Relaxing whilst doing Competition Law is not an Oxymoron

Archive for July 2016

The General Court’s Judgment in Case T-216/13, Telefónica (Counterfactual reasoning applied to fine calculation?)

leave a comment »



A few days ago the General Court issued interesting Judgments in cases T-208/13, Portugal Telecom, and T-216/13, Telefónica that have gone fairly unnoticed (possibly because they are not available in English). For the purpose of this post I will discuss the Telefónica Judgment.

The annulment proceedings relate to a Commission decision sanctioning Telefónica and Portugal Telecom (“PT”) for having included a clause under which both parties committed “to the extent permitted by law” not to compete with each other in the Iberian market in any new projects or activities in the telecommunications sector (including fixed telephone and mobile telephone services, internet access services and television services) as part of the agreement for the sale of PT’s stake in Brasilcel to Telefónica during a certain period. The clause was meant to be in force between September 2010 and 31 December 2011 but was formally withdrawn following the opening of the case in February 2011.

The decision considered this clause akin to an outright market sharing agreement, and dismissed the arguments Telefónica put forward in relation to, among others, the fact that a qualification “to the extent permitted by law” had been introduced at the end of the negotiations; the absence of actual anticompetitive effects; the alleged ancillary nature of the clause; as well as the argument that the clause was in reality considered ineffective following an ex post self-assessment undertaken by the parties’ lawyers. For the purposes of the calculation of the fine, the decision took into account the value of sales made by the parties’ in their respective home countries and applied a 2% coefficient that resulted in almost 67 million euros for Telefónica and in roughly 12 for PT.

The General Court’s detailed Judgment validates practically all of the Commission’s substantive thesis with one notable exception related to the calculation of the fine. But don’t lose the interest yet; there is some very interesting legal debate here:

-On the one hand, the Judgment confirms that, having regard to the legal and economic context, the conduct was a “by object” restriction. Telefónica had claimed that the clause was ancillary to a complex deal and that it had been imposed by the Portuguese government, so that its only choice was to limit the impact of the clause by forcing a self-assessment of any future conduct by including the phrase “to the extent permitted by law”. The Court dismisses the argument observing that Telefónica’s behaviour was autonomous and not forced by the Portuguese government (¶¶ 111-120) and that the claim that the clause was considered indispensable by the Portuguese government had not been proved (¶¶ 123-166). The Judgment does not accept Telefónica’s contention that adding the “to the extent permitted by law” qualification conditioned the validity of the clause to a subsequent self-assessment. It essentially observes that the parties themselves did have doubts as to the legality of the clause and that no one had explained why it had not been possible to clarify the matter prior to the closing of the deal or to its entry into force (¶¶ 181-192). Moreover, the Judgment does not accept the statements of the parties’ lawyers made before a notary as those do not coincide with the content of the written agreement to suppress the clause; ¶¶ 187-200).

The counterfactual. In its fourth ground of appeal Telefónica had claimed that the Commission had failed to assess in detail the structure of the markets in order to verify whether there would have been real, concrete opportunities for the parties to compete during the period affected by the clause had the latter been absent (i.e. that the Commission failed to address the counterfactual). The legal logic is correct: an agreement cannot restrict competition that would not have existed, but the argument was lost on the facts. The Judgment responds to these arguments noting that in this case it was not necessary for the Commission to assess in detail the structure of markets or potential competition given that the non-compete clause itself implied an acknowledgment of at least potential competition, that its subject-matter consisted of market-sharing, that its scope was very wide and the affected services had just been liberalized (¶¶ 201-227).

-On the other hand, however, and this is the main novelty in the case, the Judgment rules that the Commission was nevertheless required to assess potential competition between the parties for the services affected by the clause when calculating the value of sales. Since this exercise was not conducted, the Commission is ordered to make a fresh finding with regard to the calculation of fines (see ¶¶ 295-310). This relates mainly to the sales made by virtue of activities that would not have been subject to competition even absent the agreement  (e.g. services provided under monopoly conditions or others where PT’s access was impossible; see ¶ 274). The Judgment provides that the Commission should have examined the parties’ arguments seeking to establish that there was no possibility of competition for certain services, and that only after it might determine the value of sales linked to the infringement for fining purposes.

So, effectively, the Court endorses the Commission’s stance not to assess potential competition for the purposes of determining the legality of the conduct but nevertheless requires it to conduct this exercise at a later stage, when calculating the fine. And you may wonder: Why? Is this right?

Here is my off-the-top-of my-head take subject to our usual disclaimer:

  • My feeling is that the Court was also –rightly- seduced by the counterfactual logic, as it wouldn’t make sense to sanction a restriction of competition that would not have existed. However, the Court’s overall assessment of the nature of the clause and context to the case led it to conclude (like in Toshiba, also cited in the Judgment) that the agreement did restrict at least potential competition that would otherwise have existed. This is correct in general, but also ignores that the scope of the infringement would have been reduced by excluding those activities where competition was not possible (in fact, ¶ 221 seems to suggest this could have been done) (admittedly, I don’t know how the parties argued it so perhaps the Court wasn’t able to do more). Accordingly, it moves on to the next step and follows a similar logic at the stage of fine calculation.
  • In doing so, the Court encounters a problem, as there is also case-law from the ECJ (cited in 306) stating, in the context of market sharing, that one cannot uphold an interpretation whereby the Commission would be, when calculating fines, subject to obligations to which it is not subject for the purposes of the application of Article 101 (C-543/07, Prym, ¶ 64).
  • The Judgment appears to be aware of this tension and therefore observes that the ECJ’s Judgment in Prym was rendered at the time the previous Fining Guidelines were in force (306 in fine), and emphasizes that in this case it is not imposing different obligations on the Commission but merely extracting the necessary consequences of recital 13 of the current fining guidelines which was self-imposed by the Commission (and which provides that in determining the basic amount the Commission will take the value of the undertaking’s sales of goods or services to which the infringement relates). This is an interesting and certainly defensible argument; at the same time, some may claim that it does not fit squarely with other case-law (e.g. C-580/12 P, Guardian ¶¶ 57-58).
  • In my personal view, and taking for granted the Judgment’s factual assessment, perhaps the “more correct” solution, and one that would have avoided this tension, would have been to rule, in the first place, that the clause restricted competition by object but only in relation to activities where there could have been viable competition between the parties. There is abundant case-law that would have supported this reasoning (STM, European Night Services, O2, E.On –discussed in the Judgment- and others). Should that have been done, the Court would not have found the above-mentioned obstacles. Now, this is only my hastily formed opinion; happy to think it through together in case you might have comments.

If interested in all of these issues (which must be the case if you made it this far), you should know that Pablo and I are (or rather he is and I should be) working on a paper that develops the views expressed here, particularly regarding counterfactual assessments.

Written by Alfonso Lamadrid

12 July 2016 at 6:56 pm

Posted in Uncategorized

Yet another presentation on competition and big data

leave a comment »


My silence on the past few days has to do with several open fronts thanks to the Commission’s bad habits of summertime desk cleaning, but also to my bad habits of devoting  non-work time to conferences and talks (my only consolation is that Pablo has recently been by far the most active speaker in Chillin’ Competition’s flying circus).

-Some of you have asked for the presentation I used at the VUB’s very interesting debate on big data and competition law; here it is:

Competition-big-data_lamadrid 23 June

You know my views from quite a few previous posts (all links appear at the end of the ppp). The main addition this time was to discuss the joint French-German report issued last May which essentially makes general conjectures about how standard theries of harm could apply to big data (like they apply to any other asset) if the right facts were to arise in a given case. In sum, nothing new under the sun.

The change of attitude on the part of competition authorities is nevertheless remarkable. When I spoke at the EDPS closed-door workshop at the European Parliament in one of the first discussions on this matter my views were perhaps a bit anti-climatic for an audience pre-disposed to use competition law to tackle non-competition issues. But I did -logically- have the support of the only authority in the room, DG Comp. Now, however, we see not only the German Facebook case and the French-German report giving further visibility to a non-issue, but I also hear that some within DG Comp are pushing to do more on this front. That’s disconcerting.

-None of you have asked for the presentations I have used the past two Fridays at the College of Europe Summer Courses (where for the 4th year in a row I’ve lectured on Antitrust Procedure and Article 106). Lack of interest has never precluded me from posting stuff here, but since the two presentations are in Mandarin I’ll spare you the pain…

Written by Alfonso Lamadrid

12 July 2016 at 6:46 pm

Posted in Uncategorized

Some thoughts after the Intel hearing: the Court will choose between legal consistency and continuity

with 4 comments

I have read with great interest the report on the Intel hearing that Trevor Soames shared via Chillin’ Competition. I now feel I have a good sense of what went on and, more importantly, of what is at stake.

Above all, the hearing has revealed the Court of Justice will be making a choice between legal consistency and continuity. It is not a secret that courts typically lean towards stability, and thus tend to favour the latter. More often than not, it is sensible to do so, primarily because it preserves legal certainty.

Sometimes, however, continuity can be the very source of legal uncertainty. Where the case law is plagued with inconsistencies, it may not be possible to tell in advance whether a practice is lawful or unlawful. In such circumstances, favouring continuity becomes counterproductive. It may indeed provide stability, but at the price of making the case law impenetrable. If legal certainty is to be achieved, it may be necessary to refine it.

As the law stands, exclusive dealing and loyalty rebates are prohibited as abusive by object under Article 102 TFEU. These practices are only acceptable where the dominant firm is in a position to provide an objective justification. I see the attraction of continuity, and I certainly understand the reluctance to depart from a line of case law that dates back to the 1970s. Thus, I would not be surprised if the Court of Justice went for stability in Intel.

The hearing has made apparent, on the other hand, that the prohibition by object of exclusivity and loyalty rebates is a source of legal inconsistencies. It is simply not possible to reconcile this line of case law with other rulings. This is the reason why rebate cases remain controversial almost 40 years after Hoffmann-La Roche, and the reason why the Commission reviewed its enforcement priorities a decade ago.

The existing legal inconsistencies have become, if anything, more apparent after Post Danmark II. This idea is clearly illustrated by comparing the legal treatment of standardised rebates schemes (at stake in Post Danmark II) with the legal treatment of exclusive dealing and loyalty rebates (at stake in Intel).

Consider the example of a standardised rebate scheme that only covers 1% of the market. Is this practice prohibited under Article 102 TFEU? Almost certainly not. The Court explained in Post Danmark II that a standardised rebate scheme is only abusive if it is likely to have exclusionary effects. The Court also mentioned the coverage of the practice as one of the factors that determines the likelihood of such effects. And it is very unlikely that a rebate scheme that covers 1% of the market will be exclusionary.

How about an exclusive dealing agreement that only covers 1% of the market? If the supplier is dominant, this agreement would be prima facie prohibited by object. It would be very unlikely to have exclusionary effects. As the case law stands, however, this fact is irrelevant. It would still be unlawful.

I gather from the reports of the hearing that many of the questions asked by the Advocate General and the judges focused on this inconsistency. This is not surprising. It is difficult to think of a reason why loyalty rebates, on the one hand, and standardised schemes, on the other, should be treated differently under Article 102 TFEU (most probably because there is not a single one). The case law (Michelin II, British Airways, Tomra) shows that loyalty and standardised rebates are similar in their nature, purpose and (pro- and anticompetitive) effects (so much so, in fact, that it is sometimes difficult to distinguish between them in practice).

Against this background, I feel that, in this particular instance, continuity appears to do clearly more harm than good, and that, accordingly, consistency should be favoured instead. What I do not know, alas, is whether the Court will come to the same conclusion.

Written by Pablo Ibanez Colomo

7 July 2016 at 11:25 pm

Posted in Uncategorized