While many companies active in digital markets support the Digital Markets Act (“DMA”) proposal, a recurrent concern comes from the fact that the obligations contained in Articles 5 and 6 of the proposal will in principle apply to all gatekeepers independently of the core platform service (“CPS”) they deliver and their business model. While Article 5 contains “obligations for the gatekeepers”, Article 6 contains “obligations for gatekeepers susceptible of being further specified.” In the case of Article 6, the preamble suggests that the Commission may engage in a “regulatory dialogue” with gatekeepers to tailor these obligations; but it does not suggest that this “tailoring” may allow digital gatekeepers to escape them.
Thus, digital platforms concerned they could be designated as gatekeepers, but also neutral observers, regularly point to the fact that while some of the obligations contained in Articles 5 and 6 may be particularly suited to address concerns that have been expressed vis-à-vis one specific CPS (e.g., a search engine), it may not make sense to apply the same obligations to another CPS (e.g., a marketplace), and vice-versa. Indeed, some of these obligations can be associated to issues addressed by DG COMP in the context of past or ongoing competition investigations against certain large platforms (e.g., Article 6.1.(d) on self-preferencing is a direct reference to Google Shopping).
True, Article 8 provides that the Commission may, on a reasoned request by the gatekeeper, suspend in whole or in part a specific obligation laid down in Articles 5 and 6 for a CPS, but only in exceptional circumstances. In particular, Article 8 requires the gatekeeper to demonstrate that compliance with a specific obligation would endanger, due to exceptional circumstances beyond its control, the economic viability of its operation in the EU. Thus, unless a gatekeeper can show that a specific obligation would threaten its viability (arguably a very high threshold to meet), it may have to comply with an obligation that does not really fit with the services it delivers.
Before the DMA proposal was published, I had thought that this problem would be avoided by creating two lists of obligations, a “blacklist” of prohibitions that would apply to all gatekeepers, and a “grey list” of obligations that would apply to some gatekeepers, depending on the CPS they provide. Yet, for reasons I do not fully understand, this is not what happened.
Of course, some might say that the various obligations contained in Articles 5 and 6 are what they are and that we should not make things over-complicated as a bird in the hand is worth two in the bush. Some may also think that any attempt to modify Articles 5 and 6 could be used to water down the obligations they contain. These are certainly fair points, but my view is that to the extent we decide to impose obligations on digital gatekeepers, it is better that they fit the gatekeeper and its services as otherwise the DMA will not achieve its intended objective.
In fact, the risk that Articles 5 and 6 may be watered down during the legislative process may be higher if it becomes obvious that some obligations would negatively impact certain CPS providers without bringing any countervailing benefits to their business or end users. On the other hand, by tailoring the obligations to the business models of CPS providers, one could potentially impose stricter obligations than those contained in Article 5 or 6. As the CMA has recognized in its strategy to tackle large digital platforms, the right obligations need to be imposed on the right set of activities, preferring a more flexible approach than the one pursed by the European Commission.
In fact, the DMA proposal already recognizes that some obligations should only apply to some CPS providers. For instance, Article 5(g) only applies to gatekeepers providing advertising services, Article 6.1.(c) applies to gatekeepers operating app stores, while Article 6.1.(d) applies to gatekeepers operating search engines. But I still do not find this satisfactory because although some obligations contained in this list seem particularly relevant to tackle problems that have occurred in the context of a given CPS (as, for instance, illustrated by a DG COMP investigation), they would nevertheless apply to other CPSs.
There are, however, two simple ways to address this problem. First, Article 6 could make it clear that it contains obligations that may apply to some CPS providers and not to others, for instance following the regulatory dialogue that the Commission is expected to hold with the providers designated as gatekeeper. Second, Article 6 could in the alternative contain distinct lists of obligations that would be CPS-specific. Either of these approaches would be simple to implement and would not lead to a weakening of the DMA in any way.
This would also have several benefits. First, the goal of the DMA should not be to undermine gatekeepers, but to make sure they are subject to the right set of obligations to prevent them from taking advantage of business users that depend on access to their platforms and/or limiting marketing contestability. Second, forcing gatekeepers to implement obligations that are not tailored to their activities and business model will trigger a great deal of resistance, hence increasing implementation costs and reducing compliance. It will also give credence to those who criticize the DMA proposal as being unworkable. Third, it will be a good test for the legislator to match the concerns identified with respect to certain CPS with the obligations to be applied to their providers. In some cases, it might become obvious that some of the obligations contained in Article 6 do not fit, or need to be clarified, or in some cases need to be strengthened.