Tobias Lutzi, the author of this post, works at the Institute of Foreign Private and Private International Law of the University of Cologne and studies at the University of Oxford.
The ECJ’s recent decision in Gazprom (Case C-536/13) is the latest addition to a series of judgments by the Court that have considerably reduced the remedies available to claimants who seek to enforce the negative dimension of an arbitration agreement, i.e. the other party’s obligation not to initiate court proceedings. They have created a coherent framework for the protection of arbitration agreements within the EU, which has been sanctioned and complemented by the recast of the Brussels I Regulation. Yet, a number of questions still remain open – some of which are unlikely to be answered any time soon.
The current status quo
Traditionally, four types of remedies are available to parties seeking enforcement of the negative dimension of an arbitration agreement from a court. First, they may ask the court seised by the other party to stay or dismiss the proceedings. Second, they may ask another court to issue an injunction against the party in breach in order to restrain the latter from initiating or continuing litigation (so-called ‘anti-suit injunctions’). Third, they may bring an action for damages to recover the loss incurred due to the litigation. Fourth, they may apply for the foreign judgment not to be recognized and enforced.
While courts in all member states of the EU regularly dismiss or stay proceedings brought in violation of an arbitration agreement, and refuse to recognize and enforce judgments obtained in breach of such an agreement, only English courts have granted anti-suit injunctions and awarded damages for breach of an arbitration agreement in the past. Yet, as far as litigation in the courts of EU member states is concerned, all of these remedies have been affected by the harmonized regime of jurisdiction and recognition and enforcement of judgments in civil and commercial matters that has been established by the Brussels Convention and its successor regulations.
It is true, though, that regarding the first remedy, i.e. a dismissal or stay of local proceedings, there has never been much doubt that the European instruments do not require the courts of a member state to adjudicate if this would violate a valid arbitration agreement; instead, they have to send the case to arbitration, as required by Art. II(3) of the New York Convention. The ECJ’s decision in Gazprom and the first paragraph of the new recital (12) of the Brussels I Recast merely confirm that this is still the case.
Access to the second remedy, i.e. anti-suit injunctions issued by English courts to prevent a party from litigating in breach of an arbitration agreement, has however been radically restricted by the ECJ’s case law. Consistently with its reasoning in Gasser (Case C-116/02) and Turner v Grovit (Case C-259/02), the Court held in West Tankers that “even though proceedings [to enforce an arbitration agreement via an anti-suit injunction] do not come within the scope of [the Brussels I Regulation], they may nevertheless have consequences which undermine its effectiveness”, if they “prevent a court of another Member State from exercising the jurisdiction conferred on it by [the Regulation]”, which includes the decision on the jurisdictional defence based on an arbitration agreement. Accordingly, “it is incompatible with [the Regulation] for a court of a Member State to make an order to restrain a person from commencing or continuing proceedings before the courts of another Member State on the ground that such proceedings would be contrary to an arbitration agreement.”
While the new recital (12) tries to clarify the scope of the exclusion of arbitration in Art. 1(2)(d) of the Regulation, nothing in the legislative history of the Recast, which left the actual text of the regulation otherwise unchanged, suggests that it was supposed to reverse the decision of the Grand Chamber in West Tankers. Thus, it was to the surprise of many that Advocate General Wathelet, in his opinion on Gazprom, argued that “the EU legislature intended to correct the boundary which the Court [in West Tankers] had traced between the application of the Brussels I Regulation and arbitration” with the Recast. He opined that para. 2 of recital (12), which excludes decisions “as to whether or not an arbitration agreement is null and void, inoperative or incapable of being performed” from the rules on recognition and enforcement, should be understood as excluding “the verification, as an incidental question, of the validity of an arbitration agreement [entirely!] from the scope of the Brussels I Regulation”. Consequently, “the fact that the Tribunale di Siracusa [in West Tankers] had been seised of an action the subject-matter of which fell within the scope of the Brussels I Regulation would not have affected the English courts’ power to issue anti-suit injunctions in support of the arbitration because […] the verification, as an incidental question, of the validity of an arbitration agreement is excluded from the scope of that regulation.”
But as the question submitted to the ECJ concerned the pre-recast regulation (No. 44/2001), the Court – while implicitly rejecting the Advocate General’s proposition that recital (12) “in the manner of a retroactive interpretative law, explains how that exclusion must be and always should have been interpreted” – did not need to (and did not) discuss this proposition; instead, the Court simply distinguished the present question of recognition and enforcement of “an arbitral award prohibiting a party from bringing certain claims before a court of that Member State from the question of the court issuing itself “an injunction […] requiring a party to arbitration proceedings not to continue proceedings before a court of another Member State”, only the latter type of injunction being “contrary to the general principle which emerges from the case-law of the Court that every court seised itself determines, under the applicable rules, whether it has jurisdiction to resolve the dispute before it”. Yet, the fact that the Court deemed such a distinction necessary and referred repeatedly to its decision in West Tankers may be seen as an indication that it does not consider this decision to be already overruled by the Recast.
Against this background, it certainly is surprising that the third remedy, i.e. damages for the breach of an arbitration agreement, has yet to be subject to a decision of the ECJ – and has neither been affected by any paragraph of the new recital (12). As English courts may no longer issue anti-suit injunctions – a remedy expressly admitted to prevent that “the plaintiff will be deprived of its contractual rights in a situation in which damages are manifestly an inadequate remedy” (Lord Millett in The Angelic Grace  1 Lloyd’s Rep 87) – it seems very likely that damage awards will become much more prevalent in English courts. They have thus been allowed by the High Court after the ECJ’s decision in West Tankers ( EWHC 854 (Comm)) and awarded by the Court of Appeal in The Alexandros T  EWCA Civ 1010.
Regarding the fourth remedy, i.e. the refusal to recognize and enforce a judgment obtained in breach of an arbitration agreement, recital (12) now provides a clear solution, which seems to limit the ECJ’s decision in Gothaer (Case C-456/11) and to reverse recent English case law (cf The Wadi Sudr  EWCA Civ 1397). According to its paras 2 and 3, decisions as to the validity of an arbitration agreement are excluded from the provisions on recognition and enforcement, while decisions as to the substance of the dispute are subject to these provisions unless this would require a member state to violate its obligations (i.e. to enforce a valid arbitral award) under the New York Convention. This is not only a welcome step towards the legal certainty that the difficult relationship between the Regulation and the Convention indubitably requires but should also be understood as an attempt to counter-balance the absence of anti-suit injunctions within the Brussels I framework.
The case law of the ECJ and recital (12) of the Recast seem to provide a coherent and workable framework for the protection of arbitration agreements; they put a strong emphasis on the principle of mutual trust between the member states, but balance it out with their obligations under the New York Convention. Still, some questions remain open.
First, and foremost, the ECJ has held in Gazprom that the Regulation does not preclude the courts of a member state “from recognising and enforcing […] an arbitral award prohibiting a party from bringing certain claims before a court of that Member State”. But does the same apply to an arbitral anti-suit injunction restricting proceedings before a court of another member state? Several of the Court’s arguments – which are all carefully limited to the question of recognition and enforcement in the state where the relevant proceedings are brought – indicate that this might not be the case: while enforcing an arbitral award by ordering a party to stop or limit local proceedings raises “no question of an […] interference of a court of one Member State in the jurisdiction of the court of another Member State”, enforcing an award by ordering a party to stop or limit proceedings elsewhere might indeed amount to such an interference. While there is no risk “to bar an applicant who considers that an arbitration agreement is void, inoperative or incapable of being performed from access to the court before which he nevertheless brought proceedings” if they can contest recognition and enforcement in this very court, the defendant will indeed be denied access to that court if the courts of another member state enforce an arbitral award by ordering him to stay these proceedings. And while failure to comply with an arbitral anti-suit injunction “is not capable of resulting in penalties being imposed upon it by a court of another Member State”, the enforcement of such an injunction in another member state would attach to the award that exact kind of penalty. Thus, while the recognition of such an arbitral award in the member state where the proceedings are brought is no more contrary to the Brussels I Regulation than the court’s power to stay proceedings of its own motion in order to give effect to an arbitration clause, the enforcement of such an award by the courts of another member state would be much more similar to the situation which the ECJ ruled out in West Tankers.
Second, the ECJ has not yet decided on the admissibility of damage awards in view of its restrictive approach to anti-suit injunctions. English courts seem to distinguish the one from the other by treating anti-suit injunctions as a remedy for the jurisdictional dimension of arbitration agreements while considering damages as a remedy for their contractual dimension. Yet, one may argue that the practical effects of both remedies are still very similar, especially if damages are granted, as in The Alexandros T, by way of an indemnity even before litigation has finished. But although it is hard to see why the ECJ would not consider damage awards to be contrary to “the general principle that every court seised itself determines, under the rules applicable to it, whether it has jurisdiction to resolve the dispute before it” as formulated in West Tankers, it is indeed not very likely that the Court will get a chance to make such a decision after the English courts – the only courts that actually grant such awards – saw no need to submit the question in The Alexandros T.
Finally, it has been noted (by Hartley  ICLQ 843, 866) that the new rules on recognition and enforcement of decisions that have been obtained in violation of an arbitration agreement in paras 2 and 3 of recital (12) leave open one particular case, namely the situation where a court is asked to recognize and enforce both an arbitral award made within the jurisdiction (and thus not creating an obligation under the New York Convention) and a conflicting judgment on the merits from another member state. While the wording of recital (12) indicates that the court has to give effect to the judgment, this would give the arbitral award the weakest effect in its “home jurisdiction”. The better approach therefore seems to be to consider arbitral awards made within the jurisdiction as a “judgment given between the same parties in the Member state addressed” and apply Art. 45(1)(c) of Brussels I by analogy.