Dunsmuir’s Demise & The Rise of Disguised Correctness Review (The Hon. Joseph T. Robertson)

The Honourable Joseph T. Robertson, Q.C., formerly of the Federal Court of Appeal (1992-2000), the New Brunswick Court of Appeal (2000-2014) and Jurist-in-Residence with the Faculty of Law, University of New Brunswick (2014-2017).

This digital symposium marks the 10th Anniversary of Dunsmuir v New Brunswick.[1] Undoubtedly, attention will focus on whether this “transformative” decision has achieved its stated objective of simplifying the law of administrative deference. Some are likely to focus on how the Supreme Court’s post-Dunsmuir jurisprudence has generated doctrinal “incoherence” and “inconsistency.”[2] To the list may be added a dash of “doctrinal silliness”; an accusation that flows easily from cases where the Court has summarily professed deference for interpretative rulings never made or unsupported by reasons.[3] This commentary, however, is aimed at other pressing matters.

Most assuredly, Dunsmuir made a herculean effort to simplify the doctrine of administrative deference by providing a two-step framework for identifying the proper review standard and by reducing the number of deferential standards to a single standard of “reasonableness.” However, the Court’s post-Dunsmuir jurisprudence clearly reveals the application of a one-step framework; one that is incompatible with Dunsmuir’s understanding of the doctrine’s guiding principles and one that makes little room for correctness review. The deferential standard has become the go-to standard. Regrettably, the same jurisprudence also shows the deference obligation has had little impact on case outcomes and, in particular, those tied to the interpretation of the decision-maker’s enabling legislation. Too often the Supreme Court is caught applying the correctness standard under the banner of reasonableness; commonly referred to as “disguised correctness review”.

It is not difficult to show that the post-Dunsmuir jurisprudence has moved to a one-step framework for identifying the proper review standard. The analysis begins with Dunsmuir’s first step – the categorical approach. Its application leads to the general understanding that an administrative ruling will be owed deference unless the issue falls within one of the correctness categories identified in Dunsmuir. However, Dunsmuir anticipated the first step might prove “unfruitful” thus requiring a move to the second. Once known as the pragmatic and functional approach, the second step is referred to as the contextual approach or simply contextualism. It requires consideration of four factors to determine the legislature’s intent over whether the administrative ruling was owed deference.

The contextual approach required reviewing courts to look to the “nature of the issue” and to assess whether the decision-maker possessed a “relative expertise” when compared to that reasonably expected of reviewing courts.[4] Invariably, the discussion on “legislative purpose” was subsumed under the topic of expertise. Additionally, the reviewing court looked to the enabling legislation to see whether it contained either a “privative clause” or a “right of appeal”. However, neither provision was determinative of the applicable review standard. In theory, consideration of all four factors was mandatory. In practice, attention focused on expertise and the statutory directives.

While Dunsmuir made room for a contextual analysis, the Court’s subsequent jurisprudence has all but formally rejected its application. In Edmonton (City) v Edmonton East (Capilano) Shopping Centres Ltd.,[5] the majority (5/4) affirmed its allegiance to the categorical approach while summarily snubbing contextualism for generating “uncertainty and endless litigation concerning the standard of review.”[6] But there is more compelling evidence of contextualism’s demise. Not since Dunsmuir has the Supreme Court identified the review standard through a sustained analysis of contextualism’s four factors. The categorical approach has been driving the analyses.

Admittedly, other reviewing courts have clung to the vestiges of Dunsmuir to validate the selection of the correctness standard for issues falling outside its limited categories. If this means that contextualism lives on, the Supreme Court can take no credit. The Court’s decisions, post-Capilano, continue to embrace the categorical approach; the one that limits the grounds for rebutting the presumptions of expertise and deference to issues falling within a correctness categories.[7]

Those who hold a contrary opinion will undoubtedly seek refuge in Rogers Communications Inc. v Society of Composers, Authors and Music Publishers of Canada.[8] Without doubt, that case provides the most secure lifeline for those unwilling to publish contextualism’s obituary. However, in Mclean v British Columbia (Securities Commission),[9] the Court limited the precedential significance of Rogers Communications by effectively classifying it as exceptional case.[10] Admittedly, the Supreme Court’s ability to produce dissenting opinions on the standard of review issue leaves hope for those who continue to support the contextual approach.

The Supreme Court’s abandonment of contextualism explains why its post-Dunsmuir decisions are silent when it comes to comparing the decision-maker’s relative expertise with that reasonably expected of the reviewing court.[11] In short, what once was a presumption of expertise has become free-standing – only rebuttable if the issue falls within one of Dunsmuir’s correctness categories. Correlatively, it makes no difference to the standard of review issue whether the enabling legislation contains a privative clause, a right of appeal or neither. This explains why the jurisprudence no longer speaks of legislative intent as the source of the deference obligation and why the law has effectively abandoned the distinction traditionally drawn between judicial and appellate review. Free-standing expertise is what counts. In brief, Dunsmuir is on life-support so long as contextualism has a past but no future and the search for legislative intent remains a topic of historical interest.  In short, what we have been witnessing is the demise of Dunsmuir.

The assertion that Dunsmuir is on life-support may have to be revised if the “judicial” proposal for further doctrinal reform takes hold. One should not readily assume that the future of the categorical approach is secure. In Wilson v Atomic Energy of Canada Ltd.,[12] Abella J raised the option of engaging in future reform: the adoption of a single reviewing standard of reasonableness. However, no one on the Court was prepared to endorse the option. Most certainly the dissenting judges did not. Those who agreed with Abella J in the result wrote concurring opinions. Subsequently, the majority in Capilano referenced the invitation and the need to ensure that “the principles of Dunsmuir should provide the foundation for any future direction” but that “any recalibration of our jurisprudence should await full submissions.”[13] Perhaps, it is best to ask whether Dunsmuir is on death-row; a topic which Mathew Lewans has canvassed with insight.[14]

While some will insist that all obituaries relating to contextualism’s demise are premature, others will focus on the scope of Dunsmuir’s four correctness categories. Those categories embrace the situations in which the Court has been unwilling to concede that relative expertise rests with the statutory delegate. This is not to suggest that Dunsmuir’s treatment of those categories is immune from criticism. I digress here to emphasize what may not be obvious to all.

Of the four correctness categories, two remain particularly problematic (discussed below). At the same time, Dunsmuir failed to identify a fifth: the one that accounts for procedural fairness issues. There is also a sixth. Dunsmuir made no mention of the historical fact that the correctness standard automatically applies to the interpretative rulings of select tribunals. This was true of human rights tribunals at the time Dunsmuir was decided. Some post-Dunsmuir cases have overlooked the earlier jurisprudence while others have remained compliant. The conflict is palpable. Finally, there exists a general understanding that inconsistent administrative decisions do not generally attract review for correctness. That is the Supreme Court’s official stance, but in practice the Court has become proficient at judicial avoidance of its lead decision: Domtar Inc. v Quebec (Commission d’appel en matière de lésions professionnelles).[15]

As mentioned, two of the correctness categories are truly problematic. I begin with the one that has captivated the attention of every administrative lawyer for far too long: “true jurisdictional questions”. Unfortunately, too much ink has been spilled trying to articulate a legal framework for isolating such questions. Forty years have passed since CUPE v New Brunswick Liquor[16] and we are still looking for the true jurisdictional question despite the truth lurking behind Professor MacLauchlan’s incisive quip: “once you start thinking about jurisdiction it is like extracting oneself from fly-paper; it is virtually impossible to break free.”[17] More importantly, it was Rothstein J, in Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association,[18] who questioned the need for retaining the category assigned to true jurisdictional questions. Unfortunately, there was no consensus on this point.

The law’s continuing fascination with the concept of the true jurisdictional question perpetuates needless uncertainty in the law. Any debate that lasts forty years is both demoralizing and unnecessary. Surely, the time has come for the Court to officially declare hunting season closed for those intent on flushing out true jurisdictional questions hidden beneath the statutory branches of the decision-maker’s enabling legislation. Elsewhere I have argued that nearly all of the cases involving an allegation of true jurisdictional issue have turned on the interpretation of the tribunal’s enabling statute.[19] Frankly, it is difficult to see how the elimination of the jurisdictional category will diminish the ability of reviewing courts to fulfill their constitutional role of ensuring the legality of administrative action. After all, the law is no longer preoccupied with whether an administrative decision is reviewable in the face of a “full” privative clause. It is just a question of whether the decision is owed deference!

The correctness category most deserving of criticism captures those “questions of general law that are of central importance to the legal system as a whole and outside the tribunal’s specialized field of expertise” (“questions of central importance”). One could be forgiven for believing that Dunsmuir sought only to recognize a correctness category with respect to the application of common law and equitable principles while making room for the exceptional case; those where the tribunal had developed an expertise in the application of a particular common law or equitable principle. That was the law pre-Dunsmuir. [20]

In short, one would have expected that the reference in Dunsmuir to “questions of general law” was a reference to general legal principles for which tribunals could not claim relative expertise. However, a careful reading of Dunsmuir reveals that the majority of the Court was intent on establishing a stand-alone correctness category for questions of central importance. As a matter of fact, the majority held the interpretative issue surrounding the enabling legislation in Dunsmuir did not qualify as a question of central importance to the legal system. Only Binnie J anticipated the unwieldy scope of this category. The law should “prefer clarity to needless complexity and hold that that that the last word on questions of general law should be left to judges.”[21]

Binnie J’s plea of needless complexity stems from the reality that there is no practical test for identifying questions of central importance to the legal system.[22] And like the days of old, the reviewing court must now assess the relative expertise of the decision-maker in regard to the issue at hand. All of this forces one to ask why our Supreme Court continues to engage in self-flagellation. Bluntly stated, yet with great respect; the law has spent more than forty years looking for the true jurisdictional question, twenty years trying to assess the relative expertise of tribunals, ten years trying to convince everyone of a meaningful distinction between two deferential standards of review and now another possible lifetime wandering the administrative galaxy looking for questions of law of central importance to the legal system.

It is easier to accept that reasonableness is the go-to standard of review and to refocus on the task of assessing the reasonableness of an administrative decision. However, there is a more pressing matter; one that undermines the practical significance of the standard of review issue. It must be asked whether the Court’s selection of the deferential standard of review has had a discernible impact on case outcomes and, in particular, interpretative outcomes. A similar question was earlier posed by (Justice) John Evans.[23] Although speaking of reviewing courts generally, and not of the Supreme Court in particular, his response is illuminating: “Telltale signs abound that the choice of standard generally has little bearing on the outcome of cases involving a tribunal’s interpretation of its home statute.”[24] In his view, it is too difficult to discern any meaningful distinction between review for reasonableness and review for correctness. Other commentators have made similar observations: David Mullan, Matthew Lewans, Lauren Wihak and Paul Daly come quickly to mind.

A review of the post-Dunsmuir jurisprudence reveals that too often the Supreme Court is caught engaging in disguised correctness review. This means the Court offers a de novo analysis of the issue without meaningful reference to anything the administrative decision-maker may have said. Few have bothered to seek out explanations for the phenomenon. Paul Daly[25] offers valuable insights. Additionally, there is a false premise upon which the categorical approach rests.

Save for matters falling within the correctness categories, the categorical approach forces us to accept that that all administrative decision-makers possess a relative expertise on all matters. Better still, the law fails to account for the political reality that not all not all tribunals are created equal. The importance of these assertions should not be readily discounted. The post-Dunsmuir jurisprudence has extended the deference obligation beyond the adjudicative tribunal, to the interpretative decisions of statutory delegates such as government ministers and their delegates, office holders and even the Governor in Council. In short, it looks as though every statutory delegate gets to ride on the deference bicycle originally built for labour tribunals.

It is not difficult to spot cases where the Supreme Court has engaged in disguised correctness review. Accepting that the statutory decision-maker is obligated to apply the interpretative principles which the Supreme Court has otherwise affirmed, one would expect the delegate’s analysis to follow the template the Supreme Court uses. It is based on Elmer Dreidger’s modern principle of interpretation. [26] The obligation of the administrative delegate is to apply the Dreidger template so as to determine whether the statutory provision is open to two or more reasonable interpretations (Dunsmuir’s range of reasonable outcomes). If the delegate so concludes, the interpretative analysis moves to the next stage; to elaborate on the policy reasons for choosing one interpretation over the others. Hopefully, this is where the delegate’s “expertise” shines most brightly.

Now, if one turns to the Supreme Court’s analysis of the administrative decision-maker’s reasons for decision, invariably there is only a cursory reference to those reasons, save of course for the interpretative result. As Paul Daly so aptly observed: “Ideally, in my view, an unreasonableness analysis should start with the administrative decision, rather than the statutory text, and demonstrate why it is unreasonable; it should not begin with and end with the statutory text with only a cursory reference to the decision under review.”[27] In point of fact, most Supreme Court cases begin and end just as Professor Daly stated. On my scratch pad, there are close to 20 Supreme Court decisions that so qualify.[28]

Those who make their living from administrative law will take no comfort from the realization that the Supreme Court has spent nearly four decades crafting a deference doctrine that has had little impact on case outcomes and, in particular, those tied to the decision-maker’s enabling legislation. The practioner and reviewing judge must still pay careful attention to both Dunsmuir and Newfoundland and Labrador Nurses’ v Newfoundland and Labrador (Treasury Board),[29] as the law still requires formal determination as to a decision’s reasonableness.

There are only three paragraphs lodged within Dunsmuir’s majority opinion that speak to the meaning of the deferential standard, and although all three have achieved iconic status in the field of administrative deference, there is a continuing struggle with their application. Subsequent cases such as Newfoundland Nurses’ explain and amplify what is found in Dunsmuir. It is the subsequent case law that addresses the “sufficiency of reasons” issue and related matters.

With respect to the reasonableness of interpretative rulings, Dunsmuir requires they fall within a “range of reasonable outcomes.” Correlatively, there may be no range. The application of Dreidger’s principle may support only one interpretative outcome. Ironically, few bother to ask what Dreidger meant of his principle and how his successor, Ruth Sullivan,[30] has gone about recasting it. Importantly, there is a lingering debate over whether “statutory ambiguity” is or should remain the gateway to deference. For some, the ambiguity requirement is too closely aligned with the Chevron doctrine – the American equivalent of our deference doctrine. Together, these peripheral issues should engage the attention of those who must translate legal precedents into a livelihood. Obviously, Dunsmuir will be of little assistance.

* Formerly of the Federal Court of Appeal (1992-2000) and the New Brunswick Court of Appeal (2000-2014) and formerly Jurist-in-Residence with the Faculty of Law, University of New Brunswick (2014-2017).

[1] 2008 SCC 9, [2008] 1 SCR 190, (“Dunsmuir”).

[2] See Paul Daly: “Struggling Towards Coherence in Canadian Administrative Law? Recent Cases on Standard of Review and Reasonableness” (2016) 62:2 McGill Law Journal 527 and (Justice) David Stratas, “The Canadian law of Judicial Review: A Plea for Doctrinal Coherence and Consistency” (2106) 42 Queen’s Law Journal 27.

[3] See Edmonton (City) v Edmonton East (Capilano) Shopping Centres Ltd, 2016 SCC 47 and Mclean v British Columbia (Securities Commission), 2013 SCC 67, [2013] 3 SCR 895.

[4] The “nature of the issue” factor involved a preliminary determination as to whether the question qualified as one of fact, mixed law and fact or law.

[5] 2016 SCC 47, [2016] 2 SCR 293 (“Capilano”).

[6] Ibid., at para. 35.

[7] E.g., Barreau du Quebec v Quebec (Attorney General), 2017 SCC 56.

[8] 2012 SCC 35, [2012] 2 SCR 283 at para. 16.

[9] 2013 SCC 67, [2013] 3 SCR 895.

[10] Ibid., at paras. 19-22.

[11] An exception arises in the context of one of the correctness categories; questions of general law of central importance to the legal system.

[12] 2016 SCC 29, [2016] 1 SCR 770.

[13] Capilano, supra, note 5, at para 20.

[14] See Matthew Lewans, “Renovating Judicial Review” (2017) 68 UNBLJ 109.

[15] [1993] 2 SCR 756.

[16] Canadian Union of Public Employees, Local 963 v New Brunswick Liquor Corporation, [1979] 2 SCR 227.

[17] Wade MacLauchlan “Judicial Review of Administrative Interpretations of Law: How Much Formalism Can We Reasonably Bear?”, (1986) 36 UTLJ. 343 at note 3.

[18] 2011 SCC 61, [2011] 3 SCR 654.

[19] See JT Robertson, “Judicial Deference to Administrative Tribunals: A Guide to 60 years of Supreme Court Jurisprudence” in Joseph T Robertson, Peter A Gall & Paul Daly, Judicial Deference to Administrative Tribunals in Canada: Its History and Future (Toronto: NexisLexis, 2014) at 39 et seq.

[20] Dunsmuir, at para 54 citing to Toronto (City) v CUPE., Local 79, 2003 SCC 63, [2003] 3 SCR 63, at para. 72.

[21]Dunsmuir, supra, note 1, at para. 128.

[22] A question of central importance was identified in both Commission scolaire de Laval v Syndicat de l’enseignment de la region de Laval, 2016 SCC 8, [2016] 1 SCR 29, and Alberta (Information and Privacy Commissioner) v University of Calgary, 2016 SCC 53, [2016] 2 SCR 555.

[23] (Justice) John Evans, “Triumph of Reasonableness: But How Much Does It Really Matter?” (2014), 27 CJALP 101.

[24] Ibid, at 108.

[25] See Paul Daly, “The Signal and the Noise in Administrative Law” (2017) 66 UNBLJ 68.

[26] Dreidger, Elmer. A. Construction of Statutes, 2nd ed. Toronto: Butterworths, 1983.

[27] Paul Daly, Blog, Administrative Law Matters: Comment “Doing Reasonableness Review: Workplace Health, Safety and compensation Commission v Allen, 2014 NLCA 42. (Dated November 26, 2014).

[28] Notable cases include: Edmonton (City) v Edmonton East (Capilano) Shopping Centres Ltd, supra, note 3; Mclean v British Columbia (Securities Commission), supra, note 3; Wilson v Atomic Energy of Canada Inc, supra, note 12; Wilson v British Columbia (Superintendent of Motor Vehicles) 2015 SCC 47, [2015] 3 SCR 300; Agraira v Canada (Public Safety and Emergency Preparedness) 2013 SCC 36, [2013] 2 SCR 559; Canadian National Railway Co. v Canada (Attorney General) 2014 SCC 40, [2014] 2 SCR 135; Smith v Alliance Pipeline, 2011 SCC7, [2011] 1 SCR 160,; Canada (Canadian Human Rights Commission) v Canada (Attorney General), 2011 SCC 53, [2011] 3 SCR 471; Association des courtiers et agents immobiliers du Quebec v Proprio Direct Inc. 2008 SCC 32, [2008] 2 SCR 195, and Kanthasamy v Canada (Citizenship and Immigration), 2015 SCC 61, [2015] 3 SCR 909.

[29] 2011 SCC 62, [2011] 3 SCR 708.

[30] Sullivan Ruth, Sullivan on the Construction of Statutes, 6th ed. Markham, Ont.: LexisNexis, 2014.

This content has been updated on February 14, 2018 at 20:49.