Part I: Introduction
Administrative law focusses on the way in which, and the extent to which, courts should oversee the exercise of administrative authority. The law on substantive review of administrative decision-making has changed drastically over the last several decades, particularly around choice of standard of review. In the words of the Honorable John M Evans, courts have returned to this issue "with almost monotonous regularity over the last 30 years". (1) Two Supreme Court of Canada decisions from 2016, Wilson v Atomic Energy of Canada Ltd (2) and Edmonton (City) v Edmonton East (Capilano) Shopping Centres Ltd, (3) have regenerated discussion about standard of review in relation to questions of law. No less an authority than the Honourable Justice David Stratas has suggested that the Court may be "about to embark on one of its once-a-decade, wholesale revisions to the law of judicial review". (4)
To assess how Wilson and Capilano relate to the Supreme Court's last wholesale revision of the law on substantive review in Dunsmuir v New Brunswick, (5) this article: a) considers Justice Abella's suggestion in Wilson that a separate standard of correctness review is no longer needed; b) assesses the trend, developing pre-Capilano and implicitly accepted by the majority in that decision, of limiting correctness review to the four categories of legal questions identified in Dunsmuir, and c) discusses the difficulties of applying the Dunsmuir understanding of reasonableness where there are only two possible interpretations of the legislative provision in dispute (Wilson and Capilano), or where the administrative decisionmaker has not provided reasons on an issue under review (Capilano).
Before moving to an analysis of these three themes, the current law on standard of review is placed in context, followed by a brief overview of Wilson and Capilano.
Part II: Background Context
Under an earlier iteration of substantive review of administrative decision-making, questions of law decided by an administrative body were automatically subject to correctness review whenever the matter came before the courts by way of appeal, or on judicial review where the original decision was not protected by a privative clause. (6) The correctness standard was also applied if the enabling legislation included a privative clause, and the matter in contention was judged to be an issue of jurisdiction. The deferential standard of patently unreasonable was reserved for questions of law where an administrative decision maker protected by a privative clause was deciding on an issue within its jurisdiction. (7) Questions of fact were treated with significant deference; courts would intervene only if the administrative decision-maker's findings were capricious or made without reference to the evidence before it. (8) Where the exercise of discretion was challenged, courts refrained from reviewing the outcome and instead confined themselves to asking specific questions about how the decision maker went about its tasks: Courts would ask whether irrelevant considerations had been taken into account, or relevant ones ignored, whether the decision maker had acted for an improper purpose, or whether the decision maker had fettered its statutorily-delegated powers. (9) Thus, in this earlier era, substantive review of an administrative decision could be understood as involving a number of different silos, and placement in a particular silo depended on the nature of the question under review, and the route by which the matter came before the courts (appeal, judicial review without a privative clause, judicial review with a privative clause). Each of these silos attracted a specific test for judicial intervention. For issues of fact and the review of discretion, judicial intervention was applied sparingly. Not so with questions of law; such issues were reviewed on a correctness standard unless two criteria were in place: presence of a privative clause and identification of the issue as falling within the administrator's jurisdiction. (10)
The Supreme Court of Canada's last major reworking of substantive review occurred in 2008, with Dunsmuir v New Brunswick. (11) In the decade preceding Dunsmuir, the Supreme Court had already made significant changes to the scheme described above. The possibility of deference on appeal was recognized and a middle standard between correctness and patent unreasonableness created; (12) a unified approach to determining standard of review was developed, which applied regardless of whether the issue under review was characterized as a question of law, a question of fact or an exercise of discretion and regardless of whether the case involved judicial review or an appeal; (13) a four-part pragmatic and functional analysis was established for the purpose of discerning whether the legislature intended the decisions of particular administrative entities to be treated with deference or not; (14) and the concept of jurisdiction was largely relegated to the sidelines. (15)
In Dunsmuir, on judicial review of a decision by a labour adjudicator appointed under the Public Service Labour Relations Act, (16) the Court applied a reasonableness standard, but overturned the adjudicator's decision as unreasonable. The Court also took the opportunity to "develop a principled framework [of substantive review] that is more coherent and workable," (17) reassessing the approach which it had crafted less than a decade earlier in cases such as Pushpanathan and Baker. At the heart of this new framework was a desire to simplify the approach to determining standard of review. Besides melding the two deferential standards into one, Dunsmuir also: emphasized the role of precedent in establishing standards of review; identified four categories of questions of law to which correctness applies automatically (these being constitutional issues, jurisdictional issues, issues relating to the jurisdictional boundary between two specialized tribunals, and general questions of law of central importance to the legal system as a whole and beyond the expertise of the administrative decision maker); stated that deference may apply to legal questions, such as interpretation of the decision maker's home or related statutes; and identified deference as almost always the appropriate standard for questions of fact, and discretionary decision making. Where the appropriate standard of review is not immediately obvious, Dunsmuir directed reviewing courts to consider some or all of the four elements of the standard of review analysis (a renamed pragmatic and functional analysis) to determine the legislature's intention in this regard.
The law on substantive review did not stand still after Dunsmuir. For instance, the relevance of Dunsmuir to judicial review of federal administrative decision makers was confirmed; (18) certain types of constitutional analysis were carved out of the automatic-correctness classification; (19) the relationship between the two arms of Dunsmuir's description of reasonableness was explained; (20) and, as discussed further below, the four categories of correctness came frequently to be treated as exhaustive. Not only was the Dunsmuir framework refined in a number of ways, but the oversight of administrative action is necessarily nuanced (or vague, depending on one's point of view), so subsequent courts on occasion wrestled with applying various aspects of Dunsmuir. However, most commentators would probably have agreed with Justice Evans' 2013 statement that the law on standard of review was "reasonably well settled". (21) That sense of stability was at least partially disrupted by Capilano and Wilson.
Part III: Overview of Wilson and Capilano
Wilson v. Atomic Energy of Canada Ltd
Wilson raised the question of whether amendments to the Canada Labour Code (22) protect non-unionized employees from termination absent just cause. In the decades since these amendments were introduced, labour adjudicators across Canada had been at odds on this issue. (23) The adjudicator in Wilson had ruled that under the Labour Code, non-unionized employees could not be dismissed on the provision of reasonable notice or payment in lieu--just cause was required.
On judicial review, the majority of the Supreme Court applied the reasonableness standard and upheld the adjudicator's interpretation of the Code as reasonable. Had the discussion of standard of review stopped there, Wilson might not have attracted much notice beyond labour law practitioners. However, Justice Abella took the opportunity to provide "some general comments about standard of review" (24) with the aim of "simplifying] the standard of review labyrinth we currently find ourselves in." (25) Acknowledging that her comments in this regard would be obiter, Justice Abella offered her proposals "as an option only, for purposes of starting the conversation about the way forward." (26) Her first, and primary, proposal was that the standard of correctness be mothballed, and all substantive review of administrative action be approached from a deferential stance. As a back-up proposal, in case "there prove[d] to be little appetite for collapsing the two remaining standards of review," (27) Justice Abella proposed limiting the scope of correctness review, such that a "residual 'correctness' standard" would be available "only in those four circumstances Dunsmuir articulated." (28)
Four other members of the majority (Chief Justice McLachlin, and Justices Karakatsanis, Wagner and Gascon) agreed with Abella J that the labour adjudicator's decision deserved deference and that he had in fact arrived at a reasonable conclusion. On her broader comments, they thanked Justice Abella for her "efforts to stimulate a discussion on how to clarify or simplify our standard of review jurisprudence to better promote certainty and predictability," (29) but declined to "endorse any particular proposal to redraw our...