I. INTRODUCTION
It has become common practice in international business transactions to draft broadly worded jurisdiction agreements submitting all disputes arising between the parties to a predetermined forum.Footnote 1 Generally, business actors and legal practitioners appreciate the benefits of these agreements, on the one hand, in minimising the risk of litigation by reducing uncertainty over questions of jurisdiction and, on the other hand, in preventing parallel proceedings by channelling all possible claims to the same court. These expectations may nonetheless be disappointed if one of the parties successfully argues that the claim grounding the jurisdiction of the court has no connection with the contract between the parties but arises from a different legal relationship. This occurs in particular when non-contractual claims are litigated.Footnote 2 The EU law provisions regulating jurisdiction agreements address this issue by requiring there to be a connection with ‘a particular legal relationship’ in order for the court to have jurisdiction. This article seeks to clarify the content and implications of this ‘proximity’ requirement,Footnote 3 a subject that remains quite obscure in both academic literature and legislation.Footnote 4
Section II distinguishes the notion of proximity from the scope of the subject matter of a jurisdiction agreement. When it comes to addressing the scope of a jurisdiction agreement, both scholarly works and practitioners’ collections of clauses search deeply for the most appropriate interpretation of the agreement's scope, but tend to ignore the important questions of the existing limits to party autonomy and of their impact on the enforceability or effectiveness of such an agreement. This section also makes extensive reference to the recent judgments of the Court of Justice of the EU (CJEU), in particular Apple v eBiscuss.com,Footnote 5 where it was decided that the appropriate standard for determining claim proximity was whether the party against whom the jurisdiction agreement was being invoked would be ‘taken by surprise’ by its being heard by that court, on the basis that the claim arose from a relationship other than that in connection with which the jurisdiction agreement was entered into. In CJEU practice, this foreseeability test has become the standard approach to determining claim proximity, which the Court derived from the general principle of predictability of jurisdiction.
Section III sets out the main operational aspects of the foreseeability test. It clarifies the methodology supporting it and then considers its replicability outside the context of competition law infringements and its operability in more complex cases. Additionally, it considers possible alternative tests which nonetheless have been rejected in judicial practice. This analysis finds that, in order to ensure that jurisdiction is sufficiently predictable, courts may be required, depending on the complexity of the case, to engage in in-depth fact-finding by examining the substance of the claim raised by the parties, making proximity not only crucial but also potentially difficult to assess in determining jurisdiction.
II. THE PROXIMITY REQUIREMENT
A. Distinguishing Proximity from Consent
Consider the following four examples of broadly worded jurisdiction agreements:
The parties agree to bring any claims for breach of this contract exclusively in the courts of England and Wales [Clause A].
The Courts of England and Wales have exclusive jurisdiction to settle any dispute arising out of or in connection with this agreement (including a dispute regarding the existence, validity or termination of this agreement) [Clause B].
The parties agree to submit any and all disputes of whatever nature which arise between them exclusively to the courts of England and Wales [Clause C].
Jurisdiction: England and Wales [Clause D].Footnote 6
Agreements of this kind typically raise two distinct questions: (1) which disputes fall within their respective scopes; and (2) whether they are also effective in allocating jurisdiction.
The first question involves considering the purported scope of the jurisdiction agreement in order to verify whether the parties intended that a particular claim, on which they have subsequently centred their litigation, should fall within it. Scholars tend to address questions such as this when considering the theoretical grounds for examining the scope of a jurisdiction agreement—a question of interpreting the jurisdiction agreement and party consent. The second question concerns whether the party autonomy regime established by the applicable jurisdictional rules allows the parties to litigate a particular claim in the designated forum.Footnote 7 This is a question of effectiveness or workability that ultimately relates to the interpretation not of the jurisdiction agreement, but of the applicable jurisdictional rules. As each of these questions generates multiple issues, they deserve to be addressed separately.
1. The subject-matter scope of jurisdiction agreements
The question of the agreement's ‘purported’Footnote 8 scope is normally addressed using the common canons of contract construction under the law governing the jurisdiction agreement. General scholarly works and practitioners’ collections of clauses usually explain that the determination of the scope of a jurisdiction agreement is a matter of construction of the contract and interpretation of the parties’ intentionFootnote 9—an approach that courts also seem to follow.Footnote 10
More specifically, when it comes to determining whether or not the broadly worded scope of a jurisdiction agreement encompasses a non-contractual claim (eg, for negligent misstatement or misrepresentation, fraud or bribery), English courts for a long time tended to rely on a literal construction of the agreement focussing on connectors such as ‘arising under’, ‘arising out of’, ‘relating to’, or ‘in connection with’ a relevant contract.Footnote 11 Scholarly analyses of court rulings show that each of these connectors has a different meaning regarding non-contractual claims.Footnote 12 Accordingly, Clause A and Clause B above seem to be narrower in scope than Clause C and Clause D, which contain no connector and may therefore encompass claims that do not have even a minimal connection with the contract.
A distinction can therefore be drawn between ‘narrow’ and ‘wide’ scope of jurisdiction agreements.Footnote 13 At one end of the spectrum lie jurisdiction agreements which only cover contractual claims (or only some of them, such as Clause A). At the other end one finds jurisdiction agreements encompassing all possible non-contractual claims arising between the parties, even those with no connection whatsoever with the contract (as in Clause C or, less explicitly, Clause D).Footnote 14 Expressions such as ‘arising under’ appear to lie at one extreme, courts generally considering this wording to be narrower than ‘arising out of’,Footnote 15 whereas expressions such as ‘arising in connection with’ or ‘in relation to’ have been considered to be ‘words of the very widest scope’.Footnote 16
Any exercise aimed at measuring the inclusiveness of a jurisdiction agreement based on the use of a certain connector risks being worthless. On the one hand, because each case concerns only one or a few very specific non-contractual claims vis-à-vis one particular jurisdiction agreement, it is practically impossible to devise a comprehensive and precise approach. On the other hand, both the governing law and court discretion may lead to different interpretations of clauses drafted in English depending on the quality of the translation.Footnote 17 The result is inevitably approximate. Nonetheless, this exercise has the obvious advantage of allowing practitioners to know whether they have achieved their objective of drafting the widest possible jurisdiction agreement.Footnote 18 Case law may provide indications about whether certain types of claims will more reasonably be covered by certain forms of wording. For example, a rectification claim is likely to be covered by a jurisdiction agreement whose scope extends to all disputes ‘arising out of’ the contract, whereas this would not be the case for a clause using the expression ‘arising under’,Footnote 19 a formula that would instead cover claims of contract voidance following a repudiation of acceptance.Footnote 20 Furthermore, a jurisdiction agreement referring to all disputes arising ‘in relation to’ a contract is apt to include claims for damages arising from tortious interference with a party's business.Footnote 21
Despite the difference in specificity, it is hard to draw ex post any concrete indications about the intention of the parties from all these ‘linguistic nuances’.Footnote 22 This is the reason underlying the House of Lords' notorious judgment in Fiona Trust v Privalov and its extremely liberal ‘one-stop adjudication’ doctrine, requiring courts to disregard the particular connector(s) used by the parties in favour of a presumption of inclusiveness.Footnote 23 Essentially, by denying the distinction between narrow and wide jurisdiction agreements, this doctrine replaced the parties’ intention with an assumption of what ‘any sensible businessmen [would] have wished to agree’.Footnote 24 Even this inclusive approach, however, presents some flaws, as not only does it force courts into ‘second-guessing or only half-recognising the decisions of the parties’,Footnote 25 but it is also unlikely to find favour in other EU Member States, whose courts may continue to engage with the linguistic distinctions rejected by Fiona Trust.Footnote 26
Political philosophers have given a name to this sort of problem: ‘the inclusion paradox’. Analogous with the famous ‘paradox of tolerance’, the inclusion paradox reflects the idea that, regardless of one's good intentions, absolute inclusiveness remains nonetheless conceptually problematic.Footnote 27 In the context of jurisdiction agreements, this paradox is expressed by the tension between the parties’ statement that all claims be subject to the designated jurisdiction on the one hand, and the requirement that these claims have a certain connection with the contract on the other hand. In other words, broad language does not necessarily mean full inclusiveness, especially when it comes to non-contractual claims – which is why wide jurisdiction agreements are more often than not considered ‘by definition […] ambiguous’.Footnote 28 Unless more information can be deduced from the relevant contextual circumstances, and provided that this exercise is feasible under the applicable law or the contract,Footnote 29 saying whether or not the parties knew which specific disputes would directly or indirectly result from their relationship is pure speculation.
This information gap explains why it remains unclear whether even the most inclusive jurisdiction agreements (like Clause D, for instance) include non-contractual claims. It also contributes to the courts’ ‘increasing exasperation’ when interpreting the subject-matter scope of jurisdiction agreements, making this a field where ‘the technical value [of] precedents can only be seen as limited, and risks being over-stated’.Footnote 30
2. The question of effectiveness
Discussing a jurisdiction agreement's ‘effectiveness’ or ‘workability’ also requires determining whether its ‘purported’ scope would also be recognised and enforced by the law as a matter of party autonomy.Footnote 31 In this regard, the EU regime of jurisdiction agreements entitles business actors to agree on ‘the court or the courts of a Member State [that] are to have jurisdiction to settle any disputes which have arisen or which may arise in connection with a particular legal relationship’.Footnote 32 Accordingly, for a jurisdiction agreement to be enforceable under the EU regime the claim litigated in court must fulfil a proximity requirement vis-à-vis the legal relationship between the parties.
Despite the various amendments to the EU regime that have occurred over time, this requirement has remained unchanged.Footnote 33 Generally speaking, proximity is connected to the objectives of predictability and certainty of law.Footnote 34 Proximity is a substantive requirement of a jurisdiction agreement and therefore must be distinguished from consent and from the related analysis of the intention of the parties. Borrowing the words of AG Jääskinen from his opinion in Refcomp v Axa, a jurisdiction agreement ‘must not be worded in such a general manner as to include all possible disputes between the parties, irrespective of the contracts concluded by them’.Footnote 35
Two theories have been advanced to explain the rationale behind proximity. The first theory, to which scholars today almost unanimously adhere, argues that proximity must be understood in terms of predictability, limiting the danger of so-called ‘catch-all clauses’—clauses ‘which conferred jurisdiction in relation to any dispute which might arise out of any future legal relationship into which the parties may enter’.Footnote 36 The second theory, set out by AG Giuseppe Tesauro in his opinion in Powell Duffryn v Petereit, links proximity with the need to ‘prevent the party in a stronger bargaining position from imposing on the other party the jurisdiction of any other court’.Footnote 37 This second theory raises the problem of implementing protective public policy considerations outside the context of jurisdictional rules that already protect the weaker party to a contract by limiting party autonomy.Footnote 38
Regardless of the preferred theory, the proximity requirement makes clauses such as C and D problematic, if not totally unenforceable, in respect of non-contractual claims which have no connection with the contract. Both these clauses lack a proper connector with a relevant contract, thus potentially permitting jurisdiction for an unspecified number or type of claims and creating a problem of a lack of proximity. In so doing, they question the autonomy of the parties to contract over jurisdiction. The problem, then, is how to assess proximity in litigation.
B. The Foreseeability Test
To assess a claim's proximity to a particular legal relationship the CJEU has implemented a foreseeability test, which it deemed consistent with the general objective of the EU regime of jurisdiction agreements, namely the predictability of jurisdiction.
1. Early cases
The proximity problem was first addressed in Meeth v Glacetal (1978).Footnote 39 The defendant had raised a set-off defence against the claimant's initial request for payment based on the damage resulting from the alleged delay in delivering goods. The CJEU observed that respect for party autonomy and the need ‘to avoid superfluous procedure’ required the court designated in the jurisdiction agreement to adjudicate ‘a claim for a set-off connected with the legal relationship in dispute’.Footnote 40 This was an easy exercise, however, as the defendant's cross-claim was clearly connected with the contract entered into by the parties.Footnote 41Meeth therefore left unsettled the question of whether there was jurisdiction over an offset of a cross-claim arising from a different legal relationship than the main claim (for example, an alleged infringement of the duty to disclose) which in some legal systems may require separate proceedings.Footnote 42
In Powell Duffryn v Petereit (1992), the CJEU applied the foreseeability test in respect of a broadly worded jurisdiction clause contained in a company's bylaws.Footnote 43 The clause in question, similar to Clause C above, provided:
By subscribing for or acquiring shares or interim certificates the shareholder submits, with regard to all disputes between himself and the company or its organs, to the jurisdiction of the courts ordinarily competent to entertain suits concerning the company.Footnote 44
The CJEU first set out the test for assessing proximity under Article 17 of the 1968 Brussels Convention, stating that the requirement of a connection with a particular legal relationship
is intended to limit the scope of an agreement conferring jurisdiction solely to disputes which arise from the legal relationship in connection with which the agreement was entered into. Its purpose is to avoid a party being taken by surprise by the assignment of jurisdiction to a given forum as regards all disputes which may arise out of its relationship with the other party to the contract and stem from a relationship other than that in connection with which the agreement conferring jurisdiction was made.Footnote 45
The Court therefore concluded that the jurisdiction clause at issue covered all disputes arising out of the relationship between the company and the shareholders as such.Footnote 46 Yet even this test does not seem to provide a straightforward answer in respect of more difficult corporate litigation cases, such as where ‘the shareholder(s) were to take action against the company no longer as shareholders but in another capacity (or the company against them)’.Footnote 47 It should be recalled, additionally, that party autonomy is already heavily limited in the corporate setting, as certain questions concerning the relationship between a company and its shareholders are subject to the exclusive jurisdiction of the court of the Member State where the company has its seat.Footnote 48 Because of these constraints, Powell Duffryn can hardly be qualified as a leading case helpful for the understanding of the problem of claim proximity in other settings.
2. Recent cases regarding competition law claims
In more recent rulings, the CJEU has employed the foreseeability test in the context of claims arising out of competition law infringements.
In CDC Hydrogen Peroxide v Evonik Degussa and Others (2015),Footnote 49 the CJEU was asked to interpret Article 23(1) of the Brussels I Regulation in relation to an unlawful secret cartel.Footnote 50 The lawsuit before the referring court consisted of a follow-on damage action arising out of a cartel agreement between the world's leading manufacturers of hydrogen peroxide and sodium perborate, which the EU Commission had declared unlawful in 2006.Footnote 51 Acting on behalf of dozens of purchasers of hydrogen peroxide, the Belgian company Cartel Damage Claims (CDC) Hydrogen Peroxide brought a follow-on action for disclosure and damages in Germany against six cartel participants.Footnote 52 The defendants objected to the jurisdiction of the referring court on the basis of forum selection and arbitration clauses in the supply agreements between the claimants and the cartel participants.Footnote 53 The CJEU concluded that,
given that the undertaking which suffered the loss could not reasonably foresee such litigation at the time that it agreed to the jurisdiction clause and that that undertaking had no knowledge of the unlawful cartel at that time, such litigation cannot be regarded as stemming from a contractual relationship.Footnote 54
The Court added that ‘where a clause refers to disputes in connection with liability incurred as a result of an infringement of competition law’, the designated court is entitled to retain jurisdiction over the referred claim.Footnote 55
CDC was ambiguous in at least two respects. First, as the CJEU referred to competition law infringements in general, the ruling was misinterpreted as providing a clear-cut rule for all kinds of competition law claims regardless of their specific legal qualification.Footnote 56 This confusion in turn led courts to mistakenly assume that they had been absolved from applying the foreseeability test in each specific case, thus applying the CDC judgment in situations different from those which had given rise to it.Footnote 57 Second, some commentators have interpreted CDC as requiring contract practitioners to add specific competition law-related jurisdiction clauses to show that the parties had foreseen the possibility of such litigation when contracting.Footnote 58 However, these commentators failed to account for the practical implications of their proposal both in terms of costs and feasibility.Footnote 59 On the one hand, requiring drafters to insert competition law-specific jurisdiction clauses in international contracts would dramatically increase the transaction costs, multiplying the lawyer's bills and the parties’ negotiation efforts.Footnote 60 On the other hand, such specificity would disrupt the use of boilerplate clauses which have become dominant in the context of international business transactions.Footnote 61
In the subsequent case Apple v eBizcuss.com, the CJEU attempted to unravel at least some of the uncertainties generated by CDC.Footnote 62 In Apple, the referring court sought guidance from the CJEU on whether Article 23(1) of the Brussels I Regulation required the enforcement of a broadly worded jurisdiction agreement in the context of an action for abuse of a dominant position stemming from Article 102 TFEU and the corresponding national law provisions.Footnote 63 The jurisdiction agreement at issue provided:
This Agreement and the corresponding relationship between the parties shall be governed by and construed in accordance with the laws of the Republic of Ireland and the parties shall submit to the jurisdiction of the courts of the Republic of Ireland. […].Footnote 64
The CJEU first recalled the foreseeability test and then distinguished the case from CDC, stating that, whilst an unlawful cartel under Article 101 TFEU
is in principle not directly linked to the contractual relationship between a member of that cartel and a third party which is affected by the cartel, the anti-competitive conduct covered by Article 102 TFEU, namely the abuse of dominant position, can materialise in contractual relations that an undertaking in a dominant position establishes and by means of contractual terms.Footnote 65
The Court therefore concluded that an action for abuse of dominance ‘cannot be regarded as surprising’ and is therefore ‘not excluded on the sole ground that that clause does not expressly refer to [it]’.Footnote 66
Apple immediately attracted considerable criticism. Whilst one commentator observed that foreseeability ‘does not seem to have the force the Court ascribes to it’,Footnote 67 another highlighted the radical change of perspective from CDC to Apple, moving from granting access to justice for the victims of competition law infringements to protecting potential abusers.Footnote 68 At any rate, reconciling the two rulings seems difficult, particularly for legal practitioners called on to draft such agreements since the issues raised by the decision in CDC still remain unsettled after Apple.Footnote 69 The next section attempts to provide the necessary clarifications.
III. Demystifying Foreseeability
The objective of this section is to address the gap in the academic literature concerning the use of the foreseeability test as a means of determining claim proximity.Footnote 70 It has been argued that the EU regime ‘allows for jurisdiction agreements to encompass some non-contractual claims, but its scope is restricted by questions of the proximity of the claim to the contractual relationship’ and that ‘[t]he requirement that the parties not be “taken by surprise” suggests a focus on the foreseeability of the action’.Footnote 71 It follows that
It is clear that claimants cannot rely on a jurisdiction agreement in respect of proceedings that are entirely unrelated to that contractual relationship, however broadly it is drafted. It is, however, not clear how this ‘proximity’ test will be determined in difficult cases.Footnote 72
This section offers an analysis of the foreseeability test in four steps. First, it examines the proper standard under which the test should be applied (the reasonable person standard). Second, it verifies the extent to which the test is replicable outside both the corporate setting and the competition law context in which the CJEU has elaborated it. Third, it investigates how the test has been applied by domestic courts in cases with complex factual backgrounds. A final subsection mentions two tests (‘but-for’ and ‘interdependence’) that have been proposed in judicial practice as alternatives to the foreseeability test.
A. The Reasonable Person Standard
Among the ‘central pillars’ of the Brussels Regime are the certainty of law and the predictability of jurisdiction.Footnote 73 EU legislators put ‘legal protection and, hence, legal certainty’ at the core of the EU system of jurisdiction,Footnote 74 holding that ‘[t]he rules of jurisdiction must be highly predictable’.Footnote 75 Legal certainty pertains to the quality of the lawFootnote 76 and is often used as an ‘umbrella principle’ embracing other principles such as that of predictability.Footnote 77 In turn, predictability serves the economic interests of business actors, by making jurisdiction a diagnosable risk.Footnote 78 This risk-allocation rationale is an element that the provisions on jurisdiction agreements share with other aspects of law concerning international business transactions, such as damage recovery.Footnote 79
The CJEU has specified on multiple occasions that, in order to make jurisdiction predictable, it is necessary that
the applicant [must be able] to identify easily the court in which he may sue and the defendant reasonably to foresee before which court he may be sued.Footnote 80
Both the ease of identification and the reasonable foreseeability of the forum call for a notion of foreseeability that combines subjective and objective dimensions. According to this model, the subjective element would require courts to allow the party challenging the enforceability of a jurisdiction agreement to prove that the other party had actual knowledge of the particular claim to be submitted to, or excluded from, the jurisdiction of the designated court. At the same time, the objective element would require courts to verify whether a reasonable person in the same circumstances as the party against which the jurisdiction agreement is invoked would have foreseen that such a claim might be submitted to the designated court.
The CDC case is a perfect illustration of this twofold mechanism, with the CJEU acknowledging the relevance of both dimensions.Footnote 81 It also makes clear that actual knowledge trumps the reasonableness test, so that knowledge, and evidence thereof, of an existing cartel would be sufficient to connect the related damage claim with the parties’ contractual relationship and therefore make it foreseeable. On the other hand, by concluding that a claim for damages can arise from an abuse of dominance in a distributorship relationship, the CJEU in Apple acknowledged the value of the objective dimension. What is important is not the foreseeability of the claim per se, but rather the foreseeability of the jurisdiction flowing from the claim.Footnote 82 An accurate analysis of the claim is therefore essential in order to determine its proximity to the parties’ legal relationship.Footnote 83
It might be argued that hardly anyone—let alone a reasonable person—would be in a position to say what forms of claims would take them by surprise when entering into a contract. However, whilst this might be true in abstracto, it might still be possible to determine claim proximity in relation to a specific case. For example, whilst it can be maintained that a contract might not by itself give rise to a fraud claim, fraud might nonetheless be present under circumstances which may cause a reasonable person to foresee a related claim. To this end, the proposed combination of subjective and an objective element may ensure that there is the necessary flexibility ‘to accommodate a variety of factual settings and, at the same time, […] still have a discernible content’.Footnote 84
B. Case Specificity
For business actors to properly allocate jurisdictional risk, the test used to determine the effectiveness of a jurisdiction agreement should ideally be replicable outside the specific claims for which it has been elaborated, ie the corporate setting and the competition law context. Foreseeability, however, is just a ‘shortcut’Footnote 85 which assists courts in deciding on jurisdiction and, in this sense, is hardly a precise test.
In particular, courts might be tempted, in the wake of CDC and Apple, to conclude that there is proximity simply from the legal characterisation of the claim so that, although damages claims arising out of unlawful cartels have no connection whatsoever with a contract, the abuse of a dominant position means that such a connection necessarily exists.Footnote 86 This is not the case. Even after these rulings, a case-by-case assessment is still required.Footnote 87 Indeed, the nature of competition law infringements are so varied that it is hard to delineate the concrete reach of CDC and Apple.Footnote 88 The careful wording used by the CJEU in both rulings—that an infringement of Article 101 TFEU is ‘in principle’ unconnected to a contractual relationshipFootnote 89 and a violation of Article 102 TFEU ‘can’ materialise in contractual termsFootnote 90—shows that the CJEU also intended that the test be applied to the facts of individual cases. The labelling of the legal relationship is simply insufficient to determine its proximity to the claim.Footnote 91
Beyond competition law, the galaxy of possible non-contractual claims connected to a particular legal relationship is vast. Consider a false public statement released by a senior company manager which harms the reputation of one of the company's investors. Is the resulting damage claim connected, for the purpose of determining jurisdiction, with a financing contract that the victim has signed with the company's CEO? Imagine a forged document, totally unrelated to the contract in force between two business partners but which triggers a governmental investigation to the detriment of the commercial relationship between the parties. Or a bribery case, unknown at the time, that caused a company to settle in a certain way a dispute relating to a share purchase agreement. Is the tort claim arising out of the bribe subject to the jurisdiction clause contained in such a share purchase agreement? And what about a supply or a distributorship contract being the primary vehicle through which cartel participants profit from the cartel's detrimental economic effects?
Answering each of these questions is difficult because the foreseeability test is not just replicable outside the contexts in which the CJEU elaborated it. Accordingly, domestic courts must be prepared to engage in a case-by-case analysis in order to determine whether the foreseeability test is met. Neither CDC nor Apple has exempted courts from conducting such case-by-case assessments of proximity.
C. Determining the Substance of the Claim
Proximity is normally determined in the light of the parties’ formulation of a claim. This makes the outcome ultimately dependent on contingent factors such as the counsel's creativity, raising the question of the scope of the court's analysis regarding the substance of the claim.
Consider for example Hewden Tower Cranes v Wolffkran, concerning a fatal accident that occurred on a construction site in London in 2000.Footnote 92 The claimant sued in the London Technology and Construction Court arguing that the accident was caused by defects in a climbing frame it had hired from the defendant in 1997, holding the latter liable for damages. The defendant contested the court's jurisdiction by invoking a jurisdiction clause contained in a crane purchase agreement entered into with the claimant in 1999, which designated a court in Germany as the ‘sole [court] competent for any disputes arising directly or indirectly out of the contractual relation’.Footnote 93 In order to rule on jurisdiction, the English court had to establish whether the damages claim arising out of the accident was ‘directly or indirectly’ connected to the crane purchase agreement and found it was not, one reason being that the crane's involvement in the accident ‘appeared to have been pure chance’ as the crane ‘was merely the location where the defects in the [c]limbing [f]rame manifested themselves with devastating consequences’.Footnote 94
Here the connection between the claimant's tort claim and the contractual relationship between the parties was so remote that the defendant could not avail themselves of the jurisdiction clause in relation to such a claim. Notably, the analysis of the accident's factual basis played a crucial role in assisting the court in assessing the claim's proximity.
Of course, sometimes the parties’ formulation of the claim may be determinative. Take for example the ruling of the Portuguese Supremo Tribunal de Justiça (STJ) in a case very similar to Apple that involved the enforcement of a jurisdiction clause in a damages claim arising out of an abuse of dominance allegedly committed by Apple to the detriment of its nationwide distributors.Footnote 95 Notably, this ruling was issued before the CJEU's Apple judgment. The STJ decided to enforce the jurisdiction clause on the ground that ‘the initial petition shows and emphasises […] a change in behaviour in the equilibrium in the contractual framework established in the contract between the claimant and the respondent’.Footnote 96 In fact, ‘it is [the initial petition] that defines and sets the legal and factual framework of the procedural premises to the jurisdiction of the court […] and reflects the meaning and the effect that result and arise out of the contractual relationship that both parties have entered into’.Footnote 97 Decisively, the claimants’ petition, which centred on the distributors’ freedom to fix prices, only raised questions concerning the behaviour of the respondent which related to the contract, emphasising that the claim was based on competition law violations ‘committed inside the contract’.Footnote 98
Cases such as these show how the boundary between the assessment of jurisdiction and the substance of the claim may become extremely blurred, questioning the actual possibility of applying the foreseeability test in cases characterised by complex factual backgrounds. It might be therefore suggested, as a general recommendation to determine proximity, not simply to look at the claim as formulated by the parties but to dig deep into the substance of the claim by engaging in a review of the facts of the case and the related evidence provided by the parties.
Remarkably, this suggestion was recently offered by the High Court of Justice in Etihad Airways PJSC v Flöther, decided in 2019.Footnote 99 In that case, the court had to determine whether a dispute arising from the alleged breach of a comfort letter containing no jurisdiction clause originated from a previous loan agreement between the parties which provided for the jurisdiction of English courts.Footnote 100 The court acknowledged that proximity has to be decided ‘in the light of admissible evidence as a whole’ after considering ‘the substance of the claim’ and concluded it had jurisdiction over the dispute.Footnote 101
D. Complementary Tests
Finally, two tests have been considered in judicial practice to determine proximity.Footnote 102 A first test involves enquiring whether the non-contractual claim would have arisen without the existence of the contract. Under such a but-for test, the non-contractual claim does not have to be foreseeable at the time of contracting, but what is relevant is whether such a claim would not have arisen ‘but for’ the contractual relationship. When foreseeability does not help clarify proximity, considering the contract as the claim's necessary factual antecedent of the claim could fill the gap.Footnote 103 A second test—the close interdependence test—requires a close factual link between the non-contractual claim and the particular legal relationship between the parties.
Both these tests have occasionally been raised by the parties but the courts have rejected them on the ground that they have no basis either in scholarship or in judicial practice.Footnote 104 The fact that foreseeability is the only test currently used by the CJEU seems to explain its prevalence in practice. Nonetheless, courts may find some comfort in these two tests as a complement to foreseeability, when the latter test does not provide a clear answer to the question of proximity.
IV. CONCLUSION
This article has attempted to cast light on a difficult question regarding the EU regime of jurisdiction agreements: the need for there to be a sufficient connection (proximity) between a claim and the particular legal relationship which gives rise to the jurisdiction of the court. The importance of this requirement lies in the fact that both business actors and legal practitioners need clear and predictable jurisdictional rules, and, as they stand now, neither proximity nor its associated foreseeability test seem to make them so.
A few suggestions are offered here. First, proximity must be kept conceptually separate from the question of consent—the issue is not whether the parties wanted a particular claim to be submitted to the designated jurisdiction, but whether that claim objectively arises from a relationship between the parties other than that in connection with which the jurisdiction agreement was made. Second, courts should be reminded that the foreseeability test necessarily entails a case-by-case analysis which may require a factual analysis of the substance of the claim. They should not draw on the recent CJEU decisions in CDC and Apple to adopt an overly fixed rule which turns on the legal characterisation of the non-contractual claim, but respect proximity's necessary connection with the case at hand.
