Abuse of dominance is arguably the most complex area in competition law. It presupposes distinctions between anticompetitive exclusionary conduct and competition on the merits, and between legitimate and ‘exploitative’ terms or prices. Such distinctions are rarely clear in practice. The task is more challenging still where antitrust authorities apply these principles to new markets, for example, in the digital sector, where prior case law and principles developed in traditional industries may be ill suited to capture rapid changes in competitive structures and consumer demand. Small wonder that there is often divergence in the enforcement priorities and practice of competition authorities around the world.
While the legal distinctions in this area of competition law are nuanced, and investigations are fact-specific and resource-intensive, the consequences of infringing rules on abuse of dominance may be severe. Several major antitrust jurisdictions have imposed fines of several hundred million dollars in single-firm conduct cases; the European Commission recently imposed a record fine of around US$5 billion on Google for abuse of dominance in connection with its Android mobile phone operating system.
Yet despite these significant costs, in most jurisdictions and scenarios, companies cannot submit proposed conduct to competition authorities agencies for ex ante review. This guide aims to assist companies and their advisers with the complex exercise of self-assessment. In doing so, it draws on the insights of specialist counsel from a wide range of jurisdictions. They include long-established antitrust regimes, such as the US, Canada, the EU and Japan. The guide also covers the fast-evolving practice and principles in ambitious, developing jurisdictions such as China and India, and nascent antitrust regimes such as Hong Kong, offering a detailed summary of applicable rules and an overview of the enforcement climate.
A high-level summary cannot do justice to the careful contributions of the various authors of this guide. In this introduction, though, we draw attention to a number of important recent trends.
The as efficient competitor test revisited
Ever since the European Commission began its movement towards a ‘more economic approach’ to abuse of dominance issues with its 2005 white paper, lawyers, economists and agencies have debated the need to examine whether conduct would exclude as efficient rivals. This debate has intensified since the Court of Justice judgment in Intel in 2017, which made clear that the competitive assessment focuses on the ‘exclusionary effect on competitors considered to be as efficient as [the dominant firm] is’. Thus, if a firm submitted evidence that its conduct was incapable of foreclosing competition, the European Commission would be required to analyse (among other factors) the existence of any strategy to exclude equally efficient rivals. Critically, the assessment should examine the ‘intrinsic capacity of that practice to foreclose competitors which are at least as efficient as the dominant undertaking’.
Commissioner Vestager stated in a speech in 2018 that the Intel judgment confirmed the Commission ‘can presume that this sort of rebate, from a dominant company, is against the competition rules’ and ‘in practical terms, our main conclusion is that you won’t see fundamental change’. In December 2017, the Italian Competition Authority issued a decision fining Unilever €60 million in respect of allegedly anticompetitive rebates concerning single-wrap ice creams. The Authority noted that its judgment followed the Intel ruling, but did not appear to consider itself bound to apply the ‘as efficient competitor’ test. Rather, it analysed the factors listed in the Intel judgment (eg, market coverage of the practice) following Unilever’s submission of an as efficient competitor analysis. (Interestingly, the UK Competition and Markets Authority carried out its own effects analysis in the UK and found that the conduct was not abusive.)
Other commentators argue that Intel creates a legal requirement to examine the ‘as efficient competitor’ test, consistent with the Commission’s own guidance paper on abuse of dominance. Advocate General Wathelet’s Opinion in Orange Polska in February 2018 explained that the analytical framework in Intel ‘is by no means a purely procedural requirement’. Rather, he emphasised the need to show a capability or likelihood that the rebate would exclude equally efficient rivals and that article 102 of the TFEU does not ‘seek to ensure that competitors less efficient than the undertaking with the dominant position should remain on the market’.
The importance of the ‘as efficient competitor’ criterion has been confirmed in other categories of abuse too, including the 2018 Court of Justice judgment in MEO relating to discrimination and the 2018 General Court judgment in Slovak Telekom relating to an alleged margin squeeze. The precise requirements of the as efficient competitor criterion will likely continue to feature in abuse of dominance appeals, such as the Qualcomm and Google Android cases. It will also play an important role in the UK Competition Appeal Tribunal’s review of a 2018 finding by Ofcom that Royal Mail had excluded downstream rivals through proposed price changes.
Interaction between agencies: commonalities and divergence
It has long been a concern that companies face divergent legal assessments of their conduct in different jurisdictions, contrasting the traditionally more interventionist Europe with a US enforcement climate that is concerned with avoiding errors of over-enforcement (known as ‘Type I’ errors), as well as less predictable emerging antitrust jurisdictions in Asia. A prominent example is the Google Shopping case, which in 2017 led to an infringement decision and a €2.3 billion fine in Europe, whereas the case was closed without a finding of infringement in the US (as well as in Canada and Taiwan, and in similar cases in Germany and the United Kingdom).
A further example of potential divergence concerns the Court of Justice judgment in Huawei v ZTE, which set out the conditions under which standard essential patent owners could seek injunctions against implementers without committing an abuse of dominance. On one view, the prescriptive procedure laid down by the Court of Justice for negotiating SEP licences created a possibility for a harmonised framework for assessing conduct by SEP holders. In the UK case of Unwired Planet v Huawei, though, Mr Justice Birss stated: ‘I am not persuaded that the CJEU in Huawei v ZTE sought to set out a series of rigid predefined rules, compliance with which is never abusive whereas deviation from which is always abusive, all regardless of the circumstances. Abuse of dominance is a serious matter and the court will have had well in mind that circumstances can vary.’ This was confirmed by the Court of Appeal in 2018, which construed the Huawei v ZTE criteria as a ‘safe harbour’ rather than a set of mandatory conditions to avoid committing an abuse. It moreover accepted that its approach to assessing whether a proposed royalty rate was ‘discriminatory’ differed from the approach taken by courts in China and elsewhere.
In other cases agencies have reached consistent conclusions and contributed to each other’s enforcement practice. For example, the European Commission’s 2017 decision to fine Qualcomm almost €1 billion followed investigations into Qualcomm’s patent licensing practices in other jurisdictions. In 2015, China’s NDRC imposed a fine on Qualcomm of US$975 million in 2015 for failure to license its standard essential patents on fair, reasonable and non-discriminatory (FRAND) terms. Subsequently, the Korean Fair Trade Commission fined Qualcomm US$854 million for unfair patent licensing practices (Qualcomm’s appeal is ongoing). Taiwan’s TFTC found Qualcomm had committed an abuse by precluding competition through exclusivity agreements and related practices. And the US FTC filed a complaint against Qualcomm, alleging that it used its monopoly position in supplying baseband chips for mobile phones to impose anticompetitive licensing terms on SEPs. In particular, the FTC alleges that Qualcomm required customers to pay elevated royalties on products that use baseband chips made by rivals, thereby excluding competitors. The FTC won a partial summary judgment in November 2018, with the case proceeding to trial in 2019. It bears mention that the Department of Justice has adopted a more pro-licensor policy. Assistant Attorney General Makan Delrahim referred in a September 2018 speech to the Department’s ‘New Madison’ approach, which contends that ‘antitrust law should not be used as a tool to police FRAND commitments’.
More broadly, antitrust authorities and courts are increasingly confident of their ability to assess abuse of dominance issues in respect of conduct by holders of standard essential patents. This has led to disputes between the same parties generating parallel claims and actions before courts and administrative agencies of different countries, creating scope for jurisdictional disputes, anti-suit injunctions and global rate setting by national courts or agencies. By way of example, a patent dispute between Samsung and Huawei resulted in a judgment by the Shenzhen Intermediate People’s Court in January 2018 in favour of Huawei, ordering Samsung not to infringe Huawei’s SEPs through the sale of infringing devices. In April 2018, though, the District Court of Northern California granted Samsung an anti-suit injunction preventing Huawei from enforcing the Chinese court injunction until US litigation concerning the dispute had been resolved.
Increased scrutiny of digital platforms
Online platforms and services have grown at an extraordinary rate, leading to unparalleled choice, lower prices and disruption of traditional business models. In 2017, ecommerce sales in the United States reached approximately US$453.5 billion, growing 16 per cent compared to the previous year and accounting for 11 per cent of total retail sales (Department of Commerce, Quarterly Retail E-Commerce Sales, Fourth Quarter 2017). Streaming of music on services such as Spotify and Apple Music was projected to surpass physical music sales on CDs and other media in 2017 (Financial Times, June 2017). And photo apps continue to grow at unprecedented rates – Instagram was reported to have approximately 800 million active monthly users as of September 2017.
Despite the increased choice and availability of cheaper (or even free) online services, competition authorities have raised concerns about the emergence of purportedly ‘dominant’ online platforms and have scrutinised practices in online markets. The European Commission carried out its e-commerce sector inquiry and in 2017 Mexico’s COFECE opened an investigation into tying and bundling in e-commerce. In 2018 the UK government commissioned former Obama advisor Professor Jason Furman to carry out a review into digital platforms. In December 2018, Australia’s ACCC published a preliminary report in its Digital Platforms Inquiry 2018, addressing questions like ‘Do digital platforms have market power?’. And in January 2019, China’s new e-commerce law came into effect, governing issues such as the reasonableness of operators’ terms, taxation, and liability for the sale of counterfeit goods.
Concern about the power of digital platforms has led not only to reports and legislation, but also antitrust cases. In 2018, the European Commission opened an investigation into Amazon’s ‘dual role’ as a marketplace operator and as a merchant selling products on its own platform, in competition with third-party merchants. Specifically, the Commission is carrying out a preliminary review of whether Amazon collects and uses data from third-party merchants with a view to introducing its own ‘copycat’ products. These concerns about the dual role of marketplace operators were first raised in the e-commerce sector inquiry. Separately, the German Bundeskartellamt is investigating the terms and conditions that Amazon imposes on merchants – positioned as complementary to the Commission’s investigation, rather than overlapping with it, presumably to avoid the German authority losing jurisdiction if the Commission opens a formal probe. Other technology platforms have also faced (or are facing) antitrust probes, including Apple, Facebook, Google and Microsoft.
Data aggregation and data privacy
Competition authorities have expressed concerns related to the rise of ‘big data’ – broadly defined as the collection and processing of large, accurate datasets at high speed, thereby enabling firms to enhance their services relative to rivals. The European Commission and CMA are building specialist teams to review and work with big data in antitrust assessments. Reviews into the possible use and abuse of big data have been carried out (or are ongoing) by competition authorities in Canada, France, Germany, Italy and elsewhere. Two issues emerge in particular: (i) whether data aggregation gives rise to market power; and (ii) whether it raises privacy concerns that antitrust ought to address.
On one side of the ‘market power’ debate, certain competition authorities have considered that the amassing of large datasets could present a barrier to entry by smaller new entrants that lack the same scale of data. The Commission’s e-commerce sector inquiry considered a possible concern whereby an online marketplace owner might require third parties using the marketplace to provide it with their sales data, which could be used to strengthen the owner’s competing downstream service. And a joint paper by the French and German competition authorities considered possible abuses comprising refusal to supply access to ‘essential’ data and providing data access on a discriminatory basis, among others.
On the other side of the debate, commentators have pointed out that big data have facilitated high-quality services, often available for free, by allowing firms to monetise through targeted advertising rather than charging subscription fees. Moreover, while an initial pool of data helps develop an accurate algorithm and allows it to be improved and tested, data have a diminishing marginal return: adding data only helps up to a certain point, beyond which improvements to the underlying algorithm become more important. And in a series of merger cases, the Commission has dismissed data-related concerns as the data in question were non-exclusive, replicable by rivals and available from third-party data providers.
As regards data privacy, the German competition authority is investigating Facebook’s policy of conditioning access to its ‘dominant’ social networking site on users giving Facebook access to ‘data generated by using third-party websites’ that use Facebook APIs in order to ‘merge it with the user’s Facebook account’. The authority announced in December 2017 that it had sent its preliminary assessment to Facebook, objecting that:
- users would not expect data generated on third-party sites to be provided to Facebook itself;
- users do not effectively consent to Facebook’s data collection as it is a dominant social network;
- the data are personalised and valuable; and
- it is inappropriate for users to have to consent to this level of data collection, which may breach European data collection rules.
In January 2019, Andreas Mundt, President of the German competition authority, was quoted as saying that the probe would soon conclude. Close attention will be paid to the decision once it is published, in particular as regards the following questions:
- Why is antitrust intervention necessary, given the existence of data protection rules (which have themselves been strengthened by the GDPR)?
- How is Facebook’s conduct related to its dominance, given that online services without market power seem equally capable of imposing extensive data collection conditions?
- At what point does data collection go from being ‘extensive’ to ‘abusive’ (and are their parallels with ‘excessive pricing’ cases)?
- How are the pro-competitive benefits of data collection taken into account?
- Does intervention risk unintended consequences (eg, Facebook withdrawing its APIs from third-party sites)?
Continued focus on pharmaceuticals
The pharmaceuticals sector has long drawn antitrust scrutiny. The leading AstraZeneca case confirmed that ‘the illegality of abusive conduct under article 102 is unrelated to its compliance or non-compliance with other legal rules and, in the majority of cases, abuses of dominant positions consist of behaviour that is otherwise lawful under branches of law other than competition law’. Therefore, conduct of the type alleged in AstraZeneca – submitting misleading requests for supplementary patent certificates and strategic withdrawal of marketing authorisations to impede generic entry – were treated as abusive, even though patent law and regulatory rules permitted it. In Brazil – similar to the AstraZeneca case – CADE imposed a fine of 36.6 million reais in Eli Lilly (2015), for seeking to maintain its position as sole supplier of Gemzar, a cancer drug, by filing misleading and contradictory lawsuits with Brazilian courts and challenging the Brazilian Patent Office’s refusal to grant the patent of the cancer drug. This conduct was viewed as abusive ‘sham litigation’. Similarly, in February 2017, the US FTC sued Shire ViroPharma in federal court, alleging that ViroPharma engaged in a campaign of serial, repetitive and unsupported filings before the US Food and Drug Administration to delay the entry of generic competitors to Vancocin HCI Capsules. In December 2018, the 3rd US Circuit Court of Appeals in Philadelphia heard arguments from the parties concerning the FTC’s ability to sue where no further violations of antitrust law were ‘imminent’.
In recent years, antitrust challenges to pharma companies have intensified, particularly in the area of excessive pricing. In August 2017, China’s NDRC found an abuse by two Chinese active pharmaceutical ingredients companies that sold active ingredients for isoniazid at unfairly high prices. The NDRC imposed a fine of total 443,916 yuan on the two companies, equivalent to 2 per cent of their previous year’s sales in the relevant market. Shortly after that case, the NDRC published its ‘Price Conduct Guidelines for Operators of Drugs Prone to Shortages and APIs’.
Also in 2017, the Italian authority fined Aspen for charging unfair prices for various cancer drugs, which has in turn prompted the European Commission to launch its own (ongoing) probe into Aspen’s drug pricing in other member states. The Dutch competition authority appears to have prepared the legal and economic groundwork for excessive pricing probes, including through a submission to the OECD in 2018, followed shortly by a complaint submitted to the authority in respect of prices charged by Leadiant Biosciences. Outside of Europe too, pharmaceutical prices have raised concerns.
That said, the legal and economic tests for excessive pricing continue to raise challenges for enforcers. In AKKA/LAA, the Court of Justice confirmed that excessive prices need to be ‘significantly and persistently’ above the competitive level and there is a need for objective and consistent criterion in identifying the relevant comparators against which to test whether a price level is excessive. Failure properly to apply the relevant test led the UK Competition Appeal Tribunal to annul the CMA’s record fine that it imposed on Pfizer and Flynn In June 2018, although the CMA has been granted permission to appeal.
There is also much enforcement activity outside the sphere of excessive pricing. In a decision of 20 December 2017, the French authority imposed a €25 million fine on the laboratory Janssen-Cilag and its parent company Johnson and Johnson, for abusing its dominant position by repeating legally unjustified approaches to the French health agency for the purpose of convincing the authority to refuse to grant generic status to competing medicinal products. This follows previous cases of disparaging a generic rival, including charges against Sanofi and Schering Plough, which were upheld before the appellate courts in France.
Another practice raising concerns is ‘product hopping’ – the withdrawal of an existing drug from the market and introduction of a new one with a view to making entry by generics more difficult following the expiry of patent protection. In 2016, Canada’s Competition Bureau updated its Intellectual Property Enforcement Guidelines to cite product hopping as an example of a possible abuse of dominance. A series of product-hopping cases has made its way through courts in the US, including the ongoing Asacol litigation concerning the replacement of a drug used to treat ulcerative colitis.