Refocus on effects-based approach in rebates cases

The European Court of Justice (ECJ) has set aside the General Court’s judgment upholding the landmark EUR1.06 billion fine imposed on Intel for allegedly abusing its dominant position by trying to exclude its closest rival from the market. The ECJ referred the case back to the General Court without even considering all of Intel’s grounds of appeal in full, but along the way made some important statements supporting the role of effects-based analysis and efficiencies in loyalty rebates and in abuse of dominance cases more generally.


In 2009 the European Commission (Commission) found that Intel had, between 2002 and 2007, abused its dominant position in the market for x86 central processing units (CPUs). According to the Commission, Intel implemented a strategy aimed at excluding its only real rival, Advanced Micro Devices (AMD), from the market. Intel engaged in three specific types of conduct:

  • Granting rebates to four computer manufacturers (Dell, Lenovo, HP and NEC) on condition that they purchased all, or almost all, of their x86 CPUs from Intel.
  • Making payments to Media-Saturn, a European retailer, on condition that the computers it sold exclusively contained Intel x86 CPUs.
  • Making payments to three computer manufacturers (HP, Acer and Lenovo) on condition that they postponed or cancelled the launch of products with AMD CPUs, or restricted the distribution of those products (the Commission called these “naked restrictions”).

Intel was fined a (then) record breaking EUR1.06 billion for the conduct. It appealed the Commission’s decision to the General Court, which in 2014 handed down judgment supporting the Commission’s decision. Intel subsequently appealed the General Court’s decision to the ECJ. Last year, Advocate General Wahl gave his opinion that the ECJ should annul the General Court’s ruling on a number of grounds, including that it failed to assess the effects of the rebates.

The judgment: General Court gears up for round two

The immediate effect of the ruling is that the already long-running case is heading back to the General Court for its (re)consideration. The ECJ’s reason for setting aside the General Court’s ruling was that the General Court had held that it did not need to consider Intel’s criticisms of how the Commission had carried out the ‘as efficient competitor’ (AEC) test. The aim of this test is to assess whether a theoretical competitor with the same costs as the dominant undertaking would be able to compete with it.

In its original decision the Commission stated that there was no need to consider all the circumstances of the case, since the rebates in question were “by their very nature” capable of restricting competition. However, since the Commission had in any event conducted an in-depth analysis of these circumstances, including applying the AEC test, the ECJ concluded it had been important for the Commission’s finding of an infringement. The General Court should therefore have dealt with Intel’s concerns about how the AEC test had been carried out. The ECJ has instructed the General Court to re-examine the question of whether the rebates were capable of restricting competition.

The ruling is interesting on this point alone but, more crucially, the ECJ makes some key statements about how the Commission (and Courts) should assess rebates more generally.

Rebates: the return to an effects-based analysis

Rebates are still common commercial practice, and where entered into by non-dominant companies will usually raise no competition issues. The ECJ’s judgment doesn’t change anything here. In fact, the judgment clearly sets out that the purpose of Article 102 is not to stand in the way of a company’s success on a market, nor to ensure that “less efficient” competitors are able to survive on the market. The judgment goes on to include a reminder, though, that companies in a dominant position must be careful due to their “special responsibility” not to abuse that position.

In its earlier ruling, the General Court identified three categories of rebates: (1) ‘quantity rebates’ (generally a green light), (2) ‘other rebates’ that do not encompass an element of “exclusivity”, but still have a “fidelity-building” effect (highly fact dependent), and (3) ‘exclusivity rebates’ (categorised by the General Court as being a clear red flag from a competition law perspective). According to the General Court, the rebates implemented by Intel fell in this final category.

While the General Court found that exclusivity rebates amounted to inherently abusive conduct, the ECJ has decided on a more effects-based approach. In particular, the ECJ clarifies existing case law to say that if the dominant company argues that its conduct was not capable of restricting competition or producing foreclosure effects, then the Commission must take a closer look at the conduct, including all the circumstances of the case. In particular, it must assess whether the conduct forms part of a broader strategy to exclude competitors that are at least as efficient as the dominant company.

The ECJ also reiterates that rebates which have a restrictive effect can still be objectively justified. The Commission must balance possible negative effects on competition with any positive effects, such as advantages in terms of efficiency which may also benefit consumers.

The ECJ has therefore put the spotlight back on effects in Article 102 cases.

In practical terms, the judgment brings the approach to Article 102 enforcement on rebates in line with the effects-based approach taken in enforcement on other types of conduct, namely pricing abuses. The Commission had already advocated a more effects-based approach to rebates (and abuse cases more generally) in a 2009 guidance paper, and had been putting this theory into practice. The judgment suggests a helpful aligning of policy at the Courts and the Commission and acts to relieve the anxiety which resulted from the General Court’s hard line approach to loyalty rebates.

Nevertheless, dominant companies which are considering implementing rebate schemes should remain cautious - and rebates that may contain potentially exclusive or other loyalty inducing conditions should still be avoided in order to steer clear of competition issues. Even if such schemes may be able to be objectively justified in theory, compelling evidence will be required.

Interviewers take note

One aspect of the ECJ’s ruling which should make the Commission sit up and take note relates to the Commission’s handling of an interview with a Dell executive – notably its failure to make a record at the time of the interview.

The General Court found that the meeting in question was not an ‘interview’ within the meaning of the relevant regulations as it was an ‘informal’ interview. The General Court’s view was therefore that the Commission was not obliged to make a record at the time. However, neither the Advocate General nor the ECJ accepted the General Court’s arguments, since there is no distinction between ‘formal’ and ‘informal’ interviews in the regulations.

They found that the General Court was also wrong to conclude that, even if the Commission had committed a procedural error by not making a record of the meeting at the time (in breach of Intel’s rights of defence), this was cured by the subsequent disclosure to Intel of an internal note from the Commission’s file. The Advocate General thought that the General Court’s ruling should be set aside on this point. However, according to the ECJ, Intel had not shown it would have been able to use any evidence given in the meeting in its favour. The ECJ therefore confirms previous case law by requiring the company to show that the irregularity had affected the evidence available for it to use in its defence.

In any event, a clear outcome of the ECJ’s judgment is that the Commission must make a record of all interviews relevant to the case and make these available to the company in question if it wants to avoid its decisions being successfully challenged. This means that, going forward, the Commission will need to ensure proper procedure is followed.

Qualifying for jurisdiction

When establishing whether it has jurisdiction over a company’s conduct, the Commission has, to date, looked at where the conduct is implemented (the ‘implementation test’) or where the effects of the conduct are felt (the ‘qualified effects test’). Intel argued that the Commission could surely not have jurisdiction over Intel’s agreements with another non-EU party relating to products being sold in China, either on the basis of where the agreements were implemented, or where the effects were felt.

However, the ECJ took a different view and dismissed Intel’s arguments. Notably, on the application of the qualified effects test, the ECJ agreed with the General Court that it should be “foreseeable” that the conduct in question will have “an immediate and substantial effect in the European Union” in order for the Commission to have jurisdiction. When it comes to foreseeability, the ECJ held that it is enough to look at the “probable effects” of the conduct. It also stated that it is important to look at the company’s “overall strategy” when considering whether conduct is capable of affecting the EEA market.

The Commission will take heart from this aspect of the judgment. Companies, however, should be aware of the long reach of the Commission’s arm when it comes to assessing the compatibility of their conduct with EU competition law.

What next?

The judgment is now subject to reconsideration by the General Court and, whilst the final decision will be eagerly awaited, we will most likely not hear anything for a couple of years. In the meantime the ECJ’s view is clear – an effects-based economic approach is key when assessing rebates.

The full judgment can be found here.