Chillin'Competition

Relaxing whilst doing Competition Law is not an Oxymoron

On the Qualcomm (exclusivity payments) decision (by Max Kadar)

with 2 comments

Qualcomm (@Qualcomm) | Twitter

[The second installment of this series of guest blog posts is devoted to the Qualcomm decision on exclusivity payments, which was (finally) made public in June. The post has been prepared by Max Kadar, Deputy Head of Unit at the European Commission and Visiting Professor at King’s College London. He is uniquely placed to provide an overview of this landmark development. Thanks so much for accepting the invitation, Max!]

I confess that when Pablo got in touch a few days ago to invite me to write a blog post on the Qualcomm (Exclusivity Payments) decision, my instinctive reaction has been something along the lines of “after having spent the past months locked up at home, there’s no way that I am going to spend my holidays writing another article”. After some time reflecting on the matter, I ended up doing the only logical thing that I could have done in these circumstances: politely but firmly refusing the kind invitation writing the piece before going for holidays.

Before starting, I should say that I have a conflict of interest to declare, as I was part of the case team working on the Qualcomm (Exclusivity Payments) decision. In addition, as usual, I should say that my views in the following are personal and may not necessarily reflect the position of the European Commission.

The Qualcomm (Exclusivity Payments) decision adopted on 24 January 2018 (and a public version of which was published on the Commission’s website a couple of weeks ago) constitutes the first example of a Commission decision dealing with exclusivity rebates under Article 102 adopted after the Court of Justice’s Intel judgment.

As a brief reminder, in Intel, the Court of Justice clarified the Hoffmann-La Roche case law with regard to exclusivity rebates and stated that while such rebates by a dominant undertaking are subject to a presumption of anti-competitive effects, this presumption is a rebuttable one. In particular, if the dominant undertaking submits, “during the administrative procedure, on the basis of supporting evidence”, that its exclusivity rebates do not have the capability to restrict competition, the Commission will be required to engage in an analysis of such capability.

In the aftermath of the Intel judgment, academics and practitioners have engaged in an extensive debate on what exactly the Commission needs to show and how (my take on this here). Against this background, the Qualcomm (Exclusivity Payments) decision is important as it sheds light on the way the Commission interprets the standard of proof it has to satisfy under the Intel case law and in particular under paragraph 139 thereof.

The case concerns agreements between Qualcomm and one of its main customers, Apple. In a nutshell, pursuant to these agreements, Qualcomm granted significant rebates (in the form of direct payments) to Apple, in exchange for Apple exclusively incorporating Qualcomm baseband chipsets compliant with the LTE cellular communication standards (so-called “LTE chipsets”) in its iPhones and iPads. Such conduct was at the centre of the European Commission’s investigation in this case but was also pursued by the US FTC as part of its wider case against Qualcomm.

Qualcomm has historically been a leader in the supply of baseband chipsets and, at the time of the challenged conduct, was by far the world’s largest supplier of LTE chipsets, with market shares in the range of 90% or more for most of the infringement period. Although certain of Qualcomm’s customers were large companies such as Apple and Samsung, the Commission found that even these customers were to a large extent dependent on Qualcomm during the infringement period.

At first sight, one can say that the agreements between Qualcomm and Apple were similar to those challenged by the Commission in previous decisions on exclusivity rebates. On closer inspection, though, the agreements contained loyalty-inducing mechanisms that went even further. In particular, if Apple had incorporated LTE chipsets from a competitor of Qualcomm in some of its iPhone or iPad devices, it would not only (1) have lost all future payments from Qualcomm; but also (2) in the central years covered by the agreements (2013-2015), have had to return to Qualcomm a large part of the payments it had already received in the past (so-called “clawback” provision). Thus, it is perhaps not surprising that throughout almost the entire duration of the agreements, Apple sourced LTE chipsets exclusively from Qualcomm. Only in September 2016, when the agreements were about to expire and the cost of switching was reduced due to the limited future payments left and the termination of the clawback provision, did Apple start to source part of its LTE chipset requirements from one of Qualcomm’s competitors – Intel.

Building on these basic facts, the Commission decision first establishes that Qualcomm’s payments amounted to exclusivity payments or rebates within the meaning of the case law (see Section 11.3 of the decision). As a second step, the Commission analysed the capability of such payments to restrict competition on the basis of several elements (see Section 11.4 of the decision), namely: (i) the extent of Qualcomm’s dominance; (ii) the long duration (from 2011 until 2016) of the payments and their significance, both in absolute terms and as a percentage of Apple’s LTE chipset expenditure; (iii) the fact that Apple’s demand affected by the exclusivity payments covered a significant share of the LTE chipset market (up to roughly half of the worldwide market); (iv) the fact that Apple was a key customer for baseband chipset suppliers; and (v) a broad range of contemporaneous evidence, including Apple’s internal documents, showing that Apple gave serious consideration to switching part of its LTE chipset requirements for iPads to Intel but refrained from doing so, the loss of Qualcomm’s payments being a material factor in Apple’s decision-making.

As one can clearly see, the elements above are in line with those referred to by the Court of Justice at paragraph 139 of the Intel judgment to establish whether exclusivity rebates are capable of restricting competition. Notably, when it comes to the last point, one may even say that the Commission engaged in a form of actual effects analysis, which goes beyond the requirements of the case law.

As regards the so-called “as-efficient competitor” (AEC) test, the Intel judgment does not require the Commission to run such a price-cost test to prove an infringement of Article 102 (as recently confirmed by the UK CAT in Royal Mail). Against this background, the Commission did not have to rely on an “own” AEC test to establish that Qualcomm’s payments were capable of restricting competition. Nevertheless, in the decision, the Commission did analyse and rebut an AEC test that had been prepared and submitted by Qualcomm (see Section 11.5 of the decision).

In addition to submitting an AEC test, Qualcomm had also argued that by failing to carrying out an AEC test, the Commission breached Qualcomm’s legitimate expectations concerning the application of the Guidance Paper on Enforcement Priorities (see Section 11.7 of the decision). In this regard, the decision notes that Qualcomm had failed to submit contemporaneous evidence that it had indeed relied on the Guidance Paper in any way. For example, the AEC test submitted by Qualcomm was prepared specifically for the purposes of the proceedings and not at the time of the negotiations with Apple. In any case, the decision dismisses Qualcomm’s arguments, pointing inter alia to the fact that the AEC test is merely a tool that the Commission can use to establish its enforcement priorities and that in this case, there were many other clear reasons why the case should be handled as a priority (on a strictly personal level, I find it very difficult to argue that an agreement of the type at hand between two leading technology companies in a multi-billion market should not be investigated by the Commission as a matter of priority).

A final but important part of the decision’s effects analysis relates to Qualcomm’s claimed efficiencies (Section 11.6 of the decision). In essence, Qualcomm argued that its conduct was justified because the exclusivity was necessary to allow Qualcomm to recoup the alleged investments required for the design and production of tailor-made LTE chipsets for Apple. As is usual in Article 102 cases (see here), the Commission examined these claims in detail: there is indeed no doubt that efficiencies may outweigh the effects of anti-competitive conduct and it is the Commission’s task to look into this after having received a properly substantiated efficiency submission. In this case, however, the decision concludes that Qualcomm failed to prove the link between the exclusivity and the alleged need to recoup investments for tailor-made products. Once again, the decision also refers to the fact that Qualcomm failed to submit any contemporaneous evidence supporting its claims.

As a concluding remark, I note that, while the Commission adopted a Statement of Objections (SO) in this case well before the Court of Justice’s Intel judgment, the SO already contained an analysis on the capability of Qualcomm’s payments to restrict competition. This allowed the Commission to proceed to a final decision without the need to issue a supplementary SO after publication of the Intel judgment, which would have likely dragged on the proceedings for at least several more months. While the end-result was certainly good from an administrative perspective (as I hope we all share the importance of avoiding unnecessarily protracted legal procedures), it was certainly not a departure from the Commission’s recent decisional practice. Those who have been following the Commission’s approach closely know well that, regardless of whether the Commission is legally required to assess the potential effects of a given practice on competition, such an assessment is always included in the Commission’s decisions. While I would not go as far as saying that the debate between “by object” and “by effect” in Article 102 is purely of academic interest, I do think that the practical importance of such a debate should not be overstated.

What is in my mind, however, highly relevant, is the type of evidence that a competition agency can use to prove a case and the robustness of such evidence. The Qualcomm (Exclusivity Payments) decision confirms that the elements of evidence that can be taken into account to prove effects are many and varied. Furthermore, it confirms that there is no pre-determined hierarchy of importance between those elements. Depending on the specific circumstances in each case, its factual background and investigative history, a given element of the analysis may play a more important role than another. In this respect, one can say that the different types of evidence are like bricks used to build a wall: what matters is not whether they are of a given colour or shape, but rather that the wall in the end is solid and robust. Whether this was the case in this decision is a matter for the General Court to decide, in the context of the ongoing appeal of the decision by Qualcomm. In the meantime, one can only expect that exclusivity arrangements entered into by dominant firms, which have the potential of harming competition, will remain high on the Commission’s agenda, as shown for example by the ongoing Broadcom case.

Written by Pablo Ibanez Colomo

7 September 2020 at 7:36 am

Posted in Uncategorized

2 Responses

Subscribe to comments with RSS.

  1. Thanks for the interesting blog post.

    Regarding recital 529 of the decision (“Qualcomm has presented no contemporaneous evidence from the time of the negotiation and conclusion of the Agreements showing that it believed in good faith that the Commission would assess the legality of Qualcomm’s exclusivity payments in accordance with the Guidance on Enforcement Priorities”), could you please elaborate upon this element?

    Am I correct in my assumption that the Commission expects dominant companies to take into account the AEC test when assessing a possible payment scheme with a client? Isn’t an AEC test always post-factum by nature?

    Thomas Nelissen

    9 September 2020 at 12:54 pm

    • Hi Thomas,

      Thank you for your question!

      First of all, for the sake of clarity, the Decision (paras 527 and 528) recalls that an AEC test is not necessary for the legal qualification of an infringement.

      Second, as regards Qualcomm’s argument on legitimate expectations, which seems to be the focus of your question, the Commission does not “expect expects dominant companies to take into account the AEC test when assessing a possible payment scheme with a client”. What recital 529 rather says is that if a dominant company submits that it relied on a given framework of analysis, then that company should produce evidence corroborating its own assertion (see in this respect the GC judgment in T-371/17 Qualcomm and Qualcomm Europe v Commission, paras 175 and 190).

      Finally, I am not sure I agree with you that the AEC test is necessarily a “post-factum” exercise. At the time of entering into an agreement featuring rebates that are conditional on exclusivity, a dominant company should have a good degree of understanding of the market conditions and the elements to take into account for a potential forward-looking test. If on the contrary one were to assume that it is not possible for a dominant company to run a reliable AEC test before entering into such agreement, then I cannot see how in the first place one could reasonably argue that legitimate expectations are at stake.

      All the best

      Max

      Max Kadar

      10 September 2020 at 12:47 pm


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: