As the European Commission is digesting the (voluminous) results of its public consultation on the upcoming Digital Services Act (“DSA”) package and preparing a proposal for what is described in its “Inception Impact Assessment” as an “ex ante regulatory instrument for large online platforms with significant network effects acting as gate-keepers in the European Union’s internal market”, I would like to reflect on the meaning of the notion of “gatekeeper”.
Indeed, if the ex ante regulatory instrument will only apply to “gatekeepers”, it is critically important to define this concept. Of course, we don’t know yet whether the Commission will use the term “gatekeeper” to define the online platforms that will be subject to the dedicated rules that it intends to include in its ex ante regulatory instrument. A variety of labels have indeed been used in the various reports that have been produced to discuss what needs to be done with regard to large online platforms – including terms such as platforms holding a “strategic market status” (Furman review), platforms holding “bottleneck power” (Stigler report), “structuring digital platforms” (Autorité de la concurrence), etc. But to make matters simple, I will assume that ex ante regulation will apply to “gatekeepers” as it is a term that is used not only in the Commission’s “Inception Impact Assessment”, but also in the DSA package consultation.
The terms of reference (“ToRs”) of the “Platforms with Significant Network effects acting as Gatekeeper Impact Assessment Study” that was commissioned earlier this year by the Commission provide some indication of the types of platform that could fall within the scope of ex ante regulation. The ToRs suggest that the regulation should cover “online platforms, i.e. digital services that facilitate interactions via the Internet between two or more distinct but interdependent sets of users,” including “online marketplaces, app stores, search engines, social media and platforms for the collaborative economy.”
The ToRs identify a series of features that could be used to identify the online platforms that would be subject to ex ante regulation, including the fact that:
- They are “driven by strong economies of scale, and direct and indirect network effects [and] increasingly act as private gatekeepers to critical online activities for an exceptionally large population of private and business users.”
- Their “gatekeeper role is enabled inter alia by their hold over vast amounts of data and in some cases very large customer bases.”
- They enjoy a “systemic ability to cement and even expand their critical gatekeeping roles, including in other ecosystems, to raise barriers to entry and expansion for rivals and to increase their hold and leverage over their users.”
- Their “large competitive advantage due to the economies of scale and of scope, reinforced by data-driven network effects” allows them “to act as private regulators setting the rules of the game on the markets they control.”
- The ability to leverage their core abilities (data, customer base, technological assets, etc.) to enter and potentially conquer new markets with relative ease “may lead to unusually large commercial imbalances and bargaining power between platforms on the one hand and their users and rivals on the other.”
Similar factors can be found in the public consultation for the DSA package and it is fair to say that it is widely recognized that these factors are in one way or another relevant to define the “gatekeeper” platforms to which ex ante regulation should apply. The challenge is of course to identify which of those factors should be central to the definition of a gatekeeper and how the selected factors should be organized.
In my view, the core element of a definition of “gatekeeper” is that it “controls access” to critical online services which allow to reach a large category of users. For instance, the App Store controls access to iOS device users as it is the only way for app developers to distribute their apps to iOS users. But a gatekeeper is not necessarily the only access point to a given category of users. It is sufficient that it controls access to a sufficiently large category of users for access to its platform to be critical for business users to compete on a market where these users are customers. For instance, 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.
As to the other factors referred to in the Commission ToRs or the DSA public consultation, such as the presence of barriers to entry (e.g., large economies of scale and significant network effects) and market power, they can be the cause or the consequence of the gatekeeping position enjoyed by some platforms.
For instance, the App Store is a gatekeeper due to the presence of an insurmountable barrier to entry, which is that Apple does not allow other app stores on iOS devices. Hence, no rival app store can emerge. In the case of online search, while Google cannot prevent the entry of competing search engines, the presence of large economies of scale and significant network effects have made it impossible for competing search engines to grow their user base. It is in turn the presence of high barriers to entry that gives a gatekeeper platform market power on the market for the critical online services to which access is critical to compete. In the absence of barriers to entry, the platform would likely see its market power erode.
As to the ability of the platform to “leverage their core abilities” to enter and potentially conquer new markets, it directly relates to the fact that it holds market power on the market for the critical online services. As illustrated by the Google Shopping case, the market power on search can then be leveraged to a neighbouring market (comparison shopping services). Their gatekeeper market power also allows these platforms to behave like “private regulators” in that business users that need to have access to the platform have no choice but to accept the rules developed and enforced by the platform, even if such rules are unfair or discriminatory.
Now, from a policy standpoint, although they may be closely interlinked, it is not a bad thing to include several factors in the definition of a “gatekeeper”, such as the facts that (i) the platform “controls access” to critical online services for business users to compete on a relevant market, (ii) is protected by high barriers to entry and (iii) holds market power on the market for the critical services it delivers to business users. To avoid the definition to be overinclusive, I would make these factors cumulative. In order to avoid false positives, I would add a temporary dimension to the market power requirement of the above definition, by requiring that the market power enjoyed by the platform is “enduring”.
It has also been suggested that in addition to qualitative factors, such as those suggested above, the definition of a gatekeeper should also include some quantitative factors. Adding quantitative factors could be another way to avoid capturing platforms, whose role in a sector may be important, but may not justify that they be subject to ex ante regulation (because, for instance, they are small or their activities are limited to one Member State). Some may also say that this is a way to “objectivize” the definition, although if it is combined with qualitative factors it is hard to see how any definition could be sufficiently objective to be self-executing. There is no question that the identification of those platforms to be considered as gatekeepers for the purpose of the ex ante regulation will require the intervention of a “regulator” that will have to look at all the factors and take a decision.
The use of quantitative factors or thresholds is in fact common in EU law. A well-known example is merger control where turnover thresholds are used to determine whether a concentration needs to be notified to the European Commission. The question is of course which kind of quantitative factors should be used as relevant factors for the determination of a digital gatekeeper. Here are several candidates:
- The turnover of the platform concerned. This factor has the merit of simplicity. As in the case of merger control one could look at the global turnover as well as the turnover realized in the EU. The turnover threshold should be sufficiently high to avoid the definition to be overinclusive. A worldwide turnover of €10 billion and a turnover of €1 billion in at least three Member States could for instance be a starting point. Before selecting such thresholds, one would need to verify whether few or many platforms would meet these thresholds to ensure that these thresholds are neither too high nor too low. These thresholds should be higher than those in merger control because the consequence of exceeding these thresholds (while also satisfying the various qualitative factors) would be the automatic imposition of an onerous set of rules.
- Another criterion could also be the number of platform users. This factor also has the merit of being simple. The difficulty however is to agree on what should be the threshold number of users. Once again, the number should be neither too high nor too low.
- Other factors, such as the volume of data held by a platform, would be hopelessly hard to calculate and should have no place in the definition.
It has been suggested in some quarters that it would be wrong to have a definition of gatekeepers that would capture a small subset of US-based companies, such as the so-called GAFAs. To avoid looking protectionist, the definition should at least include a few non-American companies, so the argument goes. I really do not think such considerations should have any place when trying to come up with a meaningful definition of the notion of gatekeeper. Framing a definition on the basis of geopolitical considerations to ensure that it captures certain companies would be fundamentally wrong and unfair to the platforms that would be covered by the definition.
I also believe that platforms should not be designated as gatekeepers once and for all. A review process should thus be put in place to take market realities into account. For instance, while companies like Google and Facebook have certainly enjoyed enduring market power, and it is hard to see who could dethrone them in the short and medium term, this market power may fade away at some stage.
In sum, the definition of digital gatekeeper should be based on several qualitative criteria (such as those listed above), possibly in combination with easy to measure quantitative factors. The determination of which platforms would fall within the definition should be made by a regulatory authority to be based at the EU level to ensure consistency across the internal market. This regulatory authority should not necessarily be an entirely new regulatory body as setting it up would certainly delay the implementation of the ex ante framework. It would be wiser in my view to rely on existing institutions within the European Commission.
As a final point, like many Brussels observers, I have had access to several leaked documents from the Commission, including impact assessments of the DSA package and a list comprising “unfair” practices. The intentions of those leaking these documents are not always clear, but this is a bad practice. Assuming these documents are genuine, they show the strong resolve of the Commission to tackle the issues raised by digital gatekeepers. At this stage, the question is no longer whether regulation should be adopted by what it should look like. More commentary on this will follow.