Some Thoughts on Wilson v. Atomic Energy of Canada Ltd., 2016 SCC 29

There are several points of general interest raised by the Supreme Court of Canada’s most recent administrative law decision, Wilson v. Atomic Energy of Canada Ltd., 2016 SCC 29. By majority, the Court reversed the Federal Court of Appeal (see my post). For a discussion of the merits — in which the Court restored what labour lawyers saw as the status quo ante — see David Doorey. I have four comments on the administrative law issues.

First, Abella J. wrote the majority reasons on the merits. Applying a standard of reasonableness because the Adjudicator whose decision was under review was interpreting his home statute (at para. 15), she upheld the Adjudicator’s interpretation of federal labour legislation (at para. 39); indeed, she concluded that to decide otherwise would “fundamentally undermine” the statutory scheme (at para. 69), a suggestion perhaps that there was only one reasonable interpretation of the statute.

However, her more expansive comments on standard of review did not attract a majority (see para. 70 (McLachlin C.J., Karakatsanis, Wagner and Gascon JJ.) and paras. 71-73 (Cromwell J.)). Regular readers will recall that in a case called Tervita last year, Abella J. hinted that the two standards of review, correctness and reasonableness, were beginning to shade into one another. Here, she was explicit in suggesting that the Court ought to consider collapsing the two into a single reasonableness standard, the “most obvious and frequently proposed reform of the current system” (at para. 28). She set up the question as follows: “whether we need two different names for our approaches to judicial review, or whether both approaches can live comfortably under a more broadly conceived understanding of reasonableness” (at para. 24). She also highlighted the key advantage of a general reasonableness standard, that it is flexible enough to allow “a wider range for those kinds of issues and decision-makers traditionally given a measure of deference, and a narrow one of only one ‘defensible’ outcome for those which formerly attracted a correctness review” (at para. 33):

Nothing Dunsmuir says about the rule of law suggests that constitutional compliance dictates how many standards of review are required. The only requirement, in fact, is that there be judicial review in order to ensure, in particular, that decision-makers do not exercise authority they do not have.  I see nothing in its elaboration of rule of law principles that precludes the adoption of a single standard of review, so long as it accommodates the ability to continue to protect both deference and the possibility of a single answer where the rule of law demands it, as in the four categories singled out for correctness review in Dunsmuir (at para. 31).

I agree that, given where Canadian administrative law currently finds itself, it would make sense to adopt a single reasonableness standard.

Second, however, quite what reasonableness might mean is somewhat unclear. In particular, two paragraphs of Abella J.’s reasons are particularly odd when placed side by side:

Nor do I accept the position taken in this case by the Federal Court of Appeal that even if a reasonableness review applied, the Adjudicator should be afforded “only a narrow margin of appreciation” because the statutory interpretation in this case “involves relatively little specialized labour insight”. As this Court has said, the reasonableness standard must be applied in the specific context under review. But to attempt to calibrate reasonableness by applying a potentially indeterminate number of varying degrees of deference within it, unduly complicates an area of law in need of greater simplicity (at para. 18).

The range will necessarily vary. As Chief Justice McLachlin noted, reasonableness “must be assessed in the context of the particular type of decision making involved and all relevant factors” and “takes its colour from the context”: Catalyst Paper Corp. v. North Cowichan (District), 2012 SCC 2 (CanLII), [2012] 1 S.C.R. 5, at paras. 18 and 23, citing with approval Canada (Citizenship and Immigration) v. Khosa, 2009 SCC 12 (CanLII), [2009] 1 S.C.R. 339, at para. 59 (at para. 22).

Without more, these paragraphs are (as I said on Twitter) irreconcilable. The very point that the Federal Court of Appeal was making (and has been making for at least as long as I have been writing this blog) is that “the range will necessarily vary”, leading to broader or narrower ranges or margins of appreciation.

Where I think Abella J. has a problem (and I hope I am reading her unwritten subtext correctly) is with the idea that ‘narrow’ and ‘broad’ might have particular meanings, that is, that narrowness and broadness are themselves “degrees of deference” or standards of review. However, this is not actually the position the Federal Court of Appeal has developed — see notably Delios v. Canada (Attorney General), 2015 FCA 117 — which I think is simply that the narrower or broader the range the easier or more difficult it will be to justify quashing a decision which bears badges of unreasonableness.

It is true that Canadian judges have sometimes found themselves casting about for language that describes a particularly narrow or particularly broad range of reasonable outcomes. I take it that the point of paragraphs 18 and 22 is that Abella J. is unhappy with this tendency and would say simply that the range of reasonable outcomes will vary according to context (see paras. 34-35). But what this means in practice will require further refinement (see here and here).

Third, Cromwell J. concurred with the result but wrote separately to respond to Abella J.’s obiter comments on standard of review.  Unlike the concurring judges who expressed their appreciation of Abella J.’s effort to stimulate debate (at para. 70), Cromwell J. insisted “the basic Dunsmuir framework is sound and does not require fundamental re-thinking” (at para. 72). He nonetheless agreed with Abella J. that “developing new and apparently unlimited numbers of gradations of reasonableness review — the margins of appreciation approach created by the Federal Court of Appeal — is not an appropriate development of the standard of review jurisprudence” (at para. 73). Cromwell J. does not endorse the idea of a variable range of reasonable outcomes at all. For him, reasonableness simply requires a contextual analysis; it is always the same standard, to be applied taking account of a particular decision-making function and all relevant factors. This is perfectly coherent but unfortunately offers little or no guidance on the key question of when a court can intervene to quash a decision as unreasonable.

Fourth, a minority of three judges disagreed with the disposition of the appeal and would also have applied a standard of correctness, on the basis that conflicting interpretations by Adjudicators of federal labour legislation has undermined the rule of law. As Côté and Brown JJ. wrote, “leaving unresolved a divided body of arbitral decisions clouds an essential feature of the federal regime governing employment relationships” because “[f]ederally regulated employers cannot predictably determine when and how they can dismiss their employees, while employees are left in a state of uncertainty about the extent of their job security” (at para. 86).

Quite why the minority judges refused to discuss L’Heureux-Dubé J.’s reasons in Domtar Inc. v. Quebec (Commission d’appel en matière de lésions professionnelles), [1993] 2 SCR 756, the leading Canadian treatment of the issue, is a mystery (though presumably the absence of an “institutional umbrella” for Adjudicators has something to do with it (at para. 82)). They instead cited to Shaw Cable Systems, a very different case which involved conflicting obligations imposed by different tribunals. To be clear, where Adjudicator A says ‘you can fire Employee X without cause’ but Adjudicator B says ‘you need cause to fire Employee Y’, there are no conflicting obligations in the Shaw Cable Systems sense; it may be problematic to have different rules for similarly situated employees but there is no conflict, properly speaking. Moreover, in Shaw Cable Systems, the Court applied principles of deference to decide which tribunal’s view of the law ought to prevail. The case provides no authority for the minority’s approach.

Then there is this mysterious passage:

We believe, therefore, that where there is lingering disagreement on a matter of statutory interpretation between administrative decision-makers, and where it is clear that the legislature could only have intended the statute to bear one meaning, correctness review is appropriate. This lingering disagreement presupposes that both interpretations are reasonable, since, of course, a contradictory but unreasonable decision will be quashed on judicial review and no lingering disagreement can result. But we wish to make one point clear: it does not matter whether one or one hundred decisions have been rendered that conflict with the “consensus” interpretation identified by the majority. As long as there is one conflicting but reasonable decision, its very existence undermines the rule of law (at para. 89).

Question 1: is the test for intervention on a standard of correctness “lingering disagreement” or the existence of “one conflicting but reasonable decision”? Question 2: how could there be any disagreement, lingering or otherwise, “where it is clear that the legislature could only have intended the statute to bear one meaning”? Question 3: is one to take it that the additional factors contained in paragraph 91 — that this is a “matter of general importance” over which the courts share jurisdiction — simply underpin the conclusion already reached because of the “discord” and do not serve any analytical purpose in this case?

Remarkably, the dissenting judges made no effort to justify their application of the correctness standard in doctrinal terms. They made no mention whatsoever of the Dunsmuir framework! Instead, they spoke of the “rule of law and the promise of orderly governance” and the “caprice of the administrative state” (at para. 84). Indeed, Abella J. held a door open for the dissenters with the suggestion that sometimes the range of reasonable outcomes can admit of only one possible, acceptable outcome. That they refused to go through it suggests profound disagreement about standard of review and the relationship between administrative decision-makers and the courts. Whenever the Court gets around to fully re-evaluating its administrative law framework, there are likely to be deep cleavages between the justices.


This content has been updated on July 19, 2016 at 17:03.