The visibility and relevance of foreign state (or sovereign) immunity has grown significantly in recent years. States and state-related entities are playing a growing role in international investment and commerce, while seeking civil remedies against states in domestic courts is increasingly seen as an important tool in holding states accountable for torture or other breaches of human rights.
State immunity, in its most traditional formulation, is the rule that a domestic court will not implead a foreign state in its proceedings without the state’s consent. It is, in effect, the expression of judicial deference to the executive’s responsibility for international relations. A state properly requires compliance with its laws from its citizens, or other persons located or acting within its jurisdiction. By contrast, disputes between foreign states, equals under international law, are ordinarily mediated through diplomacy. However, the exceptions to state immunity have grown over time, so that the modern doctrine provides for several other exceptions besides waiver by the foreign state.
The courts of Quebec have played a particularly significant role in the development of state immunity in Canada. They have consistently decided more cases involving state immunity than courts in any other Canadian jurisdiction. They were the first (and arguably the only) in Canada to accept the “restrictive doctrine” of state immunity in customary international law, under which a state was not immune from proceedings relating to its commercial activities. The restrictive doctrine, in turn, formed the basis for the 1982 statutory codification, the federal State Immunity Act (“Act”).
Quebec courts have continued to punch well above their weight in generating jurisprudence under the Act, and the last few months have been no exception. Last week’s decision of the Quebec Court of Appeal in the case of Trudel v. Nahmiash (sub nom New Jersey (Department of the Treasury of the State), Division of Investment v. Trudel), 2009 QCCA 86 (“Trudel”) raised a number of important issues well worth exploring in further detail.
1. The facts
The dispute underlying Trudel initially arose from a number of class action lawsuits by investors against Nortel Networks in the U.S. and Quebec in 2004 and 2005. The plaintiff in one of the U.S. actions was the New Jersey Department of the Treasury, Division of Investment (“New Jersey”), an organ of the State of New Jersey and thus a “state” within the meaning of the Act.
Subsequent to a settlement in the U.S. actions in June 2006, the Quebec Superior Court approved the extension of the settlement to include the pending Quebec class action. Philippe Trudel and Bruce Johnston, the lawyers acting for the Quebec plaintiffs’ class, submitted their proposed fees; New Jersey, together with several other parties to related actions, intervened to make joint submissions questioning the amount of the fees. Trudel and Johnston, believing that the intervenors’ submissions were defamatory, then brought an action against the intervenors and their lawyer, Laurent Nahmiash.
The defamation action was filed in November of 2007. The plaintiffs Trudel and Johnston obtained leave to serve the originating process on New Jersey by registered mail. New Jersey promptly brought a motion to set aside service and, further, pleaded irrecevabilité du recours on grounds of state immunity.
The Superior Court set aside service on the basis that it failed to confirm to the requirements of the Act, but declined to decide the merits of the immunity claim on the basis that there were insufficient facts to determine at this stage whether immunity applied. The Court of Appeal reversed in part on the latter point, finding that immunity applied.
2. Service on a foreign state
Section 9(1) of the Act provides as follows:
9. (1) Service of an originating document on a foreign state, other than on an agency of the foreign state, may be made
(a) in any manner agreed on by the state;
(b) in accordance with any international Convention to which the state is a party; or
(c) in the manner provided in subsection (2) [service by diplomatic means]
9. (1) La signification d’un acte de procédure introductif d’instance à l’État étranger, à l’exclusion de ses organismes, se fait :
a) selon le mode agréé par l’État;
b) selon le mode prévu à une convention internationale à laquelle l’État est partie;
c) selon le mode prévu au paragraphe (2).
This section has raised some dispute in the caselaw because of the apparent ambiguity in the English version. “May” on its face seems permissive, suggesting that the rules of service on a state under the Act are not exclusive of those under the local rules of court. Several previous cases had touched on the issue: most agreed that s. 9(1) was mandatory, such that where the plaintiff serves an originating document by any other means, any judgment obtained against the state will be set aside.
The question had been considered in particular detail in Cegir Inc. v. Banque algérienne de développement,  R.J.Q. 1965 (C.S.), where the court contrasted the French and English versions of s. 9. As noted above, the French version states “signification se fait” (“service is made”), which, unlike the English version, clearly indicates that the service provisions are mandatory. The court in Cegir preferred the French version, which in its view better accorded with the intention of the legislator. Nonetheless, a subsequent case in Alberta, Ritter v. Donell, took the opposite view, holding (in obiter) that the English “may” must be read as permissive. The Superior Court’s decision in Trudel, which regrettably remains unreported, canvassed all the previous decisions on s. 9(1) and chose to follow Cegir.
Though the service issue was not appealed and therefore not addressed by the Court of Appeal, the meaning of s. 9(1) should, in my view, now be treated as conclusively settled. The arguments in favour of s. 9(1) being mandatory are compelling. First, with the greatest respect to the Alberta court, it is simply not correct that “may” in English is always permissive: as Sullivan and Driedger note, it may be permissive or mandatory, depending on context. Read together with the French version, the rules of bilingual interpretation would support the narrower, mandatory meaning. (It is telling that Ritter made no mention of Cegir, nor considered the French version of the Act: once again, perhaps, an example of the limitations imposed by the unilingual character of most of the English Canadian bar and bench, a perennial problem that seems to resist solution.) The surrounding context also supports the mandatory interpretation: s. 9(3)(c) of the Act specifically authorizes the use of the rules of court for serving agencies of a foreign state. The absence of any similar provision for foreign states in s. 9(1) implies that the rules of court may not be used. While s. 17 of the Act preserves the applicability of the local rules of court except where inconsistent, it is circular reasoning, as the court in Cegir noted, to use that section to interpret s. 9 – as the court did in Ritter. Section 17 is a residual provision, to be used only after interpreting and giving full effect to the Act.
3. When and how immunity is adjudicated
The Court of Appeal in Trudel reversed the Superior Court on the issue of when immunity is adjudicated, holding that being a matter of public order, immunity must be determined as early as possible on the basis of the pleadings. This point is, in fact, uncontroversial. Ordinarily state immunity will be raised by the defendant as an interlocutory motion at the earliest opportunity, and resolved at that stage.
Yet, the Supreme Court has held that even where the immunity plea is rejected on a preliminary motion, it may be raised again at a later stage of the proceedings. Clearly this implies that there may exist circumstances where the record is inadequate to establish whether immunity applies. The Court of Appeal in Trudel suggested similarly, stating that [my translation] “the judge seized with a motion pleading immunity on the part of the foreign State is required to decide the question, absent particular circumstances that do not apply here.” Arguably, then, the real difference between the Court of Appeal and the motions judge was less the stage when immunity is determined than the nature of the burden of proof to establish its (in)applicability.
The question of which party has the burden of proof with respect to state immunity is not fully clear. Pursuant to s. 3(2) of the Act, the court itself is charged with responsibility for determining whether state immunity applies, and with giving effect to it, even if the parties do not raise the issue. This might suggest neither party has the burden of proof as such. This was the position taken by the Supreme Court of Canada, albeit pre-Act, in Congo v. Venne, where it stated that whether a state act is commercial in nature “should be decided on the record as a whole without placing the burden … on either party.” The Court of Appeal in Trudel, like most of the more recent cases, took the view once the state has made out a prima facie case for immunity (in this case, establishing that the defendant is a “foreign state” within the meaning of the Act), the burden shifts to the plaintiff to establish that the criteria for an exception are met. However, a minority of cases appear to have taken the view that the burden remains with the state to demonstrate the non-applicability of any exceptions. Regardless of where the burden lies, the standard of proof is the ordinary civil standard. The Court of Appeal in Trudel, following the decision of the Ontario Court of Appeal in Schreiber v. Federal Republic of Germany (2001), 52 O.R. (3d) 577, held that a motion to dismiss based on state immunity need not meet the “plain and obvious standard”.
A somewhat unusual aspect of Trudel was the manner in which New Jersey raised its substantive immunity plea, as an motion for irrecevabilité du recours. For those unfamiliar with Quebec procedure, this bears some explanation.
Quebec civil procedure distinguishes between “declinatory exceptions”, which are preliminary motions to stay or dismiss based on absence of personal or subject-matter jurisdiction (compétence) [Quebec Code of Civil Procedure, arts. 163-164] and “exceptions to dismiss” (moyens de non-recevabilité), which are motions to stay or dismiss on other grounds: (i) res judicata or lis pendens, (ii) lack of legal capacity to sue or be sued, (iii) plaintiff’s lack of standing, and (iv) no reasonable cause of action [C.C.P., arts. 165-167]. The majority trend in previous Quebec cases has been to plead state immunity as a declinatory exception, thereby treating it as a question of compétence, not recevabilité.
Interestingly, common law cases addressing state immunity have more typically turned first to questions of personal jurisdiction, addressing immunity only once jurisdiction has otherwise been established, thereby suggesting that immunity is a question of recevabilité. However, the analysis in the common law jurisprudence is less clear, in part because civil procedure rules in the common law provinces treat exceptions based on irrecevabilité and subject matter jurisdiction similarly, or do not even distinguish them, while personal jurisdiction is addressed separately through the rules on service.
The Court of Appeal observed that ordinarily the motions judge has a margin of discretion in when exceptions based on recevabilité are adjudicated: they may be addressed as a preliminary motion, or deferred to the trial on the merits. By holding that state immunity must be addressed as soon as possible, arguably the Court may have in fact implied that the plea of immunity should have been treated as a matter of compétence, not recevabilité. Nonetheless, I am aware of no example where a court has rejected the plea of immunity solely on the basis that it was brought using the incorrect rule. In any event, as already noted, the court must raise immunity of its own motion even if the state has not made any submissions.
4. Scope of the commercial exception
The plaintiffs in Trudel argued several distinct exceptions to state immunity, which the Court of Appeal rejected. Their argument as to the applicability of the commercial activity exception is in my view the most interesting, as it raises a question that has received comparatively little jurisprudential attention.
The commercial activity exception, as previously noted, codifies the restrictive doctrine of state immunity. The restrictive doctrine drew a distinction between acts jure imperii (acts of sovereign power) and acts jure gestionis (private or commercial acts), granting immunity only to the former. According to the restrictive doctrine, where a state acts like a private commercial businessperson or investor, it should be subject to the same liabilities as any private party.
According to the Supreme Court, determining the application of the commercial activity exception requires a two-step analysis: first, determining the nature of the activity, and second, whether the proceedings were “related to” the activity. Most of the caselaw on the commercial activity exception has focused on whether the state activity at issue is “commercial”. The application of the exception in Trudel, however, turned solely on the second part of this test. The question at issue was, in effect, how far does “related to” extend?
The plaintiffs put forward a kind of “commercial-by-transitivity” argument: the defamation proceedings were “related to” New Jersey’s commercial activity because the allegedly defamatory statements were made in the context of separate proceedings where the commercial activity exception admittedly applied. The court rejected this argument, relying on the analytical framework used by the Ontario Court of Appeal in Bouzari v. Iran. In Bouzari, the plaintiff brought suit in relation to his imprisonment and torture by Iran. The plaintiff alleged his arrest resulted from Iranian government efforts to force him to pay a percentage of his consultant’s commission from a joint venture between the Iranian national oil company and a foreign consortium. In rejecting the application of the commercial exception in the circumstances, the court held that the ultimate intention or purpose of the state acts of torture did not make them “commercial” in nature; rather, the acts themselves must be commercial. The damages sought in the proceedings must be related to the alleged commerciality of the activity.
However, the proposition that the damages sought in the Canadian proceeding must relate directly to the commercial activity at issue raises a troublesome question: can the commercial activity exception properly apply to the enforcement of foreign judgments and arbitral awards? Might it not be argued – especially in light of their treatment (at least at common law) as contract debts for limitations purposes – that the proceedings to enforce them are distinct from the underlying activity that gave rise to the dispute?
As a practical matter, such an approach would render arbitral awards considerably more difficult to enforce, at least where the applicable state immunity statute lacks a specific exception for arbitration (as is the case for the Canadian Act). Moreover, numerous cases have in fact enforced commercial arbitral awards against states and state-related entities, without any suggestion that the commercial activity exception is per se inapplicable. Furthermore, the suggestion that a foreign award or judgment should be treated as divorced from its factual underpinnings is something of a fiction in any event, to be discarded when inconvenient: for example, when determining whether the Federal Court has subject-matter jurisdiction over the enforcement of a foreign judgment, it is routine to look at the character of the underlying dispute.
Another approach to avoid this concern might be to regard immunity as inapplicable to a proceeding which relates directly to another, non-immune underlying proceeding only where it is a necessary or readily foreseeable corollary of that underlying proceeding – as is the case with proceedings to enforce a foreign arbitral award, but not, presumably, with a defamation action arising from statements made in an earlier proceeding.
Though more than twenty five years have passed since the Act came into force, Canadian courts continue to be faced with quite significant and fundamental questions as to its interpretation and application. For example:
- In the recent Quebec Superior Court decision in Kuwait Airways v. Iraq, the court held that the defendant was immune from the enforcement of a U.K. judgment, though the U.K. court had held the opposite in the underlying proceedings. Though immunity is a question of the law of the forum, should the enforcing court give any deference to the determination of the tribunal that renders a foreign judgment or arbitral award, where the foreign law is materially similar to the Canadian Act?
- Questions remain as to the scope of the service provisions: does the Act permit the initiation of proceedings ex parte, as is not uncommon in the enforcement of arbitral awards? While the UK practice permits it, the Canadian position remains unsettled.
- In execution proceedings, does the Act permit courts to issue pre-judgment attachment orders? And how does one execute on assets when it is unclear what property is in fact owned by the state (a not uncommon problem in states whose governments are less than transparent)?
- Numerous actions have sought to challenge the immunity of foreign states for torture carried out outside Canada – for example, in the notorious case of Zahra Kazemi, where a motion challenging the applicability of state immunity to an action against Iran is scheduled to be heard next week in Quebec Superior Court.
As the world gets smaller, these types of questions will only multiply. Those who enjoy those strange juxtapositions found at the interfaces of domestic and international law have much to look forward to.