4.01  Subject to the rules of displacement contained in Arts 4(2) and (3), Art 4(1) provides that the law generally applicable to non-contractual obligations arising out of a tort/delict is the law of the country in which the damage occurs (lex loci damni‎).² Despite minor changes in wording, this outcome largely reflects the Commission Proposal.³ The Commission’s approach was challenged by the European Parliament during the passage of the Regulation,⁴ but was supported by the Member State delegations in the Council.

4.02  In recommending this solution, the Commission rejected four alternatives to the country of the damage as the primary connecting factor. First, to refer to the country of the ‘harmful event’ (French ‘fait dommageable‎’), reflecting the language of Art 5(3) of the Brussels I Regulation and the Brussels Convention before it. Although that solution had appealed to the authors of the 1972 draft convention,⁵ the ECJ had subsequently determined that the expression ‘place where the harmful event occurred’ in Art 5(3) of the Brussels Convention (matters relating to tort, delict, or quasidelict) referred not only to the place where the event giving rise to damage occurred but also to the place where damage occurred, giving the claimant the right to choose between the courts of those two places.⁶ In the Commission’s view, that solution was acceptable within the Brussels I Regime dealing with the jurisdiction of Member State courts, ‘but it does not enable the parties to foresee the law that will be applicable to their (p. 297) situation with certainty’.⁷ Had this approach been adopted for tort/delict in the Regulation, it would have introduced a so-called ‘principle of ubiquity’ (Günstigkeitsprinzip‎) in a form adopted until 1999 in Germany, allowing the injured person to choose the law most favourable to him.⁸ An express provision of that kind was the second alternative option rejected by the Commission. In its view, ‘this solution would go beyond the victim’s legitimate expectations and would introduce uncertainty in the law, contrary to the general objective of the proposed Regulation’.⁹ The third alternative would have been to apply the law of the country of the act or other event giving rise to damage (lex loci actus‎). Finally, the Commission rejected the possibility of applying the law of the place where the event giving rise to damage occurred.¹⁰

4.03  The Commission argued that its preference for the law of the country of the damage reflected recent developments in Member States’ conflict rules, and suggested that the solution had been applied in the Netherlands, the United Kingdom, and France, as well as Switzerland.¹¹ As appears from the comparative survey in Chapter 1, however, that statement manifestly over-simplified the rules of applicable law in force in these countries, which essentially blend the lex loci actus‎ and the lex loci damni‎ albeit (a) in different ways, (b) using different terminology, and (c) subject to different exceptions.¹²

4.04  The Commission also argued that its favoured solution ‘establishes an objective link between the damage and the applicable law’,¹³ although each of the solutions that it rejected would also have achieved this. Further, in the Commission’s view, the preference for the law of the country in which the damage occurred ‘reflects the modern concept of the law of civil liability which is no longer, as it was in the first half of the last century, oriented towards punishing for fault-based conduct: nowadays it is the (p. 298) compensation function that dominates as can be seen from the proliferation of no-fault strict liability schemes’.¹⁴

4.05  In the following commentary, the term ‘victim’ is used as a convenient shorthand for the person sustaining the damage and the term ‘tortfeasor’ is used as a convenient shorthand for the person responsible for the event giving rise to damage. The terms ‘claimant’ and ‘defendant’ are generally used to describe, respectively, the likely procedural positions of the victim and tortfeasor (or the persons representing them or liable for their conduct, as the case may be). It is, of course, possible that those positions may be reversed, for example if the victim counterclaims against the tortfeasor¹⁵ or if the tortfeasor seeks a declaration that he is not liable to the victim.¹⁶

4.06  The concept of a ‘non-contractual obligation arising out of a tort/delict’ has already been considered at length in Chapter 3, in assessing the subj ect matter scope of the Regulation.¹⁷ In summary:

1. 1. The concept of a ‘tort/delict’, like that of ‘non-contractual obligation’,¹⁸ is an autonomous one under EC law, and its content does not fall to be determined by reference to the concepts of ‘tort’ or ‘delict’, or corresponding terminology, deployed within particular legal systems, including the legal system of the forum Member State and that of the putative applicable law.¹⁹

2. 2. The concept of ‘non-contractual obligation’ potentially includes all civil obligations that are not ‘contractual obligations’.²⁰ For these purposes, a ‘contractual obligation’ is one which has as its foundation a ‘contract’, (p. 299) consisting of an intentional and voluntary act by one person towards another and (probably) some reciprocal voluntary act by that other person with those acts in combination giving rise to a binding obligation.²¹

3. 3. The Regulation also applies to non-contractual obligations arising from strict liability.²² The concept of a ‘tort/delict’ cannot, therefore, be understood as requiring fault on the part of the tortfeasor.²³

4. 4. The source of the rule generating the obligation within the legal system is not a relevant factor in its characterization.²⁴ Thus, for example, statutory rules and rules developed by the Courts of Equity in England (e.g. breach of trust, breach of fiduciary duty) may create non-contractual obligations falling within the Regulation’s scope.

5. 5. Despite the Commission’s suggestion that the Rome II Regulation should apply to all non-contractual obligations, there is a residual group of non-contractual obligations for which the Regulation does not provide any rule of applicable law. Such obligations (for example, monetary obligations attaching to immoveable property) fall outside the scope of the Regulation.²⁵

6. 6. Within the category of non-contractual obligations, the concept of ‘tort/delict’ should be given a broad construction so as to accommodate (so far as possible) non-contractual obligations that might otherwise fall outside the Regulation.²⁶

7. 7. The structure and language of the Regulation and common usage among the Member States’ legal systems, as well as the approach taken by the Court of Justice to the concept of ‘tort, delict or quasi-delict’ in Art 5(3) of the Brussels Convention, support the definition of a non-contractual obligation arising from tort/delict as:²⁷

   A non-contractual obligation establishing the defendant’s responsibility to the claimant for, or preventing the occurrence of, an act, omission or event which (1) has adverse consequences for the claimant or one or more persons represented in the proceedings by the claimant and (2) is an act or omission of the defendant or a person for whose conduct the defendant is liable, or is otherwise an act, omission or other event attributable to the defendant through a relationship with the person or thing giving rise to those consequences.

8. (p. 300) 8. It is unclear whether it is an additional requirement that the relevant act, omission, or other event must, in some sense, be unlawful, so as to exclude rules imposing responsibility for lawful acts.²⁸

4.07  As its opening words make clear, Art 4 also has a negative aspect to its scope, in that it does not apply to non-contractual obligations for which separate provision has been made elsewhere in the Rome II Regulation. Accordingly, Art 4 does not apply to:

• •  non-contractual obligations arising out of damage caused by a product (Art 5);²⁹

• •  non-contractual obligations arising out of environmental damage or damage sustained by persons or property as a result of such damage (Art 7);³⁰ • non-contractual obligations arising out of an infringement of an intellectual property right (Art 8);³¹

• •  non-contractual obligations in respect of the liability of workers, employers, or organizations representing their professional interests for damages caused by an industrial action (Art 9).³²

4.08  The relationship between Art 4, on the one hand, and Arts 6 and 10 to 12, on the other, is less straightforward. In the case of Art 6, Recital (21) states that it is a clarification of the general rule in Art 4(1) rather than an exception to it. The intended meaning and effect of this statement are unclear.³³ Further, under Art 6(2), Art 4 applies without modification in cases which affect exclusively the interests of a specific competitor. In the case of Arts 10 to 12, the headings of Chapters II (containing Art 4) and III (containing Arts 10 to 12) in the Regulation, as well as Recital (29), confirm that tort/delict, on the one hand, and unjust enrichment, negotiorum gestio‎, and culpa in contrahendo‎, on the other, are mutually exclusive categories. Nevertheless, it may be argued that the rules contained in Chapter III must, to some degree, be given priority over, or at least equal status with, those in Chapter II in determining the non-contractual obligations to which they apply. The opposite view, giving the rules in Chapter II priority over those Chapter III and leaving Arts 10–12 to pick from the residue of non-contractual obligations not arising out of a tort/delict would be liable to deprive Art 12 (culpa in contrahendo‎), at least, of much (p. 301) of its content.³⁴ It would also remove from Art 11 (negotiorum gestio‎) claims by the principal or intervener against the other to recover compensation for harm resulting from the intervention, thereby splitting the claims arising from relations between principal and intervener into two.³⁵ The relationship between Art 4 (together with the special rules for tort/delict in Chapter II) and the special rules for negotiorum gestio‎ and culpa in contrahendo‎ will be returned to in the commentaries on Arts 11 and 12.³⁶ The relationship between Chapter II and Art 10 (unjust enrichment) will be addressed in the following section, in considering the treatment of claims for gain-based remedies based on tort/delict.

4.09  The category of ‘non-contractual obligations arising out of a tort/delict’ under the Rome II Regulation is not limited to actions to obtain damages compensating the claimant for the adverse consequences of the event giving rise to damage. It extends, for example, to claims for injunctive relief to prevent future breaches of non-contractual obligations.³⁷ Art 15(d) refers to the law applicable under the Regulation ‘the measures which a court may take to prevent or terminate injury or damage’.³⁸

4.10  Further, Recital (32), concerning the overriding effect of public policy and mandatory provisions, specifically contemplates that law designated by the Regulation may provide for non-compensatory exemplary or punitive damages.³⁹

4.11  In many countries, the law provides remedies of various descriptions (e.g. disgorgement, restitutionary damages, account of profits, remedial constructive trust) that have the object or the appearance at least of depriving the defendant of the profit of his wrongful conduct towards the claimant. In some cases, an award of this kind may sometimes justifiably be (p. 302) described as compensating the claimant;⁴⁰ sometimes not.⁴¹ Commentators on English law disagree as to whether cases falling within the description⁴² of ‘restitution for wrongdoing’ should be seen as part of the law of obligations, or the law of unjust enrichment, or both. The late Professor Birks, the leading English scholar in this area, concluded that:⁴³

Unjust enrichment … requires a connection between claimant and enrichment which is independent of wrongdoing and a reason for restitution which is likewise not a wrong… . In unjust enrichment ‘unjust’ always denotes a reason for restitution which is not a manifestation of consent and not a wrong.

4.12  For Professor Burrows, on the other hand, ‘the law of restitution may be regarded as having a central divide between unjust enrichment by subtraction … and unjust enrichment by wrongdoing. This division is marked by the two meanings of “at the expense of the claimant”; “by subtraction from” and “by a wrong to”’.⁴⁴ Professor Burrows acknowledges, however, that ‘[t]he divide is one between where unjust enrichment is the cause of action or event to which restitution responds and where a wrong is the cause of action or event to which restitution responds’.⁴⁵ The difference between these two schools of thought, therefore, may be more one of terminology than substance. Each accepts that it is the wrong that triggers the law’s response in these cases.

4.13  Whether, as a matter of domestic law, a claim of this type can be fitted within the law of unjust(ified) enrichment and whatever label is attached to the remedy sought, such claims should be characterized as falling within Chapter II of the Regulation (tort/delict), not Art 10 (unjust enrichment), whenever the remedy sought responds to a tort/delict in the sense described above, i.e. an act or omission of or attributable to the defendant or other event for the consequences of which the defendant⁴⁶ is responsible to the claimant.⁴⁷ In such a case, even if it can also be said that there is ‘unjust enrichment’ within the autonomous meaning given to that term in the Regulation,⁴⁸ the defendant’s non-contractual obligation should be (p. 303) considered to be founded not on that unjust enrichment but on the act, omission, or other event that constitutes the ‘tort/delict’ within Art 4 or one of the special rules in the following provisions of the Regulation.⁴⁹ To quote Professor Birks:⁵⁰

Where a claimant identifies himself as the victim of a wrong he is relying on the wrong and, albeit in the language of unjust enrichment, asking the court whether that wrong is one which yields a right to a gain-based award. The law of unjust enrichment cannot answer that question. It belongs to the law of wrongs.

4.14  This view accommodates not only what an English lawyer might describe as ‘restitution for wrongdoing’⁵¹ but also, for example, claims under the German law concept of the Eingriffskondiktion‎ (literally ‘interference/encroachment condition’) based on an interference with the claimant’s right⁵² even though they fall under the heading of ‘unjustified enrichment’ in the German Civil Code.⁵³ It accords with Art 15(c) of the Regulation referring to the law applicable under the Regulation ‘the nature and assessment of the damage or the remedy claimed’.⁵⁴ Further, insofar as remedies that strip the defendant of some or all of the profit of his wrongful conduct serve a prophylactic purpose in disincentivizing that conduct, they can also be seen as measures ‘to prevent or terminate injury or damage’ (Art 15(d)), closely related to other measures of this kind, such as injunctions.⁵⁵ Finally, a test focusing on the basis for granting the remedy, rather than on the nature of the remedy granted, disposes of the need for further analysis as to whether, in substance, the remedy is ‘compensatory’ or ‘restitutionary’ in nature, and avoids making what may be seen as arbitrary distinctions, for example, between restitutionary damages and punitive damages against those who deliberately commit wrongs for profit.

4.15  Two aspects of the wording of Art 10(1) may be thought to provide reasons to doubt the correctness of the above conclusion. First, this subrule refers specifically to ‘payments of amounts wrongfully received’, which at first sight might be thought to cover restitutionary claims based on wrongdoing. Secondly, the sub-rule also contemplates the possibility of (p. 304) a close connection between a non-contractual obligation arising out of unjust enrichment and a relationship existing between the parties ‘such as one arising out of … a tort/delict’. The existence of that connection results in the application to the unjust enrichment claim of the law governing the tort/delict, which might be thought to make this question of characterization a largely theoretical one. As to the first aspect, this appears to be no more than an unfortunate choice of words in the English text of the Regulation, for which it appears the UK delegation on the Council’s Rome II Committee must take some responsibility.⁵⁶ The French language version (referring to ‘paiement indu‎’) and the German (referring to ‘Zahlungen auf eine nicht bestehende Schuld‎’) indicate that the legislator’s intention was to refer to claims to recover payments of a non-existing debt (condictio indebiti‎). This would appear to have been in order to remove doubts raised by Member State delegations as to whether these types of claim, dealt separately from unjust(ified) enrichment within some legal systems, were covered by the rule.⁵⁷ The second aspect is more problematic, as it is not clear what the addition of this wording, which first appeared in the draft produced by the outgoing Dutch and incoming Luxembourg Presidencies in December 2004,⁵⁸ was intended to achieve. It may be that the draftsman had in mind situations in which the prior commission of a tort/delict is an element in an unjust enrichment claim brought by the wrongdoer against the victim to prevent the latter from being overcompensated.⁵⁹ Whether that speculation is correct, the reference to a ‘relationship existing’ appears to require a liability in tort/delict of one of the parties to the other that is complete at the time of the unjust enrichment that is connected to it.⁶⁰ On this view, it would not suffice that there is factual connection between the circumstances giving rise to claims in tort/delict and unjust enrichment. The required element of separation between the tort/delict and the unjust enrichment cannot, in any event, be said to exist in the case of a claim for profit-based damages, which (p. 305) constitutes an indissociable part of the tie⁶¹ between victim and tortfeasor arising out of the latter’s conduct towards the former.

4.18  In light of this explanation, it could be suggested that, without a provision corresponding to Art 13 for other obligations within Chapter II (Arts 4–7 and 9), claims to reverse the defendant’s enrichment from a tort/delict other than an infringement of an intellectual property right must, by necessary inference, fall within Art 10. It is submitted that, having regard to the points made above, that inference is not justified. Art 13 should be considered as a clarification in the scope of Art 8, for the avoidance of doubt, and not as resolving the question as to how gain based remedies for wrongdoing should be characterized in other situations.

(p. 306) 4.19  Treating claims of this kind as falling within Chapter II not only ensures consistency between the remedies for tort/delict, whether compensatory or non-compensatory, it also prevents dilution of the special rules in Arts 6, 7, and 9 and supports the achievement of the particular objectives that they pursue. In particular, it should be noted that, even if (contrary to the author’s view⁶⁵) the reference in Art 10(1) to ‘a relationship existing between the parties, such as one arising out of … a tort/delict’ can be read sufficiently broadly to allow the conclusion that the law applicable to a tort/ delict should also be applied to a restitutionary claim based on that tort/delict,⁶⁶ that law would be subject to displacement by the flexible escape clause in Art 10(4). An escape clause of this kind was not considered appropriate for obligations falling within Arts 6, 7, or 9, and it is difficult to see why the claimant’s decision to seek a gain-based remedy for the defendant’s wrongful conduct should make a difference. Further, the law applicable under Art 10 can be displaced by party choice under Art 14, a possibility that is excluded for non-contractual obligations arising out of acts of unfair competition and restrictions of competition (Art 6(1) and (3)).⁶⁷ The wider, market-oriented objectives pursued by these rules, and in particular by the exclusion of party autonomy,⁶⁸ do not support a distinction between loss-based and preventative remedies, on the one hand, and gain-based remedies on the other. Indeed, in some cases, the availability of a remedy stripping the defendant of the profits of anticompetitive conduct may provide a strong incentive for him to tailor his conduct to comply with laws and practices regulating the market in question.

4.20  Of course, the claimant may be able to frame his claim as an unjust enrichment claim without needing to rely on the defendant’s legal responsibility for the event that caused his enrichment.⁶⁹ In such a case, as for concurrent contractual and non-contractual obligations,⁷⁰ there is no reason in principle why the same factual situation cannot, under the Regulation, give rise to both a non-contractual obligation arising in tort/delict (Art 4) and a (p. 307) non-contractual obligation arising in unjust enrichment (Art 10). The law applicable to each obligation must be determined separately: it appears unlikely, for the reasons given above, that the ‘tort/delict’ connector in Art 10(1) can be used to fix both claims with the same applicable law, although that result may be achieved through the more general rule of displacement in Art 10(4).

(p. 308) It must not, however, be assumed that the victim’s country of residence is the country in which the damage ‘occurs’.

4.24  In this connection, the Commission Proposal had referred to the country in which the damage ‘arises or is likely to arise’. The word ‘occurs’ seems preferable, being consistent with the English text of the leading decision of the ECJ in Bier v Mines de potasse d’Alsace, discussed below.⁷² It emphasizes that the focus is on the place where the event giving rise to the damage ‘produces its harmful effects on the victim’, rather than where it originates.⁷³

4.25  This aspect of the Commission Proposal was not universally well received.⁷⁴ Some Member States were critical of the vagueness of the distinction between consequences of the event giving rise to damage that are ‘indirect’ and those that are not.⁷⁵ In the main, these concerns seem to have been assuaged at an early stage in the process.⁷⁶

4.27  The references in Art 4(1) of the Rome II Regulation to ‘the country in which the damage occurs’ and ‘the event giving rise to the damage’ reflect the distinction drawn by the ECJ in Bier. In later cases, the ECJ sought to limit the effect of this decision in extending the jurisdiction of Member State courts under the Brussels Convention in matters relating to tort etc. In particular, it sought to restrict the concept of ‘damage’ by excluding the consequences of the event giving rise to damage that were thought to be indirect and too remote. That restriction was first identified in Dumez France SA v Hessische Landesbank⁸¹ and subsequently applied in Marinari v Lloyd’s Bank plc⁸² and Kronhofer v Maier.⁸³ Marinari was referred to by the Commission as supporting the exclusion of the ‘indirect consequences’ of the event giving rise to damage from the connecting factor in its proposed general rule for tort/delict. The scope and effect of that exclusion, carried forward into Art 4(1) in its final form, will be considered, with reference to the ECJ’s decisions in all three of these cases, at 4.36–4.45 below.

4.28  In Bier, the ECJ was unable or unwilling to choose between the place of the event giving rise to damage and the place of damage as a connecting factor for jurisdiction purposes. The choice of the latter in Art 4(1) of the Rome II Regulation was described by the Commission as ‘a compromise between the two extreme solutions of applying the law of the place where the event giving rise to the damage occurs and giving the victim the option’ and seems to (p. 310) have been preferred to the country of the event giving rise to damage on the basis that it was more favourable to the victim, consistently with the Commission’s view that the civil liability rules now had as their predominant object the compensation of those affected by the conduct of others rather than the punishment of fault-based conduct.⁸⁴ That view is reflected in Recital (16).⁸⁵ Nevertheless, as Recital (16) also makes clear, the primary object of the rule in Art 4(1) is to enhance the foreseeability of court decisions and to strike a reasonable balance between the interests of the person claimed to be liable and the person who has sustained damage. Thus, the concept of the ‘country in which the damage occurs’ in Art 4(1) must be applied so as, whenever possible, to base the identification of the law applicable to a tort/delict on factors objectively ascertainable by both parties.⁸⁶ To enhance legal certainty, those factors should also be applied uniformly with respect to similar types of claim. The strength of the connection to the country identifi ed is a matter to be taken into account in moulding the concepts of ‘damage’ and ‘indirect consequences’, but the absence of a real connection in a particular case should not lead the court to refuse to apply Art 4(1).⁸⁷ Instead, if there is only a tenuous connection with the country of damage, it may be considered appropriate to apply the ‘escape clause’ in Art 4(3).⁸⁸

4.29  Art 2(1) defines ‘damage’ as covering ‘any consequence arising out of tort/delict [etc]’.⁸⁹ This clarification renders it unnecessary for the ECJ and Member State courts to develop an autonomous concept of ‘damage’ for the purposes of the Regulation⁹⁰ or to refer to the putative applicable law or the law of the forum.⁹¹ Instead, it is submitted, Art 4(1) requires the court to consider the sequence of events giving rise to the claimant’s claim, not limiting itself to those pleaded by the claimant, and to identify (1) the ‘event giving rise to damage’, (2) the consequences of that ‘event’ in terms of its (adverse⁹²) effect on the victim,⁹³ and (3) which of those consequences may be considered as ‘indirect consequences’ excluded from having any (p. 311) role in determining the law applicable under Art 4(1). Member State courts must take a broad view of what, for these purposes, may constitute ‘damage’, and must not restrict the meaning of the term to interests that would be protected under the law of the forum. In VfK v Henkel, for example, the ECJ held that the concept of ‘damage’ for the purposes of Art 5(3) of the Brussels Convention was capable of covering ‘the undermining of legal stability by the use of unfair terms’.⁹⁴

4.30  The concept of ‘damage’ within the Rome II Regulation is broader, for example, than that in the Product Liability Directive⁹⁵ or developed by the ECJ in its case law concerning the non-contractual liability of the Community.⁹⁶

4.31  Consistently with the reasoning of the ECJ in Bier,⁹⁷ both Art 2(1) and the language of Art 4(1) emphasize that the ‘damage’ must, in some sense, be caused by the event giving rise to damage.⁹⁸ Nevertheless, it is important that Member State courts should not take an unduly restrictive approach to causation, for example by applying a ‘but for’ test or by requiring that the consequence be foreseeable. Nor should the words ‘consequence’ and ‘consequences’ in Arts 2(1) and 4(1) respectively be given a technical meaning, contrasted (for example) with ‘effects’ or ‘results’.⁹⁹ Indeed, for Art 4(1) to apply effectively, every ‘event giving rise to damage’ must be considered as causing some ‘damage’, the location of which will enable the law applicable to the non-contractual obligation arising in tort/delict to be identified. Art 4(1) must, therefore, be able to accommodate those cases in which the claim tests the boundaries of scientific knowledge or involves ‘an unfamiliar tale about unfamiliar events’,¹⁰⁰ as much as those in which the effect of a particular event is obvious to all. For example, if the claimant asserts that his illness resulted from exposure to asbestos while in the employ of several persons, (p. 312) it should not matter that he cannot point to the ‘guilty’ fibres or say that, but for a particular period of exposure, he would still have been healthy.¹⁰¹ The illness must, for the purposes of determining the law applicable, be taken as the consequence of each exposure. Equally, but more fancifully, if a landowner claims that the flapping of butterflies kept on his neighbour’s land resulted in damage to his property due to a change in air pressure, the court must proceed on the basis that the property damage is a consequence of the defendant’s lepidoptery. Of course, the question of causation must be reexamined at the stage of considering the defendant’s liability by reference to the law applicable to the tort/delict, once this has been identified.¹⁰²

4.32  In accordance with Art 2(3)(b), the concept of damage must also include damage that is likely to occur. This suggests an element of foreseeability as to future consequences of the event giving rise to damage although, for the reasons just given, the threshold should not be set at a high level.

4.33  The ‘event giving rise to damage’ should be the present or future¹⁰⁴ act, omission, or other event of or attributable to the tortfeasor on which the claimant’s claim is founded. This will not necessarily correspond with the fact that the claimant, or the rules of any of the countries whose law may be apply, identifies as being ‘unlawful’¹⁰⁵ or with the last act attributable to the defendant.¹⁰⁶ In the Brussels Convention case of Shevill v Presse Alliance, the ECJ held that the ‘event giving rise to damage’ in a case involving defamatory comments in a newspaper¹⁰⁷ was the issue and putting into circulation of the newspaper containing the libel (in France) rather than its publication to one or more persons (in several countries, including England).¹⁰⁸ It was able to reach that conclusion, although (a) under English law, as both the law of the forum and the law applicable to the tort in that case, it was publication of the defamatory statement to a third party that constituted the tort, and (b) the defendant had also been responsible (p. 313) for distribution of its newspaper in England after printing it and putting it into circulation.

4.34  Two English High Court decisions on the Brussels Convention support a test expressed in terms of ‘where in substance the cause of action in tort arises, or what the tort is most closely connected with’,¹⁰⁹ a test reminiscent of the approach taken by the English courts in locating cross-border torts under the former common law rules.¹¹⁰ That approach was, however, rejected by Rix J in Domicrest Ltd v Swiss Bank Corporation as not reflecting the wording of or philosophy behind the ECJ’s case law on Art 5(3) (including Shevill).¹¹¹ Reasoning by analogy with Shevill, he held that the event giving rise to damage in a claim based on misrepresentation was the making of the representation and not its receipt.¹¹²

4.35  Having identified the event giving rise to damage, it is not necessary for the purposes of the Regulation to identify where it occurred (at least for the purposes of Art 4(1), although this may be a relevant factor in carrying out the balancing exercise required by Art 4(3)¹¹³). In many cases, the event giving rise to damage and the (direct) damage will occur in the same country, and the reference to the former in Art 4(1) does no more than emphasize that its focus is not on the tortfeasor’s conduct but on the effect that conduct has on the victim. It is not, therefore, a qualification to the concept of ‘damage’ but a clarification as to the nature of the connecting factor.

4.37  The Commission referred, in this connection, to the ECJ’s decision in Marinari v Lloyd’s Bank. In that case Mr Marinari sought compensation in Italy following the Bank’s refusal to return to him promissory notes that he had lodged with its branch in Manchester and its subsequent report to the British police leading to Mr Marinari’s arrest and sequestration of the notes. He sought not only payment of the face value of the notes but also compensation for damage suffered by reason of his arrest and damage to his reputation.¹¹⁶ The ECJ, responding to a preliminary question put to it by the Italian court, concluded that the concept of ‘harmful event’ within Art 5(3) of the Brussels Convention did not extend to ‘the place where, as in the present case, the victim claims to have suffered financial damage following upon initial damage arising and suffered by him in another Contracting State’.¹¹⁷ In the Court’s view, this was inherent in the wider proposition that the connecting factor in Art 5(3)¹¹⁸ ‘cannot be construed so extensively as to encompass any place where the adverse consequences can be felt of an event which has already caused damage actually arising elsewhere’. In the preceding paragraphs of its judgment, the Court had explained the need for Art 5(3) to be applied so as to ensure a significant connection to the court having jurisdiction, and to avoid unduly favouring the courts of the claimant’s domicile.¹¹⁹ Although those objectives are specific to the Brussels I Regime, they are broadly consistent with the Commission’s reasons for choosing the lex loci damni‎ as the connecting factor under the Rome II Regulation.¹²⁰ There is little doubt, therefore, that financial or other consequences that reflect or follow from initial damage to the claimant must be treated as ‘indirect consequences’ within Art 4(1). For example, in the case of a claim for loss of earnings in State A following injuries suffered in a motor accident in State B, Art 4(1) will point exclusively towards application of the law of State B. Similarly, if the defendant’s negligence leads to the loss of assets under management in State C, it matters not that the claimant’s assets are concentrated in State D and that he feels the effects of that loss there.¹²¹

4.38  The examples given in the preceding paragraph are relatively straightforward applications of the distinction between ‘direct’ and ‘indirect’ consequences of the event giving rise to damage in Art 4(1). Other situations create greater difficulty. The second sentence of Recital (17) states that, in cases of personal injury or damage to property, the country of damage should be that where the injury was sustained or the property was damaged.¹²² Ordinarily, therefore, it would appear that personal injury or property damage must be treated as ‘direct’ damage. What if an injury occurring in State E is followed by a deterioration in the victim’s condition after his return home to State F? In Henderson v Jaouen, the English Court of Appeal (applying the reasoning in Marinari by analogy) held that although aggravation in England of the claimant’s injury gave a fresh cause of action in France (where the traffic accident which had caused the injury had occurred), it was not a separate ‘harmful event’ under Art 5(3) of the Brussels Convention.¹²³ That conclusion also seems appropriate under the Rome II Regulation, in terms of avoiding undue favour to the law of the victim’s habitual residence and enhancing foreseeability as to the law applicable. The damage should be taken to have occurred in State E.

4.41  Rushworth and Scott also argue that the replication of Dumez within the framework established by Art 4(1) of the Rome II Regulation would create insuperable difficulties in determining whether a person claims as a ‘direct’ (primary) or ‘indirect’ (secondary) victim,¹²⁹ illustrating their argument with the following example:¹³⁰

[A] football match played in Athens is negligently stewarded by the defendants, causing serious injuries to supporters at the stadium. Where the events are broadcast in England to a local claimant’s distress, which law will Art 4(1) apply to determine the defendant’s liability for the claimant’s psychiatric harm?

4.42  A short-cut to the answer to this particular example might be provided by Recital (17),¹³¹ which states that, in cases of personal injury, the country in (p. 317) which the damage occurs should be the country where the injury was sustained. This may be argued to create a strong indication that a person suffering personal injury, however it may occur, should be considered to be a primary, direct victim of the event giving rise to damage, leading in the example to the application under Art 4(1) of English law.¹³² That solution, however, creates its own problems as to the meaning of the (presumably) autonomous concept of ‘personal injury’. If nervous shock or other psychiatric illness is to be considered as falling within this category, what about emotional distress or grief?¹³³ It also fails to address the wider problem. Suppose that an opera singer is injured in a traffic accident in State G leading to the enforced cancellation of his concert in State H. The concert organizer wishes to recover his losses from the person responsible for the accident.¹³⁴ The effect of Marinari, as the Commission explains,¹³⁵ is that the opera singer’s claim to recover his own lost income, like his personal injury claim, would prima facie‎ be governed by the law of State G. There seems no reason why his employer, the concert organizer, should be treated any differently. His own loss, although apparently sustained in State H, is an adverse consequence of an event that has already caused damage elsewhere, albeit to a different, primary victim. Applying the law of State G to the secondary victim’s claim not only promotes consistency as to the law applicable to the claims of the two victims, it also links the applicable law to a tangible, recordable fact (the accident) the location of which can, without difficulty, be identifi ed by both parties after the event. The location of that fact is also foreseeable to both the tortfeasor and the primary victim at the time of the event giving rise to damage. Although it may not be foreseeable to the concert organizer, the balance between the parties seems in favour of applying the law of State G (the country of damage to the primary victim) rather than that of State H (the country of damage to the secondary victim).¹³⁶

4.43  The submission that, at least, financial or other non-material consequences to the claimant that reflect or follow from the primary victim’s damage should be treated as ‘indirect consequences’ of the event giving rise to (p. 318) damage within Art 4(1) may be supported by other examples. Thus, if the opera singer were a passenger in a car damaged in the accident, should his claim for lost earnings be treated any differently depending on whether he is injured or is the owner of the damaged vehicle or is merely delayed by the incident? What if he is driving a car caught in a traffic jam caused by an accident ahead of him in the road? In each of these cases, the opera singer’s financial loss should be seen as an indirect consequence of personal injury or damage to property sustained, whether by him or by another victim, at the time and place of the accident.

4.45  There seems no rational basis for distinguishing between relations of the primary victim and other indirect victims, such as the concert organizer in the example given above. Whether, by virtue of Recital (17) and the above statement of the Commission, a distinction can be drawn between ‘financial’ and other ‘non-material’ damage, on the one hand, and material damage (including personal injury and damage to property), on the other, with the latter always being taken into account in applying Art 4(1), remains to be seen.

4.46  Having identified the adverse consequences of the event giving rise to damage, and excluded those that are ‘indirect’, the next step is to locate the damage in a particular country (or countries¹⁴⁰) in order to identify the law(s) applicable to the non-contractual obligation arising out of tort/delict under Art 4(1). That will, in many cases, involve a straightforward (p. 319) verification of a fact. In other cases, as appears below, it will be more difficult to locate the damage for the purposes of the Regulation.

4.47  As noted, Recital (17) provides that, in cases of personal injury or damage to property, the country in which the damage occurs should be the country where the injury was sustained or the property was damaged respectively. In the case, for example, of road traffic accidents, this should cause no difficulty. The place where personal injury was ‘sustained’ or damage to property ‘occurred’ will not, however, always be certain. Two types of uncertainty may arise. First, evidential uncertainty as to the point in time and place at which the damage occurred. Secondly, legal uncertainty as to the meaning of ‘personal injury’ or ‘damage to property’ or resulting from the location of damage being outside the territory of a State. Neither the first nor, probably, the second type of uncertainty would seem to take a matter outside the scope of the Regulation (3.311–3.314 above).

4.48  Both types of uncertainty may occur in cases concerning international transport, by road, rail, air, or sea. In the case of the carriage of persons by road or rail, it will not usually be difficult to locate the place of injury arising by reason of an incident occurring on, or in relation to, the mode of transport. Such situations are not, it is submitted, isolated from the countries through which vehicles may pass merely because the passengers may not engage with the outside world. The carriage of persons by sea or air, however, seems different in character, and raises three particular issues. First, there may be a genuine practical difficulty in locating a vessel or aircraft at the time of a particular event, particularly one that occurred exclusively on board the aircraft (Issue A).¹⁴¹ Secondly, particularly in the case of aircraft, the passengers are, in a very real sense, isolated from the countries through whose territorial waters or over which they pass (Issue B). Thirdly, the ship or aircraft may pass through, or across, areas not subject to the territorial sovereignty of a State, including the high seas, and it is possible that the event may occur during that passage (Issue C).¹⁴²

4.49  It appears possible for Issue A, raising a problem of evidential uncertainty, to be dealt with under the Regulation in one of two ways. The first solution would be to require the court, by considering evidence of factual and expert witnesses and applying its rules of evidence and the burden (p. 320) of proof, to reach a conclusion as to which State’s territory the ship or aircraft was sailing through or flying over at the time that the damage occurred. The second solution would be to treat non-contractual obligations arising during the international transit of aircraft and ships as a special category and to mould the connecting factor so as to remove the evidential difficulty. Neither of these two solutions is obviously correct. From the standpoint of international law, the first solution has a powerful attraction. It accords, in particular, with Art 1 of the Chicago Convention on international civil aviation, which provides that ‘every State has complete and exclusive sovereignty over the airspace above its territory’,¹⁴³ and with Art 2 of the UN Convention on the law of the sea concerning sovereignty over the territorial sea.¹⁴⁴ Having regard, however, to the isolation of the ship or aircraft from its surroundings (Issue B), it seems artificial, especially for incidents occurring on board an aircraft, to conclude that damage occurring (for example) in an aircraft flying from Paris to Warsaw occurred ‘in’ Germany, even if it is established on the balance of probabilities that the event occurred over German territory.¹⁴⁵ Of course, in many cases, the Regulation, by providing an ‘escape clause’ such as that in Art 4(3), may allow the court to sidestep the evidential difficulty by concluding that, whatever country is identified as a result of the enquiry, the fact that the ship or aircraft was in transit significantly weakens the link with the country identified by the primary connecting factor and supports the conclusion that the non-contractual obligation is manifestly more closely connected with another country. Among the candidates for the country with which the obligation is more closely connected would be (a) the State of departure, (b) the State of (scheduled) destination, (c) the State in which the ship or aircraft is registered, or (d) the country of habitual residence of its operator.¹⁴⁶ That, however, leads one to question the sense of an evidential enquiry that requires the resources of the court to be deployed to produce a result which, if accepted, is unappealing and, if rejected, is worthless. An enquiry of this kind would also be fruitless if it were to result in a finding that the damage occurred on or over the high seas or another area not subject to the territorial sovereignty of a State (Issue C).

4.50  These issues must also be faced in the case of the carriage of goods by sea or air. The international carriage of goods by whatever means presents additional difficulties by reason of the fact that goods are not sentient and (p. 321) cannot tell or protest when they have been damaged. That damage may be discovered days, weeks, or months later, for example when a container is opened for inspection on arrival of a vessel.

4.51  Liability of carriers of persons and goods by land, sea, or air, and some other aspects of the civil liability of those concerned with international modes of transport (e.g. limitation of a shipowner’s liability¹⁴⁷), are regulated by international conventions to which EC Member States are party and which, in order to ensure their efficacy, will normally require to be given overriding (mandatory) effect by a State party, whether or not the law applicable to the obligation in question under the Rome I Regime or the Rome II Regulation (as applicable) is the law of another State party.¹⁴⁸ The law applicable under the Rome II Regulation will continue to apply to non-contractual obligations in situations not regulated in the forum Member State by international convention, or in which any relevant convention only partially harmonizes the civil liability rules. Accordingly, these international conventions do not provide a complete solution to the problems described above even within the areas to which they apply.

4.52  Issues of this kind have been addressed by the ECJ, within the framework of Art 5(3) of the Brussels Convention, in Réunion Européenne v Spliethoff’s.¹⁴⁹ In that case, pears carried by sea from Australia to The Netherlands and then by road to France were discovered by the consignee on arrival to have been damaged during transit. The consignee’s insurers, exercising their right of subrogation, sought in the French courts to recover from the issuer of the bill of lading, the sea carrier, and the master of the vessel the amount that they had paid to the consignee. The Dutch owners and master of the vessel contested jurisdiction under the Brussels Convention. The insurers argued that jurisdiction existed under Art 5(3) of the Convention (matters relating to tort, delict, or quasi-delict) on the basis that ‘damage’ had occurred in France, where the goods were eventually delivered and the damage discovered. The ECJ rejected that argument. For this purpose, (p. 322) it might well have sufficed for the ECJ to point out that it was for the French court seised of the dispute to determine when, and where, ‘damage’ had occurred,¹⁵⁰ drawing attention to the requirement under the Convention (as now under the Rome II Regulation¹⁵¹) that the damage be the ‘direct’ consequence of the event for which the defendant was said to be responsible. On any view, given that the pears had already travelled halfway around the world, it is difficult to see how the French court could have concluded that the pears were ‘damaged’ only after they were unloaded and had entered France. Whatever the outcome, however, this question of fact would not have been a matter for the ECJ to determine.

4.54  The Court explained that this place ‘meets the requirements of foreseeability and certainty imposed by the Convention and displays a particularly close connecting factor with the dispute in the main proceedings, so that the attribution of jurisdiction to the courts for that place is justified by reasons relating to the sound administration of justice and the efficacious conduct of proceedings’.¹⁵⁵ That reasoning appears consistent with the objectives underlying Art 4(1) of the present Regulation, to ‘enhance the foreseeability of court decisions and ensure a reasonable balance between the interests of the person claimed to be liable and the person who has (p. 323) sustained damage’.¹⁵⁶ With that in mind, there is every reason to believe that, if the ECJ were to be presented with a case under the Rome II Regulation involving international transport in a type of situation (such as that in Réunion Européenne) in which evidential difficulties commonly arise, it will direct its reasoning in such a way as to locate the damage in a particular country by reference to an objectively ascertainable factor present in all situations of that type that links the damage to a particular country.¹⁵⁷

4.56  An autonomous solution to locating the country of damage in cases of international transport has the added advantage that it addresses the third issue referred to above, i.e. the possibility that damage may be found, in (p. 324) fact, to have occurred on or over an area not subject to State sovereignty. In such cases, application of the law of the flag or registration State would provide a workable, if not perfect, starting point in determining the law applicable to damage resulting from events on board ships or aircraft on or over the high seas. That outcome also seems consistent with international conventions concerning criminal jurisdiction.¹⁶⁰ In the case of collisions between ships and aircraft on or over the high seas, the logic and structure of Art 4(1) would appear to favour application of the law of the flag or registration State of the damaged ship/aircraft, not that of the ship/aircraft causing damage.¹⁶¹

4.57  Otherwise, the only viable solution in cases where damage, in fact, occurs on or over an area not subject to State sovereignty would be to neutralize Art 4(1), leaving the law applicable to the obligation in question to be determined solely by reference to Arts 4(2) and (3). That approach could not, it may be noted, be extended to other rules within the Regulation that contain no flexible rule of displacement, for example in a situation in which one ship releases pollutants which damage another ship (or crew member) (Art 7) or the crew of a ship threaten industrial action while the ship is on the high seas (Art 9). If the solution of applying the law of the flag or registration State were rejected in these cases, a default rule would be required, perhaps favouring the law of the forum, faute de mieux‎. This appears a less satisfactory solution.

4.58  Similar difficulties in identifying the country of damage will also arise in cases not involving international transport. For example, consider a case in which a contractor is exposed to asbestos in State J while working temporarily away from his home in State K and, many years later, moves to State L where he is diagnosed as suffering from an asbestos related disease. He sues the owner of the building in which he was working.¹⁶² It is suggested that, in this case, the injury should be considered to have been ‘sustained’ in State J where the claimant was exposed to the asbestos, and not in State K (where the disease most likely developed), or State L (where the disease was diagnosed). This allows the applicable law to be identified by reference to an objective factor, ascertainable by both parties and appears, in particular insofar as it excludes the country of diagnosis, (p. 325) consistent with the approach of the ECJ in Réunion Européenne. The deterioration of the claimant’s health in States K and L should be considered as indirect consequences of the original injury sustained in State J.

4.59  Another problematic area concerns the unborn child. Two types of situation may be contemplated. In the first, perhaps more likely to arise in cases of product liability,¹⁶³ an event affecting the child’s mother may result in injury to the child,¹⁶⁴ with a claim being brought by it or on its behalf. If the child is en ventre sa mère‎ at the time of the event (e.g. a car accident), it seems appropriate to conclude that the child has also been injured at that time and in that country. What if, however, the child has not been conceived at the time that damage is caused to the mother (e.g. by swallowing medication)? Four possible solutions may be suggested. First, the child should be deemed to have sustained injury at the same time and in the same place as the mother. Objection may be taken to that approach, however, on the ground that it does not treat the child as a separate life. Secondly, the child should be taken to have sustained injury at the moment and in the place of its conception. Thirdly, the child should be taken to have sustained injury in the country where the mother was habitually resident at the time of conception. Fourthly, the child should be taken to have sustained injury at the moment and in the place of its birth.

4.60  In the second type of situation, an event (e.g. a sterilization or vasectomy) causes the unintended conception of a child, with a claim being brought by the parents of the child.¹⁶⁵ The second, third, and fourth options described above (i.e. country of conception, country of habitual residence at the time of conception, and country of birth) may also be suggested as candidates for the location of damage in this type of case. As a further alternative, it may be argued that the damage in this second category is purely economic, and that the conception of the child should not be treated as ‘damage’, and should be seen as an indirect consequence of the injury caused by the (failed) treatment at the time and place of treatment.

4.61  It is submitted that, in both types of case, the country of birth should be rejected, as being too remote to the event giving rise to damage.¹⁶⁶ The country of habitual residence at the time of conception has its attractions, as it may avoid problems in identifying the place of conception. It may, however, be argued to be inconsistent both with the rejection by the ECJ in (p. 326) Marinari and Kronhofer (4.37 above) of the victim’s country of habitual residence as a general connecting factor for jurisdiction purposes and with the use of common habitual residence as a connecting factor in Art 4(2). The better view, therefore, may be that the country of damage in both types of situation should be where the child was conceived, although the connection with that country may be a weak one (capable of being displaced under Art 4(3)) if, for example, the parents were travelling at the time. The economic loss suffered by the parents in supporting the child should be considered an indirect consequence of its conception, and left out of account for the purposes of Art 4(1).

4.63  That said, Art 2(1) of the Regulation defines ‘damage’ in terms of the consequences of a tort/delict and it would appear difficult to dispute that the conception of a child is a direct consequence of, for example, the negligent performance of a sterilization option. Nor does it stigmatize the child to describe its conception in this way. Further, if this seems too technical a point, the ‘damage’ to the parents may instead be described as a loss of autonomy in matters of family planning and upbringing,¹⁶⁸ a consequence that has its strongest connection to the moment, and the country, of conception.

4.64  The mother herself will sustain injury during birth. If she claims with respect to such injury, as well as pain, discomfort, and financial losses suffered earlier in her pregnancy, it may be argued that all such losses flow from the child’s conception and that all claims arising from that event should be governed by the same law, whether by locating all of the ‘damage’ in the country of conception (which seems artificial, especially in view of the wording of Recital (17)¹⁶⁹) or through application of the escape clause in Art 4(3).

4.65  The second type of situation has been considered by the Dutch courts on more than one occasion in the context of the Brussels I Regulation.¹⁷⁰ (p. 327) In 2003, a district court held that the damage sustained from the child’s birth was economic and occurred in the country where the parents’ acco unts were located.¹⁷¹ In a later case, an appellate court held that, so far as material damage is concerned, damage was suffered in the place where the child was conceived.¹⁷² The court, however, ruled that the nonmaterial damage arising from the breach of the mother’s right to selfdetermination was located in the place where the mother was when she became aware of the pregnancy. That conclusion is difficult to reconcile with the ECJ’s rejection in Reunion Européenne of the place of discovery of damage as identi fying the place of damage within Art 5(3) of the Brussels Convention.¹⁷³ On balance, the country of conception seems the more appropriate connecting factor under Art 4(1) of the Rome II Regulation.

4.66  The task of locating, for the purposes of the Regulation, the financial or other non-material consequences of an event giving rise to damage raises a different set of problems. First, these consequences may lack an immediate, physical manifestation. For example, a credit balance in an Internet bank account may be recorded on a server in a country that has no direct connection with either the customer or the bank. An electronic transfer from that account to a third party’s account consists of nothing more than a stream of electrons.¹⁷⁴ Other forms of non-material damage, such as damage to reputation and to goodwill are equally intangible. Secondly, particularly in the case of transfers of financial assets, the location of damage may be open to manipulation by both tortfeasor and victim. A fraudster may lure his victim to a country whose laws provide a lower level of protection, or a company may decide to concentrate its assets in, or channel transfers through, a particular country on the basis that the laws of that country provide stronger protection against fraud. Thirdly, the tortfeasor’s act may trigger a chain of events involving the victim. At what point is it to be said that the victim has sustained ‘damage’ and that later consequences following on from that ‘damage’ are to be considered as ‘indirect consequences’¹⁷⁵ to be left out of account in determining the applicable law?

(p. 328) 4.67  The most common situation raising problems of this kind concerns the effects of a negligent or fraudulent misrepresentation. The misrepresentation may be initiated by the tortfeasor in one country (State M), received by the victim or the victim’s representative in a second (State N), and present in the mind of the victim at the time that one or more decisions are taken in a third country (State P), leading to particular acts or omissions by the victim or its representatives in a fourth country or set of countries (State Q). Of these, only State M can be immediately disregarded under Art 4(1), as the initiation of the misrepresentation by the tortfeasor constitutes the event giving rise to damage and not damage.¹⁷⁶ The possibility of applying the law of State N can also be dismissed with relative ease, as the misrepresentation is at this stage only latent. Decisions of the English courts have, for the most part, also rejected the possibility of applying the law of State P, instead requiring some form of ‘concrete transaction’ as a result of the victim’s reliance on the misrepresentation.¹⁷⁷ There is eminent good sense in that requirement, as the adverse consequences of the misrepresentation would appear to be reversible until that point. Indeed, it is submitted that the test of ‘reversibility’ is helpful in terms of identifying the transaction or other act or omission on the faith of the misrepresentation that constitutes ‘damage’ to the victim for this purpose, and in reconciling the earlier decisions of the English courts concerning the location of the place of damage under Art 5(3) of the Brussels Convention. If, on the faith of the representation, the victim enters into a contract or other transaction with the tortfeasor, it may be appropriate to look through that transaction and instead to identify the country of damage with that in which the victim or his representative irreversibly incurred expenditure or liability or in which he lost control of assets or other resources in the performance of the contract.¹⁷⁸ The fact that the victim may undertake a liability towards the tortfeasor at the moment he signs the contract should be ignored, on the basis of a strong presumption at least that the contract, while executory, is reversible either by setting it aside or by a refusal to perform on the basis of misrepresentation. If, however, the victim performs, or fails to perform, some act towards a (p. 329) third party, whether by entering into a contract¹⁷⁹ or by transferring an asset¹⁸⁰ or otherwise,¹⁸¹ resulting in a diminution in the victim’s assets or in his incurring liability to that third party then that act or omission should be considered to have caused him ‘damage’ at the moment, and in the country, where the act or omission becomes irreversible by his own actions. In the case of entry into a contract, the country of damage should be that of the country in which the contract was concluded. Although this fact may be open to manipulation and does not in any event appear to provide a particularly powerful connecting factor,¹⁸² it can be objectively ascertained and seems preferable (for example) to the place where the third party may enforce the contract against the victim or the place of performance of the victim’s obligations towards the third party. In the case of transfer of an asset, the country of damage should be that in which the asset was located immediately prior to transfer (and not the place where the victim gives instructions for its transfer). In the case of transfers from bank accounts, this country should be that in which the branch at which the account is held is situated or, in the case of Internet or other delocalized accounts, the bank’s habitual residence determined in accordance with Art 23. The intangible nature of financial and other material loss, and the propensity for the connecting factors to be manipulated, may be factors to be taken into account in applying the ‘escape clause’ in Art 4(3) (4.87 below).

4.68  Similar problems must be addressed in other situations involving nonmaterial consequences of an event giving rise to damage. For example, what of a situation in which the first defendant (D1) in State R hacks into a computer server located in State S, steals information confidential to the claimant (C, a company habitually resident in State T), and transfers it electronically to the second defendant (D2) in State V, who then puts it to use in the manufacture of products sold in competition with C’s products around Europe. Here, the consequences of the event(s) giving rise to damage are beyond the claimant’s control. As against D1, there would appear to be a (p. 330) strong case for identifying State S as the country of damage, as this is where the confidentiality of the claimant’s information was first compromised and diluted by that defendant’s actions. That, in a case such as this, the location of the server may provide only a tenuous connection to the tort/delict is again a factor to be taken into account in applying the Art 4(3) ‘escape clause’. As against D2, it is not clear whether Art 4 applies at all, or if D2’s obligation must be considered to arise out of an act of unfair competition falling within Art 6(1) and not excluded from its scope by Art 6(2).¹⁸³ If Art 4 does apply, there are arguments for locating the ‘damage’ in (1) State S (if it is sought to hold D2 responsible for D1’s conduct), (2) State R (on the basis that this is the country from which the unauthorized disclosure was made by D1 to D2), (3) State V (on the basis that this is the country in which the confidential information was misused in the manufacture of goods), or (4) the countries in which the markets on which D2’s products were sold in competition with C’s products were located. Of these, State R, as the country in which D2’s conduct directly interfered with the obligations that D1 owed to the claimant, appears the strongest candidate (and would have the advantage of fixing the damage in D2’s case in a single country), but the position is far from clear.¹⁸⁴ The only country that can be safely excluded from consideration is State T, as any damage suffered there by the claimant must be seen as an indirect consequence of damage occurring elsewhere.¹⁸⁵

4.69  If the relevant damage is located in more than one country, the rule of applicable law in Art 4(1) requires the law of each of those countries to be applied on a distributive basis (the so-called mosaic view or Mosaikbetrachtung‎).¹⁸⁶ Having regard to the ECJ’s decision in Shevill v Presse Alliance SA,¹⁸⁷ which applies the same concept in the context of the Brussels Convention, this distributive approach would appear to require the Member State court seised of the dispute to apply a country’s law only with respect to damage occurring in the territory of that country (as well, presumably, as damage in other countries that is considered to be an ‘indirect consequence’ of the damage occurring in that country) and then to combine the (p. 331) results of applying each country’s law to determine the outcome of the case before it. That seems straightforward enough if what the claimant seeks is compensation for his injury, and such an award is available in all of the countries whose law applies. In this type of case, the court (assuming that it has jurisdiction over the entire claim¹⁸⁸) must proceed as if there were as many different obligations as the number of countries whose law applies and must combine the compensation awarded under the law of each country in a single award.

4.70  The mosaic view approach is, however, much less satisfactory in cases in which the claimant seeks a remedy, whether monetary or non-monetary, which is available and would be given to him under the rules of one or more, but not all, of the countries concerned. Particular difficulties arise in the case of claims to prohibit future breaches of non-contractual obligations or to prevent further damage from an existing breach. One specific criticism leveled at the approach is that the distributive application of laws to a non-contractual obligation is unsatisfactory in its application to claims to prevent an activity (such as marketing via the Internet¹⁸⁹) which produces its effects simultaneously in many countries, or globally, and in such a way that those effects cannot easily be controlled by reference to national boundaries.¹⁹⁰ In such cases, it is argued, fragmentation of the applicable law allows a claimant to select the law most favourable to him from among those that apply and to seek an injunction to prohibit the harmful effects of the activity in that country, an order that in practice will compel the defendant to cease his activity even in countries where it is entirely lawful. These concerns, however, may be overstated. In particular, the defendant in an injunction case of the type referred to above can legitimately seek to rely on the law of other countries where his activity produced its effects to argue that an injunction is an inappropriate measure to terminate injury or damage in the circumstances of the case.¹⁹¹ Further, the prohibition by a Member State court of cross-border activity must be consistent with the requirements of the EC Treaty and, where applicable, the E-Commerce Directive.

4.72  As a result, it may be feared that the claimant’s ability to protect his confidential information would, in practice, be limited to the lowest common denominator among the remedies available under the laws applied on a distributive basis. It would not appear open, except perhaps on the basis of public policy, for the court to award a remedy for a non-contractual obligation with respect to ‘damage’ likely to occur in a particular country that is not available under the law of that country.

4.73  In the face of the challenges of applying several potentially conflicting laws in a harmonious and just manner, it will be interesting to see whether Member State courts will seek to interpret the connecting factor in Art 4(1) in such a way as to locate the damage in a single country or will resort to the escape route provided by Art 4(3), allowing a single law to be applied with respect to all of the consequences of the tort/delict. In the case of confidential information, for example, it might be possible to identify the country of damage as being that in which the confidentiality of the claimant’s information was or would be originally compromised as a result of the defendant’s conduct and to treat later disclosures or abuses of that damage as indirect consequences of that original damage, to be left out of account. Alternatively, Art 4(3) may be relied on to displace the otherwise applicable law(s) in cases where there is a pre-existing relationship between the parties. Finally, if the obligation to protect the claimant’s confi dential information arises from an agreement between the parties, whether as an express or implied term, the matter may be treated as involving a ‘contractual obligation’ falling outside the scope of the Rome II Regulation and governed by the law applicable to the contract under the Rome I Regime.

(p. 333) 4.74  The foregoing discussion has proceeded on the basis that the Member State court has jurisdiction over the entire dispute between the parties. Under the Brussels I Regulation, however, a court other than those of the defendant’s domicile or the place where the event giving rise to damage occurred may have jurisdiction only with respect to damage occurring within its territorial jurisdiction.¹⁹⁴ In such a case, Art 4(1) will point in practice towards application of the law of the forum.

4.75  Art 4(1) opens with the words ‘unless otherwise provided for in this Regulation’. According to Recital (19) to the Regulation, ‘specific rules should be laid down for special torts/delicts where the general rule does not allow a reasonable balance to be struck between the interests at stake’. The relationship between Art 4 and the other provisions in Chapter II has been considered at 4.07–4.08 above.

4.77  In the explanatory memorandum accompanying its proposal, the Commission suggested that these rules of displacement reflect the concern that ‘application of the basic rule might well be inappropriate where the situation has only a tenuous connection with the country where the damage occurs’.¹⁹⁵

4.78  Professor Symeonides has argued that the principal defect of the rules of displacement in Arts 4(2) and 4(3) of the Rome II Regulation is that they do not allow the general rule to be departed from on an issue-by-issue (p. 334) basis (depeçage‎).¹⁹⁶ The issue-by-issue approach would have been allowed by the single rule of displacement adopted by the European Parliament in its 1st Reading Position,¹⁹⁷ following (in this respect at least) the preexisting UK statutory rules.¹⁹⁸ This aspect of the Parliament’s proposal was criticized as ‘[e]xtremely dangerous for a consistent and homogeneous appreciation of a case’,¹⁹⁹ a criticism that appears well-founded. There is, of course, no systemic reason why a legal system cannot, through its rules of private international law, combine rules originating in different legal systems in order to regulate a particular situation.²⁰⁰ Indeed, the provisions of Chapter V and Art 26 of the Regulation²⁰¹ openly contemplate the possibility of more than one law applying to different aspects of a single case, and matters of evidence and procedure remain outside the Regulation’s scope.²⁰² It is nevertheless desirable that the approach taken by a foreign legal system in regulating the civil consequences of particular events should, so far as possible, be treated as a coherent whole. The assumption, while not always reflecting constitutional reality, should be that choices made by a country’s legislature or courts in relation to one element of its regulation of these matters are reflected in choices made in relation to other elements. Thus, for example, the imposition of a stricter standard of responsibility may be counterbalanced by limits on the level of liability.²⁰³ Although cross-border situations undoubtedly present different challenges for legislatures and courts, and may justify different regulatory choices, the task of judges should be to apply the content of foreign law as it stands to the facts of the case before them and not to speculate as to whether different choices might have been made by a foreign legislature or court if it had the facts of the instant case in mind. Allowing the law applicable to non-contractual obligations to be determined on an (p. 335) issue-by-issue basis creates the risk of either a ‘slice and dice’ approach to civil liability, creating a modified obligation that is unrecognizable to its legal system of origin, or of placing an unsupportable burden on individual judges requiring them to consider whether applying a different law to a particular issue would undermine the interests of the country whose law would apply to the remaining issues. Although the ‘all or nothing’ approach is not perfect, and it has given rise to genuine concerns in the context of traffic accidents,²⁰⁴ it has the attraction not only of coherence but also of enhancing legal certainty.

4.79  In any event, it is now too late to re-write history. The current position is clear: neither Art 4(2) nor Art 4(3) permits depeçage‎, which occurs under the Regulation only to the extent that it is mandated or permitted under Chapter V or as a result of the exclusion of matters of evidence and procedure. Under Art 15, the law applicable to non-contractual obligations under the Regulation applies to all of the matters set out. It is not possible for the court to divide those matters, with one law applying to some of them and another law applying to the others.

4.81  This last statement appears too sweeping.²⁰⁶ Although the application of the law of the country of habitual residence of the relevant actors to (p. 336) determine certain aspects of non-contractual liability between them may in some circumstances reflect their reasonable expectations, particularly if they have a pre-existing relationship, it is not difficult to dispel the notion that this is a universal truth. Take, for example, the situation in which the claimant (A) is struck by a car driven by the defendant (B) while crossing a road in France. A and B are both habitually resident in England. Under the French loi Badinter‎, liability for traffic accidents is strict.²⁰⁷ Under English law, liability is based on negligence, albeit subject to a high standard of care.²⁰⁸ Damages for personal injury in England are typically higher than those in France.²⁰⁹ In these circumstances, it can scarcely be said to reflect the parties’ legitimate expectations that the basis and extent of B’s liability should be governed by English law. Even if one supposes that A is a passenger in the car and that A and B are husband and wife (or employee and employer), the application of English law to judge B’s responsibility for his conduct might well come as a surprise to A, although both parties would perhaps be more willing (at least in principle) to accept the proposition that questions of compensation should be determined by their place of common habitual residence.²¹⁰ If the facts are reversed, with the incident occurring in England and A and B being habitually resident in France, B might be similarly surprised to find that he owes a higher standard of responsibility when he runs over a fellow countryman than when he injures an Englishman or another foreigner in England.

4.82  In these, and similar circumstances, it may be argued that the ‘escape clause’ in Art 4(3) provides a solution for any situation in which Art 4(2) may be thought to depart from the parties’ reasonable expectations. The difficulty with this argument, as both the language of the Regulation and the Commission recognize,²¹¹ is that Art 4(3) imposes a very high threshold for the displacement of the law applicable under Art 4(1) or Art 4(2). That threshold would almost certainly not be reached in the second scenario contemplated above (where A is a passenger and has a relationship with B based in England). It would provide a more promising argument in (p. 337) the first scenario (where A has no prior relationship with B), but this identifi es the principal weakness of the ‘common habitual residence’ rule in Art 4(2), that it automatically displaces the general rule on the basis of a connection between the parties that is much weaker than the type of pre-existing relationship that, in Art 4(3), provides only a possible basis for displacing the otherwise applicable law. In other words, Art 4(2) gives too much weight to the commonality of the parties’ residence. It would have been preferable if a ‘common habitual residence’ had been identified in Art 4(3) as a factor that might, depending on the circumstances, justify displacement of the general rule.²¹² Again, however, it is too late to re-write the regulation.

4.83  Other aspects of this rule of displacement require further explanation, as follows:

1. 1. Habitual residence: The concept of ‘habitual residence’, partly defined in Art 23, is considered at 3.47–3.58.

2. 2. ‘Person claimed to be liable’: Art 4(2) refers to ‘person claimed to be liable’ (in the French text, ‘la personne dont la responsabilité est invoqée‎’), a term appearing also in Arts 5(1) and 17. Four possible meanings of this expression may be canvassed, each of which may produce different results in individual cases. First, that it refers, individually, to each person (usually in the position of a defendant²¹³) whose responsibility it is sought to establish in the proceedings, the test being applied on a party-by-party basis. Secondly, that, notwithstanding the use of the singular ‘person’, it is capable of referring, collectively, to all of the persons whose responsibility it is sought to establish in the proceedings. Thirdly, that it is capable of referring to every person against whom a claim for particular damage has been or might in the future be made, whether or not that person is sued in the proceedings. Fourthly, that it refers to the person (or, possibly, persons) whose conduct caused the damage, whether or not that person is sued.²¹⁴ Of these, the first is to be preferred, as being more easily reconciled (p. 338) not only with the legislator’s choice of words but also with Recital (18),²¹⁵ with the view of the Commission,²¹⁶ and with the demands of legal certainty. It also appears consistent with the approach taken by the ECJ in applying the lis alibi pendens‎ rules under the Brussels Convention.²¹⁷ The second and third possible meanings, on the other hand, would generate uncertainty: the second would potentially allow the claimant to manipulate the applicable law by choosing which defendants to sue and the third is unworkable, as there appears no basis for assessing whether a particular claim is justified or not. The fourth possible meaning seems inadequate, as a claim of responsibility in tort/delict may be based on an event that is not an act or omission of a particular person (e.g. injury caused by an animal or building). Further, Art 15(g), which applies the law applicable under the Regulation to ‘liability for the acts of another person’ appears inconsistent with this meaning.

   Even so, the need to adopt a party by party approach highlights further deficiencies in Art 4(2), as (unless the court intervenes by deploying the escape clause in Art 4(3)) the application of the common habitual residence rule may result in claims arising from the same event and with respect to the same damage being governed by different laws. In particular, in cases of joint, common, or vicarious liability, Art 4(2) may displace the generally applicable law for claims against one defendant, habitually resident in the same country as the person sustaining damage, but not another. If, in the example given above, a French resident driver (C) had also been involved in the accident or if B had been acting in the course of his employment for his French-based employer (D), Art 4(2) would apply to a claim by A against B but not against C or D. This, in turn, may raise complex issues when it comes to assessing questions of joint responsibility and contribution.²¹⁸

3. 3. ‘Person sustaining damage’: Art 4(2) also refers to the ‘person sustaining damage’.²¹⁹ At first sight, it might appear possible to argue that this choice of wording supports the view that Art 4(2) applies only to cases of damage to the person. In this connection, it may be noted that Art 7 refers to ‘damage sustained by persons or property’ in restricting the kinds of damage that fall within the scope of that Article. There is, however, no suggestion in the travaux préparatoires‎ that the habitual residence exception was intended to be so restricted. Further, the French language version of the (p. 339) Regulation uses the expression ‘la personne lésée‎’ in Art 4(2), while referring instead in Art 7 to ‘dommages … subis par des personnes ou causés à des biens‎’. Art 4(2) should apply, therefore, whatever the nature of the damage.

   Although Recital (18) refers to ‘the parties’, the choice of terminology here and elsewhere in the Regulation suggests that the identity of the ‘person sustaining damage’ will not necessarily correspond to the party (usually in the position of claimant) who seeks to establish the responsibility of another in the proceedings. In this connection, a contrast may be drawn with the expression ‘the person seeking compensation’ which appears in Arts 6(3)(b) and 7. Against this background, the expression ‘person sustaining damage’ appears capable of bearing one of two possible meanings, depending in turn on the meaning given to the term ‘damage’ in Art 4(2). If ‘damage’ bears the (broader) meaning given to it in Art 2(1), it would seem to follow that the ‘person sustaining damage’ is the person affected by the consequences of the tort/delict in respect of which the proceedings have been brought, whether those consequences are characterized as direct or indirect. If, however, ‘damage’ bears the (narrower) meaning given to it in Art 4(1) (excluding indirect consequences of the harmful event), the ‘person sustaining damage’ will be the person who is the primary victim of the tort/delict. On the view taken above as to the basis for separating direct and indirect consequences within Art 4(1),²²⁰ that person may be neither a claimant nor a person represented by a claimant.

   For consistency, it is suggested that ‘damage’ should bear the same meaning in Art 4(2) as it does in Art 4(1) and that the term ‘person sustaining damage’ should be read as referring to the primary victim of an event giving rise to damage. For example, if, in the example given above,²²¹ A had been killed in the traffic accident and A’s widow E (habitually resident in Poland) had sued B both as A’s personal representative and, in her own right, claiming damages for loss of support and bereavement, A would remain the ‘person sustaining damage’ for the purpose of all of the claims brought by E, with the consequence that Art 4(2) would apply (leading to the application of English law) notwithstanding her habitual residence in Poland. The common habitual residence of the person claimed to be liable and the actual claimant may, however, be a factor to be taken into account in applying the more flexible rule of displacement in Art 4(3).

4. 4. Timing: The test of common habitual residence must be applied at the time when the relevant²²² damage occurs, not (for example) the date of the (p. 340) harmful event or the date of proceedings. If that damage is sustained over a period of time, it is submitted that the earliest point in time should be taken. A change in habitual residence after that point in time will not affect the application of the rule of displacement in Art 4(2),²²³ although it may be a factor to be taken into account in applying Art 4(3).

That suggestion was not accepted by the Council in its Common Position,²²⁹ and did not find its way into the Regulation in its final form.

4.86  Art 4(3) requires the Member State court, in applying the rule, to consider ‘all the circumstances of the case’. For this purpose, the court should not limit itself to considering the matters relied on by the claimant to support the claim. As the following words indicate, however, those circumstances must be relevant to the question whether the tort/delict is manifestly more closely connected with a country (State B) other than that whose law applies under Art 4(1) or (2) (State A). The court cannot, therefore, apply a different test, by considering (for example) whether justice would be better served by the application of the rules of State A than those of State B.²³⁰ The reference in Art 4(3) to ‘the tort/delict’ (in the French text, ‘fait dommageable‎’)²³¹ should here be taken to refer in combination to the event giving rise to damage and all of the consequences of that event, including indirect consequences. Against this background, and having regard to Art 24 which excludes renvoi‎, it would not appear to be permissible for a court to have regard to the rules of applicable law that would be applied by the courts of State A (or State B) if they were to be seised of the matter.²³² Such rules are extraneous to the question that the court that is actually seised of the dispute between the parties must ask itself under Art 4(3).

(p. 342) 4.87  The case for applying State A or State B may be strengthened (or weakened), for example, by:

1. 1. The existence of circumstances linking the tort/delict to State A or State B, including, without limitation, (a) the presence (or absence) of factual connectors other than damage (Art 4(1)) or common habitual residence (Art 4(2)) linking the tort/delict to State A, (b) any factual connectors linking the tort/delict to State B, (c) any pre-existing relationship between the parties,²³³ and (d) the personal connections of the persons involved, including (where applicable) (i) the parties, (ii) the tortfeasor, if not the defendant, (iii) the victim, if not the claimant, (iv) any primary victim not represented by the claimant, and (v) any other person whose conduct is closely connected to the tort/delict (including a person claimed to be jointly liable).

2. 2. The permanence (or transience) of the circumstances that link the tort/ delict to State A or State B.

3. 3. The nature of the circumstances linking the tort/delict to State A or State B and, in particular, whether those linking factors were foreseeable, tangible, accidental, artificial, or open to manipulation.

4. 4. The occurrence of personal injury or damage to property (as opposed to non-material damage) in State A.

4.88  If, in all the circumstances of the case, there is any room for doubt as to whether the tort/delict is more closely connected to State A than State B, the law of State A should be applied. The burden of proving the existence of circumstances establishing a manifestly closer connection should be on the party who seeks to displace the law applicable under Art 4(1) or 4(2).²³⁴

4.89  The following additional points may be made:

1. 1. No depeçage: Art 4(3), like Art 4(2), operates on an ‘all or nothing’, and not an issue-by-issue, basis.²³⁵ The court must look for connections to the ‘tort/delict’ and not to the particular issues which the parties have presented for determination. Whether, for example, the defendant has admitted liability or quantum is agreed between the parties should not be a relevant factor in the application of the ‘escape clause’.²³⁶ It follows that Art 4(3) is significantly less flexible than the rule of displacement in s 12 of the UK Private International Law (Miscellaneous Provisions) (p. 343) Act 1995,²³⁷ which is capable of being applied to the determination of a particular issue.

2. 2. Relationship with Arts 4(1) and 4(2): Art 4(3) refers to the application of the law of a country ‘other than that indicated in paragraphs 1 or 2’.²³⁸ The use of the plural, ‘paragraphs’, might be thought to suggest that Art 4(3) cannot be invoked to displace the law applicable under Art 4(2), by reason of common habitual residence, in favour of the law originally identified by reference to Art 4(1).²³⁹ On this view, Art 4(3) also could not be applied to select one of a number of laws applicable under Art 4(1) by reason of the occurrence of damage in more than one country.²⁴⁰ Neither argument can be accepted. The Commission described this as a ‘general exception clause’ and its sphere of operation should not be restricted without clear words. The reference should be to a country different from that (or those) indicated by whichever of paragraphs 1 or 2 has identified the law from which it is sought to escape.

3. 3. Timing: Unlike Art 4(2), Art 4(3) does not specify the time when the relevant circumstances connecting the tort/delict to State A or State B are to be compared. As a result, it should be applied at the time that the law applicable to the non-contractual obligation is determined. Indeed, Art 2(3)(b) would appear to contemplate that the court will have regard not only to fresh consequences of the event giving rise to damage, occurring between the date of issue of proceedings and the date on which it finally determines the applicable law but also any further consequences that are likely to occur in the future.

4.90  Art 4(3) gives only one example of circumstances in which a manifestly closer connection might exist with a country other than that whose law applies under Art 4(1) or Art 4(2), referring to ‘a pre-existing relationship between the parties, such as a contract, that is closely connected with the tort/delict in question’. The wording of the Regulation, however, is not specific as to whether this connecting factor points towards application of the law of the country in which the pre-existing relationship is based or the law applicable to the pre-existing relationship. Although in many cases, a relationship will be governed by the law of the country in which (p. 344) it is based, that will not always be the case. Some support for the view that the reference should be exclusively to the law of the country in which the relationship is based is provided by (a) the apparent contrast in the first sentence of Art 4(3) between ‘country’ and ‘law of the country’, suggesting perhaps that a geographical rather than legal connection was intended, and (b) the undoubted fact that not every pre-existing relationship will have an identifiable governing law.²⁴² On the other hand, it may be countered that (1) Art 25 links the concept of a ‘country’ to a territorial unit of a State with its own rules of law, thereby weakening any inference to be drawn from the different terminology used in Art 4(3), and (2) more signifi cantly, the explanatory memorandum accompanying the Commission Proposal strongly supports the conclusion that the rule is capable of, and indeed will usually be taken to, connect the tort/delict with the law applicable to the relationship in question.²⁴³ Thus, the Commission Proposal refers repeatedly to ‘the law applicable to the pre-existing relationship’ as well as to a ‘choice of law clause’.²⁴⁴

4.91  Accordingly, the fact of a pre-existing relationship would appear capable of supporting a manifestly closer connection to either the country whose law applies to that relationship or the country in which the relationship is centred. That is consistent with the approach taken under the pre-existing English statutory rules.²⁴⁵ On this view, as the law applicable to a preexisting relationship²⁴⁶ cannot be determined by reference to the Regulation’s rules of applicable law for non-contractual obligations, that law must be identified by reference to the forum’s other rules of private international law. For relationships other than contractual relationships falling within the scope of the Rome I Regime, which provides a common set of rules across the Community for contractual obligations, that creates the possibility of a lack of consistency between Member State courts in their application of Art 4(3).²⁴⁷

4.93  It remains to be seen whether and, if so, how this restriction will be given effect to by Member State courts and the ECJ on the wording of Art 4(3). One possibility is that Art 4(3) must be taken to connect non-contractual obligations that are closely connected to a relationship arising from a consumer or employment contract falling within Arts 5 and 6 of the Rome Convention, or the corresponding provisions of the Rome I Regulation, not only to the country whose law is chosen by the parties to govern that contract but also to the consumer’s country of habitual residence or, as the case may be, the country whose law would apply to the employment contract in the absence of choice. That dual connection may make it more diffi cult for the non-consumer or employer to rely on Art 4(3) to displace the law otherwise applicable under Arts 4(1) and 4(2). That does not mean, however, that the rules of applicable law for consumer and employment contracts in the Rome I Regime should be applied automatically to non-contractual obligations, so as (for example) to require the application of the law of the consumer’s country of habitual residence. The flexibility inherent in Art 4(3) remains in all cases. Similarly, although the requirement in Art 14 that agreements on choice of law made before the event giving rise to damage occurred be ‘freely negotiated’²⁵⁰ cannot be read (p. 346) into Art 4(3), the fact that the relationship between the parties is governed by a non-negotiable contract containing a choice of law provision may be taken into account as a factor diluting the strength of the connection to the law applicable to that contract.

4.94  The Commission also suggested that the reference to a pre-existing relationship ‘is flexible enough to allow the court to take account of a contractual relationship that is still only contemplated, as in the case of the breakdown of negotiations or of annulment of a contract, or of a family relationship’.²⁵¹ Elsewhere in its explanatory memorandum, the Commission suggests that ‘the expression “pre-existing relationship” applies particularly to pre-contractual relationships and to void contracts’.²⁵² To a common lawyer, it may seem artificial to describe the exchanges between parties negotiating a contract as a ‘pre-existing relationship’, although the civil law doctrine of culpa in contrahendo‎ rests on the idea that initiation of contractual obligations creates a relationship of obligation belonging or related to the law of contract.²⁵³ A debate on whether this terminology is apt to describe the continuum of communications in contemplation of contract would, in any event, be largely academic as (a) the appearance in the final sentence of Art 4(3) of the words ‘in particular’ emphasizes that a connection between a tort/delict and a ‘pre-existing relationship’ is but one example of circumstances that may justify displacement of the otherwise applicable law, and (b) Art 12 contains a specific rule for ‘non-contractual obligations arising out of dealings prior to the conclusion of a contract’ (culpa in contrahendo‎),²⁵⁴ which includes non-contractual obligations originating in a contemplated contractual relationship that would otherwise lie within Art 4.²⁵⁵ So far as ‘void contracts’ are concerned, it seems appropriate to recognize a ‘preexisting relationship’ as a matter of fact even if the law’s response to the existence of the vitiating factor is to treat the contract as if it had never existed.²⁵⁶ That relationship may establish a connection either to the putative applicable law or to the law of the country in which the contract was centred. Art 10(1)(e) of the Rome Convention and Art 12(1)(e) of the Rome Convention require the consequences of nullity of a contract to be treated (p. 347) as matters of contractual obligation falling outside the scope of the Rome II Regulation.²⁵⁷ Nevertheless, there may still be a close connection between a void contract and a tort/delict committed, for example, in the course of ‘performing’ the supposed obligations.

4.95  As to the application of Art 4(3) in cases involving a pre-existing relationship, the Commission emphasized that ‘the law applicable to the preexisting relationship does not apply automatically and the court enjoys a degree of discretion to decide whether there is a sufficient connection between the non-contractual obligations and the law applicable to the relationship’.²⁵⁸ The use of the word ‘discretion’ is unfortunate, as (taken out of context) it might be thought to suggest that the court, in making its decision, may take into account matters other than the relative connections to the country whose law otherwise applies under Art 4(1) or 4(2) and the country whose law applies to the parties’ relationship (for example, the consequences of applying the rules of those two countries to the case before him²⁵⁹ or the private international law of a non-Member State²⁶⁰). The requirement that the court focus solely on the link between the relationship and the tort/delict in question is, however, reinforced by the closing words of Art 4(3) requiring that the two be ‘closely connected’. This requirement should be applied objectively and not by reference to the parties’ private motives for the performance of a particular act.²⁶¹

4.96  Against this background, this section considers certain types of claim to which Art 4 may apply, and which raise particular problems. These categories are considered mainly from the viewpoint of English law but would, of course, extend to analogous claims under the law of another country.

4.97  Unlike some language versions²⁶² of Art 5(3) of the Brussels Convention, the Rome II Regulation does not specifically refer to ‘quasi-delict’. That category is (p. 348) unknown in the English law of obligations. In Roman law, the category of quasi-delict included a residue of actions dealing, variously, with liability for persons and things, for example liability of occupiers to those on the street below, liability of ship-owners and publicans for damages caused by their business ‘because they employed unreliable persons’, the liability of animal keepers, and the liability of the paterfamilias‎ for delicts by children of the house.²⁶³ These actions appeared to have in common a stricter form of liability for events that could not, in most cases,²⁶⁴ be directly attributed to the defendant’s acts or omissions. In modern law, the term quasi-délit‎ appears in the Civil Codes of France, Belgium, and Luxembourg ‘where it has from the outset been a synonym for liability for negligence’.²⁶⁵ Whatever the common denominator of obligations described under national law as quasi-delictual, there is little doubt that they fall within the scope of Chapter II of the Regulation, and principally within Art 4, at least insofar as they concern the defendant’s responsibility for the adverse consequences of an event having harmful consequences for another person. Accordingly, the Commission’s suggestion in the explanatory memorandum accompanying its Proposal that ‘obligations relating to what in some jurisdictions is termed “quasi-delict” or “quasi-contract”’ fall outside the category of ‘tort or delict’²⁶⁶ should be rejected. As Recital (11) makes clear, the Regulation extends to non-contractual obligations arising out of strict liability. As these obligations cannot comfortably be accommodated within any of the rules in Chapter III, Art 4 appears the natural home for them.

4.98  The question whether and, if so, where the action paulienne‎ under French law, and similar claims recognized in other legal systems, fits within the Regulation has already been addressed in Chapter 3. In summary, there appears no compelling reason why these claims cannot be considered to fall within Art 4, with the place of damage for the purposes of Art 4(1) being the country in which the creditor has lost his opportunity to enforce against a particular asset or the assets generally of the defendant.²⁶⁷

4.99  It will be clear from the preceding discussion,²⁶⁹ as well as the restricted nature of the exclusion of trust obligations in Art 1(2)(e), that the Rome II Regulation may apply to obligations that an English lawyer would recognize as ‘equitable obligations’, i.e. as originating in the jurisprudence of the courts of equity prior to their union with the common law courts.²⁷⁰ Accordingly, for example, claims for breach of trust, breach of fiduciary duty, and breach of confidence²⁷¹ may fall within the scope of the Regu lation, although whether they do so in a particular case will depend on whether (a) the obligation in question is ‘contractual’ or ‘non-contractual’,²⁷² and (b) a non-contractual obligation falls within the scope of any of the exclusions in Art 1(2).²⁷³ Further, in those Member States party to the Hague Trust Convention (and, in particular, the UK and Luxembourg that have chosen to extend the Convention to trusts declared by judicial decisions) the rules of applicable law contained in the Regulation may be displaced by those in the Convention.²⁷⁴

4.100  Of those equitable claims falling within the scope of the Regulation, it seems likely, however curious the juxtaposition may seem to an English lawyer, that many²⁷⁵ will be characterized as arising in tort/delict and subject to the general rule in Art 4. In the main, the essential basis of these claims is that the defendant’s conduct falls short of that expected of him by rules of equitable origin (in terms of honesty, loyalty, diligence, or otherwise) and that he should be restrained from further ‘inequitable’ conduct and/or required to compensate the claimant or to account to him for the consequences of his conduct. This fits the pattern of responsibility for conduct that is the hallmark of a tort/delict under the Rome II Regulation.²⁷⁶

4.101  A good example of an equitable obligation falling within Art 4 is the accessory liability of a person who dishonestly assists a breach of trust or fiduciary duty to the beneficiary of the trust or duty.²⁷⁷ In Grupo Torras SA v Al Sabah, the Court of Appeal, in considering the law applicable to a claim of this kind under the pre-existing English rules, endorsed the view that equitable wrongdoing, such as dishonest assistance, had ‘marked similarities’ to a tort in that it imposed liability to pay damages for fault.²⁷⁸ Subsequently, in Casio Computer Co Ltd v Sayo,²⁷⁹ the English Court of Appeal held that a dishonest assistance claim fell within Art 5(3) of the Brussels Convention as a ‘matter relating to tort, delict or quasi-delict’. The Court rejected the argument that the claim lacked the necessary element of a causal connection between the damage suffered and an event attributable to the defendant.²⁸⁰

4.102  For this type of claim, the country of damage should be considered to be the country in which the asset misappropriated by the defaulting trustee/ fiduciary was situated at the time of the misappropriation facilitated by the defendant’s conduct. That is the country in which relations between the claimant and the trustee/fiduciary are affected by the defendant’s interference.²⁸¹ Insofar as decisions of the English courts concerning Art 5(3) of the Brussels Convention may favour (without accepting as correct) the view that the place of ‘dissipation’ of the assets is the place of damage for these purposes,²⁸² it is respectfully submitted that they should not be followed in the context of the Rome II Regulation. The laundering of assets appears to be an indirect consequence of the defendant’s involvement in their misappropriation in the first place.