ECJ Refuses to Extend the Scope of Article 5 (3) Brussels I to Coperpetrator


Vincent Richard is a Research Fellow at the Max Planck Institute Luxembourg.

On May 16th, the Court of Justice of the European Union rendered its judgment in Melzer v. MF Global UK ltd (C-228/11) in which the judges refused the extension of  the scope of article 5 (3) suggested by the Landgericht Düsseldorf.

A German individual residing in Berlin was solicited by telephone by a German company (WWH) based in Düsseldorf which opened an account for him in an English brokerage company (MF Global UK) trading in futures in return for remuneration. The investment did not go as planned; the German client lost almost all of his initial investment and decided to go to Court in order to obtain compensation for his loss.

Oddly enough, the plaintiff decided to sue only the English company in Düsseldorf and to base his claim on tortious liability. Thus, the Court in Düsseldorf needed to assess its jurisdiction in regard to article 5 (3) of Brussels I. In this case, the German court considered that the damage occurred in Berlin where the plaintiff had his assets and that the harmful events occurred in London where the English company conducted its business, and in Düsseldorf where the German company is based. But as the German company was not a party to the litigation, the court explored whether it could apply the national principle of “reciprocal attribution of the place where the event occurred”.

This principle, as understood by the CJEU, is derived from provisions of the German Civil Code (§830) and the German Code of Civil Procedure (§32). It allows a Court to retain jurisdiction insofar as it is the place where the event giving rise to the damage has been caused by a presumed joint participant or accomplice, even though this accomplice is not himself a defendant.

Unsurprisingly, the CJEU answered negatively to the question asked by the German Court and held that as an exception to article 2, article 5 (3) has to be interpreted restrictively. In the present case, it found that there was no connecting factor between the English defendant and the Court of Düsseldorf. Moreover, the CJEU ruled that the use of national legal concepts to interpret Brussels I regulation would lead to different outcomes among the Member States and thus be contrary to the objective of legal certainty.    

Finally, the Court mentioned that several others possibilities could have been used by the plaintiff who could have based his claim on contractual liability or could have sued both companies in Düsseldorf under article 6(1) of the Regulation.


Article 5(3) of Council Regulation (EC) No 44/2001 of 22 December 2000 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters 2001 must be interpreted as meaning that it does not allow jurisdiction to be established on the ground of a harmful event imputed to one of the presumed perpetrators of damage, who is not a party to the dispute, over another presumed perpetrator of that damage who has not acted within the jurisdiction of the court seised.

5 replies
  1. Jan von Hein says:

    I agree that the decision of the CJEU is not much of a surprise, but I would rather submit that it nonetheless sets a highly problematic precedent. The idea that a coperpetrator should be held responsible where his other “partners in crime” have actually committed the deed is not particular to German procedural law, as was evidenced by the opinions of the national governments in this case, which all sided with the LG Düsseldorf. Tortfeasors who share a common plan or persons who aid and abet others actually committing the tort in question cannot really claim to be unfairly surprised when they are sued in a place where the immediate tortfeasor has acted with their consent. Moreover, from the point of the plaintiff, there may be legitimate procedural reasons for not suing all possible defendants at the same time, especially in jurisdictions such as Germany where a party to a proceeding cannot be summoned as an ordinary witness. The judgment of the CJEU is a serious setback in the civil combat against internationally organized crime, not only in the financial sector. Just imagine the following fictitious case: A German and a Danish terrorist conspire to plant a bomb in the plane of a third-state diplomat. Following a plan set out by his German comrade who stays in Berlin, the Danish tortfeasor installs the bomb while the plane is staying at the airport in London. The plane then blows up somewhere over third-state territory. According to the logic of the Melzer judgment, the German coperpetrator could not be sued by the diplomat’s relatives in London under Art. 5 no. 3, because he was merely a coperpetrator who did not act personally in the U.K., and the place of damage is obviously located outside the EU. Would this result make sense? With due respect, I doubt it very much.

  2. Florian Horn says:

    One important reason why this answer of the ECJ had to be expected was also the nature of the question of the German court, which directly asked if “reciprocal attribution” of the place where the event occurred was admissible. This probably set off the alarm that autonomous interpretation of the Regulation was threatened. Indeed “reciprocal attribution” is a German law concept, foreign to many other jurisdiction. Nevertheless the underlying rationale is much more common, as Jan Van Hein rigthfully stated.

    One would hope that the judgement only excludes the possibility to apply jurisdiction over one of the perpetrators directly on another where there exists no additional factual or causal link. If the judgement were to be interpretated as to restrict any possibility to initiate proceedings against joint perpetrators even if such factual link exists, this would greatly reduce the scope of application of Art 5 (3). In this event, joint proceedings could only be introduced at the seat of one of the perpetrators according to Art 6 (1) or at the place where the damage occurred according to Art 5 (3). This would be against the case law on ubiquity of jurisdiction under this provision, practically hindering joint proceedings in a country where only the event was set by one of the perpetrators .

    Indeed in most cases which might be seen as problematic under the new judgment, a factual or causal link exists between the aprty which is not directly acting and the event set by the other party. It might be a peculiar consequence of the focus of the present proceedings on the apparently very formal national provision that no further material issues were discussed.

    I believe it has to be differentiated:

    1) If all accomplices just worked besides each other, each in their own countries setting independent events and only in the end a common damage would occur in another country, this might indeed justify restriction of the common forum under Art 5 (3) to the country where the damage occurred;

    2) If however one accomplice supported the other in his act, e.g. sending documents to a certain location etc, the first accomplice may well be seen as causing also the event set by the second, consequently allowing for jurisdiction in that country;

    3) Even more so, if one accomplice acts as instigator, it may well be seen as causing the event set by the other; also in this case jurisdiction for both seems justified in the country, where the instigated person finally acted;

    4) In general, it would be hard to argue that an accomplice, which has not contributed to a specific event allowing for jurisdiction, should be subject to this specific jurisdiction.

  3. Jan von Hein says:

    Florian Horn is right in pointing out that the reference for a preliminary ruling made by the Court of Düsseldorf left much to be desired, as it was framed in a very germanic legal terminology. It is all the more surprising, however, that the CJEU did not join the Melzer case with two other, better reasoned references on the same question made by Germany’s highest court, the Federal Court of Justice (see the cases “Hi Hotel”, June 28, 2012 – I ZR 35/11 and “Parfumflakon II” – I ZR 1/11).
    With regard to the differentiating approach suggested by Florian Horn: Although one might always argue about details, this would have been the correct way forward: Developing an autonomous (!) concept of attributing responsibility for collaborative torts within the framework of Art. 5 no. 3. Unfortunately, it seems that the CJEU either shied away from this task or did not fully grasp the implications of its ruling.

  4. Matthias Weller says:

    Thank you, Vincent, Jan and Florian, for your instructive comments. I would simply like to amend these by the following considerations of mine:

    Under section 32 German Code of Civil Procedure (ZPO) – the German forum delicti commissi, the Federal Court of Justice (BGH) reciprocally attributes the acts of any co-perpetrator in order to establish jurisdiction at the place of acting. “Co-perpetrators” are defined along the lines of section 830 German Civil Code. This provision deals with the situation that two or more persons are involved in a tort. In principle, each person would only be accountable for the damage exactly caused by him/her personally. Contrary to this principle and to the benefit of the victim, section 830 Sentence 1 German Civil Code holds each person involved accountable for the entire damage if these persons acted willfully together to commit the tort. Sentence 2 deals with the situation that it is impossible to prove who of two or more involved per-sons caused the damage. In that case, sentence 2 provides for a presumption that all of the persons involved (not necessarily acting willfully together) caused the damage. In both cases, the result is that all persons involved are joint debtors to the victim. Each debtor owes the entire damage to the victim. The victim may choose against whom he wants to raise the claim. Obviously, these rules aim to help the victim on the level of substantive law.

    There is nothing in this ratio that would directly apply to the question of where should be a forum delicti commissi. Nevertheless, the reciprocal attribution of all actions of co-perpetrators on the level of jurisdiction has been established case law for decades. Obviously, the German courts wanted to be able to assume jurisdiction as soon as possible in cross-border cases, i.e. as soon as one (minor) part of the joint action takes place in Germany, in order to help the victim which in all cases were a German private capital investor domiciled in Germany. The German courts are free to follow this approach in the German autonomous procedural law vis-à-vis third states. However, the reasoning for this approach obviously cannot justify reciprocal attribution under the co-operative system of distributing jurisdiction amongst the Member States. And indeed, the BGH expressly left open the question whether this approach applies to the European fora delicti commissi (once however, the BGH did extend this approach to Article 5 no 3 Lugano Convention but did not give any reasons why this should be justified under the Treaty provision).

    Most commentators in Germany nevertheless argued in favour of extending the con-cept of reciprocal attribution to Article 5 no. 3 Brussel I-Regulation/ Convention/ Lugano Convention. Others, including myself (M. Weller, Praxis des internationalen Privat- und Verfahrensrechts 2000, pp. 202 et seq.), argued that the ECJ would and should not further extend the scope of Article 5 No. 3 Brussels I Regulation/Convention substantially beyond the alternative heads of jurisdiction at the place of the acting and the place of the harmful event because unrestricted reciprocal attribution would multiply the heads of jurisdiction to a degree contrary to the basic principle of the Regulation/Conventions that heads of special jurisdiction should be interpreted narrowly. Instead, it was suggested limiting the attribution of the acts of co-perpetrators in order to avoid jurisdiction at places where there is little connection to the central parts of the relevant actions.
    For example, if person A acting in Member State A instigates a tort to be committed by person B in Member State B, suing B in Member State A would mean to allow proceedings in a forum with hardly any connections to the central parts of the rele-vant actions in Member State B. At the same time the proceedings in Member State A would have to deal mainly with exactly these actions in Member State B in order to assess whether B is liable.
    However, if person A supports, by actions in Member State A, a person B in commit-ting a tort in Member State B, suing person A in Member State B would result in pro-ceedings at the place where the main parts of the relevant action took place. This could be justified under the jurisdictional ratio of Article 5 no 3, the former type of jurisdiction could not.

    Therefore, it appears clear to me that unrestricted reciprocal attrib-ution should not be allowed under the Regulation/Conventions. On the other hand, jurisdiction against instigators and supporters – even in stand-alone proceedings against them – at the place where the main tortfeasor acted appears to me to be preferable to fully excluding any attribution at all. In a way, granting jurisdiction against all persons willfully involved at the “centre place” of acting mirrors the case law of the ECJ on framing the place of the harmful event in C-509/09 – eDate Adver-tising. In that case, the ECJ allowed the victim of an infringement of personality rights committed on the internet to claim the entire damage at the centre of his interests.

    It remains of course the question how jurisdiction should be assumed if two tortfea-sors jointly act together by actions of more or less equal relevance, i.e. where there are two centers of acting. Given the principle that special jurisdiction should be inter-preted narrowly, I would still opt for excluding reciprocal attribution (M. Weller, op.cit.). However, in Melzer, I would have opted for granting jurisdiction at the centre of acting by the main tortfeasor at Düsseldorf, Germany.

    A follow-up question is, how co-perpetrators should be defined for jurisdictional pur-poses. It is evident that neither section 830 German Civil Code as such nor its notion can be directly transferred on the level of European procedural law because European procedural law calls for an autonomous interpretation of the place of acting for its fora delicti commissi. One might of course think of Article VI. – 4:102 DCFR on “Collaboration” as a starting point. This provision reads: “A person who participates with, instigates or materially assists another in causing legally relevant damage is to be regarded as causing that damage”. However, the underlying notion seems to be very close to section 830 German Civil Code. If this holds true, (again) nothing can be inferred from it for the question of where should be a forum delicti commissi. Therefore, intentional acting together should be the criterion for attribution. The courts do not need to go into the merits of the case in order to determine this issue. According to the theory of double-relevant facts it is enough that the plaintiff merely pleads the facts that allow, if proven in the main proceedings, attribution on the level of jurisdiction.

  5. Andrew Dickinson says:

    I share Jan’s reading of the decision as extending to the possibility of attribution generally. It appears from the summary of the case given by the CJEU (para. 14) that MFG was sued both for its own alleged wrongdoing (in England) and for assisting WWH’s alleged wrongdoing (in Düsseldorf). It appears to have been conceded (para. 16) that the only viable basis of jurisdiction was with respect to WWH’s activities in Germany, so that a claim based directly on MFG’s failure to advise from (or negligent trading in) London could not proceed in Düsseldorf. From there, the CJEU refuses to countenance application of the German attribution rule for tortfeasors and, while acknowledging the possibility of an EU interpretation based on common principle (para. 32), holds back from attempting the development of such principle in this case. Nevertheless, it is obvious that Art. 5(3) must countenance rules of attribution in some circumstances, otherwise claims against corporate defendants (acting through employees and other representatives) would not fall within the rule. The question, therefore, is where the limits lie. On its face, MFG appears to be an attempt to turn the tort upside down – to treat a tort committed in London with facilitation from Germany, as one committed in Germany with facilitation from London. That may explain the CJEU’s restrained reasoning. Jan’s terrorism example provides a truer test. If G (the German) is sued in London for his acts of assistance and facilitation in Germany, the case may be difficult to distinguish from MFG. If, however, G is sued in London because, by a principle shown common to the legal systems of the Member States, D’s acts are treated as his own (in the same way as those of an employee or agent acting within his authority) there would appear to be scope for crafting a different solution.

Comments are closed.