Standard Of Review -- Dunsmuir Applied

In Dunsmuir v.New Brunswick, 2008 SCC 9, at para. 34, the Supreme Court of Canada determined that there should now only be two standards of review of an administrative tribunal – correctness and reasonableness. The two variants of reasonableness review, reasonableness simpliciter and patent unreasonableness, are now collapsed into a single form of “reasonableness” review: Dunsmuir at para 45.

Today’s Court of Appeal decision in Mills v. Ontario (Workplace Safety and Insurance Appeals Tribunal), 2008 ONCA 436 considers whether there is a sliding scale of reasonableness within the reasonableness standard. The short answer is no – while reasonableness has to be considered within the context of the decision made, there are only two standards of review, correctness and reasonableness.

The Court holds:

[19] In my view, by collapsing the patently unreasonable standard and the reasonable standard, the majority has not set aside the court’s earlier decision in Law Society of New Brunswick v. Ryan, nor has it signaled that courts must now puzzle over the degree of deference to give to a tribunal within the reasonableness standard. The existence of varying degrees of deference within the single reasonableness standard suggests that a decision made by a tribunal will be found to be unreasonable if the court accords the tribunal a low degree of deference but that same decision will be found to be reasonable if the court decides to accord the tribunal a high degree of deference. I do not read the decision of the majority in Dunsmuir as encompassing any such approach.

[20] As the majority explained that para. 42:

[E]ven if one could conceive of a situation in which a clearly or highly irrational decision were distinguishable from a merely irrational decision, it would be unpalatable to require parties to accept an irrational decision simply because, on a deferential standard, the irrationality of the decision is not clear enough. It is also inconsistent with the rule of law to retain an irrational decision. As LeBel J. explained in his concurring reasons in Toronto (City) v. C.U.P.E., at para. 108:

In the end, the essential question remains the same under both standards: was the decision of the adjudicator taken in accordance with reason? Where the answer is no, for instance because the legislation in question cannot rationally support the adjudicator’s interpretation, the error will invalidate the decision, regardless of whether the standard applied is reasonableness simpliciter or patent unreasonableness. …

[21] The “revised system” established in Dunsmuir was designed in part to make the approach to judicial review of administrative decisions “simpler and more workable” (para. 45). An analysis of the varying degrees of deference to be accorded to the tribunal within the reasonableness standard, as submitted by the appellant, fails to comply with this objective.

[22] My conclusion does not signal that factors such as the nature and mandate of the decision-maker and the nature of the question being decided are to be ignored. Applying the reasonableness standard will now require a contextual approach to deference where factors such as the decision-making process, the type and expertise of the decision-maker, as well as the nature and complexity of the decision will be taken into account. Where, for example, the decision-maker is a minister of the Crown and the decision is one of public policy, the range of decisions that will fall within the ambit of reasonableness is very broad. In contrast, where there is no real dispute on the facts and the tribunal need only determine whether an individual breached a provision of its constituent statute, the range of reasonable outcomes is, perforce, much narrower