SUPREME COURT OF CANADA
18 APRIL 2012
(delivered the judgment of the court - Binnie and Charron JJ. took no part in the judgment)
With a globalized world comes the sometimes poisonous gift of ubiquity. Our conduct, however local, is now more likely to have global effects. For instance, words printed in one location can inform, or offend, readers all over the world. In the present case, an Ontario-based corporation has taken legal action in Ontario against a publisher based in the province of Quebec for defamation. Can an Ontario court assume jurisdiction over this action and, if so, should it exercise its jurisdiction?
The decisions of this Court have established that a court can assume jurisdiction over an action if there is a real and substantial connection between the action and the court’s territory. In companion cases, this Court has set out a series of factors that would meet the real and substantial connection test (see Club Resorts Ltd. v. Van Breda, 2012 SCC 17). As in Club Resorts, we must determine here whether there is a real and substantial connection between the plaintiff’s claim in defamation and the chosen forum, namely, an Ontario court.
The tort of defamation presents an interesting challenge for the principles underlying the assumption of jurisdiction. At common law, the tort of defamation crystallizes upon publication of the libellous material, and publication of the libellous material is presumed when it is printed in a book. The tort of defamation will thus crystallize in all jurisdictions where the book is available. This also raises difficult issues when publication occurs through the Internet, as this Court noted recently in Crookes v. Newton, 2011 SCC 47,  3 S.C.R. 269. Is it sufficient, however, that the defamatory book be available in a jurisdiction for a court to assume jurisdiction over a defamation claim involving that book? If a court may assume jurisdiction on that basis, in what circumstances should it apply the doctrine of forum non conveniens and decline to exercise its jurisdiction in favour of another, more convenient forum? This case also raises the issue of “libel tourism”: if more than one forum can assume jurisdiction over a single instance of tortious conduct, should we prevent plaintiffs from choosing the forum of greatest juridical advantage?
For the reasons that follow, I must dismiss this appeal. The assumption of jurisdiction is justified under the real and substantial connection test where there exist appropriate objective factors connecting the plaintiff’s claim to his or her chosen forum. In the present case, the commission of the tort of defamation in Ontario satisfies the real and substantial connection test. Though it could be argued that a Quebec court is an appropriate forum, I see no reason to interfere with the motion judge’s decision not to decline to exercise the Ontario court’s jurisdiction in this matter.
B. Background Facts
The appellant Les Éditions Écosociété Inc. (“Écosociété”) is the publisher of a book by A. Deneault, with D. Abadie and W. Sacher, entitled Noir Canada: Pillage, corruption et criminalité en Afrique (2008) (“Noir Canada”), which comments on the international mining activities of some Canadian corporations, including the respondent, Banro Corporation (“Banro”). Banro brought an action in Ontario against Écosociété, Deneault, Abadie and Sacher, alleging that Noir Canada’s content is libellous and that the book accuses Banro of committing human rights violations and fraud to further its financial interests in Africa. The defendants moved to stay the Ontario action, submitting that there was no real and substantial connection between the subject-matter of the action and Ontario and that the Ontario court was not the convenient forum for the action.
Banro is incorporated under the Canada Business Corporations Act, R.S.C. 1985, c. C-44, and is engaged in the exploration and development of gold properties in the Democratic Republic of Congo (“DRC”). Banro’s activities are international in scope; it has offices and hundreds of employees located in Ontario, the United Kingdom, South Africa and the DRC. Four of Banro’s nine officers are residents of Ontario, three are residents of the DRC, one officer resides in the United Kingdom and another in South Africa. Banro’s Board of Directors is composed of eight members, two based in Ontario, two in South Africa, two in the United Kingdom, one in France and one in British Columbia. Banro’s head office is located in Toronto, and the corporation is publicly traded on the Toronto Stock Exchange.
Écosociété, the defendant in the main action, is a corporation based in Montréal, Quebec, that is incorporated under the laws of that province. Its head office is in Montréal, where all of the individual defendants work and reside. Noir Canada was published in Montréal and released to bookstores for purchase to the public on or about April 15, 2008. The defendant Alain Denault is the author, while defendants Delphine Abadie and William Sacher contributed research and editorial assistance. Two French editions of the book have been printed, totalling nearly 5,000 copies, of which 93 were distributed to bookstores in Ontario, including 27 in the city of Toronto. A number of copies are available in Ontario public libraries. Noir Canada is also available for purchase on Écosociété’s website. The defendant William Sacher gave a speech about the book at the University of Toronto on July 27, 2008. The book has also been referenced by many websites and newspaper articles available and accessible in Ontario.
Banro has no offices, affiliates or subsidiaries in Quebec, nor does it operate or own mining properties in that province. Banro’s public profile in Quebec is very limited.
C. Judicial History
(1) Ontario Superior Court of Justice, 2009 CanLII 7168 (Roberts J.)
Roberts J. delivered her reasons prior to the release of Charron Estate v. Village Resorts Ltd., 2010 ONCA 84, 98 O.R. (3d) 721 (“Van Breda-Charron”), in which the Ontario Court of Appeal revisited the multi-pronged test for the assumption of jurisdiction developed in Muscutt v. Courcelles (2002), 60 O.R. (3d) 20 (C.A.). On the basis of the latter test, Roberts J. dismissed the defendants’ motion to stay the proceedings. Roberts J. ruled that the Ontario court did have jurisdiction, owing to a real and substantial connection between the forum and the action. Roberts J. also dismissed the motion on the grounds of forum non conveniens; in her view, the defendants had not met the onus of showing that a Quebec court was the more convenient forum.
(a) Jurisdiction Simpliciter
On the basis of the criteria set out in Muscutt, Roberts J. found that the Ontario court could assume jurisdiction over the defendants. She found that there was a real and substantial connection between Ontario and the plaintiff’s claim. Roberts J. recognized that the plaintiff was based in Ontario, where its good reputation was relied on to attract investors and reassure shareholders and was also critical in its dealings with various regulators. Roberts J. also held that the tort of defamation was only complete upon publication, which occurred where libellous material was read. Libellous material printed in a book was presumed to have been published for the purposes of establishing the tort.
Roberts J. also noted that proof of the claim was not necessary at this early stage of the proceedings; rather, the plaintiff’s claim was to be taken at face value. Without determining the veracity of the facts alleged by the plaintiff in its action, she further took the view that it had pleaded sufficient facts to support a defamation claim. With respect to damages, Roberts J. stated that it was recognized in defamation law that the vindication of the plaintiff’s reputation was just as important as any monetary award that might be obtained. She wrote that the plaintiff “may be quite content with a declaration by a court in Ontario that the statements made by the defendants are untrue even if it cannot recover any damages that might be awarded to it as a consequence” (para. 27). She therefore held that there was a significant connection between Ontario and the respondent’s claim.
With respect to a connection between the forum and the defendants, Roberts J. noted that the corporate defendant was headquartered in Quebec and that the individual defendants all resided in that province. That said, she stated that “if the defendants have done anything within Ontario that bears upon the claim advanced by the plaintiff, the case for assuming jurisdiction is strengthened and may render the defendants subject to the jurisdiction” of the Ontario court (para. 32).
For Roberts J., the question was “whether it was reasonably foreseeable that the defendants’ conduct would result in harm within Ontario” (para. 32). Roberts J. relied on the principles enunciated in Barrick Gold Corp. v. Blanchard & Co. (2003), 9 B.L.R. (4th) 316 (Ont. S.C.J.), to hold that it was reasonably foreseeable that the defendants’ conduct would result in harm in Ontario, because the marketing and distribution of Noir Canada was not limited to Quebec, because the alleged defamatory statements contained in Noir Canada were published in Ontario, and because Mr. Sacher promoted Noir Canada during his visit to Toronto. Consequently, Roberts J. found that the activities of the defendants in Ontario rendered them subject to the jurisdiction of the Ontario court.
As required by the Muscutt test, Roberts J. then considered whether the assumption of jurisdiction was unfair to one of the two parties. Roberts J. was not convinced by any arguments that assuming jurisdiction would be unfair to the defendants. With respect to Banro, although she recognized that, if the action were tried in Ontario, Banro would enjoy a juridical advantage by avoiding what she viewed as the more stringent requirements of Quebec defamation law, she held that Banro had “a legitimate interest in having the falsehood of the allegedly defamatory statements proven in its home jurisdiction and to clear its name” (para. 49).
Roberts J. also considered whether the Ontario court would be willing to recognize an extraprovincial judgment rendered on the same jurisdictional basis, and found that it would without question. She was of the view that in light of this Court’s holding in Morguard Investments Ltd. v. De Savoye,  3 S.C.R. 1077, Canadian provinces were constitutionally mandated to recognize and enforce judgments rendered in sister provinces as a result of a proper assumption of jurisdiction. Roberts J. therefore found that this factor favoured the plaintiff. She further noted that, even if Quebec did not enforce an Ontario judgment, such a judgment would remain of significant value to the plaintiff in restoring its reputation in Ontario.
(b) Forum Non Conveniens
Having found that the Ontario court had jurisdiction to hear this action, Roberts J. turned to the question of whether she ought to decline to exercise it pursuant to the forum non conveniens doctrine. Roberts J. did so by applying the “well established, although not exhaustive, factors typically used to assess the connections to the parties’ respective fora” (para. 61). She concluded that the defendants had not satisfied their onus of showing that Quebec was a clearly more appropriate forum for the hearing of the action.
Roberts J. first considered the location where the tort occurred. She rejected the defendants’ argument that the action should be heard in Quebec, where the vast majority of the copies of the book were distributed. Despite the fact that only 93 copies of the book had been distributed in Ontario, Roberts J. held that this factor favoured the plaintiff. She wrote that the distribution of “fewer copies ... in Ontario than Quebec is not conclusive of the issue of damage to reputation and the ability of the person allegedly defamed to clear its good name” (para. 65). She noted that this was especially the case when the book was distributed in the home jurisdiction of the person allegedly defamed, where his or her reputation may suffer the most damage.
Second, Roberts J. considered the law applicable to the tort. She noted that the law applicable to a tort was determined in accordance with the lex loci delicti rule and that in actions for defamation, it was that of the jurisdiction where publication occurred. Since Noir Canada was distributed and accessed in Ontario, she held that Ontario law would apply to the present action. She thus found that this factor favoured the plaintiff.
With respect to the next factor, the place of residence or place of business of the parties, Roberts J. took the view that, in determining the forum conveniens in a defamation action, the place of residence of the plaintiff was significant because this was the jurisdiction in which he or she was likely to suffer the most damage to his or her reputation. She held that this factor favoured the plaintiff.
After briefly considering other factors, Roberts J. addressed the remaining two factors, namely the avoidance of a multiplicity of proceedings and the loss of juridical advantage. For reasons previously outlined in her analysis of the unfairness in assuming jurisdiction, she held that the defendants had not established the risk of inconsistent judgments. She also held that “it was reasonably foreseeable that the plaintiff would commence an action in Ontario because of the defendants’ activities in Ontario” (para. 79). On this latter point, Roberts J. was of the view that Banro had not engaged in forum shopping by commencing the action in Ontario. Owing to the real and substantial connection between Ontario and the plaintiff’s claim, she was of the opinion that Banro was entitled to bring its action in its home jurisdiction, and hence benefit from the juridical advantages of that jurisdiction.
(2) Ontario Court of Appeal, 2010 ONCA 416 (CanLII) (Weiler, Blair and Rouleau JJ.A.)
In a short judgment, the Court of Appeal endorsed the motion judge’s order and reasons. The Court of Appeal was of the view that, though the Muscutt test had been modified in Van Breda-Charron, the underlying principles for assuming jurisdiction were the same and had been properly applied by the court below. In its view, the record provided “ample support” for Robert J.’s conclusion that there was a real and substantial connection between the action and the forum chosen by the plaintiff (para. 2). The Court of Appeal also held that there was no error in the motion judge’s exercise of discretion with respect to the forum non conveniens analysis.
A. Positions of the Parties
(a) Jurisdiction Was Not Properly Assumed in this Case
The appellants raise three main arguments to support their submission that jurisdiction was not properly assumed in this case. First, they contend that, in multijurisdictional defamation cases, the constitutional principles of order and fairness restrict a court’s power to assume jurisdiction over an action to situations where there has been substantial publication in the jurisdiction. They write that “[i]n the absence of other substantial connections between the subject matter of the action or the defendant and the forum there must at least be substantial publication for a court to find a real and substantial connection” (A.F., at para. 43). They argue that should this Court fail to endorse that approach, authors or publishers whose publications circulate in multiple jurisdictions would constantly face the threat of being sued in an unlimited number of jurisdictions. That threat could easily operate as a “libel chill” and be harmful to freedom of expression in Canada.
Second, the appellants argue that there is no real and substantial connection between the forum and the defendants as, among other things, there is no evidence on the record that suggests that the respondent suffered harm to its reputation in Ontario. For that reason, the appellants contend that the motion judge wrongly relied on the respondent’s bald assertion that its reputation in Ontario was “particularly important” (A.F., at para. 62). It is the appellants’ view that, in light of the minimal distribution of the impugned book and the lack of evidence of publication, injury to reputation could not be inferred.
Third, the appellants challenge the motion judge’s conclusion that Écosociété was subject to the jurisdiction of the Ontario court because it was carrying on business in Ontario. In their opinion, the mere fact of offering books for sale in bookstores in Ontario and on a website which is accessible from Ontario cannot properly ground a finding of “carrying on business” in that jurisdiction. The appellants thus submit that Roberts J.’s finding to that effect is unsupported by the evidence and is inconsistent with the constitutional principles of order and fairness.
(b) Quebec Is the More Appropriate Forum
If this Court were to find that Ontario had jurisdiction, the appellants submit that the courts below erred in law by failing to recognize that the following four factors showed that Quebec was a clearly more appropriate forum in which to hear the present action (A.F., at para. 93):
Finally, the appellants invite this Court to clarify the choice of law rule that applies in multijurisdictional defamation cases. They submit that the law of the jurisdiction where substantial publication occurred should apply. In the appellants’ view, Quebec law should govern the present dispute regardless of whether the action proceeds in Quebec or Ontario.
(a) The Courts Below Rightly Assumed Jurisdiction Over the Appellants
It is the respondent’s view that the appellants, through this appeal, are “seeking to redefine defamation jurisprudence by introducing into a preliminary jurisdictional analysis tort-specific considerations of publication of defamatory material and a party’s ability to substantiate damages” (R.F., at para. 4). The respondent contends that “a restructuring of Canadian libel law in the manner that the appellants propose is contrary to the fundamental considerations underlying Canadian law on jurisdiction, ... and [will] distance such Canadian law from that of other common law nations” (R.F., at para. 1). Furthermore, according to the respondent, requiring plaintiffs in multijurisdictional cases to show that substantial publication occurred in the jurisdiction and to adduce evidence of sufficient harm to reputation in that jurisdiction would subject them to an evidentiary burden that plaintiffs in intrajurisdictional cases do not face.
The respondent emphasizes that the decisions upon which the appellants rely to support their assertion that plaintiffs in multijurisdictional defamation cases must show, at the jurisdictional stage, that publication in the jurisdiction was more than minimal have recently been overturned by the Ontario Court of Appeal in Paulsson v. Cooper, 2011 ONCA 150, 105 O.R. (3d) 28. The respondent notes that in that decision, the court adopted Nordheimer J.’s reasoning in Barrick Gold v. Blanchard to the effect that reasonable foreseeability of harm to reputation, rather than evidence of damage to the plaintiff’s reputation, is a relevant consideration in determining whether the foreign defendant in a defamation case has connections with the jurisdiction in which the action is brought. The respondent also asserts that the Court of Appeal in Paulsson departed from its previous decision rendered in Bangoura v. Washington Post (2005), 258 D.L.R. (4th) 341.
Relying on the Ontario Court of Appeal’s reasons in Paulsson, the respondent writes that “[as] with any tort, the reasonable foreseeability of harm to another party can serve to affix liability in defamation actions”. The respondent submits that the appellants knew that the distribution of nearly one hundred copies of Noir Canada in bookstores in Ontario and online would result in the book being read and purchased in that jurisdiction. The respondent argues that the appellants were equally aware of the respondent’s strong connections with Ontario. Therefore, according to the respondent, it was reasonably foreseeable that Noir Canada would be read and purchased in Ontario and would cause harm to the respondent’s reputation in that jurisdiction. Consequently, the respondent submits that a presumption of a real and substantial connection exists based on r. 17.02(g) of the Rules of Civil Procedure and that the fact that a tort occurred in Ontario is a significant factor supporting the assumption of jurisdiction by the Ontario court in the present action.
The respondent accordingly argues that the appellants’ core arguments ought to be rejected. In the respondent’s view, the requirement of “substantial publication” not only would be unworkable and create confusion, but runs counter to the law of jurisdiction in Canada, which clearly distinguishes between issues of standing to bring a claim in a given forum and the substantive elements of that claim. The respondent also notes that this proposed requirement is based on a “snapshot” in time. It writes that “[t]o impose this concept would absurdly require the plaintiff, upon learning of a defamatory publication, to simply await ‘substantial’ publication, while incurring the consequent damage, in order to ensure that its defamation claim would be permitted to proceed” (R.F., at para. 79). Similarly, requiring the plaintiff to prove “harm to reputation” at the jurisdictional stage is inconsistent with one of the basic tenets of defamation law in Canada, namely that the main concern in a libel action is vindication of reputation, not the damages that one can obtain. It also fails to take account of the well-established principle of defamation law that damages are presumed when the defamatory material is made available to the public at large.
(b) Ontario Is the Appropriate Forum to Hear This Action
Relying on Young v. Tyco International of Canada Ltd., 2008 ONCA 709, 92 O.R. (3d) 161, the respondent submits that a motion judge’s exercise of his or her discretion as to forum non conveniens is owed deference on appeal. According to the respondent, the party who seeks the exercise of the court’s discretion must show that the forum it proposes is clearly more appropriate, which is an “extremely heavy burden” to meet on appeal (R.F., at para. 96). In this case, the respondent takes the view that the appellants have failed to meet this burden, and further contend that the courts below reached the correct conclusion in finding that Ontario was the most appropriate forum to hear this action.
The respondent argues that the correctness of the conclusion reached by the courts below is chiefly evidenced by the following factors:
its residence is in Ontario, a factor which is a matter of significance in a defamation case;
Ontario law would govern the action;
there is no risk of multiplicity of proceedings as the respondent is not a party to the defamation action filed against the appellants by Barrick Gold Corporation (“Barrick Gold”) in Quebec; and
the respondent would suffer a major juridical disadvantage if this action were tried in Quebec in terms of burden of proof and because its action would be barred by the application of the limitation period, or rather prescription in Quebec law, pursuant to art. 2929 of the Civil Code of Québec, S.Q. 1991, c. 64.
The respondent further submits that this latter factor is sufficient to outweigh all other considerations in the forum non conveniens analysis. Finally, in response to the appellants’ argument that it is paramount that they not be deprived of the benefit of Quebec’s anti-SLAPP legislation, the respondent asserts that “there is no right to inaccurate, unbalanced and biased reporting that can unjustifiably damage reputation” (R.F., at para. 132).
B. The Issues
In my reasons in Club Resorts, I set out the analytical framework for assuming jurisdiction under the common law real and substantial connection test. I also set out the proper approach to the application of the doctrine of forum non conveniens.
Applying this analysis to the tort of defamation poses special challenges; in particular, it raises concerns about forum shopping. The tort of defamation is crystallized upon publication of the libellous material. In Canada, publication occurs when libellous material is read by a third party. In the case of libellous material printed in a book that is circulated in a library, it is possible to draw an inference of publication (R.E. Brown, The Law of Defamation in Canada (2nd ed. (loose-leaf), at p. 22-58; Hiltz and Seamone Co. v. Nova Scotia (Attorney General) (1997), 164 N.S.R. (2d) 161 (S.C.), at para. 28; Paulsson v. Cooper, at para. 37).
The defendants in this action have expressed the concern that an overly flexible application of the real and substantial connection test would render them liable in defamation in more than one jurisdiction. Indeed, given the elements of the tort of defamation, if an allegedly libellous book is distributed in more than one jurisdiction, then an inference may be drawn that the libellous material has been published in all these jurisdictions. If publication is sufficient to connect the plaintiff’s claim to a given jurisdiction, then the courts of more than one jurisdiction could potentially assume jurisdiction over the same tort.
The elements of a tort such as defamation potentially vary from one jurisdiction to another, thus making it easier or more difficult to sue depending on one’s choice of jurisdiction. That being the case, a plaintiff might make a strategic choice and sue in the jurisdiction in which he or she enjoys the greatest juridical advantage. This is the well-known problem of “forum shopping” or “libel tourism”. I propose to address this problem at the forum non conveniens stage of the analysis.
C. Application to the Facts
(1) The Assumption of Jurisdiction
The motion judge’s decision to assume jurisdiction should be upheld. When the analytical framework identified in Club Resorts is applied, it is clear that there is a real and substantial connection between Banro’s claim and Ontario.
Here, the alleged tort of defamation occurred in Ontario. Noir Canada was distributed in Ontario. At this stage of the proceedings, the plaintiff need not show evidence of harm or that the book was read. The plaintiff need only allege publication and its allegations should be accepted as pleaded unless contradicted by evidence adduced by the defendants. For the purposes of proving defamation, publication may be inferred when the libellous material is contained in a book that is circulated in a library; the new evidence adduced by Banro on consent establishes that 15 copies of Noir Canada were circulated in Ontario libraries and one copy was checked out. In addition, Banro adduced evidence establishing that its reputation in Ontario is vital to conducting business, attracting investors and maintaining good relations with regulators such as the Ontario Securities Commission.
As discussed in Club Resorts, the commission of a tort in Ontario is a recognized presumptive connecting factor that prima facie entitles the Ontario court to assume jurisdiction over this dispute. For the reasons discussed above, the defendants have not shown that only a minor element of the tort of defamation occurred in Ontario. As a result, they have not displaced the presumption of jurisdiction that arises in this case.
On this basis, I conclude that the motion judge correctly assumed jurisdiction. That said, it is then entirely appropriate for the respondent party in such a motion to raise the doctrine of forum non conveniens, and ask that factors that go beyond the objective connecting factors considered in the jurisdictional analysis be taken into account.
(2) Forum Non Conveniens
The application of forum non conveniens is an exercise of discretion reviewable in accordance with the principle of deference to discretionary decisions: an appeal court should intervene only if the motion judge erred in principle, misapprehended or failed to take account of material evidence, or reached an unreasonable decision (see Young v. Tyco International of Canada Ltd., at para. 27). I find that the motion judge made no such error. Roberts J. correctly exercised her discretion in maintaining Ontario’s jurisdiction over the plaintiff’s claim.
In this Court, the defendants put forward the following reasons for which the Ontario court should have declined to exercise its jurisdiction:
I will review each of these submissions in turn.
(a) The Cost of the Proceeding
The defendants’ first argument is one of proportionality. The defendants submit that the cost of litigating the plaintiff’s claim in Ontario far outweighs the potential for recovery. The defendants submit that the evidence necessary for litigating the claim is outside Ontario, and, consequently, that “the expense and cost of litigating those issues is out of all proportion to the potential recovery given the limited publication in Ontario and the absence of evidence of harm to reputation in Ontario” (A.F., at para. 100).
The defendants’ argument of proportionality rests on the lack of evidence of harm, which is not of concern at this preliminary stage of the proceedings. The plaintiff’s claim is assumed to be proven as pleaded unless contradicted by evidence. Furthermore, harm is typically presumed in defamation cases, and in this case Banro has adduced sufficiently compelling evidence of its reputation in Ontario. Finally, financial recovery may not be the central issue. It is conceivable that, for Banro, a declaratory judgment is as valuable to its reputation as any pecuniary award.
(b) Parallel Proceedings in Quebec
The defendants submit that “[m]ultiplicity of proceedings should be avoided because of the waste of judicial resources and the risk of inconsistent findings” (A.F., at para. 104). The defendants are referring to a defamation action brought by Barrick Gold which they were still facing in Quebec at the time of the hearing in this Court, also in relation to Noir Canada. I gather that the defendants are bearing the cost of two separate lawsuits and that litigating both actions together would minimize costs.
With respect, I cannot agree with the defendants. Firstly, it is incorrect for them to say that “[m]atters with parties or issues in common should be litigated in the same forum” (A.F., at para. 104). Rather, it should be said that matters with parties or issues in common should be litigated in the same action. Even if Ontario declined to exercise its jurisdiction, it is not a foregone conclusion that Banro’s action would be amalgamated with Barrick Gold’s action in Quebec. In Visram v. Chandarana, 2007 CanLII 28334 (Ont. S.C.J.), which the defendants cite as supporting their contention, the Ontario Superior Court had the option of staying some of the claims in an action on the basis that Ontario was not the most appropriate forum, while allowing other claims asserted against Ontario residents only to proceed in Ontario. The motion judge declined to stay part of the action, not to avoid a multiplicity of fora, but to avoid a multiplicity of proceedings arising from the overlapping claims against the same parties. In the present case, Barrick Gold’s action and Banro’s action are already separate claims, and although they arose from the same book, the claims are related to different allegedly defamatory statements.
In addition, the defendants submit that a multiplicity of proceedings would risk “Ontario and Québec courts [coming] to opposite conclusions” (A.F., at para. 107). With respect, I am not persuaded by this argument. Again, there is no guarantee that these actions would be joined, if both were litigated in Quebec. It is thus entirely conceivable that the same Quebec court would also come to opposite conclusions in respect of the two actions simply because the plaintiffs, the facts and the statements relating to each plaintiff are different.
(c) Choice of Law
One factor that must be considered in the forum non conveniens analysis is the law applicable to the tort. Restricting the available choice of laws might be a way to curb forum shopping. Indeed, there would be little strategic advantage to forum shopping if the conflicts rules were to require application of the same law regardless of where the matter is tried.
In Tolofson v. Jensen,  3 S.C.R. 1022, La Forest J. established lex loci delicti, or the place where the tort occurred, as a general principle for determining choice of law for torts. However, La Forest J. also left room for the creation of exceptions to the general rule of lex loci delicti for torts such as defamation. The rationale for the rule is that in the case of most torts, the occurrence of the wrong constituting the tort is its most substantial or characteristic element, and the injury or consequences are typically felt in the same place. In establishing lex loci delicti as a general rule, however, La Forest J. also recognized that “[t]here are situations .... notably where an act occurs in one place but the consequences are directly felt elsewhere, when the issue of where the tort takes place itself raises thorny issues. .... Difficulties may also arise where the wrong directly arises out of some transnational or interprovincial activity” (p. 1050).
La Forest J. suggested that in such cases, “it may well be that the consequences would be held to constitute the wrong” (ibid.). Significantly, La Forest J. went so far as to suggest without deciding that the tort of defamation may be just such a case: “[I]t could well be argued .... that, unlike a motor vehicle accident [the tort at issue in Tolofson], the tort of libel should be held to take place where its effects are felt” (p. 1042). La Forest J. thus left room for the creation of exceptions to the general rule of lex loci delicti for torts such as defamation.
The defendants argue against the application of lex loci delicti here. They submit that the Ontario court should have declined to exercise its jurisdiction because Quebec is the place of the most substantial publication, and therefore Quebec law is applicable. The defendants cite Olde v. Capital Publishing Ltd. Partnership (1996), 5 C.P.C. (4th) 95 (Ont. Ct. (Gen. Div.)), aff’d (1998), 108 O.A.C. 304, as standing for the proposition that we should look to the standard of substantial publication to determine the centre of gravity of the tort, and identify the most convenient forum on that basis. If I understand the argument, Ontario could assume jurisdiction on the basis that the tort occurred there; Ontario defamation law uses the traditional common law standard of publication, but substantial publication would be considered to determine whether Ontario should, in fact, exercise its jurisdiction based on whether or not its law would be applicable.
The defendants’ argument is not persuasive. First, there is a factual difference between Capital Publishing and the present case. In Capital Publishing, the libellous material was published in a magazine distributed mainly in the United States; the record showed that only one copy of the magazine was purchased in Ontario, while the vast majority were sold in the United States (Capital Publishing, at para. 3). In the present case, the majority of copies were indeed distributed in Quebec, but the number of copies available in Ontario remains substantial. Moreover, there is evidence that Écosociété actively promoted its book in Ontario, unlike the situation where a copy of the libellous material is accessed by happenstance in the forum of choice.
Second, and as the English experience demonstrates, the substantial publication requirement provides both courts and litigants with little guidance (see D. Price, K. Duodo and N. Cain, Defamation: Law, Procedure & Practice (4th ed. 2009), at p. 448). One can easily imagine a publication, such as a bestselling novel, being substantially published in more than one jurisdiction, in which case, the problem of forum shopping and the multiplicity of jurisdictions would remain.
More fundamentally, however, the use of the substantial publication requirement in England reflects England’s merits-based approach to the assumption of jurisdiction, which is arguably inconsistent with the Canadian approach of treating jurisdiction separately from the merits of a claim. The defamation law of Canada has not adopted the substantial publication standard. In Canada, the evidentiary standard for proving publication remains the traditional common law standard, according to which a single instance of publication is sufficient for the tort to crystallize. To adopt the standard of substantial publication in the context of private international law would amount to a significant change in the substantive tort. It would be anomalous to adopt a new standard in the context of private international law but to continue applying the traditional standard in the context of the substantive tort.
While the defendants’ approach cannot be accepted, the question of whether the lex loci delicti represents the proper rule for choice of law in defamation remains. Although I need not decide the question in this case, I note that one possible alternative to the lex loci delicti in defamation cases, which has gained some significant support, may be the place of most substantial harm to reputation.
It is well settled in Canadian law that the tort of defamation occurs upon publication to a third party – that is, when the allegedly defamatory material is read or downloaded by someone other than the plaintiff or the publisher. On the other hand, it is also clear that the harm occasioned by the publication of a defamatory statement is not the publication itself, but rather injury to the plaintiff’s reputation. While the constitutional right to the protection of freedom of expression must be upheld in the crafting of the law of defamation, this Court has recognized that one of the primary purposes of the law of defamation is to protect the reputation of the individual, which was elevated to quasi-constitutional status in Hill v. Church of Scientology of Toronto,  2 S.C.R. 1130 (paras. 120-21):
Although it is not specifically mentioned in the Charter, the good reputation of the individual represents and reflects the innate dignity of the individual, a concept which underlies all the Charter rights. It follows that the protection of the good reputation of an individual is of fundamental importance to our democratic society.
Further, reputation is intimately related to the right to privacy which has been accorded constitutional protection .... The publication of defamatory comments constitutes an invasion of the individual’s personal privacy and is an affront to that person’s dignity.
The importance of place of reputation has long been recognized in Canadian defamation law. For example, the importance of permitting plaintiffs to sue for defamation in the locality where they enjoy their reputation was recognized by the Ontario High Court in Jenner v. Sun Oil Co. Ltd.,  2 D.L.R. 526. In that case, McRuer C.J.H.C. found that the plaintiff would not be able to satisfactorily “clear his good name of the imputation made against him” other than by suing for defamation in the locality where he enjoyed his reputation – that is, where he lived and had his place of business and vocation in life (pp. 538 and 540).
The approach adopted by McRuer C.J.H.C. is consonant with Australia’s recently enacted Defamation Act 2005 (Qld.). Prior to 2005, the choice of law rule for the tort of defamation in Australia was lex loci delicti, as in Canada. In 2004, acting on the recommendations of the Australian Law Reform Commission, the Attorneys General of Australia’s States and Territories agreed to enact model provisions which included a defamation-specific choice of law rule. In cases where a matter is published in more than one Australian jurisdictional area, the rule establishes that the applicable law is that of the jurisdictional area most closely connected to the harm occasioned by the publication as a whole. In determining which jurisdictional area has the closest connection with the harm, courts may take the following factors into account:
(Defamation Act 2005, s. 11(3))
The Australian Law Reform Commission’s recommendations were motivated by a concern that applying the lex loci delicti rule to such claims would encourage forum shopping (Law Reform Commission, Report No. 11 Unfair Publication: Defamation and Privacy (1979), at p. 190-191). Concern about “forum shopping” has increased in recent years, not only in Australia, but also in England, the United States and Canada. There have been calls for the adoption in this country of an approach similar to the Australian one. In particular, if Canadian courts are not ready to accept the proper law of the tort as a rule of conflicts in defamation cases, Professor Castel has suggested the following:
For choice of law purposes, the tort of defamation should be deemed to be committed where the plaintiff suffered the most injury to his or her reputation, that is, where substantial damage occurred. Only one law would be relevant. For jurisdiction purposes, the plaintiff should be given a wide choice depending upon the circumstances and provided that the court hearing the case applies the proper law and not its own law as a matter of principle.
(J.-G. Castel, “Multistate Defamation: Should the Place of Publication Rule Be Abandoned for Jurisdiction and Choice of Law Purposes?” (1990), 28 Osgoode Hall L.J. 153, at p. 177 (emphasis in original); see also C. Martin, “Tolofson and Flames in Cyberspace: The Changing Landscape of Multistate Defamation” (1997), 31 U.B.C. L. Rev. 127, at pp. 149 and 158)
It should be emphasized that this proposal would not result in a change to the substantive tort of defamation. Rather, the approach already adopted in Australia and recommended by Professor Castel appears to reflect the view that when it occurs at a multistate level, the elements of the tort of defamation play different roles when the rules of jurisdiction are applied than they do when the rules of choice of law are applied. In Professor Castel’s opinion, “rules of jurisdiction and of choice of law address different concerns and .... the test of place of publication should not always be used for both purposes” (p. 154).
In the case at bar, whether we apply the lex loci delicti rule or consider the location of the most substantial harm to reputation, the applicable law is that of Ontario and not Quebec. As a result, whichever approach is adopted, this factor favours Ontario in the forum non conveniens analysis. In this case, nothing turns on the question of whether lex loci delicti ought to be abandoned as the choice of law rule in multijurisdictional defamation cases. For this reason, I believe it prudent to leave this issue for another day.
(d) Juridical Advantage
The parties have also raised issues of juridical advantage. On the one hand, the appellants submit that they would be deprived of the procedural advantages of the new anti-SLAPP provisions in the Quebec Code of Civil Procedure, R.S.Q., c. C-25 (art. 54.1 to 54.6). On the other hand, in the event that the action is stayed in Ontario and transferred to Quebec, Banro might face an argument that its claim is barred on the basis of the short one-year limitation period for defamation claims under art. 2929 of the Civil Code of Québec. Arguments about which law would govern the civil liability of Écosociété could also be raised under s. 3126 of the Civil Code of Québec and would have to be resolved by the courts of Quebec. In the end, when these factors are weighed, the balance of fairness would appear to favour Banro and I find no error, in this respect, in the judgments rendered by the courts below.
Moreover, although this claim has connections to more than one forum, given the strength of the connections between the plaintiff and Ontario, it is not at all clear that the plaintiff is engaged in libel tourism and that Quebec would be a clearly more appropriate forum.
For the reasons set out above, I would dismiss the appeal with costs.
William C. McDowell, Yashoda Ranganathan and William Amos (m/s Lenczner Slaght Royce Smith Griffin, Toronto), for the appellants.
Lorne Honickman and Rory Barnable (m/s McCague Borlack, Toronto), for the respondent.
Karim Renno, Karine Chênevert and Fady Hammal (m/s Osler, Hoskin & Harcourt, Montréal), for the intervener the Canadian Civil Liberties Association.
Jason B. Gratl and Robert D. Holmes, Q.C. (m/s Holmes & King, Vancouver), for the intervener the British Columbia Civil Liberties Association.
all rights reserved