Last month, the European Commission introduced an ambitious new set of rules for digital platforms, the Digital Markets Act. Here is what it says, and how it can succeed.


For several years, authorities across the world have tried to understand the competitive dynamics of the digital economy in order to determine how to deal with the increasing economic and informational power of the Big Tech. They are now beginning to unveil their action plans. In the US, agencies are launching antitrust cases against the biggest digital platforms such as Google and Facebook. In Europe, governments are proposing ex ante rules to increase digital markets’ contestability and fairness.

On December 15, 2020, the European Commission proposed an ambitious new set of rules, the Digital Markets Act (DMA), which could be adopted by the European legislature by mid-2022. Six months after its adoption, it will apply to ten to fifteen global digital platforms, including Google, Apple, Facebook, Amazon and Microsoft. Here is what it says and what it means. 

Features of the Digital Markets Act

The scope of the proposed DMA is limited to a closed list of eight types of digital services that are directly defined in the draft law: intermediation services such as Amazon marketplace or Google and Apple app stores; search engines such as Google Search; social networks such as Facebook; video-sharing platforms such as YouTube; communication services without phone numbers such as Facebook Messenger; operating systems such as Apple iOS; cloud computing services such as Microsoft Azure; and advertising networks, exchanges, and intermediations services such as Google AdSense.

However, the obligations of the DMA will only apply to the providers of those digital services that are designated to be “gatekeepers” on the basis of a cumulative “three criteria test”, namely:

  • their large size and impact on the EU internal market;
  • their control of an important gateway for business users to reach end-users;
  • and whether the control in question is entrenched and durable.

This “three criteria test” is presumed to be satisfied when a platform meets  the following quantitative thresholds during three consecutive years:

  • a turnover equal or above €6.5bn ($7.9bn) or market capitalization of at least €65bn ($79bn);
  • the presence in at least three of the 27 Member States of the European Union;  
  • and a reach of more than 45 million monthly active end users (which represent 10 percent of the EU population) as well more than 10,000 active business users on an annualized basis.

However, this presumption of gatekeeper status based on those crude quantitative thresholds can be rebutted by the digital platform if it demonstrates convincingly that the “three criteria test” has not been met on the basis of a series of qualitative indicators, such as the fact that their users are also using competing platforms (multi-homing) or the existence of low entry barriers for potential competitors. Thus, the gatekeeper designation can be quick (when based on quantitative thresholds) but is not dirty (as platforms may provide qualitative indicators to overcome such a designation). In practice, presumption is used cleverly to compel digital platforms to disclose relevant information and to reduce the very large information asymmetries between digital platforms and their regulators.

Those designated gatekeepers are subject to a “blacklist” of seven prohibited practices against which antitrust has proved not to be effective enough. Some prohibitions relate to specific types of conduct that limit users mobility across platforms (such Most Favor Nations clauses or anti-steering clauses) while other prohibitions relate to bundling different digital services. These prohibited practices are very detailed, favoring clear rules over the exercise of discretionary powers. However, the DMA contains an in-built dynamic mechanism which allows the European Commission to adapt the “blacklist” of prohibited practices as the many unknowns of the digital economy become known with the advent of enforcement experience.

Gatekeepers are also subject to a “grey” list of eleven prohibitions and obligations. Several practices aim to ensure market contestability, such as the prohibition of some forms of self-preferencing and business envelopment, the prohibition of conducts limiting users switching or multi-homing, and the obligations of data sharing and interoperability. Other practices aim to guarantee market fairness, such as the obligation for more transparency on AdTech intermediation services or to give access to app stores on fair, reasonable, and non-discriminatory terms. Since those prohibitions and obligations are sometimes broadly defined, they need to be specified by the European Commission on an individual platform basis and in dialogue with the regulated gatekeeper during which the costs and the benefits of the precise obligation to be imposed may be weighted.

“The European Commission… will see its powers increase substantially to the point where it is set to become the European equivalent of the US Federal Trade Commission, enjoying concurrent regulatory and antitrust powers.”

Finally, designated gatekeepers are subject to two additional transparency obligations: the need to inform the Commission—in its role as the EU’s top antitrust agency—about their planned mergers and acquisitions as well as the need to provide independently audited descriptions of the consumer profiling techniques which they are using. As sunlight is the best disinfectant, collecting such information could act as a “coercive regulatory device” and alleviate unfair and harmful conducts.

The monitoring of compliance with the blacklist and the specification of the grey list will be effected by the European Commission, which will see its powers increase substantially to the point where it is set to become the European equivalent of the US Federal Trade Commission, enjoying concurrent regulatory and antitrust powers.

Policy Choices Made by the European Commission

In proposing the Digital Markets Act, the European Commission made four key policy choices which will shape the activities of large digital gatekeepers in Europe in the years to come; three of these are in line with an European regulatory tradition which is different than the US practice while the fourth choice is new for Europe.

The first policy choice is to rely on ex ante regulation to complement—and not substitute—ex post antitrust in achieving contestability and fairness in digital markets. This is not a surprising choice. Europe has a long tradition of combining antitrust and regulation to pursue similar policy objectives with each of the two instruments focusing on their respective strengths. When Europe tries to remedy some market failures with antitrust and isn’t sufficiently effective, it then opts to pursue complementary regulation. Such moves have happened before, in the telecommunications sector with the regulation of international roaming charges or in the financial sector with the regulation of credit card interchange fees, and it is now happening in relation to digital gatekeepers.

The second choice is to allow digital gatekeepers to remain integrated vertically and across related markets in exchange for the compulsory opening of their platforms to providers of substitute and complementary services. Structural separation could only be imposed at last resort when behavioral access obligations are violated systematically. By doing so, the DMA seizes the benefits of having integrated platforms able to internalize the massive network effects of connecting users and their data while mitigating the risks of harmful corporate conduct. Again, this is not new. At the end of 1980s, when the telecommunications sector was liberalized in Europe, the Open Network Provision Programme allowed telecom operators to remain vertically integrated in exchange for the opening their networks to smaller entrants.

“Faithful to its ordo-liberal tradition, Europe views protecting the competitive process as a value in itself .”

The third choice is again not new, but is possibly more controversial. The DMA establishes market contestability, and to some extent market diversity, as specific policy objectives. Thus, in deciding upon the innovation trade-off, the DMA favors innovation by business users and new entrants over innovation by existing gatekeepers. In turn, this choice leads to prefer long term competition over short term efficiencies. There remain many uncertainties in terms of innovation theory as to whether business users and potential competitors are more innovative than existing gatekeepers but, faithful to its ordo-liberal tradition, Europe views protecting the competitive process as a value in itself. Moreover, in a European Union characterized by many small and mid-size enterprises and different cultural traditions and demographic characteristics, this is a policy choice that makes much sense to many stakeholders.

The fourth choice is a big step forward in the European integration. If adopted as proposed, the DMA will, for the first time, confer fully-fledged regulatory power to the European Commission instead of conferring these powers to the national authorities of its Member States. Europe may thus have a federal regulator for large digital gatekeepers at the Commission, next to the federal supervisor for systemic banks set up at the European Central Bank in the aftermath of the 2008 financial crisis.

This new role of the Commission will be challenging, in particular if it wants to share the same characteristics that EU law imposes upon regulatory authorities at the Member State level. In this case, the Commission should be fully independent from the companies it regulates but also from political power. The pursuit of such goals may be in tension with the apparent willingness of the Commission to play an increasing geo-political role.

The Commission should also have sufficient human and financial resources to deal with extremely complex and, at times, new issues in markets which are evolving quickly and often under high levels of uncertainty. Finally, the traditional accountability mechanisms of the Commission may need to be strengthened, for instance with more hearings at the European Parliament or stricter judicial review by the Court of Justice. 

A Need to Change Corporate Cultures

More fundamentally, the success of the DMA in creating more innovation opportunities for new entrants and fairer relationships between the digital gatekeepers and their business users will require a cooperative—rather than adversarial—relationship between the Commission and the large digital platforms.

In turn, this will arguably require a change in corporate cultures. A change for the digital platforms, which should accept being more constrained in their actions and take additional responsibility for preserving market contestability. A change for the Commission, which has to learn from the very platforms it will regulate—for instance, by developing AI tools to deal with zillions of data points which will need to be reviewed or by being more experimental in the design of its remedies. In practice, the Commission needs to move from its more traditional bureaucratic culture to something more akin to “geek” culture. This will be much more challenging than adopting a new law.