Corps de l’article


Judicial review aims to ensure that administrative action is lawful, reasonable, and procedurally fair.[1] Achieving these goals requires the development and application of doctrine, which is often complex in its design. Complexity results from the inherently technical nature of the discipline, which involves the application of general principles to substantive areas of law that differ greatly in their contours and content, combined with the need for doctrine to conform to normative commitments (such as the principles of good administration and the rule of law).[2] Sitting atop the Canadian judicial hierarchy, the Supreme Court of Canada bears the additional responsibility of developing clear and coherent doctrine, thus providing a set of tools that lower courts can confidently apply to the complex (and not-so-complex) cases that come before them.

In its decision in Dunsmuir v. New Brunswick,[3] the Court attempted to clarify and simplify Canadian judicial review doctrine.[4] I will argue in this paper that the Court got it badly wrong, as evidenced by four of its subsequent decisions. Clarification and simplicity have not been achieved. Indeed, one might playfully suggest that the Court’s decisions fail to meet the standards of justification, transparency, and intelligibility that the Court has deemed central to the conception of reasonableness in Canadian law.

Before Dunsmuir, the standard of review of administrative action was determined by application of the “pragmatic and functional” analysis.[5] During the first stage of the analysis, reviewing courts were charged with examining four factors: whether there was a privative, or conversely an appeal, clause in the decision maker’s home statute;[6] whether the decision maker was relatively more expert than the reviewing court in respect of the decision under review; what the purpose of the statutory scheme and of the particular provision or provisions at issue was; and what the nature of the question in dispute was.[7] The goal was to capture legislative intent and, particularly, the degree of deference the legislature intended that reviewing courts should accord in respect of the decision under review.

The second stage was the selection and application of the standard of review. Depending on the degree of deference indicated by the examination of the four factors, one of the following standards was applied by the reviewing court: the highly interventionist standard of correctness, the moderately deferential standard of reasonableness simpliciter, or the highly deferential standard of patent unreasonableness.[8]

Both stages of the pragmatic and functional analysis were much criticized. The first stage was said to require an overly lengthy analysis of the statutory scheme, which distracted from the merits of the challenge to the decision under review.[9] Of the second stage, it was said that the distinctions between the three standards were unclear and that the standards could be difficult to apply in practice.[10] Thus, in Dunsmuir, the Court announced that it was time to reassess Canadian judicial review doctrine.[11]

The Court made changes to the first and second stages of the pragmatic and functional analysis, itself renamed the “standard of review analysis”.[12] As to the second stage, the Court merged the standards of reasonableness simpliciter and patent unreasonableness. It sought not just to reduce the number of standards of review but also to clarify the meaning of reasonableness. Reasonableness, the Court explained, is concerned with “justification, transparency and intelligibility within the decision-making process,” and also with whether the decision falls within a range of acceptable and rational solutions.[13]

As to the first stage, the Court held that an exhaustive review of the standard of review factors would not always be necessary and could be omitted altogether where previous decisions had clearly established the appropriate standard of review.[14] In Smith v. Alliance Pipeline Ltd., Justice Fish explained the effect of Dunsmuir as the identification of a series of nonexhaustive categories. Decisions falling into these categories are subject to review for either correctness or reasonableness. Correctness applies to (1) constitutional issues; (2) questions of general law “both of central importance to the legal system as a whole and outside the adjudicator’s specialized area of expertise”; (3) the “drawing of jurisdictional lines between two or more competing specialized tribunals”; and (4) “true question[s] of jurisdiction or vires.” Reasonableness is “normally the governing standard where the question: (1) relates to the interpretation of the tribunal’s ... home ... statute or ‘statutes closely connected to its function, with which it will have particular familiarity’; ... (2) raises issues of fact, discretion or policy; or (3) involves inextricably intertwined legal and factual issues.”[15]

The four-factor approach has thus been marginalized in favour of a categorical approach. The former has not been entirely jettisoned, however; the Court has continued to mention the factors,[16] but they are now to be treated as “guideposts”[17] rather than as a road map. In short, it is clear that the categorical approach now drives the analysis.[18]

I have argued elsewhere that the Court’s reform efforts were flawed in principle.[19] In the present article, I focus closely on four of the Court’s recent decisions, which demonstrate the flaws with the Dunsmuir approach. As to the first stage of the analysis, the cases demonstrate that the categorical approach is unworkable and that a reviewing court cannot in fact apply the categorical approach without reference to the much-maligned four factors (or some variant thereon). The categories regularly come into conflict, in that decisions could reasonably be assigned to more than one category. When conflict occurs, it must be resolved by reference to some factor or factors external to the categorical approach. As to the second stage, the single standard of reasonableness is similarly impractical. It is not enough to say that “[r]easonableness is a single standard that takes its colour from the context.”[20] Different types of decisions attract different degrees of deference, and they do so on the basis of a factor or factors external to the elegant elucidation of reasonableness offered in Dunsmuir.

I will address each of the four recent cases in turn. I start with the two that exemplify the flaws of the first stage of the analysis and conclude with the two that, in addition to illustrating problems with the first stage, demonstrate the flaws of the second stage.

I. Alberta (Information and Privacy Commissioner) v. Alberta Teachers’ Association[21]

In this case, the Court artfully avoided creating a conflict between classifying the relevant issue as a jurisdictional question or an interpretation of a home statute, which prevented the flaws of the categorical approach from being fully exposed. Nevertheless, the decision demonstrates that application of the categorical approach still requires reference to external factors.

In 2005, several individuals complained to Alberta’s information and privacy commissioner that their data had been disclosed contrary to the provisions of the Personal Information Protection Act (PIPA). The commissioner commenced an investigation into the complaints.[22] At the time, subsection 50(5) of the PIPA provided for stringent timelines:

An inquiry into a matter that is the subject of a written request referred to in section 47 must be completed within 90 days from the day that the written request was received by the Commissioner unless the Commissioner

  1. notifies the person who made the written request, the organization concerned and any other person given a copy of the written request that the Commissioner is extending that period, and

  2. provides an anticipated date for the completion of the review.[23]

An investigation was commenced in October 2005. A report was issued in July 2006. Shortly thereafter, the complainants requested an inquiry. The inquiry was completed in March 2008.

Between the commencement of the investigation and the completion of the inquiry, the commissioner used its power to extend the ninety-day timeline. But this was not done until August 2007, some twenty-two months after the initial complaint had been received.

The Alberta Teachers’ Association sought judicial review of the commissioner’s decision on the basis that the commissioner had lost jurisdiction by failing to exercise its subsection 50(5) power within the ninety-day timeline. This argument prevailed at first instance and in the Alberta Court of Appeal,[24] but not before the Supreme Court.

One can easily observe how the categorical approach fell short. There is a strong argument that the issue was purely a true question of jurisdiction or vires. There were statutory preconditions with which the commissioner had to comply in order to remain seized of the matter. The commissioner had failed to do so and, by virtue of that failure, forfeited jurisdiction. Equally, however, there is a strong argument that the commissioner was interpreting its home statute, on a question with which it had some familiarity. Matters relating to compliance with timelines are likely to arise often in the course of a decision maker’s work. Different factual contexts might demand different responses from the commissioner. Whether strict compliance is necessary or whether the purposes of the statutory scheme would be furthered by a less rigorous approach is arguably a question of policy and fact. The application of the categorical approach could justify imposing either the reasonableness standard or the correctness standard, and is incapable of resolving this conflict.

Justice Rothstein, writing for the majority, purported to follow the categorical approach. He determined that the decision under review did not fall into any of the correctness categories. Instead, it involved an interpretation of the home statute and thus attracted a standard of reasonableness. The influence of the standard of review analysis is discernible, however, in Justice Rothstein’s determination that the question related to the home statute. Justice Rothstein noted that the decision was “squarely” within the “specialized expertise” of the decision maker and that the decision maker had “significant familiarity” with the issue, which was “specific” to the PIPA.[25] The expertise of the decision maker and the raison d’être of the commissioner (i.e., balancing competing interests) are factors that would have been taken into account in a standard of review analysis.

In the same passage, Justice Rothstein stated that the parties’ interests in the conclusion of inquiries in a timely manner, in being kept informed, and in the effect of automatic termination of privacy investigations all fell within the commissioner’s role, which “centres upon balancing” the rights of individuals to privacy with organizations’ needs to disclose information in certain circumstances.[26] The complexity of the commissioner’s balancing exercise would also have been considered under the standard of review analysis. This passage was critical to the decision, yet note that the factors discussed are external to the categorical approach, which the Court claims should be driving the analysis of the reviewing court.

It is notable as well that Justice Rothstein expressed significant skepticism about the categorical approach’s category of true questions of jurisdiction. He pronounced himself “unable to provide a definition” of such a question[27] but twice emphasized that true jurisdictional questions are “exceptional”.[28] While this acknowledgment was commendable,[29] it made the application of the categorical approach artificially easy. Questions relating to timelines might be frequent enough to ground an argument that the administrative approach to timelines is a general question of law of central importance to the legal system, but the true conflict here was between jurisdictional questions and questions pertaining to the decision maker’s home statute. By defining jurisdictional questions almost out of existence, Justice Rothstein avoided this conflict. It was just as well that he did, for it is difficult to think of a better example of a jurisdictional error than a failure not just to comply with a time limit (which is often excusable)[30] but also to exercise a power to extend the time limit. Justice Rothstein’s artful avoidance of a direct conflict between the category of jurisdictional questions and the category of interpretations of a home statute prevented the shortcomings of the categorical approach from being further exposed.

Justice Rothstein then assessed the reasonableness of the decision. He was presented with an interesting problem in that the commissioner had not formally given a decision on the timeline issue. The point was first raised in argument at first instance and was not, therefore, explicitly decided by the commissioner, but Justice Rothstein was satisfied that no prejudice resulted from considering it. Moreover, a finding that the failure to extend the timeline was not fatal to the ability to continue the processes of investigation and inquiry was necessarily implicit in the commissioner’s decision.[31] The key factor here was that, in light of Dunsmuir’s admonition to look to the reasons that “could be offered” for a decision,[32] a plausible basis for the commissioner’s implicit decision was readily available. In related decisions of the commissioner in respect of the PIPA and the Freedom of Information and Protection of Privacy Act,[33] the latter of which is also within the commissioner’s purview, a consistent approach had been taken to the timelines issue. Because a reasonable basis for the decision was apparent and no prejudice would be suffered, Justice Rothstein concluded that the absence of a set of formal reasons from the commissioner should not inhibit the Court from conducting an analysis of the commissioner’s approach.[34]

Justice Rothstein concluded that the implicit decision in the instant case was reasonable. First, it was consistent with the approach taken by the commissioner in similar cases. Second, the text of subsection 50(5) was ambiguous, in that the ninety-day limit could be taken as referring either to completion of the inquiry or to extension of the inquiry, such that the power of extension would not necessarily lapse after the period had expired. Third, the decision was also consistent with the purposes of the statutory scheme: subsection 50(5) aimed in a general sense at keeping parties informed, a purpose not subverted by permitting implicit extensions of the timeline after the ninety days were up. Finally, the decision was well reasoned, detailed, and transparent.[35]

As to the broader question of whether the category of jurisdictional questions ought to be retained, Justice Rothstein’s reasons indicate an openness to abolishing the category altogether.[36] This is unsurprising, given the weak theoretical basis for the category[37] and the historical difficulty in applying the concept of jurisdictional error in a clear and coherent manner.[38]

II. Nor-Man Regional Health Authority Inc. v. Manitoba Association of Health Care Professionals[39]

In Nor-Man, the categorical approach proved incapable of resolving a conflict between a question of general law and one of fact, policy, and discretion. Again, reference to external factors was necessary to determine the appropriate standard of review.

This case concerned the interpretation of a collective agreement by a labour arbitrator. Ms. Jacqueline Plaisier had been employed by the regional health authority for twenty years. For some of that period, she was a casual employee. Once Ms. Plaisier had accumulated twenty years of service, she claimed an additional week of paid leave under the collective agreement. This claim was denied by the authority, based on the fact that it had consistently interpreted the collective agreement as not permitting workers’ time as casual employees to be taken into account in determining their entitlement to additional paid leave. When Ms. Plaisier’s time as a casual employee was not taken into consideration, she had not accumulated the required twenty years of service.

Significantly, Ms. Plaisier’s union had never previously mounted a challenge to the authority’s interpretation of the collective agreement. Over five collective agreements spanning a twenty-year period, the authority’s interpretation had been consistently applied. On this occasion, however, her union adopted Ms. Plaisier’s position that time as a casual employee should be taken into account.

A single arbitrator appointed under The Labour Relations Act[40] (LRA) concluded that the union’s interpretation was correct. But the arbitrator also concluded that the union was estopped from challenging the regional health authority’s interpretation. After years of acquiescence, it would be unfair for the union to prevail. The authority was entitled to assume that the union had accepted its practice and to act on that assumption.

The union sought judicial review. It was unsuccessful at first instance, where a standard of review of reasonableness was applied in upholding the arbitrator’s decision.[41] On appeal, the union persuaded the Manitoba Court of Appeal to apply a standard of correctness. As the court of appeal commented, defining the parameters of promissory estoppel was a matter within the purview of the ordinary courts and outside that of labour arbitrators.[42] The court of appeal went on to quash the decision, on the basis that the arbitrator had wrongly applied the common law concept.[43]

It is difficult to see how the categorical approach could have resolved this conflict. On the one hand, as the court of appeal held, estoppel is a quintessentially common law principle, which arbitrators are not necessarily expert at applying. On the other hand, there is much to commend the view of the judge at first instance, as the application of concepts such as estoppel in the context of complex labour relations disputes will invariably involve important considerations of fact and policy.[44]

How did the Supreme Court of Canada resolve the impasse? Writing for a unanimous Court, Justice Fish emphasized the unique position of labour arbitrators. He first drew attention to their broad statutory and contractual mandates. He linked this to their role in fostering industrial peace and harmonious relations between management and unions. He then emphasized their expertise in dealing with labour relations issues. Finally, he noted that “arbitrators are uniquely placed to respond to the exigencies of ... employer-employee relationship[s].”[45]

Significant reliance on the standard of review analysis factors was necessary to determine the appropriate standard of review. Starring roles were played by expertise and statutory purpose. In addition, Justice Fish relied on additional factors, namely the extent of the statutory mandate and the complexity of the problem underlying the decision in issue.

But at least two dangers lie in this resort to contextualism.[46] The first is that the identification of the appropriate standard of review will turn on whatever factors the reviewing court considers relevant in the case before it. This puts litigants in a difficult position. At least with the previous standard of review analysis it was clear what factors would be relied upon. Now, it is entirely unclear. The second danger is that the factors will differ from sector to sector. In other words, what is relevant to labour arbitrators may not be relevant to ministers, municipalities, or telecommunications regulators. Canada could end up with a balkanized administrative law, in which different principles apply depending on the subject matter of the decision maker’s jurisdiction, even where the relevant legal frameworks are substantively similar. The Court should think long and hard before giving in too easily to the siren call of context.

These important concerns about clarity and balkanization notwithstanding, the central point is that, in determining the standard of review, the work was not done by the categorical approach, but by factors external to it.

Having held that the appropriate standard of review was reasonableness, Justice Fish had no difficulty in determining that the arbitrator’s decision should be upheld. It was transparent, intelligible, and, indeed, coherent; it was based on the evidence, submissions of the parties, and the arbitrator’s analysis; and it was consistent with the precedents relied upon by the arbitrator in reaching a conclusion.[47]

To the submission that the arbitrator’s application of promissory estoppel failed to follow the test laid down in Maracle v. Travellers Indemnity Co. of Canada,[48] Justice Fish held simply that the arbitrator had “adapted and applied the equitable doctrine of estoppel in a manner reasonably consistent with the objectives and purposes of the LRA, the principles of labour relations, the nature of the collective bargaining process, and the factual matrix of Ms. Plaisier’s grievance.”[49] Some relief exists in Nor-Man, then, for those who saw the court of appeal decision as an interventionist aberration.[50]

III. Canada (Canadian Human Rights Commission) v. Canada (Attorney General)[51]

This case again exposes the categorical approach’s inability to identify the appropriate standard of review where there is conflict between categories. Moreover, it lays bare some of the shortcomings of the conception of reasonableness developed in Dunsmuir.

Ms. Donna Mowat was an employee of the Canadian Forces for a period of fourteen years before she was compulsorily released in 1995. Ms. Mowat had made numerous complaints against her employers over the years and, subsequent to her release, filed a complaint with the Canadian Human Rights Commission. She alleged sexual harassment, adverse differential treatment, and failure to continue employment on the basis of sex.

Ms. Mowat’s claim was heard by the Canadian Human Rights Tribunal. After a hearing that lasted several weeks and required the consideration of voluminous evidence, Ms. Mowat was partially successful: her sexual harassment claim was upheld, though her other complaints were not. Ms. Mowat received an award of $4,000. In addition, the tribunal awarded her legal costs, in the amount of $47,000. The award of costs was made by virtue of paragraphs 53(2)(c) and 53(2)(d) of the Canadian Human Rights Act (CHRA). These paragraphs provide that, where an adverse finding has been made against a respondent, the tribunal may order

  1. that the person compensate the victim, as the Tribunal may consider proper, for any or all of the wages that the victim was deprived of and for any expenses incurred by the victim as a result of the discriminatory practice; and

  2. that the person compensate the victim, as the Tribunal may consider proper, for any or all additional cost of obtaining alternative goods, services, facilities or accommodation and for any expenses incurred by the victim as a result of the discriminatory practice.[52]

The question that occupied the reviewing courts was whether the effect of these provisions was to vest the power to award legal costs in the tribunal.[53]

What would be the appropriate standard of review? On the one hand, one might agree with the Federal Court, which held that the tribunal was interpreting its home statute and as such was entitled to deference.[54] On the other hand, the position of the Federal Court of Appeal, which held that the tribunal was answering a question of general law of central importance to the legal system as a whole, seems equally plausible.[55] Many tribunals have the power to award expenses, and a finding that the human rights tribunal is entitled to award legal costs would surely resonate throughout federal and provincial administrative structures.

At the Supreme Court of Canada, Justices LeBel and Cromwell held that the appropriate standard was that of reasonableness. In setting out the categorical approach, and presaging the view taken by Justice Rothstein in Alberta Teachers’ Association, Justices LeBel and Cromwell held that the category of jurisdictional questions was to be construed “narrowly”.[56] Ultimately, however, the conflict between the categories was resolved by reference to external factors. Justices LeBel and Cromwell held that the question of whether legal costs incurred because of discriminatory conduct constituted expenses within the meaning of the CHRA was one “within the core function and expertise” of the tribunal. They noted that the question was “inextricably intertwined with the Tribunal’s mandate and expertise to make factual findings relating to discrimination.”[57] They further added that, given the tribunal’s familiarity with making such factual determinations, the tribunal was well positioned to make assessments as to the necessity of awarding legal costs.[58] Finally, they emphasized that the question required a fact-sensitive inquiry.[59]

In determining the appropriate standard of review, Justices LeBel and Cromwell relied on the tribunal’s expertise, the factual nature of the question, and the purpose of the statutory provisions at issue. In other words, while Justices LeBel and Cromwell purported to apply the categorical approach,[60] the real drivers of the choice of standard of review were the much-maligned standard of review analysis factors.

Justices LeBel and Cromwell proceeded, however, to apply a stringent standard of reasonableness to the decision. The question, they stated baldly, was one of statutory interpretation, calling for the application of Driedger’s modern principle.[61]

Justices LeBel and Cromwell first examined the text of the provision. Significant concern was caused by the appearance of the word “expenses” in both paragraphs 53(2)(c) and 53(2)(d). Accepting the tribunal’s interpretation, Justices LeBel and Cromwell reasoned, would render “the repetition of the term ‘expenses’ redundant and [fail] to explain why the term is linked to the particular types of compensation described in each of those paragraphs.”[62] This violated the presumption against tautology (i.e., that the “legislature avoids superfluous or meaningless words”).[63] In addition, in legal parlance, the term “costs” was said to have a well-understood meaning different from that of “expenses”. Had Parliament intended to grant the authority to make costs awards, it could simply have employed the term “costs”.[64] Finally, the tribunal did not have the power to make financial awards in excess of $5,000; to allow it to make potentially unlimited awards of legal costs would do violence to an express limitation on the tribunal’s authority.[65]

Justices LeBel and Cromwell then examined the surrounding context: the legislative history, the commission’s understanding of its authority to award costs, and parallel provincial and territorial legislation. The legislative history bolstered the justices’ view that the term “costs” was a term of art that Parliament could have included had it been so minded: previous, unenacted iterations of human rights legislation have expressly included a power to award costs. Instead, the commission was given a role that could include litigating on behalf of claimants.[66] This legislative choice obviated the need to grant the tribunal a free-standing power to make costs awards. That the tribunal did not have such a power was suggested by the commission’s own interpretation of section 53: “[T]he Commission itself has consistently understood that the CHRA does not confer jurisdiction to award costs and has repeatedly urged Parliament to amend the Act in this respect.”[67] Finally, reference to provincial and territorial legislation tended “to confirm the view that the word ‘costs’ is used consistently when the intention is to confer the authority to award legal costs.”[68] For Justices LeBel and Cromwell, the absence of this magic word from the CHRA was telling.

Lastly, the justices examined the purposes of the CHRA. While they acknowledged that the legislation ought to be given a broad and liberal interpretation, the justices held that “a liberal and purposive interpretation cannot supplant a textual and contextual analysis simply in order to give effect to a policy decision different from the one made by Parliament.”[69] Here, “the text, context and purpose of the legislation clearly show that there is no authority in the Tribunal to award legal costs and that there is no other reasonable interpretation of the relevant provisions.”[70]

Justices LeBel and Cromwell essentially applied a standard of correctness: they judged the tribunal’s decision by reference to the principles of statutory interpretation, presumably—though they did not put it in these terms—because the relevant context was one of statutory interpretation. Indeed, Justices LeBel and Cromwell rebuked the tribunal for not having employed legal principles: its decision to adopt a dictionary meaning of “expenses” and to articulate a beneficial policy outcome “led the Tribunal to adopt an unreasonable interpretation of the provisions [in issue].”[71]

Even if Justices LeBel and Cromwell have (slightly) the better of the argument on the interpretation of the term “expenses”, this rebuke was not justified. First, the approach demanded by the justices is unrealistic. Administrative tribunals are not courts. They do not always have the same access as judges to legal databases, a battery of eager clerks, and most tellingly of all, to the assistance of learned counsel. It is hard to see how they can realistically be held to the highest legal standards.

Second, Justices LeBel and Cromwell’s substitution of their own judgment on the meaning of the term “expenses” was inappropriate. If administrative tribunals are required to act like courts, with a corresponding reliance on legalistic procedures, the advantages of simplicity, accessibility, and flexibility that they offer will be lost.[72] The tribunal should not be obliged to articulate its conclusions in purely legal terms. Rather, its conclusions should be informed by the overarching purposes of the statutory scheme, the context presented by individual cases, and the tribunal’s experience applying the provisions of the CHRA to diverse factual situations.

Moreover, the standard of review analysis indicated that the question was one that the legislature intended the tribunal to resolve. The very purpose of the standard of review analysis is to determine whether a question should be answered by the reviewing court (correctness) or by the decision maker in question (reasonableness). To employ a nondeferential standard of reasonableness is to undermine the standard of review analysis. No justification was offered for the intensity of the review, beyond the summary conclusion that the question was one of statutory interpretation. Given that the authority to determine the meaning of “expenses” had been vested in the tribunal and not in the courts, Justices LeBel and Cromwell’s insistence that the tribunal interpret statutory provisions as if it were a court flew in the face of legislative intent. Had Parliament intended the provision in question to be interpreted in a judicial manner, it could have charged the courts with doing so.[73]

Where the standard of review is that of reasonableness, the onus is on the applicant to demonstrate some shortcoming in the decision or decision-making process.[74] To impose a standard of review of correctness is to subvert review for reasonableness. To put the point starkly, the applicant is given a second kick at the can: an argument that failed to convince the decision maker can be run again in the alternative forum of the reviewing court. Whatever the merits of the ultimate conclusion reached by Justices LeBel and Cromwell, the route by which they arrived there is a dangerous one. Indeed, if the Court’s treatment of the category of jurisdictional questions in Alberta Teachers’ Association suggests a willingness to narrow the categories of correctness review in order to be more deferential,[75] this aim will be frustrated if the standard of reasonableness is applied in a nondeferential manner.

IV. Catalyst Paper Corp. v. North Cowichan (District)[76]

The final case is most remarkable for its reintroduction of a third standard of review, but it also provides an important lesson on the shortcomings of Dunsmuir’s unified standard of reasonableness.

Catalyst Paper is a major corporation that has four mills in Western Canada. One of these is located in the district of North Cowichan and has been there for a long time.[77] More recently, the residential population of the district has increased dramatically, straining the district’s resources. Reluctant to raise residential taxes for fear of driving out newcomers, the district instead began to incrementally raise the rates on major industrial property, effectively, Catalyst Paper’s mill. By the time Catalyst Paper applied for judicial review of the most recent municipal taxation bylaw, its tax rate was over twenty times greater than the rate that applied to residents. Both the British Columbia Supreme Court[78] and Court of Appeal[79] upheld the bylaw.

Catalyst Paper conceded that the bylaw was reviewable on a standard of reasonableness but argued that the district could not meet the standards of justification, intelligibility, and transparency mandated by Dunsmuir. In agreeing that the standard of review was that of reasonableness, the Supreme Court of Canada clarified a point that had been unclear:[80] the same general principles apply to the review of municipal bylaws as apply to the review of administrative action. Delivering the judgment of the Court, Chief Justice McLachlin overruled Thorne’s Hardware Limited v. The Queen,[81] which stood for the proposition that delegated legislation could be reviewed only to determine whether it is ultra vires and not for substantive reasonableness because the latter was related to policy and was not within the purview of the courts. This position is untenable, as Chief Justice McLachlin explained, because “[t]he rule of law insists on judicial review to ensure that delegated legislation complies with the rationale and purview of the statutory scheme under which it is adopted.”[82]

However, in determining the appropriate standard of review, the chief justice did not employ the categorical approach. Chief Justice McLachlin first noted that municipal taxation decisions “involve an array of social, economic, political and other non-legal considerations. ... In this context, reasonableness means courts must respect the responsibility of elected representatives to serve the people who elected them and to whom they are ultimately accountable.”[83]

Then, Chief Justice McLachlin held that previous approaches to judicial review of particular types of decision should be taken into account in determining the standard of review.[84] Accordingly, fastening on a suggestion made by the court of appeal,[85] the chief justice held that the applicable standard was whether the bylaw was so unreasonable that no reasonable municipality could have enacted it.[86]

As to whether or not the bylaw was unreasonable, Chief Justice McLachlin held that it was not. The municipality was not bound to take only objective considerations into account; it was entitled to consider “broader social, economic and political issues.”[87] While Chief Justice McLachlin acknowledged that the bylaw’s effect on Catalyst Paper was extremely harsh, this was not sufficient to justify striking it down. The district council was entitled to take into account the countervailing considerations, which included the impact of a sudden hike in residential property taxes on long-term, fixed-income residents, and a municipal policy of gradually working toward greater equalization of tax rates.[88] There may be much to commend the substantive outcome here, which expresses faith in the democratic process and in the ability of municipalities to reach rational results. But the route by which this outcome was reached is again problematic.

On the face of it, the Catalyst standard is easier to meet, from the point of view of the municipality, than the Dunsmuir standard. There is no mention of justification, intelligibility, and transparency, qualities that even a charitable observer would not attach to many instances of municipal decision making (or perhaps legislating tout court!).

However, the standard emerges from a curious reading of Dunsmuir. In Catalyst, the Court effectively reintroduced a third standard of review: municipal bylaws can be struck down only where they are so unreasonable that no reasonable municipality could have enacted them. This undermines the Dunsmuir project, which was to simplify judicial review of administrative action by reducing the number of standards of review and clarifying the meaning of reasonableness. When the surrounding circumstances are taken into account, the Court’s statement in Dunsmuir that previous jurisprudence is relevant to determining the standard of review must have been aimed at situations in which earlier cases had identified the appropriate standard of review for a certain type of decision taken under a particular statutory scheme.

Whether it is consistent with the spirit of Dunsmuir or not, the effect of Catalyst has been to increase the number of available standards of review. Moreover, it is inevitable that in subsequent challenges to delegated legislation or to high-level policy decisions taken by elected officials (such as the Governor-in-Council), parties will argue that this new standard should also apply to them.[89] In addition, the Catalyst standard will confuse the meaning of reasonableness. It is in substance identical to Wednesbury unreasonableness, which has been subjected to sustained attack because of its opaque nature.[90] This is all to the detriment of clarity.

One appreciates the bind in which the Court found itself. Municipal bylaws certainly do not lend themselves to an analysis based on justification, intelligibility, and transparency: elected representatives have never been judicially required to furnish written reasons for their legislative decisions. But having jettisoned the third standard of review in Dunsmuir, the Court found itself without an appropriate standard to apply in Catalyst.

I would venture that, in Dunsmuir, the Court spoke too soon. A third standard is necessary. As the court of appeal had noted in Catalyst,[91] the spectrum of factors in Catalyst cried out for a highly deferential approach. The decision was taken by an elected body, one that was capable of hearing evidence and argument from those affected. It was expert in terms of meeting the needs of those living in its area. It was addressing a complex problem. Moreover, these are surely the reasons for which the authority to pass bylaws was vested in the municipality in the first place.[92] Any reasonable reading of the home statute would indicate that a highly deferential standard of review must have been intended by the legislature. Indeed, the Court was motivated by these factors in deciding to apply its new Catalyst standard rather than the unified reasonableness standard elucidated in Dunsmuir.[93]

Finding a workable third standard of review is difficult, however.[94] In his concurring reasons in Alberta Teachers’ Association,[95] Justice Binnie suggested that varying standards of reasonableness are appropriate, based on his observation that “‘[r]easonableness’ is a deceptively simple omnibus term which gives reviewing judges a broad discretion to choose from a variety of levels of scrutiny from the relatively intense to the not so intense.”[96] With respect to administrative decisions, such as the one at issue in Canadian Human Rights Commission, which involve an interpretation of a statutory provision, a standard close to correctness will be appropriate.[97] For discretionary decisions, such as the one at issue in Catalyst, a highly deferential standard will be appropriate.[98] Once again, one can perceive the lure of the standard of review analysis factors: for Justice Binnie (and, apparently, the Court) the nature of the question is critical. Questions of law attract little or no deference; exercises of discretion in a fact- and policy-sensitive setting attract great deference.

Justice Binnie is correct that reasonableness depends on context. Nonetheless, his analysis is problematic. For Justice Binnie, the degree of deference should turn on a distinction between law and discretion. This distinction was specifically rejected as a touchstone of judicial review doctrine in Baker v. Canada (Minister of Citizenship and Immigration),[99] and properly so.[100] If it is inappropriate for judicial review doctrine to hinge on a distinction between law and discretion, it must be inappropriate for the intensity of reasonableness review to hinge on it. Otherwise, why not choose a standard of review based on the distinction between law and discretion? If one finds Justice Binnie’s approach appealing, one should defend it in terms of the distinction between law and discretion, rather than seeking to send through the back door something that was rejected at the front.

Moreover, given the difficulty that the Court has had in applying its post-Dunsmuir framework, it is by no means clear that doubling down on the categorical approach represents a workable solution. Sweeping problems under the carpet will not smooth out the fabric of administrative law; rather, it will cause it to bulge and crease. Such is the inevitable effect of a formalistic approach to judicial review:

Formalism is formal in that it requires judges to operate with categories and distinctions that determine results without the judges having to deploy the substantive arguments that underpin the categories and distinctions. Since those categories and distinctions must take on a life of their own in order to operate in this detached way, they are capable of determining results that contradict the very arguments for these categories and distinctions.[101]

The search for a workable third standard of review will have to continue.

More generally, taking Justice Binnie’s point as descriptively accurate, the Court’s approach demonstrates the flaws in the notion that reasonableness takes its colour from its context. Context has not been defined by the Court; instead, the standard of review seems to vary depending, on the nature of the question—although, in the absence of a conclusive statement from the Court, it is hard to so state with certainty.

It is telling that, despite the emphasis on the categorical approach, the Court apparently has regard to the nature of the question that was before the administrative decision maker when determining how intrusive the standard of review ought to be. It is discomfiting that the nature of the question plays such an important role in practice even though its importance has been diminished in the rhetoric of the Court. If the Court is to continue to insist that reasonableness is a single standard that takes its colour from its context, the Court ought to be more open about the factors that comprise the context.


If one were—perhaps mischievously—to judge the Supreme Court’s recent decisions on standard of review by reference to the virtues of justification, intelligibility, and transparency, one would have to conclude that the flaws in the Dunsmuir approach cause the Court to come up short.

In terms of intelligibility, it is not clear what is driving the choice of standard of review in recent cases. The Court’s rhetoric suggests the categorical approach, but a more close analysis suggests that the standard of review analysis factors, or some variant thereon, are the real motor. Moreover, apart from Justice Binnie’s concurring reasons in Alberta Teachers’ Association, no attention has been paid to the pressing question of what constitutes context for the purposes of applying the reasonableness standard.

In terms of transparency, it is not clear what factors will push the Court to choose correctness or reasonableness as the standard of review. Litigants and litigators are likely to become more confused in the future than they ever were by the standard of review analysis: at least the categorical approach’s predecessor comprised four fixed factors. Furthermore, the factors that the Court will take into account in determining context, when it comes to assessing the reasonableness of a decision, remain opaque.[102]

Finally, in terms of justification, I have tried to demonstrate in this paper that the use of the categorical approach will almost inevitably lack justification. Rather than examining the substance of the statutory scheme, the intricacies of the decision-making process, and the characteristics of the decision maker, the choice between correctness and reasonableness is made by definitional diktat. The reviewing court simply holds that a decision falls into a particular category and thus attracts the corresponding standard of review. Of course, I have pointed out that, in truth, a variety of factors influence the choice. But to hide these factors from view rather than assess them openly is to fail to meet the requirement of justification. And again, the judicial ability to rely on a mishmash of relevant factors in determining context is most disconcerting.

This is not necessarily to say that the Court is being substantively unreasonable. Calibrating judicial review doctrine is a difficult undertaking, particularly when the Court has not been presented with a raft of viable alternatives.[103] Rather, despite its good faith efforts to do so, the Court has not yet succeeded in formulating clear and coherent standard of review doctrine.