One Conduct, Two Different Outcomes on Either Side of the Atlantic
The Google Shopping ruling further confirms the divergent approach taken by the U.S. and EU competition authorities to Google’s conduct.
In the United States, the Federal Trade Commission (FTC) began investigating Google in June 2011. A further request by the Committee on the Judiciary at the U.S. Senate dated December 19, 2011 noted the Senators’ concerns that “Google’s search engine is biased in favor of its own secondary products and services, undermining free and fair competition among e-commerce websites.”
In January 2013, the FTC closed its investigation without pursuing charges against Google. The FTC noted in a succinct statement the “unanimous decision to close the portion of the investigation relating to allegations that Google unfairly preferences its own content on the Google search results page and selectively demotes its competitors’ content from those results.” In its reasoning, the FTC indicated Google’s horizontal search engine, which searches the entire internet, may be considered to operate in a different market from vertical search engines, which focus on specific categories of content. Vertical search engines had complained to the FTC about the introduction of the “Universal Search” box, which displays Google vertical search results in a prominent position and consequently demotes the general search results that list competing vertical websites. They also argued that the general search algorithm was manipulated to demote competing vertical websites.
In its assessment, the FTC considered that the introduction of the Universal Search box in general improved the quality of Google’s search results, and that any negative impact on competitors was “incidental” to that purpose and the byproduct of competition on the merits. More interestingly, the FTC illustrated this quality-enhancing objective by stating that Google would “demote its own content to a less prominent location when a higher ranking adversely affected user experience.”
While the Commissioners were unanimous, subsequent disclosures of part of the case file in 2015 and 2021 revealed that FTC staff working on the case had greater concerns about the anticompetitive harm of Google’s conduct than reflected in the final vote. In particular, the staff indicated that demotion algorithms would not apply to Google vertical sites, in contrast to what the Commissioners finally held.
In the European Union, the EC launched a parallel investigation against Google in November 2010. After several months negotiating commitments with Google, the EC finally decided to continue the infringement proceedings, which led to the 2017 decision imposing on Google a record fine of EUR 2.42 billion for abusing its dominant position as a search engine by favouring its own comparison shopping service over competing ones. The Decision further ordered Google to treat rival comparison shopping services and its own service equally, but left it to Google to decide how it would do so.
Several reasons may explain the EC’s and FTC’s divergent positions. First, the EC dedicated almost seven years to investigate Google’s practices, conducted extensive econometric analyses, and fine-tuned its theory of harm regarding self-favouring. While the FTC reviewed nine million pages of documents and held hearings, and its staff economists conducted empirical analyses of the effects of Google’s conduct, the difference in length between the two procedures remains remarkable. Second, in the context of Article 102 of the TFEU, the EC—as confirmed by the GC Judgment—is not required to prove that the conduct had actual foreclosing effects, it being sufficient for the EC to demonstrate that the conduct is capable of having or likely to have such an effect. This is in contrast to the standard of proof required by U.S. case law under Section 2 of the Sherman Act, which requires the plaintiff to demonstrate an anticompetitive effect, i.e., that the monopolist’s conduct actually harmed competition. Even though the EC conducted more extensive analyses than the FTC performed—with the EC finding evidence of traffic diversion to Google’s own shopping services—the prospect of each of the agencies succeeding in court was certainly influenced by the standard of proof in each jurisdiction. Third, the FTC arguably gave more weight to the innovation brought about by Google’s conduct; whereas the EC focused on the exclusionary effect of the differential treatment of Google’s algorithms on competing shopping services, disregarding innovation as a sufficient objective justification. Some commentators argue that the balance struck by each agency between innovation and the anticompetitive effect lies in the different welfare standards traditionally applied in the European Union and the United States.
Competition on the Merits
In its appeal before the GC, Google did not dispute the fact that it holds a dominant position in the 13 national markets for general search services in which the EC found that Google had abused such position. Therefore, the GC Judgment focused on whether that position was abused in a manner prohibited by Article 102 of the TFEU. Google’s first main argument was that the conduct at issue represents a qualitative improvement that constitutes competition on the merits and therefore cannot be treated as abusive.
The GC agreed that not every exclusionary effect is necessarily detrimental to competition. Competition on the merits may lead to the departure from the market or the marginalization of competitors that are less attractive to consumers from the point of view of, among other things, price, choice, quality, or innovation. Leveraging a dominant position into adjacent markets cannot in itself constitute proof of conduct departing from competition on the merits. In other words, “leveraging” a dominant market position is not per se abusive; it is only if competition has actually been weakened.
The conduct specifically identified by the EC as the source of Google’s abuse is, in essence, the fact that Google displayed its comparison shopping service on its general results pages in a prominent and eye-catching manner in dedicated “boxes,” without that comparison service being subject to the adjustment algorithms used for general searches. At the same time, competing comparison shopping services could appear on those pages only as general search results (blue links) that tended to be given a lower ranking than those of Google as a result of the application of adjustment algorithms.
The GC accepted that the EC had established three particular circumstances that demonstrated how this self-preferencing conduct was liable to lead to a weakening of competition in the market:
- The traffic generated by Google’s general search engine for comparison shopping services produces network effects. In other words, traffic may generate a virtuous circle, improving the relevance of results and thus attracting more users and convincing more partners to advertise their products and services.
- Users typically pay attention to the first three to five results of a search. This makes prominent positions on the first page essential for the success of a comparison shopping service.
- Competing comparison shopping services lack credible alternative sources of traffic. Traffic diverted from Google’s general results pages accounts for a large proportion of traffic to competing comparison shopping services and cannot be replaced effectively by other sources, such as other search engines, social networks, or referrals from affiliate websites.
In essence, the GC concluded that the treatment and prominent position of Google Shopping services, which are immune to demotion by adjustment algorithms otherwise applicable to competing shopping services, aggravated by the importance of Google’s general search for the success of a shopping service, qualifies as unlawful leveraging departing from competition on the merits. However, the GC goes further and gives two additional reasons that deserve closer analysis.
The first element concerns the application of the principle of equal treatment. The GC Judgment explicitly invokes the principle of equal treatment, as a general principle of EU law, which means that abuse may take the form of an unjustified difference in treatment. Some commentators point out that the case law cited in this respect refers to the behavior of public authorities and that the GC would be extending this principle to privately owned entities. However, Article 102 (c) of the TFEU declares that “applying dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage” is a type of abuse. Whether such practice stems from public undertakings exercising special rights or private companies leveraging a dominant position seems not to play a role in Article 102 (c) TFEU.
In addition, the GC relies on the statutory equal treatment obligation of internet services providers, which according to the GC cannot be disregarded when analyzing practices in the downstream market, given Google’s “ultra-dominant” position. While this argument is merely raised for completeness, this enlargement of the scope of application of existing EU law may blur the lines between the legislator and the judiciary.
The second element of the GC’s reasoning about competition on the merits that deserves closer attention is that Google’s behavior involves “a certain form of abnormality.” This abnormality results from the fact that Google’s general search engine initially had a universal vocation and was designed to index results containing any possible content. Against this background, a subsequent change of business model to promote Google’s own services over the rest would not necessarily be rational according to the GC, unless Google was so dominant, and barriers to entry so high, that adding such a limitation of internet users’ choice would not cause new competitors to enter the market.
It remains unclear how this abnormality presumption—which would raise suspicion of changes in the behavior of a dominant operator—will be construed in the future and reconciled with the natural need for any undertaking to intelligently adapt its initial business model considering the ongoing market developments. Normality is a subjective concept that provides little guidance to distinguish competition on the merits from unlawful abuses. As noted by Cristina Caffarra, “[i]ncentives remain the natural way of thinking about how to interpret conduct, and the exercise of asking whether there were profit incentives to adopt a specific conduct is much more fruitful than trying to benchmark it against some hypothetical, subjective ‘normal’.”
Legal Test Applicable to Self-Preferencing
Google argued that the Decision in fact imposed on it a duty to supply, consisting of providing competing comparison shopping services access to its improved services without satisfying the conditions identified in the case law and, in particular, those applicable to infrastructure qualifying as an essential facility. The essential facility doctrine set out in cases such as Bronner notes that a refusal to supply may be considered abusive solely with regard to infrastructure that is indispensable for carrying out a business in the market where there is no actual or potential substitute, so that refusing access may lead to the elimination of all competition. This qualified test is aimed at ensuring that a refusal to supply is abusive only in very limited cases, as otherwise a broader application of a duty to supply would ultimately impede competition by reducing a dominant undertaking’s incentive to build such infrastructure in the first place.
As a preliminary point, the GC accepts that Google’s general results page has “characteristics akin to those of an essential facility” for the three reasons stated in the last section. Yet it does not accept that the Decision imposed a duty to supply, which would require compliance with the essential facility doctrine. It recalls that most exclusionary abuses are liable to constitute implicit refusals to supply, since they tend to make market access more difficult. The GC notes that Google did not technically refuse to supply, but rather did supply but discriminated against competitors in favor of its own downstream service, and therefore the essential facility doctrine would not be applicable.
Finally, the GC clarified that even if the remedy for Google’s conduct can be equivalent to the one applied in essential facilities cases, this does not mean that the abuse has to be of the same nature. There can be no automatic link between the criteria for the legal classification of the abuse and the corrective measures enabling it to be remedied.
In this light, the GC accepts that self-preferencing amounts to an independent form of abuse of a dominant position which does not explicitly have to follow the Bronner requirements. However, in practice the GC applies those requirements to a large extent and shows how Google’s general search engine resulted in a “quasi-essential facility,” in the sense that its traffic is indispensable for competing shopping services and that access in discriminatory terms could lead to the elimination of all competition. It can be contended that the GC still largely considered indispensability and potential elimination of all competition—traditional Bronner criteria—to conclude that discrimination by Google infringed Article 102 TFEU.
A Counterfactual Analysis Is Not Required
to Establish Effects
Google argued that the EC had not established any causal link between the practices at issue and the—undisputed—decrease in generic search traffic to almost all comparison shopping services on a lasting basis. Specifically, Google contended that the EC should have conducted a counterfactual analysis to show how the market would have developed had the investigated practices not been put in place.
The GC concluded that in Article 102 TFEU cases there is no need for the EC to systematically run a counterfactual analysis to establish actual exclusionary effects, not even in response to a counterfactual analysis put forward by the undertaking being investigated. It explains that this would oblige the EC to demonstrate that the conduct at issue had actual effects, which is not required in the case of an abuse of a dominant position, where it is sufficient to establish that there are potential effects.
Nevertheless, this plea by Google was actually upheld by the GC as regards the market for general search services. The GC noted that the EC put forward too imprecise considerations to show that there were (even potential) foreclosure effects in this market by simply stating that Google was protecting the revenue from its general results pages that was generated by its own comparison shopping service—revenue which, in turn, financed its general search service. This finding led to the partial annulment of the Decision, but the GC did not reduce the fine imposed as the value of sales that the EC took into consideration for its calculation excluded revenues related to the markets for general search services and only included revenues accrued from specialized search services.
However, the GC accepted that the EC sufficiently established at least potentially exclusionary effects in the market for comparison shopping services, for which the EC had conducted a detailed analysis of the traffic diversion from rival comparison services to Google’s services as a result of Google’s conduct.
While a counterfactual analysis is not required within the context of Article 102 of the TFEU, this does not release the EC from its duty to adduce a sound theory of harm and produce sufficient evidence demonstrating at least a potential anticompetitive effect.
It is also interesting that the GC ruled out the need to establish that the exclusionary conduct affected as-efficient competitors as Google. The as-efficient-competitor test must be applied in pricing practices cases, but according to the GC Judgment it does not apply in cases in which the competition practice did not relate to pricing.
Google’s Innovation as an Objective Justification
Under EU law, it is possible for a dominant undertaking to justify behavior that would otherwise be caught by Article 102 of the TFEU by establishing either that its conduct is objectively necessary from a technical or commercial point of view or that the exclusionary effect may be counterbalanced by efficiencies that are also passed on to consumers.
The GC does not deny that Google’s practices may have improved some users’ experience, but rather focuses its criticism on Google failing to demonstrate why it could not use the same underlying processes to decide how to display the results of both its own and competing comparison shopping services. According to the EC, equal treatment was proved possible during the commitments discussion in the administrative proceeding; something Google disputes. In any event, as the onus for establishing an objective justification lies on the dominant undertaking, the GC rejected this plea.
Conclusions and Implications for the Future
In our view, three main conclusions may be drawn from the GC Judgment on assessing dominance in digital markets.
First, the GC Judgment strengthens the enforcement of Article 102 of the TFEU in digital markets.
The GC Judgment encourages authorities not to solely pursue well-established categories of abuse, but also to prosecute and develop novel theories of harm that do not squarely fit in existing precedents. It expressly enshrines the principle of equal treatment and opens the door to require non-discriminatory treatment from dominant undertakings without strictly complying with the rigid requirements laid down in the essential facilities doctrine.
The GC also explicitly releases the EC from the obligation to run a counterfactual analysis to establish an infringement of Article 102 of the TFEU. This conclusion stems from existing case law, which states that proof of actual effects is not required to establish exclusionary abuse—potential effects suffice. This departure from the “more economic approach” embraced by the EU during the last two decades could have ambiguous effects. On the one hand, it favors rapid intervention in digital markets without the need to wait until it is too late when damage can be proven but potentially not undone. After seven years of investigation and a massive array of evidence, one may wonder whether the prevailing competitive situation before the abusive conduct began can be restored. This belated response via ex post antitrust investigation seems aligned with the goal that led to the proposal for the Digital Markets Act, which will ex ante regulate gatekeepers’ conduct. On the other hand, a softened standard of proof may give rise to authorities challenging future abusive conduct on the basis of feebler theories of harm and light-effect analyses.
However, it must be borne in mind that the GC Judgment is not yet final, and some of its legal reasoning may warrant closer inspection during the appeal. “Abnormality” as a criterion to identify competition on the merits seems a subjective notion unable to render useful results. Moreover, the GC’s suspicions about Google’s apparently irrational conduct of changing its initial business plan after years fail to take into account that undertakings—even dominant ones—have the right to intelligently adapt to changing market conditions. A dominant undertaking that switches to self-preferencing may be responding to economic incentives other than excluding competition, unless it holds a quasi-essential facility or gatekeeper position. In any event, abnormality does not seem to play a decisive role in the GC Judgment reasoning. It is rather an additional argument to support the GC’s finding that the different treatment given to competitors is aggravated by the importance of Google’s general search for the success of a shopping service, which qualifies as unlawful leveraging departing from competition on the merits.
Second, the final outcome of the Google Shopping saga will shape the implementation of the future Digital Markets Act. Pursuant to its draft Article 6(d), gatekeepers shall “refrain from treating more favourably in ranking services and products offered by the gatekeeper itself or by any third party belonging to the same undertaking compared to similar services or products of third party and apply fair and non-discriminatory conditions to such ranking.” If this regulation is finally enacted in its current form, the EC will be given the power to specify how this obligation must be complied with by a given gatekeeper. The GC Judgment will set a seminal precedent in assessing future decisions adopted by the EC under the Digital Markets Act.
The Digital Markets Act could also bring about changes in terms of remedies for abusive self-preferencing. As some commentators point out, the Decision merely ordered Google to bring the infringement to an end. As of now, this seems the most frustrating part of the EC process: it developed a sound theory of harm to establish the abuse, but did not specify how to stop it. However, Article 7(2) of the draft Digital Markets Act will enable the EC to adopt a decision further specifying how the principle of equal treatment should be put into effect by gatekeepers and underpin effective remediation of the conduct.
Finally, the GC Judgment includes helpful passages for follow-up damages claims. If confirmed by the Court of Justice, the GC made clear references to the existence of a causal link between self-preferencing and the loss of traffic suffered by competing comparison shopping websites, even naming some of them. It will be for national courts to assess damages claims and potential effects, which are sufficient for public enforcement but may not necessarily be so for private enforcement. However, these excerpts may prove useful ammunition for potential follow-up litigation.
