Intel and the effects-based approach in EU competition law
On 6 September 2017, the Court of Justice of the European Union ("CJEU" or "Court") essentially held in Intel1 that the European Commission ("Commission") cannot consider rebates, and in particular loyalty rebates, as per se illegal. Rather, the Commission needs to show that a specific rebates scheme is capable of restricting competition before finding a company liable for abuse of a dominant position under Article 102 TFEU. In doing so, the Commission is required to examine all the circumstances of the case and conduct the as-efficient competitor ("AEC") test.
The CJEU also implicitly rejected the view that loyalty rebates are not pricing practices and as such subject to a less demanding test of restriction of competition than pricing practices. Instead, the Court makes clear that rebates are a form of pricing practice and subject to the same standards. This is an important development that re-introduces reason in the case law. The Court also departed from the formalistic classification of rebates into distinct groups that was previously proposed by EU courts.
Based on this, the CJEU set aside the judgment of the General Court ("GC") because it did not examine the Commission’s analysis of the AEC test and all of Intel’s arguments concerning that test, and referred the case back to the GC for a new examination taking account of the CJEU’s judgment.
In May 2009, the Commission fined Intel for abusing its dominant position in the market for x86 central processing units ("CPU") between 2002 and 2007.2 Intel was accused of granting conditional rebates to four original equipment manufacturers ("OEMs"), Dell, HP, NEC, and Lenovo. According to the Commission’s decision, these rebates were conditioned on the OEMs purchasing all or almost all of their x86 CPUs from Intel. The Commission also held Intel responsible for making payments to certain OEMs and the then largest desktop computer distributor in Europe (MSH) in exchange for them delaying the marketing of certain products equipped with CPUs supplied by Intel’s only competitor, Advanced Micro Devices, Inc. ("AMD"). The Commission imposed a fine of €1.06 billion, which was at the time the highest fine ever imposed on a single company by the Commission.
The GC rejected Intel’s appeal and upheld the Commission’s decision in its entirety.3 It held that loyalty rebates can be categorized as abusive without any analysis of the circumstances of the case or the rebates’ effects on the market.4 The GC categorized rebates into three groups: volume rebates (presumed lawful), exclusivity/loyalty rebates (presumed harmful to competition) and other rebates (which need to be examined with reference to all the circumstances of the case to determine whether they can have a loyalty-inducing effect).5 Intel appealed the judgment on a number of grounds. It argued inter alia that the GC erred in accepting the position that loyalty rebates could be considered an abuse of a dominant position without first examining all circumstances of the case and without assessing the likelihood of that conduct restricting competition. It also argued that the GC erred in its analysis concerning the capacity of the rebates and payments to restrict competition in the circumstances of the case, including in correctly assessing the relevance of the AEC test applied by the Commission.
Although the judgment also includes important principles of procedural and jurisdictional nature, this client alert concentrates on the main substantive question – the CJEU’s views on rebates.
Strong endorsement of an effects-based approach
The Intel case over the last years acquired particular prominence and symbolized the battle in EU competition law between a more effects-based approach that is based on sound economic evidence and a more formalistic approach based on mere presumptions of competitive harm. The GC’s judgment in Intel and perhaps to a lesser extent a recent CJEU judgment in Post Danmark II6 were seen as formalistic, and many commentators doubted whether the Commission’s Guidance Paper7 on exclusionary abuses was of any continuing relevance. A particular point of contention in this regard was whether the AEC had any role to play in the analysis of rebates.
So the CJEU’s judgment was eagerly awaited by enforcers, companies and their advisors. If the CJEU were to uphold the GC’s judgment, it was not unthinkable that the Guidance Paper might be withdrawn. That would have been regrettably retrogressive. Advocate General Wahl’s Opinion of last October, which contained a powerful and very clearly articulated message in favour of an effects-based approach and proposed that the Court set aside the GC’s judgment, raised the stakes and expectations.8
The Court was evidently aware of the importance of the case. It should not be ignored that the judgment is rendered by its Grand Chamber, presided over by President Koen Lenaerts. This increases the authoritativeness of the ruling. It is also remarkable that the judgment is short – only 151 paragraphs long. The relevant part on the substantive analysis of rebates is not more than 40 paragraphs long. This is indicative of the fact that the Grand Chamber saw this as a judgment of principle.
Indeed, the Grand Chamber starts its analysis reiterating a fundamental principle which it established five years ago in Post Danmark I:9 dominant companies can engage in "normal competition", or in other words, "competition on the merits" which results in consumers benefiting from lower prices, better-quality products and a wider choice of new or improved goods and services. Such conduct will not be considered an abuse, even if it might result in the exclusion of competitors from the market. The CJEU confirmed "not every exclusionary effect is necessarily detrimental to competition" and that "[c]ompetition on the merits may, by definition, lead to the departure from the market or the marginalisation of competitors that are less efficient and so less attractive to consumers from the point of view of, among other things, price, choice, quality or innovation".10 This illustrates a distinction between anti-competitive foreclosure and "mere foreclosure", which may follow from “competition on the merits”, exactly as the Commission’s Guidance Paper advocates.11 It is telling that the Court starts its analysis with these powerful quotes from Post Danmark I, and every advocate of the effects-based approach will welcome this.
Then comes paragraph 137 of the judgment, where the Court breaks new ground: “Article 102 TFEU prohibits a dominant undertaking from adopting pricing practices that have an exclusionary effect on competitors considered to be as efficient as it is itself (…) by using methods other than those that are part of competition on the merits”. Indeed, “not all competition by means of price may be regarded as legitimate”.12 These statements qualify loyalty rebates as “pricing practices” and seem to brush aside – as suggested by AG Wahl in his Opinion13 – the distinction introduced by the GC judgment in Intel and by the CJEU itself in Post Danmark II between low pricing practices (predatory pricing, margin squeeze, selective price cutting) and rebates which were dealt with under different standards, being “non-pricing” practices. This is a crucial point in the judgment: by analysing rebates as a form of pricing practice and by explicitly referring to exclusionary effects on competitors considered to be as efficient as the dominant company, the CJEU in essence supersedes Post Danmark II and strongly endorses an effects-based approach. The application of the AEC test, in particular, is central to that approach and this is confirmed by the subsequent paragraphs of the judgment.
To be clear, the CJEU does acknowledge the old case law on exclusive dealing and loyalty rebates. However, it explains that that the case law needs to be “further clarified” where undertakings submit that their conduct was not “capable of restricting competition and in particular producing the alleged foreclosure effects”.14 While it could be argued that this means that anticompetitive harm continues to be simply presumed in exclusive dealing and loyalty rebate cases unless the dominant company puts supporting evidence forward, this is not a realistic reading of the judgment. In practice, all dominant companies defending themselves before a competition authority will contest the fact that their practices are capable of restricting competition and will certainly produce economic evidence to that effect. According to the Court, the Commission must now thoroughly examine the arguments showing that the practice is not capable of having effects on competition.
Specifically, the CJEU states that the Commission is required, before concluding on whether a loyalty rebate is abusive, to assess “all the circumstances of the case”, including: a) the extent of the undertaking’s dominant position on the relevant market, b) the share of the market covered by the challenged practice, c) the conditions and arrangements for granting the rebates in question, d) the duration of rebates, and e) the level of rebates.15 The Commission must also assess whether there is “a strategy aiming to exclude competitors that are at least as efficient as the dominant undertaking from the market”, by effectively applying the AEC test.
Finally, the CJEU notes that, once the assessment of all circumstances is completed and the AEC test is applied, and if the rebates appear to fall within the scope of Article 102 TFEU, the dominant company can raise an objective justification or an efficiency defence. The company needs to show that “the exclusionary effect arising from such a system, which is disadvantageous for competition, may be counterbalanced, or outweighed, by advantages in terms of efficiency which also benefit the consumer”.16 All practitioners know how difficult it has been to make such an argument successfully to date. The CJEU now makes clear that such a balancing exercise takes place after “an analysis of the intrinsic capacity of that practice to foreclose competitors which are at least as efficient as the dominant undertaking”,17 reminding us again of the importance of the AEC test.
In the last paragraphs of the judgment, the CJEU applies these legal principles to the facts of the Intel case and scrutinizes whether the GC correctly performed its task of reviewing the Commission’s decision. It concludes that although the Commission applied the AEC test and “the AEC played an important role in the Commission’s assessment of whether the rebate scheme at issue was capable of having foreclosure effects on as efficient competitors”, the GC failed to assess it and examine Intel’s argument concerning that test.
Given that the review of the arguments put forward by Intel involves the examination of factual and economic evidence that is out of the CJEU’s jurisdiction, the Court annulled the GC judgment and referred the case back to the GC.
In conclusion, the Grand Chamber’s judgment strongly endorses the effects-based approach and the AEC test in rebates cases. At the same time, although it does not cite the Commission’s Guidance Paper, it indirectly provides support to it. The Court’s initial statements are also of relevance to all types of exclusionary abuses.
As far as rebates are concerned, the Grand Chamber departs from the tripartite categorisation of rebates introduced by the GC in Intel and followed by Post Danmark II (per se lawful, per se unlawful and the grey zone), and effectively abandons the per se unlawful category altogether. If, for simplicity, categories have some value, then there seem to be two: a) volume-based incremental generalized rebates which are essentially per se lawful and b) all other rebates, to which, according to the CJEU, we must now apply an AEC test, like we do for other pricing abuses.
The Court’s judgment, being short, does leave many questions unanswered. For example, is it impossible for all above-cost rebates to be anti-competitive? Can the other "circumstances of the case" mentioned in paragraph 139 of the judgment supersede the AEC test in a given case when deciding about the "capacity to foreclose"?18 Another uncertainty exists in relation to the AEC test itself: a competition authority’s AEC test may not necessarily be identical to a dominant company’s AEC test; it may have very different results. One would also have to see how the GC will review the Commission’s discretion in conducting its AEC test in the Intel case. This all means that rebates do remain an area where dominant companies need to be careful and take good counsel. But, at the same time, yesterday’s ground-breaking judgment offers support to those who have advocated a more economic approach in Article 102 TFEU. The Commission and national competition authorities must now re-appraise their approach in light of this seminal judgment, and apply the legal and economic standards it mandates. The first step was made with the Court’s ruling. The next step is to see how the enforcers will apply that ruling in practice.
The procedural and jurisdictional issues of the Judgment
In addition to addressing the core issue of rebates, the Court’s judgment is also interesting in respect of two other aspects: one procedural and one jurisdictional.
First, the CJEU unequivocally sided with Intel regarding the Commission’s failure to record the content of a meeting with a Dell senior executive. Contrary to what the GC had found, the CJEU held that the Commission is under an obligation to properly record all interviews that it conducts during an investigation, which it did not do in this case, and also to make such records available to the company under investigation.19 However, the CJEU found that this procedural irregularity should not, in this case, lead to an annulment of the decision, since Intel could not show that the information withheld was of such nature as to cast a different light on the evidence relied on in the decision to establish the conditionality of the practices at issue.20
Second, the CJEU ruled for the first time that the Commission can apply Article 102 TFEU to conduct implemented outside Europe, when it is foreseeable that such conduct will have an immediate and substantial effect in the European Union (the so-called "qualified effects test", which had already been applied in the area of merger control). The CJEU further clarified that "probable effects" were sufficient to demonstrate the "foreseeability criteria", and that the test should be applied to the company’s strategy as a whole, and not to individual parts of such strategy. On this basis, Intel’s argument -- that the agreements concluded with Lenovo fell outside the Commission’s jurisdiction because their effects in Europe were negligible -- was rejected by the Court, in light of the fact that such agreements were part of a broader strategy that satisfied the qualified effects test. This element of the judgment will have important repercussions, since so much of the conduct investigated in competition cases is global in nature.