Four Seasons Holdings Incorporated v Brownlie

The claimant, Lady Brownlie, is the widow of the distinguished international lawyer Sir Ian Brownlie QC. In January 2010, she and her husband were on holiday in Egypt, staying at the Four Seasons Hotel Cairo at Nile Plaza. Lady Brownlie’s evidence is that on a previous visit to the hotel, she had picked up a leaflet published by the hotel advertising safari tours which it provided. Before leaving England on the subsequent trip, she telephoned the hotel and booked with the concierge an excursion to Fayoum in a hired chauffeur-driven car. The excursion took place on 3 January, and ended in tragedy. The car left the road and crashed. The passengers, in addition to Sir Ian and Lady Brownlie, were his daughter Rebecca, and Rebecca’s two children. Sir Ian and Rebecca were killed. Lady Brownlie and the two children were seriously injured.

Lady Brownlie subsequently began proceedings for (i) damages for personal injury in her own right, (ii) damages under the Law Reform (Miscellaneous Provisions) act 1934 in her capacity as Sir Ian’s executrix, and (iii) damages for bereavement and loss of dependency under the Fatal Accidents Act 1976 in her capacity as her late husband’s widow. The First Defendant, Four Seasons Holdings Inc (“Holdings”), is the holding company of the Four Seasons hotel group. It is incorporated in British Columbia. The Second Defendant, Nova Park SAE (“Nova Park”) is an Egyptian company which was identified by Lady Brownlie’s solicitors as the owner of the hotel building. The claim form has not been served on Nova Park and, apart from the issue of the claim form, no attempt has been made to pursue the claim against them. Nor have they been represented at any stage. The present appeal is concerned only with the position of Holdings, which has applied to set aside the claim form and service thereof out of the jurisdiction so far as it relates to them.

Before permission can be given for the service of originating process out of the jurisdiction, it is necessary for the claimant to establish (i) that the case falls within at least one of the jurisdictional gateways in CPR 6BPD, para 3.1, (ii) that she has a reasonable prospect of success, and (iii) that England and Wales is the proper place in which to bring the claim. The third of these conditions reflects the principle of forum conveniens, and there is no issue about it in this case. It is accepted that England is a proper place in which to bring the present claim if the first two conditions are satisfied. So far as the claim is founded on contract, Lady Brownlie’s application for permission to serve out was based on CPR 6BPD, para 3.1(6)(a) (“the contract … was made within the jurisdiction”). So far as it was founded on tort, it was based on CPR 6BPD, para 3.1(9)(a) (“damage was sustained … within the jurisdiction”). Holdings says, first, that Lady Brownlie has not established that the contract with the hotel was made in England, but that wherever it was made, it was not made with them. Their case is that they are a group holding company whose subsidiaries provide certain central services to hotels of the Four Seasons hotel chain but neither own nor operate them. Gateway (6)(a) does not therefore apply. Secondly, they say that gateway (9)(a) does not apply because the damage which is the basis of the claim in tort was not sustained in England. Thirdly, they say that Lady Brownlie does not satisfy the requirement of CPR 6.37(1)(b) that there should be a “reasonable prospect of success”. It is common ground that any relevant contract for the services of the car and driver was governed by Egyptian law.

Some of the jurisdictional gateways in CPR 6BPD merely require that the claim should be of a particular character. For example it is a claim for an injunction regulating conduct within the jurisdiction. Others, including gateways 6(a) and 9(a) on which Lady Brownlie relies, depend on the court being satisfied of some jurisdictional fact. A relevant contract must, for example, have been made or breached in England or relevant damage sustained there. There are two closely related problems about this. The first is a legal one, namely that none of the law’s established evidential standards satisfactorily meets the case. The second is a practical one, namely that some jurisdictional facts, for example the existence of the contract said to have been made or breached in England, may be in issue at trial if the case is allowed to proceed, when they will in all probability be determined on fuller material than is likely to be available at the interlocutory stage. The same is true of the more general requirement that if it proceeds the claimant should have a reasonable prospect of success.

The leading modern cases are the decisions of the House of Lords in Vitkovice Horni a Hutni Tezirstvo v Korner[1951] AC 869 and Seaconsar Far East Ltd v Bank Markazi Jomhouri Islami Iran[1994] 1 AC 438. Vitkovice was about the evidential standard to be applied to the applicability of the jurisdictional gateways. It concerned what was then RSC order 11, rule 1(e) (“the claim is brought in respect of a breach committed within the jurisdiction of a contract made within or out of the jurisdiction”). The Appellate Committee held that each element of the gateway’s factual requirements had to be established, namely the contract, the breach and its geographical location. However, it rejected the view expressed by Lord Goddard CJ in Malik v Narodni Banka Ceskoslovenska[1946] 2 All ER 663 that the evidential standard for establishing that one of the jurisdictional gateways applied was the civil burden of proof, on the ground that such a test “in effect amounted to a trial of the action or a premature expression of opinion on its merits”: see Lord Simonds, at p 879. It also rejected the suggestion that it was enough to show a prima facie case, because that test depended on the legal adequacy of the factual case advanced by the claimant. The application of such a test would not be consistent with the practice, which allowed a factual challenge to the evidence led by the claimant on the point. Lord Simmonds (p 880), with whom Lord Normand agreed, adopted from Counsel the expression a “good arguable case”, which appeared to him to import more than a prima facie case but less than a balance of probabilities. Lord Radcliffe, with whom Lord Tucker agreed, spoke of a “strong argument” or “a strong case for argument” (pp 883, 884, 885). In Seaconsar Far East Ltd v Bank Markazi Jomhouri Islami Iran[1994] 1 AC 438, Lord Goff, with the agreement of the rest of the Committee, endorsed Lord Simmonds’ formulation as applied to the gateways, and suggested that Lord Radcliffe’s formulation meant the same thing. At the same time, he held that the existence of a reasonable prospect of success fell to be determined according to a lesser standard, namely that there should be a “serious issue to be tried”. This has been held to correspond to the test for resisting an application for summary judgment: Altimo Holdings and Investments Ltd v Kyrgyz Mobil Tel Ltd[2012] 1 WLR 1804, para 71.

Since Lord Goff considered that the evidential standard applicable to jurisdictional facts relevant to the availability of the gateway was derived from RSC order 11, rule 4(2) (“No such leave shall be granted unless it shall be made sufficiently to appear to the court that the case is a proper one for service out of the jurisdiction under this Order”), he must also have thought that the standard was the same whether the jurisdictional fact in question would or would not be in issue at a trial on the merits. I think that that must be right, and equally true of the current rules, although the language of CPR 6.36, which limits the court’s jurisdiction to cases falling within the gateways, is not precisely the same.

An attempt to clarify the practical implications of these principles was made by the Court of Appeal in Canada Trust Co v Stolzenberg(No 2) [1998] 1 WLR 547. Waller LJ, delivering the leading judgment observed at p 555:

The choice of defendants in this case was difficult because of the diffused character of the Four Seasons hotel chain, and the complex and undisclosed contractual arrangements governing the relationship between individual hotels and the Four Seasons group.

On 7 June 2010, Lady Brownlie’s solicitors Kingsley Napley wrote to the legal department of “Four Seasons Hotels and Resorts” in Toronto, outlining the basis of the proposed claim. They asserted that the contract was made with “Four Seasons Hotel Group” without specifying the particular entity to which they were referring. They invited them to accept liability or failing that to identify their defence and disclose certain documents. They received a reply from Ms Marilyn Waugh, Corporate Legal Adviser to “Four Seasons Hotels and Resorts”, saying that their letter had been passed to the Four Seasons Hotel Cairo at Nile Plaza. This was followed by a letter from a firm of lawyers in Cairo, dated 22 August 2010, acting for the Cairo hotel, who denied liability. They asserted that the driver was employed by an independent car hire company, the role of the hotel being simply to relay Lady Brownlie’s request to them. Accordingly, they said, the contract was made with the car hire company and the hotel was under no liability. Kingsley Napley eventually wrote back on 9 May 2011 to Ms Waugh in Toronto and to the Egyptian lawyers in Cairo saying that they did not accept this analysis. Their letter to the Cairo hotel’s lawyers asked for particulars of the relevant corporate entities:

The claim form was issued in December 2012. Holdings was sued on the footing that it was the owner and the manager of the Cairo hotel business and the provider of the driver’s services, or alternatively the agent for an undisclosed principal who provided the driver’s services. In due course, an application was made to Master Yoxall for permission to serve it on Holdings out of the jurisdiction. This was supported by a witness statement in which that company was described as a “corporate entity engaged, among other things, in the ownership and/or operation and/or organisation of a chain of international hotels which includes the Four Seasons Cairo at Nile Plaza Hotel, Cairo, Egypt.” It exhibited draft Particulars of Claim in which that statement was repeated. It was alleged that the contract for the excursion was made with Holdings and that they were vicariously liable for the negligence of the driver of the car. Master Yoxall gave permission for service out. Service was effected on Holdings in Canada and, for good measure, on the Four Seasons Hotel in Park Lane, London.

At this point, Messrs Kennedys came on to the scene, acting for Holdings. They applied to set aside the order of Master Yoxall. Mr Newman of that firm made a witness statement in which he said that hotels of the Four Seasons chains were owned by different owners, who entered into agreements with “a number of Four Seasons entities” covering “licensing, management and advisory issues.” The Cairo hotel was owned by Nova Park. Holdings was a “management company” which did not own either the Cairo or the Park Lane hotel and had no contractual relationship with either of them. Mr Donovan responded by reasserting that the contract was made with Holdings. The basis of this assertion was said to be that internet research suggested that Holdings was the parent company of the Four Seasons group, that it operated a central reservation system and website for the worldwide chain, and that it was the owner and licensor of the trade marks used by the Cairo hotel. Master Cook set aside the order for service out on the ground that in the face of Mr Newman’s evidence these assertions were not enough to support the contention that Lady Brownlie had contracted with Holdings or that Holdings was vicariously liable for the driver of the car.

On appeal from Master Cook, the matter came before Tugendhat J, who allowed the appeal and restored the order for service out. Tugendhat J was clearly irritated by the failure of any Four Seasons company to answer the question put to the lawyers for the Cairo Hotel in Kingsley Napley’s letter of 9 May 2011 about the identity of the corporate entity responsible for taking the booking for Lady Brownlie’s excursion. His irritation may have coloured his assessment of the evidence. His reasons for allowing the appeal were, in summary, that Mr Newman’s evidence that the Hotel was owned by Nova Park was not to be taken at face value, first because it was expressed in the present tense and did not necessarily relate to the position in 2009 and, secondly, because he did not identify the source of his information. He concluded that in the absence of acceptable evidence about who owned the Cairo hotel, the inferences of Kingsley Napley were enough to support the case that Holdings was vicariously liable for the negligence of the driver. He observed that having failed to supply acceptable evidence, “Four Seasons Hotels and Resorts” could have no complaint if the court ignored points that might be made at trial.

The judge’s reasoning on this point was endorsed by the Court of Appeal, but I confess to finding it rather unsatisfactory. On the face of it, even assuming that individual Four Seasons hotels used a central reservation system operated by Holdings and centrally owned trade marks owned by Holdings, that would not identify Holdings as the owner or operator of the hotel or the employer of the concierge who took the booking. It is true that Mr Newman’s evidence was technically defective, but ultimately the party who would lose by discarding it was Lady Brownlie. If Holdings did not own or operate the hotel, this would inevitably become apparent at trial, with the result that her claim would be dismissed after a great deal of additional delay and expense. This would be in the interest of neither party, and certainly not in the interests of justice.

For that reason, this court took the exceptional course of inviting Mr Palmer QC, who appeared for Holdings, to take instructions on the precise distribution of corporate responsibility for the operation of the Cairo hotel and to serve more circumstantial evidence on the point in a form which complied with the rules. The result was a witness statement of Ms Barbara Henderson, Vice President, Corporate Finance of an associated company of Holdings, setting out the position in detail, with supporting exhibits. It is entirely clear from this material that Holdings is a non-trading holding company. It neither owns nor operates the Cairo hotel, which has at all material times been owned by Nova Park, a company with no corporate relationship to any Four Seasons company. A Dutch subsidiary of Holdings called Four Seasons Cairo (Nile Plaza) BV entered into an agreement with Nova Park to operate the hotel on behalf of Nova Park, although at the material times the actual operator was an Egyptian subsidiary of Holdings, FS Cairo (Nile Plaza) LLC, which assumed the contractual obligations of the operator by assignment. Other subsidiaries of Holdings supplied advice and specific services such as sales, marketing, central reservations and procurement, and licensed the use by Nova Park of the Four Seasons trade marks.

It follows that on the information now available, which substantially corresponds to that given more summarily in Mr Newman’s witness statement before Master Cook, there is no realistic prospect that Lady Brownlie will establish that she contracted with Holdings, or that Holdings will be held vicariously liable for the negligence of the driver of the excursion vehicle. Lady Brownlie’s claim does not satisfy the specific factual requirements of the gateways. A fortiori, it does not satisfy the general requirement that there should be a reasonable prospect of success.

Since Holdings was not party to the relevant contract, it is unnecessary to deal with the question where that contract was made, which may in due course have to be determined as against other parties. But I think it right to draw attention to the artificial nature of the issue as the law currently stands. The argument on the point turned on the question who uttered the words which marked the point at which the contract was concluded and where the counterparty was physically located when he or she heard them. This is the test which has for many years been applied where the contract was made by instantaneous exchanges, eg by telephone: see Entores v Miles Far East Corpn [1955] 2 QB 327 (CA). It differs from the test applied to contracts made by post, which are complete when and where the letter of acceptance is posted: Adams v Lindsell (1818) 1 B & Ald 681, Dunlop v Higgins(1848) 1 HLC 381. These rules were adopted for reasons of pragmatic convenience, and provide a perfectly serviceable test for determining whether a contract has been concluded at all. However, their deployment for the purpose of determining when or where a contract was made is not at all satisfactory. It depends on assumptions about the point at which an offer is accepted or deemed to be accepted, which are particularly arbitrary when the mode of communication used is instantaneous (or practically so). It also gives rise to serious practical difficulties. The analysis of an informal conversation in terms of invitation to treat, offer and acceptance will often be impossible without a recording or a total recall of the sequence of exchanges and the exact words used at each stage, in order to establish points which are unlikely to have been of any importance to either party at the time. This may be unavoidable under the current wording of gateway 6(a). But the whole question could profitably be re-examined by the Rules Committee.

In those circumstances, the correct interpretation of the tort gateway in CPR 6BPD, para 3.1(9)(a) does not arise, and anything that may be said on the subject is obiter. If there had been sufficient reason for treating Holdings as the owner and operator of the hotel, I would in any event have held that the order for service out could not stand so far as it concerned the claims in tort.

In the case of the claim for bereavement and loss of dependency under the Fatal Accidents Act 1976, the reason is that that Act operates as part of the proper law of the tort, and has no application to a tort which is not governed by English law: Cox v Ergo Versicherung AG[2014] AC 1379. Since it is accepted that the proper law governing the driver’s negligence was that of Egypt, Lady Brownlie’s claim under the Act of 1976 has no prospect of success. The only sustainable pleaded claims are her claims on behalf of her husband’s estate and in respect of her own injuries. It may well be that other claims, including a claim for bereavement and loss of dependency, will be available to her under Egyptian law, but as matters stand no such claims are advanced.

The more difficult question is whether the English court has jurisdiction in respect of the claims for damages for personal injury caused by negligence. This depends on whether the claim falls within PD6B, para 3.1(9), which currently permits the English court to assume jurisdiction if:

Leaving aside the statutory causes of action under the Fatal Accidents Act 1976, the losses claimed are (i) funeral, memorial, repatriation and probate expenses and reimbursement of the cost of the holiday, claimed on behalf of Sir Ian Brownlie’s estate; (ii) medical expenses occasioned by Lady Brownlie’s injuries; and (iii) non-pecuniary damage for Lady Brownlie’s pain, suffering and loss of amenity. All of these can loosely be described as after-effects of the accident. It may be assumed that they were experienced, at least in part, in England. Accordingly, the question at issue on this appeal is whether, when a tortious act results in personal injury or death, “damage” is limited to the direct damage, ie the physical injury or death, or extends to the indirect damage, ie the pecuniary expenditure or loss resulting. On the latter view, the English court would have jurisdiction. The 2015 amendment extends the test to prospective torts and prospective damage, but will not affect the question what “damage” means in this context.

Rules substantially similar to CPR 6BPD, para 3.1(9)(a) have been interpreted in Canada and New South Wales as extending jurisdiction to the court of the place where the financial consequences of physical damage were experienced: see, as to Canada, Skyrotors Ltd v Carrière Technical Industries(1979) 102 DLR (3d) 323 (Ont) and Vile v Von Wendt(1979) 103 DLR (3d) 356 (Ont CA); and as to New South Wales Challenor v Douglas[1983] 2 NSWLR 405 and Flaherty v Girgis[1984] 1 NSWLR 56, [1985] 4 NSWLR 248. These decisions have been followed in England. In Booth v Phillips[2004] 1 WLR 3292 Nigel Teare QC, sitting as a deputy judge of the Queen’s Bench Division, held that jurisdiction in respect of a fatal accident in Egypt was properly established in England by virtue of the fact that the widow’s loss of dependency under the Fatal Accidents Act 1976 and the cost of the deceased’s funeral had been sustained in England where she lived. This decision was followed by Tugendhat J in Cooley v Ramsey[2008] ILPr 27 and applied to non-fatal injuries sustained in a road accident in Australia but leading to significant care costs in England, where the claimant lived. Both cases were followed by Haddon-Cave J in Wink v Croatia Osiguranje DD[2013] EWHC 1118 (QB) (unreported), by Sir Robert Nelson in Stylianou v Toyoshima[2013] EWHC 2188 (QB) (unreported) and by Stewart J in Pike v Indian Hotels Co Ltd[2013] EWHC 4096 (QB) (unreported). The English cases were all decided at first instance, and they have been questioned on appeal. In Erste Group Bank AG (London Branch) v JSC “VMZ Red October”[2015] EWCA Civ 379, paras 104-105, the Court of Appeal considered that they gave an “extraordinarily wide” effect to the tort gateway and expressed “serious misgivings” as to whether they were right, but did not decide the point. In the present case, the Court of Appeal effectively overruled them. I think, although for somewhat different reasons, that they were right to do so.

The main reason given by Arden LJ, giving the leading judgment in the Court of Appeal, was based on an analogy with article 4 of the Rome II Regulation EC 864/2007 on the Law Applicable to Non-contractual Obligations. Article 4 provides:

There is, however, a more fundamental reason for concluding that in the present context “damage” means direct damage. It concerns the nature of the duty broken in a personal injury action and the character of the damage recoverable for the breach. There is a fundamental difference between the damage done to an interest protected by the law, and facts which are merely evidence of the financial value of that damage. Except in limited and carefully circumscribed cases, the law of tort does not protect pecuniary interests as such. It is in general concerned with non-pecuniary interests, such as bodily integrity, physical property and reputation which are inherently entitled to its protection. Of these, bodily integrity has been described as “the first and most important”: Parkinson v St James and Seacroft University Hospital NHS Trust[2002] QB 266, at para 56 (Hale LJ). Where these interests are deliberately or negligently injured, the tort is complete at the time of the injury, notwithstanding that damage is an essential element of it. This is the basis of the rule that all the damage flowing from bodily injury or damage to property must be claimed in one action, which may be brought as soon as the claimant has been injured or his property damaged. And, although damage is an essential element of the cause of action in tort, the limitation period in respect of any damage flowing from the breach will run from that time. I would readily accept that that “damage” as that word is used in the rule is not necessarily limited to the damage which serves to complete a cause of action in tort. But the two concepts are clearly related, even if they are not coterminous.

These points may be illustrated by reference to tortious damage to physical property, another interest which is inherently protected by the law of tort. The law in this area has been largely made in the context of collisions at sea and, more recently, damage to road vehicles. The measure of damages in a collision action is the resulting diminution in the value of the ship and its earning potential. The damage is sustained as soon as the collision occurs, notwithstanding that at that stage there has been no out of pocket pecuniary loss or expense. The cost of repair is no more than the prima facie measure of the diminution of value of the ship. The injury to the ship’s earning potential arising from the physical damage is represented by the amount of the earnings lost or the cost of hiring a replacement in order to avoid loss of earnings. These points were made by Lord Hobhouse in Dimond v Lovell[2002] 1 AC 384, 406 in the context of collision damage to a car:

At an emotional level, it might be thought to belittle the gravity of bodily injury suffered by a human being to treat it as analogous to physical damage to a mere chattel or the profits derived from it. But the analysis is essentially the same. The law protects the claimant’s bodily integrity from deliberately or negligently inflicted harm. The damage to that interest is suffered as soon as the bodily injury has occurred, even if subsequent events are relevant to determine the pecuniary measure of that damage. Thus, until the position was altered by statute in 1982, a right to damages for loss of expectation of life was held to accrue at the moment of the accident although the victim was killed. It was therefore to be recoverable for the benefit of the estate under the Law Reform (Miscellaneous Provisions) Act 1934: Rose v Ford[1937] AC 826. As Lord Roche put it (at p 857), “the initial bodily injuries [carry] with them from the outset a diminished expectation of life.” The same principle was applied to damages representing the loss of prospective earnings, before that too was changed by statute: Gammell v Wilson[1982] AC 27. In principle, the same must apply to pain, suffering and loss of amenity. These are consequences of greater or lesser duration which (to borrow Lord Roche’s phrase) are carried with the bodily injury sustained at the time of the accident. It follows, as in the case of damage to property, that the damage to the interest protected is sustained in country A where the claimant has been injured or killed. The pecuniary measure of that damage may depend on things that happened elsewhere. For example, medical or care costs may be incurred in country B, or earnings may be lost which would have been earned in country C or paid in country D, but the damage has not been sustained in these places.

Where the interest protected by the relevant legal duty is a purely financial interest, the same distinction will usually fall to be made between the damage sustained by the interest which the law protects and the expenditure which is merely evidence of its amount. Where the relevant duty is not to cause a purely financial loss, the relevant interest may be located and damaged in more than one country, something which is conceptually harder to envisage in the case of bodily injury or injury to property. But the fact that the amount of that damage falls to be measured by expenditure which has been incurred somewhere else is irrelevant. Thus, if I carry on a single business in France and Germany, and an actionable conspiracy damages that business, the interest protected is the business, and it may sustain damage for the purpose of the rule in both countries. But the fact that I am an English company whose balance sheet suffers in England, or that I incur expenditure in England to make the damage to my business good, is of no relevance. This point may be illustrated by the facts of the European cases considered in para 29 below.

It would have been possible for the draftsman of the Practice Direction to provide that “damage” should extend to the financial or physical consequences of the damage, but there is nothing in the language to suggest that he has done so, and two policy considerations which to my mind strongly suggest that nothing like that was intended.

The first is that in different ways all the jurisdictional gateways in the Practice Direction are concerned to identify some substantial and not merely casual or adventitious link between the cause of action and England. This is a purpose which is better served by locating jurisdiction in the place where the relevant interest of the claimant was damaged than by asking where he or she experienced the effects of the damage. To revert to the example of a ship damaged in a collision, the physical damage sustained at the time of the collision has a real connection with the country in which it happened, whereas the connection with the country where it was repaired or would have earned freight is likely to be adventitious. In the context of personal injury, a principle which located damage in the place where the pecuniary consequences of the accident were felt or where any continuing pain, suffering or loss of amenity were experienced would in the great majority of cases confer jurisdiction on the country of the claimant’s residence. It would confer on the English courts what amounts to a universal jurisdiction to entertain claims by English residents for the more serious personal injuries suffered anywhere in the world. Yet that would be far too wide to be consistent with principle. English law has never asserted a jurisdiction for its courts on the basis of the English identity of the claimant, whether by virtue of residence, domicile or nationality. Personal connections between the parties and England are generally relevant to jurisdiction only in the case of the defendant, for example because the claim form can be served on him there or because CPR 6 BPD, para 3.1(1) applies by virtue of the defendant’s English domicile. This is the fundamental reason why I am unable to accept Baroness Hale’s analysis of this issue. It appears to me to produce a test for jurisdiction so wide as to conflict with the purpose of the rule.

The second policy consideration concerns the history of the tort gateway and its relationship with article 5.3 of the Brussels Convention and Regulations. Article 5.3 is one of a number of provisions for special jurisdiction. It authorises proceedings in tort “in the courts for the place where the harmful event occurred or may occur”, notwithstanding the general rule that suit must be brought in the jurisdiction of the defendant’s domicile. The place where the “harmful event” occurred was interpreted by the Court of Justice in Handelskwerij G J Bier v Mines de Potasse d’Alsace SA (Case C-21/76) [1978] 1 QB 708 as referring at the plaintiff’s option either to the place where the damage was sustained or to the place (if different) where the act was done that gave rise to it. The issue in Bier arose out of the wrongful emission of pollutants into the Rhine in France which damaged the plaintiff’s seed-beds in the Netherlands. Since the physical damage and its financial consequences were all suffered in the Netherlands, it was unnecessary for the Court to consider what losses or expense were encompassed by the word “damage”. That question did, however, arise in Netherlands v Ruffer (Case C-814/79) [1995] ECR I-3807, where the Court of Justice adopted precisely the same distinction as I have done between the damage sustained by the interest which the law protects, and the expenditure which serves as the measure of that damage. The facts were that a barge had sunk, allegedly by the carelessness of its German-domiciled owner, in waters that were deemed for the relevant purpose to be part of Germany. The Dutch state sought to claim in its own courts the cost of raising and disposing of the wreck. It argued that the “harmful event” had occurred in the Netherlands because that was where it had incurred the cost of disposal and suffered the financial losses associated with it. Advocate General Warner, at p 3836, rejected that contention because (i) the cost of disposal merely quantified a loss consisting in the blockage of the waterway; and (ii) acceptance of the argument “would be tantamount to holding that under the Convention a plaintiff in tort had the option of suing in the courts of his own domicile, which would be quite inconsistent with the scheme of article 2 et seq of the Convention.” The Court was able to deal with the matter without reference to this point. In Société Commerciale de Réassurance v Eras International Ltd (The Eras Eil Actions) [1992] 1 Lloyd’s Rep 570, 591, however, Mustill LJ, delivering the judgment of the Court of Appeal, treated the Advocate General’s analysis as “unanswerable” and equally applicable to the tort gateway under the Rules of the Supreme Court. In Dumez France SA v Hessische Landesbank (Case C-220/88) [1990] ECR I-49, the Court of Justice adopted the same analysis. The plaintiffs had sought to recover in France the loss which they claimed to have sustained there as a result of the insolvency of their German subsidiaries, said to have been caused by the defendants’ wrongful acts in Germany. The Court of Justice held that the damage had been sustained in Germany. The harm alleged to have occurred in France was “merely the indirect consequence of the financial losses initially suffered by their subsidiaries” (para 13). The Court expanded and clarified this statement in Marinari v Lloyds Bank Plc (Case C-364/93) [1996] QB 217. Mr Marinari sought to sue the defendant bank in Italy for the act of staff at its Manchester branch in impounding certain promissory notes which he had deposited with them, asserting that he had suffered the financial consequences in Italy, where he was domiciled. The Court rejected this contention, holding that “damage” in article 5.3

It is fair to say that the construction of the Brussels Convention and Regulations depends in part on policy considerations which are irrelevant in the context of the English rules governing jurisdiction over non-EU defendants. Both the Convention and the English rules recognise the possibility that there may be more than one eligible jurisdiction for a given dispute. But the Brussels Convention and Regulations are a code for allocating jurisdiction between EU member states. Acceptance of jurisdiction allocated in accordance with them is mandatory, and not merely permissive as it is under the English rules. Nonetheless, I consider that the principle adopted by the Court of Justice should be followed for two reasons. The first is that they embody an analysis of what constitutes “damage” which, like Mustill LJ in the Eras Eil Actions, I regard as unanswerable. It is not so much a proposition of law as the application of basic logic to the facts. The second reason is that in its current form, the jurisdictional gateway in the English rules for claims in tort was deliberately drafted so as to assimilate the tests for asserting jurisdiction over persons domiciled in an EU member state and persons domiciled elsewhere. Before 1987, service out of the jurisdiction was permitted by RSC order 11, rule 1(1)(h) where the action was “founded on a tort committed within the jurisdiction”. This expression was interpreted as referring to the place where in substance the wrongful act was done: Distillers Co (Biochemicals) Ltd v Thompson[1971] AC 458. The location of the damage (if different) was irrelevant. Under the pre-1987 rules, the English court would plainly not have had jurisdiction to hear Lady Brownlie’s claims in tort. Article 5.3 of the Brussels Convention, as interpreted by the Court of Justice, was broader. In relation to actions against persons domiciled in the EU, it conferred special jurisdiction on the courts of the place where the damage was sustained as well as the place where the wrongful act was done. Effect was given to the Brussels Convention in England by the Civil Jurisdiction and Judgments Act 1982 and by amendments to the Supreme Court Rules (SI 1983/1181) which were expressed to take effect when the Act came into force (in the event, 1 January 1987). The changes effected by these instruments might have been limited to the cases covered by the Convention, which did not extend to actions brought against persons domiciled outside the EU. In fact, they were not. The new rules of court expanded the tort gateway in RSC order 11 so as to correspond with article 5.3 of the Convention as interpreted in Bier. The new RSC order 11, rule 1(1)(f) provided that jurisdiction could be exercised in a non-Convention case where “the claim is founded on a tort and the damage was sustained or resulted from an act committed within the jurisdiction”. Although the language changed when the gateways were transferred to a Practice Direction in 2000, the substance of the rule remained the same, except for the omission of the definite article before “damage”. That omission appears fairly clearly to have been intended to exclude the suggestion that all the damage had to be sustained within the jurisdiction, thus allowing for the possibility that jurisdiction might be founded on the occurrence of some of the damage in England. At any rate, the result is that RSC order 11, rule 1(1)(f) and the corresponding provisions of CPR 6BPD, para 3(9)(a) have generally been construed in the light of the case law of the Court of Justice: see Metall und Rohstoff AG v Donaldson, Lufkin and Jenrette Inc [1990] 1 QB 391, 424 (CA); Société Commerciale de Réassurance v Eras International Ltd (The Eras Eil Actions) [1992] 1 Lloyd’s Rep 570, 589 (Mustill LJ); Bastone & Firminger Ltd v Nasima Enterprises (Nigeria) Ltd[1996] CLC 1902 (Rix J); ABCI v Banque Franco-Tunisienne[2003] 2 Lloyd’s Rep 146, at paras 41, 43 (Mance LJ). It would be strange if the effect of expanding the gateway to match the wider special jurisdiction authorised in Convention cases had been to make it very much wider than even the Convention authorised.

Tugendhat J in Cooley v Ramsey and Haddon-Cave J in Wink v Croatia Osiguranje rejected the two policy considerations which I have described because they considered that the risk that the gateway would be too wide could be managed through the court’s overriding discretion jurisdiction as to forum conveniens. The scheme of the Brussels Convention and Regulations is different, it is said, because its mandatory character excludes discretion: see Owusu v Jackson (Case C-281/02) [2005] QB 801. That view of the matter derives energetic support from Professor Briggs in his book Civil Jurisdiction and Judgments, 6th ed (2015), para 4.73, and in various articles. Indeed, Professor Briggs has gone further, proposing that in the light of my own comments in Abela v Baaderani[2013] 1 WLR 2043, the time has come to “downgrade” and eventually abolish the jurisdictional gateways and make forum conveniens (and presumably reasonable prospect of success) the sole criteria for service out: see “Service out in a shrinking world” [2013] LMCLQ 415. In my opinion, this approach is contrary to principle, and is not warranted by anything that was said in Abela v Baaderani. The jurisdictional gateways and the discretion as to forum conveniens serve completely different purposes. The gateways identify relevant connections with England, which define the maximum extent of the jurisdiction which the English court is permitted to exercise. Their ambit is a question of law. The discretion as to forum conveniens authorises the court to decline a jurisdiction which it possesses as a matter of law, because the dispute, although sufficiently connected with England to permit the exercise of jurisdiction, could be more appropriately resolved elsewhere. The main determining factor in the exercise of the discretion on forum conveniens grounds is not the relationship between the cause of action and England but the practicalities of litigation. The purpose of the discretion is to limit the exercise of the court’s jurisdiction, not to enlarge it and certainly not to displace the criteria in the gateways. English law has never in the past and does not now accept jurisdiction simply on the basis that the English courts are a convenient or appropriate forum if the subject-matter has no relevant jurisdictional connection with England. In Abela v Baaderani, I protested against the importation of an artificial presumption against service out as being inherently “exorbitant”, into what ought to be a neutral question of construction or discretion. I had not proposed to substitute an alternative, and equally objectionable, presumption in favour of the widest possible interpretation of the gateways simply because jurisdiction thus conferred by law could be declined as a matter of discretion.

I would allow the present appeal and declare that Holdings not being the owner or operator of the Four Seasons Hotel at Nile Plaza Cairo, or vicariously liable for the driver of the car, the court has no jurisdiction to try any of the claims presently made in this action. In those circumstances it is unnecessary to make any order on the cross-appeal which Lady Brownlie has brought against the Court of Appeal’s order that there was no jurisdiction to entertain her claims in tort for personal injury to herself and on behalf of Sir Ian’s estate. The parties should make submissions in writing on the form of order and on costs within 21 days. I would remit all other consequential matters to the High Court, so as to enable Lady Brownlie to make such applications as she may be advised to join additional parties, amend the draft Particulars of Claim or seek other relief. I express no opinion, even tentative, about the prospects of any such applications.

As we agree that this action cannot continue against the current defendant, everything which we say about jurisdiction is obiter dicta and should be treated with appropriate caution. For what it is worth, I agree (1) that the correct test is “a good arguable case” and glosses should be avoided; I do not read Lord Sumption’s explication in para 7 as glossing the test; and (2) that the action in tort is governed by Egyptian law and so the Fatal Accidents Act 1976 cannot apply to it, although Egyptian law may in fact allow for a similar claim, should permission ever be given to plead it.

Also for what it is worth, (3) this is not the place to cast doubt upon the longstanding rule in Entores Ltd v Miles Far East Corpn [1955] 2 QB 327, nor could the Rules Committee change that rule by changing the rules relating to jurisdiction in contractual claims; but it could consider avoiding the factual problem which has arisen in this case by adopting a broader formulation of the rule in CPR 6BPD, para 3.1(6)(a); the inclusion of contracts made “by or through an agent trading or residing within the jurisdiction” in para 3.1(6)(b) suggests that this would not be wrong in principle.

Above all, however, (4) I wish to sound a note of special caution as to the correct interpretation of the gateway for claims in tort, contained in para 3.1(9) of the Practice Direction, which currently reads (the words in square brackets having recently been added):

Although this has been done many times before, it may be helpful to trace the genesis of this rule. Before the advent of the Civil Procedure Rules 1998, service out of the jurisdiction was governed by the Rules of the Supreme Court, order 11, rule 1(1)(f). Before the Civil Jurisdiction and Judgments Act 1982 came into force, this referred only to cases “founded on a tort committed within the jurisdiction”. It was amended, with effect from the date when that Act came into force, to read “The claim is founded on a tort and the damage was sustained, or resulted from an act committed, within the jurisdiction”. This broadened the gateway, because the Brussels Convention on Jurisdiction and Enforcement of Judgments in Civil and Commercial Matters 1968, to which the United Kingdom acceded in 1978, and which was incorporated into United Kingdom law by the 1982 Act, provided a special jurisdictional rule, in article 5.3, that a person domiciled in a contracting state could be sued in another contracting state in matters relating to tort, delict or quasi-delict “in the courts for the place where the harmful event occurred”; in Bier v Mines de Potasse d’Alsace (Case C-21/76) [1978] 1 QB 702, the European Court of Justice had interpreted this phrase to refer both to the place where the damage occurred and the place of the event giving rise to it, so that the claimant could choose between them; it appears that the words “harmful event” were deliberately chosen because it was not considered appropriate for the Convention to be specific between the two. In that case, damage had been done to horticultural nurseries in the Netherlands by the discharge into the Rhine of saline waste from operations in France.

In Bier, the damage was all sustained in one place. In Dumez France SA v Hessische Landesbank (Case C-220/88) [1990] ECR I-49, the French Companies were complaining of financial losses suffered because of the insolvency of their German subsidiaries, brought about by the suspension of construction projects in Germany allegedly because the defendant German banks had cancelled loans to finance the projects. The European Court of Justice pointed out that article 5.3 was an exception to the general rule that defendants were to be sued in their country of domicile. The aim of the Convention was to avoid a multiplicity of jurisdictions, with the heightened risk of irreconcilable decisions creating problems for the mutual recognition and enforcement of judgments. So article 5.3 did not permit a claimant claiming for damage, which was the consequence of harm suffered by other persons who were the direct victims of the harmful act, to bring proceedings in the place where the claimant sustained the damage.

In Marinari v Lloyds Bankplc (Case C-364/94) [1995] ECR I-2715, the Grand Chamber affirmed both Bier and Dumez and took the latter a stage further. The claimant brought proceedings in Italy alleging financial loss and damage to his reputation caused when the defendant bank reported him to the police in England because promissory notes he had lodged with them appeared suspicious; this led to his arrest and the confiscation of the promissory notes. The court held that article 5.3 did not cover every place where adverse consequences of an event which had already caused actual damage elsewhere could be felt. It did not refer to the place where the victim claimed to have suffered financial loss consequential on actual damage arising and suffered by him in another member state.

The Brussels Convention was replaced by Council Regulation (EC) No 44/2001 (the Brussels I Regulation) which was in turn replaced by Regulation (EU) No 1215/2012 (the recast Brussels I Regulation). Article 7.2 repeats the wording of article 5.3 of the Convention. Marinari is still the authoritative interpretation of “where the harmful event occurred” in European law. It goes without saying, however, that we are not here concerned with a claim which is governed by the jurisdictional rules of European law. We are dealing with a claim against a defendant who is not domiciled in a member state, which is therefore governed by the jurisdictional rules of the law of England and Wales, now contained in the Civil Procedure Rules 1998 (CPR).

Under the CPR, the equivalent rule to RSC order 11(1)(f) was contained in CPR 6.20(8): “a claim is made in tort, where (a) damage was sustained within the jurisdiction; or (b) the damage sustained resulted from an act committed within the jurisdiction”. The definite article was omitted from (a), in line with the holding of the Court of Appeal in Metall und Rohstoff AG v Donaldson Inc [1990] 1 QB 391, at p 437, that (a) did not require all the damage to be sustained in England; it was enough if “some significant damage” had been sustained here; and similarly that (b) did not require that all the acts constituting the tort be committed in England; it was enough if the tort was in substance committed here. Neither the Rules of the Supreme Court nor the Civil Procedure Rules required that permission be given to serve out of the jurisdiction if the relevant gateway applied; there was always a discretion not to do so, exercised in accordance with the principles laid down in SpiliadaMaritime Corpn v Cansulex Ltd [1987] AC 460. In the CPR, this was reflected in rule 6.21(2A): “the court will not give permission unless satisfied that England and Wales is the proper place in which to bring the claim”.

There is a consistent line of first instance decisions holding that, in a case which is not governed by the European jurisdictional rules, a claim in tort may be brought in England if damage is suffered here as a result of personal injuries inflicted abroad. The first is Booth v Phillips [2004] EWCA 1437 (Comm), a decision of Nigel Teare QC (as he then was). This was a widow’s claim in negligence for her own loss of dependency and the funeral expenses of the estate of her deceased husband who had died while working as chief engineer on a vessel in Egypt. The judge rejected the argument that “damage” referred to the damage which completed the cause of action. This was not what the rule said. The words used should be given their “ordinary and natural meaning, namely, harm which has been sustained by the claimant, whether physical or economic” (para 35). Dropping the definite article reflected the decision in Metall that it was enough that some significant damage had been sustained here. He also rejected the argument that this was “improbably wide”, because the court had also to be satisfied that it was appropriate, in Spiliada terms, to exercise jurisdiction.

It does not appear that any argument based upon the Brussels Regulation was advanced in Booth, but it was advanced most vigorously, as it happens by Mr Howard Palmer QC, before Tugendhat J in the next case, Cooley v Ramsay [2008] EWHC 129 (QB). He accepted that RSC order 11, rule 1(1)(f) had been changed to give effect to the 1982 Act, but Parliament had not fully assimilated the rules relating to non-party states with those relating to the European member states. It had left in the significant difference that there was no discretion in the Convention and the Regulation, but there was such a discretion under the CPR. The object of the Convention and Regulation was to provide a clear and certain attribution of jurisdiction, but the CPR were more flexible. Hence a claimant who was severely disabled, with continuing needs for care, support and medical attention in this country as a result of a road accident in New South Wales, could bring his claim here.

By the time of the next case, the CPR had been amended. CPR rule 6.36 now refers to the various jurisdictional gateways set out in Practice Direction 6BPDCPR (no doubt to increase flexibility), but rule 6.37(3) repeats the rule that the court will not give permission unless satisfied that England and Wales is the proper place in which to bring the claim. In Harty v Sabre International Security Ltd [2011] EWHC 852 (QB), the claimant was severely injured in a road accident in Iraq while working as a security consultant with the defendant. The defendant did not challenge the gateway, no doubt anticipating that MacDuff J would follow Booth and Cooley, and so the argument focussed on the discretion.

In the next case, Wink v Croatia Osiguranje DD [2013] EWHC 1118 (QB), where the claimant had been seriously injured in a road accident while on holiday in Croatia, a spirited attack upon the correctness of Booth and Cooley was mounted before Haddon Cave J, arguing that 6BPD should be interpreted consistently with European law, so that in a claim where both direct and indirect damage is alleged it is only the place where the direct damage is sustained which is relevant. The judge pointed to a number of obvious problems with this argument: there are no such limiting words in 6BPD, para 3.1(9)(a); the natural and ordinary meaning of “damage” is any damage; the defendant’s argument was tantamount to saying that damage was sustained only where the injury occurs, which is plainly not so in many cases; it was this construction rather than that in Booth which required re-writing (paras 33-35). Agreeing with the “comprehensive” analysis in Cooley, he held that the two schemes - in the Regulation and the Rules - were “fundamentally different in structure and policy” (para 41).

In Stylianou v Toyoshima [2013] EWHC 2188 (QB), the claimant was very severely injured in a road accident in Western Australia and repatriated six weeks later. This time, the defendants argued that Booth and Cooley were incorrect, because they were decided before Regulation (EC) 864/2007 of the European Parliament and Council on the law applicable to non-contractual obligations (the Rome II Regulation) came into force. Article 4.1 provides that the applicable law shall be the law of the country “in which the damage occurs, irrespective of the country in which the event giving rise to the damage occurred and irrespective of the country or countries in which the indirect consequences of that event occurred”. Sir Robert Nelson rejected the argument that the CPR should be interpreted in the same way. He pointed out that article 2.1 of the Rome II Regulation provides that “damage shall cover any consequence arising out of the tort/delict …”, so that article 4.1 was expressly excluding what would otherwise be included in the word “damage”. There was no reason to interpret “damage” in the CPR as in the specific article 4.1 rather than in the general article 2.1. In any event, Rome II was not about jurisdiction and did not override the CPR. The court’s discretion was a “valuable safety valve rendering unnecessary a narrow definition of damage” (para 53).

Trying another tack, the defendant in Erste Group Bank AG (London Branch) v JSC “VMT Red October” [2003] EWHC 2926 (Comm), argued that the judges in Cooley and Wink had failed to appreciate that the Rules Committee was intending to mirror the Brussels Convention as interpreted by Professor Jenard in his preparatory report. Flaux J rejected that argument as “hopeless”: the consistency argument had been rejected because the English rules were wider and that would not have been affected by anything that Professor Jenard had said (para 147). This was an action, inter alia, for the torts of conspiracy and interference with contract allegedly resulting in the failure of a Russian company to meet its obligations under a loan agreement. The case was taken to the Court of Appeal: [2015] EWCA Civ 379. Were it not for the string of first instance decisions to the contrary, the Court would have regarded as “very attractive” the submission that the tort gateway was intended to reflect the European jurisprudence (para 103) and expressed “very serious reservations” as to whether those decisions were right. But they preferred not to decide the point as they did not need to do so, having found that the damage was all sustained in New York.

Finally, in Pike v Indian Hotels [2013] EWHC 4096 (QB), where the claimants had been injured trying to escape from the Taj Mahal Palace in Mumbai during the terrorist attack, Stewart J agreed with Sir Robert Nelson’s “comprehensive demolition” of arguments based on European Union law and held that outside the European context the previous decisions were correct.

It is clear from reading these first instance decisions that each of these judges was not slavishly following the decisions which preceded his own. They carefully considered and rejected the ever more sophisticated arguments against them but clearly considered that they were correct. So do I.

In the first place, I entirely agree with Lord Sumption that the argument based on the Rome II Regulation, accepted by the Court of Appeal in this case, should be rejected. Applicable law and jurisdiction are two different matters. There is no necessary coincidence between the country with jurisdiction and the country whose law is applicable. It is accepted that in this case Egyptian law is applicable to the tort claims. Furthermore, there can only be one applicable law, whereas even in European law there can often be more than one country with jurisdiction.

Indeed, I see no reason to think that those who framed the RSC and CPR intended them precisely to mirror the interpretation later given to the Brussels Convention. The language used in the Rules, although no doubt intended to widen the gateway so as to encompass the cases covered by the Brussels Convention, is quite different from the language of the Convention. The Dumez and Marinari decisions came afterwards, to restrict the scope of the language used in the Convention, but they do not override the language of the Rules in non-EU cases. They are of no help in construing Rules which have remained in essentially the same language ever since. If the Rules Committee had wanted to assimilate the Rules after the decisions in Booth and Cooley, they could easily have done so, and now more easily, as the gateways are contained in a Practice Direction rather than a Rule.

It is also necessary to bear in mind the difference between the two schemes. The European scheme deliberately eschews any discretion in favour of clear and certain rules, in the context of a scheme which governs, not only jurisdiction, but also recognition and enforcement of the resulting judgments. No doubt that is why the Court of Justice was anxious to restrict the scope of the Bier decision by drawing the direct/indirect distinction. That is not a feature of the English scheme, which retains the “valuable safety valve” of discretion, a discretion which need not be limited to the Spiliada principles, but can concentrate on the real question, which is “the proper place for the resolution of the dispute” (as Professor Briggs puts it).

I also have great difficulty with the approach to “damage” adopted by Lord Sumption. He appears to equate “damage” in the Rule with the damage which completes the cause of action. It is true that damage is an essential component of some torts, such as the tort of negligence and the economic torts, so that it is necessary to know if and when the cause of action is complete for purposes such as limitation. But damage is not an essential part of every cause of action in tort. There are many torts which are actionable per se, without proof of damage: trespass to the person, trespass to goods, libel and some slanders are the obvious examples. There is no particular reason to think that completion of the cause of action is what the framers of the Rules had in mind when they used the word “damage”. They are more likely to have had the ordinary and natural meaning of the word in mind. I would be very reluctant to disagree with the several first instance judges who held that this refers to the actionable harm caused by the wrongful act alleged. In this they have the support of a very distinguished Court of Appeal in New South Wales, in Flaherty v Girgis (1985) 63 ALR 466, where Kirby P and Samuels JA agreed with McHugh JA, construing a similar jurisdictional rule, that “damage, therefore, is to be contrasted with the element necessary to complete the cause of action; it includes all the detriment, physical, financial and social which the plaintiff suffers as a result of the tortious conduct of the defendant” (p 482).

Furthermore, it is quite clear that damage can be suffered by the same person in more than one place, just as the wrongful acts can be committed in more than one place. The Court of Appeal in Metall must have been right to say that the Rules contemplated the possibility of there being jurisdiction in more than one place. Nor do I find the distinction between direct and indirect damage easy to draw in all cases. If I am seriously injured in a road accident, the pain, suffering and loss of amenity that I suffer are all part of the same injury and in cases of permanent disability will be with me wherever I am.

I do, of course, take the point that the claimant should not be in the position of choosing where to bring the claim. But in my view the discretion should be robust enough to prevent that. It is looking for a substantial reason to allow a claim against a foreign defendant to be brought in the courts of this country and the courts have always treated such cases with caution. And it is important to bear in mind that, in a tort claim, the applicable law will be the law of the country where the events took place.

I was for a while attracted by a middle course, which would restrict “damage” to the continuing bodily (physical or psychological) effects of the wrongful act, because these are part and parcel of the initial injury, but excluding consequential financial losses. But it is difficult to find a warrant for that in the language used and in some torts the damage is wholly financial, so that separating out the direct and the consequential would be even more difficult. In the end, therefore, I would adopt the ordinary and natural meaning of the language used in the Rules.

I agree with the judgment of Lady Hale and therefore with those parts of the judgment of Lord Sumption with which she agrees.

It may, however, be appropriate for it to be no part of the actual decision of this court today that, as a majority of us considers, the claimant’s claims for personal injury both to herself and, as his executrix, to her late husband (“the two tort claims”) fall within para 3.1(9)(a) of Practice Direction 6B in the Civil Procedure Rules (“the CPR”). For, had it been part of the decision, it would have been far-reaching; and the need for the court at the hearing of this appeal to address other issues, in particular, in an exercise uncharacteristic of it, at last to extricate the facts which have established the impossibility of any recovery against the particular company within the Four Seasons group which is presently sued, may have led to less full argument about the meaning of para 3.1(9)(a) than its importance requires.

In para 22 above Lord Sumption refers to Regulation EC 864/2007 (“the Rome II Regulation”). It requires a member state which determines a claim in tort to apply the law there identified even when such is not the law of another member state. Were these two tort claims to proceed in our courts, it would require them to be determined by reference to Egyptian law. The law of a foreign state is more easily applied in the courts of that state; and in what I will call “the appropriate forum inquiry”, namely into whether our courts are clearly the appropriate forum for the trial of an action, also described in rule 6.37(3) of the CPR as “the proper place in which to bring the claim”, any requirement for it to apply foreign law will always be a negative factor and sometimes a powerful one: see the Spiliada case, cited at para 40 above at pp 478B and 48IH. But the Rome II Regulation is irrelevant to the existence of the jurisdiction of the courts of the member states; and I agree that the Court of Appeal was, with respect, wrong to hold otherwise.

What, by contrast, can be relevant to the existence of the jurisdiction of the courts of member states is Regulation (EU) No 1215/2012 (“the recast regulation”). It recast Council Regulation (EC) No 44/2001, which in turn had replaced the Brussels Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters 1968 (“the 1968 convention”). Importantly, however, by article 4 the recast regulation governs the jurisdiction of those courts only when the defendant is domiciled in the EU. In relation to a case such as the present, in which the defendant is not so domiciled, article 6 of the regulation recognises that the jurisdiction of the courts of a member state is governed by its own law. Limited, as it therefore is, to the allocation of jurisdiction between the courts of member states in relation to claims against persons domiciled in the EU, the regulation is understandably prescriptive. A member state can, for example, rely on the fact that each of its fellow-members is obliged by article 6 of the European Convention on Human Rights, and in relation to the assertion of rights under EU law by article 47 of the EU Charter of Fundamental Rights, to afford to litigants a fair hearing of their claims within a reasonable time; and there is, in the words of recital 26 to the regulation, “mutual trust in the administration of justice in the Union”. It permits no collateral attack upon its allocation of jurisdiction to the courts of one member state by reference to any conclusion that in all the circumstances it would be more appropriate for the case to proceed in the courts of another member state. Articles 4 and 5 are specific: a person domiciled in a member state may be sued in the courts of another member state “only” by virtue of the rules which then follow and, subject to them, he must be sued in the courts of the state in which he is domiciled.

It has therefore been necessary for our procedural rules in respect of service of claims outside England (and, which will go without saying, also Wales) to be wide enough to permit service in circumstances in which the recast regulation and its predecessors have allocated jurisdiction to English courts to determine a claim against a person domiciled elsewhere in the EU. In 1978 the Court of Justice in Luxembourg determined the Bier case, cited and explained in para 29 above, which disclosed a rare situation in which an allegedly unlawful physical act in one member state caused direct physical damage only in a second member state. The court’s construction of the location of the “harmful event” in what was then article 5(3) of the 1968 convention, namely that it had occurred in the second state as well as the first and that it was for the claimant to choose in which of them to bring his claim, therefore required an amendment, which came into force in 1987, to what was then rule 1(1)(f) of Order 11 of the Rules of the Supreme Court. The rule then began to provide for service out of the jurisdiction if, among other things, “the damage was sustained … within” England as well as if it “resulted from an act committed” here.

Our procedural rules for such service have therefore needed to be wide enough to enable us to comply with our duties under EU law. But it does not follow that, even if the natural construction of our rules indicates a wider gateway to service out of the jurisdiction in the case of a claim unconstrained by EU rules of jurisdiction, construction of them should be narrowed to the size of the gateway set by the EU rules, as interpreted by the Court of Justice.

In the Metall und Rohstoff case, cited at para 30 above, the 1968 convention did not apply to the issue of the court’s jurisdiction. The defendants were domiciled in the state of New York, and argued that, were any action to be brought against them, it should be brought there. But the Court of Appeal held that the English court had jurisdiction to determine one group of the various tort claims made against them. The court considered whether, for the purpose of rule 1(1)(f), the alleged torts within the group “resulted from [acts] committed” in England or alternatively in New York. It was enough, so the court held at p 449D, that “as a matter of substance” the acts were committed here. But the court also considered whether “the damage was sustained” in England or alternatively in Switzerland or Belgium. It observed at p 437C-D:

The passage of the court’s judgment in the Metall case set out above leads (and entitles) Lord Sumption at para 30 above to cite the case as exemplifying construction of rule 1(1)(f) and its successors in the light of the case law of the Court of Justice. But it is, I suggest, of greater significance that, as Lord Sumption explains in para 29 above by reference in particular to the judgment of the Grand Chamber in the Marinari case, the Court of Justice has rejected any suggestion that the requisite “harmful event” has occurred in a member state in circumstances in which only a significant part of the damage has been sustained there. If, unlike in the Bier case, damage is sustained in the state in which the causal act took place, the recast regulation does not confer jurisdiction upon the courts of a second state even if significant further damage is sustained there: see paras 14 and 15 of that judgment. Where, by contrast, the jurisdiction of the English court is not governed by EU law, the decision in the Metall case demonstrates that our rules create a gateway wider, as is now clear, than EU law would permit.

I, for my part, would not interpret the word “damage” in para 3.1(9)(a) of Practice Direction 6B by reference to “the damage” which violates the interest protected by the law and which completes a cause of action in tort. The absence of the definite article demonstrates the contrary; and, in that it therefore has to be accepted (as Lord Sumption accepts in para 23 above) that “damage” for the purposes of the paragraph can be wider than the damage which violates the interest and which completes the cause of action, I find the relevance of the latter concept, whatever its importance in the substantive law of tort, to be elusive.

It would, to put it at its lowest, be legitimate to interpret the word “damage” as extending to the secondary damage which the claimant and her husband’s estate sustained in England and which flowed from the primary damage sustained in Egypt. Rule 1.2(b) of the CPR obliges a court which interprets another of its rules to seek to give effect to the overriding objective of enabling it to deal with cases justly. So, if an otherwise legitimate interpretation better serves the ends of justice, it ought in principle to be adopted. Take the case of Pike v The Indian Hotels Co Ltd cited in para 21 above. Mr Pike, an English tourist, was staying at the Taj Mahal Palace in Mumbai on the night of the terrorist attack in 2008. He suffered spinal injuries which rendered him paraplegic. Following his return home, he aspired to sue the operator of the hotel in England. Stewart J found at para 58 that, were Mr Pike to sue the operator in the courts of India, the case would not be concluded for 15 to 20 years; and he held at para 71 that it would therefore be a denial of justice to prevent him from suing the operator in England. The judge held that the word “damage” extended to the secondary damage sustained by him in England, which was also “the proper place to bring the claim”. The facts of Mr Pike’s case were no doubt extreme but they illumine the injustice to which any narrow interpretation of the word “damage” can give rise.

Is it possible that proponents of the narrow interpretation fail to invest due confidence in the appropriate forum inquiry? Is not that inquiry sufficiently muscular to exclude claims founded only on a tenuous amount of damage sustained in England? Lord Sumption contends in para 31 above that the main factor which determines such an inquiry is the practicality of litigation. But in the Spiliada case, cited at para 40 above, Lord Goff of Chieveley held

at p 474F-G that the question was not one of mere practical convenience;

at p 480B-C that the court had to take into account the nature of the dispute as well as the legal and practical issues which it raised; and

at p 480G that the fundamental requirement was to identify the forum in which the case might suitably be tried in the interests of all the parties and of the ends of justice.

The relevance of the jurisdiction of the courts of Ontario and New South Wales to entertain a claim in tort on the basis only of secondary damage sustained there is necessarily limited. But the long-standing existence of the jurisdiction there should allay fears that a broader interpretation of para 3.1(9)(a) would encourage abuse. A claim which requires service of the form out of the jurisdiction will not lightly be brought, not least because of the likely complexity of attempts to enforce any judgment ultimately obtained; and a rigorous exercise of the appropriate forum inquiry should in my view yield the proportionate outcomes which all of us, on both sides of what in the present case reduces only to a discussion, no doubt intend that our law should achieve.

In so far as there are issues between Lady Hale and Lord Wilson on the one hand and Lord Sumption and Lord Hughes on the other, I prefer the reasoning of Lady Hale and Lord Wilson for the reasons they give. In particular, like Lady Hale and Lord Wilson, I prefer the reasoning in the various decisions of first instance judges to which they refer. In particular I agree with Lord Wilson in his para 64 that, in the absence of the definite article in para 3.1(9)(a) of Practice Direction 6B, it has to be accepted that “damage” for the purpose of the paragraph can be wider than the damage which violates the claimant’s interest and which completes the cause of action.

Further, I agree with Lady Hale’s analysis of the various first instance decisions to which she refers. I would endorse the last three sentences of her para 52 as follows: