(This post was written by Jacco Bomhoff of Leiden University on his Comparative Law Blog, and is reproduced here with his permission.)
It’s official; dozens of private international law commentators, including such luminaries as professors Briggs (UK), Gaudemet-Tallon (France) and Geimer (Germany), have for years completely misread the ECJ. At least, that is what the Court’s Third Chamber suggests in last week’s ruling in Case C-98/06, Freeport/Arnoldsson. According to the new judgment, when the Court said, in its classic Brussels Convention decision in Réunion Européenne and others that:
two claims in one action for compensation, directed against different defendants and based in one instance on contractual liability and in the other on liability in tort or delict cannot be regarded as connected,
it didn’t actually mean that,
two claims in one action for compensation, directed against different defendants and based in one instance on contractual liability and in the other on liability in tort or delict cannot be regarded as connected.
Right. Of course. So, what is really going on?
The heart of the controversy is a single paragraph in the ECJ’s 1998 judgment Réunion Européenne and others. Although the questions referred to the ECJ by the French Cour de cassation in that case did, in fact, only concern articles 5(1) and 5(3), the ECJ, almost in passing, offered a sweeping statement on art. 6(1) of the (then) Brussels Convention on jurisdiction over multiple defendants at the domicile of one of them. The Cour de cassation’s reference did not touch upon art. 6(1), probably because the court was keenly aware of the fact that as the relevant proceedings were not brought in the court of the domicile of one of the defendants, that article could never apply. The Cour de cassation did, however, want to ask the ECJ more generally to rethink its narrow conception of when a single court could take jurisdiction over several related claims, in particular as French private international law allowed joinder of claims in many more cases. ‘We know’, the French court seems to say, ‘of the strict Convention requirements for jurisdiction over multiple defendants when cases are merely related, but could you allow an exception for cases where, quote: “the dispute is indivisible, rather than merely displaying a connection?”
The ECJ began by pithily remarking that “the Convention does not use the term `indivisible’ in relation to disputes but only the term `related'” (par. 38). The Court went on to refer to art. 6(1) as one of the articles that allow defendants to be sued in the courts of another Contracting state than the one in which they are domiciled. This article could not apply because the proceedings in question had not been brought before the courts for the place where one of the defendants was domiciled (par. 44-45). The acknowledged inapplicability of art. 6(1), however, did not stand in the way of the following general statement on the provision:
“48 (…) the Court held in Kalfelis that, for Article 6(1) of the Convention to apply there must exist between the various actions brought by the same plaintiff against different defendants a connection of such a kind that it is expedient to determine the actions together in order to avoid the risk of irreconcilable judgments resulting from separate proceedings.
49 In that connection, the Court also held in Kalfelis that a court which has jurisdiction under Article 5(3) of the Convention over an action in so far as it is based on tort or delict does not have jurisdiction over that action in so far as it is not so based.
50 It follows that two claims in one action for compensation, directed against different defendants and based in one instance on contractual liability and in the other on liability in tort or delict cannot be regarded as connected.”
The ruling in Réunion was condemned almost immediately and virtually universally. Briggs and Rees labeled the decision as “extraordinary and, one is driven to conclude, simply wrong” (Civil Jurisdiction and Judgments 2002, 175) and Gaudemet-Tallon called the Court’s conclusion “trop catégorique” (Rev. crit. Dr. int. priv. 1999, 339). Courts in different Member States took divergent approaches to the unwelcome statement in Réunion. The English Court of Appeal, for example, in Brian Watson v. First Choice Holidays (25 june 2001,  I.L.Pr. 1) said:
“It seems to us that, although paragraph 50 of Réunion Européenne is undoubtedly clear, the full implications of the position there set out may possibly not have been considered by the Court”.
The Court of Appeal did ultimately refer a question on Réunion’s paragraph 50 to the ECJ, but that reference was withdrawn. In other cases, courts took creative courses of action such as characterizing claims according to national law (rather than according to autonomous European standards, as usually required) (see English High Court, Andrew Weir Shipping v. Wartsila UK and Another, 11 june 2004,  2 Lloyd’s Rep. 377). Other courts, such as the French Cour de cassation ignored Réunion completely (Société Kalenborn Kalprotect v. Société Vicat and others, ). During all of this, only the Irish High Court, as far as I’m aware, at one point explicitly indicated that there was no suggestion that the ECJ in Réunion had had the “radical intention” of laying down a broad principle (Daly v. Irish Group Travel, 16 May 2003,  I.L.Pr. 38). And now we have Freeport/Arnoldsson:
“43 As the Commission has rightly pointed out, that judgment [Réunion] has a factual and legal context different from that of the dispute in the present main proceedings. Firstly, it was the application of Article 5(1) and (3) of the Brussels Convention which was at issue in that judgment and not that of Article 6(1) of the Convention.
44 Secondly, that judgment, unlike the present case, concerned overlapping special jurisdiction based on Article 5(3) of the Brussels Convention to hear an action in tort or delict and special jurisdiction to hear an action based in contract, on the ground that there was a connection between the two actions. In other words, the judgment in Réunion Européenne and Others relates to an action brought before a court in a Member State where none of the defendants to the main proceedings was domiciled, whereas in the present case the action was brought, in application of Article 6(1) of Regulation No 44/2001, before the court for the place where one of the defendants in the main proceedings has its head office.
45 It was in the context of Article 5(3) of the Brussels Convention that the Court of Justice was able to conclude that two claims in one action, directed against different defendants and based in one instance on contractual liability and in the other on liability in tort or delict cannot be regarded as connected (Réunion Européenne and Others, paragraph 50).
47 Having regard to the foregoing considerations, the answer to the first question must be that Article 6(1) of Regulation No 44/2001 is to be interpreted as meaning that the fact that claims brought against a number of defendants have different legal bases does not preclude application of that provision.”
I can only say, with all due respect: if you say so. Because this reading of Réunion seems to me, again with all due respect, fairly implausible. As to the substance, the clarification/reversal of the infamous paragraph 50 is, on the whole, to be welcomed. But Freeport/Arnoldsson does create new questions and leaves meany old ones still unanswered. If the contract/delict divide is abandoned (at least as a rigid rule), it would seem to follow that national courts will have significantly more leeway when assessing possible jurisdiction over multiple defendants, based on art. 6(1). This discretion seems all the more considerable given that the Court, elsewhere in its new judgment, rejects a basic notion of ‘abuse’. This would seem to mean that a claim against a defendant potentially liable for 99% of all damages at the domicile of a co-defendant potentially liable for the remaining 1% will be allowed under the Brussels Regulation. It seems likely that the Court will, over the coming years, have to revisit this vexed issue.