10 April 2008

Parallel trade in pharmaceuticals and the quixotic search for meaning in Article 82

On April Fools' Day Advocate General Colomer delivered his Opinion in yet another variation of the GlaxoSmithKline parallel trade saga (Joined Cases C-486/06-C-470/06). The English translation is not yet out, I have had a go at translating a few key passages below. You can get other language versions at the European Courts' website.

This time GSK's refusal to provide Greek wholesalers with sufficient medicines to allow them to export these abroad to Member States where the price is higher was challenged in the Greek courts as an abuse of a dominant position. This dispute had been aired before in Case C-53/03 when the Greek competition authority referred a number of question to the ECJ on the application of Article 82. The ECJ then refused to answer them ruling that the Greek competition authority was not entitled to make a reference. However AG Jacobs had delivered a strong Opinion on the merits of the issues, and it is the merits that AG Colomer also addresses. It will be useful to compare the approaches of the two Advocates General.

1. Refusal to supply as an abuse (paras 39-46)
The AG rightly notes that the closest precedents are two old cases, Commercial Solvents and United Brands: (a) like the facts at hand, there was a refusal to continue to supply (not a refusal ab inito), (b) like in Commercial Solvents whose termination was followed by its entry in the downstream market, GSK had reorganised its sales through its exclusive distributor in Greece; (c) United Brands set out a principle that a dominant firm cannot terminate a contract with a long-standing client whose orders are not abnormal. Drawing these strands together at para 46: "... we can infer that an undertaking holding a dominant position that refuses to supply goods - especially in the absence of alternative products..., reserving to itself the market for parallel imports adops a form of behaviour that is an abuse under Article 82."

This passage is closer in spirit to Commercial Solvents than United Brands. This is to be welcomed because the ratio of United Brands is more akin to a contract doctrine than a competition law rule. The AG instead notes that the refusal to continue to supply has an adverse effect on the market for parallel trade. (We can debate whether this is an adverse effect but at least he looks for one...). Note also that the passage requires that the dominance held is significant - the wholesaler must be unable to find a substitute. This is different from United Brands where Olesen could easily find alternative suppliers of bananas. So market definition is crucial.

2. Intention and abuse (paras 47-54)
This is a strange part of the Opinion. The AG starts by noting that abuse is an objective notion but this is subject to two important aspects: (1) subjective elements can indicate that the dominant firm is pursuing ends contrary to competition; (2) (with reference to British Airways)since Article 82 does not contain any reference to the anticompetitive effect of the practices, it is sufficient to show that the abusive conduct tends to restrict competition, or in other words can have anticompetitive effects. I don't see how point (2) has to do with intent; nevertheless the AG continues to say that GSK never denied that its real objective was to eliminate parallel exports and (paras 53-54): "It is obvious that the intetion of GSK is in contrast with the obectives of the Treaty, because it affects the freedom of trade among Member States in a manner that can damage the internal market, according to the case law and Article 3(1)(g) EC, because it indisputably isolates national markets and modifies the structure of competition in the common market. In sum, these observations show that GSK has committed a grave violation of the Treaty that would deserve the label of an abuse 'in itself and per se' given that the basis of the behaviour has no other economic foundation than the intetion of elimitaing parallel trade from its competitors, the Greek wholesalers."

This analysis is unnecessary but the AG is obviously keen to participate in the debate on the reform of Article 82, so he has a go here at doing two things: first it seems he tries to find a role for intent to reconcile the case law; and this is to be welcomed, esp in that the observation about 'gravity' might suggest that intent is relevant also to the fine imposed. But second he seems to be crafting a theory of abuse akin to the 'no economic sense' test by which conduct is an abuse if it is not rational for the defendant absent a tendency to eliminate competition. (See
Gregory J. Werden 'Identifying Exclusionary Conduct Under Section 2: The "No Economic Sense" Test' (2006) 73 Antitrust Law Journal 413). I don't know if he intended to do this or not, but if he did, then it seems to run against the Commission's view in the Discussion Paper on Article 82 that seemed to apply an 'as efficient competitor' test to determine abuse (at least for predatory pricing, see para 63).
Also interesting how the AG construes the market relations: he treats GSK and the wholesalers as competitors in the market for selling pharmaceuticals to health care providers. So the abuse harms intra-brand competition. This is obviously a concern when there is an absence of inter-brand competition. But this has nothing to do with intent, instead it has to do with effects. Would it not be better to ground abuse on anticipated effects than upon intent?

3. No per se abuse (paras 62-77)
First the AG notes that the case law seemed to have classified three abuses as per se unlwful (exclusive dealing, loyalty rebates and predatory pricing under the AKZO rule), but that the more recent case law (British Airways) suggests that firms may offer justifications even for what may be per se abuses (loyalty rebates in that case). So the question of whether any practice is abusive per se is open. He then quickly shuts the door on any suggestion that we should develop a category of per se abuses, on legal and economic grounds.
Legal grounds: the absence of a provision for exemption like Art 81.3 meants that (para 67): "the analysis of abusive behaviour reqires a dialectical debate between the dominant undertakings and the competent authorities, whether national or EC, and also interested parties. ... (para 69): "In this scenario if certain behaviour were always togive rise to an irrebuttable presumption of abuse, dominant firms would be denied the right of defence."
Economic grounds: (1) per se abuses would be contrary to the need to examine each case on teh basis of its economic context; (2) the approach would commit the 'sin' of 'excessive formalism' (para 72). There is a third ground at para 74-5 which is somewhat opaque and possibly circular and seems to say that because there are two types of abuse (exploitative and exclusionary) and there is no hierarchy, the defence for a dominant firm must be based on the economic effects.

I disagree with this analysis. Per se abuses can be useful as a time saving device: granted we can formalistically condemn some efficient behaviour but the reason we put behaviour in a per se illegality box is because experience teaches us that normaly such behaviour is harmful and it is cheaper to condemn some procompetitive behaviour than run a full scale economic inquiry in all cases. The sound administration of justice can sometimes justify denying defendants a right to offer a justification, after all any legal system tolerates strict liability rules. A contrario we should also tolerate safe harbours, or per se legality.

4. Objective justification (paras 78-119)
This is the good bit. We start with something remarkable at para 79: "There are three justifications that can be invoked by firms in a dominant position that are accused of abuse: reasons related to the market in which they operate; legitimate defence of their commercial interestsl and proof of a positive balance." The first one is related to one suggested in the Discussion Paper (para 80 thereof) but the AG wants to widen it. This is remarkable because it is a bold, exhaustive formulation of what objective justification means which seems to be wider than the existing jursprudence.

(a) market conditions: The suggestion is that the Court should be willing to take into consideration the way in which a market is regulated (e.g. here the fact that prices are fixed by Member States and that there are statutory requirements about the quantity of medicine that should be available). The AG says this can be a plausible defence but it fails on the facts because while prices are set by the Member State, the undertakings are involved in discussions over price, and the obligation to keep minimum quantities does not explain the refusal to deal with wholesalers. (This defence seems to be a variation of the state action defence whereby if the Member State requires an infringement, then the undertaking is not responsible (the MS is under Arts 10 & 81/82)).
(b) defence of legitimate interests: This is apparently a well defined defence (para 99, exemplified by BP v Commission and United Brands) but this is subject to the principle of proportionality. I prefer to see BP as a case of force majeure, see Ekaterina Rousseva's analysis. This line of cases is applied to the following argument: that parallel trade reduces the profits of pharmaceutical companies and this does not allow them to recoup the R&D expenditures. On the facts this is rejected for lack of evidence of a causal link between parallel trade and the damage to the firm's profits. Also it is suggested that GSK could have chosen a different method of distribution, full vertical integration, which would then perhaps not have constituted an abuse. Further, it is argued that this line of argument is more about the incentive effects (that is, parallel trade may dent the incentives to invest in future medicines), but this is rebutted by stating that the EU offers a friendly environment for firms and R&D...
(c) positive economic balance: this is an efficiency defence, and GSK raises the argument that parallel trade does not benefit patients, nor the health service providers who buy medicines, but only serves to benefit the wholesaler. The response is amusing (para 118): "Aside from the description of the 'horrors' provoked by parallel trade, GSK does not indicate any positive aspect resulting from its decision to limit supplies to wholesalers save the recovery of its profit margins, something which is irrelevant." And so this fails for lackof evidence.

Brief comments: first it is nice to see a willingness to develop the law, especially on defences but also (see above parts 2 and 3) on theories of abuse the AG seems to be happy to give dominant firms a chance to prove they are not guilty but the scepticism over the arguments of GSK is a matter of concern, perhaps the national court will have a different view of the evidence. However the concern si that there are defences in theory but not in practice. Second it is unfortunate that the AG does not engage directly with the contrasting Opinion of AG Jacobs (who has slightly different views on the role of intention, a more subtle way of integrating defences (see para 72), and diametrically opposite views on the merits of the justifications proffered by GSK); third I think categories (b) and (c) of the defences are overlapping, it is best to treatthe cases in (b) as instances where the dominant undertaking is faced with no choice but to terminate, so leaving all the economic cost-benefit arguments for the efficiency defence. Moreover I think if you look at the three defences separately you miss the point of GSK's argument - It is a combination of the arguments in the three categories that gives rise to the defence, AG Jacobs' Opinion had recognised this. Accordingly I feel that the analysis of defences is flawed, and the Opinion does not solve the mysteries of Article 82.


Anonymous said...

On the theory of abuse, you could argue that the AG is actually following the Discussion Paper. This is because according to the Commission there a refusal to continue to supply makes no economic sense because it must have been 'efficient' to supply in the first place. This is akin to the Aspen Skiing case as interpreted in Trinko. so while the as efficient competitor test might be used in some abuses, for contract termination the Discussion Paper uses the no economic sense test.

Michael said...

Hi - EU law blog posted on a recent Appellate court judgment involving Glaxo. The general link to the blog is ---

Interesting - is the Paris court just out of it, or is the commission approach here posing problems for developing more coherent standards?