The DMA’s Little Piece of Heaven: The General Court Dismisses ByteDance’s Appeal Against its Designation Decision (Case T-1077/23, ByteDance Ltd v. European Commission)
The General Court dismissed ByteDance’s action seeking the annulment of the European Commission’s designation decision of its TikTok service. The Extended Composition of the Eighth Chamber of the General Court (GC) confirmed ByteDance’s status as a gatekeeper.
The ruling is the first of its kind: the first ruling to address head on the Digital Markets Act’s legal standards. ByteDance’s appeal basically revolved around three main pleas in law: i) the infringement of Articles 3(1) and 3(5) DMA; ii) the violation of its rights of defence; and iii) the infringement of the principle of equal treatment. The GC dismissed all three pleas. The judgment is most enlightening with respect to the first plea because it conflicts with ByteDance’s rhetoric that it is not to be captured via the DMA due to a vast array of reasons. To name a few, the gatekeeper argued it does not have an ecosystem to leverage its data and power across different services or end users substantially multi-home on TikTok and across other social networks.
Whilst agreeing with the European Commission in nearly every aspect of its designation decision, the General Court drew its attention to fleshing out some of the meanings of the DMA’s provisions and thresholds of intervention.
Articles 3(1), 3(5) and Recital 23 DMA: the rebuttal of the quantitative presumption
The DMA sets a specific regulatory framework governing the designation of gatekeepers, which is particularly characterised by two features. First, the European Commission performs designation within a short period of time through the presumptions set out under Article 3(2) DMA. By this means, if the undertaking surpasses the thresholds enshrined in that provision, then it is presumed to be a gatekeeper unless it manages to rebut it by presenting arguments to that effect. Article 3(5) establishes the legal standard for the rebuttal of that presumption.
Second, the designation process is subject to strict requirements, both at the procedural level and as regards the burden of intervention imposed on the undertakings as well as the applicable standard of proof. ‘Presumed’ gatekeepers must adhere to procedural requirements to submit their rebuttals and must meet a particular threshold of intervention to escape the DMA’s obligations. Article 3(5) highlights that the gatekeepers may present sufficiently substantiated arguments to demonstrate that, exceptionally, although it meets the thresholds in Article 3(2), due to the circumstances in which the relevant CPS operates, it does not satisfy the requirements listed in Article 3(1) DMA. Those arguments must, at least, manifestly call into question the presumption.
The standard of proof of the rebuttal
Against this background, ByteDance submitted a wide range of arguments to try to escape the DMA’s regulatory grasp. Quantitative and qualitative considerations were included in the exercise. Due to the regulation’s wish to avoid past mishaps from antitrust, Recital 23 narrows down the reasons which can go into rebutting the presumption (i.e., the interpretation provision only establishes quantitative considerations that may go into rebutting the presumption).
On its first plea, ByteDance argues that on the basis of this rationale, the European Commission incorrectly rejected the qualitative arguments and evidence it had submitted to fall outside of the DMA’s scope of application (para 36 of the ruling). Those arguments mainly related to its representation as an important gateway for business users to reach end users in line with Article 3(1)(b) DMA, namely TikTok’s lack of an ecosystem, the significant degree of multi-homing on its service, its scale vis-à-vis other services of the same CPS category and the minimal level of engagement of its advertisers. These are the most contentious sections of the ruling where the General Court substantially reviews whether the EC applied correctly the legal standard of Article 3(5) DMA.
Prior to that, however, the GC elucidated, once and for all, what the standard of proof for the rebuttal is: that of submitting sufficiently substantiated arguments manifestly calling into question the quantitative presumption (para 61). In the Court’s own words, the standard of proof is high. Undertakings must be capable of showing with a high degree of plausibility that the presumptions laid down in Article 3(2) are called into question (para 71). In turn, the GC recognises that there still is sufficient scope for leeway for gatekeepers to rebut their status as such. Any other interpretation would make the quantitative presumption de facto irrebuttable, which would be contrary to Article 3(5) DMA (para 50). As a matter of fact, a number of services have escaped the EC’s scrutiny by successfully rebutting that presumption, namely Microsoft’s Edge, Bing and Ads or Alphabet’s Gmail and Outlook.com.
By this same token, the General Court upholds that the content of Recital 23 is not particularly limited to quantitative considerations. It only excludes the possibility of gatekeepers submitting a justification on economic grounds seeking to enter into market definition or demonstrate efficiencies (para 46). Nothing else, in the GC’s words, should be discarded a priori as irrelevant because it is not expressed in figures, i.e., because it is not quantitative (para 47). In fact, the GC understands that the separation between qualitative and quantitative arguments playing a role in the rebuttal is merely artificial (para 40). The only limitation with the arguments that gatekeepers may submit is that they must be related to the quantitative thresholds laid down in Article 3(2) DMA (para 51).
The application of a standard of proof based on high plausibility
The General Court examines whether the standard of proof the European Commission applied to its designation decision is consistent with these findings (para 67). In short, the Court agrees that the EC correctly interpreted the DMA in applying the legal standard across the decision. However, it disagrees with the EC’s view when it drew into assessing the quantitative thresholds relating to measuring ByteDance’s significant impact on the internal market, as per Article 3(1)(a) DMA.
For fleshing out the fulfilment of this requirement, Article 3(2)(a) DMA provides an alternative set of thresholds. Either the EC must demonstrate that the undertaking meets the threshold of annual turnover or the market capitalisation/fair market value threshold. Surpassing any one of them will do for the presumption to apply. In ByteDance’s designation decision, the EC proved that ByteDance met the fair market value (accounting for its value at a global scale) although it did not meet the turnover threshold. However, ByteDance attempted to disprove the quantitative presumption by arguing that it did not meet the turnover threshold by far, as a supporting line of reasoning in the sense of Article 3(5) DMA.
At this stage, the EC did conflate the alternativeness of both thresholds and their relevance in terms of the rebuttal of the presumption. The General Court sets out that although the gatekeeper must not meet both to be captured by the presumption (because it would lead them to be de facto cumulative thresholds), that does not necessarily mean that ByteDance’s turnover in the Union is irrelevant for sustaining the rebuttal (paras 81-86). For instance, the undertaking may demonstrate that it has only a limited presence in the internal market by these means (para 90). This is the only instance within the whole judgment where the General Court asserts that the EC erred in law by taking a particular position regarding its interpretation of Article 3(5). Notwithstanding, since the argument would not cause a decisive effect on the designation’s outcome, the EC’s error does not bring further practical consequences to the undertaking (para 117).
Economic grounds, market definition, efficiencies vis-à-vis effects
The General Court also backed the EC’s interpretation of its impact in the internal market by not considering as sufficiently substantiated the fact that most of ByteDance’s revenue derives from China (paras 103-105). In this context, the General Court indicates that ByteDance could have demonstrated the existence of cultural and regulatory barriers in the EU segmenting its EU business from its Asian business (with some success, at that). However, the General Court points out that ByteDance failed to produce any evidence to that particular effect (paras 106 and 107).
In fact, most of the rest of the General Court’s assessment pivots around this same idea. Although the considerations ByteDance brought to the table were legitimate and could have played a part in the rebuttal, the gatekeeper failed to demonstrate the initial premiss of each one of them (paras 145, 201 and 202). At face value, however, the General Court’s assessment is, at least apparently, at odds with the intention of Recital 23. According to the GC, Recital 23 hinders the gatekeeper from submitting “a justification on economic grounds seeking to enter into market definition or to demonstrate efficiencies” (para 46). It is, thus, quite clear that the DMA does away with economics, market definition and efficiencies for capturing undertakings into its scope of application.
In reading the General Court’s interpretation of ByteDance’s failure to provide sufficient evidence to account for the qualitative arguments it submitted in the rebuttal process, both statements seem to be at odds one with another. The General Court’s appraisal of the ecosystem argument is good proof of that. In the notification of its gatekeeper status, ByteDance argued it did not have an ecosystem and, therefore, that this was a sufficiently substantiated argument to manifestly call into question the presumption. The European Commission disagreed with it and stressed that the concept is not included within the DMA’s scope.
Following the discussion, the General Court disagrees with the European Commission’s view and highlights that the concept of ecosystem is not explicitly included, but that it can be inferred from Recitals 3, 32 and 64 as playing a role in the regulation’s context. Having a digital platform ecosystem may be a relevant factor for the purposes of assessing whether the undertaking is a gatekeeper and whether it is an important gateway for business users to reach end users (para 131). The General Court, however, draws additional nuance into how the argument should operate in practice. Proving the existence or absence of an ecosystem is not sufficient to reach that conclusion. Instead, undertakings must prove the benefits and disadvantages associated with its existence or absence, which are the elements that make it possible to assess whether the CPS is an important gateway, in particular with regard to contestability (para 134). The applicant should have substantiated to the requisite legal standard that it did not have an ecosystem and specify its effects (para 139).
In my own mind, those ‘effects’, ‘benefits’ and ‘disadvantages’ wholly resemble precisely what Recital 23 is trying to back away from: factoring effects into the mix of the DMA’s analysis. As a matter of fact, how could an undertaking prove the existence of those effects, benefits and disadvantages in an ecosystem (or lack thereof) without taking recourse to economics?
In a similar vein, the General Court draws into the same kind of reasoning when demonstrating that ByteDance did not submit sufficient evidence to the effect of accounting for its CPS’ multi-homing. In the GC’s own words, multi-homing must be examined in relation to the circumstances in which the relevant CPS operates. To this effect, undertakings must present the specific and concrete characteristics of multi-homing of their services. This idea derives from the fact that the presence of multi-homing does not directly mean that all platforms of the same CPS category bear the same (and equal) importance to users. Instead, the GC asserts, the most relevant tenet of multi-homing is that of the intensity of its use and its importance for certain categories of users (paras 177 and 178). ByteDance did submit some evidence on multi-homing, but it did not concern its intensity. Therefore, it fell below the threshold of evidence required for this particular purpose (paras 189, 199 and 202).
Secondary findings and how to access the rebuttal: a holistic assessment?
Despite ByteDance’s failure to put forward evidence to demonstrate the validity of its qualitative considerations for succeeding in the rebuttal of the quantitative presumption when one observes the ruling from a broader perspective, the designation decision could not have resulted in any other outcome. Even though the General Court reviews the weight and scope of each of the concepts that can go into the rebuttal, it establishes (following the EC’s initial determination) the same conclusion over and over again. None of the arguments were, by themselves, sufficiently substantiated to call into question the quantitative presumption. This was the case for the interpretation of ByteDance’s significant impact on the internal market (paras 109 and 113), its lack of an ecosystem (para 161) and the significant proportion of TikTok uses that multi-home across services (para 185).
The General Court’s line of reasoning in confirming the EC’s interpretation is quite striking since it repeats that those arguments are not ‘by themselves’ sufficiently substantiated to call into question the presumption. However, Article 3(5) reads that the undertaking must submit those types of arguments to demonstrate it exceptionally should not be captured due to the circumstances the CPS operates in. At face value, the provision is quite unclear on how the EC should perform the analysis, i.e., whether it should take each argument individually and check whether it surpasses the standard of proof or whether all considerations should be analysed as a whole.
A range of alternative scenarios derive thereof. First, it may well be the case that undertakings must engage in a ‘silver bullet’ type of rebuttal. That is, the ‘presumed’ gatekeeper must present, at least, one argument that, taken individually, surpasses the standard of proof. That reason then may act as a lever to the rest of the considerations the undertaking may want to allege once it has accessed the market investigation procedure under Article 17(3) DMA. Second, the undertaking may have to ensure that every single one of its reasons meets the standard of proof taken individually. In other words, each one of them must meet, by themselves, the high plausibility standard of proof. Or third, the standard of proof works as a holistic assessment of all collated arguments presented for the rebuttal. The EC may, thus, determine whether those arguments considered as a whole meet the high standard of proof. Needless to say, the second scenario is the most demanding in terms of the legal standard required, whereas the third alternative places a more manageable threshold of intervention.
ByteDance raised this same point in the appeal by sustaining that the EC took a piecemeal and siloed approach in the decision by failing to carry out a holistic assessment of the evidence (para 329). In response to the allegation, the General Court highlights that all arguments did not meet the thresholds, whether taken individually or as a whole in the context of the presumptions applied to the CPS (para 334). A few lines after this general statement, the General Court seems to covertly recognise that the EC applied the second scenario set out above by stating that the “applicant does not put forward any other specific argument capable of demonstrating that the conclusion reached by the Commission would have been different had it assessed its arguments and evidence as a whole” (para 336, added emphasis).
In the plea in law relating to the EC’s potential infringement of ByteDance’s rights of defence, however, the contrary seems to be true. The undertaking alleged that the EC relied on matters of fact and law in relation to which it did not have the opportunity to submit its observations during the administrative procedure (para 338). Two of those were, precisely, ByteDance’s rebuttals surrounding the lack of an ecosystem and the presence of multi-homing. Despite that the undertaking was not given a real chance of contradicting the EC’s views, the enforcer holds that none of both were decisive findings with respect to the CPS’ designation and that they were secondary in nature in the decision’s scheme (paras 354 and 361). In this context, the EC hints, in some way, at the first scenario presented above: some considerations to rebut the presumption may impact designation in a silver-bullet-like fashion, whereas secondary findings cannot go into meeting the standard of proof by themselves. Therefore, for the requisite legal standard to be met, a combination of the first and second scenarios must be given so that there is, at least, one argument meeting the standard of proof and the secondary findings should be backed at least with convincing evidence.
Key takeaways
The General Court’s ruling demonstrates its deference towards the European Commission’s interpretation of the DMA’s legal standards. This is nothing particularly new to competition policy, but it is relevant in the context of the DMA’s departure from EU competition law standards (para 237). The regulation’s backing away from efficiencies and economics requires more (and not less) from the undertakings in terms of their procedural obligations. Anything that the gatekeeper did not include nor mention within the administrative procedure before the EC cannot sustain its further appeal before the EU courts (para 233). Again, this idea is not particularly surprising, since the President of the General Court already recognised the doctrine of exhaustion of administrative remedies applies in the judicial review of the DMA when dismissing ByteDance’s request for interim measures following its appeal (see a comment here).
Following this line of reasoning, the General Court has clarified some of the main points of contention surrounding the requisite legal standard for designation, i.e., a high degree of plausibility of the arguments in manifestly calling into question the application of the presumption. The question of how that degree of plausibility operates in practice remains, however, elusive in terms of the procedural requirements imposed on the European Commission.
