In v Goldhar 2018 SCC 28, the Canadian Supreme Court considered jurisdiction and forum conveniens in a multi-jurisdictional Internet libel claim. The Court was divided – allowing the news organisation’s appeal 6:3. All of the Justices concluded that Canada had jurisdiction to hear the claim, but a majority found that Israel was a clearly more appropriate forum.


Mitchell Goldhar is a prominent Canadian businessman. He also owns Israeli football club Maccabi Tel Aviv FC. In 2011, an Israeli news organisation – Haaretz – published an article alleging that Mr Goldhar’s management of the club had been “penny-pinching” and lacked long-term planning. The article was published in print in Israel and online. The majority of readers were in Israel, but at least 200 people read the article online in Canada.

Mr Goldhar commenced a defamation action against Haaretz in Ontario. Haaretz brought a motion to have the claim stayed, arguing that the Canadian court lacked jurisdiction to hear the claim, or alternatively that Israel was a clearly more appropriate forum. The motion judge held that Ontario did have jurisdiction and that Israel was not a clearly more appropriate forum. A majority of the Court of Appeal in Ontario agreed. Haaretz appealed to the Canadian Supreme Court.


The Canadian Supreme Court (6:3) granted the news organisation’s motion to stay Mr Goldhar’s defamation claim in Ontario. The majority were all agreed as to the outcome of the case, but they differed in some aspects of their reasoning.

The majority (1) – Côté, Brown and Rowe JJ

Côté, Brown and Rowe JJ gave a judgment together. They recognised that the Internet increased the potential for multi-jurisdictional publication, which brought with it a risk of ‘libel tourism’ – allowing a claimant to shop for the forum they consider strategically most beneficial to them [1]. The three Justices considered, however, that the pre-existing law on jurisdiction was able to address the challenges of the Internet Age, so long as the underlying principles of stability and fairness were borne in mind [2].

The Justices explained that a court would have jurisdiction where there is a “real and substantial connection” between the forum and the subject matter of the litigation [27]. To achieve order and predictability, the courts rely on a set of presumptive connecting factors at the jurisdiction stage (Club Resorts Ltd v Van Breda 2012 SCC 17 [78]) [29]. As there had been publication in Ontario, Mr Goldhar had established a presumptive factor connecting the claim to Canada [38]. The burden then shifted to Haaretz to rebut the presumption. Côté J et al. held that Haaretz could reasonably have expected to be called to answer legal proceedings in Ontario [43], [45]. Consequently, the news organisation had not rebutted the presumption of jurisdiction.

On the question of forum, Côté, Brown and Rowe JJ held that Israel was a clearly more appropriate forum [95]-[97]. They considered that in Internet libel cases, where jurisdiction could easily be established, there had to be a “robust and carefully scrutinized review” of forum non conveniens [48]. Courts of appeal “should not normally interfere” with the analysis of a motion judge, but there were limits to deference; where the motion judge erred in principle or failed to take account of material evidence, appellate courts could intervene [49].

Côté J et al. held that the motion judge had fallen into error in his forum conveniens analysis [50]. Haaretz had established that it would face substantial unfairness and inconvenience if a trial were held in Ontario. Factors which favoured Israel included the comparative convenience and expense for the parties and witnesses, fairness (Mr Goldhar had a significant business interest and reputation in Israel) and enforcement (Haaretz had no presence or assets in Ontario) [96].

The majority (2) – Karakatsanis, Abella and Wagner JJ

Karakatsanis, Abella and Wagner JJ agreed with the conclusion of Côté J et al., but disagreed with some aspects of their reasoning on forum. Each of Karakatsanis, Abella and Wagner JJ gave a separate judgment.

Abella and Wagner JJ advocated modifying the choice of law question in Internet libel claims, replacing the principle of lex loci delicti (the place where the tort occurs) with a test based on where the most substantial harm to the plaintiff’s reputation occurred [109], [144]. Applying their preferred approach, Abella and Wagner JJ concluded that the place of most substantial harm to Mr Goldhar’s reputation was Israel, so Israeli law should apply [131]. They agreed with Côté J et al. in respect of the rest of the forum analysis, concluding that Israel was the clearly more appropriate forum [136], [150].

Karakatsanis J gave a short judgment in which she took issue with aspects of Côté J’s reasoning on forum [99]-[103]. She considered (inter alia) that Mr Goldhar’s reputation in Israel was not material to fairness, as that factor was concerned with the Mr Goldhar’s “interest in vindicating his reputation in the jurisdiction where he enjoys it” [101]. However, Karakatsanis, Abella and Wagner JJ all agreed with Côté J et al. in the result, which meant that Haaretz’s appeal was allowed and the claim stayed in Canada.

The dissent – McLachlin CJ and Moldaver and Gascon JJ

The dissenting Justices would have dismissed Haaretz’s appeal. They stated:

“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 — is he 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 answer of the motion judge and of the majority in the Court of Appeal was yes. We agree…” [151].

McLachlin CJ and Moldaver and Gascon JJ considered that the current test for jurisdiction readily accommodated multi-jurisdictional Internet defamation cases and considered that it was “more than reasonably foreseeable that Haaretz would be sued in Ontario” [172]. The dissenting judges stated that, if analysis at the rebuttal stage is “done properly, with an adequate consideration of reasonable foreseeability”, there would be “no need to apply a robust and carefully scrutinized forum non conveniens analysis”, as advocated by Côté J et al. The dissentients were concerned that this “new standard” would “frustrate the predictability and stability” at the core of the forum framework [174].

McLachlin CJ and Moldaver and Gascon JJ would have held that the factors in the instant case did not show that Israel was a clearly more appropriate forum than Ontario. The considered that the only factor which favoured Israel was comparative convenience and expense for the parties and witnesses. They concluded that the “key factors” of applicable law and fairness weighed heavily in favour of Ontario [238].


This case provided an opportunity for Canada’s highest court to address certain complexities arising from multi-jurisdictional Internet defamation claims. The potential for ubiquitous publication creates novel questions about the boundaries of libel and prompts senior courts to grapple with the risk of forum shopping.

The case provoked an interesting debate among the Justices over the question of applicable law as a factor in the forum conveniens analysis.

As noted above, Abella and Wagner JJ considered that the principle of lex loci delicti should not continue to provide the basis for choice of law in Internet libel claims. Abella J stated that the “standard approach” did not adequately respond to the “unique issues and challenges” of Internet defamation, where a single download could determine applicable law [105]. They considered that an alternative test – the place of the most substantial harm to reputation – would focus on the place with the most significant connection to the tort, allowing forum conveniens to be determined “on a more principled basis” [145].

The other judges considered that it was not appropriate to adopt a new test in place of the lex loci delicti principle. Côté, Brown and Rowe JJ considered that such a change could create legal uncertainty [91]. Similarly, the dissenting judges thought that the ‘place of most substantial harm’ test should not be adopted instead of lex loci delicti. They considered that such a test would be “highly subjective” and would not always clearly point towards one jurisdiction. They were also concerned that such a test would lead to complex preliminary motions and increase delay and expense [200][201].

As these discussions demonstrate, Internet publication continues to pose challenges to traditional doctrines in libel and provides fertile ground for innovative arguments in litigation.

Natasha Holcroft-Emmess is a trainee barrister at Matrix Chambers.