This paper – which can be found here – reviews the European Commission’s decision in the Google case, and the remedy that the Commission imposed in that decision.
It argues that this decision follows settled law regarding anti-competitive extensions of dominance from a primary market to a distinct, but related, secondary market. It also seeks to refute the argument that the decision created a novel rule that a dominant company may not favour its own services – instead, it is argued that this requirement is merely the remedy that the Commission imposed to bring Google’s infringement to an end.
The paper is structured as follows:
- A first section provides an overview of the decision and some critical reactions to it.
The Commission fined Google for having abused its market dominance as a search engine by promoting its comparison shopping service, Google Shopping, and demoting rival services. Describing the abuse, the EC explained that it: “objects to the fact that Google has leveraged its market dominance in general internet search into a separate market, comparison shopping.” In addition, the decision imposed a remedy that required Google to: “respect the simple principle of equal treatment in its search results for its own comparison shopping product and rival comparison shopping products”.
Such an approach has been criticised on a number of levels. The main arguments are that: (i) this was inconsistent with the current understanding that even dominant companies are not per se required to treat third parties equally to its own subsidiaries; (ii) a duty to treat a company’s own services like external services may only be envisaged when dealing with access to ‘essential facilities’, which does not apply to the facts of this case.
- Section II makes the simple argument that the list of abusive practices under Art 102 TFEU is not exhaustive. In fact, the lack of an exhaustive list of examples of abuse is an essential feature of Article 102 TFEU, as it ensures crucial flexibility when dealing with new types of anti-competitive conduct.
- Section III argues that the non-discrimination remedy imposed on Google does not imply a new general duty under Article 102 TFEU to treat third-party businesses and one’s own subsidiaries similarly. Instead, the abuse identified in this was the leveraging of monopoly power. The non-discrimination obligation imposed by the Commission in its decision was only an effective instrument, in this particular case, to bring the identified monopoly leveraging abuse to an end.
- Section IV argues that the issue with Google’s conduct was not the promotion of its own services. Instead, the problem was that Google changed its business conduct on the primary market for general search services in order to enter a secondary market for comparison shopping service in a way that infringes the conditions set out in the case law for finding an abuse of a dominant position. This section merits a bit more attention, and I will discuss it in a bit more detail.
The author argues that leveraging market power is prohibited under both US and EU law. It is said that there is a general principle under EU competition law that market power on a primary market may not be extended to a separate, but related market by means that deviate from competition on the merits on the primary market. Two main reasons for condemning such leveraging strategies are: (i) to spare competitors on the secondary market, M2, from having to vertically integrate backwards into the primary market, M1, in order to compete; (ii) to make it more difficult for the dominant company on M1 to gain unjustified advantages on M2 by simply using its dominance on M1 as a lever to foreclose competition on M2.
Classic strategies of monopoly expansion, which have been condemned on both sides of the Atlantic, are: (i) bundling and tying, (ii) margin squeeze, and (iii) particular types of refusal to deal. Such classic leveraging cases are all characterised by the same economic situation: a company is: (a) dominant in an existing primary market (M1); (b) it tries to maximise its profits by vertically or horizontally (as in the case of bundling/tying) integrating its business into a distinct secondary market (M2); and (iii) it changes its business strategy in M1 in a way that established competitors on M2 are hindered. Hindrance means that the change in the dominant player’s business strategy in M1 has an exclusionary effect in M2, and that there is no counterbalancing economic or other objective justification for the conduct.
It is explained that Google tried to maximise its profits by vertically integrating into the neighbouring markets for comparison shopping services (and other specialised search markets), which were separate and ‘secondary’ or ‘downstream’ to the general search market. To do this, Google (as in classic leveraging cases) changed its business strategy on the primary market for general search services in a way that foreclosed competitors on the secondary market for comparison shopping services.
In particular, this behaviour was thought not to make sense from a commercial perspective. Google argued that the change from relevance-based to self-promoting search results made sense because the rich format of the self-promoting search results boxes (with product images and price information) enhanced users’ general search experience and thus improved the product – but this does not explain why Google’s new design exclusively favoured its own downstream services even if there were more relevant results from competing services. The shift from equal-treatment to self-promotion can be explained only by Google’s objective to gain market share on the downstream markets (i.e. for comparison shopping) in order to increase total profits. Furthermore, Google was only in the position to accept the risk of disappointing its users on the general search market with less relevant results due to its near monopoly in this market throughout Europe. The lack of competition rendered large-scale shifts of users to other search engines unlikely.
Importantly, existing competitors on the secondary market could not, on a timely basis, effectively counter the sudden changes on the primary market. Comparison shopping services rely to a large extent on traffic to be competitive’. If competitors wished to secure the same traffic advantages that Google secured for its own service, they would have had to integrate into the upstream market for general search services and, more importantly, build up similarly large user bases. Due to strong network effects in the market for general search services, following the market’s ‘tipping’ towards Google between 2001 and 2004, this appeared economically unviable for any company.
- Section V explains why the case is not concerned with refusal to supply, and that in any event the essential facility criteria introduced in the case law (by Magill, Bronner and IMS Health) would not be in point.
In practice, this section builds on the previous one – and tries to explain why leveraging can amount to a competition infringement. It is argued that, for a finding of abuse under EU law, it has been considered necessary, but also sufficient, that the dominant company’s conduct has an anti-competitive effect, which may potentially exclude competitors who are at least as efficient as the dominant undertaking on the secondary market. However, it is not required that the leveraging conduct eliminates all competition on that market, nor that it must be generally impossible for competitors on the secondary market to create alternatives for the upstream service provided by the dominant company. This is supported by a discussion of case law on bundling/tying, margin squeeze and, in more detail, refusal to supply.
- Section VI then explains how the imposed remedy of equal treatment effectively brings to an end the identified leveraging by self-promotion of Google of its Shopping service. The remedy seeks to eliminate or neutralise the anti-competitive effects of the infringements on the market. In monopoly leveraging cases, the typical approach to effectively remedy an identified abuse is a cease and desist order that obliges the dominant company to stop pursuing the business conduct on the primary market that was used to leverage market power. In the light of this, imposing the equal treatment remedy on Google is a natural fix to the identified leveraging abuse.
Comment: I have a number of problems with the arguments submitted in this paper (which are similar to some doubts I have about the reasoning in the decision). A fundamental disagreement is that I do not think there is a general prohibition of leveraging of market power in either EU or US law. In my view, the paper actually seems to reflect this view by actively grappling with the effect that a leveraging practice must have before it is deemed anticompetitive.
This is important because there are indeed some anticompetitive practices – such as tying, bundling, margin squeeze and refusal to supply, which are discussed in the paper – which involve leveraging. However, a relationship between a conduct in the dominant market and an (anticompetitive) effect in the secondary market must be demonstrated. One such effect may well be to ‘foreclose’ competitors in the secondary market (i.e. an exclusionary effect). I think that a finding of foreclosure effect may be implicit in the Commission decision, but when I read it I did not find any section that explained what the anticompetitive effect was, or how it related to a conduct in the primary market, in terms of a fully articulated theory of harm (having said that, I may have missed – I read the decision only once, and quickly).
A second issue I have with the paper is that I’m not really convinced by its attempt to defend the decision under a “traditional” leveraging theory. In particular, it is still unclear what anticompetitive effect the author is thinking about when he mentions ‘foreclosure’ – after all, the supposedly foreclosed competitors are still in the market. This, I think, is quite an important point: if leveraging in itself is not an abuse, but only if it has some anticompetitive effect on the secondary market, then the question is: what should that effect be? Should it be complete foreclosure? Or merely substantive foreclosure? Or should we focus on whether the conduct increases rivals costs to such an extent that they cease to compete effectively?
I have no ready answer to this, by the way. But I think that the answer provided in the article (i.e. “Regarding the effects of the conduct on the comparison shopping market, the EC found that rivals have suffered ‘very substantial losses of traffic on a lasting basis’, whilst traffic to Google’s own service ‘increased significantly’.98 For example, traffic to certain rival sites dropped ─ up to 85% in the United Kingdom and up to 92% in Germany, whilst Google’s comparison shopping increased its traffic 45-fold in the United Kingdom and 35-fold in Germany. Considering the relevance of traffic for competition on the comparison shopping market, the conduct was thus capable of excluding at least equally effective competitors on this market) does not really answer this question.
I think the CJEU’s price discrimination decision in MEO provides important guidance in this respect. I actually think this has the potential to become a very important decision, which will be relied in defending the Google decision before the General Court.
Before we do so, however, I must be clear: I am not disputing the correctness of the Commission decision. I am merely pointing out that the decision is not very clear about what its theory of harm is (as opposed to the practical effects of Google’s conduct, which are discussed extensively), and that I think that this paper does not really add much to our understanding of that matter.