Cases Considered: Dunsmuir v. New Brunswick, 2008 SCC 9

PDF Version:  Dunsmuir: Much Ado about Nothing

The majority judgment in Dunsmuir, written by Justices Bastarache and Lebel JJ. (writing also for Fish, Abella, and McLachlin JJ.), begins by setting out its grandiose intention to re-examine judicial review principles in Canadian administrative law with the view to making them more workable and coherent. In an initial glance, one is immediately struck by how such an immense and significant task is built upon a seemingly insignificant set of facts. The appellant, a former non-unionized provincial employee who was dismissed with pay in lieu of notice, sought to uphold a grievance arbitrator’s ruling that his employment be reinstated. In dismissing the appeal, the Supreme Court judgment follows that of both the New Brunswick Court of Queen’s Bench and Court of Appeal. One cannot also help but notice that in purporting to reformulate the pragmatic and functional approach to substantive judicial review, Dunsmuir consists of three concurring but inconsistent sets of reasons. Indeed, it is difficult to envision Dunsmuir as a defining moment in Canadian administrative law along the lines of CUPE Local 963 v. New Brunswick Liquor Board, [1979] 2 SCR 227, Nicholson v. Haldimand-Norfolk Police Commissioners, [1979] 1 SCR 311, Knight v. Indian Head School Division, [1990] 1 SCR 653, Pushpanathan v. Canada (Minister of Citizenship and Immigration), [1998] 1 SCR 982, or Baker v. Canada (Minister of Citizenship and Immigration), [1999] 2 SCR 817. This is because Dunsmuir falls well short of its lofty ambitions. Binnie J.’s reasons aside, Dunsmuir is little more than formal acknowledgement of recent shifts in, and deficiencies with, the Supreme Court’s attitude towards substantive judicial review.

The majority opinion justifies the need to merge reasonableness simpliciter with patent unreasonableness on now familiar grounds that: (i) the two standards are impossible to distinguish in application, despite good intentions in selecting a “middle ground” standard where pragmatic factors point both for and against judicial deference; and (ii) patent unreasonableness contemplates judicial endorsement of an “unreasonable” administrative decision. Their reformulated reasonableness standard retains key features of its predecessors: (i) applicability to questions that lend themselves to more than one answer; and (ii) judicial review limited to an assessment of the administrative decision-maker’s reasons on the basis of its justification, transparency, and legitimacy.

Dunsmuir reduces the choice of standards to correctness or reasonableness. It remains to be seen, however, whether the refined reasonableness standard will be any more coherent in application than its predecessors. Bastarache and Lebel JJ., and Binnie J. in his separate reasons, contemplate a range of deference within a reasonableness review. Binnie J. calls for a collapse of the two step approach (select the standard of review and then apply it) such that the 4 pragmatic and functional factors (privative clause/statutory right of appeal; relative expertise; purpose of the statutory scheme; nature of the question) are applied to assess the level of deference within a reasonableness review.

Dunsmuir affirms the Supreme Court’s continued reluctance to apply judicial deference as originally contemplated by Dickson J. in his 1979 CUPE Local 963 v. New Brunswick Liquor Board judgment and later followed by Wilson J. in her National Corn Growers Assn. v. Canada (Import Tribunal), [1990] 2 SCR 1324 dissent. In recognition that the judiciary is not as well placed to decide questions entrusted by the legislature to specialized delegates to decide as part of a regulatory framework, Dickson J. asserted that the judiciary should be reluctant to substitute its views with those of the delegate. Where an administrative decision-maker is protected by a strong privative clause and is deciding a matter within its sphere of “expertise” (including statutory interpretation and other questions of law), substantive judicial review is thus limited to whether the administrative decision-maker’s interpretation of the law is so patently unreasonable that it cannot be rationally supported. This type of review does not include an assessment as to the reasonableness of the decision-maker’s conclusions.

The level of deference called for by a reasonableness standard as originally contemplated in 1979 has rarely been endorsed by the Supreme Court; it is simply too deferential to be accepted by those that insist on holding themselves out as having the final word in our legal system. As such, problems in applying the reasonableness standard will persist.

Dunsmuir acknowledges the nature of the question as the most important factor in selecting the appropriate level of deference in substantive judicial review; endorsing what has been the reality for years despite claims that relative expertise is the most influential factor. The case law is full of examples where the Supreme Court splits over the nature of the question rather than disputing relative expertise. In Trinity Western University v. British Columbia College of Teachers, [2001] 1 SCR 772, the issue was whether the College of Teachers exceeded its jurisdiction by refusing to accredit Trinity Western on grounds the institution employed discriminatory policies. The majority judgment accorded no deference to the College of Teachers by constructing the College decision as one concerning human rights, an area firmly within the judiciary’s traditional territory. The dissent, in contrast, constructed the College decision as that concerning the management of the teaching profession in British Columbia and hence accorded significant deference to the College.

Note that even within Dunsmuir itself, the nature of the question divides the Court. Deschamps J. (writing for Charron and Rothstein JJ.), constructs the question faced by the grievance arbitrator as involving an interpretation of the common law and, accordingly, accords no deference to arbitrator’s decision (correctness standard). Meanwhile, the majority judgment as well as Binnie J. purport to be more deferential (reasonableness standard) as they see the question as concerning the arbitrator’s interpretation of its governing legislation. As was the case pre-Dunsmuir, how the reviewing court constructs the question is key towards the level of deference that will be afforded to the administrative decision.

None of the opinions in Dunsmuir challenge the received view on the role of substantive judicial review being that of mediating a tension between legislative supremacy and the rule of law by policing the boundaries of administrative authority. In numerous instances throughout Dunsmuir, the Supreme Court describes the role of a reviewing court as ensuring administrative decision-makers, including those with discretionary authority, do not exceed their authority in making decisions. This rigid focus on ensuring decision-makers act within their statutory authority liberates them from legal scrutiny on their refusal or failure to exercise authority granted – for example, in cases where an administrative body charged with having to consider the social and environmental effects that will result from its decision, refuses or fails to do so. Canadian judicial review seems ill-equipped to address this sort of administrative law issue. This is increasingly troublesome in an era where economics increasingly dictates administrative decision-making at the expense of fairness or justice interests.

For other opinions on Dunsmuir, see David Corry and Alice Woolley’s blogs on this site.