RE: Information exchange=cartel?
Some days ago I wrote a post on the change of approach towards exchanges of information set out in the new EU horizontal guidelines, in which I challenged the assertion that this practices should (or could) be equated to cartels (an assertion which, as I see it, has come out of the blue) and expressed my concern over the possibility that the statements made by the Commission in that document could be interpreted in a excessively wild wide manner by overzealous enforcers.
Since then, I have received various comments on that post. Since we´ve always liked the idea of fostering as much interaction as possible on this blog and most of those comments are not visible here, I´ve decided to provide an overview of what some of them said (other must be kept confidential) and, where necessary, to reply to some of the questions they raise. I have checked with their authors and have only mentioned their names where theu have given their express consent.
This will be lengthy, so, if interested, click here to keep reading.
– The first comment is the only one which was actually made on the blog. “M” wonders: “what about the T-Mobile judgment”?
Well, as I said in the post, T-Mobile has the potential to enlarge the concept of concerted practices to an extent that is yet uncertain, and endorses characterization of certain information exchanges as infringements by object.
I have two main comments in this regard:
Firstly, nowhere in that Judgment does the Court give leeway for the identification of such infringement, or of restrictions by object in general, with a cartel. Resale price maintenance, for instance, is also a restriction by object but no one has ever thought of treating such conduct as if it were a cartel.
Secondly, I also object to the characterization of certain information exchanges as infringements by their object, since in order to characterize something as a restriction by object we should be able to say, in the abstract, that they are anticompetitive under any circumstance. What the Court did there, and what the Commission is doing now, was/is basically to say, in a legally questionable way, that an exchange of information is a restriction by object when it is a restriction by its effect. If we were to accept that an infringement by object is a form of collusion capable of having restrictive effects (see paragraph 31 of the T-Mobile Judgment), how should we then define a restriction by effect? As a form of collusion which is not capable of having restrictive effects???. In a way, you could label this as ex-ante effects analysis. And in my view, the “spirit” of a restriction by its object is hardly compatible with subjecting its existence to multiple “ifs” and buts”.[Those interested on this point can jump to the discussion that is taking place in the comments to this post]
– The second comment came from Brazil, and reveals that the spill-over effects that I feared in other jurisdictions are not actually limited to EU Member States. The comment from our reader reads as follows: “I couldn’t agree more with you in relation to this post. Brazilian authorities have historically been completely lost with respect to information exchanges, as one can notice from their “guidelines” with respect to this topic (which basically say companies cannot do anything). It is clear that the new EU Guidelines will only complicate things in Brazil even more. If EU officials – who have more expertise than most other officials around the world– are capable of equating the two things, what to think of the rest of the world?”. Very interesting.
– An official of the Spanish Competition Authority wrote saying that their decision (see extended summary in English ) would probably give us more grounds to voice out our criticism than the press release did. According to this person our comments are welcome and much more useful than any praise, because that´s what spurs them to get better. I could not have hoped for a better attitude on the part of some of the staff within the CNC.
– One of the lawyers involved in the case before the CNC also suscribed what was said here and added that the decision in that case was not much better than the press release. (once again, I have no opinion on the decision itself since I haven´t had the chance to read it yet).
– Another comment comes from Pablo Figueroa, a Spanish Brussels-based lawyer who, in addition to stating that he shared what was said in the post, noted that the Spanish Competition Act characterizes all horizontal anticompetitive agreements as “very serious infringements” and who hinted, tongue in cheek, that, accordingly, the exchange of these sort of emails between us lawyers as well the blog itself could constitute a “very serious ingringement” under spanish law”. 🙂
– Nicolas Petit (the name might ring a bell..) also had some comments: I will now make you privy to our private email exchanges:
Nicolas firstly referred me to case-law regarding exchanges of future information (namely, IFTRA Rules on Glass Containers, Vegetable Parchments, and even Woodpulp (arguing that had there been an agreement in Woodpulp, it would have been sanctioned as a restriction by its object).
My reply email said exactly the following:
Since our subsequent meetings have taken place at bars and it´s clearly impolite to discuss competition law while having a beer, I had not yet replied to that observation, so I´ll do it now:
I totally agree on the fact that the main error lies in the use of the term “cartel” in the Guidelieslines, but I do believe that the concept of “cartel” is clearly a legal concept (the Spanish Competition Act, for instance, defines what it means by “cartel” and especifically says that leniency can only be open to cartel cases.
Even at the EU level, although (as with almost everything else) the Treaty does not specifically refer to cartels as such, I believe that the concept of cartel has, at least nowadays, the status of a legal concept to which particular consequences are attached. Some examples:
(i) Settlements, at least at the EU level (I understand that the OFT has a different stance), only apply to cartel cases (see “Commission Regulation (EC) No 622/2008 of 30 June 2008 amending Regulation (EC) No 773/2004, as regards the conduct of settlement procedures in cartel cases“, or the Notice on the “conduct of settlements in cartel cases” (which in footnote 2 defines what a cartel is).
(ii) The Commission’s leniency programme only applies to cartel cases (see the Commission Notice “on Immunity from fines and reduction of fines in cartel cases” and the statements made in the first paragraphs thereof).
(In fact, the enlargement of the notion of cartel may have a lot to do with leniency; let´s say you´re a competition authority and someone presents you evidence of an information exchange which could run afoul of competition law, what do you do? a) reject the application replying that the leniency mechanism is there for very particular infringements and use other investigatory instruments; or b) accept the case in order not to loose a chance to show the world how tough you are. The ideal would be a), but b) seems to be the most common in some jurisdictions).
(iii) Sanctions in cartel cases have traditionally been in a completely different dimension than those imposed so far on information exchange cases. This applies both to pecuniary sanctions as well as, very particularly, to alternative sanctions such as director disqualification of criminal conviction. Would it be proportionate to apply those to an incredibly wide array of possible information exchanges?
(iv) The disclaimer in paragraph 9 of the guidelines gives a hint on the fact that “cartel” is a concept with its own legal meaning. That disclaimer is, by the way, quite surprising and reveals that at least someone within the Commission must have shared the concerns expressed here and tried to control part of the damage.
In sum:
Equating information exchanges with restrictions by their object is very arguable, although the EU Courts seem to have validated such approach (we should probabñy discuss this on an ad hoc post). Doing that no matter what the circumstances are can be very counterproductive (even if convenient for enforcers) as it would remove the in-depth assessment that those practices really deserve. To be sure, the European Commission has not said that info exchanges shall be treated as object restrictions in all instances, but it has provided bold statements which other authorities may use as an alibi enabling them to label anything as a restriction by its object.
This is certainly a problem. An example: many trade associations have been exchanging information in ways which perhaps were/are not strictly in line with what the competition rules would demand. On many occasions (not all) the deviation from what´s strictly legal doesn´t stem from the will to attain any anticompetitive goal, but rather from them not being conscious of what they were doing. Those deviations can surely be reprehended, but to the extent they are not very significant they should not automatically be seen as very serious infringements by their object.
Competition authorities should be required, as they´ve always been pursuant to the case-law of the EU Courts, to prove that an info exchange is restrictive by its effects. After all, if something is so blatantly anticompetitive, it shouldn´t be difficult to prove that it can potentially give rise to anticompetitive effects (see amongst others para. 77 of the John Deere Judgment). We´re getting too used to see how shortcuts are used by NCAs so as to avoid thorough assessments. As I said on my previous post, let´s hope the Commission keeps an eye open and effectively supervises its partners.
Moreover, going beyond there, equating info exchanges or for that matter restrictions by their object with cartels (as both the European Commission and at least the CNC have done) is, in my view, also plainly wrong, and even more worrying.
In addition, as I explained on my previous post (and the CNC has proved true), the moment you say that some info exchanges of info are a cartel, then someone´s going to come and argue that all of them are a cartel (remember the “expansionist” tendency of legal concepts?).
In light of the foregoing I insist: enlarging the legal concept of cartel so as to encompass information exchanges, thereby applying to them all the instruments that were developed and justified in the framework of the fight against cartels is bad policy and a dangerous development.
Looking forward to more debating!
Chillin'Competition 
I’ve always felt uneasy with the object characterisation. In the end I believe it is exactly as you put it, object infringement are anticompetitive by their nature, which means you are pretty confident that they are restrictive on the basis of an ex-ante effects assessment – which for hard core restriction is simply skipped since it is too obvious to be bothered in the first place.
Now, let me put it this way: if on an ex-ante basis there are no reasonable efficiency rationales justifying the info exchange the main plausible explanation is anticompetitive in nature, since I don’t buy your story that trade bodies or managers are just naive or don’t have a clue of why they are exchanging info – i.e. the tasks of gathering and elaborating info reliably is time and resource consuming, it doesn’t occur by accident.
Think of this as a Chicago School’s critique by reverse – the original goes something like: if there are no anticompetitive rationales it must be efficiency enhancing; whereas this one goes: if there are no efficiency rationale it must be bad for competitive rivalry.
And let’s not forget, there is still art. 101(3) to rebut this presumption – or at least this is the way it ought to be, which to me it is the main reason why you ought to keep certain info exchanges in the object box but not treat them as cartel infringement.
Paolo Siciliani
30 March 2011 at 9:54 am
Paolo, we both seem to agree on the fact that treating info exchanges as cartels is wrong (although I don´t fully understand what you mean in your last paragraph, since 101(3) is theoretically also open to cartel infringements). That was my main point and the issue at the origin and at the core of this discussion.
Now, with regards to the object characterisation it´s true that the discussion can get a little more tricky. You say:
“object infringement are anticompetitive by their nature, which means you are pretty confident that they are restrictive on the basis of an ex-ante effects assessment – which for hard core restriction is simply skipped since it is too obvious to be bothered in the first place”.
In my view, the ex-ante effects assessment must be done in the abstract, as you propose to do for “hardcore restrictions”. Maybe our disagreement here lies on the fact that I do not see the distinction between hard core restrictions and object infringements; to me both are the same thing, that is, precisely those restraints for which the effects analysis “is simply skipped since it is too obvious to be bothered in the first place”.
Now, it is true that the object of effect characterisation becomes blurry when we address some kinds of practices. That´s the problem of all or nothing/ black or white characterisations; they are almost always inherently flawed. The current approach towards info exchanges is a further step along the path opened by T-Mobile, which allows to cover the middle ground between the two by endorsing a practical enlargement of the “object box”.
Contrary to that, to the extent it is not absolutely clear (in the abstract) that a practice will certainly bring about anticompetitive effects, I think it would be more correct to apply an “effects” assessment (which, btw, should be much easier to conduct than in most other cases!). If something is so blatantly anticompetitive why should we avoid competition authorities the little trouble of motivating how those effects may arise??
Even if in theory art. 101 (3) is theoretically there (in practice NCA´s also use the statements in the Commission guidelines on the unlikelyhoof that object restrictions might be redeemed by paragraph 3), one must acknowledge that the burden of proof here is key, and I don´t think it´s good for anyone (although it certainly is more convenient to some) to move towards easy object characterisations (as I said the other day, take a look at how many national cases are characterised as infringements by their effect and you´ll realise the perils of choosing that path).
As I write I´m thinking that also in the US they´ve faced this object-effect/ per se-rule of reason conundrum, and if you take a look at the Supreme Corut´s decision in California Dental v FTC you´ll find a good explanation of how they´ve decided to confront it (in that case the Court acknowledged that there was a sliding scale and stated that unless an “observer with even a rudimentary understanding of economics could conclude that the arrangements (…) would have an anticompetitive effect on customers and markets”, it was preferable to move towards the effect-side of that scale).
Surely it´s not always easy to identify bright lines, but that´s exactly why we should prefer thoroughness over shortcuts.
PS. My example of trade associations shall be read in its context. I took care to say that the circumstances to which I referred concur “on many occassions (not all)”, and to make clear that “small” infringements should also be reprehended. My point there is that not all minor deviations (if and when they are indeed minor) should be considered as an object infringement.
Alfonso Lamadrid
30 March 2011 at 11:53 am
In point of principle, I don’t see any distinction either, as I said I would simply see object as a case where the effects assessment ex-ante can be easily argued “in the abstract” – hard-core cases are slam-dunks, hence the fact that para 3 is only in principle open to the defendants, but as you said it is a matter of sliding scales.
Anyway, being agreements (as opposed to unilateral decisions), I find compelling the Chicago school line of attack by reverse that absent plausible efficiency rationales “in the abstract”, the burden of proof could shift on the defendants to disprove that the agreement is not restrictive by nature.
Paolo Siciliani
30 March 2011 at 1:25 pm
Company X purchases plant A and plant B from Company Y. The transaction is subject to merger control approval.
While the authority is reviewing the transaction, Company X ask for some volume figures linked to plant A. Company Y submit such figures. No transition / “clean team” process has been set-up between the 2 companies.
The authority request the divesture of plant A in order to clear the transaction.
The info exchange is now outside of the perimeter of the transaction and the data is strategic (accoridng to new Guidelines on horizontal agreements).
A cartel case?
Un abrazo,
Manuel
Manuel
31 March 2011 at 9:54 am
Very good question. Also this comes up at due diligence procedures, if a competitor wants to allow another competitor to inspect its files, contracts etc. for the purpose of a future transaction. Under the Guidelines I see that – theoretically – it can be a problem under 101, especially if the merger does not go through. I have found a ppt on this (and possible solutions to avoid any risk from a competition law point-of-view)
Click to access slides-cormac-odaly-information-exchange-through-competitor-contacts.pdf
Still, I think it would be tough for business players e.g. to be obliged to prove that their purpose was a future (but evenually failed!) transaction and not a restrictive agreement.
Gabor
24 September 2012 at 3:22 pm