Several recent Commission cases and Court judgments suggest a new way of looking at some abuse cases – as a form of estoppel.
The Court of Justice’s preliminary ruling in Telia Sonera, holding that a margin squeeze can be abusive even absent a duty to supply, has come in for some heavy criticism. If there is no antitrust obligation to supply, critics argue, how can the terms of supply be abusive? Telia Sonera contrasts markedly with the position of the US Supreme Court in LinkLine (PDF) which holds that margin squeeze is not a self standing abuse under the Sherman Act, though the minority opinion of Breyer does leave open the possibility that a margin squeeze might be unlawful if there was an obligation to supply.
The criticism that if there is no antitrust obligation to supply, how can the terms of supply be abusive is a fair point, but there’s another way of looking at the case: a dominant company may not have an obligation to supply, but if it unilaterally chooses to supply, it must supply on terms at which it at least could compete (let’s assume that the relevant test for margin squeeze is the “as efficient competitor” test).
Why might the voluntary supply make a difference relevant to the competition rules? Perhaps because the purchaser makes commercial decisions based on the offer to supply. If the purchaser is faced with a make or buy decision, for example, then its incentives are obviously different depending on whether the buy option is or is not available.
Perhaps the dominant company was carrying out a bait and switch strategy – enticing purchasers to buy its wholesale product, rather than build a competing wholesale product, but then tweaking the terms of supply so that purchasers are squeezed.
It’s arguable that there’s Court of Justice precedent going against this reading. In the Bronner refusal to supply preliminary ruling, the Court was asked first under the circumstances of the case, could there be an obligation on a dominant company to supply, and second whether conditioning supply on certain terms and conditions could be regarded as tying or bundling.
The Court replied to the first that there could be no obligation to supply, then continued:
48 In its second question, the national court asks whether the refusal by that undertaking, in the circumstances mentioned in the first question, to allow the publisher of a rival daily newspaper to have access to its home-delivery scheme where the latter does not at the same time entrust to it the carrying out of other services, such as sale in kiosks and printing, constitutes an abuse of a dominant position within the meaning of Article 86 of the Treaty.
49 Given the reply to the first question, there is no need to answer the second.
If this should be read as saying, “if there’s no obligation to supply, then there’s no obligation to supply in a particular way” then it seems to contradict Telia Sonera. (Bronner wasn’t cited in the case.) But I think you can make a reasonable argument that that aspect of Bronner is at best per curiam.
There are other cases that could also fit well into a form of estoppel analysis: the Commission has pursued a number of cases in relation to standard essential patents.
First, in IPCom – a case similar to the US Federal Trade Commission’s NData Section 5 case – the Commission press release welcomed IPCom’s declaration that it would take over the FRAND commitments made by Bosch, the previous patent owner. The press release noted that, “The transfer of FRAND commitments after the sale of standard-essential patents is important from a competition law perspective.”
Second, in Rambus, the Commission concluded an Article 9 decision, accepting commitments from Rambus (PDF link) on the pricing of certain patents included in a JEDEC standard. The Commission was concerned that Rambus had carried out a “patent ambush” in the standardisation process, which resulted in its patented technology being included in a standard without the knowledge of the participants in the standard setting process.
Third, the Commission has sent Statements of Objections to Samsung and Motorola expressing concerns that the companies were seeking injunctions against potential licensees’ use of their standard-essential patents notwithstanding having given commitments to license those patents on FRAND terms.
In the IPCom, Samsung and Motorola cases, there is no allegation that the companies concerned were obliged to enter into the standardisation process. They – at least on the facts set out – were free to withhold their IP from the standards, in which case they would presumably have been free not to license on FRAND terms or any other (unless they fell within the exceptional circumstances test).
But one reading of the IPCom press release and the press release on the Statements of Objections is that when they did commit their IP to the standardisation process, and did engage in FRAND commitments and they are dominant, then failure to comply with the FRAND commitments they gave may be potentially abusive.
Rambus is similar: the Article 102 concern arises because of third party reliance on their behaviour (the alleged ambush and the inclusion – as it turned out – of the IP in the standard).
This might also suggest that a withdrawal of supply case – such as Commercial Solvents – may, depending on the circumstances, be regarded differently to refusal to supply de novo, an argument consistent with Justice Scalia’s comments on Aspen Skiiing in the US Supreme Court case of Trinko. (One of the few aspects of the ruling I agree with.)
As a first comment the quote from Bronner was raised before the ECJ as a basis for asserting that the ECJ had in fact ruled on the issue allready – i.e. going down the Linkline route well before the US supreme court. This reading was however dismissed by the ECJ as a missunderstanding.
As a second comment, the big danger with the prelimnary ruling in the TeliaSonera case is not that it leaves an opening that a margin squeeze may – in certain circumstances – constitute an abuse (bait and (sudden) switch strategies a possible example).
The danger is that the ECJ’s judgment is interpreted too widely, i.e. that every instance where the margin between an input and a derivative product is “too low” could be held to constitute an abuse.
It is clearly bad for competition and economic development if undertakings that believe that they may be deemed dominant decide not to supply competitors where supply makes economic sense both for themselves and the competitors at issue.
The TeliaSonera ruling makes it much more likely that possibly dominant undertakings use refusal to supply as their default reaction whenever they feel uncertainty as to the correct LRAIC costs in its retail operation or the need to maintain freedom to compete even below this level of cost in the retail market for the derivative product.
The possibility to use competition law to condemn certain activities that may be pro competitive always comes at a price.
Great post! but the only possible abuse would be the dominant MODIFIYING its terms, i.e. a hold-up. but then, it requires high switching costs for the counterparty, i.e, specific investments that can be expropriated by the dominant. Otherwise, the counterparty could terminate the contract with the dominant.
In any case, the counterparty could protect himself through a long-term contract in order to forbear the deployment of its own network