After Wilson v Servier Canada Inc. (2000), 50 OR (3d) 219 (Gen. Div.), ¶57-95, defendants ought to have stopped bringing national class “constitutional” motions. In that case, Sheila Wilson sued Servier Canada Inc., its parent corporation Biofarma S.A., and others for their role in making and selling Ponderal, a weight loss drug that caused “life threatening diseases, including primary pulmonary hypertension, valvular heart disease and valvular regurgitation.” Cumming J. considered the defendants’ motion to disallow the certification of a “national class” on the basis of ss. 92(13), (14), and (16) of the Constitution Act, 1867. They submitted that the Ontario Superior Court lacked jurisdiction over non-resident claims on the basis that the Class Proceedings Act was ultra vires the legislature of Ontario to the extent that it allowed a national class. They submitted that non-residents did not have a real and substantial connection to Ontario.
After reviewing Morguard and Hunt and the decisions where Canadian courts certified “national” classes, Cumming J. dismissed the motion. His analysis is just as compelling today as it was then, yet the Ontario defence bar continues to make the same pitch in other provinces to local judges through satellite firms.
My answer to the defence position on the national class engages the following five theses.
(i) The “constitutional imperatives” are found in The BNA Act and Charter.
In Morguard Investments Ltd. v De Savoye, [1990] 3 SCR 1077, Morguard Investments Limited and the Credit Foncier Trust Company sought to enforce an Alberta judgment against Douglas De Savoye in British Columbia. While residing in Alberta, Mr. De Savoye borrowed money from them and mortgaged his Alberta situated property to secure the loan. He later moved to British Columbia, and after he defaulted on the loan, Morguard and Credit Foncier foreclosed on the mortgage, sold the land, applied the proceeds of sale to Mr. Savoye’s debt and obtained a judgment against him for the deficiency. They sued Mr. De Savoye in British Columbia to enforce the Alberta judgment. Mr. De Savoye had not contracted to have mortgage disputes heard in Alberta, and he did not attorn to the proceedings (he took no steps to defend them at all). Morguard and Credit Foncier could not satisfy any of the traditional Symon categories for assuming jurisdiction, which had been created by English courts for Englishmen doing business all over the globe in a bygone era,. So La Forest J. considered whether, within the Canadian common market federation, the Alberta Court had any basis beyond Symon to assume jurisdiction in a way that would require a British Columbia court to give the Alberta judgment “full faith and credit”.
The “real and substantial connection” test is not the “rule in Morguard.” The real “rule in Morguard” is with respect to judgment recognition – one provincial court must enforce a judgment from another provincial court if the court properly exercised jurisdiction. “Real and substantial connection” was not established as a constitutional or indeed any broad test in Morguard. It was a rule made not by judges, but elevated beyond its scope by some academics in the years after Morguard was decided. (I suspect Toronto lawyers latched onto that language because it sounded like “minimal contacts” in American constitutional law.) At p. 325 of Hunt v T&N plc, [1993] 4 SCR 289, La Forest J. tried to put out those fires:
In Morguard, a more accommodating approach to recognition and enforcement was premised on there being a “real and substantial connection” to the forum that assumed jurisdiction and gave judgment. Contrary to the comments of some commentators and lower court judges, this was not meant to be a rigid test, but was simply intended to capture the idea that there must be some limits on the claims to jurisdiction.
At p. 326, he reproached any rigid test:
I need not, for the purposes of this case, consider the relative merits of adopting a broad or narrow basis for assuming jurisdiction and the consequences of this decision for the use of the doctrine of forum non conveniens.... Whatever approach is used, the assumption of and the discretion not to exercise jurisdiction must ultimately be guided by the requirements of order and fairness, not a mechanical counting of contacts or connections.
La Forest J. cited “real and substantial connection” test as just one example from the matrimonial category of cases where jurisdiction was “reasonably” assumed. Other categories are:
| category | quote | cite |
| matrimonial | “Indeed I observed (at p. 1104) that the “real and substantial connection” test was developed in Indyka v. Indyka, [1969] 1 A.C. 33, in a case involving matrimonial status (where sound policy demands generosity in recognition), and that in a personal action a nexus may need to be sought between the subject matter and the territory where the action is brought.” | Hunt, p. 325 |
| tort | “I then considered the test developed in Moran v. Pyle National (Canada) Ltd., supra, for products liability cases as an example of where jurisdiction would be properly assumed.” | Hunt, p. 325 |
| contract | “...Moran v. Pyle National (Canada) Ltd., [1975] 1 S.C.R. 393, through a tort action, is instructive as to the manner in which a court may properly exercise jurisdiction in actions in contracts as well. ... ...barring express or implied agreement, the reasoning in Moran is obviously relevant to contracts; indeed, the same activity can often give rise to an action for breach of contract and one in negligence....” | Morguard, pp. 1105, 1107 |
| criminal | “It is interesting to observe the close parallel in the reasoning in Moran with that adopted by this Court in dealing with jurisdiction for the purposes of the criminal law; see Libman, supra.” | Morguard, p. 1107 |
| other | “Morguard was careful to indicate, however, that a court must have reasonable grounds for assuming jurisdiction. ...The exact limits of what constitutes a reasonable assumption of jurisdiction were not defined, and I add that no test can perhaps ever be rigidly applied; no court has evr been able to anticipate all of these.” | Hunt, p. 325 |
So the “real and substantial connection” test was not said to be, nor did it become, a “constitutional rule.” La Forest J. referred to it as just one example of where courts appropriately assumed jurisdiction in the matrimonial context, while recognizing that jurisdiction was context-sensitive.
Recently, Van Breda v Village Resorts Ltd., 2012 SCC 17 dealt with the test as a conflicts rule, but in obiter dicta, LeBel J. may have elevated the test to a constitutional rule or at least a catch-all phrase for the aggregate effect of a series of provisions which actually do have a textual basis in the BNA Act. There is still support in the SCC’s conflicts jurisprudence that the correct question to ask is whether the forum is a “reasonable” place for the action to be heard. La Forest J. had more to say about a “reasonable” or “proper” assumption of jurisdiction than he did about “real and substantial connection”. In Beals v Saldanha, [2003] 3 SCR 416, ¶22, Major J. affirmed that: “Modern ideas of order and fairness require that a court must have reasonable grounds for assuming jurisdiction where the participants to the litigation are connected to multiple jurisdictions.” LeBel J. in Van Breda did not have to examine the constitutional underpinnings that form expression in the phrase “real and substantial connection”. A regular return to the fundamentals as stated in Morguard is preferable to the tireless gossiping amongst academics and lobbying by defence lawyers as to what Morguard means.
What were the “constitutional imperatives” in Morguard? In Hunt, La Forest J. referenced the “principle” (p. 332) and “principles” (p. 326-27, 331) in Morguard” and the “constitutional imperatives” rooted in the BNA Act and Charter. They include:
| imperative | constitution | quote | morguard |
| common market | ss. 91(2), 91(10), 121, & “pogg” | “In particular, significant steps were taken to foster economic integration. One of the central features of the constitutional arrangements incorporated in the Constitution Act, 1867 was the creation of a common market. Barriers to interprovincial trade were removed by s. 121. Generally trade and commerce between the provinces was seen to be a matter of concern to the country as a whole; see Constitution Act, 1867, s. 91(2) |
¶36 |
| common judiciary | ss. 92(14), 96, 100, & 101 | “The Canadian judicial structure is so arranged that any concerns about differential quality of justice among the provinces can have no real foundation. All superior court judges – who also have superintending control over other provincnial courts and tribunals – are appointed and paid by the federal authorities. And all are subject to final review by the Supreme Court of Canada, which can determine when the courts of one province have appropriately exercise jurisdiction in an action and the circumstances under which the courts of another province should recognize such judgments.” | ¶37 |
| common bar | none | “Any danger resulting from unfair procedure is further avoided by subconstitutional factors, such as for example the fact that Canadian lawyers adhere to the same code of ethics throughout Canada. In fact, since Black v. Law Society of Alberta, supra, we have seen a proliferation of interprovincial law firms.” | ¶37 |
| interprovincial mobility | s. 6 | “A common citizenship ensured the mobility of Canadians across provincial lines, a position reinforced today by s. 6 of the Charter; see Black v. Law Society of Alberta, [1989] 1 S.C.R. 591” | ¶36 |
As this table illustrates, each of these factors (except the common bar which La Forest referred to as “sub-constitutional”) comes from text within the Constitution. Where does “real and substantial connection” appear in the Constitution? Nowhere.
Argument on the national class should be informed by these true “constitutional considerations” that underpin Morguard and Hunt. With a single court system, in a class action arising from a federally imported and regulated product sold throughout a common market to an interprovincially mobile population, there are no non-residents. All class members are ‘in the jurisdiction’.
The common judiciary is a key to the national class issue. At pp. 34-35 of Aetna Financial Services v Feigelman, [1985] 1 SCR 2, Estey J. recognized a provincial superior court’s national jurisdiction:
An initial question, therefore, must be answered, namely, what is meant by “jurisdiction” in a federal context? It at least means the jurisdiction of the Manitoba court. ... In the Canadian federal system, the appellant is not a foreigner, nor even a non resident in the ordinary sense of the word. It is capable of ‘residing’ throughout Canada and did so in Manitoba. ... some ways, ‘jurisdiction’ extends to the national boundaries, or, in any case, beyond the provincial boundary of Manitoba.
See Meeking v Cash Store Inc. (Sept. 9), 2013 MBCA 81 (Chartier, Beard, Cameron JJ.A.).
National jurisdiction is supported by Muscutt v Courcelles (2002), 60 OR (3d) 20 (Ont. C.A.), ¶95- 100, where Sharpe J.A. acknowledged that it is easier to justify assuming jurisdiction over nonresidents in interprovincial cases than international ones. When Canadian judges preside over national common market class actions, they sit on a federalist bench. There is one jurisdiction, not thirteen. Indeed, the word “Canada” is the first word on court documents in some provincial superior courts. That is a reminder of their overlapping federal jurisdiction.
The phenomenon of an interprovincially mobile population supports class definitions that are not limited to current residents of a single province. Class members, like hydrocarbons, are fugacious. While Tiboni v Merck Frosst Canada Ltd. (2009), 95 OR (3d) 269 (Div. Ct.) has discouraged a rush to be the first court to “strike oil” and extract class members from other jurisdictions, there is an analogy between the “rule of capture” in petroleum law, and the “first-to-certify” rule in multijurisdictional class actions throughout North America. While recent private international law jurisprudence from the Supreme Court of Canada has facilitated the flow of wealth, skill, and people across state lines, s. 6 of the Charter constitutionally guarantees Canadians the right to interprovincial mobility. There are many permutative situations arising from the exercise of those mobility rights, and many more when there is an international component. An elderly user of a defective pharmaceutical may reside, purchase their drug, and sustain injury in Saskatchewan in 2010, become a resident of Nova Scotia in 2011, and domicile to Arizona in 2012, where they may reside for the purpose of assessing their capacity to opt into a class action certified in British Columbia.
Tort based products liability class actions in particular possess other constitutional factors that support a national class.
(ii) The “real and substantial connection” test does not apply in every jurisdictional challenge.
At ¶19 of Muscutt, supra, Sharpe J.A. described “three ways in which jurisdiction may be asserted against an out-of-province defendant”:
As a rule of private international law rather than a constitutional rule, the “real and substantial connection" test is not the governing consideration in every jurisdictional challenge. It does not oust traditional presence and consent based grounds for assuming jurisdiction: Ward v Canada (Attorney General), 2007 MBCA 123; Club Resorts Ltd. v Van Breda, 2012 SCC 17, ¶79. It does not supplant longstanding checks on the territorial limits of provincial legislatures. Moreover, there is “order and fairness”. At p. 326 of Hunt, supra, La Forest J. said, “the assumption of and the discretion not to exercise jurisdiction must ultimately be guided by the requirements of order and fairness, not a mechanical counting of contacts or connections.”
One aspect of “order and fairness” is fairness to the plaintiffs. That recognizes their right to choose a forum, a right that should not be upset unless a defendant shows great hardship. They rarely can in a modern world with fast planes, electronically scanable documents, and video-conferencing. Defendants cannot show hardship in concurrently defending common issues against residents and non-residents. That is because there is no added burden by a large class, and indeed, there are many benefits to the largest class possible. Harrington v Dow Corning Corp. (1997), 29 BCLR (3d) 88 (S.C.), ¶19; Nantais v Teletronics Proprietary (Canada) Ltd. (1995), 127 DLR (4th) 552 (Ont. S.C.J.), ¶79.
There are benefits to including all Canadians in a single class. Nantais v Telectronics Proprietary (Canada) Ltd. (1995), 127 DLR (4th) 552 (Ont. S.C.J.), ¶82; Wilson v Servier Canada Inc. (2000), 50 OR (3d) 219 (Ont. S.C.J.), ¶93; Ring v Canada (Attorney Canada), 2007 NLTD 213, ¶34-35. “Order and fairness” recognizes the efficiencies of a national class action. From the perspective of judicial economy, extra provincial residents come at no extra effort or cost. In Harrington v Dow Corning (1997), 29 BCLR (3d) 88 (S.C.)
[19] Once jurisdiction is established, the question remaining is forum conveniens. I am satisfied that the Class Proceedings Act facilitates the efficient litigation of multiple claims and this jurisdiction is therefore a convenient forum. In Nantais, both levels of court stressed that a certification order can be varied if unexpected problems arise. The same flexibility is available under the B.C. statute. The common issue will not be made any more complicated by the inclusion of non-resident class members. The defendants may be deprived of the opportunity of trying that factual issue separately in several jurisdictions but, if that is prejudicial, it is outweighed by the advantage to the class members of having a single determination of a complex issue that can only be litigated at substantial cost.
(iii) The “real and substantial connection” test examines connection to the litigation as a whole, without requiring a connection to each party.
The “real and substantial connection” test does not require a connection between the forum and each party. In Muscutt, Sharpe J.A. in reviewing Morguard referred to connections between the “subject matter of the action”, the “wrongdoing”, “damages suffered”, the “defendant”, the “transaction”, and the “parties” to the forum. Even if class members were “parties” (and they are not), the absence of a connection between the forum and any one of many parties is just one consideration.
(iv) The “real and substantial connection” test in class actions should be based on the “administration of justice” approach, not the “personal subjection” approach.
Defendants’ statement of the “real and substantial connection” test is contrary to the broader “administration of justice” approach that Muscutt accepted. Personal subjection is not an acceptable approach. If “the defendant” were replaced with “each class member”, the following extract of LeBel J. from Beals v. Saldanha, supra, describes the broader approach to jurisdiction:
[179] But there may be good reasons why jurisdiction should be recognized even where there is little or no connection to the defendant, particularly when other considerations, such as fairness to the plaintiff and the importance of administering the ustice system in an efficient manner, are taken into account along with the interests of the defendant. It is not unusual for cross-border litigation to arise out of complex transactions involving a number of parties with connections to several jurisdictions. Watson and Au, supra, point out, at p. 200, that when litigation involves “multiple defendants in different jurisdictions, insisting on a substantial connection between each defendant and the forum can lead to a multiplicity of actions and inconsistent findings”. In such circumstances, a test that recognizes jurisdiction based on a connection to the subject matter of the action seems better suited to identifying whether the forum is a reasonable place for the action to be heard.
In Muscutt, for cases of “assumed jurisdiction”, Sharpe J.A. rejected a “personal subjection” approach in actions that requires each defendant to be connected to the province. It would be ironic if another court accepted the “personal subjection” approach on the plaintiff side of the equation for class actions, where class members make up a cohesive body with a common interest in the outcome of the litigation. Sharpe J.A. based his eight part checklist on the broader “administration of justice” approach which considers the “involvement of other parties to the suit” and whether the litigation is interprovincial or international as two factors to consider along with the connection of the forum to the plaintiff and defendant.
If that “administration of justice” approach were applied to class actions on the basis that “class members” are “parties”, their connection to the jurisdiction would be merely one factor that the Court would consider in whether to assume jurisdiction over the national class action as a whole. Defendants seek to elevate the residence of class members from a factor to a requirement. That would make jurisdiction simpliciter more restrictive in class actions than in individual actions.
(v) Defining a class does not require assuming jurisdiction over class members.
Defendants rely on the territorial limitations on provincial legislatures set out in s. 92. Legislative power is territorially limited. But, as discussed above, judicial jurisdiction in many ways extends beyond the provinces. Courts regularly assume jurisdiction over non-residents. Moreover, territoriality, which limits legislative power, and which governed the 19th century territoriality based recognition and enforcement theories, is not the only nor primary constitutional consideration in Canadian judicial jurisdictional disputes.
Drafting a class definition does not require that a court assume jurisdiction over non-residents. It does not make class members parties to litigation. Class actions legislation does not require any reference to residence when stating a class’s identifying characteristics in a certification order. Class members can opt out of the action – they are not compelled to answer legal process served upon them or have judgment granted against them. Defendants who are brought into the jurisdiction, in contrast, cannot opt-out. Fairness requires a higher standard of jurisdictional scrutiny over party defendants who cannot opt out than over non-party class members who can. Zuber J. in Nantais v Telectronics Proprietary (Canada) Ltd. (1995), 129 DLR (4th) 110 (Ont. Div. Ct.), ¶12:
It is clear that the Ontario legislature and the Ontario courts are not simply imposing jurisdiction on non-residents. Those outside the jurisdiction who are included in the class are free to opt out in the same manner as those inside Ontario may do.
Including non-residents in a class definition is not a determination that they will be bound by the common issues resolution. Other courts at the enforcement stage will later determine whether they are. Nantais v Telectronics Proprietary (Canada) Ltd. (1995), 129 DLR (4th) 110, ¶13; Spar Aerospace Ltd. v American Mobile Satellite Corp., [2002] 4 SCR 205 (Le Bel J.), ¶64; Nantais v Teletronics Proprietary (Canada) Ltd. (1995), 25 OR (3d) 331, ¶79-80. Certification judges need not decide what another Court will do at the enforcement stage and then make that a factor in deciding jurisdiction. Comity is better observed by that Court considering how it wants a common issues order to affect in personam rights of its residents. Including non-residents in the class definition would invite non-residents to participate and would invite another court to accept that a common issues judgment would be binding, but would not decide the issue on behalf of the foreign court. Whether another court will recognize and enforce a common issues judgment is better decided by the foreign court if faced with an identified claimant who attempted to bring an individual action which asserts a different position on the common issue than that set out in the judgment. In Robertson v Thomson Corp. (1999), 43 OR (3d) 161:
The defendants do not suggest that Ontario is not an appropriate forum for this case or that Ontario otherwise lacks jurisdiction. Nor is it disputed that virtually all members of the proposed class would be subject to Ontario jurisdiction. The defendants take no issue with the Canada-wide class definition. As I have already noted, the proposed class must be defined in objective terms and may well include individuals who, in the end, will have no claim. The question is whether the Australian freelancer, who did not opt out of this action, would be bound by the result elsewhere. That would be an issue for the foreign court in which the Australian freelancer brought proceedings. In my view, the possibility that such question might arise elsewhere with respect to an atypical class member cannot be sufficient to defeat this claim from proceeding in Ontario.
In short, the consequences of purporting to bind non-residents with a common issues judgment or settlement approval order should be considered after trial by another Court.
CLASS ACTIONS BIBLIOGRAPHY
These cases address jurisdiction over class action defendants and non-resident class members and issues arising from parallel proceedings across provincial borders. I have set them out in chronological order, but not because there is an evolutionary superiority in the later cases, and indeed, the first 10 cases alone contain all of the themes in the remaining. A running theme is that many core issues relating to the assumption of jurisdiction, choice of law, and judgment recognition continue to be deferred rather than resolved.