Chillin'Competition

My paper discussing the implications of the Court of Justice’s judgment in Intel is now available for download on ssrn (click here). As usual, I would very much welcome your thoughts on it.

Has something changed? Maybe nothing, but a lot needs to change

Has something changed with Intel? Strictly speaking, nothing has changed. It really is a clarification. The point of law introduced in Intel was there for all to see before the ruling. It was just waiting for the occasion to come out.

The above does not mean, however, that everything people were saying before Intel is correct, or that courts and authorities can carry on as usual as if the judgment had never been delivered. It may be true that nothing has changed in theory, but a lot needs to change in practice.

Exclusive dealing and loyalty rebates continue to be ‘by object’ infringements – in the sense that they are prima facie prohibited irrespective of their effects – after Intel. A practice is only prohibited by object if it is capable of having restrictive effects. However, capability in relation to these practices is presumed. In this sense, it is true that nothing has changed.

On the other hand, the Court made it explicit in Intel that dominant firms can challenge the presumption of capability. This is true of all ‘by object’ abusive practices, not just exclusive dealing and loyalty rebates. We knew it was possible to rebut the presumption of capability under Article 101 TFEU (Murphy). We now know for sure that the same can be done under Article 102 TFEU.

An arrêt-cadre, not an arrêt-loi

I explain in my article that Intel is an arrêt-cadre, not an arrêt-loi. I mean by this that Intel lays down a framework, but does not develop it. I have come to the conclusion that this is a positive aspect of the judgment: trying to craft a whole system from the top-down may be counterproductive. It is better to wait for concrete issues to arise – this is after all the whole philosophy behind the EU Treaties.

What are the main questions that need to be clarified? I identify a few of these in the article:

What does capability mean?

The first key question relates to the meaning of capability – that is, what we mean when we say that a practice is capable of restricting competition.

Is capability the same as likelihood? Where does the threshold of capability lie?

In this regard, I argue in the paper that the threshold of capability is a fairly low one. This low threshold makes sense in relation to abuses ‘by object’, which are (and/or should be) those that according to economics and experience are liable to result in harm to competition with no offsetting benefits.

What authorities need to prove is that anticompetitive effects are plausible – not even likely. I also explain why it is important to distinguish between capability and likelihood.

The plain meaning of these words (‘capability’, ‘likelihood’) is – as highlighted by several speakers during our Chillin’ conference – already suggestive of different thresholds. I explain in the paper that mixing up the two concepts would blur the line between ‘by object’ and ‘by effect’ infringements. And such an outcome would be undesirable.

If the line between object and effect became blurred, either all ‘by object’ conduct would become subject to a case-by-case effects analysis (which would make no sense) or all ‘by effect’ conduct would become, for all practical purposes, prohibited by its very nature – thereby rendering the case-by-case assessment meaningless in practice.

What level of evidence do firms need to satisfy?

We know after Intel that dominant firms can rebut the presumption of capability. The judgment is silent, however, about the level of evidence that dominant firms need to meet in this regard. Do they need to prove the absence of capability beyond reasonable doubt? Do they need to provide ‘convincing evidence’ within the meaning of Tetra Laval?

This question, I sense, is likely to be very relevant in practice. DG Laitenberger’s speech on how to apply Intel going forward certainly focused on evidence, as did Alfonso’s post commenting on that speech.  A competition authority, instead of assessing the capability of a practice to restrict competition, may attempt to argue that the evidence adduced by the dominant firm is insufficient to rebut the presumption.

The more I think about it, the more I am persuaded that it would be enough if the evidence brought forward by the dominant firm has an ‘air of reality’, which is a relatively low level. Why? Alfonso and my colleague Andriani Kalintiri argued – separately and persuasively – in my discussions with them that this level is consistent with the presumption of innocence.

Does Intel require the application of the ‘as efficient competitor’ test?

I do not think so. The ‘as efficient competitor’ test has always been a proxy, and the Court refers to it as a proxy.

This said, if a dominant firm shows, in light of the ‘as efficient competitor’ test, that anticompetitive effects are implausible – in the sense that rivals would not have to sell below cost – a competition authority cannot avoid engaging with the question. In such circumstances, the authority may show, for instance, that there are other factors suggesting that the practice is capable of driving an equally efficient rival out of the market.

But even if Intel does not require the use of the ‘as efficient competitor’ test, the Guidance Paper does.  But we can leave the role of the Guidance as a pre-commitment device for another day. In fact, Alfonso has an interesting view that I challenge him to write up here. While we wait for his post, you can take a look at the paper and send your comments my way!