Tort Choice of Law Rules in Cross-border Multi-party Litigation under European and Chinese Private International Law
Tort Choice of Law Rules in Cross-border Multi-party Litigation under European and Chinese Private International Law Read more
Tort Choice of Law Rules in Cross-border Multi-party Litigation under European and Chinese Private International Law Read more
In Agostinho da Silva Martins v Dekra Claims Services Portugal SA (C-149/18), between Mr Agostinho da Silva Martins, who suffered damages in a car accident, and the insurance company Dekra Claims Services Portugal SA, the CJEU was called to rule on two different issues of qualification: one related to the interpretation of Article 16 of the Rome II Regulation on overriding mandatory provisions and the other related to interpretation of Article 28 of Directive 2009/103 on protection of victim in case of a motor vehicle accident.
Regarding the overriding mandatory provisons under the Rome II Regulation, the CJEU refers to the definition in Article 9(1) of the Rome I Regulation and reasons that in order to qualify a national rule on statutory limitation period as an overriding mandatory the national court has to be satisfied that there exist “particularly important reasons, such as a manifest infringement of the right to an effective remedy and to effective judicial protection arising from the application of the law designated as applicable”. The relevant part of the CJEU holding uses careful phrasing suggesting restrictive interpretation of overriding mandatory rules: a rule
cannot be considered to be an overriding mandatory provision, […] unless the court hearing the case finds, on the basis of a detailed analysis of the wording, general scheme, objectives and the context in which that provision was adopted, that it is of such importance in the national legal order that it justifies a departure from the law applicable.
Regarding the conflict of law nature of Article 28 of Directive 2009/103, which regulates the Member States’ obligation to provide measures guaranteeing that the victim of a road traffic accident and the owner of the vehicle involved in that accident are protected, the CJEU states that this is not the conflict-of-law provision and that, consequently, it does not take precedence over the Rome II Regulation under Article 27 of the latter.
The Court of Appeal for Ontario has upheld a decision of the Superior Court of Justice dismissing a $2 billion claim against Loblaws relating to the 2013 collapse of the Rana Plaza building in Savar, Bangladesh. In Das v George Weston Limited, 2018 ONCA 1053 (available here) the court concluded that the claims were governed by the law of Bangladesh (not Ontario). It went on to conclude that most of the claims were statute barred under the Bangladeshi limitation period and that it was “plain and obvious” that the remaining claims would fail under Bangladeshi tort law.
Unlike some of the recent cases in this area, this was not a case about a Canadian parent corporation and the operations of its own foreign subsidiary. It was a case about a contractual supply relationship. Loblaws bought clothes (to sell in its Canadian retail stores) from corporations whose workers manufactured the clothes in Rana Plaza.
The key conflict of laws point was the choice of law issue. The rule in Ontario is that tort claims are governed by the law of the place of the tort: Tolofson v Jensen,  3 SCR 1022. The plaintiffs had argued that they were suing Loblaws for negligent conduct that exposed those working in Rana Plaza to harm. They argued that Loblaws had, by adopting corporate social responsibility policies and hiring Bureau Veritas to conduct periodic “social audits” of the workplace, assumed a degree of responsibility for the safety of the workplace in Bangladesh (para 20). They argued that the key steps and decisions by Loblaws took place in Ontario rather than in Bangladesh and therefore Ontario was the place of the tort (para 80). The court rejected these arguments. It held that the place where the alleged wrongful activity occurred was Bangladesh (para 85), that the alleged duty was owed to people in Bangladesh (para 87) and that the injury suffered in Bangladesh “crystallized the alleged wrong” (para 90).
The court also refused to apply Tolofson‘s narrow exception to the place of the tort rule. One reason the plaintiffs raised for triggering the exception was the lack of punitive damages under the law of Bangladesh. The court noted that the lower court’s decision had suggested such damages might actually be available under that law, but in any case “the absence of the availability of punitive damages is not the type of issue that offends Canadian fundamental values” (para 95). The court raised no basis on which to disagree with this analysis.
Because the applicable law was that of Bangladesh, and because some of the claims were not statute-barred, the court was required to do a detailed analysis of Bangladeshi tort law on the duty of care issue in order to determine whether those claims were to be dismissed as not viable. This aspect of the decision may be the most disquieting, since there was little if any on-point authority in the Bangladeshi jurisprudence (para 130). The court had to rely on experts who were relying on a considerable volume of Indian and English cases and then debating the extent to which these would impact the issue if determined by a Bangladeshi court. Ultimately the court concluded that under Bangladeshi law the claims could not succeed.
In common law Canada there is a clear separation between the question of a court having jurisdiction (jurisdiction simpliciter) and the question of a court choosing whether to exercise or stay its jurisdiction. One issue discussed in the Supreme Court of Canada’s recent decision in Haaretz.com v Goldhar (available here) is the extent of that separation. Does this separation mean that a particular fact cannot be used in both the analysis of jurisdiction and of forum non conveniens? On its face that seems wrong. A fact could play a role in two separate analyses, being relevant to each in different ways.
Justice Cote, with whom Justices Brown and Rowe agreed, held that “applicable law, as determined by the lex loci delicti principle, should be accorded little weight in the forum non conveniens analysis in cases where jurisdiction is established on the basis of the situs of the tort” (para 90). She indicated that this conclusion was mandated by the separation of jurisdiction and staying proceedings, which extends to each being “based on different factors”. So if the place of the tort has been used as the basis for assuming jurisdiction, the same factor (the place of the tort) should not play a role in analyzing the most appropriate forum when considering a stay. And since the applicable law is one of the factors considered in that analysis, if the applicable law is to be identified based on the connecting factor of the place of the tort, which is the rule in common law Canada, then the applicable law as a factor “should be accorded little weight”.
In separate concurring reasons, Justice Karakatsanis agreed that the applicable law “holds little weight here, where jurisdiction and applicable law are both established on the basis of where the tort was committed” (para 100). In contrast, the three dissenting judges rejected this reason for reducing the weight of the applicable law (para 208). The two other judges did not address this issue, so the tally was 4-3 for Justice Cote’s view.
As Vaughan Black has pointed out in discussions about the decision, the majority approach, taken to its logical conclusion, would mean that if jurisdiction is based on the defendant’s residence in the forum then the defendant’s residence is not a relevant factor in assessing which forum is more appropriate. That contradicts a great many decisions on forum non conveniens. Indeed, the court did not offer any supporting authorities in which the “double counting” of a fact was said to be inappropriate.
The majority approach has taken analytical separation too far. There is no good reason for excluding or under-weighing a fact relevant to the forum non conveniens analysis simply because that same fact was relevant at the jurisdiction stage. Admittedly the court in Club Resorts narrowed the range of facts that are relevant to jurisdiction in part to reduce overlap between the two questions. But that narrowing was of jurisdiction. Forum non conveniens remains a broad doctrine that should be based on a wide, open-end range of factors. The applicable law, however identified, has to be one of them.
Shortly before Christmas the UKSC released its decision on jurisdiction in Brownlie v Four Seasons Holdings Incorporated (available here). Almost all the legal analysis is obiter dicta because, on the facts, it emerges that no claim against the British Columbia-based holding corporation could succeed (para 15) and the appeal is allowed on that basis. I suppose there is a back story as to why it took a trip to the UKSC and an extraordinary step by that court (para 14) for the defendant to make those facts clear, but I don’t know what it is. On the facts there are other potential defendants to the plaintiffs’ claim and time will tell whether jurisdictional issues arise for them.
The discussion of the value of the place of making a contract for jurisdiction purposes is noteworthy. In para 16 two of the judges (Sumption, Hughes) are critical of using the traditional common law rules on where a contract is made for purposes of taking jurisdiction. This has been the subject of debate in some recent Canadian decisions, notably the difference in approach between the Court of Appeal for Ontario and the Supreme Court of Canada in Lapointe Rosenstein Marchand Melançon LLP v Cassels Brock & Blackwell LLP, 2016 SCC 30 (available here). The SCC was fine with using the traditional rules for this purpose. In Brownlie, I do not think it is clear as to what view the other three judges take on this point.
Even more interestingly, the UKSC judges split 3-2 on how to understand the idea of damage in the forum as a basis for jurisdiction. Three judges (Hale, Wilson, Clarke) retain the traditional broad common law view – the position in many Canadian provinces prior to Club Resorts Ltd v Van Breda, 2012 SCC 17 (available here) – that ongoing suffering in the forum in respect of a tort that happened abroad is sufficient. Two judges (Sumption, Hughes) reject that approach and adopt a more narrow meaning of damage in the forum (it must be direct damage only).
This 3-2 split is closer even than it might first seem, since Lord Wilson (para 57) suggests that in a different case with fuller argument on the point the court might reach a different result.
Canadian law does not get a fair description in the UKSC decision. The court notes twice (para 21 and para 67) that Canada’s common law uses a broad meaning of damage for taking jurisdiction. Club Resorts, and the change to the law it represents on this very issue, is not mentioned. This is yet another illustration of the importance of being careful when engaging in comparative law analysis.
The Supreme Court of Canada has released its decision in Lapointe Rosenstein Marchand Melancon LLP v Cassels Brock & Blackwell LLP, 2016 SCC 30 (available here). The decision builds on the court’s foundational decision in Club Resorts Ltd v Van Breda, 2012 SCC 17, which altered the law on taking jurisdiction in cases not involving presence in the forum or submission to the forum.
In Club Resorts the court held that to take jurisdiction in service ex juris cases the plaintiff had to establish a presumptive connecting factor (PCF) and it identified four non-exhaustive PCFs for tort claims. The fourth of these was that a contract connected with the dispute was made in the forum. This was viewed as unusual: there was very little precedential support for considering such a connection sufficient to ground jurisdiction in tort cases. Commentators expressed concern about the weakness of the connection, based as it was on the place of making a contract, and about the lack of a clear test for determining whether such a contract was sufficiently connected to the tort claim. Both of these issues were squarely raised in Lapointe Rosenstein.
The majority (6-1) agreed with the motions judge and the Court of Appeal for Ontario that this PCF was established on the facts of this case. Justice Cote dissented, concluding both that the contract was not made in Ontario and that it was not sufficiently connected with the tort claim.
The facts are somewhat complex. After the 2008 financial crisis the Canadian government bailed out General Motors of Canada Ltd (GM Canada). In return for this financial support, GM Canada agreed to close dealerships (ultimately over 200) across Canada. Each dealership being closed was compensated under a Wind-Down Agreement (WDA) between GM Canada and the dealer. The WDA was governed by Ontario law and contained an exclusive jurisdiction clause for Ontario. The WDA required each dealer to obtain independent legal advice (ILA) about the consequences of signing the WDA.
Some time after the dealerships closed over 200 dealers brought a class action in Ontario against GM Canada disputing the legality of the WDAs. They also sued Cassels Brock & Blackwell, the lawyers for the Canadian Automobile Dealers Association, for negligent advice to the dealers. In turn, Cassels Brock brought third-party claims against 150 law firms which had provided the ILA to the dealers. Many of the law firms, including those in Quebec, challenged the court’s jurisdiction over the third-party claim. Cassels Brock argued that the WDAs were contracts made in Ontario and that the WDAs were connected with the tort claim Cassels Brock was advancing in the third-party claim (which was for negligence in providing the ILA).
The court had the chance to adjust or move away from this PCF, given the criticism which it had attracted (see para 88). But it affirmed it. Worse, the Court of Appeal for Ontario had at least expressed a willingness to be flexible in determining the place of making of the contract (which in part got around the central weakness in this PCF). In contrast the majority stresses the “traditional rules of contract formation” (para 31). Insisting on the traditional rules is what gives rise to the core difference between the majority (Ontario: paras 42-43) and the dissent (Quebec: paras 74-80) on where the WDAs were made. Those rules mean the dissent is right to point out (para 81) that related connections between the WDAs and Ontario (such as the applicable law and the jurisdiction clause: see para 48) do not, strictly speaking, have anything to do with where the contract is made and so must be ignored on that issue. The more robust approach of the Court of Appeal allows more to be assessed and thus for an easier (more consensual) conclusion that the WDAs were “made” in Ontario. There is reason to be quite concerned that the Supreme Court of Canada’s approach will lead to more disputes about where a particular contract has been made, focusing on technical rules, which is unwelcome.
The court also splits on whether the contract, if made in Ontario, is connected to the tort claim. I am inclined to think the majority gets it right when it finds that it is. Note, though, that I think it is wrong to claim, as the majority does (para 47 last sentence), that somehow the law firms were brought “within the scope of the contractual relationship” by providing the advice about it. The best part of the dissent is the demolition of that claim (para 86). The real problem is that a close enough connection should be available to be found even in the absence of bringing the defendant “within” that contractual relationship. This PCF, if the misguided narrow focus on place of contracting could be overcome, can be broader than that and thus broader than the dissent would make it (para 87).
Here a local Quebec law firm is asked by its local client to provide it with advice about the client’s entering into the WDA. The terms of the WDA expressly say that to so enter into it the client has to get that advice. The WDA is clearly very connected to Ontario. It seems to me right to say that the WDA is a contract related to any subsequent negligent advice claim the client would advance against the firm. The WDA is not just context, bearing peripherally on the advice. The advice entirely centers on the WDA and whether the client should enter into it. The WDA is what the advice is about. The majority gets all of this right in para 47 except for its last sentence. Of the 11 judges who addressed this issue in the three levels of court, only Justice Cote finds the connection between the contract and the tort claim to be insufficient.
So I think the decision is right but the majority errs by stressing the traditional rules of contract formation for assessing the place of making and by using the “within the scope of the contractual relationship” test for the requisite connection.
Some smaller points:
1. I am somewhat puzzled by the idea (para 31) that parties would expressly think about how they would go about making their contracts so as to have them made in a particular place so as to get to subsequently take advantage of this PCF. Do parties think like that? Did they before this PCF was created? I suppose it is easier to say they now do think like that since they are being told to do so by the court.
2. For future debates about where contracts are made, I worry about some of the court’s language. One example is para 40’s reference to where the acceptance “took place”. Is that compatible with the postal acceptance rule which looks, for some contracts, at the place of posting rather than place of receipt? Would we say the acceptance in such a case “took place” at the place of posting? See in contrast para 73.
3. Justice Cote’s dissent could be seen as a covert attempt to eliminate this PCF. She insists on a very tight connection between the contract and the tort claim. She refers to circumstances in which “the defendant’s breach of contract and his tort are indissociable” (para 95; emphasis in original) and states that this PCF “only provides jurisdiction over claims where the defendant’s liability in tort flows immediately from the defendant’s own contractual obligations” (para. 90). In such cases, this PCF (tied to the place of contracting) might safely be abolished and replaced with other, better PCFs relating to tort and contract claims (especially in light of para 99 of Club Resorts). It would not be needed for the court to be able to take jurisdiction, as it was on the facts of Club Resorts and Lapointe Rosenstein. I am sympathetic to a desire to eliminate this PCF, but I think that result needed to be confronted directly rather than indirectly. In the wake of the majority decision, it is now unlikely to happen at all.
After 13 months the Supreme Court of Canada has finally released its decisions in four appeals on the issue of the taking and exercising of jurisdiction. The main decision is in Club Resorts Ltd v Van Breda (available here) which deals with two of the appeals. The other two decisions are Breeden v Black (here) and Editions Ecosociete Inc v Banro Corp (here).
The result is perhaps reasonably straightforward: in all four cases the court upholds the decisions of both the motions judges and the Court of Appeal for Ontario. All courts throughout held that Ontario had jurisdiction in these cases and that Ontario was not a forum non conveniens.
The reasoning is more challenging, and it will take some time for academics, lawyers and lower courts to work out the full impact of these decisions. The court’s reasoning differs in several respects from that of the courts below.
The court notes that a clear distinction needs to be drawn between the constitutional and private international law dimensions of the real and substantial connection test. This is an interesting observation, particularly in light of the fact that the court’s own decision is not as clear on this distinction as it could be. I expect that going forward there will be different interpretations of what the court is truly saying on this issue.
The court is reasonably clear that the real and substantial connection test should not be used as a conflicts rule in itself. It is not a rule of direct application. Rather, it is a principle that informs more specific private international law rules governing the taking of jurisdiction. This is a change from the approach used by provincial appellate courts, especially the Court of Appeal for Ontario, which arguably had been using the real and substantial connection test as its rule, at least in part, for establishing jurisdiction in service ex juris cases.
The court states that it is establishing the framework for the analysis of jurisdiction. Going forward, a real and substantial connection must be found through a “presumptive connecting factor” which is a factor that triggers a presumption of such a connection. The presumption can be rebutted. If the plaintiff cannot establish such a presumption, the court cannot take jurisdiction. This last point is perhaps the largest change made to the law. On the law as it stood, the plaintiff could establish jurisdiction through a variety of non-presumptive factual connections that collectively amounted to a real and substantial connection to the forum. That approach is rejected by the Supreme Court of Canada.
The court does not purport to set out a complete list of presumptive connections. It confines itself to identifying some such connections that could apply in tort cases, namely that (a) the defendant is domiciled or resident in the forum, (b) the defendant carries on business in the forum, (c) the tort was committed in the forum, and (d) a contract connected with the dispute was made in the forum. It is quite open, on the language in the decisions, as to what other presumptive connections lower courts will need to be finding in other cases. One possible solution is that lower courts will largely continue to follow the recent approach of the Court of Appeal for Ontario that the enumerated bases for service ex juris, subject to some exceptions, amount to such presumptive connections.
The decisions also address the test for the doctrine of forum non conveniens. Three points can be made about that analysis. First, the language suggests the burden is always on the defendant/moving party. Second, emphasis is placed on “clearly” in “clearly more appropriate”, suggesting that it will be harder to displace the plaintiff’s choice of forum. Third, the court cautions against giving too much weight to juridical advantage factors. Judges should avoid invidious comparisons across forums and refrain from “leaning too instinctively” in favour of the judge’s own forum.
The decisions are not a radical break with the earlier cases but they do change the law on taking jurisdiction in several respects. In addition, the court makes several points along the way, as asides, that will impact other aspects of the conflict of laws. For example, the court confirms the propriety of taking jurisdiction based on the defendant’s presence in the forum.
In Kazemi (Estate of) v. Islamic Republic of Iran, 2011 QCCS 196 (available here) the estate of Zahra Kazemi and her son, Stephan Kazemi, sued Iran and certain state officials in Quebec, alleging that in 2003 Ms. Kazemi was tortured and assassinated in Iran. The defendants argued that the claim could not succeed due to state immunity.
Much of the court’s analysis involves the provisions of the State Immunity Act, R.S.C. 1985, c. S-18. The court has to consider whether this statute is a complete code on the issue of state immunity or whether it is open to courts to create exceptions to the statutory immunity beyond those listed in the statute. The court also has to address whether aspects of the statute are constitutional.
The court ends up concluding that the estate has no claim because the wrongs done to her occurred in Iran and so are covered by the immunity under the statute. However, the court allows the claim by Stephan Kazemi, a claim for his own losses arising from hearing the reports of what was done to his mother, to continue since his losses were suffered in Quebec, not Iran, and so the immunity does not cover them (see section 6 of the statute).
The decision is lengthy (220 paragraphs), and yet it does not mention the recent decision of the Supreme Court of Canada on state immunity: Kuwait Airways Corporation v. Republic of Iraq from October 2010.
Brill / Martinus Nijhoff has recently published The Rome II Regulation on the Law Applicable to Non-Contractual Obligations: A New International Litigation Regime. The book is edited by John Ahern and William Binchy of Trinity College Dublin. Full details of the book are available here. It can be ordered through this link from the publisher or web sites like Amazon.
The book is the result of a conference held in Dublin in June 2008. It contains fifteen chapters by authors from across Europe and North America.
The province of Manitoba’s Law Reform Commission has released a report on Private International Law (available here). It considers three central issues:
1. Should legislation be adopted to modify the common law choice of law rule for torts as formulated in Tolofson v. Jensen?
2. Should legislation be adopted regarding the characterization of limitation periods?
3. Should Manitoba adopt the Uniform Law Conference of Canada’s model Court Jurisdiction and Proceedings Transfer Act?
A secondary question under the first issue is how similar the legislation should be to the English PIL(MP)Act 1995.