
As readers of this blog know, the core idea behind (and the principle underpinning) the DMA proposal is that a small number of large digital platforms have become “gatekeepers” in that they operate one or more important “gateways” for business users to reach end users. In this context, the DMA rightly proposes to impose on such gatekeepers a series of obligations to protect the contestability of the core platform services concerned, as well as to prevent them from adopting unfair behaviour vis-à-vis the business or end users that are dependent on these services. The DMA proposal does so by providing for a mechanism to designate gatekeepers (using a combination of qualitative and quantitative criteria) and laying down a list of obligations in Articles 5 and 6.
In this post we would like to touch on what we consider an alarming issue, namely the apparent expansion of the DMA – largely resulting from amendments brought in by MEPs (the Council is largely remaining on course) – beyond the principled architecture of the original proposal. While any draft Regulation going through the various steps of the legislative procedure will be subject to changes, it is important that the final product (which enforcers will have to implement, and the courts will have to interpret) retains its effectiveness and its overall coherence.
While the DMA proposal was tightly crafted by the Commission to respond to the abovementioned gatekeeper concern in a targeted and principled manner, it has progressively been inflated to comprise a greater number of core platform services (although in some cases, such as browsers and digital voice assistants, this extension is desirable) and an ever-wider range of (at times vague) obligations. As the legislative procedure often involves political bargains and compromises between the various parties (“you give me this and I’ll give you that”), there is a concern the final DMA will stray from the original principled approach of the Commission. However, the greater the scope of the DMA in terms of regulated undertakings and obligations imposed, the higher the risk it will be of reduced effectiveness in practice.
For example, some voices in the European Parliament have pushed the view that the DMA should capture a wider range of undertakings, so that companies such as Booking.com and AirBnB would fall within its scope (e.g., MEP Evelyne Gebhardt recently told Bloomberg that excluding AirBnB and Booking.com from the DMA would be a “red line”). However, as I have explained elsewhere, it is hard to see how such companies could be considered as “gatekeepers” when there is a significant degree of multi-homing on both sides of the platform; hotels often offer rooms in more than one platform (plus they operate their own direct channel), whereas users can easily compare offers across platforms before making a reservation.
A more bizarre example is the various efforts in the European Parliament to include within the list of “core platform services” services that do not involve any intermediation between business and end users in the first place, such as video streaming or music streaming [note that such services were considered but rejected by the Commission as they lacked multi-sided character; see the impact assessment supporting the DMA proposal, page 38]. Others such as ECON propose to include within the scope of the DMA “software as a service” (software accessed online through a subscription rather than being locally installed on a device) and impose specific obligations on software providers. Again, it is far from clear how software as such intermediates between business and end users; the provision of software is not a platform activity and cannot credibly serve as an important gateway for customers.
Turning to the obligations in Articles 5 and 6, most amendments submitted in the European Parliament seem to aim at clarifying the various obligations imposed on gatekeepers. Still, I was surprised to read certain amendments imposing wide-ranging and rather vague obligations. For example, according to the ITRE amendments gatekeepers should refrain from “unduly” tying or bundling services or requiring the acceptance of “supplementary conditions or services that, by their nature or according to commercial usage, have no connection with and are not necessary for the provision of the platform or services to its business users” (a wording reminiscent of Article 102(d) TFEU). Needless to say, it is hard to see how the regulated undertaking could comply with such obligations and self-assess its conduct ex ante. Even if these obligations are subject to a regulatory dialogue with the Commission, there are doubts as to whether this would suffice (after all, it appears that the Commission would have discretion as to whether to engage in the dialogue) or whether this would effectively grant unfettered discretion to the Commission to prohibit or authorize conduct when enforcing the DMA.
Overall, there are several risks to this expansion of the DMA discussed in this post, which lawmakers should carefully consider.
- First, an ever-expanding DMA will create new interpretative difficulties (especially with respect to openly worded obligations) and increase the risk of inconsistent application of the rules, giving rise to costly and lengthy litigation.
- Second, the DMA would no longer be a principled response to the well-identified problem of firms acting as gatekeepers between consumers and businesses and taking advantage of this position to impose unfair conditions and/or reduce contestability. Instead, it would become an inflated and potentially incoherent piece of legislation designed to address any (perceived) problem linked to digital markets.
- Third, the above may raise questions of proportionality should the DMA, once adopted, be subject to judicial review.
- Fourth, an inflated DMA will stretch the scarce resources of the Commission and negatively impact its implementation and enforcement. This would ultimately backfire and undermine the effectiveness of the DMA.
There is no question there is a need for the DMA and I am a strong supporter of the legislation. However, the EU institutions involved (the Commission, the Council, and the Parliament) should ensure that the DMA remains principled, targeted to well-identified problems, and enforceable. The DMA is not a general-purpose Regulation applying across the economy, nor can it solve all the problems in digital markets. It has a specific remit and goals. The DMA should stay true to its principles, lest it ends up a missed opportunity.