I. INTRODUCTION
The ‘Italian Torpedo’ has long vexed the Brussels Regulation on jurisdiction and the recognition and enforcement of judgments.Footnote 1 This phenomenon involves the use of litigation with the primary purpose of delay. An action is filed in Italy, despite the existence of a valid agreement between the parties for the matter to be heard in another Member State. The Italian courts, not known for their speed, could take a long time to make a determination on jurisdiction. Any action filed in the appropriate courts—the courts that were agreed between the parties—must wait for the determination of the Italian courts that they have no jurisdiction before proceeding. The resulting delay can be of significant litigious advantage to one party and cause significant unfairness to the other.
A recast of the Brussels I Regulation was finalized in December 2012, and will take effect from 10 January 2015.Footnote 2 One of the most significant elements of this reform—perhaps the most significant—is the change made to accommodate cases where the Italian Torpedo is used to frustrate a choice-of-court agreement. When the regulation comes into force, this problem should, in theory, be solved. However, we believe there is a potential problem with the solution adopted which might allow the continuation of Torpedo litigation to frustrate choice-of-court agreements, or allow troublesome and pointless, though perhaps not abusive, delays to litigation taking place in the appropriate forum. In this article, we want to explore the effectiveness of this change in the Recast, using a recent case from the Irish Supreme Court as an illustration of these potential difficulties.
In February 2014, the Irish Supreme Court delivered judgment in the case of Websense v ITWAY.Footnote 3 In this case, the Supreme Court felt it was bound by the provisions of the Brussels I Regulation to stay Irish proceedings between the parties while the Italian courts decided if they had jurisdiction to hear proceedings that were related to, but not identical to, the Irish action. They felt bound to do this notwithstanding that an agreement existed between the parties giving the courts of Ireland exclusive jurisdiction in the matter at the heart of the Irish proceedings, and that the case would have to return to Ireland should the Italian courts ultimately decide that jurisdiction should not be taken by the Italian courts. The Supreme Court had some reservations about this result and hoped, in the concluding paragraphs of the judgment, that reform would be introduced to stop this happening again. The Supreme Court seems to have been unaware that reform in this area has, in fact, already taken place, albeit in any event too late to avail the plaintiff in Websense. Does the Recast therefore make Websense a dying breed, the last case of its kind? Perhaps not. The Recast, while addressing the problem of delay to litigation involving choice-of-court agreements in some cases, is not entirely clear in its scope. Its new provisions may not apply to cases that are related rather than identical. Therefore, Websense illustrates a possible failure of the Recast to avoid problematic delays in honouring choice-of-court agreements. It may even show how Torpedo actions could circumvent the Recast's attempts to prevent them.
This article proceeds in four parts. In Part II, we examine the 2001 Brussels Regulation rules on lis pendens and choice-of-court agreements and look at how these resulted in the rise of the Italian Torpedo and delay litigation. In Part III, we look at the provisions of the 2012 Recast designed to solve problems associated with the Italian Torpedo. In Part IV, we set out the facts of Websense and the reasons for the Irish Supreme Court's judgment in that case. In Part V, we raise the possibility that cases like Websense—where the litigation is related to, but not identical to, litigation underway in Italy—might not be covered by the new provisions of the Recast Regulation. We conclude that unless these provisions are interpreted as including related actions—which we think unlikely—the Recast will be only a partial solution to the problem of delay in honouring choice-of-court clauses, and may fail to put an end to the Italian Torpedo.
II. THE BRUSSELS REGULATION, LIS PENDENS, AND THE ITALIAN TORPEDO
Regulation 44/2001/EC—known as the Brussels Regulation—harmonized the rules on jurisdiction and recognition and enforcement of judgments in civil and commercial matters across the Member States of the European Union. The Regulation was itself a successor to the Brussels Convention on Jurisdiction and Recognition and Enforcement.Footnote 4 It made the rules of the Convention, in modified form, directly applicable in, and binding on, Member States.Footnote 5 The Regulation sought to produce a set of clear and uniform rules that would be consistent and predictable. Divergence in jurisdictional rules, and the consequent effect on recognition and enforcement of judgments, was thought to endanger the smooth operation of the internal market. The core purpose behind the Brussels Regime was to harmonize these rules in order to enable the free movement of judgments.Footnote 6 In part because of this priority, the interpretation of the provisions of the Regulation has been very strict. There is very little flexibility in the way that its rules are applied. This gave rise to the possibility of the Brussels Regime allowing obstructionist litigation.Footnote 7
The most problematic of these litigation tactics was a phenomenon that came to be known as the Italian Torpedo. First posited in 1997,Footnote 8 this tactic used the Regulation's rules on pending litigation within the EU in an ingenious way to delay litigation. Under Article 27 of the Regulation, if there is litigation pending in the court of a Member State, the court of any other Member State which is asked to hear the same action must stay the matter until the court first seised could determine whether or not it had jurisdiction to hear the matter.Footnote 9 Article 28 contains a similar rule, with slightly more discretion, where the litigation is not identical, but is closely related. Here, any Court other than the court first seised may decline jurisdiction in such cases. The actions are related if they are ‘so closely connected that it is expedient to hear and determine them together to avoid the risk of irreconcilable judgments'.Footnote 10 These rules exist to avoid conflicting judgments from different Member States, and show the degree of comity and trust expected from Member State courts: if a court is seised of a matter inappropriately, another court should wait for that court's own determination of the matter before proceeding. It should not decide on the jurisdiction matter itself.
Though well intentioned, these provisions transpired to be ripe for exploitation. A phenomenon grew up whereby preemptive actions were filed in courts that were unrelated to the action in question for the purposes of delay. The court where the matter was first initiated would be entitled to determine its own jurisdiction. If one filed the proceedings in a court system that was notoriously slow in its determination of jurisdiction issues—Italy was the usual choice, where pace of court determinations on jurisdiction is glacialFootnote 11—then the delay could be substantial indeed. Meanwhile, under the lis pendens rules, an action in any other place would have to be stayed until the Italian courts made their determination that they had nothing whatsoever to do with the matter in question, and had no reason to take jurisdiction.
This practice seemed most unfair when there was a clear and obvious answer to the jurisdiction question: the parties had agreed to a prima facie valid choice-of-jurisdiction clause in a contract. Under Article 23 of the Regulation, parties are fully entitled to make such choices unless the Regulation provides otherwise.Footnote 12 Despite the seeming injustice of this, the European Court of Justice held that this tactic was a legitimate use of the Regulation's provisions.
In Gasser v MISAT,Footnote 13 proceedings involving the same cause of action between the same parties were brought in Italy by MISAT and, eight months later, in Austria by Gasser. A regional Austrian court stayed its proceedings until the Italian court, as the court first seised, had established whether or not it had jurisdiction. Gasser appealed to a national Austrian court on the ground, inter alia, that the Italian court did not have jurisdiction according to a prima facie valid agreement that gave the Austrian courts exclusive jurisdiction over the matter. On application by the Austrian court for a preliminary ruling on the interpretation of the Brussels Convention, the ECJ held that the lis pendens provisions of the Convention should be interpreted broadly and would cover any pending litigation in another contracting state. A court other than the court first seised could not, therefore, hear an action that was underway elsewhere, even if that court appeared to have exclusive jurisdiction under a choice-of-court agreement.Footnote 14 The ECJ was unmoved by concerns that the delay of the Italian courts was so significant as to cause serious detriment to Gasser. The Court held that the delay of a court in hearing a matter was a concern more appropriately addressed to the European Court of Human Rights.Footnote 15
The position taken by the European Court in these cases is understandable. The Brussels Regime is founded on the bedrock of cooperation, mutual trust and mutual respect between courts. Any measure that results in the courts of one Member State questioning the processes of another is incompatible with the general scheme of the Regulation. However, the Court's tolerance of the Italian Torpedo was hugely problematic.Footnote 16 It undermined another core tenet of the regime: that parties should have the right to choose for themselves what jurisdiction was to govern their disputes, and that this choice should be respected. The Italian Torpedo allowed litigants to frustrate such choices, and to exploit the mechanisms of the Regulation for litigious advantage. As Hartley puts it, Gasser ‘seriously jeopardized the effectiveness of choice-of-court agreements in the European Union’.Footnote 17 Strong objections can be made to Gasser on the basis of its undermining of party autonomy alone, and any potential solution could be judged through the lens of how well it mollified these objections. In this article, however, we will focus primarily on the practical issue of delay, and the unfairness that can result from this, in assessing how well the problems of Gasser have been addressed.
III. THE RECAST
The problems created by Gasser were finally addressed in the Recast of the Brussels I Regulation,Footnote 18 which was finalized in December 2012. This followed a review of the operation of the Regulation in the late 2000s,Footnote 19 which considered some radical proposals in the course of negotiations,Footnote 20 but ultimately adopted more modest reforms. Perhaps the most significant changes to the Regulation are in the areas of choice-of-court agreements and lis pendens. Party choice has been respected to an even greater extent by removing the requirement that, for a valid choice, one party must be domiciled in a Member State; any party can now choose a Member State's jurisdiction under the new Article 25.Footnote 21 Some controversial issues around choice have not been addressed, such as choice of a non-Member State forumFootnote 22 and unilateral or one-way jurisdiction clauses.Footnote 23 However, the recast is a significant improvement in the way that lis pendens is dealt with, and a significant step towards respecting choice-of-court agreements.
The intention of the Recast is made clear from its recitals. In Recital 22, it is apparent that the amendment to the lis pendens provisions is designed to stop the sort of abuse that can be seen in Italian Torpedo cases:
in order to enhance the effectiveness of exclusive choice-of-court agreements and to avoid abusive litigation tactics, it is necessary to provide for an exception to the general lis pendens rule in order to deal satisfactorily with a particular situation in which concurrent proceedings may arise. This is the situation where a court not designated in an exclusive choice-of-court agreement has been seised of proceedings and the designated court is seised subsequently of proceedings involving the same cause of action and between the same parties.Footnote 24
In this Recital, the drafters of the Recast acknowledge that more needs to be done to ensure that choice-of-court agreements are respected. In the particular circumstances of exclusive choice-of-court agreements, the Regulation's priorities need to be the reverse of those prioritized by the ECJ in the Gasser case. The general respect which the Regulation expects between courts and between states should not abide one party exploiting the mechanisms of the Regulation for improper litigious advantage. The Recast prevents such exploitation by allowing the Court named under the choice-of-court agreement to proceed notwithstanding the fact that litigation is underway in a court first seised, and notwithstanding the fact that that other court has not yet stayed its proceedings.
This is achieved by means of the new provisions on lis pendens. Articles 29 and 30 of the Recast replicate, without substantive change, the contents of Articles 27 and 28 of the previous Regulation. The modification is found in the new provisions of Article 31(2). It reads:
Without prejudice to Article 26, where a court of a Member State on which an agreement as referred to in Article 25 confers exclusive jurisdiction is seised, any court of another Member State shall stay proceedings until such time as the Court seised on the basis of the agreement declares that it has no jurisdiction under the agreement.Footnote 25
Article 31(3) clarifies that where the court designated by agreement establishes that it has jurisdiction, other States should then decline jurisdiction entirely. Under this new provision, unless a defendant acquiesces to another jurisdiction,Footnote 26 the jurisdiction agreed under a choice-of-court agreement takes priority, and the court designated under that agreement is the one that makes the determination as to the validity of the agreement.
Though it is not made explicitly clear in the text of Article 31, Recital 22 indicates that the designated court does not have to wait for the other jurisdiction to stay proceedings: ‘The designated court should be able to proceed irrespective of whether the non-designated court has already decided on the stay of proceedings.’Footnote 27 Essentially, Member State courts no longer have to wait for Italian courts—or anyone else—to acknowledge the fact of their exclusive jurisdiction to hear a matter under a choice-of-court agreement. They can, when the new Regulation is applicable, continue to hear the matter and make a determination on the issue of whether the choice-of-court agreement does in fact confer exclusive jurisdiction on the courts of that Member State. Nothing can delay or derail this. Litigation filed elsewhere solely for the purpose of delay and disruption of a choice-of-court agreement will no longer be tolerated by the machinery of EU law. The Torpedo is to be disarmed.
Article 31 thus seems like a fairly comprehensive solution to the problem of the Italian Torpedo in choice-of-court agreements. Some other minor issues have been raised, such as the fact it applies only to exclusive jurisdiction conferred by agreement, so non-exclusive clauses are outside the scope of the new mechanism.Footnote 28 These issues are small in the scheme of things and the new provisions have been welcomed.Footnote 29 However, we believe the facts presented by a case before the Irish Supreme Court highlight a potential problem with the effectiveness of this provision: it is unclear whether the new rules will apply to related proceedings, or only to identical proceedings. We will now outline the facts and holding of that case, and then address the questions it raises for the application of the Recast.
IV. WEBSENSE v ITWAY
A. Facts
The facts of Websense are Byzantine in their complexity, but need only to be outlined here. The case arose from a commercial relationship that was in place between the Websense Group and Itway whereby Websense supplied information technology products to Itway. Itway then distributed these products to resellers in a number of countries. The substantive issue between the parties to the Irish proceedings was a dispute as to whether or not monies were due and owing from Itway to Websense under a distribution agreement dated 26 April 2010. The distribution agreement contained an exclusive choice-of-court clause, indicating that any action related to the agreement should be brought in Ireland. Proceedings were issued, a month apart, in Italy and in Ireland: the Italian proceedings were issued by Itway on 22 March 2012 and the Irish proceedings were issued by Websense on 19 April 2012. The Italian court was, therefore, the court first seised. This raised a preliminary question around jurisdiction, with Websense bringing the matter before the Irish Supreme Court to ascertain whether the Irish courts had jurisdiction to hear the case or whether they had to stay the case, pending deliberation of the question of jurisdiction by the Italian court, which could take some time. The delay was likely to be significant, as the Italian court had declined to examine the jurisdiction question as a preliminary matter, and instead decided that it would be heard alongside the substantive issues in the case.Footnote 30
The Websense case is of a type slightly different to Gasser because it involved lis pendens under Article 28 rather than Article 27: it was an allegedly related action, rather than an identical action, that was underway in Italy.Footnote 31 The action in Italy was against Websense Italia, a related company to the Irish plaintiff Websense International, an Irish registered company. Though similar, the subject matter of the two cases was somewhat different. In the Irish proceedings, Websense sought payment from Itway under the distribution agreement. Though also involving breach of the distribution agreement, and concerning the non-payment at the heart of the Irish case, the Italian proceedings involved broader claims. Itway alleged that Websense had used its superior commercial position to manipulate companies ordering Websense products from Itway. They also claimed, amongst other things, that Websense had made a subsequent agreement with Itway, which Websense denied, and that Websense's failure to honour that agreement incurred significant financial loses for Itway. The Supreme Court noted that if Itway's version of events was correct, the Irish proceedings on payment under the agreement were only one part of a broader dispute between the parties.Footnote 32
Article 28 rules are different to Article 27. Where the cases are identical, involving the same parties and the same issues, any court other than the Court first seised must stay the matter. However, if they are related—which is defined as being ‘so closely connected that it is expedient to hear and determine them together to avoid the risk of irreconcilable judgments resulting from separate proceedings’Footnote 33—then the other court may stay the matter, but is not obliged to do so. There is a discretion for the court to refuse to stay the matter.
Websense resisted the application for a stay under Article 28 on two grounds: first, that the two proceedings were not ‘related actions’ under Article 28; and secondly, if they were related, that the Court should use its discretion to refuse to grant the stay, as the parties had deemed Ireland as the appropriate jurisdiction under a choice-of-jurisdiction clause in the distribution agreement. They were unsuccessful on both points.
It is important to note that there was no indication from the Irish courts that the Italian proceedings were a Torpedo action, issued solely for the purposes of delay. The Supreme Court never characterized the Italian action in that way. Itway might have significant claims to pursue against Websense and there may be reasons for some of these to be heard in Italy. However, even if not a Torpedo action, the potential unfairness is the same: if the choice-of-court clause in the distribution agreement was valid, it is possible that any dispute related to the agreement, including Itway's various claims, should be pursued in Ireland. If this is the case, the delay, and the potential unfairness, is the same, even if it is done in good faith and not for purely strategic reasons.
B. Related Action
In determining whether the cases were related, the Court had to consider if they were so closely connected that it is expedient to hear and determine them together to avoid the risk of irreconcilable judgments resulting from separate proceedings. The Court noted that the ECJ had held that the concept should be interpreted broadly, in order to ‘cover all cases where there is a risk of conflicting decisions’.Footnote 34 Having referred approvingly to English authority,Footnote 35 MacMenamin J found that a sufficiently close connection existed between the Italian and Irish proceedings to make them related actions. Crucial to this was the fact that the dispute at issue in both cases emanated from the same distribution agreement; that both cases would have to address the alleged variation of the agreement in 2011; that the parties overlapped to some degree; and that the witnesses would likely be the same.Footnote 36 Perhaps most importantly, ‘a determination in one case would have a significant bearing on the ultimate outcome of the two cases, having regard to claims and potential counterclaims’;Footnote 37 that is, the judgments could easily be in conflict. This result was ‘almost self-evident’ from the facts presented.Footnote 38
C. Exclusive Jurisdiction Clause and Use of Article 28 Discretion
In considering the question of whether or not a stay was appropriate, the Court cited the ECJ in Gasser to the effect that the court second seised is not entitled to ignore the existence of prior proceedings between the parties, despite the existence of a binding exclusive jurisdiction clause designating a relevant forum. It was, therefore, for the court first seised—in this instance, the Italian court—to establish whether or not it is the appropriate forum. However, unlike Gasser, as an Article 28 matter, the Court had discretion to stay the proceedings or not in this instance, and Websense asked that the Court use its discretion.
The Court adverted obliquely to the strategic use of the Italian actions as a litigious tactic designed to stall or frustrate proceedings, and suggested that ‘in circumstances such as the initiation of “negative declaration proceedings” designed solely to forestall proceedings elsewhere, careful scrutiny might possibly be warranted’.Footnote 39 The Court considered that, perhaps, ‘proven gross mala fides or evidence of abuse of process in the initiation of proceedings elsewhere might be sufficient’Footnote 40 to warrant the use of the Court's exercise of its discretion to refuse to grant a stay. Nonetheless, even this was to be considered in light of the decision in Gasser and the risk of conflicting judgments; the Court could not agree with English authority, which suggested that Gasser was of limited relevance in Article 28 matters.Footnote 41 The Court concluded that its discretion could not be used simply because the litigation in question operated contrary to a seemingly valid choice-of-court clause. It followed the approach set down by the ECJ in Gasser, where it stated that ‘[n]owhere does the Convention provide that courts may use the pretext of delays in procedure in other contracting States to excuse themselves from applying its provisions’.Footnote 42
While the Court felt bound to stay the case in accordance with the Regulation, it was not entirely happy about this. One gets the sense the Court believed that it would be preferable if the case could have been heard in Ireland. The Court noted the distinct possibility that these proceedings will find their way back to an Irish court when the Italian courts have adjudicated on the choice-of-jurisdiction clause. MacMenamin J described this as a ‘most unsatisfactory state of affairs’,Footnote 43 and opined that the Regulation did not allow for circumstances such as these to be considered. He concluded his judgment with a call for reform:
[T]here are strong grounds for suggesting that a more nuanced approach to the rules contained within the Regulation is necessary if the commitment to certainty is not to lead to impractical situations being mandated. This Court can do no more than suggest that, for reasons such as those identified not only in this case but in other cases which have come before the Irish courts, a review of the rules in respect of jurisdiction contained within the Regulation is highly desirable.Footnote 44
The Supreme Court seemed unaware that some reform of the sort it called for in this passage has already been put in place, although not yet come into force. The Recast seems to provide the ‘more nuanced approach’ to the Regulation's provisions on lis pendens that the Supreme Court desired. But is it actually? Does it apply to a case like Websense, or would the new provisions in fact be useless in a situation such as this one? It is regrettable that the Court did not have its attention drawn to the Recast, as it might have used the opportunity to consider this issue.
V. RELATED ACTIONS AND THE RECAST
A. Article 31 of the Recast and Related Actions
Clearly the intention and effect of Article 31 of the Recast is to reverse the result in cases such as Gasser, where there is an exclusive jurisdiction clause, and identical proceedings are filed elsewhere for the purposes of delay. There remains a question mark, however, over whether it will have the effect of reversing the outcome in cases such as Websense. Obviously, the reform came too late to make any difference in the Websense case; the Recast will only apply from 10 January 2015. However, there is a more fundamental problem: it is not clear how the new rules are to operate in respect of related proceedings, like those in Websense, rather than identical proceedings, like those in Gasser.
One might think that, as a matter of policy, such cases should be included. However, the language of the Recast does not strongly support this position. The language of Article 31 itself is neutral on the point; it does not clearly limit itself to identical actions, nor does it expressly include related actions. However, two textual arguments point toward the exclusion of related actions. First, the relevant recital, which can be important for interpretation, suggests that these cases were not contemplated in the situation imagined in Article 31. Recital 22 speaks of ‘proceedings involving the same cause of action and between the same parties’.Footnote 45 It says nothing of related proceedings. These recitals guide the interpretation of the substantive provisions of the Regulation, though they are not decisive.Footnote 46 Secondly, Article 29(1)—the provision on lis pendens for identical cases—starts with the phrase ‘Without prejudice to Article 31(2)’, specifically suggesting that the rules for such cases are subject to this exception. No such qualification is included in Article 30, dealing with related actions.
It is open for interpretation, then, as to whether these new rules will apply to related actions.Footnote 47 Cases like Websense might well be excluded from the new rules in Article 31 and Websense illustrates two separate problems that would arise if this were the case. The first is the unfairness that can result from delaying an action proceeding in the appropriate forum, even when related actions are taken in good faith. The delay in Websense, should the Italian courts ultimately hold that they cannot determine any matter related to the distribution agreement, will be just as unfair and pointless as a case where the matters were identical. The reform sought by the Irish Supreme Court would not be provided in the Recast. The second problem is the possibility that parties could misuse this related claims exception to rearm the Italian Torpedo. It would be an invitation to those who wish to file Torpedo litigation to alter slightly the substance of the claim, or alter the parties involved, and thereby avoid the strictures of Article 31. This does not seem like it would be difficult to do. It might also be very difficult for a court to know, on the face of it, if this were a stratagem or a legitimate stating of a broader case.Footnote 48 It could be a very simple way to nullify the Recast's attempts to prevent Torpedo actions.
B. Can This Problem Be Avoided?
These problematic outcomes are not inevitable, and might be avoided by the courts in one of two ways. First, the courts might interpret the provisions of Article 31 to apply to closely related proceedings as well, so that a court should stay any action, or any part of an action, that on its face was governed by a choice-of-court agreement indicating another forum, even if that case before them were slightly different in form or substance to those before the other court. It is difficult to predict if this interpretation will be adopted. It is a plausible reading of the text of Article 31. The broader intention of the provision—to stop the delay of litigation that is governed by a choice-of-court clause—might be thought to point towards the more inclusive reading. However, the wording of the Recital and Article 29 militates against it. In addition, there is clearly a difference between identical and related claims: it will surely always be the fair result to stay the former, whereas the latter may be more complicated, and there may be reasons for both claims to coexist. To put it another way, the filing of an identical claim in a jurisdiction other than the one stated by an exclusive jurisdiction agreement will almost always be abusive; the filing of a related claim might not be.
It would not be surprising if the Regulation therefore considered identical and related claims to be different for this purpose. One might expect that if the Article 31 rule was to apply to related claims, it would explicitly provide for this, and provide for a discretion rather than a mandatory stay, as is done in Article 30 on lis pendens stays for related litigation. In Hartley's terms, Article 31 is a ‘sort of reverse lis pendens rule’.Footnote 49 If it is a mirror image of those rules, the lack of a mirror image of the discretion to stay related claims might indicate that they are intended to be excluded from the scope of Article 31. We believe, therefore, that the interpretation of Article 31 to include related actions is unlikely.
The second way of avoiding these problems would be for courts facing requests to stay related litigation under Article 30 (formerly Article 28 of the 2001 Regulation), like the Irish Supreme Court faced in Websense, to exercise their discretion to refuse a stay on the basis of the exclusive jurisdiction agreement. Though the Court declined to follow English authority such as JP Morgan v Primacom,Footnote 50 might it, after the Recast, change its view on the exercise of this discretion in light of the existence of the new Article 31? The Court in Websense was unwilling to proceed with related proceedings because the Regulation, before the Recast, did not envisage that it would ever be permissible for a Member State court to usurp the jurisdiction adjudicating function of the court first seised simply because a jurisdiction agreement existed. From 2015, this will no longer be true: Article 31 of the new Regulation will now envisage that the courts seised by reason of exclusive jurisdiction agreements have a sort of priority over any other court seised. Even if that provision did not explicitly apply to related actions, it could be argued that the existence of this new provision would give courts a valid reason to proceed with the litigation rather than wait for other courts to make determinations. It might empower courts to exercise discretion where previously they would not have done so.
This is an appealing option, but it is not without problems. One of the main reasons that the Irish Court was unwilling to follow the English authority in this case was the risk of irreconcilable judgments that came from exercising the discretion to stay related actions, which was apparently not at issue in the JP Morgan case.Footnote 51 Avoiding irreconcilable judgments from different Member State courts is a key priority of the Brussels Regime. The provisions in Article 31 are designed to avoid the new rules leading to irreconcilable judgments; the court first seised must stay its proceedings to allow the court designated by the agreement to make the determination.Footnote 52 However, this obligation to stay proceedings would not apply if the courts were simply using their discretion under Article 30. The risk of conflicting judgments would be markedly higher in a case such as Websense; the Irish court might consider it proper to hear the action under the agreement, but the Italian court would not be under any obligation to stay the proceedings taken in Italy, and two conflicting judgments might be issued. Therefore, there would be a good case to say that this solution would be too different from the scheme envisaged by Article 31 to be acceptable. The Irish Supreme Court's reasons for refusing to follow JP Morgan and to exercise its discretion persist after the Recast; courts will still have to consider the risk of irreconcilable judgments as a core priority in the use of this discretion and might find, as the Irish Supreme Court did, that this risk is too great to allow a related action to proceed.
At this juncture, we cannot know for sure if either of these interpretive solutions will be adopted. It will require a judgment of the ECJ in a case with similar facts to those presented in Websense for this matter to be fully addressed. However, given the strictness of the ECJ's past interpretations of the Brussels Regulation, and its desire to avoid irreconcilable conflicting judgments, we suggest that the most likely result that the new provisions of the Recast will be held to make no change in respect of related actions.
VI. CONCLUSION
It is unclear at this juncture where the limits of the Article 31 reform are to be drawn. If it is held not to include cases such as Websense, which involve related actions, this would be a significant and potentially problematic omission. Such an omission might have been intended, and there may have been good reasons for the distinction being drawn, or it may have been overlooked and its consequences not fully considered. Either way, it will result in substantial unfairness in certain cases, and we cannot say, therefore, that the problems around the frustration of choice-of-court agreements have been fully resolved by the Recast. Moreover, if it opens the way for Torpedo actions to circumvent Article 31, it will be a failure so significant and consequential that it could undermine one of the core purposes of the Recast. It seems doubtful that the interpretive options that would avoid the problematic consequences of excluding related actions will prove to be acceptable to the ECJ. On this point, only time will tell. In the event that there is no interpretive solution, there is little to be done. Given that the Recast is still a very recent development, additional reform in this area is unlikely for the foreseeable future. Despite the promise of the Recast, the Brussels Regime may have to tolerate the unfair frustration of choice-of-court agreements for some time to come.
