Brussels I Review – Illmer and Steinbruck on the Interface Between Brussels I and Arbitration


Martin Illmer and Ben Steinbrück are research fellows at the Max Planck Institute for Comparative and International Private Law, Hamburg. They have both published in the area of international arbitration (including their Ph.D. theses).

In our brief discussion of the interface between Regulation (EC) No 44/2001 (Brussels I) and arbitration we will focus on the proposals in the Heidelberg Report to include a new Art. 22(6) and a new Art. 27A.

Exclusive Jurisdiction for State Court Support (Art. 22(6))

1. The suggestion that exclusive jurisdiction for state court proceedings in support of arbitration be granted to the courts of the place (or seat) of the arbitration triggers problems in several areas.

2. An exclusive jurisdiction rule is only appropriate for a limited number of supportive measures, such as the appointment of an arbitrator. In this case, support by one single court is usually sufficient in order to set up the arbitral tribunal. Indeed, any other jurisdictional regime could lead to parallel ancillary proceedings that might produce conflicting decisions. The courts at the arbitral seat are well suited to assist in the establishment of the tribunal at the beginning of the arbitration since in most cases the lex arbitri, governing the arbitral proceedings, will be the law of the arbitral seat. Thus, the appointment procedure will usually fulfil the requirements set out by Art. V(1)(d) of the New York Convention. It follows that, at least in this respect, the future enforcement of the arbitral award is guaranteed.

3. It appears that most national arbitration laws in the EU provide for this kind of state court support. Thus, a party to an arbitration agreement will usually find its juge d’appui at the seat of the arbitration if the opponent is refusing to cooperate in the establishment of the tribunal. Hence there is no need for a harmonised mandatory rule to this effect in the Brussels I Regulation.

4. An exclusive jurisdiction regime will also lead to major problems regarding other supportive measures. The most serious consequences concern the arbitral tribunal’s establishment of the facts and the taking of evidence. State court support in this field has to be granted in the state where the evidence is located. In international disputes this state is usually not the state where the seat of the arbitration is located. Parties tend to choose a neutral place in a third state as the arbitral seat. The crucial evidence is often located in their home countries. If the courts at the seat of the arbitration were to have exclusive jurisdiction to assist the tribunal in the taking of evidence, the parties would not be able to directly request judicial assistance in the state where the evidence is located. They would have to apply to the courts at the seat to issue an official request for cross-border judicial assistance. Even under the Evidence Regulation such a procedure is burdensome and time-consuming. Consequently, it is practically never used in international arbitration.

5. Being sensitive to the problem some national legislators have enacted rules that provide for cross-border court assistance in the taking of evidence. English, German and Austrian arbitration laws, to mention a few, explicitly enable their national courts to support the taking of evidence in aid of foreign arbitrations. These provisions are widely praised as promoting the efficiency of the arbitral process.

6. Other national arbitration laws should therefore adopt similar rules rather than being subjected to an out-dated regime of exclusive court jurisdiction that flies in the face of modern arbitration practice.

7. It seems that the proposed new Art. 22(6) would not affect the state courts’ power to grant interim relief in relation to foreign arbitration proceedings. The need for cross-border interim measures is self-evident in international disputes. When a party is about to dissipate its assets or to create a fait accompli, a state judge will often be the only authority to grant effective relief to the other party. In most cases, these assets will not be located in the state of the arbitral seat but in other jurisdictions.

8. However, the existing case law in this field suggests that some state courts might consider applications for interim relief as “ancillary proceedings concerned with the support of arbitration” within the meaning of Art. 22(6) and thus refuse to grant interim measures to parties to a foreign arbitration. Even in jurisdictions that provide explicitly for cross-border interim relief in arbitration, courts have held that only the courts at the seat of the arbitration were competent to order these measures (OLG Nürnberg, (2005) 3 German Arbitration Journal (SchiedsVZ) 50). These decisions confuse a “neutral” arbitral seat with an “exclusive” forum for ancillary proceedings in support of the arbitral process. There is a serious threat that an enactment of the proposed Art. 22(6) would increase the number of such misconceived decisions.

9. The European Commission should therefore refrain from enacting an exclusive jurisdiction rule for supportive state court measures as proposed in the Heidelberg Report. By effectively ruling out cross-border judicial assistance, an exclusive jurisdiction rule in this field would be contrary to the interests of international arbitration (for a detailed analysis of the topic see Steinbrück, Die Unterstützung ausländischer Schiedsverfahren durch staatliche Gerichte, Mohr Siebeck, forthcoming in July 2009).

Determination of the validity of the arbitration agreement (Art. 27A)

10. We generally support the proposal to include a new Art. 27A that would provide for a mandatory stay of proceedings on the merits before a Member State court once a court in the Member State at the place (or seat) of arbitration is seized for declaratory relief in respect of the existence, validity or scope of the arbitration agreement.

11. If the issue of the existence, validity or scope of the arbitration agreement arises in parallel proceedings, a mechanism for allocating jurisdiction is required. The issue does not call for the exclusive jurisdiction of one court ab initio but once parallel proceedings arise, one court has to be exclusively competent to decide the issue with res iudicata effect upon any other Member State court. Otherwise there would be no legal certainty for the parties to the alleged arbitration agreement from the very beginning of their dispute up until the enforcement stage. Contradicting decisions would be inevitable – a highly undesirable result.

12. The Heidelberg Report suggests that the courts at the place (i.e. seat) of the arbitration take precedence over the court first seized with binding force upon other Member States’ courts achieved by way of recognition of the declaratory judgment pursuant to Art. 32 of the Regulation.

13. In our view this mechanism is superior to the other two possibilities for the allocation of jurisdiction: neither a lis pendens rule giving priority to the foreign court seized in breach of the arbitration agreement nor the French doctrine of the negative effect of Kompetenz-Kompetenz is as effective in protecting the parties’ interest in an early binding decision on the existence, validity or scope arbitration agreement.

14. If the foreign court seized in breach of the arbitration agreement were to determine the issue (other courts being barred by the lis pendens-rule of Art. 27(1) of the Regulation), there would be no remedy against torpedo proceedings. After the ECJ has now put an end to practice of anti-suit injunctions in West Tankers if the foreign court seized is a Member State court, the threat of torpedo actions requires a solution.

15. If the arbitral tribunal were to determine the issue (barring any decision on the matter by a state court), the risk of an unenforceable arbitral award is imminent. If the arbitral award is to be enforced in another country, Art. V(1)(a) of the New York Convention provides for non-recognition if the court determining recognition regards the arbitration agreement as non-existent, invalid or as not covering the dispute in question. In the end, it will always be a state court that will have the final say on the existence, validity or scope of the arbitration agreement. Only the moment in time of such final say differs.

16. If the state court’s final say is limited to the recognition phase, considerable time and money may have been wasted by the parties in obtaining a practically unenforceable award. Cross-border enforcement requires recognition, such recognition is only available through a state court and the New York Convention empowers the state court to rule on the existence, validity and scope of the arbitration agreement. Arbitration is not a purely transnational process, somehow detached from national laws. At the enforcement stage at the latest, the state courts enter the field.

17. If in contrast, the state court renders a decision on the existence, validity or scope of the arbitration agreement even before the arbitral process was initiated, legal certainty and procedural economy are fostered. State court intervention is indispensable in the West Tankers scenario – the earlier, the more convenient, faster and cheaper it is for the parties.

18. If the courts at the place of arbitration were to determine the issue exclusively (once seized for declaratory relief) and if this court’s decision was to be recognized by the courts of the other Member States under the Regulation’s scheme of recognition, as it is suggested by the Heidelberg Report, the torpedo scenario would be addressed very practically and the difficulties and inconvenience of the French doctrine of the negative effect of Kompetenz-Kompetenz would also be avoided.

19. The advantages of the declaratory relief mechanism are numerous: (i) The court first seized in breach of the arbitration agreement has to stay its proceedings (according to the proposed Art. 27A in order to ensure exclusive jurisdiction of the courts at the arbitral seat) so that there is no risk of contradicting decisions. (ii) It is widely accepted internationally that the courts at the seat of the arbitration are the natural forum for supervisory jurisdiction (in contrast to supportive jurisdiction, see under I). (iii) The parties achieve legal certainty at an early stage saving time and costs. (iv) The application will usually be dealt with much faster than an application to set aside the arbitral award afterwards which will often include other grounds for non-recognition prolonging the setting aside proceedings. (v) Excluding an appeal against the state court decision might even speed up the process. (vi) If the proceedings before the foreign court first seized were not initiated as a torpedo in bad faith, this court would still be competent to determine the existence, validity and scope of the arbitration agreement. This is because the scenario of parallel proceedings is unlikely to arise. The other party will usually not seise another court for declaratory relief since it can rely on the foreign court first seized to determine the issue in a reasonable time and with due care. Therefore, he will rather invoke the defence of the existing arbitration agreement and plead its validity before the foreign court.

20. Approving the suggested solution of the Heidelberg Report one should stress the following point: the proposed Art. 27A does not interfere with the national arbitration laws regarding the power of the national courts to grant declaratory relief. It merely provides for an exclusive jurisdiction if the national law chooses to grant such power and gives binding force to the declaratory judgment. It is entirely and without caveat up to the Member States to determine whether they want to empower their courts to grant such declaratory relief or not (available in England and Germany, not available in France or Austria). This solution respects different systems and peculiarities of the national arbitration laws. In English law, for example, the application to the state court for a preliminary determination of the tribunal’s jurisdiction depends on the permission by the other party or the tribunal (sec. 32 Arbitration Act 1996). German law, in contrast, does not provide for such a (sensible) restriction. Leaving the autonomy of national procedural laws and arbitration laws untouched it enables a competition for the best place of arbitration by means which appear to be more in line with most Member States’ laws and the Regulation itself than anti-suit injunctions.

The arbitration exception in Art. 1(2)(d) – keep it or delete it?

21. A final, brief remark on the proposed deletion of the arbitration exception in Art. 1(2)(d) by the Heidelberg Report: many commentators on the Heidelberg Report have so far rejected the proposed deletion of the arbitration exception. They mainly go with the adage “If it ain’t broke, don’t fix it” and fear problems of unintended consequences. However, as indicated above, the system is broken with regard to the issue of parallel proceedings, in particular the West Tankers scenario. Anti-suit injunctions are no longer available; torpedo proceedings are easy to initiate for an obstructing party. Against this background active steps to remedy the situation are required. The solution proposed by the Heidelberg Report in Art. 27A with the duty to recognise a declaratory judgment by the courts at the arbitral seat is such an active step (which we endorse). Moreover, no one has come up with a better solution so far.

22. Including a new Art. 27A does, however, require opening up the arbitration exception at least to some extent. It appears possible to open only one slot in the arbitration exception with regard to the particular problems identified after five years of operation of Regulation (EC) No 44/2001 while leaving the arbitration exception as such untouched. Taking up the initially mentioned adage, we would suggest to fix it only to the extent it is broken.