Edmund Morris, The Rise of Theodore Roosevelt; Theodore Rex; and Colonel Roosevelt (all: New York City: Random House, 2010). A nuanced treatment of what it meant to be a “progressive” president is given in John Morton Blum, The Progressive Presidents: The Lives of Theodore Roosevelt, Woodrow Wilson, Franklin D. Roosevelt, and Lyndon B. Johnson (New York City: W.W. Norton, 1982).

[4] Kathleen Dalton, “Changing Interpretations of Theodore Roosevelt and the Progressive Era” in A Companion to the Gilded Age and Progressive Era (Hoboken, NJ: Wiley, 2017), pp. 296-307.

[5] Ron Chernow: Titan. The Life of John D. Rockefeller Sr. (New York City: Random House, 1998).

[6] For full quotes of Rockefeller’s arguments, refer to Chernow (1998).

[7] Stéphanie Novak, Olivier Rozenberg, and Selma Bendjaballah, “Enduring consensus: why the EU legislative process stays the same,” Journal of European Integration. Vol. 43, No. 4 (2021), pp. 475-493,

[8] DMA, Recital, (para. 18).

[9] Philipp Bongartz, Sarah Langenstein, and Rupprecht Podszun, “The Digital Markets Act: Moving from Competition Law to Regulation for Large Gatekeepers,” Journal of European Consumer and Market Law, Vol. 10, No. 2 (2021), pp. 60-67,

[10] DMA, Explanatory Memorandum, p. 3.

[11] Ibid., p. 1.

[12] Ibid., p. 4.

[13] DMA, para. 1(1).

[14] DMA, Explanatory Memorandum, p. 1.

[15] Ibid, pp. 6, 15 and 18.

[16] Ibid, p. 1.

[17] Ibid, p. 20 and recital, para. (3).

[18] DMA, para. 3(1).

[19] Ibid., para. 3(6).

[20] Martin Effert, Axel Metzger, Heike Schweitzer, and Gerhard Wagner, “Taming the giants: The DMA/DSA package,” Common Market Law Review, Vol. 58, No. 4 (2021), pp. 987-1028,

[21] DMA, Explanatory Memorandum, p. 6.

[22] Ibid, p. 15.

[23] DMA, para. 5(a)-(g).

[24] Ibid. para. 6(1)(a)-(k).

[25] Ibid., para. 6(2).

[26] This table expands on one published by, a web publication of the Centre for Economic Policy Research (CEPR), which is generally supportive of European Commission regulation. Cristina Caffarra and Fiona Scott Morton, “The European Commission Digital Markets Act: A translation,”, January 5, 2021,

[27] A provision in a contract in which a seller (or licensor) agrees to give the buyer (or licensee) the best terms it makes available to any other buyer (or licensee).

[28] DMA, para. 2(4).

[29] Ibid., para. 2(3).

[30] Directive (EU) 2015/1535, para. 1(1)(b).

[31] Ibid.

[32] George Westerman, Didier Bonnet, and Andrew McAfee, Leading Digital: Turning Technology into Business Transformation (Cambridge, MA: Harvard Business Review Press, 2014), p. 42.

[33] Henrique Schneider, Creative Destruction and the Sharing Economy: Uber as Disruptive Innovation (Northampton, MA: Edward Elgar Publishing, 2017), p. 63.

[34] Autorité de la Concurrence, Decision FNAC/Darty, 16-DCC-111, and Decision 18-D-16, July 27.

[35] 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, Vol. 8, No. 1 (2017), pp. 30-35,

[36] DMA Impact Assessment, Part 1.

[37] This sub-section is largely based on Schneider (above, Fn. 32) and David Teece and Henry Kahwaty, “Is the Proposed Digital Markets Act the Cure for Europe’s Platform Ills? Evidence from the European Commission’s Impact Assessment,” BRG Institute, April 12, 2021,

[38] Nick Kostov and Sam Schechner, “GDPR Has Been a Boon for Google and Facebook,” The Wall Street Journal, June 17, 2019, Mark Scott, Laurens Cerulus, and Steven Overly, “How Silicon Valley gamed Europe’s privacy rules,” Politico, May 22, 2019,

[39] Alexandre de Streel and Pierre Larouche, “The European Digital Markets Act Proposal: How to Improve a Regulatory Revolution,” Concurrences, Vol. 21, No.2 (2021), pp. 46-63,

[40] Geoffrey Parker, Marshall W. Van Alstyne, and Xiaoyue Jiang. “Platform Ecosystems: How Developers Invert the Firm.” Boston University Questrom School of Business Research Paper No. 2861574 (2016),

[41] Herbert, Hovenkamp, “The Looming Crisis in Antitrust Economics,” Boston University Law Review, Vol. 101 (2021), pp. 489-545,

[42] Adrian, Kuenzler, “Competition law enforcement on digital markets—lessons from recent EU case law,” Journal of Antitrust Enforcement, Vol. 7, No. 2 (2019), pp. 249-278,

[43] Katie Suominen, “On the Rise: Europe’s Competition Policy Challenges to Technology Companies,” Report, Center for Strategic and International Studies, October 26, 2020, Introduction,

[44] De Streel/Larrouche, Fn. 38, p. 52.

[45] Sumit Majumdar, “Stick Versus Carrot: Comparing Structural Antitrust and Behavioral Regulation Outcomes,” The Antitrust Bulletin, Vol. 66, No. 3 (June 23, 2021),  pp. 431-455,

[46] Mark Jamison, “Applying antitrust in digital markets: Foundations and approaches,” Boston College Intellectual Property and Technology Forum, April 20, 2020,

[47] Regarding Facebook: While the German Bundeskartellamt considers Facebook a monopolist and therefore applies remedies to curb its market-power, the Dutch competition authority insists that since there is no abuse in the economic sense, whatever market-power Facebook might have does not warrant antitrust intervention. Anne C. Witt, “Excessive Data Collection as a Form of Anticompetitive Conduct: The German Facebook Case,” The Antitrust Bulletin, Vol. 66, No. 2 (2021), pp. 276-307,