Case considered: Dunsmuir v. New Brunswick, 2008 SCC 9
For a recent session of the Canadian Bar Association’s administrative law sub-section we reviewed Alberta Court of Appeal decisions with respect to the use of Dunsmuir v. New Brunswick, 2008 SCC 9 from May 2008 to May 2009. Here we share some preliminary analysis from our findings.
Identification of the Standard of Review – Overview
We begin by summarizing the main points from the majority judgment in Dunsmuir with respect to how to identify the standard of review. The point of Dunsmuir was to simplify the judicial review analysis. The majority said that you can usually determine the standard of review by looking at the nature of the issue or question in the case. The existence of a privative clause is a significant factor, however, as is prior case law. Indeed, the first step in identification of the standard of review is whether “the jurisprudence has already determined in a satisfactory manner the degree of deference to be accorded with regard to a particular category of question” (Dunsmuir, para. 62). It is only after this point that the “standard of review analysis” should take place. The standard of review analysis is the pragmatic and functional test renamed.
When considered in light of Alberta Court of Appeal judgments issued subsequent to Dunsmuir, it seems to us in hindsight that Dunsmuir does not create a single direction for a reviewing court to follow in identifying the appropriate standard of review. Prior to Dunsmuir the standard of review was determined by application of the pragmatic and functional analysis in every case. After Dunsmuir, courts are potentially left with a wide variety of approaches to use:
1. Once the nature of the question is identified and characterized, the standard of review should simply follow; it is only if it is unclear, or some other factor makes it non-determinative, that further analysis is necessary;
2. The determination of the standard of review should be based on cases assessing how different types of questions are reviewed for this particular administrative decision-maker; if such cases do not exist, then the standard of review analysis must be used;
3. The determination of the standard of review should be based on cases establishing how different questions are to be dealt with, without particular focus on the specific administrative decision-maker in question; it is only if this is unclear or non-determinative (as in 1) that further analysis is necessary; or,
4. Normally the standard of review is determined by the standard of review analysis, unless one of the other factors, and in particular very closely related cases, exist.
This variable interpretation of Dunsmuir becomes clear when looking at the Alberta Court of Appeal’s jurisprudence. When viewed in light of those cases, Dunsmuir does not seem so much to have simplified judicial review, as to have uncaged it from the strictures of the pragmatic and functional analysis. In effect, the case has given reviewing courts a whole variety of ways to consider the level of deference that should apply to particular administrative questions brought before it.
Cases Using the Standard of Review Analysis
Only eight of the decisions we read conducted a standard of review analysis for at least one issue raised by the application for judicial review. Interestingly, the standard of review analysis appears unchanged from its former incarnation as the pragmatic and functional analysis. And it has the same problem of that test: courts note where the different factors push deference – either to more or to less – and then “hey, presto!” out pops a standard of review. Thus in Bishop v. College of Optometrists, 2009 ABCA 175, the court states there is a full right of appeal which means less deference, the College has a role in standard setting and ensuring standards are upheld, which suggests more deference, the questions were central to its jurisdiction and had a significant factual component, which suggests more deference. In the end, the Court holds that review should be on the basis of reasonableness. Just like with the pragmatic and functional analysis; however, there is no real explanation of why the statutory right of appeal factor weighs less than the others – is it because it is 1 factor pointing against deference while the other 3 point towards it, or is it because it is less significant or important?
Cases using precedent
The majority of the cases resolved the question of standard of review through precedent. How they did so was, however, quite variable.
The most common approach was to look at past jurisprudence with respect to a particular tribunal and the particular questions. So in a number of cases the Court sets out a list indicating the various kinds of questions that can come before this decision-maker, and indicates what prior case law has determined with respect to how each of them gets reviewed. In Gahir v. Alberta (Workers’ Compensation Appeals Commission), 2009 ABCA 59, for example, the Court says that prior cases indicate that fact, credibility and mixed fact and law decided by the Commission are to be reviewed for reasonableness, as are interpretation and application of policies, interpretation of law within the Commission’s expertise and application of law.
The most extreme version of this approach is where the Court is very narrow in how it looks at precedent. In several cases the Court required that there be a case right on point – with this tribunal and this precise type of question – without which they went to the standard of review analysis. Thus in a Labour Relations Board case Conrad J.A. (who dissented, but decided the appropriate standard) said there was “no relevant jurisprudence relating to the degree of deference owed to the Board when interpreting the collective agreement for the purpose of applying section 147(3) of the Code” with the result that the past jurisprudence could not determine the standard (Finning v. Intern’l Assn. of Machinists and Aerospace Workers, Local Lodge No. 99, 2009 ABCA 55, para. 80).
By contrast, in other cases the Court looked at past jurisprudence more generally, not with respect to the particular tribunal, but simply with respect to how past cases indicate questions of the nature at issue before the Court should be reviewed. Thus in Sincennes v. Alberta (EUB), 2009 ABCA 167, O’Brien J.A. simply said: “The standard of review with respect to a tribunal’s application of its public interest mandate is reasonableness; determination of what is in the public interest has been held to be a matter of administrative discretion and a formulation of opinion… To the extent, however, that the issue requires the determination of the test for what constitutes public interest, the standard of review is correctness” (para. 29). He cited two cases, only one of which dealt with the Alberta Energy and Utilities Board, and he did not appear to be citing it for that reason. The most extreme version of this approach is where the Court barely uses precedent at all. Thus in Wachtler v. College of Physicians and Surgeons of the Province of Alberta, 2009 ABCA 130, the Court essentially identified the question and the standard of review.
A couple of other points about the use of precedent are worth noting. In two cases, the Court appeared to use precedent as part of how it applied the standard of review analysis. Thus in Border Paving Ltd. v. Alberta (Occupational Health and Safety Council), 2009 ABCA 37, the Court said that the past cases on the Council, which noted its expertise and function, suggested more deference, but the statutory right of appeal suggested less. Further, in one case the Court declined to rely on a precedent in part because it predated Dunsmuir (Gift Lake Metis Settlement v. Metis Settlements Appeal Tribunal (Land Access Panel), 2009 ABCA 143). We suggest that neither of these uses of precedent is consistent with the direction of the Court in Dunsmuir. It would render the decision nonsensical to preclude precedent that comes before it, and the incorporation of prior cases into the standard of review analysis complicates the court’s task, it does not simplify it.
True Questions of Jurisdiction and Procedural Fairness
Two smaller points from the post-Dunsmuir jurisprudence are worth noting. First, in none of the cases did the court identify the issue as being a true question of jurisdiction subject to correctness review. In fact in several of the cases the Court drew quite careful and fine distinctions to show how the question was not one of jurisdiction, although a superficial analysis could have let it seem that way. Thus in Macdonald v. Mineral Springs Hospital, 2008 ABCA 273, the issue under appeal was a decision by the Hospital Privileges Appeal Board that it had no jurisdiction to hear an appeal from an Operating Room Committee decision that it would not increase a physician’s operating room time. This looks at first glance like a jurisdictional question – after all, they are deciding whether they have jurisdiction to do something. However, as the Court rightly points out, that isn’t sufficient to make something a jurisdictional question in the relevant sense. A jurisdictional question would be one that says, can the Hospital Privileges Appeal Board even ask this question? Are they even entitled to decide whether they have the power to hear the appeal of the Operating Room Committee on this issue? That type of question isn’t raised here. They do have that power, and the only question is as to the answer – can they hear the appeal or not? That question is a straightforward matter of statutory interpretation, and is reviewed on a reasonableness standard. This is a tricky point, and it is nice to see the Court addressing it properly.
With respect to procedural fairness, as Alice Woolley has discussed elsewhere on ABlawg (see Don’t you forget about me: Remembering the rest of administrative law after Dunsmuir), a few Queen’s Bench decisions appear to think that Dunsmuir applies to questions of procedural fairness. The Court of Appeal in two cases – Nortel Networks Inc. v. Calgary (City), 2008 ABCA 370 and Hennig v. Institute of Chartered Accountants of Alberta (Complaints Inquiry Committee), 2008 ABCA 241 – has properly noted that it does not. In such cases the question for the court is simply, was the procedure fair?
In merging the standards of reasonableness simpliciter and patent unreasonableness, Dunsmuir responded to criticisms that these two standards of deference were indistinguishable in practice. The question remaining subsequent to Dunsmuir is: What does it mean in practice to assess the “justification, transparency, and intelligibility within the decision-making process” and “whether the decision falls within a range of possible acceptable outcomes which are defensible in respect of the facts and law” (Dunsmuir at para. 47)? We set out to assess how the Court of Appeal is answering this question, keeping in mind that being truly deferential is hard work. Simply put, in reviewing the decision of someone else it is very difficult to resist the temptation to impose your own view of matters.
Of the 34 judgments reviewed for this entry, 29 apply the reasonableness standard to review the administrative decision. And of those 29 judgments, 21 defer to the administrative decision (i.e., the Court finds the administrative decision is reasonable). These statistics suggest the Court of Appeal has adopted a deferential approach to substantive judicial review post-Dunsmuir. A closer look at how reasonableness is applied by the Court, however, suggests a more intrusive Court.
In several cases that purport to apply reasonableness, the Court’s reasoning deliberates over whether it agrees with the administrative decision. To borrow from Justice Binnie’s caution in his concurring Dunsmuir opinion (at para. 141), in its reasonableness review the Court of Appeal often reweighs the inputs that led to the administrative decision as if it were the Court of Appeal’s view of reasonableness that counts, rather than limit itself to identifying the outer boundaries of reasonable outcomes. This is most subtle in cases where the Court of Appeal finds the administrative decision to be reasonable, but does so by essentially agreeing with the administrative outcome.
For example, the Court of Appeal’s decision in Calgary (City) v. Alberta (Municipal Government Board), 2008 ABCA 187, involves a reasonableness review of the Municipal Government Board’s decision that the Hudson’s Bay Company is entitled to file a complaint to a city tax assessment on property (a shopping mall) in which the Bay is a tenant. The Municipal Government Board made its decision by interpreting applicable provisions of the Municipal Government Act, R.S.A. 2000, c. M-26, including the defintion of “assessed person” in the legislation. While the Court of Appeal finds the Board’s decision to be reasonable and is purportedly deferential, Madam Justice Elizabeth McFadyen effectively re-decides the interpretive issue by applying various rules of statutory interpretation to the Municipal Government Act and ending with the conclusion that the Board’s interpretation to include the Bay as an “assessed person” is reasonable (paras. 26-37). Perhaps the most telling hint from McFadyen J.A. that she would be deciding the matter for herself is in how she frames the issue for the Court: does s. 460(3) of the Municipal Government Act limit the right of complaint about a property assessment to “the owner of the assessed property?” (at para 12). Rather than canvass the range of possible outcomes here, it seems McFadyen J.A. is intent from the outset to decide for herself whether the Board’s decision is correct.
More obvious illustrations of an intrusive court that purports to be deferential will be those cases where the Court of Appeal sets aside the administrative decision as unreasonable. Thus in United Association of Journeymen and Apprentices of the Plumbing and Pipefitting Industry of the United States and Canada, Local 488 v. Bantrel Constructors Co., 2009 ABCA 84, the Court quashed a labour arbitration panel decision that upheld an employer drug testing policy with more stringent employee testing than the policy set out in the applicable provisions of the collective agreement. In making its decision, the arbitration panel considered the collective agreement, the drug testing policy incorporated into the collective agreement, and policy on employee drug testing generally. The panel’s decision rested on its conclusion that the collective agreement drug testing provisions were not a complete code, and could be augmented by additional requirements. While the Court of Appeal expresses that it is applying reasonableness to review the panel’s decision (at paras. 26-30), the Court’s subsequent application confirms a correctness review. After reviewing the panel’s interpretation of the collective agreement guidelines, the Court bluntly states its view that the panel incorrectly interpreted such guidelines to conclude they were not a complete code on drug testing (at para 37). Based on its view of an incorrect interpretation, the Court finds no line of analysis in the panel’s reasons to support its conclusion (at para 39). In summary, the Court’s disagreement with the panel’s interpretation leads the Court to find the panel’s decision as unreasonable. There is no discussion as to whether the panel’s decision was intelligible, justifiable, and transparent, and no discussion as to whether the panel’s discussion falls within a range of outcomes.
Other notable examples where the Court of Appeal purports to be deferential but is actually quite intrusive are ATCO Gas and Pipelines v. AEUB, 2008 ABCA 200, and Alberta v. Alberta Union of Provincial Employees, 2008 ABCA 258.
This is not to suggest that there are no instances where the Court of Appeal is truly deferential. Good examples of deference applied in the context of a reasonableness review include the Court of Appeal’s decisions in Craig v. Bighorn (Municipal District No. 8), 2009 ABCA 119, The Brick GP Ltd. v. Calgary (City), 2008 ABCA 356, Anderson v. Alberta Securities Commission, 2008 ABCA 184, and MacDonald v. Mineral Springs Hospital, 2008 ABCA 273. One common feature in each of these judgments is the relatively short length (at most, several paragraphs of text) of the Court’s application of the reasonableness standard to the impugned administrative decision, generally finding that either the decision falls within a range of possible outcomes or that there was evidence to support the decision.
Has Dunsmuir simplified the identification of the standard of review? We think not. While it has removed the mechanical, and often somewhat arbitrary, pragmatic and functional analysis from its dominance in selecting the standard of review, Dunsmuir replaces it with a different sort of complexity. Now lawyers and their clients will question what approach the reviewing court will take to selecting the standard. Will it look narrowly at precedent, focusing on the standard of review analysis? Or will it look at precedent more broadly, which in the end means essentially deciding the standard based purely on the nature of the question? These are not simply methodological differences; they are also substantive differences. Application of the standard of review requires consideration of the question and the relative expertise of the tribunal and the court to decide it, given the specific nature of this tribunal. Focusing on precedent and the nature of the question more generally takes away the analysis of specific tribunal competence and expertise.
Similarly, it seems that confusion continues to reign over how to apply reasonableness despite the fact there is now only one deferential standard. Lawyers and their clients are left to speculate as to whether the reviewing court will be more or less deferential (or deferential at all!) in its application of reasonableness. Arguably, the complexities in selecting the standard of review pre-Dunsmuir have simply been shifted into complexities in applying the standard of review post-Dunsmuir.
In the end, we wonder why the Supreme Court does not simply focus the standard of review analysis directly on the point which the old pragmatic and functional analysis was supposed to address: given the nature of this question, who is best positioned to decide it, the court or the decision-maker? And if it is the decision-maker, can the decision be rationally supported by the governing legislation? Everything else – the use of precedent, the analysis of the question, the use of the standard of review analysis – are simply methodologies for answering these questions. Of course, the thorny issues underlying judicial review — such as the separation of powers and judicial versus legislative supremacy – remain as contested as ever. The inability, or perhaps unwillingness, to address these issues head on might explain why the Supreme Court remains focused on methodology rather than substance. For this reason alone, Dunsmuir cannot be the last word on remaking substantive judicial review.