3.1. The frustration account

Richard Fentiman has argued that strong cause will be established, inter alia, in ‘exceptional circumstances in which a court accepts that changed circumstances have in effect frustrated the agreement’.Footnote ⁴¹ Call this the ‘frustration account’. This account may seem to explain why only unforeseeable factors can amount to strong reasons (foreseeable events generally do not frustrate contracts),Footnote ⁴² and why self-induced related proceedings cannot (frustration cannot be self-induced).Footnote ⁴³

However, the frustration account proves inaccurate upon closer inspection, which reveals that foreseeability is much less crucial under the strong cause test than frustration. Reasonably foreseeable events rarely frustrate contracts.Footnote ⁴⁴ This is because frustration occurs when an event’s occurrence makes contractual performance ‘radically different’ from what was promised, and an event’s foreseeability ‘informs the parties’ knowledge, expectations, assumptions and contemplations, in particular as to risk’ when they contracted.Footnote ⁴⁵ This proposition is not qualified in the event that the foreseeable event has onerous consequences for the parties. Instead, the opposite seems true: if the parties do not expressly provide for an onerous yet foreseeable event, the proper inference is usually that they ‘accepted the risk’.Footnote ⁴⁶

By contrast, the foreseeability of some of the strong cause test’s factors—of an unfair trial in the chosen court and related proceedings involving third parties—seems less consequential.Footnote ⁴⁷ In Zephyrus Capital Ltd v Fidelis Underwriting Ltd, aircraft owners sued insurance companies in the UK for losses caused by Russia’s detention of their aircraft upon its invasion of Ukraine. Some policies contained English jurisdiction agreements, others contained Russian jurisdiction agreements, and of the latter group, some insurers submitted to the English proceedings, while others sought a stay. Henshaw J denied the stay on two alternative grounds. First, Russian state inference would likely deprive the owners of a fair trial there.Footnote ⁴⁸ Such interference might have been somewhat foreseeable, but that only meant that this factor’s weight would be reduced, not that it would be negligible.Footnote ⁴⁹ Second, requiring the owners to sue only some of the insurers in Russia would multiply related proceedings, since all claims raised similar factual issues and many claims would already proceed in the UK.Footnote ⁵⁰ Even though the prospect of related Russian proceedings was ‘foreseeable and self-induced’ by the owners, who had voluntarily entered into policies containing conflicting jurisdiction agreements, it remained a factor which could ‘properly be taken into account’.Footnote ⁵¹

Another way in which foreseeability differs in its relevance between frustration and the strong cause test concerns the effect of exclusion clauses. Frustration can be contractually excluded, as is clear from how courts give effect to clauses expressly purporting to do soFootnote ⁵² and force majeure clauses.Footnote ⁵³ While courts may read these clauses contra proferentem,Footnote ⁵⁴ a clearly worded clause purporting to exclude the operation of frustration will be given effect. The logic here is that the parties, by expressly stating that an event will not affect enforceability or by expressly allocating the risk of that event occurring, have demonstrated that they foresaw that event, which means it cannot trigger frustration.Footnote ⁵⁵

The strong cause test, by contrast, is incapable of being contractually excluded. This is demonstrated by cases involving jurisdiction agreements coupled with what are called ‘forum non conveniens waivers’.Footnote ⁵⁶ The label is slightly misleading: the parties agree to waive any objection to the chosen court’s exercise of jurisdiction on grounds of ‘convenience’ factors,Footnote ⁵⁷ under either forum non conveniens or the strong cause test.Footnote ⁵⁸ Importantly, courts never give these clauses their full effect: strong cause can still be established on the basis of convenience factors. As Field J acknowledged in Bank of New York Mellon v GV Films, ‘where a party has expressly agreed not to rely on convenience arguments in resisting the jurisdiction of the nominated court … especially strong grounds will be required before the exclusive jurisdiction clause can be departed from on grounds founded on convenience’.Footnote ⁵⁹ This is not contra proferentem but a qualification based on principle. ‘If even an exclusive jurisdiction clause will not trump a stay application, then at least a similar approach must follow in respect of a stay application brought … in breach of an FNC waiver clause.’Footnote ⁶⁰ In other words, forum non conveniens waivers cannot preclude the strong cause test being applied on the basis of convenience factors. It is irrelevant that parties, by expressly purporting to waive forum non conveniens, demonstrate that they foresaw the inconvenience of their chosen forum—if the inconvenience is ‘especially strong’, it still amounts to strong reasons.

3.2. The remedial discretion account

Courts have a general discretion to regulate the remedy for breaches of contract.Footnote ⁶¹ Adrian Briggs, among others,Footnote ⁶² has argued that when strong reasons are established, the court acknowledges that a jurisdiction agreement has been breached, but then exercises its remedial discretion to deny specific relief and award damages instead. Call this the ‘remedial discretion account’. As Briggs puts it:

The remedial discretion account has its merits. For one, it may explain why the interests of third parties affected by related proceedings are weighty factors.Footnote ⁶⁴ Third-party interests cannot typically ground defences to private law claims, but they can support the denial of an injunction and the award of damages instead.Footnote ⁶⁵

However, the remedial discretion account ultimately also does not work, because it struggles against the weight of contrary practice and authority. In particular, while courts do sometimes award damages for breaches of jurisdiction agreements, they never do so after making a finding of strong cause. Instead, courts award such damages in only three categories of cases. First, damages may be awarded in addition to other specific relief (e.g. the exercise of jurisdiction and/or an anti-suit injunction), if specific relief cannot satisfactorily remedy the claimant’s loss.Footnote ⁶⁶ Second, damages may be awarded to remedy the breach of a jurisdiction agreement in circumstances where the English court completely lacks the power to order specific relief (e.g. when EU Regulations prohibited anti-suit relief restraining proceedings in EU Member States), to reverse the losses suffered in the foreign proceedings.Footnote ⁶⁷ Third, damages may be awarded when the claimant would have been entitled to anti-suit relief from an English court, but delayed in seeking such relief.Footnote ⁶⁸ Against this, the only ‘authority’ that might be cited in favour of courts awarding damages after strong reasons are established are obiter comments in Donohue v Armco, which restated parties’ concessions to that effect without actually endorsing them.Footnote ⁶⁹

Thus, when courts find strong cause, they will not only refuse specific relief, but also refuse to award any other remedy, like damages for breach of the jurisdiction agreement.Footnote ⁷⁰ This rules out the possibility that the strong cause test is simply the exercise of remedial discretion. After all, remedial discretion in private law involves a choice between remedies, not a choice to award no remedy:Footnote ⁷¹ the claimant must still receive some relief—a suspended injunction,Footnote ⁷² or at least nominal damagesFootnote ⁷³—not nothing at all. It is no response to say that most parties who rely on jurisdiction agreements do not seek damages as an alternative remedy: at least some damages in lieu of specific relief can be awarded without being specifically pleaded.Footnote ⁷⁴ Nor is it sufficient to say that such damages are generally not awarded, in the event of a finding of strong cause, because they would be difficult to quantify:Footnote ⁷⁵ in that case negotiating damages,Footnote ⁷⁶ or at least nominal damages, would be the appropriate remedial response. A better response might be that, in strong cause cases, damages are usually not actively pursued in court. Yet, the fact that parties almost never take up the point—and, indeed, that judges never ask to be addressed on it—itself suggests that, when strong cause is found, no such entitlement to damages exists.

To be clear, it is not denied that damages should be available as a remedy for breaches of enforceable jurisdiction agreements. That follows inexorably from the contractual account. The point is simply that, when strong reasons are established, the jurisdiction agreement is rendered unenforceable, not that it remains enforceable with specific relief withheld and damages awarded instead.

3.3. The qualified contractual account

The frustration and remedial discretion accounts start from the premise that jurisdiction agreements are as enforceable as ordinary contracts, and explain the strong cause test within the confines of that logic. A third account starts from a different premise but attempts to reach the same end.

The starting point here is that an English court always has a discretion to decide whether it should exercise jurisdiction, and that contracts which purport to ‘oust’ or interfere with that discretion are unenforceable.Footnote ⁷⁷ Ordinarily, the court exercises this jurisdictional discretion by applying forum non conveniens. The strong cause test, it is said, is an alternative test of jurisdictional discretion which applies when parties have entered into a jurisdiction agreement, which is ‘similar but different from’ forum non conveniensFootnote ⁷⁸ and couched in terms that respect parties’ contractual bargain. As Lord Bingham put it in Donohue:

In short, where parties have entered into a jurisdiction agreement, the court should generally exercise its jurisdictional discretion in a way that respects ‘the contractual bargain’. The strong cause test is just the exception to this general position.

This account of the strong cause test—call it the ‘qualified contractual account’—is reflected in many judgmentsFootnote ⁸⁰ and academic commentaries,Footnote ⁸¹ most of which simply endorse the above passage in Donohue. In a vague and general sense, the qualified contractual account is accurate: the strong cause test is indeed an exception to a general rule holding parties to their jurisdiction agreement. But this is really only a description of the test, rather than an explanation of or justification for it.

If one probes further, the qualified contractual account proves unhelpful, because it is question begging. Why, in fact, should the parties’ contractual bargain be respected? The answer cannot simply be that jurisdiction agreements are contracts and contracts should be upheld. As the Supreme Court of Canada held in Douez v Facebook Inc, ‘because forum selection clauses encroach on the public sphere of adjudication … courts do not simply enforce them like any other clause’.Footnote ⁸² After all, the starting premise here is that the court’s jurisdictional discretion exists to protect some interest other than freedom of contract, like the ‘efficient administration of justice’,Footnote ⁸³ which necessarily means that contracts relating to the exercise of jurisdiction cannot be ‘enforced as of right’.Footnote ⁸⁴ Indeed, insisting that jurisdiction agreements should be upheld just like ordinary contracts forces the conclusion that the strong cause test should be abolished.Footnote ⁸⁵

Perhaps, one might respond, the strong cause test does not prioritise parties’ contractual bargain over the efficient administration of justice but, rather, strikes an appropriate balance between those competing values. As Edwin Peel has argued:

But this response itself raises further questions. For one, it is unclear how such a balance might rationally be struck. After all, freedom of contract and the administration of justice are not commensurable values, meaning that they cannot be weighed and compared against each other on some objective scale. Any attempt to strike a balance between the two values will thus simultaneously be capable of being described as an ‘appropriate’ compromise while also being open to the charge of ‘inappropriately’ favouring one value over the other. In any event, the strong cause test does not, in fact, require an even balance to be struck between those two values. Rather, it requires courts to lean very much in favour of giving effect to parties’ agreement, and to require strong reasons to depart therefrom. One is still left without an explanation for why that should be so.

4.1. Appropriate dispute resolution

Forum non conveniens is a test that compares available courts as potential fora for litigation. It is commonly described as a two-stage test. At the first stage, the English court will ascertain whether it, or a foreign court identified by the party or parties seeking trial abroad, is prima facie the appropriate forum for trial. Here, courts will consider factors of ‘practical convenience’,Footnote ⁸⁸ such as the location of witnesses and evidence, the connections that parties have with either jurisdiction, the governing law of the main claim(s) pleaded and the prospect of parallel proceedings in the foreign court.Footnote ⁸⁹ At the second stage, the English court will then consider whether there are ‘circumstances by reason of which justice requires’ that trial should not occur in the prima facie appropriate forum.Footnote ⁹⁰ A significant consideration is whether there is a real risk that ‘substantial justice’ would be unavailable to one of the parties in that forum, because that party might face a biased or prejudiced coramFootnote ⁹¹ or might lack effective means to pursue his case.Footnote ⁹² This two-stage account of forum non conveniens, however, is more a useful shorthand description of its structure, rather than an analytic truth about its content. In reality, the first and second stage bleed into each other, and are really part of a single enquiry.Footnote ⁹³ That explains why the factors considered at the first stage also remain relevant at the second stage, where ‘all the circumstances of the case’ must be considered.Footnote ⁹⁴

Why do English courts apply forum non conveniens? It is sometimes said that English courts, in staying proceedings in favour of foreign courts, are motivated by a desire to extend ‘judicial comity’.Footnote ⁹⁵ Yet, ‘comity’ here cannot involve deferring to foreign courts for purely diplomatic or prudential reasons.Footnote ⁹⁶ As many have observed, a suggestion by an English court that a foreign court should hear a case instead may reasonably be construed by the latter as an unwelcome instruction rather than politeness and respect.Footnote ⁹⁷ Instead, ‘comity’ involves the English court deferring to a foreign court only because the latter is better placed to adjudicate on a particular case.Footnote ⁹⁸ Deference is thus accorded to the foreign court not for its own sake, but because that foreign court is the ‘appropriate forum’. Conversely, when the English court is itself the ‘appropriate forum’, no deference is accorded.

The purpose of forum non conveniens is thus to ensure that trial occurs in the appropriate forum or, in other words, to achieve appropriate dispute resolution. ‘Appropriateness’ is a term of art. In Spiliada Maritime Corporation v Cansulex Ltd, Lord Goff said that the appropriate forum for a dispute is where it might be tried ‘more suitably for the interests of all the parties and the ends of justice’.Footnote ⁹⁹ And in Lubbe v Cape plc, Lord Hope said that the appropriate forum cannot be determined by considering matters of ‘public interest or public policy which cannot be related to the private interests of any of the parties or the ends of justice in the case which is before the court’.Footnote ¹⁰⁰ The impression that emerges is that ‘appropriateness’ refers to private justice, or ‘justice as between the parties’,Footnote ¹⁰¹ rather than justice in service of broader public interests. To be more precise, ‘appropriateness’ is measured in three particular senses. First, an appropriate court is effective: it should have the power to facilitate its own proceedings and render decisions which determine parties’ rights and liabilities. Second, an appropriate court is convenient: proceedings before it should be quicker and incur less expense for litigants than alternatives. Third, an appropriate court will resolve the dispute accurately on the merits: the resulting decision should be more correct on the facts and the law than alternatives.

That forum non conveniens assesses appropriateness in these three senses—effectiveness, convenience and accuracy—is confirmed by the content of the relevant factors considered under forum non conveniens. The location of witnesses and evidence is relevant because documentary and testimonial evidence is more easily and accurately understood in their original language,Footnote ¹⁰² and because unwilling witnesses can only effectively be compelled to testify where they are located.Footnote ¹⁰³ The connections of parties to competing fora is relevant only insofar as both parties are resident in the same jurisdiction, given the ease of proceeding thereFootnote ¹⁰⁴ and, relatedly, the likelihood of proceedings culminating in a trial on the merits.Footnote ¹⁰⁵ The governing laws of the issues in dispute are relevant because local judges will determine their law’s content more efficiently and accurately than foreign judges.Footnote ¹⁰⁶ Related or parallel proceedings are inconvenient because they multiply expenses, would compromise accuracy since no court will have the ‘whole picture’ of the dispute and might undermine the effectiveness of resulting judgments given the possibility of conflicting judgments.Footnote ¹⁰⁷ The fact that one party would not receive a fair trial in a given court, because that court would be biased against it or because that party would lack means to properly run his case there, obviously demonstrates that that court would not decide accurately on the merits.

However, while it is true that forum non conveniens does not consider broader public interests as relevant factors in identifying the appropriate forum, it would be inaccurate to say that forum non conveniens as a doctrine does not serve important public policies. Instead, the very pursuit of appropriate dispute resolution in itself furthers at least three important public interests. Appropriate dispute resolution helps to alleviate forum shopping, which avoids the risk of parallel proceedings and inconsistent results with negative rule of law consequences.Footnote ¹⁰⁸ Moreover, forum non conveniens’ second stage also ensures that procedural injustice is not perpetrated against litigants, which upholds common law rules of natural justiceFootnote ¹⁰⁹ and the right to a fair trial.Footnote ¹¹⁰ Most fundamentally, by allocating disputes to appropriate fora and keeping them away from inappropriate fora, forum non conveniens facilitates the efficient use of judicial resources. It ensures that English courts try only those disputes that they are well-equipped to adjudicate, which spares judicial time and energy for cases that should be tried in the UK on the merits. This last-mentioned public interest featured as a ‘crucial point’ in Spiliada itself, where Lord Goff reasoned that, in the circumstances of that particular dispute, proceeding in the UK ‘would contribute to efficiency, expedition and economy … [and] to promoting a possibility of settlement’.Footnote ¹¹¹ This was not merely a matter of ‘financial advantage’ to the litigants but, rather, was ‘in the objective interests of justice’.Footnote ¹¹²

Some may deny that forum non conveniens serves public interests in addition to private interests, because the doctrine is not applied when neither party invokes it.Footnote ¹¹³ But that latter feature of forum non conveniens can be explained on grounds that, in adversarial legal systems, the appropriate forum cannot usually be accurately identified unless litigants supply arguments or evidence suggesting why one forum is more appropriate than another. The public interests that forum non conveniens serves, then, are simply public interests which can only be effectively furthered if some litigant chooses to pursue his own private interest in appropriate dispute resolution.Footnote ¹¹⁴

4.2. Agreements and appropriateness

Thus, forum non conveniens facilitates appropriate dispute resolution—effective, convenient and accurate dispute resolution—which in turn furthers public interests like the efficient use of judicial resources, the avoidance of forum shopping and the prevention of procedural injustice. Having established this, it may now be asked: why should jurisdiction agreements be given effect under forum non conveniens?

The answer is that parties who enter into jurisdiction agreements will generally be better placed than the English court to determine the appropriate forum for trial of their disputes. After all, the appropriate forum can only be identified once the ‘nature and shape of the dispute’ is properly understood,Footnote ¹¹⁵ which in turn requires an appreciation of the relevant facts and parties’ legal arguments. An English judge who adjudicates a dispute must typically work hard to understand the nature and shape of that dispute. However, sophisticated commercial parties (and their legal advisors), taken together, will typically already have a good idea of the facts they will adduce and the legal arguments they will make in any dispute foreseeable at the time they enter into a jurisdiction agreement. It thus seems natural to assume that the parties will generally be better-placed to determine the court which can most appropriately (that is, effectively, conveniently and accurately) try their dispute.

This, then, is the reason why courts should generally give effect to jurisdiction agreements under forum non conveniens. As Alex Mills has argued, another potential justification for ‘party autonomy in choice of forum’ is that it:

What matters here is not that jurisdiction agreements are enforceable contracts; they usually are, but that is beside the point. Instead, the important point is that jurisdiction agreements reflect parties’ well-informed view that their chosen court is the appropriate forum, which means that giving effect to the agreement facilitates appropriate dispute resolution—precisely the goal that forum non conveniens is meant to achieve.

One might, however, question whether jurisdiction agreements do indeed reflect parties’ well-informed views that their chosen court is the appropriate forum. It should be emphasised that the argument here is only that jurisdiction agreements will generally reflect such views, not that they invariably will. For reasons explored in Section 5, in exceptional situations where parties are not well-informed on, or otherwise clearly wrong about, the appropriateness of their chosen court, their agreement should not be given effect. But perhaps even the general assumption that jurisdiction agreements reflect parties’ well-informed views on appropriateness might be doubted. These doubts may take three forms.

It might first be said that parties may have broader views on what constitutes an appropriate forum for trial than the English court. Parties may view their chosen court as appropriate for additional reasons beyond effectiveness, convenience and accuracy. This may be accepted. It just means that parties view their chosen court as, among other things, the most appropriate forum for trial, which does not affect the conclusion that jurisdiction agreements should be given effect under the forum non conveniens analysis.

It might then be said that parties may choose a court for reasons entirely unrelated to its appropriateness as a forum for trial. Perhaps one of the parties might ‘price in’ a choice of forum that favours their counterparty, in return for the latter agreeing to undertake other obligations for the former’s benefit. This, however, may not be a credible explanation for parties’ motivations when entering into symmetric jurisdiction agreements—the paradigm with which this enquiry is concerned—since these impose the same benefits and burdens on both parties. And even for asymmetric agreements, the ‘pricing in’ hypothesis is suspect. Dispute resolution clauses are often seen as ‘non-price’ terms, which ‘cannot realistically be traded off against price terms, but rather present a distinct set of provisions that are bargained for separately’.Footnote ¹¹⁷

Finally, it might be said that, since jurisdiction agreements are often boilerplate clauses, they are unlikely to reflect parties’ actual considered views on their chosen court’s appropriateness. However, at least for jurisdiction agreements contained in contracts between parties of equal bargaining power,Footnote ¹¹⁸ this may be inaccurate. For starters, the assumption that jurisdiction agreements are usually boilerplate clauses is doubtful. While some studies suggest that contract drafting precedents may heavily influence parties’ choice of court,Footnote ¹¹⁹ others suggest that that choice often reflects parties’ genuine preferences.Footnote ¹²⁰ Moreover, even if parties do often use boilerplate jurisdiction agreements, that may well reflect a genuine choice to adhere to prevailing industry standards, based on a considered belief that those standards generally reflect parties’ interests in appropriate dispute resolution. This seems a sound decision-making tactic, especially if a precise prior assessment of the costs and benefits of alternative fora would be difficult to make.Footnote ¹²¹ Parties may thus choose a court on the basis of its popularity among other commercial parties, because they rationally believe that popularity is evidence of the court’s appropriateness as a forum for trial of commercial disputes.

4.3. Giving weight

If it is accepted that forum non conveniens’ underlying logic of appropriate dispute resolution requires courts to give effect to jurisdiction agreements under that test, the final question is: how should these agreements be given effect? A court cannot categorically defer to parties’ choice of court, for two reasons. First, since forum non conveniens serves certain public interests in appropriate dispute resolution, a court which defers unquestioningly to parties’ pre-existing contractFootnote ¹²² when exercising that discretion would be abdicating its function as guardian of those interests. That is the whole logic of the ouster objection. Second, and more importantly, a court which always defers to parties’ jurisdiction agreements runs the risk of making wrong decisions in those cases where parties, well-informed though they might generally be, are exceptionally and obviously wrong about the appropriateness of their chosen court.

Nevertheless, a court in applying forum non conveniens may still legitimately give effect to a jurisdiction agreement in a more qualified manner. That is: the court may treat that jurisdiction agreement, reflecting the parties’ view that their chosen court is the appropriate forum for trial, as a weighty piece of evidence or a weighty factor, to be considered alongside all other evidence or factors, in the court’s process of reaching its own answer. The notion that one should accord weight to the views of others, on matters where those others are more epistemically well-equipped, is commonsensical, and is also an approach that courts routinely adopt, for example, when relying on the testimony of experts at trial.Footnote ¹²³ The primary responsibility for reaching its own conclusions remains with the court, but the court can and should treat the views of those more well-informed than it as a weighty piece of evidence in reaching its own conclusion. However, the court can and should depart from expert evidence which is obviously wrong, as may be apparent from the fact that the other evidence before the court is overwhelmingly to the contrary. This is because the expert is only there to assist the court, and while their view is usually persuasive, it is not authoritative, and the court must disagree when it feels compelled to.

5.1. Even-handedness

The strong cause test applies even-handedly, that is, to both English and foreign jurisdiction agreements governed by either English or foreign law. The frustration account struggles to explain this. On that account, the strong cause test should only apply when the jurisdiction agreement is governed by English law or a materially similar foreign law, which will not necessarily be the case even for English jurisdiction agreements, since these are usually governed by the law of the main contract.Footnote ¹²⁴ The remedial discretion account fares better, since English private international law may still view remedies as governed by the lex fori.Footnote ¹²⁵ Yet, it cannot explain why the strong cause test also operates even-handedly in many other common law jurisdictions where remedies are generally governed by the lex causae, like Canada,Footnote ¹²⁶ SingaporeFootnote ¹²⁷ and Hong Kong.Footnote ¹²⁸

By contrast, on the forum non conveniens account, the strong cause test’s even-handedness is self-explanatory.Footnote ¹²⁹ The exercise of jurisdiction is always a question for the lex fori, and forum non conveniens is always applied with the same factors considered. A jurisdiction agreement should generally be one of those relevant factors, because it generally reflects parties’ well-informed views about the appropriateness of their chosen court. The jurisdiction agreement’s governing law is irrelevant here: that informs the agreement’s enforceability and the remedies that flow from its breach, but not the test applied to determine the exercise of jurisdictional discretion.

5.2. Convenience

Under the strong cause test, ‘convenience’ factors can generally amount to strong reasons only if unforeseeable. The frustration account can explain this general position, but not the exception where foreseeable inconvenience which is ‘especially strong’ establishes strong cause.Footnote ¹³⁰ The qualified contractual account fares similarly, since holding parties to foreseeable but extremely inconvenient outcomes which they bargained for would respect rather than undermine their contractual bargain. The remedial discretion account fails to explain the relevance of foreseeability to the ‘convenience’ factors at all. Whether the inconvenience of contractual performance is grounds to deny specific relief depends on its gravity—that is, whether it amounts to excessive hardshipFootnote ¹³¹—rather than its foreseeability.

On the forum non conveniens account, by contrast, it is clear why ‘convenience’ factors will typically, but not invariably, establish strong cause only if unforeseeable. Forum non conveniens generally gives significant weight to parties’ views on their chosen court’s appropriateness because parties are generally more familiar with the facts and legal arguments relevant to any subsequent disputes between them. However, parties’ actual familiarity with the facts and legal arguments relevant to their current dispute will depend on the extent to which this actual dispute was foreseeable when they contracted. Insofar as it was unforeseeable, parties’ views on appropriateness should count for less in the forum non conveniens analysis.

‘Convenience’ factors—the location of witnesses and evidence, and the governing law—are precisely those factors which are relevant under forum non conveniens because they concern the facts and legal arguments relevant to a dispute. To say that ‘convenience’ factors were foreseeable is thus to say that parties were probably familiar with the facts and legal arguments relevant to their current dispute when they contracted. If so, parties’ views on appropriateness should be given significant weight, and this will usually outweigh the ‘convenience’ factors themselves in the forum non conveniens analysis. Conversely, to say that ‘convenience’ factors were not foreseeable is to say that parties were probably unfamiliar with the facts and legal arguments relevant to their current dispute when they contracted. If so, parties’ views on appropriateness should not be given significant weight, and so will not usually outweigh the ‘convenience’ factors in the forum non conveniens analysis.Footnote ¹³²

However, it should be noted that on the forum non conveniens account ‘convenience’ factors, foreseeable or otherwise, do not cease to be relevant factors under the forum non conveniens enquiry, independent of the jurisdiction agreement. That is because the jurisdiction agreement, representing parties’ views on appropriateness, is still only one (weighty) factor informing the court’s view on appropriateness, to be weighed alongside all the other relevant factors under the forum non conveniens analysis.Footnote ¹³³ This should mean that, even where ‘convenience’ factors were foreseeable, it should theoretically remain possible (albeit improbable) that they might outweigh the jurisdiction agreement. And this explains why English courts acknowledge that, although generally foreseeable ‘convenience’ factors cannot establish strong reasons, they still may do so in ‘rare’ situations of extreme foreseeable inconvenience.Footnote ¹³⁴

5.3. Fair trial

The prospect of an unfair trial in the chosen court will generally amount to strong reasons, even if that prospect was foreseeable at the time the jurisdiction agreement was entered into. The frustration account cannot explain this, since foreseeable events rarely frustrate contracts.Footnote ¹³⁵ The qualified contractual account faces the same problem, for it would respect the parties’ bargain to send them to their chosen court, fair or otherwise. The remedial discretion account fairs better, since an unfair trial probably counts as excessive hardship justifying the denial of a specific order to subject oneself to such a trial.

The forum non conveniens account squarely explains why the ‘fair trial’ factor should typically, but not invariably, establish strong cause even if the prospect of an unfair trial in the chosen court was foreseeable. An unfair trial is, by definition, inconvenient for one party and inaccurate for all parties, and so it is strong evidence that the relevant court is an inappropriate forum. This means that the ‘fair trial’ factor should always be a weighty factor militating against a conclusion that the chosen court is the appropriate forum.

What logically follows from this? It does not follow that English courts should now treat jurisdiction agreements as having negligible weight in the forum non conveniens analysis, if parties could reasonably foresee when contracting that they would likely get no fair trial in their chosen court. To the contrary, if parties could foresee that risk, then their view that their chosen court remains the appropriate forum would have been well-informed by that risk, and should still be given significant weight. However, what does follow is that, even if the jurisdiction agreement remains a weighty factor, it should still be (more than just theoretically) possible for it to be outweighed by the ‘fair trial’ factor, because the ‘fair trial’ factor is itself a weighty factor. And this explains why English courts suggest that the ‘fair trial’ factor, even if foreseeable, can amount to strong reasons by itself.

Yet, even if the ‘fair trial’ factor might outweigh the jurisdiction agreement, the converse—that the agreement might outweigh the ‘fair trial’ factor—remains a distinct possibility as well. Everything will depend on the weight of the jurisdiction agreement, which should in turn depend on the extent to which the prospect of an unfair trial in the chosen court was foreseeable to the parties. And that explains why English courts also acknowledge that the chance of establishing strong reasons on the ‘fair trial’ factor will be ‘reduced’ if parties foresaw the actual risk that they would not get a fair trial in the chosen court.Footnote ¹³⁶

5.4. Abuse of process

Courts will refuse to give effect to foreign jurisdiction agreements on the ground of ‘abuse of process’ only when the defendant seeks a stay of English proceedings despite conceding both liability and quantum. The frustration account cannot explain this, since anyone who foresees litigation with one’s contracting counterparty must certainly foresee the possibility that that counterparty may sue frivolously. The remedial discretion account may seem more plausible, since specific remedies may be withheld to prevent equity from acting in vain. Yet, a stay of English proceedings, despite a concession of both liability and quantum, may not be a fruitless remedy: it may be that losing in the foreign court holds certain advantages for the defendant unrelated to the content of his liability to the claimant.Footnote ¹³⁷ The qualified contractual account fares worst here. On that view, a defendant should be entitled to insist on a stay of English proceedings on the basis of a foreign jurisdiction agreement, even if the writing is manifestly on the wall, since that is his contractual right. After all, ‘parties do not choose a contractual forum for dispute resolution only because they want to win in that forum; they choose a forum because they want to win or lose in that forum’.Footnote ¹³⁸

By contrast, on the forum non conveniens account, the content of the ‘abuse of process’ factor is easily explained. Once the defendant seeking the stay has conceded both liability and quantum, it becomes irrelevant that the foreign chosen court was more appropriate for trial of the parties’ dispute, or that the parties held that view when entering into the jurisdiction agreement. There is now literally no dispute, nothing by way of appropriate dispute resolution to be gained by staying English proceedings in favour of the foreign appropriate forumFootnote ¹³⁹ and, thus, no justification for giving effect to the foreign jurisdiction agreement.

5.5. Related proceedings

Related proceedings in a non-chosen court will generally be insignificant if they involve only the parties to the jurisdiction agreement, but will likely amount to strong reasons if they involve third parties even if foreseeable. The frustration and qualified contractual accounts cannot explain this, at least when the related proceedings are foreseeable. The remedial discretion account also falls short. It is true that English courts may deny specific remedies to protect third-party interests.Footnote ¹⁴⁰ However, where strong reasons are established on the basis of related proceedings involving third parties, the English court is not necessarily acting to safeguard the interests of those third parties, who may be facing trial in those related proceedings in any event. Instead, strong reasons may also be established when giving effect to the jurisdiction agreement would force one of the contracting parties to litigate in the chosen court against their counterparty, while also participating in related proceedings in a non-chosen court against the third parties—and here, the court’s disregard of the jurisdiction agreement furthers that contracting party’s interests, not those of third parties.Footnote ¹⁴¹

The forum non conveniens account, however, explains exactly why related proceedings are treated as they are. Where related proceedings involve only parties to the jurisdiction agreement, it is difficult to see these as materially different from other ‘convenience’ factors—at most, they compromise the accuracy and effectiveness of adjudication. Thus, the prospect of these related proceedings, even when foreseeable, will rarely establish strong cause, for reasons explored above relating to ‘convenience’ factors.Footnote ¹⁴²

By contrast, where the related proceedings in the non-chosen court involve third parties to the jurisdiction agreement, it makes sense that this should be capable of establishing strong cause by itself, even when those proceedings are foreseen by parties to the jurisdiction agreement. Forum non conveniens is concerned not only with the appropriateness of trial for contracting parties, but for all parties to the ‘larger, complex, dispute’.Footnote ¹⁴³ And when there are related proceedings involving third parties, this means that the issues in dispute between the contracting parties will either be identical to those in dispute between the contracting parties and third parties, or the former will depend on the resolution of the latter or vice versa. All proceedings will thus form part of the same broader dispute. Now, if a jurisdiction agreement is a weighty factor under forum non conveniens because it reflects the parties’ well-informed views on the appropriate forum for trial of the dispute, when not all parties to the broader dispute are parties to the agreement, that agreement should be given far less weight. This will then lead to the conclusion that the most appropriate forum for the broader dispute is simply the court which can receive the best evidence and arguments from all (or at least most of)Footnote ¹⁴⁴ the parties to the broader dispute, since that is the court which can ‘make a reliable, comprehensive judgment on all the matters in issue’.Footnote ¹⁴⁵

5.6. Unequal bargaining power

When a jurisdiction agreement is between parties of unequal bargaining power, strong reasons seem easier to establish. The frustration and qualified contractual accounts struggle here, at least for English jurisdiction agreements, since English contract law routinely enforces contracts between parties of unequal bargaining power. The remedial discretion account also faces difficulties, as it is the hardship eventuated by specifically performing an obligation, rather than the quality of the consent in undertaking it, that is relevant to the exercise of remedial discretion.Footnote ¹⁴⁶

On the forum non conveniens account, however, the relevance of unequal bargaining power becomes clear. A jurisdiction agreement should be a weighty factor under forum non conveniens because it reflects the parties’ collective, well-informed view on their chosen court’s appropriateness. That, however, assumes that all contracting parties will generally have an equal ability to form and express their own views on that matter. Conversely, when the parties’ abilities in that regard are unequal, any contract between them would simply reflect the stronger party’s self-serving interests rather than a collective, well-informed judgment on the chosen court’s appropriateness.Footnote ¹⁴⁷ Thus, a jurisdiction agreement’s weight under forum non conveniens should logically decrease—and strong cause should logically be easier to establish—the more parties lacked equal ability, in terms of knowledge or opportunity, to form and contribute well-informed views on their chosen court’s appropriateness.

5.7. Public policy

English courts have never found strong reasons to disregard jurisdiction agreements on the basis of public policy, despite obiter comments to the effect that jurisdiction agreements may be denied effect on public policy grounds. The frustration account sheds no light on this issue, public policy being a non-sequitur in that context. Under the remedial discretion account, by contrast, public policy considerations should sometimes amount to strong reasons, since English courts do sometimes consider public interests in the exercise of remedial discretion.Footnote ¹⁴⁸ The qualified contractual account would also support a broader role for public policy. If the court’s jurisdictional discretion may be influenced by the parties’ contractual bargain, understood here as a consideration external to the logic of forum non conveniens, there seems no reason why that discretion cannot be influenced by other external considerations like fundamental public policies.

On the forum non conveniens account, by contrast, public policy should have no role under the strong cause test. After all, forum non conveniens ignores any ‘public interest or public policy which cannot be related to the private interests of any of the parties or the ends of justice in the case which is before the Court’.Footnote ¹⁴⁹ If the strong cause test is just forum non conveniens, public policy should also be irrelevant thereunder. This, however, does not mean that jurisdiction agreements cannot be denied effect on grounds of public policy—only that that should be viewed a reason for non-enforcement which is separate from the strong cause test. Courts may still cite public policy to deny enforcement of an English jurisdiction agreement under the English doctrine of illegality, and may do the same for foreign jurisdiction agreements under private international law’s public policy exception.Footnote ¹⁵⁰

6.1. Non-exclusive jurisdiction agreements

Must a party to a non-exclusive jurisdiction agreement establish strong reasons to stay proceedings in, or to bring proceedings outside, the non-exclusively chosen court? The law on point is not the picture of clarity but, in general, the answer seems to be ‘no’. Instead, a non-exclusive agreement is simply one factor in favour of the non-exclusively chosen court, to be considered alongside other factors in the forum non conveniens analysis. Thus, despite earlier decisions to the contrary,Footnote ¹⁵¹ the modern position with regards to English non-exclusive agreements is reflected in Toulson LJ’s comments in Deutsche Bank AG v Highland Crusader Offshore Partners LP:

A similar position for foreign non-exclusive agreements is captured by Colman J’s statement in BP plc v Aon Ltd that:

Are English courts correct to treat all non-exclusive agreements in this way—as factors leaning against, but not as requiring strong reasons for, a stay of proceedings in or the bringing of proceedings outside the non-exclusively chosen court? Canadian courts have adopted a similar position.Footnote ¹⁵⁴ However, courts in SingaporeFootnote ¹⁵⁵ and Hong KongFootnote ¹⁵⁶ have treated forum non-exclusive agreements differently from foreign non-exclusive agreements: the former imposes a requirement of strong reasons for stays of forum proceedings, but the latter does not do the same for the institution of forum proceedings. Which approach is correct? The answer here ultimately turns on the proper understanding of why courts enforce jurisdiction agreements, and what the strong cause test represents in that analysis.

Under the frustration and remedial discretion accounts, the strong cause test functions as a full or partial excuse for breach of contract, and under the qualified contractual account, the test reflects a general inclination to give effect to the parties’ contractual bargain. Either way, parties’ contractual obligations are the paramount consideration. And so, whether strong reasons must be shown by the litigant who seeks to stay or bring English proceedings reduces to this question: what obligations do non-exclusive agreements impose, and are they breached by the litigant’s application?Footnote ¹⁵⁷

In this regard, it has been said that while exclusive agreements impose both a positive obligation to submit to the chosen court and a negative obligation not to sue elsewhere, non-exclusive agreements impose only the positive obligation to submit to the chosen court. Thus, where the English court is non-exclusively chosen, an application to stay English proceedings does breach the positive obligation to submit, and strong reasons are required to obtain a stay.Footnote ¹⁵⁸ By contrast, when a foreign court is non-exclusively chosen, the maintenance of English proceedings does not breach the positive obligation to submit, so strong reasons are not required to maintain them.Footnote ¹⁵⁹ This differential treatment—the English non-exclusive agreement requires strong reasons to stay English proceedings, but the foreign non-exclusive agreement seems legally irrelevant to the exercise of the English court’s jurisdiction—inevitably carries ‘unwelcome overtones of parochialism’.Footnote ¹⁶⁰ Yet, it is the logical consequence of focusing purely on parties’ contractual obligations.Footnote ¹⁶¹

Under the forum non conveniens account, however, the reasoning changes. Now, the proposition that a non-exclusive jurisdiction agreement imposes a positive obligation to submit to the non-exclusively chosen court is irrelevant or at least inconclusive. Instead, the important questions are: is the non-exclusively chosen court the appropriate forum for trial? And what views did the parties express on that?

Properly understood, when the parties make a non-exclusive choice, they are expressing the view that their chosen court, but not only that court, can appropriately try their dispute. In other words, the parties view their chosen court as ‘an appropriate forum’,Footnote ¹⁶² not the appropriate forum, for trial. The proper judicial response should then be to treat a non-exclusive jurisdiction agreement as simply one factor suggesting that the non-exclusively chosen court is the appropriate forum under the forum non conveniens analysis. Thus, the litigant seeking a stay of English proceedings despite there being an English non-exclusive agreement need not show strong reasons to depart from the positive obligation to submit, since that obligation will only be given effect when the English court is indeed the appropriate forum. Similarly, the litigant arguing for the exercise of English jurisdiction in the face of a foreign non-exclusive agreement need not establish strong reasons to secure that, since the English court will exercise jurisdiction so long as it is the appropriate forum. Yet, in both situations, the English court’s assessment of the appropriateness of competing courts will be informed by the parties’ view that their chosen court is appropriate, which goes some way toward establishing that it is more appropriate than the non-chosen court.

6.2. Contractual anti-suit injunctions

When applicants seek ‘contractual anti-suit injunctions’ to enforce English jurisdiction agreements, they only need show that the foreign proceedings sought to be restrained are indeed brought in breach of an enforceable English jurisdiction agreement.Footnote ¹⁶³ These are contrasted with what are commonly termed ‘non-contractual anti-suit injunctions’, issued to restrain foreign proceedings which are ‘vexatious and oppressive’.Footnote ¹⁶⁴ Both contractual and non-contractual injunctions are said to require proof of some kind of wrongdoing,Footnote ¹⁶⁵ like a breach of contract (a jurisdiction agreement) or an equitable or procedural wrong (vexatious foreign proceedings).Footnote ¹⁶⁶ However, while non-contractual injunctions require the applicant to show that the English court would be the appropriate or natural forum for trial,Footnote ¹⁶⁷ contractual injunctions will issue unless there are strong reasons not to enforce the English jurisdiction agreement.

Why the apparent difference? In the context of non-contractual injunctions, Lord Goff held in Airbus Industrie GIE v Patel that ‘comity’ requires the English issuing court to be the appropriate forum, ‘to justify the indirect interference with the foreign court which an anti-suit injunction entails’.Footnote ¹⁶⁸ By contrast, for contractual injunctions, courts have said that they are only seeking ‘to secure compliance with the contractual bargain’,Footnote ¹⁶⁹ and that ‘[c]omity … has little if any role to play’.Footnote ¹⁷⁰ But this is unconvincing: foreign courts often take offence at English contractual anti-suit injunctions,Footnote ¹⁷¹ and English courts respond in kind ‘when the boot is on the other foot’.Footnote ¹⁷² Yet, if comity does remain relevant in contractual injunction cases, should it not impose the same requirement as in non-contractual injunction cases, that the English issuing court be the appropriate forum for trial?

It is submitted that comity does, in fact, impose the same requirement: the English issuing court must always be the appropriate forum for trial. Yet, in contractual injunction cases, this requirement will always be satisfied when the English court finds no strong reasons to disregard an English jurisdiction agreement. After all, on the forum non conveniens account of the strong cause test, that jurisdiction agreement reflects the parties’ well-informed view that the English court is the most appropriate forum for trial, which should be a weighty factor under forum non conveniens. So, where an English court finds no strong reasons not to enforce an English jurisdiction agreement, the court is still asking itself whether the UK is the appropriate forum for trial—just that the court also ultimately agrees with the parties’ view that it is.

This analysis may seem purely academic for contractual injunctions to enforce English jurisdiction agreements. Yet, it has real consequences for the issuance of contractual injunctions to enforce foreign jurisdiction agreements by restraining proceedings in a third State. CourtsFootnote ¹⁷³ and commentatorsFootnote ¹⁷⁴ generally recognise that courts should not grant such contractual anti-suit injunctions in aid of foreign proceedings. But why not, if comity is not a concern, and the English court need not be the appropriate forum before contractual anti-suit injunctions issue? That logic, it should be noted, is precisely what supported the UK Supreme Court’s decision in UniCredit Bank GmbH v RusChemAlliance LLC. Footnote ¹⁷⁵ There, a German bank obtained an anti-suit injunction restraining a Russian company from maintaining Russian proceedings brought in breach of a Parisian arbitration agreement. Lord Leggatt reasoned that, while the UK had to be the ‘proper place in which to bring the claim’, it did not matter ‘whether a forum other than the English court is more appropriate or suitable for the trial of the [underlying] action.’Footnote ¹⁷⁶ Instead, ‘[t]he basic principle applied is “pacta sunt servanda”’.Footnote ¹⁷⁷

In fairness, Lord Leggatt did acknowledge in UniCredit that English courts should not issue anti-suit injunctions unless that would be ‘consistent with comity’.Footnote ¹⁷⁸ Yet, that concept was reinterpreted there in a way that denuded it of its original import. ‘Comity’ only meant that the English anti-suit injunction should not ‘clash with any exercise of [supervisory] jurisdiction by the French courts’, which would not occur on the facts, since French courts could not issue anti-suits to assist Parisian arbitrations.Footnote ¹⁷⁹ The analogy here seems to be with cases where English courts are asked to issue freezing injunctions to aid foreign proceedings,Footnote ¹⁸⁰ in which context ‘comity’ only requires that such aid should not interfere with the main proceedings.Footnote ¹⁸¹ In other words, in UniCredit and the freezing injunction cases, ‘comity’ refers to a concern about interference with the proceedings of the court assisted (here, the Parisian court and tribunal), not, as in Airbus v Patel, those of the court restricted (here, the Russian court). But this analogy is inapposite,Footnote ¹⁸² for in freezing injunction cases, there is usually only a court assisted, while in anti-suit injunction cases, there is always also a court restricted. Reinterpreting ‘comity’ in this way side-steps, rather than addresses, the concerns that anti-suit injunctions raise about interfering with foreign proceedings in the court restricted.

The tenor of the post-UniCredit literature is thus that the decision should be viewed with suspicion and confined to the arbitration context.Footnote ¹⁸³ To do this, English courts must affirm that ‘comity’ comports a requirement that the English court be the appropriate forum for trial before it issues any anti-suit relief, contractual or non-contractual. Contractual injunctions to enforce foreign jurisdiction agreements would then usually be unavailable, because those agreements suggest that a foreign court is the appropriate forum for trial.Footnote ¹⁸⁴ Contractual injunctions to enforce English jurisdiction agreements would, however, remain readily available. This is because, on the forum non conveniens account, when an English court finds no strong reasons to depart from an English jurisdiction agreement, it is agreeing with parties’ view that it is the appropriate forum for trial.