Haaretz.com, Haaretz Daily Newspaper Ltd., Haaretz Group, Haaretz.co.il, Shlomi Barzel v. David Marouani, 2018 SCC 28

SUPREME COURT OF CANADA Citation: Haaretz.com v. Goldhar, 2018 SCC 28, [2018] 2 S.C.R. 3 Appeal Heard: November 29, 2017 Judgment Rendered: June 6, 2018 Docket: 37202 Between: Haaretz.com, Haaretz Daily Newspaper Ltd., Haaretz Group, Haaretz.co.il, Shlomi Barzel and David Marouani Appellants and Mitchell Goldhar Respondent - and - Samuelson-Glushko Canadian Internet Policy and Public Interest Clinic Intervener Coram: McLachlin C.J. and Abella, Moldaver, Karakatsanis, Wagner, Gascon, Côté, Brown and Rowe JJ.

Reasons: (paras. 1 to 98) Côté J. (Brown and Rowe JJ. concurring) Concurring Reasons: (paras. 99 to 103) Karakatsanis J. Concurring Reasons: (paras. 104 to 143) Abella J. Concurring Reasons: (paras. 144 to 150) Wagner J. Joint Dissenting Reasons: (paras. 151 to 240) McLachlin C.J. and Moldaver and Gascon JJ.

Haaretz.com v. Goldhar, 2018 SCC 28, [2018] 2 S.C.R. 3 Haaretz.com, Haaretz Daily Newspaper Ltd., Haaretz Group, Haaretz.co.il, Shlomi Barzel and David Marouani Appellants v. Mitchell Goldhar Respondent and Samuelson-Glushko Canadian Internet Policy and Public Interest Clinic Intervener Indexed as: Haaretz.com v. Goldhar 2018 SCC 28 File No.: 37202. 2017: November 29; 2018: June 6.

Present: McLachlin C.J. and Abella, Moldaver, Karakatsanis, Wagner, Gascon, Côté, Brown and Rowe JJ. on appeal from the court of appeal for ontario Private international law — Choice of forum — Court having jurisdiction — Forum non conveniens — Libel action commenced in Ontario in respect of statements published in Israeli newspaper available electronically in Canada — Defendants bringing motion to stay action on grounds that Ontario court lacks jurisdiction or, alternatively, that Ontario court should decline to exercise its jurisdiction on basis of forum non conveniens — Whether situs of tort is reliable basis on which to presume real and substantial connection between chosen forum and subject matter of litigation in Internet defamation cases — If so, whether presumption of jurisdiction can be rebutted — Whether choice of law factor in forum non conveniens analysis for Internet defamation cases should be based on place where plaintiff suffered most substantial harm to reputation.

G is a prominent Canadian businessman who also owns one of the most popular professional soccer teams in Israel. H is Israel’s oldest daily newspaper, which is published in print and online. H published an

article about G, which the latter alleges to be libellous. The main subject of the

article is G’s ownership and management of his Israeli soccer team, but the

article also references his Canadian business and his approach to management. While the

article was not distributed in print form in Canada, it was available electronically. G commenced an action for libel in Ontario, alleging damage to his reputation. H brought a motion to stay the action, arguing that Ontario courts lacked jurisdiction or, alternatively, that Israel was a clearly more appropriate forum. The motion judge dismissed H’s motion, finding that Ontario courts had jurisdiction and refusing to decline to exercise this jurisdiction in favour of Israeli courts. A majority of the Ontario Court of Appeal dismissed H’s appeal.

Held (McLachlin C.J. and Moldaver and Gascon JJ. dissenting): The appeal should be allowed and the motion to stay the action granted. Per Côté, Brown and Rowe JJ.: While multijurisdictional defamation claims are not new, the exponential increase in multijurisdictional publications over the Internet has led to growing concerns about libel tourism and the possible assumption of jurisdiction by an unlimited number of forums. The current rules for the assumption and exercise of jurisdiction are able to address these challenges so long as the underlying principles of stability and fairness are kept in mind.

In this case, while the motion judge properly determined that he had jurisdiction (under the jurisdiction simpliciter test), he committed multiple errors in his forum non conveniens analysis. On a robust and careful assessment of the relevant factors tainted by these errors, Israel is a clearly more appropriate forum. Central to a proper understanding of the conflicts rules of Canadian private international law is an appreciation of the distinct roles played by jurisdiction simpliciter and forum non conveniens , and how these must be understood and analysed as a cohesive whole.

The jurisdiction simpliciter analysis is meant to ensure that a court has jurisdiction. This will be the case where a real and substantial connection exists between a chosen forum and the subject matter of the litigation. This test prioritizes order, stability and predictability by relying on objective connecting factors for the assumption of jurisdiction. The forum non conveniens analysis, on the other hand, is meant to guide courts in determining whether they should decline to exercise that jurisdiction in favour of a clearly more appropriate forum.

This doctrine emphasizes fairness and efficiency by adopting a case-by-case approach to this question. At the jurisdiction simpliciter stage, in determining whether a real and substantial connection exists between a chosen forum and the subject matter of the litigation, a court must first consider whether the existence of a recognized presumptive connecting factor has been established. The situs of the tort, which is one such recognized factor, is a reliable basis on which to presume a real and substantial connection, even in Internet defamation cases.

Raising doubt as to the value of the situs of the tort as a presumptive

connecting factor in such cases, because of the ease with which publication can be established, would significantly undermine the objectives of predictability and order at the jurisdiction simpliciter stage. Concerns relating to the insufficiency of a presumptive connecting factor should be addressed either at the rebuttal stage of the jurisdiction simpliciter analysis or during the forum non conveniens analysis. In this case, the tort of defamation was committed in Ontario, and therefore a presumptive connecting factor has been established.

As a result, the Court must consider whether H has successfully rebutted the presumption. The ability to rebut the presumption of jurisdiction where there is only a weak relationship between the subject matter of the litigation and the forum serves as an important check on jurisdiction. A careful examination of this question is therefore of particular importance in Internet defamation cases, where a presumptive connecting factor can easily be established. Presumptive connecting factors must not give rise to an irrebuttable presumption of jurisdiction.

In order for a defendant to succeed in challenging jurisdiction, the circumstances must demonstrate that the relationship between the forum and the subject matter of the litigation is such that it would not be reasonable to expect that the defendant would be called to answer proceedings in that jurisdiction. Assuming that these principles are properly applied, the situs of the tort will not give rise to an irrebuttable presumption of jurisdiction in Internet defamation cases. In the case at bar, H could have reasonably expected to be called to answer a legal proceeding in Ontario.

As such, the presumption of jurisdiction is not rebutted. At the forum non conveniens stage, the burden is on the defendant to satisfy the motion judge that the alternative forum is clearly more appropriate. While the normal state of affairs favours exercising jurisdiction in the forum where it is properly assumed, this should never come at the cost of one party facing unfair or clearly inefficient proceedings. Given the ease with which jurisdiction may be established in a defamation case, in a motion for a stay, a judge must conduct a robust and carefully scrutinized review of the issue of forum non conveniens .

The establishment of a presumptive connecting factor is virtually automatic in Internet defamation cases. As the rebuttal stage of the jurisdiction simpliciter analysis fails to address all the consequences of this fact, it is appropriate for motion judges to be particularly attuned to concerns about fairness and efficiency during the forum non conveniens analysis in these types of cases. This should not be understood as imposing a different standard or burden for defamation cases.

As the forum non conveniens analysis is inherently factual in nature, courts of appeal should not normally interfere with a motion judge’s factual findings. However, there are limits to deference. Where the motion judge has erred in principle, misapprehended or failed to take account of material evidence, or reached an unreasonable decision, courts of appeal may intervene. In the case at bar, the motion judge committed several errors, which tainted his forum non conveniens analysis on each of the factors they affected as well as his overall weighing of these factors.

As a result, no deference should be afforded to these aspects of the motion judge’s analysis and the Court may intervene. Ultimately, H has established that holding a trial in Israel would be fairer and more efficient. Israel is clearly the more appropriate forum. A robust and careful forum non conveniens analysis of the relevant factors indicates that H would face substantial unfairness and inefficiency if a trial were held in Ontario. Comparative convenience and expense for the parties and comparative convenience and expense for the witnesses favour Israel.

Loss of legitimate juridical advantage favours a trial in Ontario, but this factor should not weigh too heavily in the analysis. Fairness favours Israel, namely in view of G’s significant business interest and reputation in that country and the significant unfairness that a trial in Ontario would impose on H. Enforcement slightly favours Israel as H has no presence or assets in Ontario.

Finally, while applicable law, as determined by the lex loci delicti principle — the place where the tort occurs —, favours Ontario in this case, this factor 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. In those circumstances, lex loci delicti will inevitably also point to the chosen forum on the question of applicable law. This would not be an appropriate case to adopt the place of most substantial harm to reputation test for choice of law instead of lex loci delicti .

Although in Internet defamation actions, where a tort may have occurred in multiple jurisdictions, the lex loci delicti rule may allow courts in multiple forums to assume jurisdiction and apply their own law, the Court should be reluctant to make such changes to the existing private international law framework, as this may create legal uncertainty in a manner contrary to the objectives of conflicts rules. Per Karakatsanis J.: There is agreement with Côté J.’s conclusion and much of her reasoning. However, there is disagreement with two aspects of her analysis relating to forum non conveniens .

When considering the applicable law factor, assessing what law would apply in the alternative jurisdiction is not helpful, as the ultimate question that motivates this factor is whether the plaintiff’s chosen jurisdiction would be applying foreign law. Further, G’s Israeli reputation is not material to the fairness factor, which is concerned with the plaintiff’s interest in vindicating his reputation in the jurisdiction where he enjoys it. Ultimately, the overall conclusion reached by Côté J. on forum non conveniens does not turn on any of these elements, and therefore, the appeal should be allowed.

Per Abella J.: There is agreement with Côté J. that the appeal should be allowed. However, there is disagreement with her that the lex loci delicti rule should continue to serve as the basis for choice of law under the forum non conveniens analysis in cases of multijurisdictional Internet defamation. This standard approach to choice of law does not adequately respond to the unique issues and challenges raised by Internet defamation, where a single download can determine which law applies under a strict application of the lex loci delicti rule.

The framework for choice of law should therefore be modified by replacing lex loci delicti with a test based on the place where the most substantial harm to the plaintiff’s reputation occurred. This new approach would narrow the range of potentially applicable law in a rational way and would displace the law of the place of publication of the defamation with the law of the place with the most significant connection to the tort.

It would also ensure that the choice of law rule reflects protection of reputation, which is at the core of the tort of defamation, and that the reasonable expectations of the publisher of the statement alleged to be defamatory as to where it could expect to be sued are properly considered, while at the same time striking a better balance between freedom of expression and harm to reputation. Since there are symmetrical concerns between how the choice of law analysis proceeds and how jurisdiction is determined in Internet defamation cases, the same approach should be applied to determining jurisdiction.

The current approach seems to make the

assumption of jurisdiction automatic based on a single download. Since the essence of the harm in defamation is damage to reputation, the framework for determining jurisdiction should focus on where the plaintiff suffered the most substantial harm to his or her reputation. Such an approach allows the presumption of jurisdiction to be rebutted if the defendant can show that the most harm to the plaintiff’s reputation occurred elsewhere. Adopting the most substantial harm test for determining the choice of law under the forum non conveniens analysis, the place of most substantial harm to G’s reputation is clearly Israel, and as a result, Israeli law should apply. The

article in question is essentially about G and his conduct in Israel: it was about G’s soccer team, one of Israel’s most popular soccer teams, G’s involvement in his team’s management, and G’s relationship with his players, coaches and trainers in Israel. It was researched, written and edited in Israel, addressed to an Israeli audience, and focused on someone who is a public figure there. Although G spends most of his time in Canada, he maintains an apartment in Israel and his connection to Israel is significant. Accordingly, the

article would have a far greater impact on his reputation in Israel than in Canada. As for the rest of the forum non conveniens analysis, on the basis that Israeli law applies, there is agreement with Côté J. that Israel is the clearly more appropriate forum. All of the remaining factors — the comparative convenience/expense to the parties and witnesses, juridical advantage, fairness and enforcement — favour Israel. Per Wagner J.: There is agreement with Côté J. that the appeal should be allowed.

However, as set out in the reasons of Abella J., the choice of law rule during the forum non conveniens analysis should be modified for the tort of Internet defamation, from lex loci delicti to a test based on the place where the most substantial harm to the plaintiff’s reputation occurred. Although it may be that in certain cases it would be challenging to identify the place of most substantial reputational harm, the range of possibly applicable law for a given dispute would be much narrower than with lex loci delicti and would be determined on a more principled basis.

Adopting this new test for choice of law would have several positive effects and would not result in a heavy evidentiary burden for the parties. With respect to the jurisdiction simpliciter analysis, a Canadian court should not conclude that it does not have jurisdiction over a dispute with significant connections to Canada, including potentially significant reputational harm suffered in Canada, simply because greater reputational harm occurred elsewhere.

As a result, concerns raised by the unique nature of Internet defamation are best addressed by changes to the choice of law rule, rather than by changes to the jurisdiction simpliciter stage of the analysis. The inquiry at that stage is simply whether there is a real and substantial connection between the dispute and the Canadian forum, not whether this connection is greater than that between the dispute and any other forum. There is no reason why this should be different in the context of Internet defamation.

In this case, when the most substantial harm test is applied to the facts, Israel is the clearly more appropriate forum. Per McLachlin C.J. and Moldaver and Gascon JJ. (dissenting): The appeal should be dismissed. When a Canadian citizen is allegedly defamed for his Canadian business practices — in an

article published online in his home province by a foreign newspaper — he is entitled to vindicate his reputation in the courts of the province where he lives and maintains his business, and where the sting of the article’s comments is felt. The current rules that govern the application of the test for jurisdiction simpliciter readily accommodate multijurisdictional defamation cases, even in the Internet age.

The commission of a tort in the jurisdiction remains a sound presumptive connecting factor on which to establish prima facie jurisdiction even in the context of Internet defamation cases, because the sting of the defamation is felt in the place where it is read. In this case, it is not contested that the allegedly libellous

article was consulted by 200 to 300 people in Canada; therefore, a tort of defamation was committed in Ontario. There is no valid reason to reconsider or set aside this clearly established presumptive connecting factor. While a presumptive connecting factor may be established virtually automatically in Internet defamation cases, a court does not necessarily assume jurisdiction. If there is no real and substantial connection between the action and the forum, the presumptive connecting factor would be rebutted.

Reasonable foreseeability is central to the rebuttal step of the analysis: the strength of the relationship between the subject of the litigation and the forum is informed by the reasonable foreseeability of the claim proceeding in that jurisdiction. Without this important check of reasonable foreseeability of being sued in the jurisdiction, the presumptive connecting factor of the commission of a tort in the jurisdiction could raise concerns of forum shopping.

Reasonable foreseeability is therefore an important limit on the ease with which jurisdiction can be presumptively assumed in defamation cases, especially over the Internet. In the present case, it was more than reasonably foreseeable that H would be sued in Ontario. The

article was highly critical of G’s management style, allegedly imported from his Canadian business. Furthermore, H made the

article readily available to readers worldwide through online publication. It is entirely foreseeable that a Canadian citizen and resident would want to vindicate his Canadian reputation as the owner of his Canadian businesses in a Canadian court. Therefore, the presumption of jurisdiction was not rebutted and Ontario courts have jurisdiction. The facts undeniably reveal a real and substantial connection between this case and Ontario.

If the analysis at the rebuttal stage is done properly, with an adequate consideration of reasonable foreseeability, there is no need to apply a robust and carefully scrutinized forum non conveniens analysis, as suggested by Côté J. This new standard would frustrate the predictability and stability that is at the core of the applicable framework. The basis of the forum non conveniens analysis is the clearly more appropriate forum test, which sets a high threshold for displacing the forum chosen by the plaintiff.

This purposefully stringent and consistently upheld threshold should not be lowered, whether through lenient application or through a robust and carefully scrutinized review. Furthermore, a motion judge’s discretionary decision whether or not to decline jurisdiction on the basis of forum non conveniens is entitled to considerable deference, and having appellate courts apply the proposed robust and carefully scrutinized approach would disregard the discretionary nature of forum non conveniens decisions.

A motion judge’s exercise of his discretionary power or assessment of the evidence should not be interfered with where it is not tainted by any error, or when only tainted by errors that have no impact on the result. In this case, an assessment of the factors in the forum non conveniens analysis indicates that they do not meet the test of showing that Israel is a clearly more appropriate forum than Ontario. Only the factor of comparative convenience and expense for the

parties and witnesses favours Israel, and this only slightly so with respect to the witnesses. The enforcement of judgment factor does notweigh heavily in the analysis. The factor of loss of legitimate juridical advantage weighs in favour of Ontario, and, most importantly, thekey factors of applicable law and fairness to the parties weigh heavily in favour of Ontario. With respect to the applicable law factor, the most substantial harm test to determine the applicable law in multijurisdictionalInternet defamation cases should not be adopted in place of lex loci delicti.

Such a rule is highly subjective, and will not reliably point toone jurisdiction. It does not provide a clear answer where a person lives and maintains an important reputation in one jurisdiction, butacts — and is the subject of defamatory statements — in another jurisdiction. It would also lead to complex preliminary motionsrequiring substantial evidence which would increase delay and expense.

In terms of the proper approach to balancing this factor in theforum non conveniens analysis, it is entirely appropriate for courts to only look at the chosen forum in determining the applicable law.Requiring courts to assess the choice of law rules of a foreign jurisdiction may require extensive evidence, needlessly complicating thepre-trial motion stage of the proceedings. Where jurisdiction is based on the situs of the tort, the applicable law under lex loci delicti willindeed point to the forum.

This does not mean that the applicable law factor should be granted little weight in the forum non conveniensanalysis; rather, giving due weight to this factor reflects the notion that a case should proceed in a forum that properly has jurisdictionover the matter unless another forum is clearly more appropriate. Holding that the applicable law should be given little weight ignoresthe importance of the territorial jurisdiction of the chosen forum, and distorts the forum non conveniens analysis in favour of the foreignjurisdiction.

The lex loci delicti rule directs courts to apply their domestic law after having found that the tort of defamation occurredwithin their jurisdiction. Defamation law is directed to the protection of reputation. For choice of law purposes, it is therefore logical thata court of a jurisdiction where publication occurred is entitled to apply its own law. This remains true even if a tort took placesimultaneously in another jurisdiction. In this case, as the applicable law is that of Ontario, this factor strongly favours Ontario overIsrael.

With respect to fairness, this factor, along with the efficient resolution of disputes, is the cornerstone of forum nonconveniens. The Court has repeatedly emphasized the importance of plaintiffs being allowed to sue for defamation in the locality wherethey enjoy their reputation. In the instant case, G has a real and long-standing reputational interest in Ontario. His reputation in Israel isnot material to the analysis. It is therefore not unfair that Ontario be the forum deciding the dispute.

Following the forum non conveniens analysis, Israel has not emerged as a forum that would be more appropriate thanOntario to hear the case, much less a clearly more appropriate forum. Cases Cited By Côté J. Applied: Club Resorts Ltd. v. Van Breda, 2012 SCC 17, [2012] 1 S.C.R. 572; Tolofson v. Jensen, (SCC),[1994] 3 S.C.R. 1022; considered: Éditions Écosociété Inc. v. Banro Corp., 2012 SCC 18, [2012] 1 S.C.R. 636; Moore v. Bertuzzi, 2014ONSC 1318, 53 C.P.C. (7th) 237; referred to: Breeden v.

Black, 2012 SCC 19, [2012] 1 S.C.R. 666; Lax Kw’alaams Indian Band v.Canada (Attorney General), 2011 SCC 56, [2011] 3 S.C.R. 535; Muscutt v. Courcelles (2002), (ON CA), 60 O.R.(3d) 20; Charron Estate v. Village Resorts Ltd., 2010 ONCA 84, 98 O.R. (3d) 721; Hunt v. T&N plc, (SCC), [1993] 4S.C.R. 289; Crookes v. Newton, 2011 SCC 47, [2011] 3 S.C.R. 269; Crookes v. Holloway, 2007 BCSC 1325, 75 B.C.L.R. (4th) 316,aff’d 2008 BCCA 165, 77 B.C.L.R. (4th) 201; Barrick Gold Corp. v. Blanchard & Co. (2003), (ON SC), 9 B.L.R.(4th) 316. By Karakatsanis J. Referred to: Éditions Écosociété Inc. v.

Banro Corp., 2012 SCC 18, [2012] 1 S.C.R. 636; Breeden v. Black, 2012 SCC 19,[2012] 1 S.C.R. 666. By Abella J. Considered: Tolofson v. Jensen, (SCC), [1994] 3 S.C.R. 1022; Club Resorts Ltd. v. Van Breda, 2012 SCC17, [2012] 1 S.C.R. 572; referred to: Éditions Écosociété Inc. v. Banro Corp., 2012 SCC 18, [2012] 1 S.C.R. 636; eDate AdvertisingGmbH v. X, C-509/09, C-161/10, [2011] E.C.R. I-10302; Beals v. Saldanha, 2003 SCC 72, [2003] 3 S.C.R. 416; Moran v. Pyle National(Canada) Ltd., (SCC), [1975] 1 S.C.R. 393; Morguard Investments Ltd. v.

De Savoye, (SCC), [1990] 3S.C.R. 1077; Lapointe Rosenstein Marchand Melançon LLP v. Cassels Brock & Blackwell LLP, 2016 SCC 30, [2016] 1 S.C.R. 851. By Wagner J. Considered: Éditions Écosociété Inc. v. Banro Corp., 2012 SCC 18, [2012] 1 S.C.R. 636; referred to: Club Resorts Ltd. v.Van Breda, 2012 SCC 17, [2012] 1 S.C.R. 572. By McLachlin C.J. and Moldaver and Gascon JJ. (dissenting) Club Resorts Ltd. v. Van Breda, 2012 SCC 17, [2012] 1 S.C.R. 572; Lax Kw’alaams Indian Band v. Canada (AttorneyGeneral), 2011 SCC 56, [2011] 3 S.C.R. 535; Morguard Investments Ltd. v. De Savoye, (SCC), [1990] 3 S.C.R. 1077;Breeden v.

Black, 2012 SCC 19, [2012] 1 S.C.R. 666; Éditions Écosociété Inc. v. Banro Corp., 2012 SCC 18, [2012] 1 S.C.R. 636;Paulsson v. Cooper, 2011 ONCA 150, 105 O.R. (3d) 28; Barrick Gold Corp. v. Blanchard & Co. (2003), (ON SC),9 B.L.R. (4th) 316; Lapointe Rosenstein Marchand Melançon LLP v. Cassels Brock & Blackwell LLP, 2016 SCC 30, [2016] 1 S.C.R.851; Amchem Products Inc. v. British Columbia (Workers’ Compensation Board), (SCC), [1993] 1 S.C.R. 897; Egbertv. Short, [1907] 2 Ch. 205; St. Pierre v.

South American Stores (Gath and Chaves), Limited, [1936] 1 K.B. 382; Rockware Glass Ltd. v.MacShannon, [1978] 2 W.L.R. 362; Spiliada Maritime Corporation v. Cansulex Ltd., [1987] 1 A.C. 460; Spar Aerospace Ltd. v.American Mobile Satellite Corp., 2002 SCC 78, [2002] 4 S.C.R. 205; Unifund Assurance Co. v. Insurance Corp. of British Columbia,

2003 SCC 40, [2003] 2 S.C.R. 63; Tolofson v. Jensen, (SCC), [1994] 3 S.C.R. 1022; Bou Malhab v. DiffusionMétromédia CMR inc., 2011 SCC 9, [2011] 1 S.C.R. 214; Jenner v. Sun Oil Co., (ON SC), [1952] 2 D.L.R. 526. Statutes and Regulations Cited Defamation Act 2005 (N.S.W.), s. 11(3). Rules of Civil Procedure, R.R.O. 1990, Reg. 194, rr. 1.08, 47.01, Tariff A. Authors Cited Australia. Law Reform Commission. Unfair Publication: Defamation and Privacy. Canberra, 1979. Blom, Joost, and Elizabeth Edinger. “The Chimera of the Real and Substantial Connection Test” (2005), 38 U.B.C. L. Rev. 373.

Brown on Defamation: Canada, United Kingdom, Australia, New Zealand, United States, 2nd ed. by Raymond E. Brown. Toronto:Carswell, 1994 (loose-leaf updated 2017, release 5). Canadian Encyclopedic Digest, vol. 10, Ontario 4th ed. Toronto: Carswell, 2009 (loose-leaf updated 2018, release 2). Castel, J.-G. “Multistate Defamation: Should the Place of Publication Rule be Abandoned for Jurisdiction and Choice of Law Purposes?”(1990), 28 Osgoode Hall L.J. 153. Castel, Jean-Gabriel. “The Uncertainty Factor in Canadian Private International Law” (2007), 52 McGill L.J. 555.

Castel, Matthew. “Jurisdiction and Choice of Law Issues in Multistate Defamation on the Internet” (2013), 51 Alta. L. Rev. 153. Cruz Villalón, Pedro. Opinion of Advocate General Cruz Villalón, C-509/09, C-161/10, [2011] E.C.R. I-10272. Dicey, Morris and Collins on the Conflict of Laws, 15th ed. by Lord Collins of Mapesbury. London: Sweet & Maxwell/ThomsonReuters, 2012. Downard, Peter A. The Law of Libel in Canada, 4th ed. Toronto: LexisNexis Canada, 2018. Kain, Brandon, Elder C.

Marques and Byron Shaw. “Developments in Private International Law: The 2011-2012 Term — TheUnfinished Project of the Van Breda Trilogy” (2012), 59 S.C.L.R. (2d) 277. Law Commission of Ontario. Defamation Law in the Internet Age: Consultation Paper, Toronto, 2017 (online: http://www.lco-cdo.org/wp-content/uploads/2017/12/Defamation-Consultation-Paper-Eng.pdf; archived version: http://www.scc-csc.ca/cso-dce/2018SCC-CSC28_1_eng.pdf). Martin, Craig. “Tolofson and Flames in Cyberspace: The Changing Landscape of Multistate Defamation” (1997), 31 U.B.C. L. Rev. 127. Pitel, Stephen G. A., and Nicholas S. Rafferty.

Conflict of Laws, 2nd ed. Toronto: Irwin Law, 2016. Schmitz, Sandra. “From Where are They Casting Stones? — Determining Jurisdiction in Online Defamation Claims” (2012), 6 MasarykU. J.L. & Tech. 159. APPEAL from a judgment of the Ontario Court of Appeal (Simmons, Cronk and Pepall JJ.A.), 2016 ONCA 515, 132 O.R.(3d) 331, 349 O.A.C. 132, 401 D.L.R. (4th) 634, [2016] O.J. No. 3471 (QL), 2016 CarswellOnt 10242 (WL Can.), affirming a decisionof Faieta J., 2015 ONSC 1128, 125 O.R. (3d) 619, [2015] O.J. No. 1084 (QL), 2015 CarswellOnt 3080 (WL Can.). Appeal allowed,McLachlin C.J. and Moldaver and Gascon JJ. dissenting.

Paul B. Schabas, Kaley Pulfer and Brittiny Rabinovitch, for the appellants. William C. McDowell, Julian Porter, Q.C., and Brian Kolenda, for the respondent. Jeremy de Beer, Marina Pavlović and David Fewer, for the intervener. The reasons of Côté, Brown and Rowe JJ. were delivered by Côté J. — I. Introduction [1] This appeal has to do with the rules for the assumption and exercise of jurisdiction in the context ofmultijurisdictional defamation claims.

While these types of claims are not new, the exponential increase in multijurisdictionalpublications over the Internet has led to growing concerns about libel tourism and the possible assumption of jurisdiction by an unlimitednumber of forums. [2] For the reasons set out below, I find that the current rules are able to address these challenges so long as theunderlying principles of stability and fairness are kept in mind. [3] While the motion judge in this case properly determined that he had jurisdiction (under the jurisdiction simplicitertest), he committed multiple errors in his forum non conveniens analysis.

On a robust and careful assessment of the relevant factors

tainted by these errors, I conclude that Israel is a clearly more appropriate forum for this claim to be heard. [ 4 ] The appeal should be allowed. II. Background and Facts [ 5 ] The respondent, Mitchell Goldhar, is a prominent Canadian businessman who owns and operates SmartCentres Inc. in Ontario. He also owns the Maccabi Tel Aviv Football Club (“Maccabi Tel Aviv”), one of the most popular professional soccer teams in Israel.

Goldhar, who has been described as a celebrity in Israel, maintains a residence there and travels there every few months. [ 6 ] The corporate appellants publish Israel’s oldest daily newspaper in both English and Hebrew, in print and online. It has a distribution of about 70,000 print copies in Israel. The individual appellants are, respectively, the newspaper’s former sports editor and the author of the

article alleged to be libellous. Collectively, the appellants are referred to as “Haaretz”. [ 7 ] On November 29, 2011, Haaretz published an

article about Goldhar, which the latter alleges to be libellous. The main subject of the

article is Goldhar’s ownership and management of Maccabi Tel Aviv. That being said, it also references his Canadian business and his approach to management, as follows: Though he spends most of his time in Canada, Maccabi Tel Aviv owner Mitch Goldhar runs his club down to every detail. But could his penny pinching and lack of long term planning doom the team. . . . Crises are par for the course at Maccabi Tel Aviv, even when the club appears to be on an even keel.

Most of the crises don’t make it onto the public’s radar, but they have one thing in common: their connection to [ sic ] way that Canadian owner Mitch Goldhar runs the club. . . . Goldhar’s management model was imported directly from his main business interest — a partnership with Wal-Mart to operate shopping centers in Canada. . . . Within the club, however, there are those who believe that Goldhar’s managerial culture is based on overconcentration bordering on megalomania, penny-pinching and a lack of long-term planning. . . .

Goldhar boasts to his business contacts in Toronto that he is not only the owner of Maccabi Tel Aviv but also its soccer director. (Reproduced in 2016 ONCA 515 , 132 O.R. (3d) 331, Appendix “A”.) The

article was researched, written and edited in Israel, primarily in reliance on Israeli sources. [ 8 ] The

article was published in print and electronically in Hebrew and English. While it was not distributed in print form in Canada, it was available electronically. The motion judge found it likely that 200 to 300 people in Canada read the article; by comparison, the evidence showed that approximately 70,000 people read the

article in Israel. Two affiants, both employed by SmartCentres Inc., have stated that they read the

article and that it came to the attention of most of their approximately 200 co-workers. There is no evidence that those who read the

article thought less of Goldhar as a result. [ 9 ] On December 29, 2011, Goldhar commenced an action for libel, alleging “damage to his reputation in his business and personal life”. His amended statement of claim states that “[t]he plaintiff conducts business in Israel, Canada and the United States, and will continue to suffer damages in these countries and elsewhere” (reproduced in A.R., vol. II, pp. 1-8, at para. 12). [ 10 ] Haaretz brought a motion to stay the action, arguing that Ontario courts lacked jurisdiction or, alternatively, that Israel was a clearly more appropriate forum.

[ 11 ] The motion judge dismissed the motion, finding that Ontario courts had jurisdiction and refusing to decline to exercise this jurisdiction in favour of Israeli courts. In doing so, he relied on two undertakings by Goldhar’s counsel. First, Goldhar would not seek damages at the trial of the action for reputational harm suffered in Israel or anywhere else outside of Canada.

Second, Goldhar would pay for the travel and accommodation expenses of Haaretz’s witnesses at the rates provided in the Rules of Civil Procedure , R.R.O. 1990, Reg. 194. [ 12 ] The majority of the Ontario Court of Appeal dismissed the appeal. III. Judicial History A.

Ontario Superior Court of Justice — 2015 ONSC 1128 , 125 O.R. (3d) 619, per Faieta J. (March 6, 2015) [ 13 ] The motion judge dismissed the motion to stay the action and added that, in the event that the action proceeded in Ontario, Goldhar’s claim would be limited to damages for reputational harm suffered within Canada and he would pay for the travel and accommodation expenses of Haaretz’s witnesses at the rates provided in the Rules. He also expressed the view that the lawsuit was far from being an abuse of process by Goldhar. [ 14 ] The motion judge found that he had jurisdiction.

The parties agreed that, as the evidence established that the

article had been read in Ontario, a presumptive connecting factor existed. The motion judge found that Haaretz had failed to rebut the presumption.

In particular, he did not view the absence of substantial publication of the libellous material in Ontario as rebutting the presumption, and he considered proof of harm to reputation irrelevant for the purposes of determining whether a minor element of the tort had occurred in Ontario. [ 15 ] Further, the motion judge refused to decline to exercise jurisdiction, finding that Israel was not a clearly more appropriate forum after weighing the following factors: • Comparative convenience and expense for the parties favoured a trial in Israel. The Haaretz defendants were all based in Israel.

Goldhar visited Israel regularly and there was no evidence that a trial in Israel would cause him inconvenience or expense. • Comparative convenience and expense for the witnesses slightly favoured a trial in Israel. Goldhar had not filed any evidence regarding the witnesses that he would call at trial. Haaretz proposed to call 22 witnesses, 18 of whom lived in Israel. However, the relevance of the testimony of some of Haaretz’s witnesses was questionable. Compelling the attendance of these witnesses in Ontario could be accomplished through the use of letters rogatory, also called letters of request.

Moreover, foreign witnesses could testify via videoconference. Finally, Goldhar’s undertaking to pay for the travel and accommodation expenses of Haaretz’s witnesses at the rates provided by the Rules addressed any additional expense. • Applicable law favoured a trial in Ontario. Regardless of which choice of law rule applied, the lex loci delicti (the place where the tort is committed) rule or the “most substantial harm to reputation” rule, Ontario law was applicable to this case. Ontario was the locus delicti of the tort of libel.

Further, there was no comparative evidence of reputational harm to Goldhar in Israel and Ontario as a result of the publication, and there was limited evidence regarding Goldhar’s reputation. In light of this evidence, Goldhar’s undertaking not to seek damages for reputational harm outside of Canada was a very significant factor which led to the conclusion that the most substantial harm to his reputation was in Ontario. • Loss of juridical advantage favoured a trial in Ontario.

The availability of a jury trial in Ontario was a juridical advantage of which Goldhar would be deprived if the case were tried in Israel. Any juridical advantage Goldhar might enjoy under Israeli defamation law was irrelevant, since the proper question was whether the plaintiff should be denied the benefits of his decision to select a forum that was appropriate under the conflicts rules. • Fairness to the parties favoured a trial in Ontario.

Given the importance of place of reputation in Canadian defamation law, there was no surprise or injustice to Goldhar’s attempt to vindicate his reputation in Ontario, where he lives and works. B. Ontario Court of Appeal — 2016 ONCA 515 , 132 O.R. (3d) 331, per Simmons and Cronk JJ.A. (Pepall J.A. Dissenting) (June 28, 2016) [ 16 ] The majority of the Court of Appeal dismissed Haaretz’s appeal. The majority was of the view that the motion judge had not erred in failing to find that Haaretz had successfully rebutted the presumption of jurisdiction. As the

article “draws a link between Goldhar’s management model and his Canadian business” (para. 41), there was a significant connection between the subject matter of the action and Ontario, and it should not have come as a surprise to Haaretz that Goldhar would seek to vindicate his reputation in Ontario. The question at the rebuttal stage of the jurisdiction simpliciter analysis was whether, objectively speaking, Ontario had a real and substantial connection to the subject matter of the action, not whether there was another forum that could also assume jurisdiction over the action.

In the absence of evidence demonstrating no reputational harm, evidence of actual reputational harm was not necessary to establish jurisdiction. [ 17 ] The majority also found no basis on which to interfere with the motion judge’s conclusion that Israel had not been shown to be a clearly more appropriate forum. Its analysis focused on the following factors: • It was reasonable for the motion judge to conclude that the convenience and expense for the witnesses slightly favoured a trial in Israel.

The motion judge erred in law by suggesting that letters rogatory could be used to compel the attendance of Israeli witnesses in Ontario.

Despite this error, the motion judge reasonably based his analysis on the availability of letters rogatory for compelling testimony from witnesses outside of Ontario via videoconferencing, Goldhar’s undertaking to fund foreign witnesses’ travel and accommodation expenses, and the lack of evidence concerning the likely testimony of Haaretz’s proposed witnesses. • The motion judge reasonably found that applicable law favoured Ontario irrespective of which choice of law rule was applied — the lex loci delicti rule or the “most substantial harm to reputation” rule. Haaretz’s reliance on the extent of publication in

Israel would turn the “most substantial harm” approach into a proxy for the “substantial publication” rule rejected in Éditions Écosociété Inc. v. Banro Corp. , 2012 SCC 18 , [2012] 1 S.C.R. 636. Furthermore, Goldhar’s undertaking not to seek damages for reputational harm sustained outside of Canada confirmed the significance to him of his reputation in Ontario and the importance to him of vindicating his reputation here. • While loss of juridical advantage was a neutral factor rather than a factor that favoured a trial in Ontario, this error was not significant in terms of the motion judge’s overall conclusion.

The motion judge erred in accepting that Goldhar would suffer a loss of juridical advantage. As he had failed to deliver a jury notice, Goldhar was not entitled to claim a loss of juridical advantage. However, the motion judge correctly found that potential juridical advantages to a plaintiff in the alternative forum are irrelevant to the forum non conveniens analysis. • There was no basis on which to interfere with the motion judge’s conclusion on the question of fairness. The motion judge considered it important that Goldhar lives and works in Ontario and that Haaretz chose to write an

article about him impugning his management of an Israeli soccer team in a manner that implicated his Canadian business practices. [ 18 ] Pepall J.A. agreed with the majority on the issue of the jurisdiction simpliciter test but would have allowed the appeal and stayed the action, finding that Israel was clearly the more appropriate forum. Given the ease with which jurisdiction may be established in a defamation case, she expressed the view that a “robust and carefully scrutinized review” of the issue of forum non conveniens was required (para. 132).

In light of the errors committed by the motion judge, as identified by the majority, the motion judge’s conclusion was unreasonable. Pepall J.A. weighed the following factors: • Comparative convenience and expense for the parties clearly and overwhelmingly supported a trial in Israel. There was no evidence of any inconvenience or undue expense for Goldhar associated with a trial in Israel. • Comparative convenience and expense for the witnesses overwhelmingly favoured a trial in Israel.

The motion judge’s error on letters rogatory, his failure to consider the purport of Tariff A of the Rules — providing for rates significantly below the actual cost of travel and accommodation — when dealing with Goldhar’s undertaking, and his failure to consider the fact that Goldhar had not identified any prospective witnesses all served to cause him to erroneously conclude that this factor only slightly favoured Israel. • Applicable law favoured a trial in Israel. As pleaded, the tort occurred in both Ontario and Israel. The most substantial harm test favoured a trial in Israel.

The evidence was that the extent of publication and any harm suffered were much more significant in that forum. Furthermore, the

article was written in Israel about an Israeli soccer team and was aimed at an Israeli audience. The motion judge erred by treating Goldhar’s undertaking to limit his claim for damages to Ontario as determinative. This undertaking was materially different than the one this Court considered in Breeden v. Black , 2012 SCC 19 , [2012] 1 S.C.R. 666. • At most, juridical advantage was a neutral factor. The motion judge erred in accepting that Goldhar would suffer a loss of juridical advantage with regard to the availability of a jury trial in Ontario. • Fairness clearly favoured a trial in Israel.

The motion judge focused on vindication of Goldhar’s reputation while failing to mention the burden a trial in Ontario would impose on Haaretz or the ability of a trial in Israel to achieve the vindication sought by Goldhar. • Enforcement favoured a trial in Israel. The motion judge said nothing about enforcement. The only evidence before him was that Haaretz had no assets in Ontario, whereas it could be inferred that Goldhar did have assets in Israel. IV. Issues [ 19 ] This appeal raises the following issues and sub-issues: 1. Did the motion judge err in assuming jurisdiction? (

a) Is the situs of the tort a reliable basis on which to presume a “real and substantial connection” in Internet defamation cases? (

b) Under what circumstances, if any, can the presumption of jurisdiction be rebutted? 2. Did the motion judge err in finding that Israel is not a clearly more appropriate forum than Ontario? Notably, should the “most substantial harm” test rather than the lex loci delicti test apply to determine the applicable law in defamation actions? V. Analysis A. The Scope of Goldhar’s Claim [ 20 ] At the outset, it is important to define the scope of Goldhar’s claim; the relevance of any given consideration to the jurisdiction simpliciter and forum non conveniens analyses is dependent on the scope of the claim.

In my view, a careful review of Goldhar’s amended statement of claim reveals that his action was never limited to libellous statements pertaining to his Canadian business or damage to his Canadian reputation — with respect, my colleagues McLachlin C.J. and Moldaver and Gascon JJ. err by limiting the claim in this way. [ 21 ] It is well established that the statement of claim, which in this case was amended by experienced counsel, defines the issues and informs the opposing parties of the case they have to meet ( Lax Kw’alaams Indian Band v.

Canada (Attorney General) , 2011 SCC 56 , [2011] 3 S.C.R. 535, at para. 41 ). It frames the action for the purposes of analysing the assumption and exercise of jurisdiction. [ 22 ] I cannot conclude from Goldhar’s amended statement of claim, as my colleagues in dissent do, that Goldhar is

particularly “concerned about the impact on his Canadian business reputation” or that the “sting” of libel underlying his claim relates tohis reputation in Ontario (paras. 213-14). While the amended statement of claim states that he is a business owner and operator and anactive community member in Toronto, it refers directly to only one of his business enterprises, Maccabi Tel Aviv, and does not evenmention SmartCentres Inc.

Similarly, para. 9 of his amended statement of claim, which sets out what he considers the natural andordinary meaning of the article, fails to identify any connection to his Canadian business. Furthermore, para. 10, which lists allegedfactual errors and fabrications in the article, does not identify any such errors or fabrications relating to Goldhar’s Canadian businesspractices, but it does specifically identify statements pertaining to his management of Maccabi Tel Aviv (for example, “Goldhar does nothave a long term plan for the team”).

Most notably, the amended statement of claim makes no mention of the article’s claim that“Goldhar’s management model was imported directly from his main business interest — a partnership with Wal-Mart to operateshopping centers in Canada”, even though this is the passage that is said to provide the connection between the allegedly libellousstatements and Goldhar’s Canadian business reputation. Finally, in describing the damage he suffered, Mr. Goldhar clearly states that he“conducts business in Israel, Canada and the United States” (para. 12).

Canada is never singled out as the forum where reputational harmhas been suffered for the purposes of this action. [23] For these reasons, I am satisfied that Goldhar’s action was never limited to damage to his reputation in Ontario or tostatements pertaining to his business in Ontario. Moreover, his undertaking before the motion judge to seek damages only forreputational harm suffered within Canada should not be allowed to narrow the scope of his pleadings.

This undertaking, which does notpreclude a future action from being commenced in Israel to recover damages there, is materially different than the one considered inBlack and, as observed by Pepall J.A., the failure to preclude other actions “detracts from one of the relevant factors for forum nonconveniens enumerated in [Black]: the avoidance of a multiplicity of legal proceedings and conflicting decisions” (para. 162). [24] In light of these comments, I am of the view that Goldhar’s Israeli reputation and statements identified in hisamended statement of claim pertaining to his Israeli business are also relevant to the assumption and exercise of jurisdiction.

Toconclude that “his reputation in Israel is not material to the analysis” ignores the claim as formulated by Goldhar before Haaretz broughta motion to stay (McLachlin C.J. and Moldaver and Gascon JJ., at para. 218). This was the case Haaretz had to meet (Lax Kw’alaams, atpara. 43). Neither Goldhar nor my colleagues McLachlin C.J. and Moldaver and Gascon JJ. (see notably paras. 151, 172, 213, 214 and225) may now redefine Goldhar’s action so that it better responds to Haaretz’s motion to stay. [25] I now turn to the principles underlying the assumption and exercise of jurisdiction. B.

Fundamental Principles Underlying the Conflict of Laws: Balancing Order and Fairness [26] In Club Resorts Ltd. v. Van Breda, 2012 SCC 17, [2012] 1 S.C.R. 572, LeBel J., for a unanimous Court, carefullyexplained the jurisdiction simpliciter analysis, which applies to the assumption of jurisdiction, as well as the forum non conveniensdoctrine, which is meant to guide courts in deciding whether to exercise their jurisdiction.

These principles, along with those relating tothe recognition of foreign judgments, represent the common law conflicts rules of Canadian private international law and must beunderstood and analysed as a cohesive whole (Van Breda, at para. 16). [27] Central to a proper understanding of the conflicts rules of Canadian private international law, and to the resolution ofthis appeal, is an appreciation of the distinct roles played by jurisdiction simpliciter and forum non conveniens (Van Breda, at paras. 46and 56, affirming the reasoning of Sharpe J.A. in Muscutt v.

Courcelles (2002), (ON CA), 60 O.R. (3d) 20 (C.A.),and Charron Estate v. Village Resorts Ltd., 2010 ONCA 84, 98 O.R. (3d) 721). The jurisdiction simpliciter analysis is meant to ensurethat a court has jurisdiction. This will be the case where a “real and substantial connection” exists between a chosen forum and thesubject matter of the litigation.

The forum non conveniens analysis, on the other hand, is meant to guide courts in determining whetherthey should decline to exercise that jurisdiction in favour of a “clearly more appropriate” forum. [28] The importance of maintaining this distinction flows from the discrete concerns underlying each analysis and thenature of the relevant factors at each stage. The “real and substantial connection” test at the jurisdiction simpliciter stage prioritizesorder, stability and predictability by relying on objective connecting factors for the assumption of jurisdiction.

Conversely, the forumnon conveniens analysis emphasizes fairness and efficiency by adopting a case-by-case approach to identify whether an alternativejurisdiction may be “clearly more appropriate”.

I will briefly elaborate on the principles underlying each analysis. [29] In defining the content of the “real and substantial connection” test for the assumption of jurisdiction, this Court wasfaced with a choice between an approach based on objective connecting factors and a case-by-case approach (Van Breda, at para. 30).This choice was characterized by the tension between predictability and consistency, on the one hand, and fairness and efficiency, on theother (Van Breda, at para. 66).

Ultimately, the Court decided to prioritize order and predictability at the jurisdiction simpliciter stage, inthe following terms: Given the nature of the relationships governed by private international law, the framework for the assumption of jurisdiction cannot bean unstable, ad hoc system made up “on the fly” on a case-by-case basis — however laudable the objective of individual fairness maybe. As La Forest J. wrote in Morguard, there must be order in the system, and it must permit the development of a just and fair approachto resolving conflicts.

Justice and fairness are undoubtedly essential purposes of a sound system of private international law. But theycannot be attained without a system of principles and rules that ensures security and predictability in the law governing the assumption ofjurisdiction by a court.

Parties must be able to predict with reasonable confidence whether a court will assume jurisdiction in a case withan international or interprovincial aspect. (Van Breda, at para. 73) To achieve this order and predictability, the Court opted to rely on a set of defined presumptive connecting factors at the jurisdictionsimpliciter stage (Van Breda, at para. 78). [30] This objectively ascertainable and relatively low bar to establishing that a chosen forum has jurisdiction, on a prima

facie basis, reflects the constitutional imperative underlying the jurisdiction simpliciter stage, as described in Van Breda: From a constitutional standpoint, the Court has, by developing tests such as the real and substantial connection test, sought to limit thereach of provincial conflicts rules or the assumption of jurisdiction by a province’s courts. . . .

In its constitutional sense, it places limitson the reach of the jurisdiction of a province’s courts and on the application of provincial laws to interprovincial or internationalsituations. [Emphasis added; para. 23.] The constitutional purpose of the jurisdiction simpliciter test is to establish a minimum threshold for the assumption of jurisdiction inorder to prevent improper assumptions of jurisdiction (Van Breda, at para. 26; see also Hunt v. T&N plc, (SCC), [1993]4 S.C.R. 289, at p. 325).

Its objective is to delineate circumstances in which a court has jurisdiction, not circumstances in which it shouldexercise it (which is the purpose of forum non conveniens).

The prioritization of order and predictability at the jurisdiction simpliciterstage is also consistent with the principle of comity, which is central to Canadian private international law (Van Breda, at para. 74). [31] This prioritization of order and stability at the jurisdiction simpliciter stage, through the adoption of objectivepresumptive connecting factors, is meant to work in tandem with a flexible case-by-case approach to forum non conveniens.

Once it isestablished that a court has jurisdiction, the forum non conveniens doctrine requires a court to determine whether it should exercise suchjurisdiction. [32] The purpose of the forum non conveniens analysis is to temper any potential rigidity in the rules governing theassumption of jurisdiction and “to assure fairness to the parties and the efficient resolution of the dispute” (Van Breda, at para. 104). Thisis necessary given this Court’s recognition that jurisdiction “may sometimes be established on a rather low threshold” (Van Breda, atpara. 109).

By focusing “on the contexts of individual cases”, the forum non conveniens stage plays an important role in striking abalance between order and fairness (Van Breda, at para. 105). [33] Bearing these principles in mind, I turn to the case at bar. C. Did the Motion Judge Err in Assuming Jurisdiction? [34] In determining whether a “real and substantial connection” exists between a chosen forum and the subject matter ofthe litigation, courts are required to consider two issues.

First, a court must consider whether the existence of a recognized presumptiveconnecting factor has been established (Van Breda, at para. 80). If so, the court must consider whether the party challenging theassumption of jurisdiction has successfully rebutted the presumption (Van Breda, at para. 81).

(1) Existence of a Presumptive Connecting Factor [35] The judges in the courts below agreed that a presumptive connecting factor had been established. Haaretz, however,submits that the situs of the tort is an unreliable basis on which to presume a “real and substantial connection” in Internet defamationcases.

In its view, the ease with which publication can be established in such cases gives rise to only a “weak relationship” with thechosen forum. [36] As previously discussed, in Van Breda, the Court stressed the importance of determining jurisdiction “on the basis ofobjective factors” establishing a relationship between the subject matter of the litigation and the chosen forum (para. 82). The Courtidentified the following presumptive connecting factors grounding a court’s assumption of jurisdiction: (

a) the defendant is domiciled or resident in the province; (

b) the defendant carries on business in the province; (

c) the tort was committed in the province; and (

d) a contract connected with the dispute was made in the province. [Emphasis added; para. 90.] The tort of defamation, which is a tort of strict liability, is committed where material has been “communicated” to, that is, conveyed toand received by, at least one person other than the plaintiff (Crookes v. Newton, 2011 SCC 47, [2011] 3 S.C.R. 269, at paras. 1 and 16).This was recognized by this Court in Banro: “. . . a single instance of publication is sufficient for the tort to crystallize” (para. 55).

In thecase of Internet communications, the publication of defamatory statements occurs when they are read or downloaded by the recipient(Black, at para. 20; see also P. A. Downard, The Law of Libel in Canada (4th ed. 2018); Brown on Defamation: Canada, UnitedKingdom, Australia, New Zealand, United States (2nd ed. (loose-leaf)), by R. E. Brown, at pp. 7-17 to 7-25).

Accordingly, the situs ofInternet-based defamation is the place where the defamatory statements are read, accessed or downloaded by the third party (Crookes v.Holloway, 2007 BCSC 1325, 75 B.C.L.R. (4th) 316, at para. 26, aff’d 2008 BCCA 165, 77 B.C.L.R. (4th) 201; Brown, at pp. 7-122 to 7-126; M. Castel, “Jurisdiction and Choice of Law Issues in Multistate Defamation on the Internet” (2013), 51 Alta. L. Rev. 153, at p. 156). [37] Insofar as it attempts to raise doubt as to the validity of the presumptive connecting factors identified in Van Breda,Haaretz’s argument must be rejected.

This Court has found that “[t]he situs of the tort is clearly an appropriate connecting factor” and thatthere is no difficulty “in acknowledging the validity of this factor once the situs has been identified” (Van Breda, at para. 88 (emphasisadded)). Raising doubt as to the value of the situs of the tort as a presumptive connecting factor would significantly undermine theabove-noted objectives of predictability and order at the jurisdiction simpliciter stage. Indeed, courts should be cautious in carving outexceptions to conflicts rules, as “[a]ny exception adds an element of uncertainty” (Tolofson v.

Jensen, (SCC), [1994] 3

S.C.R. 1022, at p. 1061). It is therefore preferable to address any concerns relating to the insufficiency of a presumptive connecting factor either at the rebuttal stage of the jurisdiction simpliciter analysis or at the forum non conveniens stage. [ 38 ] For these reasons, I conclude that a presumptive connecting factor has been established in the circumstances of this case, and I turn to the issue of whether the presumption has been rebutted.

(2) Rebutting the Presumption [ 39 ] At this stage, it is appropriate to take into account the legitimate concerns raised by Haaretz about the ease with which a presumptive connecting factor may be established in Internet defamation cases. This Court has previously recognized the risk of jurisdictional overreach in these types of cases: 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 . . . .

This also raises difficult issues when publication occurs through the Internet . . . . ( Banro , at para. 3) Pepall J.A., dissenting at the Court of Appeal, expressed her reluctance to accept the motion judge’s assumption of jurisdiction in light of similar concerns: To succeed in an action for defamation, the plaintiff must prove on a balance of probabilities that the defamatory words were communicated to at least one person other than the plaintiff: see Crookes v. Newton , [2011] 3 S.C.R. 269, [2011] S.C.J. No. 47, 2011 SCC 47 , at para. 1 .

As well, at the jurisdiction stage of the proceedings, the plaintiff’s pleadings are accepted as true unless contradicted by evidence adduced by the defendants: see Banro , at para. 38. Accordingly, all that is needed for the presumptive connecting factor to be found is for the plaintiff to plead that the alleged defamatory material was communicated to at least one person in Ontario other than the plaintiff.

While this is easy to establish in any defamation case, it is virtually automatic in a case of defamation on the Internet , where online publications are readily shared and accessed by users across the world. [Emphasis added; para. 127.] [ 40 ] The ability to rebut the presumption of jurisdiction where there is only a weak relationship between the subject matter of the litigation and the forum serves as an important check on jurisdiction ( Van Breda , at para. 95).

A careful examination of this question is therefore of particular importance in Internet defamation cases, where a presumptive connecting factor can easily be established. [ 41 ] Having recognized the importance of the ability to rebut the presumption of jurisdiction, I turn to consider Haaretz’s submission that, based on the analyses of the motion judge and the majority below, rebutting the presumption of jurisdiction does not seem possible at all in these types of cases. [ 42 ] This Court has recognized that presumptive connecting factors must not give rise to an irrebuttable presumption of jurisdiction.

A defendant may argue that a given connection is inappropriate in the circumstances of a particular case: The presumption of jurisdiction that arises where a recognized connecting factor — whether listed or new — applies is not irrebuttable. The burden of rebutting the presumption of jurisdiction rests, of course, on the party challenging the assumption of jurisdiction.

That party must establish facts which demonstrate that the presumptive connecting factor does not point to any real relationship between the subject matter of the litigation and the forum or points only to a weak relationship between them. ( Van Breda , at para. 95; see also para. 81) [ 43 ] In order for a defendant to succeed in showing that “a given connection is inappropriate in the circumstances of the case”, the circumstances must demonstrate that the relationship between the forum and the subject matter of the litigation is such that it would “not be reasonable to expect that the defendant would be called to answer proceedings in that jurisdiction” ( Van Breda , at paras. 81 and 97; see also para. 92).

To satisfy this test, the party challenging the court’s jurisdiction should rely on factors other than those considered at the forum non conveniens stage: “. . . the factors that would justify a stay in the forum non conveniens analysis should not be worked into the jurisdiction simpliciter analysis . . .” ( Van Breda , at para. 56). [ 44 ] Assuming that these principles are properly applied, the situs of the tort will not give rise to an irrebuttable presumption of jurisdiction in Internet defamation cases.

While it is not appropriate to propose an exhaustive list of factors that can rebut the presumption of jurisdiction in these types of cases, it is not difficult to imagine circumstances in which it would not be reasonable to expect that the defendant would be called to answer a legal proceeding in a chosen forum. For example, evidence that a plaintiff has no reputation in the chosen forum may be a factor tending to rebut the presumption of jurisdiction in a defamation action.

As the protection of reputation is the primary purpose of defamation law ( Banro , at paras. 57-58), absence of reputation would tend to point to a weak relationship between the forum and the subject matter of the litigation.

Indeed, this Court, in Banro , relied in part on the plaintiff’s reputation in the chosen forum to conclude that it would be inappropriate to find that the presumption of jurisdiction had been rebutted in the circumstances of that case (para. 38). [ 45 ] In the case at bar, the evidence fails to establish that Haaretz could not have reasonably expected to be called to answer a legal proceeding in Ontario. The pleadings indicate that Goldhar lives and operates his businesses in Ontario. Haaretz had knowledge of this fact, and the allegedly libellous

article directly references Goldhar’s Canadian residency and Canadian business practices. As such, this is not a case where the presumption of jurisdiction is rebutted.

D. Did the Motion Judge Err in Finding That Israel Is Not a Clearly More Appropriate Forum Than Ontario? [ 46 ] Having established jurisdiction (pursuant to the jurisdiction simpliciter analysis), the motion judge properly considered the question of forum non conveniens .

At the forum non conveniens stage, the burden is on the defendant to satisfy the motion judge that the alternative forum is “clearly more appropriate” by establishing that it would be fairer and more efficient to proceed in that forum: The use of the words “clearly” and “exceptionally” should be interpreted as an acknowledgment that the normal state of affairs is that jurisdiction should be exercised once it is properly assumed.

The burden is on a party who seeks to depart from this normal state of affairs to show that, in light of the characteristics of the alternative forum, it would be fairer and more efficient to do so and that the plaintiff should be denied the benefits of his or her decision to select a forum that is appropriate under the conflicts rules. The court should not exercise its discretion in favour of a stay solely because it finds, once all relevant concerns and factors are weighed, that comparable forums exist in other provinces or states. It is not a matter of flipping a coin.

A court hearing an application for a stay of proceedings must find that a forum exists that is in a better position to dispose fairly and efficiently of the litigation. But the court must be mindful that jurisdiction may sometimes be established on a rather low threshold under the conflicts rules.

Forum non conveniens may play an important role in identifying a forum that is clearly more appropriate for disposing of the litigation and thus ensuring fairness to the parties and a more efficient process for resolving their dispute. [Emphasis added.] ( Van Breda , at para. 109) [ 47 ] While the normal state of affairs favours exercising jurisdiction in the forum where it is properly assumed, this should never come at the cost of one party facing unfair or clearly inefficient proceedings.

The purpose of forum non conveniens , as discussed above, is to temper any potential rigidity in the rules governing the assumption of jurisdiction and to “assure fairness to the parties and the efficient resolution of the dispute” ( Van Breda , at para. 104). Where the evidence indicates that the alternative forum is in a better position to dispose fairly and efficiently of the litigation, the court should grant the stay ( Van Breda , at para. 109).

This is especially true in cases where the evidence raises doubt as to whether proceeding in the chosen forum will provide the defendant with a fair opportunity to present its case. [ 48 ] In light of the purpose of forum non conveniens , I agree with Pepall J.A. that, “given the ease with which jurisdiction simpliciter may be established in a defamation case, in a motion for a stay, a motion judge must conduct a robust and carefully scrutinized review of the issue of forum non conveniens ” (para. 132).

It is true that defamation cases involve a particularly rigid application of the rules governing the assumption of jurisdiction. As discussed above, the establishment of a presumptive connecting factor is “virtually automatic” in Internet defamation cases (Pepall J.A., at para. 127). Where there is no “real and substantial connection” to the chosen forum, a proper analysis at the rebuttal stage will alleviate some of the consequences of the rigid application of the rules governing the assumption of jurisdiction.

That being said, there are some other consequences to the rigid application of these rules that can only be addressed in the forum non conveniens analysis. For example, where a plaintiff enjoys a reputation in multiple forums, publication may allow jurisdiction to be properly assumed in all of them, without regard to how fair or efficient it may be to proceed in the chosen forum. This is to be expected as, again, “the factors that would justify a stay in the forum non conveniens analysis should not be worked into the jurisdiction simpliciter analysis” ( Van Breda , at para. 56).

As the rebuttal stage fails to address all the consequences of the “virtually automatic” presumption of jurisdiction in defamation actions, it is appropriate for motion judges to be particularly attuned to concerns about fairness and efficiency at the forum non conveniens stage in these types of cases.

This should not be understood as imposing a different standard or burden for defamation cases. [ 49 ] I acknowledge that a motion judge’s decision on a stay motion is entitled to deference: 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). ( Banro , at para. 41) As the forum non conveniens analysis is inherently factual in nature, courts of appeal should not normally interfere with a motion judge’s factual findings. That being said, there are limits to deference, as recognized in Banro . Where the motion judge has “erred in principle, misapprehended or failed to take account of material evidence, or reached an unreasonable decision”, courts of appeal may intervene. [ 50 ] Bearing these principles in mind, and for the reasons set out below, I conclude that this Court may intervene in the case at bar.

The motion judge committed the following errors (with references to each factor impacted by the error, as discussed below): • He erred in finding that letters rogatory could be used to compel Israeli witnesses to testify in Ontario (Comparative Convenience and Expense for the Witnesses, and Fairness). • He erred by giving significant weight to Goldhar’s undertaking to fund the travel and accommodation expenses of the foreign witnesses in accordance with the rates provided in the Rules (Comparative Convenience and Expense for the Witnesses). • He erred by unreasonably discounting Haaretz’s proposed witnesses and the relevance of their evidence (Comparative Convenience and Expense for the Witnesses). • He erred by failing to consider Goldhar’s significant reputation in Israel (Fairness). • He erred by failing to weigh Goldhar’s interest in vindicating his reputation in Ontario against the significant unfairness

that a trial in Ontario would impose on Haaretz (Fairness). • He erred by failing to consider the question of enforcement (Enforcement). In committing the first error, the motion judge misapprehended the role of letters rogatory entirely. As I will explain, the remaining errors involved a complete misapprehension of, or failure to consider, material evidence, and not merely, as my colleagues in dissent argue, an unsatisfactory weighing of that evidence. These errors tainted the motion judge’s forum non conveniens analysis on each of the factors they affected as well as his overall weighing of these factors.

As a result, no deference should be afforded to these aspects of the motion judge’s analysis. [ 51 ] I turn then to an assessment of each factor raised by Haaretz.

(1) Comparative Convenience and Expense for the Parties [ 52 ] The motion judge concluded that the comparative convenience and expense for the parties favoured a trial in Israel (para. 36). Neither party disputes this. [ 53 ] The motion judge reached this conclusion on three bases. First, there was no evidence that a trial in Israel would cause any inconvenience or expense to Goldhar (para. 35). Second, holding a trial in Ontario would place a strain on the Israeli defendants (paras. 31-33). And finally, portions of the trial might need to be conducted in Hebrew with interpreters (para. 34). I would not disturb the motion judge’s conclusion on this factor.

(2) Comparative Convenience and Expense for the Witnesses [ 54 ] Goldhar did not file any evidence regarding the witnesses he would call at trial, while Haaretz filed a list of 22 witnesses and described, in its factum, what each of the witnesses “may speak to” (motion judge, at para. 41). Furthermore, a supplemental affidavit described the evidence that eight of these witnesses could give to assist Haaretz at trial (motion judge, at para. 43).

Of the 22 witnesses, 18 resided in Israel. [ 55 ] The motion judge concluded that the comparative convenience and expense for the witnesses, a factor distinct from the comparative convenience and expense for the parties, only slightly favoured a trial in Israel (para. 45). This conclusion was wholly unreasonable in light of the evidence before him. [ 56 ] While the motion judge did not specifically identify the basis on which he concluded that the comparative convenience and expense for the witnesses only slightly favoured a trial in Israel, he did highlight four considerations.

First, he considered that Goldhar had filed an expert opinion to the effect that “many of the witnesses do not have relevant evidence” (para. 41). Second, he dismissed Haaretz’s concern about being unable to compel unwilling witnesses living outside of Ontario. He found that these witnesses would remain unwilling to testify even if the trial were held in Israel and that “compelling the attendance of these witnesses to a court in Ontario can be accomplished through the use of letters rogatory” (para. 42).

Third, the motion judge considered that arrangements could be made to have foreign witnesses testify by videoconferencing technology, pursuant to rule 1.08 of the Rules (para. 44).

Finally, he found that Goldhar’s undertaking to pay for the travel and accommodation costs of foreign witnesses in accordance with the rates provided in the Rules addressed any additional expense related to holding a trial in Ontario (para. 44). [ 57 ] The Court of Appeal correctly recognized that the motion judge had erred in law by suggesting that letters rogatory could be used to compel the attendance of Haaretz’s witnesses in Ontario, but the majority concluded that this error did not make the motion judge’s overall assessment of this factor unreasonable. It gave three reasons for reaching this conclusion.

First, videoconferencing could be used to obtain the testimony of witnesses who were unwilling or unable to come to Ontario: Contrary to Haaretz’s arguments, in my view, the motion judge was entitled to accept that reluctant foreign witnesses could be compelled to provide evidence in Israel through the use of letters of request and that videoconferencing was a potential means of obtaining the evidence of any witnesses unwilling to come to Ontario. These are available methods, under the Rules of Civil Procedure , for dealing with witnesses outside the jurisdiction.

Haaretz led no evidence to undermine Goldhar’s submissions that these methods would be available in this case. Haaretz bore the burden of demonstrating that Israel is the clearly more appropriate forum. On this record, it was not unreasonable for the motion judge to accept that Ontario letters of request would be honoured by Israel and that videoconferencing would be available in that jurisdiction. Further, in the absence of evidence or adverse judicial commentary, the use of technology and interpreters cannot be viewed as undermining the fairness of a civil trial.

We live in an age of international communication and commerce. Multi-jurisdictional parties — and witnesses who do not speak either of Canada’s official languages — are to be expected. Courtroom procedures must accommodate testimony by videoconferencing. Interpreters have long been a common feature of the Canadian judicial system.

The motion judge’s implicit conclusion that using these procedures would not undermine the fairness of the trial was not unreasonable. [Footnote omitted; paras. 69-71.] Second, the majority of the Court of Appeal was of the view that Goldhar’s undertaking to fund travel and accommodation expenses relieved any additional expense of holding the trial in Ontario.

Finally, there was a lack of evidence concerning the likely testimony of Haaretz’s proposed witnesses: While many of Haaretz’s proposed witnesses could have information about relevant matters, the record contains scant information about what particular witnesses are actually likely to say. Importantly, Mr. Marouani, the reporter who wrote the article, did not provide an

affidavit on the motion. Nor did Haaretz produce any witness statements or even any notes of conversations with the proposed witnesses.

In these circumstances, the motion judge was entitled to treat Haaretz’s proposed witness list with caution. [para. 73] [ 58 ] There is no doubt that the motion judge erred in finding that letters rogatory could be used to compel Israeli witnesses to testify in Ontario. [ 59 ] First, by dismissing evidence that Haaretz’s witnesses would not testify voluntarily on the basis that “[t]his concern will exist even if the trial is held in Israel” (para. 42), the motion judge ignored the very concern raised by Haaretz, namely that Israeli witnesses, while compellable in Israel, could not be effectively compelled to testify if the trial proceeded in Ontario.

Se

[…]