On April 24, 2024, the German Federal Court of Justice (FCJ) rejected Amazon’s appeal against the decision of the Federal Cartel Office (FCO) of July 2022, designating Amazon as a gatekeeper under national competition law, Section 19a ARC. The ruling clarifies important aspects of this provision designed to capture large digital companies (see translation of the FCO decision here). The ruling is long for the FCJ (117 pp.), reflecting that here the court exceptionally acts as first and only appeal instance regarding Section 19a ARC and deals with questions of law and facts.
Broad scope of appeal
Amazon raised a broad scope of challenges against the FCO decision, claiming that Section 19a ARC itself is illegal, because unconstitutional and infringing EU law, i.e., the DMA, EU directives in the field of Information Society services, the EU Charter of Fundamental Human Rights and the ECHR. Amazon also said that it does not meet Section 19a ARC’s material conditions, and that the subsequent designation of the company as a gatekeeper and of some of its activities as core platform services (CPS) under the DMA would supersede Section 19a ARC. Amazon argued that the facts have changed in the meantime due to the now applicable DMA obligations and the commitments Amazon gave to the Commission in the Spotify case in December 2022. The FCJ rejected all of these challenges and confirmed the FCO’s designation decision. This blog will focus on the parts of the ruling dealing with Section 19a ARC and its relation to the DMA.
Background: Section 19a ARC
Under this provision, the FCO can designate companies as gatekeepers to be subject to specific abuse rules under national law for a duration of 5 years. Once a company is designated, the FCO can intervene and prohibit certain conduct, limited to specific practices set out in Section 19a(2) ARC (e.g., self-preferencing, etc.). The designation has two conditions: (i) the company must be active to a significant extent in multi-sided markets referred to in Section 18(3a) ARC, and (ii) the company is of paramount significance for competition across markets. Amazon argued that it does not meet either condition. The FCJ rejected these points of appeal.
Active to a significant extent on multi-sided markets
The FCJ refers to the legislative recitals that the “significant extent” condition aims at excluding companies for which platform or network activities only play an entirely subordinate role in relation to (i) the companies’ other activities or (ii) competitors in the relevant markets.
Amazon’s intermediation services Amazon Marketplace and Amazon Advertising count as relevant activities in multi-sided markets. The court confirms that they are “significant” based on their worldwide and German turnover, the number of active 3rd party sellers and monthly active end users on Amazon Marketplace, and the number of items offered by 3rd party sellers. The FCJ also refers to Amazon’s gatekeeper notification to the Commission under the DMA in 2023 that the two services met the core platform service conditions under Art. 3(2) lit. b DMA.
Amazon argued that its primary business Amazon Store (its own online retail offers on its marketplace) is not a multi-sided market activity, and that the turnover with Amazon Marketplace (services for 3rd party sellers) accounted for only less than 30% of its total annual turnover. However, the FCJ says that there is no minimum 30%-turnover threshold in this context. That Amazon Marketplace alone accounted for 22% of Amazon’s total worldwide turnover in 2021 is deemed sufficient for not being “entirely subordinate”.
Paramount significance for competition across markets test
The court says that Section 19a(1) ARC requires a comprehensive review of all aspects in determining whether a company has such a position, including the criteria listed in the provision: dominance, financial strength, vertical integration, access to data relevant for competition, relevance for third-party access to markets and related influence on third parties’ business activities. The FCJ confirms that the list is not exhaustive, that the criteria do not need to be met cumulatively and that their order of mention is irrelevant. Before dealing with these criteria, the FCJ rejects several of Amazon’s general pleas related to the test.
No requirement of actual risk for competition
Amazon claimed that paramount significance for competition across markets requires finding that the company’s conduct creates actual risks for or already harms competition. The FCJ says the provision and the listed criteria only relate to strategic and competitive capabilities of a company, rather than specific market conduct. The rule’s objective is to enable the FCO to exercise a more effective abuse control over a few large digital companies, given that the characteristics of digital platform markets and gatekeepers may require swift intervention. Section 19a(1) ARC addresses these special, potential risks resulting from increased vertical and conglomerate possibilities to abuse economic power.
The FCJ also points out the distinction between designating a gatekeeper under para. 1 and intervening against specific practice, para. 2 of Section 19a ARC. Only an intervention requires identifying specific conduct and related actual risk. In contrast, designating a gatekeeper relates to the abstract risk potential for competition based on its strategic and competitive possibilities.
Test sufficiently specific
The FCJ rejects Amazon’s plea that the provision violates the principles of norm clarity. The use of abstract terms in laws is permissible, if the party concerned can capture the legal situation and adjust its conduct accordingly. This is the case with the notion of paramount significance for competition across markets, as it is further defined by the listed criteria. In addition, the court notes that the term is not vaguer than other statutory antitrust terms, like abuse of dominance, market definition or SIEC.
Not a relevant digital ecosystem?
The FCJ rejects Amazon’s claim that Section 19a ARC cannot apply to it, because it does not operate a digital ecosystem within the meaning of the rule. Amazon argued that Section 19a ARC only covers ecosystems characterized by highly complementary goods and services and high switching costs, preventing users from multi-homing. Amazon said it does not control the interoperability of the various products/services offered, and that users are not locked in.
The FCJ clarifies that Section 19a ARC does not only cover the type of ecosystem Amazon described, i.e., a multi-product ecosystem. Multi-actor ecosystems also fall within Section 19a ARC, i.e., ecosystems in which the operator of the intermediation platform has a dual role (platform operator and competitor to business users on the platform), allowing the operator to gain comprehensive and exclusive access to data of end users and of its competitors, which may facilitate expanding the operator’s activities into adjacent markets. Amazon is said to be a prime example for such a multi-actor ecosystem.
Amazon meets various criteria for paramount significance for competition across markets
Vertical and conglomerate integration
The court refers to Amazon’s many up- and downstream or neighboring activities around its online marketplace, which are interconnected and typically accessible for end users via a single user account. Amazon can use the data collected across services, benefit from indirect network effects and from its dual role. Amazon’s counterarguments, including that cloud computing is operated separately from its online marketplace, and that there is still (residual) competition in some areas, would not speak against the abstract risk for competition resulting from Amazon’s various integrated services and activities.
Financial strength/access to data relevant for competition
The FCJ notes that– unlike in dominance cases – under Section 19a(1) ARC these criteria require an analysis across markets, not in relation to specific markets. The FCJ confirms that Amazon has paramount financial strength, based on key financial indicators (turnover, turnover growth profits, cashflow). Amazon has inter alia used its financial strength to acquire more than 100 companies since 1989.
The FCJ finds that Amazon has paramount access to data relevant for competition. Amazon collects vast amount of data from users, on the products offered on Amazon Store/Marketplace and on and from third parties. Amazon combines the data across services and devices, using a variety of machine learning tools to analyze them. The FCJ rejects Amazon’s argument that competitors would do the same: the scope of data available to Amazon is much broader and deeper, as competing online marketplaces do not have comparably integrated activities. The FCJ says that the FCO did not need to provide a case-by-case quantification how the data can increase market power vis-á-vis competitors or how the data can be used to leveraging market power into other activities.
Relevance for third-party access to markets
The court says this is a key criterion, relating to the competitive capabilities created by the company’s intermediation and regulatory power. It is not necessary to prove that the company indeed uses its powers in any anti-competitive, unfair or discriminatory manner. The FCJ confirms the FCO’s findings that Amazon has significant relevance for third-party access to markets and for their business activities in three areas:
Amazon Marketplace: the general t&cs for 3rd party sellers set access conditions to the platform, as well as rules for the transactions between end users and sellers. Amazon also exercises influence through the selection criteria for the Buy Box. The court says that this may be done to serve users’ interests to quickly find a suitable offer does not speak against regulatory power.
Similarly, through intermediating logistics services for 3rd party sellers on the demand side (Fulfilment by Amazon or Seller Fulfilled Prime programs) and for logistics service providers on the supply side (Buy Shipping) Amazon sets the applicable participation conditions and is relevant for market access to logistics services.
Through its voice assistant Alexa, Amazon is relevant for access to the IoT: Alexa is among the top three voice assistants, provides access to >100k voice apps and is compatible with >140k smart home devices. Voice assistant providers set the framework regarding the compatibility of 3rd party devices with their services. The court says that voice assistant providers can thus influence the contacts between app developers, device suppliers and end users, respectively, as well as access to customer-related data. Amazon argued due to its limited device portfolio and Alexa’s prospects, Amazon does not control access to the IoT. Inter alia barriers erected by Apple and Google would speak against an Amazon gatekeeper position in this field. The FCJ rejects that line of reasoning, pointing out that Section 19a(1) ARC requires an overall comprehensive review across markets.
Dominance
The FCJ confirms that Amazon is dominant in the market for online marketplaces services for merchants in Germany, affirming the FCO’s market definition. The FCJ rejects that other distribution channels, i.e., stationary retail shops or own online shops, are substitutable form the merchants’ perspective, due to different geographic and customer footprint, availability of service for 3rd party sellers, visibility, investment, etc. Selling via online shops is not comparable to doing so via marketplaces, even when using marketing tools like comparison websites, paid search ads, ads on social media platforms or search engine optimization.
The court refers to the Commission’s Amazon Marketplace/Buy Box decision in 2022, defining the provision of online marketplace services to merchants as a separate product market, and notes that the new Commission notice on market definition would not lead to a different result. The court confirms its standing caselaw that the SNIPP test is only an auxiliary model that may provide indications for the market definition but is not decisive. The FCJ rules that if the concept of demand-side substitutability leads to the clear result that other products are not substitutable, like here, the SNIPP test cannot justify a different assessment.
Amazon’s (redacted) shares in the defined market are reportedly very high, with a significant gap to other online marketplaces. The court mentions positive indirect network effects and barriers to entry resulting from the large number of end customers and 3rd party sellers using Amazon Store, reinforced through Amazon Prime. That there might be multihoming (between different online marketplaces) does not speak against dominance, because 3rd party sellers cannot achieve the same or similar turnover on other marketplaces like on Amazon.
No misuse of discretion
The FCJ concludes that an overall assessment justifies the finding that Amazon is a gatekeeper and subject to Section 19a ARC – even if Amazon were only found to have very strong market power in Germany instead of dominance. The FCJ also refers to the fact that two of Amazon’s activities, Amazon Marketplace and Ads were designated as CPS under the DMA.
The court rejects the claim that the FCO misused its discretion when designating Amazon as a gatekeeper: the designation under Section 19a(1) ARC does not create any immediate obligations for Amazon. The FCO did equally not need to consider carving out certain business activities from the designation: the court clarifies that the provision clearly concerns the entire undertaking in the antitrust law sense. Given the rule relates to a position across markets, the designation cannot be limited to specific markets, and the FCO may need to intervene in various markets in the future. At the same time, the court confirms that the FCO cannot prohibit business activities in third countries (territoriality principle).
No significant, relevant factual changes after the FCO’s decision’s adoption
Amazon’s commitments in the Commission Art. 102 TFEU proceedings Amazon Marketplace and Amazon Buybox of December 2022 are not considered to affect Amazon’s gatekeeper position under Section 19a ARC. The court says they inter alia limit Amazon’s conduct regarding the use of 3rd party seller data and regarding the Featured Offer (Buybox). While this may become relevant in the context of an intervention into specific practices under Section 19a(2) ARC, it cannot eliminate Amazon’s paramount position for competition across markets. Statutory behavioral limits like those resulting from Art. 102 TFEU or the national equivalent cannot eliminate the dominant position, which is the very foundation for these obligations to apply. Similarly, the DMA obligations that now apply to Amazon’s CPS cannot eliminate its gatekeeper position under Section 19a(1) ARC either.
DMA does not eliminate applicability of Section 19a ARC
The FCJ finds that Section 19a ARC does not infringe Art. 1(5) DMA, i.e. that Member States must not impose additional obligations on designated DMA gatekeepers by way of laws or regulations for ensuring contestable and fair markets. Section 19a ARC falls under Art. 1(6) (2) lit. b DMA, i.e., it is a permissible national competition law provision imposing further obligations on a DMA gatekeeper.
The court rules that Section 19a ARC is not only formally competition law, but also in substance: its purpose is to protect competition against the special risk potential of large digital companies, as well as the competitive process against further concentration and expansion carried out through competition that is not based on the merits. Unlike under the DMA, the designation is not based on formal turnover or user number thresholds but requires a case-by-case analysis of the gatekeeper’s significance for competition. Importantly, the rule does not provide per se prohibitions, but any intervention depends on the FCO’s discretion, and gatekeeper can always objectively justify their conduct, even if it meets the specific abuse practices listed in Section 19a(2) ARC.
The court concedes that there are overlaps between Art. 5-7 DMA and Section 19a(2) ARC, but the latter also imposes further obligations on gatekeepers, e.g., self-preferencing is not limited to ranking, indexing and scrawling; pre-installation/default setting of its own services can be prohibited beyond the scope of Art. 6(3) DMA, and the possibility to prohibit any impediment of third party access to markets goes beyond the DMA. The DMA obligations only apply to designated CPS, whereas Section 19a ARC applies to the entire undertaking, i.e. the FCO can in principle prohibit practices under Section 19a(2) ARC in relation to any of the gatekeeper’s activities.
That Amazon has been designated as a DMA gatekeeper and is thus subject to DMA obligations would not eliminate its gatekeeper designation under Section 19a(1) ARC. The DMA explicitly allows for parallel application of national competition law. Accordingly, the claim that a DMA designation as such would eliminate the application of Section 19a ARC contradicts the legislative objective of Art. 1(6) lit. b DMA. The DMA obligations may play a role, however, in the context of an intervention against specific practices under Section 19a(2) ARC.
No request for preliminary ruling on DMA/Section 19a ARC
The FCJ rejects Amazon’s request for a preliminary ruling to Luxemburg, stating that in this case, the correct application of EU law is so obvious as to leave no scope for any reasonable doubt, i.e., this is one scenario under EU caselaw, in which there is no obligation to refer to the Court of Justice. The FCJ considers that Section 19a ARC is a national competition law provision within the meaning of Art. 1(6) (2) lit. b DMA is obvious, based on the latter’s wording, its objective as illustrated in the relevant recitals, and in particular regarding its legislative history: Art. 1(6) (2) lit. b DMA was introduced into the DMA with a specific view to Section 19a ARC.
Comments
This is a pretty forceful rejection of Amazon’s appeal – and as broad as the appeal’s scope. The court clarifies that Section 19a ARC is flexible and covers various forms of digital ecosystems, i.e., companies with different business models can qualify as gatekeepers – a point that is not surprising but important.
The court made it clear that incentive- and effects-based type of arguments, which are known from abuse of dominance cases, are not relevant in the gatekeeper designation process under Section 19a(1) ARC. It is sufficient to show that the company has the requisite capabilities – it is not required to show that it will also likely use those and that those would then have negative effects on competition. Apple voiced generally similar type of arguments in the FCO’s Apple designation process (see an English version of the decision here), and it has also appealed its designation. Based on the current ruling, Apple will likely not be successful with a similar approach.
Does this mean that the intervention under Section 19a(2) ARC will be the real playing field for effects-based arguments? To some extent maybe, but differently than under traditional abuse of dominance rules: several specific practice examples under Section 19a(2) ARC do not require the FCO to find any negative effects – the fact that the gatekeeper’s conduct meets the example is sufficient, as it is deemed to lead to negative effects and be abusive. There is a shift of the burden of proof: gatekeepers need to demonstrate that the conduct is not harmful or objectively justified. Once the FCO will intervene and issue a decision, it will be interesting to see how the FCJ will continue and develop the foundations of Section 19a ARC it laid out in this designation ruling.
On the conditions: the significant extent criterion, which theoretically opens a backdoor to escape the gatekeeper designation, will likely also be relevant in Apple’s appeal. Based on the FCJ’s findings here, it seems unlikely that Apple can escape a gatekeeper designation by saying that its main business is hardware sales.
The FCJ’s finding on dominance is interesting, because the FCO took a more cautious approach and left that question ultimately open (dominance or strong market power). The court takes a clear stance here (a bit like the unequivocal finding of dominance in the Facebook abuse of dominance ruling of June 2020). This part has a direct impact on Amazon’s conduct in Germany beyond Section 19a ARC: Amazon is now officially subject to the obligations under dominance rules (non-discrimination, exploitation, exclusion – or only competition on the merits), and third parties do not need to prove dominance any longer in litigation.
________________________
To make sure you do not miss out on regular updates from the Kluwer Competition Law Blog, please subscribe here.
