TY - JOUR
AU1 - Romano, Subiotto QC,
AU2 - R, Little, David
AU3 - Romi, Lepetska,
AB - Key Points In 2017, the Commission maintained its robust enforcement of Article 102, conducting four unannounced inspections, opening two new investigations (excessive pricing and export restriction concerns), and issuing one Statement of Objections (import restrictions). In its most significant Article 102 enforcement action of the year, the Commission issued a record €2.42 billion fine in the high-profile Google Search (Shopping) case, finding that Google had committed an abuse by displaying its own comparison shopping service (‘CSS’) more favourably than rival CSSs. In Intel, the European Court of Justice clarified the treatment of exclusivity and loyalty rebates under Article 102, rejecting the per se unlawful treatment of exclusivity rebates, setting aside the General Court judgment and referring the case back to the lower court for re-consideration. I. Overview of 2017 and expectations for 2018 The Commission has continued to enforce Article 102 TFEU actively. In 2017, the Commission conducted four unannounced inspections in relation to possible Article 102 infringements.1 The Commission has opened two formal investigations,2 and issued one SO.3 Three investigations in which the Commission issued a Statement of Objections (‘SO’) in 2016 remained open at the end of 2017.4 The Commission’s high-profile investigations into Gazprom5 and Qualcomm6 were also ongoing at the end of 2017.7 In Google Search (Shopping), the Commission found that Google had abused its dominant position on the search engine market by giving preferential treatment to its comparison online retail service, Google Shopping, and demoting rivals in its search results.8 The Commission fined Google €2.42 billion for this infringement – the largest fine issued against a single company in the history of EU antitrust enforcement. Google and its parent company, Alphabet, have challenged the decision.9 The Commission issued a prohibition decision against Lithuanian rail company Lietuvos geležinkeliai (‘LG’), fining it €27.87 million for hindering competition on the rail freight market by removing a rail track connecting Lithuania and Latvia.10 It also published two rejection decisions,11 one decision to terminate commitments,12 and closed proceedings in two cases.13 The Commission accepted Article 9 commitments from Amazon in its E-Book MFN case.14 Amazon offered the commitments in order to allay the EC’s provisional concerns that parity clauses in Amazon’s e-book distribution agreements (also known as ‘most-favoured-nation’ clauses) were anticompetitive, as they required publishers to inform Amazon when they offered more favourable terms to its competitors and/or to offer Amazon similar (or better) terms. Amazon committed not to enforce the clauses in its existing agreements, not to include them in new agreements, and to allow publishers to terminate e-book contracts. The General Court (‘GC’) dismissed five challenges against Commission decisions, four of which concerned the Commission’s rejection of complaints.15 The Court of Justice (‘ECJ’) handed down a judgment on abuse of dominance through excessive pricing in AKKA/LAA,16 and set aside the GC’s decision in Intel on exclusivity and loyalty rebates.17 Both ECJ judgments are discussed in greater detail below. In addition to the action lodged by Google, appeals were also brought before the GC and the ECJ by Qualcomm and Agria Polska, respectively.18 2018 may see the conclusion of several long-running Commission investigations, including the investigation into Gazprom (which has been active for around seven years) and the Google Android and Search (AdSense) cases (in both of which the Commission issued SOs in 2016). The Commission has a further two outstanding SOs (BEH gas market and ABInbev (Limes)); each of these investigations is likely to see further progress in the months to come. II. Summary of selected decisions and judgments from 2017 In line with previous surveys, this section provides a thematic overview of last year’s Commission decisions and European Court judgments with Article 102 elements. In this survey we focus on the following themes: the implications of the Google Search (Shopping) case for the analysis of abusive discrimination under Article 102 (c); the treatment of exclusivity clauses and exclusivity rebates where used by dominant companies (E-Books, the ECJ’s judgment in Intel); and the rules governing excessive pricing (AKKA/LAA). A. Favourable positioning and display (Google Search (Shopping)) On 27 June 2017, almost seven years after opening its investigation, the Commission adopted a prohibition decision (‘Decision’) in the Google Shopping case.19 The Decision claims that Google Inc. (‘Google’) committed an abuse by positioning and displaying its own comparison shopping service, the Google Shopping website, more favourably in its general search result pages, compared to rival CSSs. The Commission imposed a €2.42 billion fine on Google and its parent company, Alphabet Inc. It is the highest fine the Commission has ever imposed on a single undertaking for breach of EU competition law. The Commission also ordered Google to cease the infringement and remedy the abuse within 90 days by ensuring that, should Google continue to show Shopping Units, it has to position and display results from CSSs using the same underlying processes and methods. CSSs are online search services that allow users to search for products and compare their prices and characteristics across offers from different merchants. The Decision concludes that CSSs constitute a distinct product market that does not include merchant platforms like Amazon and eBay (on which users can both search for products from different merchants and also buy items). The Decision finds that Google holds a dominant position in the national markets for general search services in all 31 EEA countries, with market shares mostly exceeding 90%. The Commission concludes that the infringement took place in 13 of these countries. The conduct started in 2008 in Germany and the UK, and at various later dates in the other 11 countries. At the time of the Decision, the conduct was still ongoing in each of the 13 countries. The Decision claims that Google favoured Google Shopping by showing Product Universals and, later, Shopping Units on its general result pages.20 The Decision alleges that the two types of results, while not themselves CSSs, are a ‘means’ to favour the Google CSS—the separate, standalone Google Shopping website—because their display on Google’s general result pages benefits that service. First, the Decision describes how CSSs are positioned and displayed in Google’s search results. The Decision finds that Google’s approach to ranking generic results meant that CSSs were ‘prone’ to having their ranking reduced. Product Universals and Shopping Units were not subject to these generic ranking algorithms. The Decision does not object to Google’s algorithm and the ranking of rival CSSs in generic results but, rather, to the fact that such algorithms did not apply to its own specialised search results. The Decision also finds that, while Google showed Product Universals and Shopping Units prominently at the top of the results, competing CSSs could only be displayed as generic search results and did not enjoy richer format options, with pictures and additional information. Second, the Decision alleges that users tend to click more on links that are more visible on the general search results page, irrespective of their relevance. The Commission cites experiments suggesting that the higher the rank of a search result, the higher the click-through rate, even if the result is less relevant. Third, the Decision compares the evolution of the visibility of a subset of CSSs and the evolution of generic search traffic from Google to these services. The Decision finds that Google’s conduct decreased traffic from Google to competing CSSs and increased traffic from Google to Google Shopping. It concludes that CSSs have not replaced lost generic search traffic lost through alternative traffic sources, such as advertising, mobile apps, or direct traffic. Fourth, the Decision finds that Google’s conduct is capable of having, or is likely to have, anticompetitive effects. The Decision identifies potential foreclosure in a market for CSSs that excludes merchant platforms, such as Amazon and eBay. The Decision accepts that Amazon and eBay, like CSSs, allow users to search for and compare products. But it claims that such platforms do not compete with CSSs because they also allow users to buy products.21 The Decision claims that Google’s conduct has the potential to foreclose CSSs, which may lead to higher fees for merchants, higher prices for consumers, and less innovation. But the Decision does not set out evidence of these effects (despite the conduct allegedly lasting for nine years). The Decision characterises the alleged abuse as a practice that extends dominance from general search into comparison shopping. The Decision focuses on the concept of ‘favouring’. A claim of favouring implies discrimination. But the Decision does not attempt to reconcile the Commission’s analysis with the established case law on discrimination. For example, Google argues that product ads in Shopping Units are improved ads that serve to monetise Google’s general result pages. These product ads are different to generic results for CSSs. Google contends that it legitimately treats them differently to generic results as part of its ad-funded business model. The Decision does not explain, however, why showing Product Universals and Shopping Units differently to generic results for CSSs treats comparable situations differently.22 The Commission also rejects the application of the Bronner criteria because it claims that (i) the alleged abuse is not a passive refusal to grant CSSs access to Shopping Units, and (ii) to end the infringement, Google is not required to enter into agreements with CSSs.23 The Decision therefore creates the impression that Google’s ‘leveraging’ infringement is largely sui generis, or at least a substantial variation on more established themes. On 11 September 2017, Google challenged the Decision to the GC. Practitioners, enforcement agencies and businesses alike will be watching the proceedings keenly to see how the GC frames its review.24 B. Parity clauses (E-Book MFNs) Parity, or most-favoured-nation (‘MFN’), clauses are contractual provision which oblige a seller to offer a buyer the most favourable terms made available to the buyer’s competitors (or vice versa). MFN clauses can create efficiencies and benefit consumers. For example, MFN clauses may reduce transaction and/or (re)negotiation costs.25 However, MFN clauses may also adversely affect competition. For example, MFN clauses may precipitate coordinated effects by preventing a supplier from offering selective discounts to certain buyers, thereby softening price competition between buyers. Where sellers cannot offer selective discounts this may also have an exclusionary effect as rivals and/or prospective entrants cannot negotiate lower prices from the seller in order to compete with a dominant purchaser on price (and potentially other parameters of competition, depending on the scope of the MFN). European competition authorities have investigated parity clauses in a number of recent cases, including in relation to the supply of e-books26 and online hotel booking.27 The Commission first opened an investigation into Amazon’s MFN clauses in its agreements with various e-book publishers in June 2015.28 The agreements at issue included both price and non-price parity clauses. Two non-price related clauses29 required publishers to offer Amazon the same or similar distribution terms for e-books, and to offer Amazon equivalent functionality for any e-books with new features, as the publishers agreed with competing distributors of e-books. Publishers were required to inform Amazon of alternative terms offered to Amazon’s competitors and to offer the same or similar terms to Amazon. The Commission provisionally found that the non-price related parity clauses reduced competing e-book distributors’ and suppliers’ ability and incentive to develop new business models. Any new business models would need to be notified and offered to Amazon. By preventing distributors from differentiating their services from Amazon through the introduction of new business models, the clauses had the potential to soften competition between suppliers and stifle new entry, thereby strengthening Amazon’s dominant position. Amazon’s contracts also included five price-related parity clauses.30 The clauses granted Amazon the right to the same or similar discounts, commissions, or agency-set prices as competing e-book distributors. Publishers were also required to inform Amazon of pricing and distribution terms offered to rival distributors. The non-price related clauses also reduced rival distributors’ and suppliers’ ability to differentiate themselves on content, availability, format, and features and reduced innovation, quality, and choice. These clauses had a particularly dampening effect on the development of e-books with new features, such as animated e-books or special editions, as Amazon would be able to ‘free-ride’ on the distributors’ development efforts. Similarly, suppliers would be unwilling to develop new features for non-Amazon distributors as they would then be obliged to help Amazon implement those features. The price-related clauses, on the other hand, reduced distributors’ incentives to compete on the commission charged on e-books. Absent such parity clauses, competing distributors selling under agency agreements would have competed on the commission charged to their e-book suppliers, such suppliers having an incentive to direct sales to those distributors by offering them better retail prices. By doing so, such distributors would have increased the volume of e-books sold on their platform, growing their business. In its preliminary investigation, the Commission found that all parity clauses taken together deterred competition and innovation in the publishing sector, as they reduced distributors’ and suppliers’ ability and incentive to experiment with alternative business models and pricing schemes, and prevented effective competition at the distribution level by discouraging or even preventing competitors from undercutting Amazon’s prices. The Commission found that each category of parity clause individually constituted an abuse of dominance. Additionally, the clauses had mutually reinforcing effects as they covered each aspect of competition between e-book distributors. While Amazon disagreed with the Commission’s concerns, it offered not to enforce its MFN clauses and notification requirements with e-book suppliers or to introduce new ones. The initial commitments were revised after a market test to clarify the definitions and thereby prevent circumvention. On 4 May 2017, the Commission accepted the commitments, which will be in force for five years and apply to all e-books distributed by Amazon in the EEA. The investigation provides confirmation that parity clauses remain an area of interest for European competition authorities. Moreover, the commitments show that—at least in the case of digital markets—this interest is not limited to price parity arrangements but extends to other areas in which suppliers compete, including innovation in design and user experience. C. Excessive pricing (AKKA/LAA) Standalone excessive pricing cases have been relatively rare in competition enforcement in Europe. Over the past years, however, European competition authorities have shown a renewed interest in pursuing such investigations. The UK Competition and Markets Authority (‘CMA’) fined Pfizer and Flynn £90 million in December 2016 and currently has two ongoing investigations that focus on similar concerns.31 The French competition authority announced a sector inquiry in November 2017 to, amongst others, review pricing concerns.32 In May 2017, the EU opened an investigation into Aspen’s alleged excessive pricing practices for cancer medication.33 In addition to already having been fined for this behaviour by the Italian competition authority in October 2016,34 the Spanish competition authority also announced that it started a sanction procedure against Aspen in February 2017.35 Director General of Competition Johannes Laitenberger has commented that, for the EC, ‘keeping prices low for consumers remains a top priority across the board.’36 However, due in part to limited case law in this area, and in part to the significant discretion afforded by the applicable legal test, it has traditionally been difficult for companies to anticipate when their pricing might be deemed ‘excessive’. A preliminary reference by the Latvian Supreme Court provided the ECJ with an opportunity to give directions. The case concerned fees charged by the Latvian collecting society AKKA/LAA37 for the exploitation of its writers’ musical works in Latvia. The Latvian competition authority found that the prices charged by the collecting society were appreciably higher than those charged by collecting societies in neighbouring countries. The authority therefore imposed a fine for abuse of dominance by excessive pricing. The ECJ had previously determined in United Brands that charging a price which is ‘excessive because it has no reasonable relation to the economic value of the product supplied’38 can be abusive and introduced a two-step assessment. First, it should be determined whether the difference between the costs actually occurred and the price charged is excessive.39 If so, it should then be determined whether the excessive price is either unfair in itself or when compared to competing products.40 AKKA/LAA’s appeal led the Latvian Supreme Court to refer five questions related to the assessment method to the ECJ. Following AG Wahl’s opinion on 6 April 2017,41 the ECJ handed down judgment on 14 September 2017. In its first question, the referring court asked whether trade between Member States was capable of being affected by the collecting society’s behaviour, allowing for the application of Article 102. As the collecting society also collected remuneration in respect of works of foreign authors in Latvia, the ECJ confirmed, in line with its previous judgments, the application of Article 102.42 In its second, third and fourth questions, the ECJ considered the appropriateness of the comparison method43 used by the Latvian competition authority to assess whether the license fees were excessive. Reiterating its test, the ECJ explained that United Brands also recognised that there were other methods to determine whether a price may be excessive.44 As it had already been used in previous cases, a ‘comparison with the prices applied in the Member State concerned with those applied in other Member State must be considered valid.’45 In those cases, the ECJ had held that where comparison showed that the prices were appreciably higher and the comparison had been made on a consistent level, the difference should be regarded as indicative of an abuse of a dominant position.46 In his opinion, the AG stressed that competition authorities should aim to use multiple methods in their assessment ‘in the absence of an ubiquitous test and given the limitations inherent in all existing methods.’47 While the ECJ did not refer to this – possibly because the Latvian authority had only used a cross-country comparison – its clear recognition that multiple tests may be used suggests that it may not strictly oppose the AG’s suggestion. In response to the question of whether the authority’s comparison had been sufficiently representative, the ECJ held that ‘there was no minimum of markets to compare and [that] the choice of appropriate analogue markets depends on the circumstances of each case.’48 In his opinion, the AG had raised the concern that a competition authority could arbitrarily exclude countries from the comparison or, ‘worse still, because they produced data which did not ‘suit’ the case.’49 To remedy this concern, the AG had proposed to ensure that the sample of comparison countries was as broad as possible. The ECJ did not follow this proposal, but did agree with the AG that a comparison with only a number of Member States could be sufficiently representative where such reference Member States had been selected in accordance with ‘objective, appropriate and verifiable criteria,’ which may include ‘consumption habits and other economic and sociocultural factors, such as gross domestic product per capita and cultural and historical heritage.’50 The referring court was tasked with determining, in light of the circumstances, whether the criteria were relevant. As to the comparison itself, the ECJ held that it should be made on a consistent basis. While the difference in purchasing power parity should be accounted for (using the Purchasing Power Parity (‘PPP’) index), the referring court should review whether the method of calculating rates was comparable in the different countries. Lastly, the referring court had also asked whether it was appropriate to compare different user segments, such as shops and service centres of a specific surface area, as opposed to the average rate for all segments. Reiterating that the authority has ‘a certain margin of manoeuvre’ and that there was no ‘single adequate method,’ the ECJ held that ‘it [is] permissible to make a comparison within one or several specific segments if there are indications that the possibly excessive nature of the fees affects those segments.’51 This assessment differs from the AG’s simple solution to only consider assessment by segment where those segments constitute separate product markets for the purposes of Article 102. Commentators have rightly raised concerns that both the allowed ‘margin of manoeuvre’ and the ‘indications’ are unnecessarily vague compared to the AG’s clear solution.52 Perhaps most interesting for future excessive pricing cases were the fifth and sixth questions in which the referring court asked above what threshold a rate was to be considered excessive, and what evidence could be adduced to demonstrate that rates were not excessive. In response to the fifth question, the ECJ held that there is ‘no minimum threshold above which a rate must be regarded as ‘appreciably higher’, given that the circumstances specific to each case are decisive in that regard.’53 Instead, a difference could be appreciable where it ‘is both significant and persistent on the facts, with respect, in particular, to the market in question.’54 The ECJ concurred with the AG in holding that, for a price difference to be abusive, it must be ‘significant’, ‘persist for a certain length of time’, and ‘must not be temporary or episodic’.55 The AG set an even higher burden for intervention, which was not followed by the ECJ. He held that an authority should only intervene when it feels sure that the price difference is ‘of such a magnitude that almost no doubt remains as to [its] abusive nature.’56 On the other hand, the AG considered that it should be easier for an authority to discharge its burden of proof the more significant the price difference and the longer it had been applied.57 The ECJ also clarified that all these factors were ‘merely’ indicative of an abuse of dominance. At that point, it is for the dominant undertaking to ‘show that its prices are fair by reference to objective factors.’58 For a collecting society, such factors could relate to differences between the Member States used in the comparison, and any factors that could impact the management expenses or the remuneration of rights holders, such as local regulations. While the AG agreed that a dominant undertaking could show that its prices were fair, it was important to assess the objective reasons behind its prices. In line with United Brands, the AG distinguished between prices that were unfair by themselves, and where the abuse was therefore revealed by the price itself, and prices that were unfair in comparison to other products. While the former category was rare (e.g., where customers did not receive a service in return), the second category allowed for additional factors to explain the high price. For example, the dominant undertaking may have higher costs or the customer may perceive the quality of its products as being higher. In his conclusion, the AG continues to suggest a high burden for intervention: ‘it is only when no rational economic explanation—other than the mere capacity and willingness to use market power even when abusive—can be found for the high price applied by a dominant undertaking that that price may be qualified as abusive under Article 102 TFEU.’59 In summary, the judgment provides helpful guidance on the analytical framework for assessing excessive pricing behaviour. In particular, the conclusion that such prices must be significant and persistent is welcome, and provides a commercially pragmatic degree of flexibility for dominant companies. The ECJ nevertheless refrained from endorsing other clear proposals in the AG’s opinion. The content of the excessive pricing abuse therefore remains relatively poorly defined within the broader body of Article 102 case law. The Commission’s Aspen investigation will hopefully provide greater clarity for businesses. D. Exclusivity and loyalty rebates (Intel) The Intel judgment handed down by the ECJ on 6 September 2017 provided clarity on the treatment of exclusivity and loyalty rebates under Article 102. The ECJ held that the GC had erred in failing to examine all of Intel’s arguments calling into question the anticompetitive effects of the rebates under an ‘as efficient’ competitor analysis. The ECJ referred the case back to the GC, so that it may examine whether the Commission was correct to conclude that the rebates at issue were capable of restricting competition. Rebates are not necessarily unlawful under EU competition law. Earlier case law distinguished between three categories: (1) volume-base discounts, which were presumptively lawful; (2) exclusivity rebates, which required customers to obtain ‘all or most’ of their requirements from the dominant supplier and were considered unlawful ‘by their very nature’; and (3) a third category of rebates, which were not exclusivity-based but had a ‘fidelity inducing effect’, and were neither lawful or unlawful but should be assessed in ‘all the relevant circumstances’.60 The Commission 2009 Guidance Paper on its Article 102 enforcement priorities, set out the effects-based approach to analyses of exclusivity or conditional rebates, including by assessing whether the practice was capable of foreclosing a hypothetical ‘as-efficient’ competitor (‘AEC’).61 The Commission fined Intel a then-record €1.06 billion for infringement of Article 102. The Commission’s decision applied the form-based test and found Intel’s practice of offering conditional rebates to manufacturers to be unlawful per se, but bolstered its conclusion with an effects-based analysis of the restrictive effects on competition, including through lengthy application of the ‘AEC test’.62 Intel challenged the Commission’s decision, and the case became the subject of heated legal debate. Proponents of the formalistic approach argued, among other things, that it offered legal certainty and procedural simplicity, and that the AEC test was unsuited to an assessment of conditional rebates.63 Advocates of the effects-based approach maintained that the traditional approach was unnecessarily static and constrained competition by ruling certain practices unlawful without assessing their actual economic effects.64 On 12 June 2014, the GC came down firmly on the side of the formalists, confirming that Intel’s rebates were indeed unlawful ‘by their very nature’, and stating that the Commission’s application of the AEC test had been superfluous.65 Intel brought six grounds of appeal before the ECJ, three of which concerned the legal qualification and treatment of exclusivity rebates.66 It argued that the GC had applied the wrong legal standard and had failed to examine all the relevant circumstances when determining whether the rebates were capable of restricting competition.67 In his Opinion of 20 October 2016, AG Wahl recommended that the ECJ uphold all three grounds, set aside the judgment, and refer the case back to the GC.68 The ECJ handed down its judgment on 6 September 2017. It followed the AG’s Opinion and challenged the GC’s formalistic position that the use of exclusivity rebates by a dominant undertaking was per se abusive under Article 102. While upholding the presumption of illegality set out in Hoffman-La Roche,69 the ECJ clarified that a dominant undertaking could rebut the presumption by submitting evidence during the Commission’s investigation that the relevant conduct was not capable of restricting competition or of producing the alleged foreclosure effects.70 In such instances, the Commission is required to carry out an effects-based analysis to decide whether the conduct is capable of producing anticompetitive effects. Relevant circumstances included: (i) the extent of the undertaking’s dominant position on the relevant market; (ii) the share of the market covered by the contested practice; (iii) the conditions and arrangements for granting the rebates in question, their duration, and their amount; and (iv) the possible existence of a strategy aimed at foreclosing AECs.71 The ECJ also confirmed the principle set out in Post Danmark72 that the purpose of Article 102 was not to prevent an undertaking from attaining a dominant position or to prevent the marginalisation or foreclosure of less efficient competitors.73 But it also confirmed the relevance of the AEC test set out in the Commission’s Guidance Paper. It held that, while an AEC test is not essential in every case, it may be necessary to counter prima facie evidence submitted by the dominant undertaking that its conduct had no anticompetitive effect. As the Commission had undertaken a lengthy AEC analysis in its decision, the ECJ ruled that the GC was required to examine Intel’s submission on appeal that the test had been applied incorrectly. 74 The fact that the GC had refused to do so because it considered the rebates to be per se abusive was held to be sufficient ground to refer the case back to the lower court.75 The Intel decision also resolved an inconsistency in the treatment of exclusive vertical arrangements under Articles 101 and 102. Under Article 101, if an undertaking’s market share exceeds 30%, the Commission must carefully assess whether the practice under consideration has restrictive effects on competition. Under Article 102, prior to Intel, where an undertaking’s shares exceeded 50% (triggering a rebuttable presumption of dominance), exclusivity rebates were considered per se infringements, removing the option of an effects-based analysis. The Intel decision has resolved this tension. The ECJ also aligned certain aspects of the analysis of ‘by object’ restrictions under Article 101 with the per se abuses under Article 102. Under Article 101, a by object restriction may be lawful if other economic and legal considerations indicate that it is not liable to impair competition.76 While a similar exclusion was previously unavailable under Article 102, the Intel judgment held that an effects-based analysis could be used to determine whether the exclusionary effect of the rebates could be ‘objectively justified’, i.e., counterbalanced or outweighed by efficiencies benefiting the consumer, by taking into account ‘all the circumstances’.77 Lastly, the judgment also created parallels between the analytical framework used to assess predatory pricing practices and the analysis of loyalty rebates. When considering whether exclusivity arrangements are capable of restricting competition, the Commission is required to assess the existence of an ‘overall strategy’ aimed at foreclosing competitors.78 Footnotes 1 Commission STATEMENT/17/285, 15 February 2017 (relating to the electricity sector in Greece); Commission press release IP/17/1108, 25 April 2017 (relating to the mobile telecommunications sector in Sweden); Commission STATEMENT 17/1910, 4 July 2017 (relating to the motor insurance market in Ireland); Commission MEMO/17/3761, 6 October 2017 (relating to online access to bank account information by competing service providers). 2 Case AT.40394—Aspen, Commission press release IP/17/1323, 15 May 2017 (excessive pricing of medication); Case AT.40335—Romanian gas interconnectors, Commission press release IP/17/1501, 1 June 2017 (gas export restrictions). 3 Case AT.40134—Limes, Commission press release IP/17/5041, 30 November 2017. The Commission took the preliminary view that AB InBev abused its dominant position on the Belgian beer market by hindering cheaper imports from the Netherlands and France into Belgium. 4 Case AT.40099—Google Android, Commission MEMO/16/1484, 20 April 2016; Case AT.40411—Google Search (AdSense), Commission MEMO/16/1484, 20 April 2016; České Drahy, Commission press release IP 16/3656, 10 November 2016. 5 Case AT.39816—Upstream gas supplies in Central and Eastern Europe, Commission press release IP/15/4828, 22 April 2015. 6 Case AT.39711—Qualcomm (predation), Commission press release IP/15/6271, 8 December 2015; and Case AT.40220—Qualcomm (exclusivity payments), Commission press release IP/15/6271, 8 December 2015. 7 Gazprom offered commitments under Article 9 in December 2016. The Commission launched a market test on March 16, 2017 (Case AT.39816—Upstream gas supplies in Central and Eastern Europe, Commission press release IP/17/555, 13 March 2017). In October 2017 the Commission deemed them insufficient and invited Gazprom to make revisions. See MLex, ‘Gazprom told to improve commitments in EU antitrust probe after market test’, 13 October 2017. There were no further developments in either of the two Qualcomm cases (predation and exclusivity payments) in 2017. However, in January 2018, the Commission announced that it would fine Qualcomm €997 million for abuse of dominance in connection with the exclusivity payments investigation. The EC’s findings will be summarised in next year’s survey. 8 Case AT.39740, Google Search (Shopping)—Commission decision of 27 June 2017 (‘Google Search (Shopping)’). 9 Case T-612/17, Google and Alphabet v. Commission, action lodged on 11 September 2017. 10 Case AT.39813—Baltic rail, Commission decision of October 2, 2017. LG is challenging the Commission’s decision before the General Court (‘GC’). 11 Case AT.40205—Krefelder, Commission decision of 24 July 2017; Case AT.40273—Spanish tobacco, Commission decision of 14 June 2017. 12 Case AT.39317—E.On gas foreclosure, Commission decision of 24 March 2017. 13 Case AT.40170—Exhaust systems, Commission decision of 28 April 2017; Case AT.39822—Refrigerants, Commission decision of 25 October 2017. While, the Commission had originally opened the Refrigerants case to investigate both Article 101 and 102 related complaints, the SO was ultimately brought only under Article 101 (Practical Law, ‘Commission sends statement of objections to Honeywell and DuPont’, 21 October 2014), Commission press release IP/14/1186, 21 October 2014. 14 Case AT.40153—E-book MFNs and related matters (Amazon), Commission decision of 28 July 2017; and Commission press release IP/17/1223, 4 May 2017. 15 The GC rejected a challenge by RF against the Commission’s rejection of its complaint in the freight rail sector on 15 September 2016 for being inadmissible due to RF’s breach of procedural deadlines (Case T-880/16, RF v Commission, order of 13 September 2017, ECLI:EU:T:2017:647); the GC also dismissed a challenge by Deutsche Telekom against a Commission decision of 17 February 2015 refusing it access to documents on grounds of confidentiality (Case T-210/15, Deutsche Telekom v Commission, judgment of 28 March 2017, ECLI:EU:T:2017:224); the GC dismissed in its entirety a challenge brought by Agria Polska against the Commission’s rejection of their complaint alleging that they had been the target of a campaign by, among others, Dow and BASF to shut them out of the market for plant-protection products (Case T-480/15, Agria Polska and Others v Commission, judgment of 16 May 2017, ECLI:EU:T:2017:339); the GC also dismissed in its entirety a challenge brought by Contact Software against the Commission’s decision of 9 October 2015 to reject its complaint against Dassault Systemes and PTC (Case T-751/15, Contact Software v Commission, judgment of 14 September 2017, ECLI:EU:T:2017:602); and the GC rejected in its entirety a challenge by the European Confederation of Watch Repairers (‘CEAHR’) against the Commission’s second decision to reject its complaint. (Case T-712/14, CEAHR v Commission, judgment of 23 October 2017, ECLI:EU:T:2017:748). 16 Case C-177/16, Autortiesību un komunicēšanās konsultāciju aģentūra / Latvijas Autoru apvienība v Konkurences padome (‘AKKA/LAA’), judgment of 14 September 2017, ECLI:EU:C:2017:689. 17 Case C-413/14 P, Intel v. Commission, judgment of 6 September 2017, ECLI:EU:C:2017:632 (‘Intel—ECJ appeal’). 18 Qualcomm brought an action before the GC to annul a request for information issued by the Commission on 31 March 2017 on the grounds that it infringed the principles of proportionality, necessity, sound administration and the right to avoid self-incrimination, lacked adequate reasoning, and reversed the burden of proof (Case T-371/17, Qualcomm v. Commission, judgment of 13 June 2017, ECLI:EU:T:2017:485). Agria Polska and others challenged the GC judgment in Case T-480/15, judgment 16 May 2017, ECLI:EU:T:2017:339 (see footnote 19) to the ECJ on grounds of manifest error of assessment, infringement of the principle that full effect must be given to EU law, and infringement of the principle of effective judicial protection (Case C-373/17 P, Agria Polska and Others v. Commission, judgment of 20 June 2017). 19 Google Search (Shopping). 20 Product Universals provided specialised search results for products on Google’s general search result pages. These were often accompanied by images and other information, such as price. These results were free: Google did not charge merchants for clicks on them. Shopping Units are a specialised ad format for product ads. These results lead users directly to the websites of Google’s merchant partners, who would pay for the traffic received. 21 The Decision also points to the fact that merchant platforms purchase product ads on Google as a reason to exclude them from the market. But the Decision does not explain why a customer relationship is relevant for demand substitution and users’ choices. 22 Case C-322/81, Michelin, judgment of 9 November 1983, EU:C:1983:313, para. 90. 23 Case C-7/97, Oscar Bronner GmbH & Co. KG v. Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co, judgment of November 26, 1988, ECLI:EU:C:1998:569, para. 41. 24 The Official Journal published the summary of Google’s action on October 31, 2017. See Official Journal of the European Union, Case T-612/17 Google and Alphabet v. Commission, OJ C 369, 30.10.2017, 37–38. 25 Other pro-competitive effects include the reduction of so-called ‘free-riding’ and the mitigation of ‘hold-ups’. For a more detailed analysis of MFN clauses, see F.E. González-Díaz and M. Bennett, ‘The law and economics of most-favoured nation clauses’, (2015) 1(3) Competition Law & Policy Debate, 26. 26 Case AT.40153—E-book MFNs and related matters. In addition to manufacturing its own e-book readers and tablets (Kindle), Amazon is also vertically active in the e-books sector. It is both a publisher of its own publications (upstream) and acts as a retailer for e-book suppliers, self-publishing authors, and its own licensed e-books (downstream). 27 The national competition authorities of Belgium, Czech Republic, France, Germany, Hungary, Ireland, Italy, Netherlands, Sweden and the UK together with DG COMP, initiated a monitoring exercise to measure the effects of parity clauses used by online agents in contracts with hotels. This report was triggered by a series of investigations into Expedia and Booking.com, that led to diverging results in various EU Member States. The report is available at: http://ec.europa.eu/competition/ecn/hotel_monitoring_report_en.pdf 28 In the e-books segment publishers acquire rights from authors, which are then commercialised through agreements with online distributors, such as retailer or agents. 29 Including business model parity clauses, and selection parity clauses. 30 Including agency parity clauses, discount pool provisions, promotion parity clauses, wholesale price clauses, and agency commission parity clauses. 31 On 21 November 2017, the CMA issued a statement of objections alleging that Concordia had charged excessive and unfair prices in relation to the supply of thyroid medication (liothyronine) in the UK. The CMA’s investigation into Actavis UK’s alleged excessive and unfair pricing practices in relation to the supply of hydrocortisone in the UK is also currently ongoing. The CMA issued a statement of objections on 16 December 2016. 32 French Competition Authority, press release, 20 November 2017, available at: www.autoritedelaconcurrence.fr/user/standard.php?id_rub=662&id_article=3067&lang=fr. 33 Case AT.40394—Aspen, Commission press release IP/17/1323, 15 May 2017. Aspen had already been fined by the Italian competition authority for this behaviour in 2016. 34 Italian Competition Authority, press release, 14 October 2016, available at: www.agcm.it/en/newsroom/press-releases/2339-a480-price-increases-for-cancer-drugs-up-to-1500-the-ica-imposes-a-5-million-euro-fine-on-the-multinational-aspen.html. This decision was confirmed on appeal by the Italian administrative court TAR Lazio on 14 June 2017. See also Practical Law, ‘TAR Lazio upholds Italian competition authority’s decision to fine Aspen over EUR 5 MILLION (Italy)’, RESOURCE ID W-010-0947. 35 Spanish Competition Authority, press release, 3 February 2017, available at: www.cnmc.es/2017-02-03-la-cnmc-incoa-expediente-sancionador-contra-el-grupo-aspen-y-su-distribuidor-en-espana. It is unclear whether this behaviour relates to the same cancer medication. 36 DG for Competition Johannes Laitenberger, speech at the CMS EU Competition Conference Brussels, 19 October 2017, available at: http://ec.europa.eu/competition/speeches/text/sp2017_18_en.pdf. 37 In full Autortiesību un komunicēšanās konsultāciju aģentūra/Latvijas Autoru apvienība (‘AKKA/LAA’) 38 Case 27/76, United Brands Company v. Commission of the European Communities, judgment of 14 February 1978, ECLI:EU:C:1978:00207 (‘United Brands’), para. 250. 39 United Brands, para. 252. 40 Ibid., para. 252. 41 Case C-177/16, Autortiesību un komunicēšanās konsultāciju aģentūra / Latvijas Autoru apvienība v. Konkurences padome, Opinion of AG Wahl, 6 April 2017, ECLI:EU:C:2017:286 (‘AKKA/LAA Opinion’). The Latvian Supreme Court referred a total of seven questions to the ECJ. The sixth and seven question (concerning the method for fine calculation and jurisdiction) will not be discussed in this article. 42 AKKA/LAA, para. 28 and case law cited. 43 The authority had compared the fees charged in Latvia to those in neighbouring countries Lithuania and Estonia, and supported its analysis with a comparison with the rates charged in other Member States adjusted by a price parity index to account for differences in citizens’ purchasing power. 44 AKKA/LAA, para. 37 referring to United Brands, para. 253 (‘Other ways may be devised – and economic theorists have not failed to think up several – of selecting the rules for determining whether the price of a product is unfair.’) 45 AKKA/LAA, para. 38. 46 AKKA/LAA, para. 39 and case law cited. 47 AKKA/LAA Opinion, para. 43. 48 AKKA/LAA, para. 41. 49 AKKA/LAA Opinion, para. 65. 50 AKKA/LAA, paras. 41–42. 51 AKKA/LAA, paras. 49–50. 52 See e.g., A.M. Waksman ‘A missed opportunity: AKKA/LAA v Competition Council’, (2018) 39(2) European Competition Law Review, 77–80. 53 AKKA/LAA, para. 55. 54 Ibid. 55 AKKA /LAA, para. 56. 56 AKKA /LAA Opinion, para. 112. 57 Ibid. 58 AKKA /LAA, para. 61. 59 AKKA/LAA Opinion, para. 131. 60 Case C-85/76, Hoffmann-La Roche v. Commission, judgment of 13 February 1979, ECLI:EU:C:1979:36 (‘Hoffmann-La Roche’); Case C-459/10 P, Tomra and Others v. Commission, judgment of 19 April 2012, ECLI:EU:C:2012:221; Case C-23/14, Post Danmark, judgment of 6 October 2015, ECLI:EU:C:2015:651 (‘Post Danmark II’); Case T-286/09, Intel v. Commission, judgment of 12 June 2014, ECLI:EU:T:2014:547 (‘Intel—GC challenge’). See also Practical Law, ‘A practical approach to rebates’, RESOURCE ID W-010-9828, law stated as at 23 October 2017. 61 Although the effects-based approach was not intended to replace the form-based legal test (it was a test of prioritisation), it did reflect a growing trend towards more rigorous assessment of the economic effects of a given practice. 62 The AEC test aims to determine whether the dominant company is pricing below its own cost to the extent that an equally efficient competitor would be foreclosed from the market. 63 See, e.g., W. Wils. ‘The judgment of the EU General Court in Intel and the so-called ‘more economic approach’ to abuse of dominance’, (2014) 37(4) World Competition, 28–30. 64 See, e.g., P. Rey and J.S. Venit, ‘An effects-based approach to Article 102: a response to Wouter Wils’, (2015) 38(1) World Competition, 28–29. 65 Intel—GC challenge, para. 99. 66 The six grounds of appeal were: (1) the GC erred in law by failing to examine the rebates at issue in the light of ‘all the relevant circumstances’; (2) the GC erred in law in its assessment of the finding of an infringement, notably on the issue of market coverage; (3) the GC erred in law as regards the legal characterisation of the exclusivity rebates which Intel concluded with HP and Lenovo; (4) the GC wrongly concluded that there was no material procedural irregularity affecting Intel’s rights of defence; (5) the GC misapplied the tests relating to the Commission’s jurisdiction over the agreements concluded between Intel and Lenovo in 2006 and 2007; and (6) Intel asked the ECJ to reduce or annul the fine in accordance with the principles of proportionality and non-retroactivity. 67 Intel—ECJ appeal, paras. 109–113. 68 Case C-413/14 P, Intel v Commission, Opinion of AG Wahl of 20 October 2016, ECLI:EU:C:2016:788 (‘Intel—AG Opinion’). For a more detailed review of AG Wahl’s Opinion, see R. Subiotto, D.R. Little and R. Lepetska, ‘The application of Article 102 TFEU by the European Commission and the European Courts’, (2017) 8(4) JECLAP, 263. 69 Hoffmann-La Roche, para. 89. 70 Intel—ECJ appeal, para. 138. The ECJ offered little explicit guidance on legal and evidential burden. While it may not be sufficient for the dominant undertaking to merely assert that its conduct is not capable of restricting competition, it would arguably be enough for it to submit prima facie evidence challenging the presumption of exclusionary effects. It would then fall to the Commission to prove that the rebates were in fact capable of restricting competition, based on a review of all the factual and economic evidence. 71 Ibid., para. 139. 72 Case C-209/10, Post Danmark, judgment of 27 March 2012, EU:C:2012:172, para. 21. 73 Intel—ECJ appeal, para. 133. 74 Intel—ECJ appeal, paras. 142–144. 75 Ibid., paras. 140–147. 76 Case C-67/13 P, Groupement des Cartes Bancaires v. European Commission, judgment of 11 September 2014, EU:C:2014:2204, para. 69 77 For example, in relation to price, quality, choice, or innovation. Ibid., paras. 139–141. 78 Cleary Gottlieb Alert memorandum, ‘Modernising Abuse of Dominance – the CJEU’s Intel judgment, 16 October 2017, available at https://www.clearygottlieb.com/~/media/organize-archive/cgsh/files/2017/publications/alert-memos/modernising-abuse-of-dominance-the-cjeus-intel-judgment-10-17-17.pdf. © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - The Application of Article 102 TFEU by the European Commission and the European Courts
JF - Journal of European Competition Law & Practice
DO - 10.1093/jeclap/lpy042
DA - 2018-09-01
UR - https://www.deepdyve.com/lp/oxford-university-press/the-application-of-article-102-tfeu-by-the-european-commission-and-the-kqdi3sitUI
SP - 476
VL - 9
IS - 7
DP - DeepDyve
ER -