Ontario – Award unreasonable where tribunal failed to follow binding law on frustration – #832

In Taseko Mines Limited v. Franco-Nevada Corporation, 2023 ONSC 2055, the Ontario Superior Court of Justice (Commercial List) granted an appeal from an arbitral award due to, among other things, the arbitrator’s failure to apply binding precedent on frustration of contract. Although the Court applied a deferential reasonableness standard, it concluded the arbitrator’s departure from binding jurisprudence rendered the award unreasonable.

Background – The case facts are complex. In the interest of brevity, I will focus on the key background and the main issue of interest in the arbitration for our purposes.

Taseko, a mining company holding rights to a mineral deposit in British Columbia, concluded a contract with Franco-Nevada under which it would sell Franco-Nevada a portion of the gold extracted from a mining project. Franco-Nevada was to benefit from a set price for a 40-year period. In exchange, Franco-Nevada provided a US$ 350 million deposit, a form of project financing permitting Taseko to develop its mine. 

The contract required the parties to submit disputes to arbitration. It allowed either party to appeal on a question of law or mixed fact and law. The contract’s governing law was the law of Ontario and any applicable federal laws.

The project required both provincial and federal government approval under applicable environmental assessment legislation. The authorities denied all of Taseko’s attempts to secure that approval, largely based on a First Nation’s opposition to the project. Taseko lost in its judicial review and appeals from the decision denying approval. Taseko continued to negotiate with the First Nation, but those negotiations ultimately failed.

Taseko took the position that the government’s refusal to issue the required environmental approvals had frustrated the contract. Franco-Nevada disagreed. The parties submitted the dispute to arbitration.

Arbitration – The main issue before the arbitrator was whether Taseko’s inability to obtain government approval for the mining project rendered the contract frustrated. The arbitrator found that it did. Based on the Court’s account of the award, the arbitrator undertook a detailed review of the facts underlying both the contract and the events leading to the denial of environmental approval. 

With respect to the legal test for frustration, the arbitrator viewed as the primary authority the Supreme Court of Canada’s decision in Naylor Group Inc. v. Ellis-Don Construction Ltd., 2001 SCC 58 [Naylor]. Naylor states, “frustration occurs when a situation has arisen for which the parties made no provision in the contract and performance of the contract becomes ‘a thing radically different from that which was undertaken by the contract’”. 

The arbitrator also instructed himself on the legal test by referring to the Supreme Court of Canada’s earlier decision in Peter Kiewit Sons’ Co. v. Eakins Construction Ltd., [1960] S.C.R. 361, which in turn relied on the House of Lords’ decision in Davis Contractors Ltd. v. Fareham Urban District Council, [1956] A.C. 696 (H.L.). In particular, the arbitrator relied on the reasons of Lord Reid (in concurrence), who stated that “on this view there is no need to consider what the parties thought or how they or reasonable men in their shoes would have dealt with the new situation if they had foreseen it. The question is whether the contract which they did make is, on its true construction, wide enough to apply to the new situation: if it is not then it is at an end”. This passage appears to have played a key role in the arbitrator’s analysis.

Franco-Nevada relied on a line of cases, including three cases from the Court of Appeal for Ontario, that hold frustration can only apply when the supervening event—the event that makes performance “radically different” from what the parties undertook in their contract—was unforeseen to the parties at the time of contract. Franco-Nevada also noted foreseeability was part of the reasoning in Naylor. This was important because, as Franco-Nevada argued, the government authority denying project approval was foreseeable at the time of contract. If so, Taseko’s frustration argument should fail.

The arbitrator determined that the Court of Appeal’s jurisprudence wrongly included foreseeability as part of the frustration analysis. He gave several reasons for this based in his understanding of the principles in play. He also found that the reference to foreseeability in Naylor was not key to the Supreme Court’s reasons since it found frustration did not apply on a different and independent basis. 

Appeal – The parties agreed that the reasonableness standard applied. The Court quoted from several decisions, including Canada (Minister of Citizenship and Immigration) v. Vavilov, 2019 SCC 65 [Vavilov], explaining the deferential standard it intended to apply in reviewing the award. 

Despite the deference the Court determined was due, it concluded that the arbitrator’s analysis was unreasonable for failing to follow the Court of Appeal’s binding precedents on the test for frustration. The Court noted that since the Supreme Court and Court of Appeal case law includes foreseeability as part of the frustration analysis, it was not open to the arbitrator to disregard it: 

[76] The first four of the five reasons (listed above in para. 25) given by the arbitrator for rejecting the approach taken to the role of foreseeability in the applying the doctrine of frustration in these three Ontario decisions are all different ways of saying that these decisions were wrongly decided.

[77] The arbitrator was bound to apply the law of Ontario and applicable federal law. The law of frustration in Ontario is as articulated by this province’s highest court, read together with any binding authority on the matter issued by the Supreme Court of Canada. For reasons I have set out in paras. 57-67 above, there is no conflict between what the Supreme Court said in Peter Kiewit and Naylor, or what the Court of Appeal for Ontario said in Capital Quality, and a consideration of whether, in determining if a supervening event makes performance of a contract radically different from what was contemplated, the supervening event was foreseen.”

The Court went on to note that the parties’ contract was governed by Ontario law. As such, the arbitrator had a duty to apply Ontario law as it currently stood. The Court held that the Supreme Court and Court of Appeal “…have said unequivocally that foreseeability is a key element in the analysis of frustration. This is because, whether an event was reasonably foreseen is important, if not necessary, in determining whether performance has become something radically different from what was agreed to in the contract.”

The Court also took issue with the arbitrator’s finding that the contract was frustrated because it made no express provision for what would happen if government authorities denied environmental approval. The Court noted that, while this was true, no contract provides for all potential contingencies. In that respect, the arbitrator set up an “impossible standard for whether provision has been made in the agreement for supervening events”. If this were required to preclude a finding of frustration, the doctrine would apply all the time. That would be wholly unsatisfactory given contracts are presumptively enforceable, and frustration is a “limited exception” to that general rule.

The Court analyzed other arguments Franco-Nevada raised, and agreed with its position that the arbitrator’s decision was unreasonable on those grounds as well. Although I will stop here, I commend the full decision to interested readers.

Contributor’s Notes:

First, I have made my views on the standard of review question well known here. That view is in line with the majority of jurisprudence (including the only appellate case law on point), which holds that the appellate standard of review should apply to appeals from arbitral awards. This means a court hearing an appeal from an arbitral award should review questions of law for correctness, not reasonableness. The arbitral tribunal’s findings of fact or mixed fact and law should stand absent palpable and overriding error (assuming questions of fact or mixed fact and law are reviewable at all under the parties’ agreement and the applicable law).

That debate aside, this case demonstrates how, in many instances, the standard of review question is rather academic. This is especially so when the question of law invokes binding precedent on a point of common law, or a statutory provision’s interpretation. 

As an illustration, consider the Court of Appeal for Ontario’s case law stating that it is an error of law for a court (or arbitral tribunal) to decide a limitations issue without considering discoverability [Presley v. Van Dusen, 2019 ONCA 66]. If a party raises discoverability, and the tribunal determines a claim is statute-barred without considering it, there is a reversible error of law. You could say the award’s outcome in that case was unreasonable for failing to respect the relevant legal constraints [Vavilov, paras. 90, 99-101 and 105 et seq.]. But in substance, one has to squint pretty hard to differentiate this from saying the tribunal applied the incorrect legal test, and thus committed a reversible error of law.

The same analysis arises in this case. In fact, here, it could be argued that the law was somewhat less clear than the example provided above. A more charitable reading of the arbitrator’s award would conclude that he did not so much refuse to apply binding precedent as misinterpret it. That said, the difference is more or less immaterial; the Court examined the arbitrator’s reading of the case law, and set the award aside upon finding he got it wrong.

Second, and relatedly, although the Court stated it was applying a deferential standard, it had no trouble substituting its reading of the case law for that of the arbitral tribunal. Some might say that, in doing so, the Court overstepped. 

After all, the parties submitted their dispute to arbitration, which evidences an intention to remove it from the courts. The parties also selected their arbitrator (though unclear from the reasons, the Court does not suggest the arbitrator was appointed other than through the parties’ express agreement). On the reasoning in Sattva Capital Corp. v. Creston Moly Corp., 2014 SCC 53, this is a basis for presuming expertise, generating another reason for deference. However, a more probing analysis shows these two perennial battle cries for deference do not withstand scrutiny. 

On the first point, it is trite that the parties’ agreement to arbitrate evidences an intention to remove the dispute from the Court. Reams of cases say this, and those cases are not wrong. However, where the parties also bargain for an appeal, that statement is overly simplistic to the point of being misleading. Where the arbitration agreement provides for an appeal, the parties have chosen to exclude the courts in the first instance. But the very same exercise of party autonomy that pushed the court away in the first instance includes an intention to permit either party to pull the court back in to hear an appeal. 

Party autonomy is the cornerstone of arbitration. Nowhere is this expressed more clearly than in the legislature’s choice to allow parties to decide whether they want an appeal and, if so, on what types of issues. This is truly exceptional. The court’s jurisdiction is a question of law, and is ordinarily not subject to party agreement [see for e.g.: Hupacasath First Nation v. Canada (Foreign Affairs and International Trade Canada), 2015 FCA 4, para. 38]. While the statute book might contain another act that permits private parties to dictate the court’s jurisdiction, I am unaware of any. This and other legislative design choices show that the courts must respect the parties’ will as expressed in their arbitration agreement. That means respecting the whole agreement, including the desire to have the option to appeal the award to a court.

On the second point, there is no doubt parties often appoint arbitrators for their expertise. But this too is overly simplistic. Much of the time, the expertise the parties are after relates to an arbitrator’s knowledge or experience in a certain area of legal practice (e.g., franchise law or insurance law), or industry (e.g., commercial shipping or mining). Where that expertise is in play, there is good reason to accord deference. However, that expertise generally (though maybe not exclusively) arises where the arbitrator deals with factually suffused issues. For example, in a construction dispute, an experienced construction arbitrator is better placed than most judges to determine whether a particular delay should fall within the owner’s or contractor’s sphere of risk. On appeal, those determinations should only be disturbed in the face of palpable and overriding error (unless there is some extricable error of law, but that is a discussion for another case note). 

That said, arbitrators are typically no more expert than courts in applying general law, such as the law of estoppel, limitations, set-off or, as in this case, frustration of contract. Indeed, all of that law flows from the courts.

Even in judicial review of administrative action, where deference is the rule, the Supreme Court has reaffirmed that a court will not defer on questions of law on which  the court and administrative decision-maker share concurrent jurisdiction in the first instance [Society of Composers, Authors and Music Publishers of Canada v. Entertainment Software Association, 2022 SCC 30, paras. 26-28]. This is perforce the case when it comes to arbitrators applying the common law and most statutes. After all, but for the parties’ arbitration agreement, the dispute Taseko and Franco-Nevada referred to arbitration would have gone to court. 

Finally, on this point, arbitral tribunals are required to apply the law governing the dispute, unless the parties dispense with that requirement and empower the tribunal to decide as amiable compositeur or ex aqueo et bono. That was not the case here (and it rarely is in complex commercial disputes). A decision according to law is not a decision according to a reasonable approximation of the law. In that regard, where the parties want their disputes decided according to law, there is good reason to conclude they want the arbitral tribunal to get the law right (or “correct”).

This is a different rationale for applying correctness on questions of law than the one applicable to appeals from court decisions. The courts are public bodies. Although they decide disputes between private persons, they owe a more general duty to the public to get the law right for the sake of consistency in the administration of justice [Housen v. Nikolaisen, 2002 SCC 33, para. 9]. Arbitral tribunals are private decision-makers; they do not owe a public duty in this sense, since their decisions only bind the parties who appoint them, and thus create no law [Desputeaux v. Éditions Chouette (1987) inc., 2003 SCC 17, para. 62]. It is from that private, contractual duty that their obligation arises. In other words, they must be correct on the law because that is what the parties asked of them. Where the parties contract for a right of appeal, they provide themselves something of a judicial insurance policy to make sure they get what they paid for. This is not to say reviewing judges cannot make mistakes of their own. However, parties who want appeals make a procedural design choice. That design choice signals their view that a court should have the last word on the law. And in arbitration, it is the parties’ design choice that matters first and foremost.