Skip to Content

Arbitration: Standards of review tailored for the parties

How final is an award in a commercial arbitration? A lot depends on the parties who can set the limits of the right of appeal. Even when it comes to the standard of review.

Handshake between business woman and business man to illustrate parties agreeing

How final is an award in a commercial arbitration? A lot depends on the parties who can set the limits of the right of appeal.  Unless, it is often assumed, there are errors of law, and then only if granted leave from a court.  On this count, however, the parties have even more say than some might think.  Indeed, they can set the terms of standard of review of an award, and the courts must apply them.

How is this so? Well, the courts must give effect to the rights of appeal specified in the arbitration agreement. This can play out in different ways. The parties can agree the award should be appealed on questions of law, with or without leave, or on questions of fact or mixed fact and law. Or, if an arbitration agreement is silent on the matter, there is the default right of appeal on questions of law – again, with leave of the court.  With few exceptions and some minor variations, this is all provided for in provincial arbitration laws.   

Must we assume though that the standard of review to be applied is determined by the common law? Not necessarily so. If party autonomy rules in arbitration, then the parties to an arbitration agreement must also be allowed to specify different standards of review.  

The Supreme Court of Canada first addressed standards of review applicable to arbitration awards in Sattva Capital Corp. v. Creston Moly Corp., a 2014 ruling that dealt with an appeal under the Arbitration Act of British Columbia.

While acknowledging there are differences between judicial review of statutory and arbitral tribunals, the SCC noted similarities between the two – not least of which is that courts are called upon to review the decisions of non-judicial decision makers. What’s more, specialized tribunals have expertise, but arbitrators are also chosen either based on their expertise or qualifications. So, the judicial review framework it developed in its landmark Dunsmuir ruling may not be “entirely applicable to the commercial arbitration context.” But it can serve as helpful guidance.

Ultimately, the top court concluded that a standard of reasonableness almost always applies to questions of law arising in appeals of commercial arbitrations. In rare cases, where there is a question of a constitutional nature or of central importance to the legal system that falls outside the adjudicator’s expertise, it would be the correctness standard.

Post-Sattva, in a 2016 ruling, the Ontario Court of Appeal had to decide a priority dispute. It involved a question of mix fact and law between two insurers arising from the statutory accident benefits regime under the Insurance Act. The court concluded that the administrative law framework applied and that “decisions rendered by non-judicial decision makers should be reviewed under that framework, even when that review is styled as an appeal.” As in Sattva, it relied on the arbitrator’s presumed expertise, knowing that the parties could not appeal on questions of fact, which limited their right of appeal as would a privative clause.  The court also applied the reasonableness standard.

What these decisions confirm is that in both administrative and arbitral contexts the courts will invariably apply the administrative law standard of review.  And regardless of the nature of the question arising from the award, parties to an arbitration agreement can expect the courts to apply a deferential standard of review.

In that sense the parties to an arbitration agreement have powers akin to a legislature that creates a statutory tribunal. They confer jurisdiction on the arbitral tribunal, appoint its member or members, specify the procedural rules and the substantive law that will apply to the dispute, and determine the scope of any appeal rights.  Parties to an arbitration agreement can also contract out of most of the Arbitration Act’s provisions.  Just as legislatures can specify the applicable standard of review, so can parties in an arbitration agreement.  Nothing in provincial arbitration legislation prevents it.

Having said that, there is some authority to suggest that the courts are not bound by the standard of review chosen by the parties in an arbitration agreement.  This case was, however, wrongly decided: The Dominion of Canada General Insurance Company v. Unifund Assurance Company. While this legal conclusion may be true in a public law context where courts are bound by the applicable standard of review, this is not the case in private arbitration where party autonomy rules.  The court will have to consider this distinction in the future.  For the moment, including a standard of review provision may be controversial in the province of Ontario.

Typically, parties favour arbitration because it is a fast way to resolve disputes, among other reasons.  But they might want to insist that any resulting award be correctly decided.  They can do this by calibrating the level of court review applicable to their arbitral award and specifying a correctness standard for any question of law.  

Another advantage of arbitration is having the option to select an expert decision maker.  Even so, many disputes are multifaceted and interdisciplinary, and a single arbitrator may not have expertise in all the areas involved.  For example, a dispute may raise questions of a scientific nature and others that are strictly legal.  An arbitrator with a specific scientific background may be ideal, but the parties may want guarantees that questions of law are decided correctly.  In such a case, they may specify a very deferential standard of review on factual / scientific issues (like patent unreasonableness) and a correctness standard on questions of law.

The beauty of arbitration is that it is consensual, where the parties have ultimate control over the process, up until the very end.