1. Introduction
Canada is widely recognized as a pro-arbitration jurisdiction, with the judiciary supporting the recognition and enforcement of international arbitral awards and upholding the autonomy of arbitral proceedings.Footnote 1 While Canadian governments (both federal and provincial/territorial) have consistently promoted a pro-arbitration stance for decades — such as through the accession to the New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards (New York Convention) and the adoption of the UNCITRAL Model Law on International Commercial Arbitration (UNCITRAL Model Law) — the action of the Canadian federal government, including the delayed ratification of the Convention on the Settlement of Investment Disputes between States and Nationals of Other States (ICSID Convention), reveal a more ambivalent approach.Footnote 2 This tension invites a closer examination of the contribution of Canadian judges within this framework, particularly in contexts where judicial oversight remains active, such as in investor-state arbitrations that do not take place at the International Centre for Settlement of Investment Disputes (ICSID). In the context of investor-state dispute settlement (ISDS) — particularly, under the North American Free Trade Agreement (NAFTA) Chapter 11 — the role of courts remains both significant and complex.Footnote 3 Unlike ICSID arbitration, which operates within a self-contained international legal framework, NAFTA arbitrations conducted under arbitration rules such as the United Nations Commission on International Trade Law’s (UNCITRAL) Arbitration Rules or the Permanent Court of Arbitration’s Arbitration Rules are subject to domestic judicial review — most notably, through the set-aside mechanism provided in Article 34 of the UNCITRAL Model Law. Footnote 4
The Canadian case is particularly compelling in this regard.Footnote 5 Although Canada ratified the ICSID Convention only in 2013 — nearly two decades after NAFTA Chapter 11 had come into force — claims under NAFTA continue to be governed by non-ICSID arbitration mechanisms.Footnote 6 This delayed accession is not merely a procedural detail; it underscores a deeper tension within Canada’s approach to investment treaty arbitration. By entering the ICSID system so late, Canada effectively prolonged its reliance on domestic courts for the review of arbitral awards, in contrast to the fully internationalized enforcement and annulment regime offered by ICSID. This divergence reveals a nuanced posture: while Canada publicly affirms its support for international arbitration, its delayed commitment to ICSID suggests a continued preference for retaining a measure of domestic judicial oversight over investment treaty disputes. At the same time, it is important to clarify the scope of the article. The focus on NAFTA Chapter 11 awards is deliberate. While the UNCITRAL Model Law governs both international commercial arbitration and non-ICSID investment arbitration, the only set-aside proceedings concerning investment treaty awards in the Canadian context have arisen in relation to NAFTA Chapter 11. These cases therefore provide a uniquely concentrated dataset through which to explore the Canadian judiciary’s approach. Moreover, because NAFTA awards are not subject to the ICSID Convention’s self-contained annulment system, they fall under the UNCITRAL Model Law and the New York Convention, where domestic courts retain a greater role in supervising arbitral awards. This makes NAFTA cases especially valuable for analyzing how Canadian judges understand and exercise deference towards investment treaty arbitration as a distinct legal order.
I contend that the legacy of this policy not only impacts how Canadian judges perceive their role in the international investment regime but also invites closer scrutiny of how they conceptualize international investment law and investment treaty arbitration as a distinct legal order. Rather than viewing the application for setting aside before domestic courts merely as a procedural safeguard, this article approaches them as a window into how judges interpret their role vis-à-vis the international legal framework composed of investment treaty arbitration and the UNCITRAL Model Law. This is particularly important in the Canadian context, where the delayed ratification of the ICSID Convention and continued reliance on non-ICSID mechanisms expose a deeper tension between transnational and domestic legal orders. Unpacking how judges navigate this tension — how they situate investment treaty arbitration within the hierarchy of legal norms and reconcile it with domestic legal principles governing applications for the setting aside of international awards — is essential. It reveals not only the contours of Canada’s pro-arbitration stance but also the interpretive frameworks that shape the country’s engagement with investment treaty arbitration. This, in turn, has significant implications for how Canada is perceived as a seat of arbitration — not just in terms of neutrality and legal certainty but also in terms of how its judiciary represents and interacts with the international legal order itself.
To do so, this article examines how Canadian judges have approached these set-aside proceedings, with a particular focus on the standard of review applied. This examination is essential to unpack the degree of deference that courts accord to international investment law and investment treaty arbitration as a distinct legal order. In particular, the way in which judges interpret and apply the grounds for annulment under Article 34 of the UNCITRAL Model Law reveals not only their understanding of the arbitral process but also their broader posture towards the legitimacy and autonomy of the ISDS framework. To draw conclusions in this regard, the article analyzes the ten NAFTA Chapter 11 arbitral awards that were challenged before Canadian courts through applications to set aside pursuant to Article 34 of the UNCITRAL Model Law. Footnote 7 This analysis draws on Martti Koskenniemi’s work on the language and structure of international legal argument to deconstruct the ten decisions.Footnote 8 The study reveals the underlying assumptions on judicial deference shaping their reasoning. It finds that, while Canadian judges promote a pro-arbitration stance, this does not necessarily imply deference in the deeper sense of ‘respecting’ the international legal order constituted by investment treaty arbitration and the UNCITRAL Model Law. Deference, in this context, should be understood as respect for the international character of the UNCITRAL Model Law and the need to align with the global consensus on how the grounds for this exclusive recourse against non-ICSID awards are interpreted and applied by judges in other UNCITRAL Model Law countries.
The argument advanced in this article is that judicial deference to investment treaty awards requires recognition of the international character of the UNCITRAL Model Law. This, in turn, translates into the adoption of a standard of review that aligns with how courts in other UNCITRAL Model Law jurisdictions interpret and apply it. This does not imply that courts must invariably uphold awards; rather, I contend that deference in this context entails a principled internationally informed approach to review. Setting aside an award — when done on legitimate and narrowly defined grounds consistent with the global consensus on the interpretation of the UNCITRAL Model Law — can, in fact, reinforce the integrity of the arbitral process and strengthen a genuinely pro-arbitration legal framework. While judicial restraint may often be perceived as pro-arbitration, it is the underlying rationale for that restraint that ultimately determines whether a judicial posture should be characterized as deferential or genuinely pro-arbitration.Footnote 9
In this regard, I argue that Canadian judges have at times demonstrated limited or even timid deference to this international framework. In short, pro-arbitration does not mean deference. This is reflected in their reliance on domestic administrative law analogies when applying the standard of review under Article 34 of the UNCITRAL Model Law — an approach that risks distorting the distinct legal nature of international arbitration. Moreover, judges have often failed to engage with comparative case law or to utilize interpretive tools provided by UNCITRAL, such as the Case Law on UNCITRAL Texts (CLOUT) database and the Analytical Commentary, which are designed to promote a harmonized global understanding of the UNCITRAL Model Law. Footnote 10 Yet the explanation for this judicial posture may be far more mundane: it is possible that judges — who often rely heavily on the framing and arguments presented by counsel — are simply unaware of the interpretive significance of Article 2A of the UNCITRAL Model Law, which emphasizes the importance of international uniformity and the observance of good faith. Although foundational, this article of the UNCITRAL Model Law is often overlooked in practice, perhaps because it was only introduced in the 2006 version of the UNCITRAL Model Law.
In this article, I have chosen to rely on the set-aside decisions of NAFTA Chapter 11 awards as a basis for the analysis, operating under the assumption that judges had reasonable access to the UNCITRAL Model Law and to the interpretive tools provided by UNCITRAL, including the CLOUT system.Footnote 11 This assumption is not meant to absolve or critique individual judges but, rather, to frame the doctrinal analysis within the institutional context in which these decisions are made. Before turning to the next sections, two preliminary warnings are necessary. First, I deliberately use the term “judges” rather than “courts” to emphasize the agency and discretion of the individual decision-makers. I reject the conflation of judicial policy and the deference shown by judges with that of courts as institutions. From a critical perspective, this conflation contributes to obscuring the distinct roles and responsibilities of the individuals within the legal system.
Second, my analysis focuses on the interface between investment treaty arbitration and the Canadian legal order. Hence, this article looks exclusively at set-aside proceedings of investment treaty awards in Canada, which have only arisen under NAFTA Chapter 11.Footnote 12 It does not aim to provide a comprehensive examination of set-aside decisions under Article 34 of the UNCITRAL Model Law, as other scholars have done in recent years.Footnote 13 Indeed, although the UNCITRAL Model Law governs both investment and commercial arbitration, and recent set-aside decisions occasionally cross-reference precedents from both contexts,Footnote 14 important distinctions remain. These cross-citations highlight areas of conceptual overlap, but they also underscore the need to account for the unique features of investment treaty arbitration, particularly its public law dimensions and international legal character. The distinct public law dimensions of ISDS warrant a separate analytical focus as my objective here is to provide an intelligible understanding of how Canadian judges engage with the unique legal framework posed by ISDS.
The analysis provided in this article offers an opportunity to reflect on the claimed pro-arbitration stance of Canadian courts and the extent to which judicial deference is genuinely accorded to the arbitral framework in the context of investment treaty arbitration.Footnote 15 By scrutinizing judicial reasoning in these cases, the article seeks to illuminate whether this deference is principled and consistent, or whether it reveals underlying tensions between domestic legal norms and the international arbitral order. Section 2 introduces the central debate that shapes current scholarship on the set-aside proceedings of international arbitral awards by Canadian courts. Section 3 presents the theoretical framework underpinning the analysis of key judicial decisions. Section 4 examines these decisions in detail, highlighting the tension between Canadian judges’ stated commitment to a pro-arbitration policy and their judicial practice in reviewing investment treaty awards. Section 5 explores the contradictions embedded in this judicial discourse, revealing the active — yet often understated — role of courts in shaping the contours of investment treaty arbitration. Section 6 synthesizes the article’s findings, demonstrating that Canada’s pro-arbitration stance does not consistently translate into judicial deference as respect in the context of set-aside proceedings. Section 7 concludes.
2. The debate on the degree of deference: the administrative law inclination of the Canadian judiciary
The fate of non-ICSID arbitral awards depends on the degree of deference accorded by domestic courts to the decisions rendered by international investment arbitrators, expressed in the standards of review applied by domestic judges to international investment awards. Existing studies have shown that the standards of review applied by domestic judges encapsulate their understanding of the interplay between international investment arbitration and their domestic legal orders.Footnote 16 This understanding can significantly influence the outcome of the review of international investment arbitral awards as it plays an important role in the degree of deference that they apply in set-aside cases.Footnote 17 As Gabrielle Kaufmann-Kohler and Michele Potestà have noted, domestic courts play a fundamental role in the non-ICSID context as they retain jurisdiction over setting aside proceedings.Footnote 18 The place of arbitration, therefore, carries significant implications for the predictability and scope of judicial review. As such, they argue that not only may the grounds for setting aside differ from one jurisdiction to another, but the experience and knowledge of domestic judges with regard to public international law may influence their review of an international investment award.Footnote 19 A complementary empirical study is provided by Joshua Karton and Samantha Wynne, who examined Canadian courts’ interpretation of international uniform law instruments, including the New York Convention and statutes modelled on the UNCITRAL Model Law. Their quantitative and qualitative analysis revealed that Canadian judges seldom pursue internationally uniform interpretation; instead, grounding their reasoning primarily in domestic legal categories. Although their dataset is now somewhat dated, their findings provide an important empirical corroboration for the argument advanced here: Canadian courts’ engagement with the UNCITRAL Model Law reflects a pragmatic domestication of international instruments rather than a wholesale embrace of the transnational interpretive community envisaged by UNCITRAL.Footnote 20
While the UNCITRAL Model Law provides a harmonized framework, its application is mediated through domestic judicial reasoning, which may vary in terms of legal culture, familiarity with public international law, and interpretive approach. The Canadian judiciary’s engagement — or lack thereof — with the international character of the UNCITRAL Model Law and the tools provided by UNCITRAL can influence not only the outcome of specific cases but also broader perceptions of Canada’s legal environment for investment arbitration. This makes the Canadian case a valuable lens through which to assess how domestic courts shape the interface between international and national legal orders in the arbitration context. The interaction between domestic law and international investment arbitration is multifaceted. This study shows that the interaction encompasses both procedural and substantive aspects as well as international and transnational dimensions. This complex interplay highlights the meaningful role that domestic courts play in shaping the outcomes of international investment awards in set-aside proceedings. In this regard, the Canadian case raises significant interest as a case study because of the arguably low level of deference that Canadian courts have shown towards international investment awards.
Existing scholarship regarding set-aside proceedings in Canada has largely focused on the administrative law inclination of the Canadian judiciary. Entering into a dialogue with Canadian reviewing courts, this body of scholarship is notably structured around the debate on the appropriate standard of review that judges have used as opposed to those that they should use for reviewing such awards. Scholars are divided on whether Canadian courts should adopt standards inspired by Canadian administrative law or the standards used for the review of international commercial arbitration. While some scholars, such as Céline Lévesque, think that using “correctness” as the standard to review investment awards for true questions of jurisdiction by Canadian courts is appropriate,Footnote 21 others, like Maureen Irish, argue that the use of this standard developed in the Canadian administrative law context is unsuitable for reviewing international investment awards. Scholars have also pointed to the tendency of Canadian courts to conflate investor-state arbitration and international commercial arbitration when reviewing international arbitral awards. For instance, Irish argues that the review of international investment awards should be distinguished from that of international commercial awards. According to her, while both should align with the global context in which the interpretation of grounds of review is found in relevant instruments, Canadian courts should be more cautious about ensuring consistency in the field of international investment arbitration. This is because the consistency expectation does not exist in the field of international commercial arbitration as opposed to international investment arbitration given its private nature.Footnote 22
At the same time, it must be recognized that neither the New York Convention nor the UNCITRAL Model Law makes an explicit distinction between international commercial arbitration and non-ICSID investment arbitration. This textual silence reflects, at least in part, the difficulty of drawing a clear line between the two. Investor-state arbitrations may arise from treaties, but they may also be grounded in contracts between investors and states or in investment authorizations issued by state entities — contexts that can blur the boundary between commercial and treaty-based disputes. Reasonable observers might therefore disagree about whether a given case belongs within ISDS or commercial arbitration. For this reason, there are both legislative and practical grounds for treating the two types of arbitration alike when their awards come before national courts. To argue otherwise requires confronting this overlap directly: either by showing how ISDS’s distinct public law features warrant different judicial treatment, notwithstanding the legislative texts, or by proposing reforms to the UNCITRAL Model Law or related legislation to expressly address non-ICSID ISDS awards.
Other scholars writing on the review of international investment awards in the Canadian context have discussed the broader environment in which set-aside proceedings occurs, highlighting the global conversation that should take place between reviewing courts in various jurisdictions. One conclusion from this scholarship is that Canadian courts have ignored the approaches adopted by the reviewing courts in other countries, despite the nature of the UNCITRAL Model Law and the tools provided by UNCITRAL to encourage such dialogue.Footnote 23 In this regard, Frédéric Bachand and Fabien Gélinas criticize the Canadian courts’ approach, arguing that not only does it mirror unsuitable standards developed in Canadian administrative law, but it also ignores what other reviewing courts in other jurisdictions are doing. According to Bachand, this approach goes against the spirit of harmonization found in the UNCITRAL Model Law, which should be interpreted following a so-called “internationalist” method.Footnote 24
This article breaks from existing scholarship on the review of international arbitral awards in the Canadian context by sidestepping the debate on the appropriate standard of review and approach to set-aside cases. Instead, it takes a critical legal approach by deconstructing the set-aside decisions of NAFTA Chapter 11 arbitral awards issued by Canadian courts. The article reveals that Canadian set-aside decisions often do not align with Canada’s pro-arbitration stance. While Canadian courts claim to support a pro-arbitration policy for investment treaty arbitration that aligns with the Canadian government’s policy, in reality, they show little deference to NAFTA Chapter 11 arbitration as an autonomous legal order from the Canadian legal order. This is evident in two takeaways. First, Canadian courts have interpreted set-aside grounds in strict isolation from the harmonization objectives of the UNCITRAL Model Law. Second, they have consistently defended their ability to set aside NAFTA Chapter 11 arbitral awards, relying on administrative law standards of review developed by the Supreme Court of Canada in the Canadian administrative law context. This article finds that the pro-arbitration stance adopted by Canadian courts does not equate to deference as respect to international treaty arbitration and the UNCITRAL Model Law. In short, pro-arbitration does not mean deference.
3. A critical framework to analyze Canadian decisions on annulment proceedings
This article focuses its analysis on the international investment awards initiated under NAFTA Chapter 11 that have been challenged before Canadian courts on set-aside grounds. When domestic judges review these awards, they are faced with the question of the extent to which the Canadian legal order should defer to international investment arbitration in set-aside proceedings. To a certain extent, the degree of deference that domestic judges accord to the arbitral awards reveals their vision of the role of the domestic legal order in international investment arbitration disputes. The degree of deference that domestic judges accord to arbitral awards reveals their underlying conception of the domestic legal order’s role within the broader framework of investment treaty arbitration.Footnote 25
In the Canadian context, this is especially visible in set-aside proceedings under the UNCITRAL Model Law, where judicial reasoning often reflects deeper assumptions about the relationship between national courts and investment arbitral tribunals. The choice of standard of review — frequently shaped by analogies to administrative law — signals more than just a legal preference; it encapsulates a judicial posture towards the nature of international investment law and investment treaty arbitration. While these choices may appear doctrinal, they are often influenced by subjective and institutional factors, including the judiciary’s familiarity with public international law, international investment law, and the extent to which UNCITRAL’s interpretive tools are known and utilized. The limited engagement with these tools suggests that deference is not merely about outcome (that is, whether an award is set aside) but also about the method and mindset through which judges approach their review function. Understanding how Canadian judges interpret the UNCITRAL Model Law’s setting-aside grounds is therefore essential — not only to assess the outcomes of specific cases but also to evaluate Canada’s broader suitability and predictability as a seat for non-ICSID investment treaty arbitration.
Apart from the reference to Canada’s so-called pro-arbitration policy, the decisions analyzed in this article do not contain any express mention of ideology or doctrine followed by Canadian courts in reviewing international investment awards;Footnote 26 hence, the need to critically analyze them to uncover these considerations. The omission from the courts to develop the intellectual foundations of their judicial policy is not surprising. The legitimacy of courts generally relies on the appearance of neutrality, the expertise of judges, and the determinacy and objectivity of the law as opposed to whatever subjective consideration motivates their decisions.Footnote 27 Emmanuel Gaillard shows that arbitrators are also driven by subjective considerations. He argues that arbitrators adhere to various “representations” of international arbitration.Footnote 28 One dominant representation is that of the “arbitral legal order,” defined by Gaillard as a transnational legal order that is autonomous of other legal orders.Footnote 29 According to this representation, he contends that arbitrators show little deference to domestic legal orders since the existence of international arbitration is deemed autonomous from the latter. The fundament of this representation of international arbitration “corresponds to the international arbitrators’ strong perception that they do not administer justice on behalf of any given State, but that they nonetheless play a judicial role for the benefit of the international community.”Footnote 30
Turning back to Canadian courts and their set-aside decisions, I contend that examining the notion of deference instead of focusing the debate on the appropriate standard of review is a fertile ground to capture how Canadian courts positioned themselves (and the Canadian legal order) in relation to international investment arbitration. This deconstruction effort relies on Koskenniemi’s work on the language and structure of (international) legal argument. I analyze the language and structure of set-aside decisions as indicators of how Canadian judges have approached the interplay between the Canadian legal order and international investment arbitration and how this interplay is translated into legal terms. Through this deconstruction process, the underlying set of assumptions relied on by Canadian judges are reconstructed intelligibly. The article contends that the degree of deference accorded by Canadian courts to international investment awards in set-aside proceedings is indicative of the judges’ perception of the authority and legitimacy of arbitrators’ decisions.Footnote 31 Adapting the words of Koskenniemi to this aim, I “shall look [to the language, the voice, and the discourse used by domestic judges] in order to see what kind of understanding it mediates of [international investment arbitration].”Footnote 32
In the next section, the set-aside decisions are systematically examined for elements of language and fragments of structure as traces of their understanding of international investment arbitration. The elements of language, such as the words used, reflect how they have translated international investment awards into an intelligible construction for the Canadian legal order. The structural fragments, such as the intellectual foundations on which they base their analysis, reveal the paradigm under which they operate. The analysis proceeds in chronological order and groups the set-aside decisions by trends.
4. From Metalclad to Bilcon and beyond: The Canadian annulment decisions of NAFTA Chapter 11 arbitral awards
Since the entry into force of NAFTA in 1994, ten NAFTA Chapter 11 arbitral awards have been challenged before Canadian courts through applications to set aside pursuant to Article 34 of the UNCITRAL Model Law (or the relevant federal or provincial legislation integrating the UNCITRAL Model Law) .Footnote 33 The legal framework for the review of NAFTA Chapter 11 arbitral awards by Canadian courts is established by the namesake chapter and the Canadian legislation governing international arbitration. Article 1136 of NAFTA contemplates the possibility of the judicial review of Chapter 11 arbitral awards, and Article 34 of the UNCITRAL Model Law provides the grounds available for setting aside an award.Footnote 34 Under Article 34(2) of the UNCITRAL Model Law, a domestic court may set aside an arbitral award when the applicant proves (1) the invalidity of the arbitration agreement; (2) the lack of proper notice of the arbitration proceedings; (3) the recognition that the award deals with a dispute not contemplated by, or not falling within, the terms of the submission to arbitration; and (4) an arbitral composition not in accordance with the agreement of the parties to arbitration.Footnote 35
In addition, under Article 34(2)(b) of the UNCITRAL Model Law, domestic courts may set aside the award if they find that (1) the subject matter of the dispute is not capable of being enforced under the law of its state and that (2) the award is in conflict with the public policy of its state.Footnote 36 While providing grounds for setting aside international awards, Article 34 of the UNCITRAL Model Law does not provide for any standards of review of international awards.Footnote 37 The determination of the applicable standard of review is left to domestic courts in accordance with the law of the place of arbitration.Footnote 38 At the same time, it is important to note that the New York Convention harmonizes only the grounds on which recognition or enforcement of awards may be refused (Article V), while it is silent on the substantive grounds for annulment, which remain governed by domestic arbitration law. This asymmetry means that, in set-aside proceedings under Article 34 of the UNCITRAL Model Law, Canadian legislatures and judges have broader discretion and are more likely to draw on their own legal traditions and “knowledge and experience,” even though both annulment and non-enforcement typically involve similar categories of grounds for review.Footnote 39
This section applies the critical analytical framework to the ten set-aside decisions of NAFTA Chapter 11 arbitral awards challenged before Canadian courts. It reveals that Canadian courts have adopted a contradictory discourse by stressing the pro-arbitration policy of Canada while showing little or timid deference to the legal framework created by investment treaty arbitration and the UNCITRAL Model Law.
A. Mexico v Metalclad: the first NAFTA award to be subject to a Canadian annulment proceeding
The Mexico v Metalclad decision is the first NAFTA award to be subject to a set-aside proceeding in Canada. It is also the only case in which a Canadian court (partially) set aside an arbitral award. In its decision, the reviewing judge emphasized a policy of judicial deference to international investment arbitral awards — notably, by discarding the use of administrative law analogies to determine the appropriate standard for the review of international investment awards given the international nature of NAFTA. However, the judge engaged in a substantive interpretation of NAFTA provisions and set aside the award on jurisdictional grounds, based on the finding of a public policy violation.Footnote 40 Some authors have noted that this decision “amounted to de novo review, equivalent to a full appeal.”Footnote 41 This decision presents a typical scenario of inconsistencies between what the reviewing judge says he is doing (the desire to show deference to the arbitral award) and what he ultimately does (to conduct a complete review of the award on its merits). As the analysis of the following cases shows, the conflict between what reviewing judges state and what they ultimately do in reviewing international investment awards is a recurring feature in Canadian set-aside decisions, albeit on different grounds.
B. After Metalclad: the judicial quest to find the appropriate standard of review
After the widely criticized Metalclad set-aside decision, Canadian courts embarked on a new path to determine the appropriate standard of review to apply to international investment awards. This quest was notably rooted in criticism of the previous decision. Consequently, subsequent decisions rejected the approach adopted in the Metalclad set-aside decision, instead relying on Canadian administrative law standards of review. This practice of referring to analogies found in Canadian administrative law to determine the appropriate standard of review of international investment awards became an innovation of Canadian courts in set-aside proceedings that was subsequently also applied to the review of international commercial awards.Footnote 42 As pointed out by Denis Lemieux and Sabine Mekki, the administrative law approach taken by Canadian courts marked the intrusion of domestic law in an international investment dispute,Footnote 43 highlighting the indeterminate nature of set-aside grounds found in relevant instruments.
Starting with Canada v S.D. Myers, Canadian judges began to systematically adopt the structure and language of domestic administrative law — notably, by importing the standard of review framework developed by the Supreme Court of Canada in landmark decisions such as Pushpanathan v Canada (Minister of Citizenship and Immigration) Footnote 44 and Dunsmuir v New Brunswick. Footnote 45 Subsequent decisions constantly referred to the concept of a “high degree of deference” (developed by the Supreme Court of Canada in Pushpanathan) to illustrate the attitude that they should adopt towards international investment awards. However, following the Supreme Court’s decision in Dunsmuir, there was a marked shift: judges increasingly applied the “correctness” standard as an analytical lens for assessing these awards. Despite this evolution in the standard of review — from “deference” to correctness — their reasoning remained firmly anchored in administrative law analogies, underscoring a persistent reliance on domestic public law frameworks to navigate the complexities of international investment arbitration.
C. Feldman and Bayview: reinforcing the administrative law turn to annulment review
Notwithstanding the administrative law turn to set-aside review, subsequent decisions continued to emphasize the necessity of adopting a policy of judicial deference towards international arbitration and awards. In this regard, the reviewing courts in United Mexican States v Marvin Roy Feldman Karpa and Bayview Irrigation District No 11 v United Mexican States stressed the need to use their power of review sparingly, given the notions of “international comity” and the “reality of the global marketplace.” These notions highlight the international nature of international investment arbitration. Despite the reviewing judges stated emphasis on judicial deference, the Feldman and Bayview decisions effectively collapsed the distinction between NAFTA Chapter 11 arbitration and domestic administrative adjudication. By applying the standard-of-review framework developed by the Supreme Court of Canada for administrative decisions, these courts treated international arbitral tribunals as being analogous to “specialized tribunals” within the Canadian administrative law context.Footnote 46 This approach goes beyond a mere interpretive preference; it reflects a deeper conflation of legal categories. The standards in question were crafted to govern judicial oversight of domestic administrative bodies, not international arbitral tribunals operating under a distinct statutory and treaty-based regime. As such, their transposition risks undermining the legal autonomy of international arbitration and distorting the enforcement framework established for arbitral awards.
In Feldman, the reviewing judge relied on the structure of reasoning and language found in Pushpanathan to assert “a high degree of deference for decisions of specialized tribunals generally and for awards of consensual arbitration tribunals in particular.”Footnote 47 In Pushpanathan, the Supreme Court of Canada listed four factors to determine the degree of deference to be accorded to a “tribunal subject to judicial review,” which include: (1) the presence of a privative clause; (2) the expertise of the tribunals; (3) the purpose of the governing legislation; and (4) whether the issue to be determined is a question of law or fact.Footnote 48 The reviewing judge in Feldman directly applied the test set out by the Supreme Court of Canada and used the exact same language, without any adaptation.
With regard to the first factor — that is, the presence of the privatize clause — the reviewing judge referred to the ICSID Additional Facility Rules to conclude that “there does not exist what has sometimes been referred to as a ‘full’ privative clause.”Footnote 49 In the Canadian administrative law context, privative clauses are found in acts of Parliament or provincial statutes to remove all or part of the decisions of administrative bodies or specialized tribunals from the oversight and control of courts.Footnote 50 In Feldman, the reviewing judge equated the ICSID Additional Facility Rules with privative clauses, ignoring the international nature and scope of the rules. With regard to the second factor — namely, the expertise of the tribunal, the reviewing judge noted the expertise of the arbitrators appointed pursuant NAFTA Article 1124(4), which he decided militates for increased deference in Canadian administrative law.Footnote 51 With regard to the last two factors, the reviewing judge took into consideration the counsel of the investor’s assertion that the “statutory purpose [of NAFTA Chapter 11] is to take the resolution of these [investment] disputes out of the hands of domestic courts” and that the “matters decided by the tribunal in this case were heavily fact laden.”Footnote 52
Just as in Metalclad, the reviewing judge in Feldman adopted a reasoning that ultimately contradicted the position stated at the outset. However, the comparison warrants nuance: unlike Metalclad, which was not framed in administrative law terms, Feldman was explicitly analyzed through that lens. By assimilating international investment arbitration to specialized domestic tribunals, the reviewing judge in Feldman refused to acknowledge the autonomous and distinct character of NAFTA Chapter 11 arbitration. This approach, as noted by Bachand — now a judge at the Quebec Court of Appeal — raised a “red flag” concerning Canada’s commitment to a pro-arbitration policy.Footnote 53
Similarly, in Bayview, the judge relied on the reasoning found in the Feldman set-aside decision to conclude that a “high degree of deference” should be accorded to international awards reviewed by a domestic court.Footnote 54 While referring to the two new standards developed by the Supreme Court of Canada in Dunsmuir,Footnote 55 the reviewing judge slightly nuanced the role of courts with regard to international awards compared to decisions of administrative tribunals. The judge stated that the “role on judicial review is not to conduct a hearing de novo of the merits of the Tribunal’s decision on jurisdiction. The standard of review is narrower in scope than that governing the review by a domestic court of the decision of a domestic administrative tribunal.”Footnote 56 Even though, in Bayview, the Court slightly adapted the standard of review developed in Dunsmuir, it still relied on the same structure of reasoning and imported the same language that was adopted by the Supreme Court of Canada in this landmark decision.
The Feldman and Bayview decisions underscore a significant disconnect between the Canadian judiciary’s professed pro-arbitration policy towards international investment arbitration and the doctrinal reasoning it actually employs. While courts claim to support a pro-arbitration stance, their persistent reliance on administrative law standards of review reveals a deeper structural issue. This is not merely a matter of favouring one interpretive approach over another within a pro-arbitration framework. Rather, it reflects a more fundamental misalignment: the transposition of doctrines from administrative law into the international arbitration context risks distorting the legal architecture specifically designed for arbitral enforcement. Treating international arbitral awards as if they were administrative decisions undermines the distinct legal status of arbitration and erodes the integrity of the statutory framework governing recognition and enforcement.
Since Feldman and Bayview, Canadian courts have continued to apply domestic administrative law analogies,Footnote 57 thereby entrenching a practice that contradicts both the spirit and the structure of international arbitration law and further distancing judicial reasoning from the stated pro-arbitration policy.
D. Cargill: the reconciliation of a tension between the domestic courts “oversight power” and the deference to NAFTA tribunals
In Mexico v Cargill, Incorporated, Justice Kathryn Feldman of the Ontario Court of Appeal dismissed the application to set aside the NAFTA Chapter 11 arbitral award.Footnote 58 In this landmark decision, the Court of Appeal decided to “define the standard of review to be applied by a court in reviewing an arbitral decision on the grounds set out in” Article 34(2)(a)(iii) of the UNCITRAL Model Law (or its relevant integration provision).Footnote 59 The Cargill decision confirmed the transition initiated in Bayview towards the correctness standard, the standard of review developed by the Supreme Court in Dunsmuir. The judges who ruled on the Cargill set aside held that the correctness standard must be understood in the “[s]ense that the decision [of the NAFTA Chapter 11 tribunal] ha[s] to be correct[ed] in its determination that it ha[s] the ability to make the decision it made.”Footnote 60
Repeatedly throughout the decision, the reviewing judges in Cargill used a language supporting the “oversight power” of reviewing courts over NAFTA Chapter 11 arbitral awards. According to her reading of Article 34(2)(a)(iii) of the UNCITRAL Model Law, “courts should interfere only sparingly or in extraordinary cases, [but a] … court must have some basis to test whether the panel acted beyond its jurisdiction.”Footnote 61 In other words, the Court of Appeal questioned the degree of deference that Canadian courts should accord when reviewing international investment awards. According to the Cargill judges, the starting point of this reasoning is found in the drafting of Article 34(2)(a)(iii) of the UNCITRAL Model Law, which is used for “determining the appropriate standard of review to be applied” to NAFTA Chapter 11 arbitral decisions.Footnote 62 Pursuant to the Court of Appeal’s interpretation of the UNCITRAL Model Law, the reviewing judges in Cargill recognized the “oversight power” of Canadian courts over international investment awards. This reasoning echoes the judicial review power vested to Canadian superior courts over specialized tribunals in administrative law.Footnote 63
After having considered criticisms of the use of “various ill-defined standards of review” for international investment arbitration by Canadian courts in S.D. Myers, Feldman, and Bayview,Footnote 64 the reviewing judge rejected the idea of directly importing the standards of review developed by Canadian courts in the domestic context and especially the standards of review developed by the Supreme Court of Canada in Dunsmuir. Footnote 65 The reviewing judge stated: “I also agree that importing and directly applying domestic concepts of standard of review, both from administrative law and from domestic review by appeal courts of trial decisions, may not be helpful to courts when conducting their review process of international arbitration awards under art. 34 of the UNCITRAL Model Law.”Footnote 66 Despite this caveat reiterated several times in Cargill, the decision nonetheless contains many references to the reasoning of the Supreme Court of Canada in Dunsmuir. Footnote 67 This is particularly true in the judge’s finding that correctness is the applicable standard for the review of international awards regarding the “true question of jurisdiction”Footnote 68 because “the decision [of the NAFTA Chapter 11 tribunal] had to be correct in its determination that it had the ability to make the decision it made.”Footnote 69 In sum, the structure of the legal reasoning of the Ontario Court of Appeal in Cargill regarding the determination of the appropriate standard for the review of NAFTA Chapter 11 arbitral awards remains embedded in Canadian administrative law analogies and the structure of reasoning.
Indeed, by referring to the notion of the “true question of jurisdiction” as developed in Dunsmuir, the reviewing judge applied Supreme Court of Canada Justice Louis LeBel’s finding that courts have to take a “narrow view of what constitutes a question of jurisdiction and to resist broadening the scope of the issue to effectively decide the merits of the case.”Footnote 70 Drawing from this assertion, the reviewing judge concluded that, in the “international arbitration context,” this “approach is magnified.”Footnote 71 In other words, pursuant to Article 34(2) of the UNCITRAL Model Law, the “role of the reviewing court is to identify and narrowly define any true question of jurisdiction.”Footnote 72 Ultimately, the reasoning of Feldman J led to a standard of review akin to the one developed by the Supreme Court of Canada in Dunsmuir — the latter being “magnified” — in the sense that, by concluding that if there is a “true question of jurisdiction,” no deference shall be accorded to the NAFTA Chapter 11 tribunal if it decided outside the scope of its jurisdiction.Footnote 73 Put simply, the question is whether the investment tribunal was correct in determining its jurisdiction over the claim or whether it was not.
The reviewing judge in Cargill decided that it could not overrule its “oversight power” over the “true question of jurisdiction.”Footnote 74 In other words, the Cargill decision highlights a source of tension in the exercise of determining the appropriate standard of review between the “oversight power” of Canadian courts over arbitral awards pursuant to Article 34(2)(a)(ii) of the UNCITRAL Model Law and the expertise and authority of NAFTA Chapter 11 tribunals to decide over the merits of the investment dispute. The reviewing judge referred to this tension as follows: “One challenge for a reviewing court is to navigate the tension between the discouragement to courts to intervene, on the one hand, and on the other, the court’s statutory mandate to review for jurisdictional excess, ensuring that the tribunal correctly identified the limits of its decision-making authority. Ultimately, when deciding its own jurisdiction, the tribunal has to be correct.”Footnote 75 Ultimately, the Cargill decision showcases an exercise of reconciliation between these conflicting powers. As other judges did in other cases, the reviewing judge found the source of reconciliation in drawing from the standards of review developed in the domestic administrative law context. However, this decision uniquely sourced reconciliation by proposing an adaptation of the correctness standard to the context of international investment awards.Footnote 76
The reasoning in the Cargill decision clearly expresses the reviewing judge’s understanding of the interplay between domestic courts and international investment arbitration. He recognized that NAFTA Chapter 11 tribunals are the appropriate forum for adjudicating the merits of the investment dispute, given that it has the authority and expertise over the dispute and that its role is to “intervene only sparingly or in extraordinary cases” if the “panel has exceeded its jurisdiction.”Footnote 77 Nonetheless, the reviewing judge chose to draw on Canadian administrative law analogies to determine that correctness is applicable to the review of NAFTA Chapter 11 arbitral awards and according to which “no deference per se is owed to the tribunal under review.”Footnote 78
E. Mobil, Bilcon, Burr, and Nelson: a continuity of the Cargill precedent
Canadian reviewing courts in Mobil, Bilcon, Burr, and Nelson followed the approach developed by Cargill and applied correctness as the standard of review applicable to NAFTA Chapter 11 awards.Footnote 79 In these three decisions, the reviewing judges dismissed the applications to set aside the NAFTA Chapter 11 arbitral awards. The reasoning of the Court in the judicial review of Bilcon is worth noting here. The reviewing court agreed with the more recent precedent established in Cargill that correctness was the appropriate standard of review for NAFTA Chapter 11 arbitral awards as it was adapted to the NAFTA context.Footnote 80 However, in an obiter dictum, the judge went further by engaging in a “dialogue”Footnote 81 with the Bilcon arbitrators. The judge held that the majority’s decision “raised significant policy concerns” by considering “its effects on the ability of NAFTA Parties to regulate environmental matters within their jurisdiction,” which would potentially “chill” the environmental assessment in Canada.Footnote 82 While the content of this obiter sent an important message to the NAFTA tribunals and the government of Canada, the Court refrained from any intervention and did not set aside the decision.Footnote 83
More recently, in Clayton v Canada (Attorney General), in an application for setting aside the damages award, the Ontario Court of Appeal recognized the “strict limits on the ability of courts to interfere with arbitration awards” and concluded that the standard of review for true questions of jurisdiction is correctness.Footnote 84 According to Justice Grant Huscroft,
[18] [t]here is no room for doubt as to this court’s approach to commercial arbitration: review under s. 34(2)(a)(iii) of the Code or s. 46(1)(3) of the Arbitration Act is not an appeal. It is not an occasion for courts to review final and binding arbitration awards for either correctness or reasonableness. In order to intervene, the court must identify what Cargill described as a “true jurisdictional question”. That term, and in particular the modifier “true”, is intended to emphasize the limited scope of judicial oversight. The modifier “true” is necessary because of the problems inherent in the concept of jurisdiction itself, in particular its amenability to manipulation.
[19] Courts must not permit final and binding arbitration awards to be undermined by spurious jurisdictional arguments. The limited review contemplated by s. 34(2)(a)(iii) of the Code cannot be permitted to expand beyond its legitimate boundaries.
5. What do Canadian annulment decisions reveal of Canada’s relationship with the international dispute settlement system?
This section reflects on the elements of language and fragments of structure found in the decisions analyzed. It highlights four sets of references relied on by Canadian judges that are discussed in this section: (1) the analogies with Canadian administrative law; (2) the references to previous decisions over applications to set aside international commercial awards (that is, between two private parties) pursuant to Article 34 of the UNCITRAL Model Law; (3) the references to the specific nature of international investment arbitration and the NAFTA context; and (4) the references to the UNCITRAL Model Law’s international origin and general principles.
A. The analogies to Canadian administrative law
The first set is made of references and analogies to Canadian administrative law. These analogies often take the form of references to the standards of review developed by the Supreme Court of Canada in the context of the review of specialized tribunals’ findings or agencies’ decisions by superior courts. Since Canada (Attorney General) v S.D. Myers, all subsequent reviewing judges who have been called upon to review a Chapter 11 arbitral award pursuant to Article 34(2)(a)(iii) of the UNCITRAL Model Law have used analogies with Canadian administrative law. Indeed, the courts in S.D. Myers and Feldman referred to the Pushpanathan’s four factors to be considered when determining the three existing standards of review that the courts should apply in the Canadian administrative law context — namely, correctness, reasonable simpliciter, or patent unreasonableness. Moreover, the courts in Bayview referred to the framework developed by the Supreme Court of Canada in Dunsmuir, and the Court in Cargill applied a “magnified” version of the correctness standard of review developed by LeBel J in Dunsmuir. After the ruling in Cargill, the subsequent judges applied the “magnified” version of the correctness standard for the review of the NAFTA Chapter 11 arbitral awards in Mobil, Bilcon, Burr, and Nelson.
It is interesting to note that, until Cargill, the references made by Canadian judges to the administrative law standards of review have evolved over time alongside the development of new analytical frameworks for determining the appropriate standard of review by the Supreme Court of Canada. Interestingly, after Cargill, judges have continued to refer to the correctness standard of review developed in Dunsmuir,Footnote 85 albeit “magnified” in the context of Article 34(2)(a)(iii),Footnote 86 notwithstanding the issuance of Vavilov in 2019, which “[abolished] the shibboleth category” of “true questions of jurisdiction.”Footnote 87 Irish has remarked that the change that occurred in Canadian administrative law with Vavilov in 2019 is “especially unsuitable for reviewing international arbitral awards” since it “no longer recognize[d] true questions of jurisdiction as a category for correctness.”Footnote 88 Her article concluded based on this remark that “[f]uture Canadian courts will need to decide how much of the Cargill decision survives this change.”Footnote 89 In this regard, it appears that Canadian judges in the judicial review of the NAFTA Chapter 11 arbitral award United Mexican States v Burr, which was rendered one year after Vavilov, in 2020, decided to completely ignore the Vavilov decision and to continue to cite the reasoning found in Cargill, which is based instead on Dunsmuir. The Government of the United Mexican States v Nelson decision, rendered in 2022, indirectly refers to the reasoning found in Cargill, by referring to a set-aside decision of an international commercial award.Footnote 90
This choice by Canadian judges is representative of the consistent adherence of the Canadian judiciary to the standard developed by Feldman J after Cargill. Indeed, as pointed out by Lévesque in her thorough study of the correctness standard of review applied by Canadian judges in decisions over applications to set aside NAFTA Chapter 11 arbitral awards published in 2014, the choice of this standard “appears to strike the right balance between preventing courts from overreaching while exercising the duty of control under the Model Law.”Footnote 91 On the other hand, as Irish commented, “[t]his approach is unfortunate, as it invites a court to find a wide description of the matter at issue and then fall into an overly deferential analysis of a tribunal’s actual decision.”Footnote 92
Thus, with respect to the first set of references, Canadian judges who have been called upon to decide over applications to set aside NAFTA Chapter 11 arbitral awards pursuant to Article 34(2)(a)(iii) of the UNCITRAL Model Law have referred to Canadian administrative law analogies not with a sense of obligation but, rather, with the aim of relying on the most appropriate standards of review to the set aside of arbitral award context, which they have thus far considered to be found in Dunsmuir through the concept of the “true question of jurisdiction.”Footnote 93 Moreover, since NAFTA Chapter 11 arbitral awards potentially have an “effect … on the ability of the NAFTA Parties to regulate [public importance] matters within their jurisdiction,” Canadian judges have been more prompt in using administrative law analogies since the administrative law framework is dedicated to the oversight of agencies and the decisions of specialized tribunals by superior courts in Canada.
As explained earlier, analogies with Canadian administrative law have been used by the reviewing judges to determine the appropriate standard of review for NAFTA Chapter 11 arbitral awards pursuant to Article 34(2)(a)(iii) of the UNCITRAL Model Law. Footnote 94 Importantly, such analogies did not occur before S.D. Myers in the context of the review of international commercial arbitral awards (that is, between private parties), pursuant to the same Article 34(2)(a)(iii) of the UNCITRAL Model Law. Footnote 95 This is an indicator that Canadian judges have distinguished international investment arbitration — separating matters of public importance from purely private matters — from the international commercial arbitration context.Footnote 96 In this context, Canadian judges, while acknowledging that the standards developed in the domestic context were not useful if directly applied,Footnote 97 have used the deference framework developed in the domestic context, which provided Canadian judges with an existing “mental construct” that they already knew and understood.Footnote 98 Moreover, by using administrative law analogies in the structure of the legal arguments regarding the determination of the appropriate standard of review to be applied, Canadian judges applied a degree of deference similar to the one they were applying in domestic law for administrative agencies and specialized tribunals, but “magnified” in the context of a judicial review of a NAFTA Chapter 11 arbitral award.Footnote 99 Notably, this structure of the legal arguments used by Canadian judges also indicates that they view NAFTA Chapter 11 tribunals as something akin to “specialized tribunals.”Footnote 100
B. The international commercial arbitration analogies
The second set of references include the findings of previous decisions over applications to set aside international commercial awards (that is, between two private parties) pursuant to Article 34 of the UNCITRAL Model Law. Several of these decisions are consistently cited by Canadian judges called upon to decide over applications to set aside NAFTA Chapter 11 arbitral awards.Footnote 101 The most cited judicial reviews of international commercial arbitral awards by Canadian judges reviewing NAFTA Chapter 11 arbitral awards are Quintette Coal v Nippon Steel Corporation and Corporacion Transnacional de Inversiones v Stet International. Footnote 102 These decisions notably established a “powerful presumption” that “international arbitral tribunals act within their jurisdiction and that as a matter of respecting international comity and the global marketplace, domestic courts should use their powers to interfere only sparingly.”Footnote 103 This set of references were used by Canadian judges to determine and assess their expected adjudicating behaviour in deciding over applications to set aside international arbitral awards pursuant to Article 34(2)(a)(iii) of the UNCITRAL Model Law. Footnote 104 These references echo Canada’s pro-arbitration policy. For instance, when referring to Quintette Coal and Corporacion Transnacional, Canadian judges reviewing NAFTA Chapter 11 arbitral awards have used a language acknowledging the necessity for “international comity,” “predictability in decisions,” and “respect for autonomy of the parties’ chosen panel.”Footnote 105
An instructive development is found in Vento Motorcycles, Inc v Mexico, where the Ontario Court of Appeal was asked to determine whether the composition of the arbitral tribunal or the arbitral procedure was in accordance with the agreement of the parties, pursuant to Article 34(2)(a)(iv) of the UNCITRAL Model Law. Notably, this ground for setting aside is less concerned with a “true question of jurisdiction” and more focused on procedural compliance and adherence to party agreement. In his reasons, Huscroft JA did not engage with the broader debate surrounding the appropriate standard of review and instead treated investment treaty arbitration as functionally equivalent to commercial arbitration. This was evident in his use of language that drew parallels between the two regimes and in his reliance on precedent from set-aside decisions involving commercial arbitral awards. According to Huscroft JA,
[35] [a]lthough the courts have a role in overseeing commercial arbitration, that role is strictly limited by art. 34(2) of the UNCITRAL Model Law to matters that are, in general, not concerned with the substance of the arbitration award. For example, an award may be set aside because of a party’s incapacity or the invalidity of the arbitration agreement itself (art. 34(2)(a)(i)), or because of a failure of notice or breach of procedural fairness (art. 34(2)(a)(ii)). In exceptional circumstances the court may consider the substance of an award, but only for the limited purpose of ensuring that the tribunal did not go beyond the issues the parties agreed to submit to arbitration (art. 34(2)(a)(iii)). As this court has emphasized repeatedly, courts have no authority to scrutinize an award for an error of law that is otherwise immune from appeal or review and use any such error as a pretext to set aside an award on the basis of “jurisdictional error”. See Clayton v. Canada (Attorney General), 2024 ONCA 581, leave to appeal to S.C.C. requested, 41473; Mensula Bancorp Inc. v. Halton Condominium Corporation No. 137, 2022 ONCA 769, leave to appeal to S.C.C. refused, 40546; and Alectra Utilities Corporation v. Solar Power Network Inc., 2019 ONCA 254, 145 O.R. (3d) 481, leave to appeal to S.C.C. refused, 38665.
[36] Although art. 34(2)(a) does not refer specifically to bias, Mexico accepts that the court can set aside an award on the basis of a reasonable apprehension of bias. This flows from art. 34(2)(a)(iv), which states that an award may be set aside if “the composition of the arbitral tribunal or the arbitral procedure was not in accordance with the agreement of the parties”. This provision gives effect to the requirement in art. 18 of “equal treatment” of the parties in the arbitration process: an arbitral procedure tainted by bias fails to treat the parties equally and so is not in accordance with their agreement. See also Jacob Securities Inc. v. Typhoon Capital B.V., 2016 ONSC 604, at para. 33; and Aroma Franchise Company Inc. et al. v. Aroma Espresso Bar Canada Inc. et al., 2023 ONSC 1827, at para. 29, rev’d on other grounds, 2024 ONCA 839.
In sum, Canadian judges are increasingly citing set-aside decisions from both investment treaty and commercial arbitration contexts while interpreting Article 34 of the UNCITRAL Model Law. Footnote 106 This cross-referencing reflects a growing judicial effort to develop a coherent interpretive approach under the UNCITRAL Model Law, even as it raises questions about the appropriateness of assimilating these distinct arbitral regimes.Footnote 107
C. The nature of international investment arbitration
The third set of references concerns the specific nature of international investment arbitration. While Canadian judges almost consistently have referred to the judicial review of international commercial arbitral awards Quintette Coal and Corporacion Transnacional, they have referred to the nature and specific context of international investment arbitration although they are not necessarily expressed directly. For example, in Metalclad, the judge referred to the investment law context by, notably, engaging with the findings of other NAFTA Chapter 11 claims adjudicated under the ICSID Convention. Footnote 108 In Bilcon, the judge expressed in an obiter dictum some “policy concerns” regarding the “ability of NAFTA Parties to regulate environmental matters.”Footnote 109 This obiter dictum reveals that Justice Anne Mactavish understood that matters of public importance and the actions and regulations of Canadian jurisdictions (whether federal or provincial) were being challenged and adjudicated outside the domain of the Canadian courts (although she did not set aside the award) and that this fact deserved to be expressed publicly by the Federal Court of Canada.
D. The international origin and general principles of the UNCITRAL Model Law
The fourth set of references relates to the UNCITRAL Model Law’s international origin and general principles. Although some Canadian judges reviewing NAFTA Chapter 11 arbitral awards have made some references to the specific international investment context, they have not, to this date, considered the international context of investment arbitration in a comprehensive and systematic manner. This is despite Article 2A of the UNCITRAL Model Law, the calls from recognized Canadian legal scholars for judicial internationalism, and all the tools made available by UNCITRAL to take into consideration transnational interpretative rules and “global consensuses.”Footnote 110 According to Bachand, the “duty” of domestic judges in set-aside proceedings is to consider the “global consensuses” and to apply transnational interpretative rules.Footnote 111 In the context where UNCITRAL places several tools at the disposal of the public — such as the CLOUT database — to disseminate information on court decisions and arbitral awards relating to the UNCITRAL Model Law, the quasi absence of references to foreign decisions on applications to set-aside international investment awards pursuant to Article 34(2)(a)(iii) of the UNCITRAL Model Law suggests that there exists a judicial policy shared among Canadian judges not to take into consideration the internationalized aspect of the arbitral awards in their interpretation of the UNCITRAL Model Law. Footnote 112
The only exception to this generalized attitude of Canadian judges towards the international origin and the general principles of the UNCITRAL Model Law is found in Cargill. In that decision, the judge referred to Dallah Real Estate and Tourism Holding Co. v Ministry of Religious Affairs of the Government of Pakistan, a decision rendered by the UK Supreme Court in 2011.Footnote 113 The UK Supreme Court used the approach “to address the issue de novo, rather than as a review of the decision of the tribunal.”Footnote 114 The judge in Cargill referred to a foreign court decision to distinguish the case from the one it had in bar.Footnote 115 In Dallah, the UK Supreme Court was asked to enforce the international award on the basis of Article V of the New York Convention, after having heard evidence and expert evidence. The Court ultimately “[decided] the validity of the agreement issue de novo … and concluded that the agreement was not valid, and therefore, the arbitration panel had no jurisdiction.”Footnote 116 In contrast, in Cargill, the “jurisdiction issue [was] quite different” since the application to set aside was under Article 34(2)(a)(iii) of the UNCITRAL Model Law. In this context, by distinguishing the case at issue with that of a foreign court, the Cargill judge reasoned that he did not have “to address the issue de novo.”Footnote 117 Although the reviewing court in Cargill referred to another domestic reviewing court, it remains a minor trend in the reviewed decisions.
6. The analysis of Canadian annulment decisions shows that a pro-arbitration policy does not translate into deference from the Canadian judiciary
The above analysis reveals the original voice of Canadian judges towards international investment arbitration, confirming the need to explore the role of the judiciary in contentious international proceedings. Indeed, Canada claims itself as a pro-arbitration jurisdiction that benefits from an independent and cutting-edge judiciary. However, the analysis shows that Canadian judges struggle, on the one hand, to uphold the pro-arbitration policy of the Canadian government and, on the other hand, to perform their reviewing power as established by the general principles of the UNCITRAL Model Law. Footnote 118 While the term “pro-arbitration” has a “wide range of meanings,” this article relies on George Bermann’s definition, according to which pro-arbitration is a “policy or practice … that favours the achievement of international arbitration’s purposes.”Footnote 119 In Canada, the pro-arbitration policy is supported by all branches of government. Parliament and several provincial legislative assemblies have implemented the UNCITRAL Model Law. The Canadian government has also entered into several international investment agreements that provide for ISDS mechanisms (for example, NAFTA). Arbitration taken in its broad sense is also supported by the Supreme Court of Canada. In its landmark decision Desputeaux v Éditions Chouette (1987), the Supreme Court of Canada accepted arbitration (in this case, internal commercial arbitration) as a possible means to settle disputes within the Canadian jurisdiction by referring to the relevant legislation governing arbitration.Footnote 120
Regarding international investment arbitration, Canadian courts have taken a pro-arbitration policy stance by accepting to leave decision-making responsibilities in the hands of arbitral tribunals and to ultimately “favour … the achievement of international arbitration’s purposes.”Footnote 121 Bachand provides an account of the “pro-arbitration” policy adopted by Canadian courts despite certain “red-flags” with regard to the review of international arbitration awards pursuant to Article 34(2) of the UNCITRAL Model Law from the early 2000s to 2012.Footnote 122 Yet the judiciary has almost consistently adopted the correctness standard of review for NAFTA Chapter 11 arbitral awards.Footnote 123 The use of this standard of review has been entrusted by Canadian judges reviewing power over international investment awards in the “rare circumstances” where they found that the NAFTA Chapter 11 tribunal was incorrect in the determination of the “terms of the submission to arbitration, … the scope of the submission to arbitration.”Footnote 124 This approach reveals the lack of deference that Canadian judges accord to international investment arbitration, albeit none of the judges that used this reasoning actually set aside a NAFTA Chapter 11 arbitral award. Following the reasoning in Cargill, which was followed by subsequent Canadian judges reviewing NAFTA Chapter 11 arbitral awards pursuant to Article 34(2)(a)(iii) of the UNCITRAL Model Law, a court intervention should only occur according to the following circumstances:
[I]f the submission to arbitration, agreed to by both parties, makes a claim for damages suffered in the years 2007 and 2008, and the tribunal awards damages for 2009 and 2010, that would be an “award … not falling within the terms of the submission to arbitration”. Another example might be where the “investment” made by an investor from one Party is located in Brazil — i.e., not in the territory of another Party — and the tribunal awards c. 11 damages for losses suffered by the investor in Brazil even though c. 11 defines an investment as being located in the territory of another Party to the NAFTA, i.e., Canada, the United States or Mexico.Footnote 125
In other words, an incorrect determination of the “terms of the submission to arbitration, … the scope of the submission to arbitration” by a NAFTA Chapter 11 tribunal should only occur in rare circumstances.Footnote 126 However, if such an incorrect determination was to occur, the current standard used by Canadian judges pursuant to Article 34(2)(a)(iii) of the UNCITRAL Model Law would lead to the arbitral award being set aside. Indeed, in the Canadian context, such an intervention has never occurred since judges have been using Canadian administrative law analogies for determining the appropriate standard for the review of NAFTA Chapter 11 arbitral awards.Footnote 127 In other words, even if Canadian judges never set aside a decision pursuant to the application of Canadian administrative law standards of review, it does not mean that they have shown deference to investment arbitration. By stubbornly applying Canadian standards, even though other standards were more appropriate and available, the Canadian judiciary has decided to favour its own legal order over the international arbitral regime.
Through their quest to identify the appropriate standard for the review of arbitral awards, Canadian judges have sought to identify the appropriate degree of deference that should be accorded to NAFTA Chapter 11 arbitral awards pursuant to Article 34(2)(a)(iii) of the UNCITRAL Model Law. The concept of deference is notably understood as the respect one institution shows for the functions and autonomy of another institution.Footnote 128 In the Canadian administrative law context, Lebel J has suggested a definition of the concept of deference as respect — that is, “the judicial response to the legislative choice to take a matter out of the hands of the courts and put it primarily in the hands of a specialized tribunal.”Footnote 129 LeBel J questioned the concept of deference as respect by answering the following question: “[R]espect for whom or what?”Footnote 130 For him, deference as respect implies that courts respect the expertise of the administrative bodies and their decision-making process.Footnote 131 Furthermore, according to LeBel J, the legislative choice to transfer decision-making responsibilities from the courts to administrative bodies and to the different roles of administrative bodies must be appreciated by the courts.Footnote 132
This article draws inspiration from the concept of deference as respect, as articulated by LeBel J, in exploring how judges conceptualize their role in reviewing investment treaty arbitration awards. However, in the investment arbitration context, the foundation for deference is more complex. Unlike administrative tribunals, where legislative assignment alone justifies deference, investment arbitration involves a dual consent structure: the state consents through a treaty, and the investor consents by initiating arbitration. This dual-layered consent imbues the arbitral process with a unique legitimacy rooted in both sovereign will and party autonomy. As such, judicial respect for arbitral decisions in this context must account not only for institutional expertise but also for the consensual and international legal foundations of the process. This distinction suggests that the normative basis for deference in investment arbitration diverges from that in administrative law, shifting from institutional design to a framework grounded in mutual consent and international legal commitments.
Applying the concept of deference as respect to the set-aside context of international investment awards before Canadian courts, I conclude that the reviewing judges have not shown any deference to NAFTA Chapter 11 arbitration. Indeed, the analysis demonstrates that Canadian judges have broadly rejected the idea of an autonomous “arbitral legal order.”Footnote 133 They consistently reject the “judicial internationalist” method when interpreting Article 34(2)(a)(iii) of the UNCITRAL Model Law,Footnote 134 although the international community generally agrees that this approach is commanded by Article 2A.Footnote 135 On the contrary, Canadian judges have broadly interpreted the UNCITRAL Model Law and the relevant legislation integrating its provisions through Canadian administrative law analogies, without referring to the UNCITRAL Model Law “in the spirit of international harmonization.”Footnote 136 As a result, the structure of their reasoning remains embedded in the Canadian administrative law reference, highlighting their understanding of international investment arbitration as subordinated to the review of Canadian courts in the context of set-aside proceedings in accordance with Canadian standards of review. This result is particularly revealing of the way in which Canadian judges conceive international investment arbitration law and the interplay between their conception of what is domestic law, on the one hand, and what is international and transnational law, on the other hand.Footnote 137 However, deference is not intended to amount to “subservience or blind reverence.”Footnote 138
On the contrary, Canadian judges have expressed concern about waiving their reviewing power under Article 34(2)(a)(iii) of the UNCITRAL Model Law. Although Canadian judges have contributed to the “pro-arbitration trend,”Footnote 139 they almost consistently have defended their ability to set aside NAFTA Chapter 11 arbitral awards — notably, through the reliance on the “magnified” correctness standard of review.Footnote 140 Canadian judges have not considered non-ICSID arbitration as an independent legal order from domestic courts in the context of non-ICSID set-aside proceedings of NAFTA Chapter 11 arbitral awards. This is supported by the observation that Canadian judges have consistently asserted, since Cargill, that an incorrect determination of the “terms of the submission to arbitration, … the scope of the submission to arbitration”Footnote 141 could lead to the intervention of Canadian judges and to the setting aside of NAFTA Chapter 11 arbitral awards. This finding is the result of a constant interpretation of Article 34(2)(a)(iii) of the UNCITRAL Model Law based on Canadian administrative law analogies since Cargill, eclipsing the harmonizing objective of the UNCITRAL Model Law.
Indeed, Canadian judges have referred to Canadian administrative law analogies since Cargill in the interpretation of Article 34(2)(a)(iii) of the UNCITRAL Model Law with the aim of implementing the Canadian judicial pro-arbitration policy. However, the implementation of this pro-arbitration policy has not resulted in deference towards NAFTA Chapter 11 arbitral awards. Deference would have been truly achieved if Canadian judges had not ignored the harmonizing objective of the UNCITRAL Model Law, which is notably encapsulated in the so-called “judicial internationalist” method of interpretation,Footnote 142 and if they had engaged in a genuine dialogue with arbitrators and domestic judges from other jurisdictions.Footnote 143 Deference, as reflected in judicial reasoning, can be understood not only as a matter of institutional restraint but also as an interpretive posture — one that, in the context of international investment arbitration, may benefit from greater attentiveness to the international character of the legal instruments involved.
When Canadian judges engage with the UNCITRAL Model Law, for instance, their interpretive approach could more fully reflect its international origin. This includes engaging with foundational materials such as the Report of the United Nations Commission on International Trade Law on its eighteenth session in Vienna on 3–21 June 1985, the Analytical Commentary prepared by the Secretary-General, and practical tools like the CLOUT database. These resources offer insight into the UNCITRAL Model Law’s intended uniformity and international coherence — features that may not always be fully captured through a purely domestic legal lens. Observing how judges navigate these materials can reveal much about their orientation towards international legal norms and the extent to which they internalize the UNCITRAL Model Law’s international logic.
7. Conclusion
As outlined at the outset, this article does not seek to critique the inconsistencies in Canada’s broader pro-arbitration policy. Rather, it examines how Canadian courts have responded to the legal and institutional ambiguity created by Canada’s support for international investment arbitration, particularly in the context of non-ICSID set-aside proceedings. Through an analysis of judicial decisions under Article 34 of the UNCITRAL Model Law, this article has shown that Canadian judges play a pivotal role in shaping the interface between investment treaty arbitration and domestic legal principles.
The case law surrounding NAFTA Chapter 11 set-aside proceedings reveals how Canadian judges perceive their role in the international investment regime and how they conceptualize international investment law and investment treaty arbitration. Canadian judges have asserted a judicial posture that reflects national legal values and procedural expectations. While Canadian judges often invoke a pro-arbitration stance, this does not necessarily translate into deference as respect towards investment treaty arbitration as a distinct legal order. Their decisions suggest a cautious approach, one that prioritizes judicial oversight and procedural integrity over the automatic endorsement of arbitral outcomes. This judicial stance has broader implications.
I conclude that, despite Canada’s international commitments, its judiciary continue to speak primarily from a domestic legal perspective. In doing so, they have yet to fully engage in a transnational judicial dialogue with other annulment bodies or foreign courts reviewing investment awards. This inward-looking approach may limit Canada’s influence in shaping the evolving architecture of investment treaty arbitration. The analysis invites reflection on whether this judicial posture is a product of deliberate governmental design. By leaving interpretive discretion to the courts, the UNCITRAL Model Law may have implicitly delegated the task of reconciling its pro-arbitration rhetoric with the realities of domestic legal control. As such, this article encourages further inquiry into the role of domestic judiciaries in mediating the relationship between international dispute settlement mechanisms and national legal orders.
In closing, it is important to acknowledge the limits of doctrinal analysis in fully capturing the motivations behind judicial reasoning. Empirical research — such as interviews with judges and counsel — could provide deeper insight into whether this lack of engagement arises from unawareness, institutional constraints, or other practical considerations. Such research would enrich the doctrinal critique advanced in this article by anchoring it in the lived realities of legal practice.
