Partner Adam Rooney assesses the recent Court of Justice of the European Union decision in Intel’s long-running abuse of dominance case, in Commercial Dispute Resolution.
Adam’s article was published in Commercial Dispute Resolution, 15 September 2017, and can be found here.
The recent judgment of the European Court of Justice (ECJ) requiring the General Court to look again at whether Intel abused its dominance of the market in CPUs marks another step in a dispute that is fast approaching its tenth year. Intel was fined €1.06 billion in 2009, following an investigation by the European Commission into loyalty rebates and payments which Intel offered to various PC manufacturers to only use its chips in their computers. Whilst Intel’s appeal is understandable, given the size of the fine, does the ECJ’s decision offer hope to Intel as we enter the final stage of the appeal process, or is this another example of a US tech giant taking steps to simply delay the inevitable?
According to the Commission’s decision of 13 May 2009, between 2002 and 2007, Intel sought to push its only real competitor, Advanced Micro Devices Inc, out of the market for x86 CPUs. The chips were commonly used in PCs produced by the likes of Dell, Lenovo, HP and NEC. Although Intel already enjoyed a 70% market share, it sought to increase that by offering substantial rebates if those manufacturers agreed to buy all (or almost all) of the CPUs only from Intel. It also made exclusivity payments to Media-Saturn in return for agreement to sell only PCs containing its chips. Following an investigation, the Commission concluded that these type of rebates and payments, given Intel’s dominant position, were of a type which – of themselves – could be assumed to restrict competition. It therefore concluded that EU antitrust law had been breached and levied what – at the time – was a record-breaking fine.
Intel’s appeal of that decision took every conceivable point. In the end, in June 2014, the General Court agreed with the Commission that the rebates were of a type that led to the presumption that competition was indeed restricted. In the Hoffmann-La Roche case, the ECJ had itself held that exclusivity rebates were abusive by their nature.
Undeterred, Intel made another extensive appeal to the ECJ. Its complaints that the Commission lacked territorial jurisdiction to penalise the abuse and alleging procedural irregularities that affected its rights of defence were roundly rejected. The ECJ accepted, however, that the issue of whether the rebates were capable of restricting competition should be looked at again. It therefore remitted the case back to the General Court with an instruction that the judges there should consider the rebates in light of an “as efficient competitor” (AEC) test. In layman’s terms, the ECJ was directing the General Court to consider whether, as a matter of fact, these rebates did restrict competition: could an “as-efficient” competitor offer the same prices and remain viable?
The “as-efficient-competitor” is a hypothetical competitor having the same costs as the dominant company. Foreclosure of an as-efficient-competitor means that the dominant company is pricing below its own costs.
The AEC test is controversial. Since the as-efficient-competitor-test is a cost-benchmark-test, the choice of the appropriate cost benchmark is crucial for the analysis. In other cases, the EU Commission has chosen the average avoidable costs (AAC) or the long-run incremental cost (LRAIC) as benchmarks. According to the EU Commission, failure to cover the AAC indicates that the dominant undertaking is sacrificing profits in the short term and an equally efficient competitor cannot serve the targeted customers without incurring a loss. Failure to cover the LRAIC indicates on the other hand that the dominant undertaking is not recovering all the (attributable) fixed costs of producing the good or service in question and that an equally efficient competitor could be foreclosed from the market. Other effects also need to be considered, such as the existence of economies of scale and scope, learning curve effects or first mover advantages which may favour declining costs. The exact application of the as-efficient-competitor-test as well as the chosen cost benchmarks vary, therefore, on a case-by case basis. This gives rise to scope to argue that the benchmarks or considerations used are wrong or skewed: an opportunity which Intel has taken and (following the ECJ decision) continues to take.
The ECJ’s reference to the AEC test in Intel’s case is interesting, because in other contexts its use has been underplayed on the basis that it can be unreliable. In Post Danmark II, the Danish Postal Service (which enjoyed a dominant position) was not allowed to rely solely on the AEC test to show that it was not abusing that position. This was on the basis that the test was seen to be “inefficient”. That case also revealed the tension between a form – and effects – based test. It may be that an as efficient competitor could compete; thus, an economist would approve the rebate scheme (effects). Yet, at the same time it may be that the rebate scheme is designed to foreclose access to the market; hence, the lawyer will not approve the rebate scheme (form).
Although the Commission’s decision against Intel did not strictly depend on any application of an AEC test, it did find – as a matter of fact – that the rebates were restrictive of competition. As the ECJ itself observed, the Commission had carried out an in-depth examination of the circumstances of the case leading to the conclusion that an as efficient competitor would have had to offer prices which would not have been viable and that, accordingly, the rebate scheme at issue was capable of foreclosing such a competitor. According to the Commission, the rebates and payments induced the loyalty of the four manufacturers and a seller who agreed to sell only PCs with Intel chips, and thus significantly diminished the ability of Intel’s competitors to compete on the merits of their x86 CPUs. Intel’s anti-competitive conduct thereby resulted in a reduction of consumer choice and in lower incentives to innovate.
Both the ECJ and the Commission have acknowledged that assessing a rebate scheme is rarely a standard assessment. All facts and circumstances must be included in the assessment before it can be established whether a rebate is anti-competitive or not. Given the findings of fact made by the Commission in Intel’s case, however, it is difficult to see the General Court reversing its previous decision.
So what exactly was the scale of Intel’s recent victory? The most likely impact is on how alleged abuse of a dominant position is to be assessed. The AEC test can be used. If it is, however, it would be prudent to also consider the test in reverse i.e. looking at the rebate scheme from the competitor’s perspective to see whether it is likely, based on the market conditions, to cause them to be foreclosed even though an economist may demonstrate that an as-efficient competitor can match the price offered by the dominant undertaking. In this way, both effects and form can be judged and – it is hoped – the right conclusions drawn.
Despite Intel’s best efforts, the appeal is akin to winning a battle, but not the war. As far as Intel is concerned, this may prove to be a pyrrhic victory in a long war where the odds are firmly stacked against them.