Chillin'Competition

Bananas have traditionally been an important product in competition law. Among others, they provoked the peculiar market definition at issue in United Brands (vitiated by the toothless fallacy :“the banana has certain characteristics , appearance , taste , softness , seedlessness , easy handling , a constant level of production which enable it to satisfy the constant needs of an important section of the population consisting of the very young , the old and the sick”), and they also inspired Kevin Coates’ “exploding banana hypothesis”.

Most recently they were the subject of the ECJ’s Judgment in Dole. A few weeks ago Pablo commented on this case focusing on how the Judgment illustrates that the “object” label is not about formal categories nor about a presumption of effects. I don’t disagree with Pablo’s views, but I think that they only tell one part of the story.

The facts

In a nutshell, employees of companies active in the banana trade apparently had numerous bilateral calls to discuss/disclose pre-pricing information (namely factors relevant for the setting of quotation prices for the forthcoming week or price trends). These exchanges of views were in a sense pure gossip, and were not liable to affect real market prices because quotation prices were neither actual prices nor the basis for the negotiation of the actual prices. Moreover, the Commission had not contested that the employees taking part in these discussions did not have the authority to set the quotation prices.

The legal issue

Against this background, the legal question raised by Dole’s third ground of appeal was notably whether it is possible to characterize the information exchanges as an infringement by object. According to Dole, the information exchange was in no way capable of reducing uncertainty on the market regarding actual prices.

The ECJ validated the General Court’s conclusions in this regard, observing essentially that information had been exchanged, that the information could be relevant to infer “signals, trends or indications”, that accordingly the exchange of info created abnormal conditions of competition, and that a given practice may have an anti-competitive object even if it does not have a direct link with consumer prices.

Why I think this is bananas

Many of you may think that there’s nothing new here, and that all this was already present in T-Mobile (and partly in the guidelines on horizontal agreements), and you would be right. This is not so much a novelty as an additional (and particularly illustrative) step in a very wrong direction.

The point I want to make today is not about whether the object label was rightly applied or not (a matter on which I have doubts, particularly if one takes seriously the requirement on the “sufficient degree of harm” set out in para 58 of Cartes Bancaires).

My point is that even if the object categorization were correct, this should only entail a procedural consequence: that the Commission would be dispensed of the burden of proving effects.  In spite of my doubts, I can see how the Commission could regard these practices as being more restrictive than not and lacking a “legitimate objective” (which was the sensible point made by Pablo in his post on the Judgment).

In my view, the widespread misconception lies in the automatic identification of “object restriction” with “very serious infringement” and even with a “cartel”. In other words, the way I see it, “object” is about obviousness, not about gravity.

Even if the practices at issue were labelled as object and not considered objectively justifiable or redeemable under 101(3), they –apparently- were little more than gossip of irrelevant employees with regard to quotations far removed from actual prices. Is that really so serious as to deserve a 60 million euro cartel fine? I don’t think so. And would the Commission have characterized it equally had it not received a leniency application? I doubt it.

A cartel is something else and is subject to a whole different level of reproach (even criminal in some jurisdictions); companies and individuals know when they are engaging in a cartel, and do not engage in it unconsciously; a cartel does have effects; a cartel is the “supreme evil of antitrust” (I’m using Scalia’s words in Trinko), and an exchange of information like this, which appears as practically irrelevant at all levels, might not be right, may be a restriction by object, but it certainly is not a cartel deserving a quasi-criminal fine. It is a venial sin, not a mortal one.

Holding the contrary is not only at odds with traditional (pre-T Mobile) case law, it also is at odds with economic reality and with the principles underlying any sanctioning regime; it is, in sum, bananas.

For earlier and more developed related views on my part, see here and here.