Becoming the DMA: The Parliament’s Compromise (for now)

Becoming the DMA: The Parliament’s Compromise (for now)

Writing about the DMA means jumping on a passing train (with a parachute). The draft Regulation is in constant flow: The Commission’s first serve was shelled with 1199 amendments proposed to Andreas Schwab, the Rapporteur for the European Parliament Committee on the Internal Market and Consumer Protection (IMCO) by his colleagues. Once Parliament agrees on a compromise text, the trilogue negotiations with the Commission and the Council will start. The IMCO’s compromise text of Oct. 3 (hereinafter Compromise Amendments) was discussed on Wednesday. About time for Philipp Bongartz to report on progress and submit his wish list for the upcoming platform rulebook.

Time to nut up or shut up

Originally, the DMA was meant to create harmonised rules ensuring contestable and fair markets in the digital sector. Critics pointed to a lack of vision: Which ultimate goal is the DMA meant to serve? Consumer welfare? Internal Market integration? Promoting European small and medium sized enterprises (SMEs)?

Accordingly, the  Compromise Amendments tweak Art. 1 (1) DMA in two ways. First, the DMA shall contribute to the proper functioning of the Internal Market. This pledge to the legal basis (Article 114 TFEU) is well known: The avoidance of Internal Market fragmentation through diverging national rules for platforms was included in the Commission’s draft already. It merely moved from Recital 6 to the main body. Second and more interestingly, the Compromise Amendments add that these goals are pursued so as to foster innovation and increase consumer welfare. The amendments were inserted from two different proposals (amendment 383 and 384). Notably, the mention of European undertakings, including SMEs, as also suggested in amendment 383, was rejected.

Consumer welfare, including innovation, is a trusted friend for (many) competition lawyers. Still the DMA insistently pursues an objective complementary but different from that of competition law (Recital 10). That is hardly credible if the DMA aims to achieve fairness only as a means to promote consumer welfare (so does Article 102 (a) TFEU). The DMA understood in that way is little more than a new tool to crack the same nut. Apparently, the legislator is hesitant to openly advocate a redistribution of rents (to the detriment of US companies). The Americans got the point long ago and found their own name for it: protectionism (see here and here). So please, DMA, get out of the closet.

Blurred: The DMA’s goal

Relationship status: it’s complicated

Another sensitive topic is how the DMA interacts with Member States’ platform rules, such as Sec. 19a GWB (German Act against Restrictions of Competition). Pursuant to Article 1 (6) DMA the Regulation is without prejudice to national competition rules for unilateral behaviour (beyond the abuse of a dominant position) insofar as they are applied to undertakings other than gatekeepers or amount to imposing additional obligations on them. To put the point another way, Sec. 19a GWB will only remain applicable to gatekeepers if (1) it is viewed as a national competition rule and only (2) insofar as it exceeds the obligations under the DMA.

Concerning (1): The notion of national competition rules is explained in Recital 9, i.e. rules,

“that are based on an individual assessment of market positions and behaviour, including its likely effects and the precise scope of the prohibited behaviour, and which provide for the possibility of undertakings to make efficiency and objective justification arguments for the behaviour in question.”

There are reasonable doubts as to the assessment of anticompetitive effects under Sec. 19a GWB. Most importantly, the anticompetitive effect of the practices illustrating the provision is presumed (instead of investigated on a case-by-case basis). Arguably, however, a rebuttable presumption is deemed sufficient in Recital 10 DMA as well as Recital 9 Regulation 1/2003. As a consequence, Sec. 19a GWB would qualify as a national competition rule.

Regarding (2): It is not exactly clear, what “additional obligations” national competition rules are allowed to impose. Only obligations for other practices than those caught under the DMA? Or stricter obligations for the same practices, as well? As Recital 9 states, national competition rules

“should not affect the obligations imposed on gatekeepers under this Regulation and their uniform and effective application in the Internal Market.”

Leaving this phrase largely untouched, the Compromise Amendments add, the DMA is without prejudice to the ability of Member States to impose the same, stricter or different obligations on gatekeepers in order to pursue other legitimate public interests (such as consumer protection or freedom of expression). Sticking to the Commission’s logic that the DMA and competition law aim at different objectives, that reads like a blank cheque for Member States to tighten the rules for DMA practices in their jurisdiction to the preferred level. This is in line with the justification of amendment 147, where the change to Recital 9 was proposed. MEPs Geoffrey Didier and Nathalie Colin-Oesterlé (EPP) pointed to the uncertainty described above and suggested a clarification in favour of national competition rules.

While doubts remain (e.g. rules pursuing “other legitimate public interest” seem to refer only to Article 1 (5) DMA), the above interpretation is supported by the newly addded Article 31d DMA. There, the coordination between the Commission and NCAs is put into more detail. NCAs launching investigations on gatekeepers under national competition rules as defined in Article 1 (6) DMA shall inform the Commission thereof. Where the investigations results in measures against gatekeepers, the Commission must be notified again. If the Commission concludes that the measure runs counter to obligations, decisions taken or contemplated in initiated proceedings under the DMA, the NCA shall refrain from adopting the measure. Apparently, the Compromise Amendments resolve the relationship between the DMA and national competition rules on the procedural level.

MEP drafting Recital 9 DMA.

Over the line

The circle of undertakings caught under the DMA is one of the hottest topics. There are two separate points of interest:

The first concerns the list of core platform services (CPS) to which the DMA applies. The Commission’s draft named eight: online intermediation services, search engines, social networks, video-sharing platforms, (number-independent interpersonal) communication services, operating systems, cloud computing and advertising services. The Compromise Amendments amplify the catalogue to web browsers, voice assistants and connected TVs. There was a wide range of suggestions that didn’t make it, such as streaming services (amendment 416, 417, 430), mobile payment services (419), embedded digital services in vehicles (428) and collaborative economy services (441). Equally rejected were proposals seeking to indroduce exceptions tailored to certain communication and cloud computing services (e.g. amendment 422). Bravo! Most notably, however, Article 3 (1) (b) of the Compromise Amendments expands the notion of “gatekeeper” to platforms serving as gateways for end users to reach other end users, thereby drawing C2C networks (like WhatsApp) into the Regulation’s scope (and aligning it with Sec. 19a GWB).

The second and even more furious controversy orbits the quantitative thresholds that trigger the rebuttable presumption of a gatekeeper position. Starting with the back-of-the-envelope calculations by Cristina Caffarra and Fiona Scott Morton, several authors speculated evaluated which undertakings would fall under the thresholds. Among the recurring names were those of European platform operators such as (NL), SAP, Zalando (DE) and Orange (FR).

On a closer look the GAFA companies’ names appear.

Following the advice of the targeted companies, their funded advocates, the governments of the Netherlands, Germany and France but also the German Monopolies Commission, the Rapporteur proposed narrowing down the scope of addressees by raising the thresholds and introducing an Uncle Sam ecosystem criterion. Pursuant to the latter, the quantitative thresholds will be complemented by a fourth condition i.e., operating at least two CPS. The point is to focus the Commission’s ressources on “platforms that play an unquestionable role as gatekeepers” (amendment 37). The proposal entered the Compromise Amendments.

The Commission has considered that criterion before. In the Impact Assessment (para. 148) it estimated that the circle of addressees would thereby drop from 10-15 to 5-7 companies. As main drawback of tightening the scope, it identified the possible temporary fragmentation resulting from the automatic designation of some gatekeepers while others are appointed via a longer market investigation (para. 369). According to Article 15 (1) DMA, the market investigation required to designate gatekeepers by qualitative criteria takes up to twelve months. As a solution, the Compromise Amendments suggest to cut the Commission’s deadline down to six months. There are reasonable doubts as to the feasibility of this period: The designation under Sec. 19a GWB was estimated to take twelve months. From the ongoing four designation procedures – the first started in Jan. 2021 – none has been concluded. In any event, it seems far from clear whether trimming back the presumption will disburden the Commission (or translate into more market investigations).

Echoing the controvery of this, the MEPs decided in Wednesday’s debate to postpone their vote on the Compromise Amendments (originally scheduled for Nov. 8), as they could not agree on the scope of addressees. According to Bloomberg, MEP Evelyne Gebhardt (PES) drew the exclusion of and Airbnb as a “red line”.

To the core

At the heart of the DMA lie the (originally) eighteen dos and don’ts imposed by Articles 5 and 6. Reporting on all changes to these obligations would be equal to writing George RR Martin’s sixth book. Subsequently, only the five most astonishing proposals of the Compromise Amendments will be picked.

(1) The per-se prohibition of Article 5 (b) DMA shall extend to narrow parity clauses. That is perfectly consequent: the divergent treatment of parity clauses across the Member States was a major argument for harmonising platform rules (see Impact Assessment, Annex 5.4). While the Commission’s draft banned wide parity clauses only, however, some Member States (such as Germany, France, Austria, Italy and Belgium) took action against narrow parity clauses, as well. Adding to this, other obligations such as Article 5 (c) would essentially run dry if narrow parity clauses were available to gatekeepers.

(2) In Article 6 there is a new obligation to allow providers of equivalent CPS to interconnect with gatekeepers’ communication services and social networks. To put it more plainly, WhatsApp and Facebook shall ensure “interoperability” with their competitors’ services (e.g. Signal). Apparently, this duty is owed to the Joint Research Centre noting that the interoperability requirements under Article 6 (f) DMA do not apply to instant messaging and social media (p. 22).

(3) The obligation of app stores to use FRAND terms of access (Article 6 (k) DMA) is expanded in two ways: First, it shall apply to all CPS. Second, the conditions may not be less favourable than the ones applied to the provider’s own services. Together, these expansions amount to a general ban of self-preferential conditions of access to CPS. You have no power here, essential facility doctrine.

(4) Article 6 shall contain a new prohibition: combining personal data for the purpose of targeted advertising without a clear, explicit, renewed informed consent. Especially the CMA (at 2.18) pointed to Google’s and Facebook’s competitive advantage in the ad tech sector as they are able to round out user profiles with data from different sources (e.g. search engine, web browser, social network, communication service).

(5) The most stunning amendment is a ban on framing decisions of users. Under Article 5 (self-executing!) gatekeepers shall be prohibited from manipulating users’ autonomy, decision-making or choice via the structure, design, function or manner of operation of their online interface or any part thereof. The prohibition was proposed in amendment 599 by MEPs Martin Schirdewan and Anne-Sophie Pelletier (The Left). Originally, it was meant as an extension of Article 5 (a) (roughly, the Facebook case). In that way, gatekeepers should be prevented from obtaining their users’ consent to data commingling by framing user decisions accordingly – a risk identified in RP’s unrecent paper. Probably few expected that this would evolve to a stand-alone prohibition. Well, I’m dashed.

A pastry commonly associated with decision framing.

To round up: The Compromise Amendments do not contain the much debated general clause for new practices. In Article 22 (2a) DMA, however, it provides for interim measures to avoid serious and immediate damage resulting from new practices (amendment 1069 proposed by Renew Europe). Having advocated this in our policy paper, Rupprecht Podszun, Sarah Langenstein and me could barely read the passage without weeping for joy. Humble bragging over.


According to the Commission’s draft, the obligations should leave very little room for exceptions. Articles 8 and 9 DMA allow for a suspension or exemption only under exceptional circumstances, i.e. a threat to economic viability or overriding reasons of public interest, respectively. In the Commission’s view, the designation procedure should not provide a forum for efficiencies either. As stated in Recital 23, the Commission will discard any efficiency defense invoked to rebut the presumption of being a gatekeeper.

Upon suggestion of the Rapporteur (amendment 3) this part of the Recital was deleted from the Compromise Amendment. Undertakings falling under the presumption shall have the chance to present “sufficiently compelling arguments” (including effiencies, apparently) to rebut the presumption of holding a gatekeeper position.

That mirrors not exactly what even proponents of the efficiency defense recommend. Take for example the Monopolies Commission (p. 59 et seq.): Its draft wording for an efficiency defense can only relieve from individual obligations, not from the gatekeeper position as a whole. Shaking off the gatekeeper status means, the CPS is not subjected to any obligations at all. This all or nothing option is probably not the kind of flexibility expected from the efficiency defense.

A stride with seven-league boots

Alongside the circle of addressees, the role of national competition authorities (NCAs) in enforcing the DMA is the crossroads where minds divide. While some authors point to the Commission’s expertise and the harmonising effect of central enforcement, others endorse a stronger involvement of NCAs, as their investigations have revealed some DMA practices in the first place.

The Commission’s draft placed the Member States’ representatives in a Digital Markets Advisory Committee with the power to, well, express their opinion. In addition, they may request a market investigation for designating gatekeepers. If submitted by three or more Member States. With no binding effect. In this architecture, the involvement of NCAs is more symbolic than real.

Buckingham Palace: prospective seat of the Digital Markets Advisory Committee.

That triggered calls from both Member States (Vol. 1 & 2) and the European Competition Network (ECN) for more participation. The governments of the Netherlands, France and Germany originally suggested to expand the Member States right to request market investigations (e.g. into new practices). In their follow-on position paper, partially in line with the ECN, the governments designed a referral system in the style of the EU Merger Regulation. This would allow the Commission to outsource investigations and monitoring while retaining sole responsibility for gatekeepers’ designation, granting exemptions and regulatory dialogue.

In an earlier version, the Compromise Amendments contained a paragraph partly resembling the proposed referral system (Article 31c (2b) DMA). It was however removed upon advice of the European Parliament’s Legal Service. Instead, NCAs may support the Commission’s market investigations (also for new practices, new services and systematic non-compliance) as well as certain proceedings, including regulatory dialogue (!) and monitoring. The latter means, the NCAs may – under coordination of the Commission – request information, carry out interviews and on-site inspections.

On top of that, the above mentioned request for a market investigations receives a threefold upgrade: (1) It requires not three, but only two NCAs. (2) It expands to market investigations into new practices, new services, systematic non-compliance and the request to adopt a non-compliance decision. (3) Lastly, the Commission must reason and publish its decision to reject a request.

In several contributions (here, here and here), the editor of this blog (commonly known as RP) pointed to the blatant lack of provisions on private enforcement in the DMA and proposed the establishment of a Platform Complaints Panel. Some elements of this seem to be reflected in Article 31c (2a) of the Compromise Amendments. NCAs shall collect business users’, end users’ and competitors’ complaints about practices falling under the DMA. Subsequently, the complaints will be passed on to the Commission.

In summary, NCAs and private parties gained more rights under the Compromise Amendments. Compared to the Commission’s draft, this means a remarkable shift.

Coming up

The European Parliament has so far reached an ambitious compromise. The vote thereon is delayed, however, because social democrats and greens (reasonably) disagree with the downsized circle of adressees. In the upcoming trilogue negotiations the final text will clash with the ones of the Commision and the Council. The delay will be used by Parliamentarians to listen to Facebook whistleblower Frances Haugen – probably not an event that will ease pressure on Facebook Meta and the like.

Reportedly, the Council’s compromise is far less aspiring. To name some examples: Web browsers and voice assistants are not included in the list of CPS. The Council prefers a narrow notion of “gatekeepers”. There are no major amendments to the obligations. The role of NCAs is less developed, but rather limited to certain investigations in their territory.

So once the IMCO has got its compromise straight, there will be much on the line in the trilogue. Do you remember the feeling of waiting for the next episode (before Netflix)? There it is.

Philipp Bongartz is a fully qualified lawyer and a researcher at the Chair for Civil Law, German and European Competition Law at Heinrich Heine University Düsseldorf. He is pursuing his doctoral studies with Professor Rupprecht Podszun.

Leave a Reply

Your email address will not be published. Required fields are marked *