306

Chapter 9 Standard of Review: Back to the Future?

in s. 18.1(4) of the Federal Courts Act.135 The statute was silent about the applicable standard
of review, except to state that erroneous findings of fact warranted relief if made in a perverse or capricious manner or without regard for the material before it.136 Although Rothstein J. agreed with the majority that the appropriate standard of review was reasonableness,
he insisted that the basis for deference lay exclusively in the language of the statutory provision dealing with errors of fact, not in any broader deferential stance derived from the common law.
According to Rothstein J., where Parliament intended a deferential standard of review
in s. 18.1(4), it used clear and unambiguous language. The necessary implication is that
where Parliament did not provide for deferential review, it intended the reviewing court to
apply a correctness standard as it does in the regular appellate context.137 So, unless a statute contains a privative clause governing the entire statute or a provision stipulating deference in relation to a particular ground of judicial review (as in errors of fact under the
Federal Courts Act), the default standard of review is correctness.
The majority of the Court did not share Rothstein J.s nostalgia, but nor did it resolve the
question of how to assign unique weight to the privative clause. Instead, the majority judgment in Khosa takes a perplexing detour through judicial discretion to deny prerogative
relief en route to its conclusion. According to Binnie J., the standard of review structures
residual judicial discretion to deny prerogative relief. Judicial deference to administrative
decisions is thus recast as the method by which judges determine whether and why to withhold a remedy. The majority determined that deference was warranted in the case at bar and
the appropriate standard of review is reasonableness, but not before reviving the Pushpanathan factors and offering the following gloss on the standard-of-review analysis: A privative
clause is an important indicator of legislative intent. While privative clauses deter judicial
intervention, a statutory right of appeal may be at ease with it, depending on its terms.138
Rather than clarify the weight that a privative clause exerts in favour of deference, Khosa adds
uncertainty about the weight that a statutory appeal exerts against it. If nothing else, the
application of the standard-of-review analysis in Khosa suggests equivocation on whether
Pushpanathans multifactor balancing approach can co-exist alongside Dunsmuirs defeasible
rule methodology.139 Nevertheless, they remain difficult to reconcile methodologically.

B. Whatever Happened to Jurisdiction?

Recall that the jurisdictional question arose as a judicial escape hatch from the strictures of
privative clauses: a decision, determination, or order that exceeded the jurisdictional boundaries conferred by statute was a nullity and, therefore, not a genuine decision, determination,
135 R.S.C., 1985, c. F-7.
136 Ibid., s. 18.1(4)(d).
137 Khosa, supra note 133 at para. 117.
138 Ibid. at para. 55.
139 For other examples of equivocation, see Nolan v. Kerry (Canada) Inc., 2009 SCC 39, [2009] 2 S.C.R. 678; NorMan Regional Health Authority Inc. v. Manitoba Association of Health Care Professionals, 2011 SCC 59, [2011]
3 S.C.R. 616.

VI. Dunsmuir: And Then There Were Two

307

or order insulated from judicial review. But the demotion of privative clauses cast into question the raison dtre of its foil, the jurisdictional question. After Southam, a court could both
justify deference in the absence of a privative clause and justify correctness scrutiny in the
presence of a privative clause. Eventually, the jurisdictional question lost its formal, conceptual moorings and became merely a label affixed to the outcome reached by a judicial balancing of the four factors summarized in Pushpanathan. Serious attention to formal attributes
of jurisdiction (authority over subject matter, parties, or remedy) virtually disappeared.
As applied by subsequent courts, Dunsmuir seemed to relinquish the Pushpanathan
balancing test. But the judgment also revived the formal idea of jurisdiction as a boundarydrawing concept capable of rebutting a presumption of deference. The majority also invoked
the dictum in CUPE that urged the courts to be sparing in their resort to the formal claim
of jurisdiction. Thus far, the post-Dunsmuir Supreme Court seems committed to exercising
restraint in labelling an issue as jurisdictional and thereby subject to the stricter standard of
correctness. The best proof lies in Dunsmuir itself. Without expending much effort, the
Court could have transformed the question does the statute authorize the adjudicator to
inquire into the existence of cause for dismissal? into does the adjudicator have jurisdiction to inquire into the existence of cause for dismissal? Yet, the Court refrained from even
posing the question in jurisdictional terms. The adjudicator had jurisdiction over the parties
(the employer and employee) and over the subject matter (discharge, suspension, or other
financial penalty), and that sufficed. In Smith v. Alliance Pipeline,140 the Court swiftly disposed of an argument that the definition of costs under an expropriation statute is jurisdictional. Arbitration committees doubtless have the authority to make the inquiry
whether costs under s. 99(1) refer solely to costs incurred in the proceedings before them,
a determination that plainly falls within their statutory grant of power.141
The post-Dunsmuir Court has been so alert not to brand something as jurisdictional that
the question before the courts is no longer is this thing a jurisdictional question?, but is
there such thing as a jurisdictional question? In Northrop Grumman Overseas Services
Corp. v. Canada (Attorney General),142 the Court listed recent pre-Dunsmuir cases that
treated as jurisdictional the question of whether a U.S. supplier had standing to make a
complaint to the Canadian International Trade Tribunal (CITT). The Court noted that all
parties accepted that the earlier case law remained authoritative in imposing a standard of
correctness. Because the earlier jurisprudence determined the standard of review in a satisfactory manner, the Court was relieved of the task of conducting a fresh standard-of-review
analysis.143 Later, in Alberta (Information and Privacy Commissioner) v. Alberta Teachers

140 2011 SCC 7, [2011] 1 S.C.R. 160 [Smith].

141 Dunsmuir, supra note 112 at para. 36.
142 2009 SCC 50, [2009] 3 S.C.R. 309 [Northrup Grumman].
143 Ibid. at para. 10. For an interesting application of the correctness standard of review to the discharge of the
Crowns duty to consult a First Nation before approving the grant of surrendered lands for agricultural purposes to a non-Aboriginal applicant, see Beckman v. Little Salmon/Carmacks First Nation, 2010 SCC 53,
[2010] 3 SCR 103. While the adequacy of consultation was subject to correctness, the Court found no breach
of the duty to consult, nor a breach of procedural fairness. The ultimate exercise of discretion regarding the
grant application was reviewable on the basis of reasonableness.

308

Chapter 9 Standard of Review: Back to the Future?

Association,144 a majority of the Court distanced itself from Northrop Grumman, claiming
that it was based on pre-Dunsmuir jurisprudence applying a correctness standard to this
type of decision, not on the Court finding a true question of jurisdiction.145 It seems
troubling to rely on precedent to defend a standard of review based on a superceded legal
test, while hinting that the current test would have yielded a different result. Precedent typically binds unless and until the legal test that produced it changes, and not thereafter. Indeed, only two years later, Alberta (Information and Privacy Commissioner) directly confronted the post-Dunsmuir endurance of jurisdictional questions in the context of the
standard-of-review analysis.
Writing for six judges, Rothstein J. ventured that the time has come to reconsider
whether, for purposes of judicial review, the category of true questions of jurisdiction exists
and is necessary to identifying the appropriate standard of review.146 The core of RothsteinJ.s proposal resides in the admission that decades of administrative law jurisprudence
have left him unable to provide a definition of what might constitute a true question of
jurisdiction.147 Preserving a concept that is theoretically compelling, but practically unworkable and even superfluous seems only to invite the type of arcane and indeterminate
legal wrangling that Dunsmuir sought to avoid. Because the Supreme Court already cast the
cloak of constitutional protection over judicial review (thereby foreclosing any literal application of a privative clause), and Dunsmuir identified other criteria for applying the correctness standard, extinguishing the category of jurisdictional question jeopardizes neither
the resilience of judicial review nor correctness scrutiny. In a technical sense, the majority
leaves the issue unresolved, concluding instead that jurisdictional questions are exceptional
and none have come before it since Dunsmuir.
A sympathetic reading of the majority judgment might proffer the hypothesis that the
other post-Dunsmuir grounds for correctness review really amount to exemplars of situations typically regarded as jurisdictional in pre-Dunsmuir case law. In a jurisprudence
chiefly notable for its lack of predictability, the correctness standard was most consistently
applied to issues of procedural fairness, constitutionality, the jurisdictional lines between
competing specialized tribunals, and to questions of law elevated to central importance to
the legal system as a whole and outside the adjudicators expertise.148 One could argue
that the work done by jurisdiction pre-Dunsmuir is performed post-Dunsmuir by these
exceptions to the default presumption of Dunsmuir reasonableness, thereby rendering
jurisdiction itself otiose.
Cromwell J. emphatically disagrees with the majority on the fate of jurisdiction, warning
that the position espoused by Rothstein J. threatens to undermine the foundation of judicial review of administrative action.149 As Rothstein J. notes, however, Cromwell J.s objec-

144 2011 SCC 61, [2011] 3 S.C.R. 654 [Alberta (Information and Privacy Commissioner)].
145 Ibid. at para. 10.
146 Ibid. at para. 34.
147 Ibid. at para. 42.
148 Ibid. at para. 30.
149 Ibid. at para. 92.

VI. Dunsmuir: And Then There Were Two

309

tion fails to take into account the bases for application of a correctness standard apart from
the jurisdictional question.
Cromwell J.s version of the standard-of-review analysis tempers the inclination toward
reasonableness with a more thorough examination of legislative intent when a plausible
argument is advanced that a particular provision falls outside the presumption of reasonableness review and into the exceptional category of correctness review.150 Cromwell J.
does not actually conduct a thorough examination of legislative intent in the case at bar, or
indicate what a plausible argument should contain, confining himself to the conclusory
statement that the legislature did not intend a correctness standard to apply because the
power to extend time is granted in broad terms in the context of a detailed and highly specialized statutory scheme which it is the Commissioners duty to administer and under
which he is required to exercise many broadly granted discretions.151 This quick concession
seems curious, because CromwellJ. could have identified the fact that the interpretation of
a statutory provision about timelines essentially concerns the process of investigation and
adjudication, and the Court has consistently applied a correctness standard to matters of
procedural fairness.
Binnie J. (Deschamps J. concurring) stakes out a conciliatory middle position between
Rothstein J. and Cromwell J. He agrees with the latter that the concept of jurisdiction is
fundamental, but endorses Rothstein Js initiative to euthanize the issue on account of its
practical disutility.152 Binnie J.s middle ground consists of two propositions. The first is a
reiteration of his prediction that reasonableness will entail a spectrum of intensity of scrutiny, with the implication that the application of a reasonableness review may, in appropriate
cases, look very similar to correctness review. The second is a revision of the question of
central importance to the legal system as a whole exception to deference. Here, Binnie J.
offers a broader and more generic exception for questions of law that raise matters of legal
importance beyond administrative aspects of the statutory scheme under review and do
not lie within the core function and expertise of the decision maker.153 If adopted, BinnieJ.s reformulation would appear to enlarge this exception.
It remains unclear whether the jurisdictional question will die a peaceful death or simply
lay dormant until the Supreme Court decides to resurrect it. The boundary metaphor that
underwrites jurisdiction is at once irresistible and impracticable. Perhaps it is no coincidence that the vocabulary of jurisdiction feels most natural when invoked in respect of entities that also happen to be geographically bounded, such as municipalities, provinces, and
states. This makes all the more notable the 2012 judgment in Catalyst Paper Corp. v. North
Cowichan (District),154 which concerned a municipal tax bylaw. The Supreme Court resolutely avoided the term jurisdiction, or vires, and consistently spoke of deference and
reasonableness.

150 Ibid. at para. 99.

151 Ibid. at para. 101.
152 Ibid. at para. 88.
153 Ibid. at para. 89.
154 2012 SCC 2 [Catalyst].

310

Chapter 9 Standard of Review: Back to the Future?

It is perhaps a fitting postscript to the narrative of jurisdiction in Canadian administrative law that even its contemporary defender, Cromwell J., took the opportunity to overrule
Bell v. Ontario Human Rights Commission.155 In Halifax (Regional Municipality) v. Nova
Scotia (Human Rights Commission),156 the Court considered an appeal from an order of
prohibition by a Nova Scotia chambers judge that had the effect of invalidating referral of a
human rights complaint to a board of inquiry, thereby preventing the board of inquiry from
considering whether the complaint fell within the purview of the Nova Scotia human rights
statute. A unanimous Supreme Court rejected Bells notion of preliminary question as
both anachronistic and insufficiently respectful of the considered opinion of the tribunal
on legal questions, whether the tribunals ruling is ultimately reviewable in the courts for
correctness or reasonableness.157 The Court also ruled that the appropriate standard of review in the case at bar was reasonableness.

C. Whatever Happened to Patent Unreasonableness?

Pity the legislators: long ago, they gave up trying to refine the privative clause in order to
persuade judges that when they told the courts to get out and stay out, they really meant
it. If the privative clause was an exercise in communicating legislative intent about the role
of the courts, suffice to say that the message was, if not lost, then reformulated in translation. The Court rationalized this through the constitutionalization of judicial review and a
negotiation between the rule of law and an idea of parliamentary sovereignty envisioned as
external to, or counterpoised against, the rule of law.
Sometime after Southam, a few legislators switched tactics and decided instead to direct
the courts on the legislatures intended standard of review by explicitly stating whether a
reviewing court should apply correctness, unreasonableness simpliciter, or patent unreasonableness. For example, the B.C. Administrative Tribunals Act158 (ATA) itemizes grounds of
judicial review applicable to the tribunals subject to the ATA and matches each ground with
a standard of review. In the case of judicial review of discretion, the ATA also lists factors
relevant to determining whether discretion was exercised in a patently unreasonable way.159
In Ontario, the Human Rights Code contains a privative clause stipulating that a decision
of the [Human Rights] Tribunal is final and not subject to appeal and shall not be altered or

155 Supra note 7.

156 2012 SCC 10.
157 Ibid. at para. 37.
158 S.B.C. 2004, c. 45.
159 If a tribunals enabling statute has a privative clause, the standard of review is patent unreasonableness for
questions of law and fact or the exercise of discretion for all matters over which the tribunal has exclusive
jurisdiction (ATA, s. 58(1)). Common-law rules of natural justice and procedural fairness must be decided
having regard to whether, in all of the circumstances, the tribunal acted fairly (ATA, s. 58(2)(b)), and a
standard of correctness applies to all other matters. Where the enabling statute has no privative clause, findings of fact are reviewable on the basis of no evidence or unreasonableness, questions of law are reviewable on
correctness, the exercise of discretion is subject to a standard of patent unreasonableness, and procedural fairness is decided having regard to whether, in all of the circumstances, the tribunal acted fairly (ATA, s.59(5)).

VI. Dunsmuir: And Then There Were Two

311

set aside in an application for judicial review or in any other proceeding unless the decision
is patently unreasonable.160
And then Dunsmuir came along, and out went patent unreasonableness.
The Supreme Court in Khosa acknowledges the predicament for parties dealing with statutes that incorporate the now obsolete common-law standard of patent unreasonableness:
Generally speaking, most if not all judicial review statutes are drafted against the background
of the common law of judicial review. Even the more comprehensive among them, such as the
British Columbia Administrative Tribunals Act,[161] can only sensibly be interpreted in the common law context because, for example, it provides in s. 58(2)(a) that a finding of fact or law or
an exercise of discretion by the tribunal in respect of a matter over which it has exclusive jurisdiction under a privative clause must not be interfered with unless it is patently unreasonable.
The expression patently unreasonable did not spring unassisted from the mind of the legislator. It was obviously intended to be understood in the context of the common law jurisprudence, although a number of indicia of patent unreasonableness are given in s. 58(3). Despite
Dunsmuir, patent unreasonableness will live on in British Columbia, but the content of the
expression, and the precise degree of deference it commands in the diverse circumstances of a
large provincial administration, will necessarily continue to be calibrated according to general
principles of administrative law. That said, of course, the legislature in s. 58 was and is directing
the B.C. courts to afford administrators a high degree of deference on issues of fact, and effect
must be given to this clearly expressed legislative intention.162

How have courts interpreted statutory standards of patent unreasonableness post-

Dunsmuir?
In British Columbia (Workers Compensation Board) v. Figliola,163 the Supreme Court of
Canada considered whether the B.C. Human Rights Tribunal exercised its discretion in a
patently unreasonable fashion when it decided to adjudicate a Human Rights Code complaint that had already been rejected by a review officer of the B.C. Workers Compensation
Board. The Court was not required to interpret the term patently unreasonable post-
Dunsmuir, because the ATA already defined it according to the traditional indicia of abuse
of discretion.164 In Shaw v. Phipps,165 the Ontario Divisional Court held that patently unreasonable in the Ontario Human Rights Code should be interpreted against the legislative
intent at the time of enactment. The Divisional Court reasoned that the Ontario Legislatures intent was to confer the highest level of deference available under general principles

160 Human Rights Code, supra note 8, s. 45.8.

161 Supra note 159.
162 Khosa, supra note 133 at para. 19.
163 2011 SCC 52, [2011] 3 S.C.R. 422 [Figliola].
164 ATA, supra note 158, s. 59(4) defines a patently unreasonable exercise of discretion as one where the discretion
(a) is exercised arbitrarily or in bad faith,
(b) is exercised for an improper purpose,
(c) is based entirely or predominantly on irrelevant factors, or
(d) fails to take statutory requirements into account.
165 2010 ONSC 3884.

312

Chapter 9 Standard of Review: Back to the Future?

of administrative law on the Human Rights Tribunal. In 2006, that standard was patent
unreasonableness. The Supreme Court of Canada subsequently declared the highest level of
deference available under general principles of administrative law to be reasonableness.
Therefore, according the highest degree of deference to the tribunals determination of liability and remedy post-Dunsmuir meant respecting those questions within the specialized
expertise of the Tribunal unless they are not rationally supportedin other words, they
are unreasonable.166

D. What Is a Question of Central Importance to the Legal System as a

Whole (and Outside the Decision-Makers Area of Expertise)?
The standard of review approach articulated in Dunsmuir does not place expertise in the
foreground as a discrete focus of the evaluation. Given the primacy of expertise in earlier
jurisprudence, this seems odd. Perhaps the better reading of Dunsmuir is that the unmistakable tilt of the standard-of-review analysis toward deference presupposes the superior expertise of administrative decision-makers in interpreting and applying their home statutes,
absent some indicator to the contrary.167 In a sense, Dunsmuir deems administrative
decision-makers as expert in applying their constitutive statutes, unless one of the exceptions applies. This posture contrasts with the view expressed by U.S. appellate judge (and
scholar) Richard Posner, whose critique of poor-quality decision making by U.S. immigration
adjudicators commenced with the declaration, Deference is earned; it is not a birthright.168
Questions of jurisdiction and constitutionality might be understood as illustrations of matters that lie outside the expertise of administrative decision-makers. Questions of central
importance to the legal system as a whole are assigned to correctness review only if they
are also outside the specialized area of expertise of the administrative decision maker.169
The Dunsmuir majority is also concerned about precedent: Because of their impact on the
administration of justice as a whole, such questions require uniform and consistent answers.170 Issues concerning the jurisdictional boundaries between different administrative
tribunals seem more ambiguously tied to expertise, but also link back to the allocation of
questions of true jurisdiction to the correctness standard.
In Celgene Corp. v. Canada (Attorney General),171 and again in Smith v. Alliance Pipeline,172 the majority affirmed that interpretation of the home statute will usually attract a
reasonableness standard of review.173 One expects that most administrative decisions will
166 Ibid. at para. 42.
167 Rothstein J. strenuously objected to this presumption in Khosa, supra note 133, and reiterated precisely the
opposite view: If the legislature intended to protect expert decision-makers from review, it did so through a
privative clause (at para. 86).
168 Kadia v. Gonzales, 501 F.3d 817 at 821 (7th Cir. 2007).
169 Dunsmuir, supra note 112 at para. 60.
170 Ibid.
171 2011 SCC 1, [2011] 1 SCR 3 [Celgene v. Canada].
172 Smith, supra note 140.
173 Ibid. at para. 28.

VI. Dunsmuir: And Then There Were Two

313

satisfy this criterion most of the time, thereby assuring an expansive scope for the default
position of deference.
In Smith, the majority went on to state that the nature of the task (discretionary award of
costs), and the intertwining of fact and law, buttressed its conclusion that reasonableness is
the operative standard, and no countervailing correctness factors were present.
In his Dunsmuir concurrence, Binnie J. expressed concern that the exceptional justification for intensified judicial scrutiny where an issue is of central importance and beyond the
decision-makers expertise would unleash needless and distracting debate in the lower
courts. Thus far, few cases in which the parties have disputed whether a given legal question
fits within the exemption have percolated upward. One might find clues as to its meaning in
the pre-Dunsmuir case of Toronto (City) v. C.U.P.E., Local 79.174 The Supreme Court considered the standard of review applicable to the relitigation of a criminal conviction in the
course of a grievance arbitration. LeBel J. concurred with the majoritys assessment that the
question concerned common-law doctrines that went to the administration of justice. He
agreed that the appropriate standard of review was correctness, because the issue concerned
a question of law that is both of central importance to the legal system as a whole and outside the adjudicators specialized area of expertise.175
In Mowat v. Canada,176 the Supreme Court considered the interpretation of s. 53(2)(c) of
the Canadian Human Rights Act.177 The provision authorizes a human rights tribunal to
order the offending party to compensate the victim 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. The question was whether any expenses included the complainants legal fees.178
Although the Dunsmuir methodology presumes that the outcome of the pre-Dunsmuir
pragmatic and functional analysis will align with the outcome of the post-Dunsmuir
standard-of-review analysis, the situation of human rights tribunals gives the Supreme
Court pause. As noted earlier, prior to Dunsmuir, human rights tribunals attracted little
deference from the courts. The Supreme Court of Canada repeatedly insisted that the superior expertise of human rights tribunals relates to fact-finding and adjudication in a
human rights context and does not extend to general questions of law.179 Interpretation of
the home statute by human rights tribunals attracted no deference. So, if one relied on past
jurisprudence about human rights tribunals expertise, one would probably incline toward
the standard of correctness. Indeed, if one focused more narrowly on the power of the tribunal to award costs, one might even characterize the issue as jurisdictional and thus
subject to correctness. Yet, if one took seriously Dunsmuirs recognition of the expertise of

174 Supra note 110.

175 Ibid. at para. 62.
176 Canada (Canadian Human Rights Commission) v. Canada (Attorney General), 2011 SCC 53, [2011] 3 S.C.R.
471, Canada (Attorney General) v. Mowat, 2009 FCA 309, [2010] 4 F.C.R. 579 [Mowat].
177 R.S.C. 1985, c. H-6.
178 The legal issue in Mowat resembled the issue in Smith v. Alliance Pipeline, supra note 140namely, the power
to award costs. Although the Supreme Court adopted a standard of review of reasonableness in both cases,
query whether the Court deferred equally to the respective administrative tribunals.
179 Mowat, supra note 176 at para. 29.

314

Chapter 9 Standard of Review: Back to the Future?

decision-makers in interpreting and applying their home statute, one would be pulled in the
direction of reasonableness.
The Supreme Court acknowledges the dilemma, and resolves it by casting the issue of
costs as neither jurisdictional nor a question of central importance to the legal system and
outside the tribunals expertise. The Court carefully steered a path that would enable it to
arrive at a reasonableness standard of review with respect to costs without requiring it to
disavow the pre-Dunsmuir jurisprudence that subjected human rights tribunals to correctness review on almost all other matters:
There is no doubt that the human rights tribunals are often called upon to address issues of very
broad import. But, the same questions may arise before other adjudicative bodies, particularly
the courts. In respect of some of these questions, the application of the Dunsmuir standard of
review analysis could well lead to the application of the standard of correctness. But, not all
questions of general law entrusted to the Tribunal rise to the level of issues of central importance to the legal system or fall outside the adjudicators specialized area of expertise. In this
case, there is no doubt that the Tribunal has the power to award compensation for any expenses incurred by the victim as a result of the discriminatory practice . The issue is
whether the Tribunal could order the payment of costs as a form of compensation.. Although the respondent submitted that a human rights tribunal has no particular expertise in
costs, care should be taken not to return to the formalism of the earlier decisions that attributed
a jurisdiction-limiting label, such as statutory interpretation or human rights, to what is in
reality a function assigned and properly exercised under the enabling legislation by a tribunal.
The inquiry of what costs were incurred by the complainant as a result of a discriminatory
practice is inextricably intertwined with the Tribunals mandate and expertise to make factual
findings relating to discrimination. As an administrative body that makes such factual findings
on a routine basis, the Tribunal is well positioned to consider questions relating to appropriate
compensation under s. 53(2). In addition, a decision as to whether a particular tribunal will
grant a particular type of compensationin this case, legal costscan hardly be said to be a
question of central importance for the Canadian legal system and outside the specialized expertise of the adjudicator.180

The Courts standard-of-review analysis in Mowat remains silent on a problematic issue

that preoccupied the Federal Court of Appeal when it decided the case.181 In Domtar,182 the
Supreme Court ruled that inconsistent interpretations of a given statutory provision did not
supply an independent ground for stricter judicial scrutiny where the appropriate stance
was otherwise deferential. Although Domtar concerned the interpretation of a common
phrase by two bodies constituted under different statutes, the reach of the Supreme Courts
dictum was limited only by the possibility of direct operational conflict.183 In Mowat, the
180 Ibid. at paras. 23-26.
181 Mowat (FCA), supra note 176.
182 Supra note 30.
183 In Domtar, ibid., the Court had rejected the assertion that the precedential value of resolving inconsistency
within or between tribunals constitutes an independent basis for adopting a correctness standard of review,
unless the divergent decisions create an operational conflict, whereby compliance with one order would necessitate violation of the other. A significant concern for the Court in Domtar was the risk that real or appar-

VI. Dunsmuir: And Then There Were Two

315

Federal Court of Appeal was presented with conflicting interpretations of the same statutory provision by different panels of the same tribunal and, subsequently, by different Federal Court judges on judicial review. The Federal Court of Appeal described the problem as
follows:
The question has not been answered consistently by the Tribunal and is the subject of diverse
opinions in the Federal Court. It comes before the Court for the first time. It is difficult, if not
impossible, to conclude that the answer (either yes or no) can be said to fall within a range of
possible acceptable outcomes. There is much to be said for the argument that where there are
two conflicting lines of authority interpreting the same statutory provision, even if each on its
own could be found to be reasonable, it would not be reasonable for a court to uphold both.184

The Federal Court of Appeal adopted the comments of Feldman J.A. in Abdoulrab v.
Ontario (Labour Relations Board),185 and declared that it accords with the rule of law that
a public statute that applies equally to all affected citizens should have a universally accepted
interpretation.186 The values of certainty and consistency for the affected parties and the
public at large led the Court of Appeal to characterize the question whether a human rights
tribunal can order the losing party to pay the legal costs of the complainant as a general
question of law of central importance to the legal system as a whole and one that is outside
the specialized area of the Tribunals expertise.187 It set aside the Human Rights Tribunals
affirmative response to the question as incorrect.
The Supreme Court of Canada in Mowat suppresses the issue of conflicting decisions and
does not advert to it; one might interpret the silence as affirmation of Domtar. Quaere
whether it should matter for purposes of designating a question as one of central importance that the legal question has already been the subject of conflicting decisions by the
tribunal and by reviewing courts. If you were a member of that tribunal or representing a
party before it, would you prefer that it be left to your peers to address divergent interpretations through institutional consistency mechanisms, or would you rather that the courts
resolve the matter definitively by applying a correctness standard? Would it be legitimate to
pre-empt future conflict by asserting a standard of correctness the first time the interpretation of a legal provision is contested? If not, when does it become appropriate to do so?
In the result, the Supreme Court in Mowat decided that the Human Rights Tribunals
inclusion of legal costs as expenses was unreasonable. Of course, the Supreme Court of
Canadas decision had the convenient effect of ruling out one of only two possible interpretations of the statutory provision. The Supreme Court thus provided definitive guidance to
subsequent decision-makers.

ent inconsistencies within or between tribunal decisions might become a pretext for undermining fidelity to
the principles underlying curial deference. The Court also adverted to internal mechanisms available to tribunals to encourage consistency, and downplayed the virtues of consistency in relation to other important
values served by deference.
184 Mowat (FCA), supra note 176 at para. 45.
185 2009 ONCA 491, 95 O.R. (3d) 641, 95 Admin. L.R. (4th) 121.
186 Mowat (FCA), supra, note 176 at para. 46.
187 Ibid. at para. 47.

316

Chapter 9 Standard of Review: Back to the Future?

VII. Spin-Offs
A. The Charter, Discretion, and the Standard of Review
Where a judicial review application raises several discrete issues, reviewing courts have
sometimes calibrated the standard of review separately for each issue. Segmentation arises
whenever one link in a decision chain attracts a different standard of review than other links
in the chain. Dunsmuir offers relief from the complexity of this process by expanding the
range of decisional steps to which deference will presumptively apply, but the problem remains where one or more elements of the decision attracts a standard of review of correctness.
In Dunsmuir, Binnie J. described segmentation in the following terms:
Mention should be made of a further feature that also reflects the complexity of the subject
matter of judicial review. An applicant may advance several grounds for quashing an administrative decision. He or she may contend that the decision maker has misinterpreted the general
law. He or she may argue, in the alternative, that even if the decision maker got the general law
straight (an issue on which the courts view of what is correct will prevail), the decision maker
did not properly apply it to the facts (an issue on which the decision maker is entitled to deference). In a challenge under the Canadian Charter of Rights and Freedoms to a surrender for
extradition, for example, the minister will have to comply with the Courts view of Charter
principles (the correctness standard), but if he or she correctly appreciates the applicable law,
the court will properly recognize a wide discretion in the application of those principles to the
particular facts. The same approach is taken to less exalted decision makers. In the jargon of the
judicial review bar, this is known as segmentation.188

Unfortunately, neither Binnie J. nor his colleagues say anything further in Dunsmuir (or in
subsequent cases) about the dilemmas posed by segmentation, or how to resolve them.
The intersection of discretion and the Charter provides an object lesson in some of the
complexities of segmentation. Consider Suresh.189 A provision of the Immigration Act (now
IRPA) grants the minister discretion to deport a non-citizen who is deemed to be a threat
to national security. The s. 7 constitutional issue was whether the minister could exercise his
or her discretion to deport a non-citizen to a country where that person faced a substantial
risk of torture. A deportation decision in this context consisted of various subquestions:

1. What is the meaning of national security?

2. Is the non-citizen a threat to national security?
3. What does torture mean?
4. Does the non-citizen face a substantial risk of torture?
5. Does deportation of a non-citizen to torture violate s. 7 of the Charter?

The Court does not articulate a standard of review for questions 1 and 3, but emphasizes
that questions 2 and 4 attract deference, while question 5 is explicitly subject to correctness.

188 Dunsmuir, supra note 112 at para. 142.

189 Suresh v. Canada (Minister of Citizenship and Immigration), 2002 SCC 1, [2002] 1 S.C.R. 3.

VII. Spin-Offs

317

Had the Court said yes to question 5, then the determinations under questions 1 to 4
could have proceeded sequentially and discretely from one another. If questions 4 was
reached and the answer was yes, that would be the end of the story: no deportation. If the
answer was no, then the minister could deport to torture.
But the Courts answer to question 5 was neither yes nor no. It was no, unless there
are exceptional circumstances. The Court effectively constrained but did not eliminate the
discretion to deport to torture. The so-called Suresh exception directed the minister to consider the following additional question:
6. Do the benefits to Canadas national security of deporting the non-citizen outweigh
the harm of deporting the non-citizen to torture?
What standard of review should apply at this stage? The Court is silent on this matter. On
the one hand, the decision not to prohibit deportation to torture absolutely amounts to a
decision that a decision-maker possesses discretion to deport to torture in some imaginable
case. The exercise of balancing multiple factors usually attracts deference because of its indeterminacy, the interplay of factors, and the courts stated reluctance to reweigh evidence
on review. On the other hand, the Court itself invented the Suresh exception, and each balancing exercise under the rubric of exceptional circumstances risks the most extreme violation of a specific individuals Charter rights if the ministers calculus is mistaken. It seems
inimical to the normative and institutional foundation of the Charter to defer to a ministers
own determination of whether his or her discretionary decision violates an individuals
Charter rights, but there is precedent indicating judicial willingness to do precisely that.190
In Dor v. Barreau du Qubec,191 the Supreme Court offered a path out of this methodological dilemma: When Charter values are applied to an individual administrative decision,
they are being applied in relation to a particular set of facts. Dunsmuir tells us this should
attract deference.192 The courts should recognize the distinct advantage that administrative
bodies have in applying the Charter to a specific set of facts and in the context of their enabling legislation.193 Rather than attempt to retrofit the s. 1 Oakes test into the exercise of a

190 In practice, courts called on to review a ministerial decision to deport to torture have applied a standard of
reasonableness but found the decision unreasonable. See e.g. Jaballah (Re) v. Jaballah (Re), 2005 FC 399,
[2005] 4 F.C.R. 359. See also Lake v. Canada (Minister of Justice), 2008 SCC 23, [2008] 1 S.C.R. 761 [Lake], in
which the Supreme Court deferred to the minister of justices determination about whether surrender of a
fugitive for extradition would violate the fugitives ss. 6 and 7 rights under the Charter on the grounds that
the decision in question was largely a political decision, not a legal one and a fact driven inquiry (paras.
37 and 38). This is a deeply problematic judgment that has the potential to insulate rights violating state conduct from the correctness standard, where the breach arises from the exercise of a broad statutory discretion
that does not, on its face, violate the Charter. Many, if not most, individualized exercises of discretion can
easily be characterized as fact-driven. The claim that a decision is political rather than legal is conclusory
and unhelpful, insofar as many political decisions have a legal dimension. Moreover, it is precisely because
the violation of individual rights may be politically expedient that the Charter places legal limits on the exercise of political power.
191 2012 SCC 12.
192 Ibid. at para. 36.
193 Ibid. at para. 48.

318

Chapter 9 Standard of Review: Back to the Future?

case-specific discretion, the Court proposes a similar proportionality analysis that balances
the severity of the interference of the Charter protection with the statutory objectives.194
If the outcome of that balancing falls within a range of possible, acceptable outcomes, then
it merits deference. The concluding declaration of the Court is that [i]f, in exercising its
statutory discretion, the decision-maker has properly balanced the relevant Charter values
with the statutory objectives, the decision will be found to be reasonable.195
What this will mean in practice, of course, remains to be seen. How wide a margin of
appreciation will the courts grant administrators in properly balancing the Charter
against other objectives? Though the Court has repeatedly claimed to eschew re-weighing
of factors in deferential review of discretion, it is difficult to conceive of a proportionality
analysis that does not inquire into the appropriate weighting of the Charter right against
other interests.
We live in an era where most governments take advice from government lawyers in
drafting legislation in order to avoid flagrant unconstitutionality. It is also the case that
many contemporary statutes look increasingly skeletal. What goes on the bones of the
statute is fleshed out through regulatory authority delegated to the governor in council or
expansive and vague grants of statutory discretion to administrative decision-makers (including ministers). If Charter issues are increasingly likely to emerge in the exercise of
broad discretion rather than in the text of a law, the scope and intensity of judicial oversight
of Charter-impacting discretion will have profound implications for the level of rights protection within the Canadian legal order.

B. Reasoning About Reasons

Reasons straddle procedure and substance. As Grant Huscroft explains in Chapter 5, From
Natural Justice to Fairness: Thresholds, Content, and the Role of Judicial Review, the duty
to give reasons is a component of fairness. Reasons serve a number of purposes, not least of
which is to communicate that the decision-maker has genuinely heard and considered the
evidence and arguments presented. As the Supreme Court remarked in Lake v. Canada
(Minister of Justice), The purpose of providing reasons is twofold: to allow the individual to
understand why the decision was made; and to allow the reviewing court to assess the validity of the decision.196 One might contend that reasons that do not adequately communicate
the basis of the decision do not meet some minimal threshold that qualifies the text as reasons. They fail as a matter of form.
Reasons also disclose the findings of fact, interpretations of law, applications of law to
fact, and exercises of discretion that are the substance of the decision. Reasons contain the
evidence of the reasonableness (or correctness, as the case may be) of those exercises of
statutory authority. As you will see in Sheila Wildemans Chapter 10, Pas de Deux: Deference and Non-Deference in Action, measuring the substantive reasonableness of a decision

194 Ibid. at para. 56.

195 Ibid. at para. 58.
196 Lake, supra note 190 at para. 46.

VIII. Review of Standard of Review

319

post-Dunsmuir includes assessing the justification, transparency, and intelligibility of the

reasoning process.
There is obvious potential for overlap between assessing the formal adequacy of reasons
as a matter of procedural fairness and evaluating the substantive content of reasons as a
matter of merits review. Framing the ground-of-reasons review in terms of procedure rather
than substance potentially invites a greater degree of judicial intervention via the correctness standard. This also depends on the ambit-of-adequacy review. The more a court demands of reasons in order to satisfy the procedural duty of fairness, the greater the potential
for expansive and intrusive judicial review.
Finding a consistent break point between the form of reasons and the content of reasons proved challenging for lower courts, and different jurisprudential currents began to
emerge.197 Without adverting to this incipient debate, the Supreme Court of Canada abruptly terminated it.
In Newfoundland and Labrador Nurses Union v. Newfoundland and Labrador (Treasury
Board),198 Abella J., writing for the Court, stated: I do not see Dunsmuir as standing for the
proposition that the adequacy of reasons is a stand-alone basis for quashing a decision, or
as advocating that a reviewing court undertake two discrete analysesone for the reasons
and a separate one for the result.199 Later, she explicitly minimized the scope of the duty to
give reasons to bare form.
It strikes me as an unhelpful elaboration on Baker to suggest that alleged deficiencies or flaws
in the reasons fall under the category of a breach of the duty of procedural fairness and that
they are subject to a correctness review. As Professor Philip Bryden has warned, courts must
be careful not to confuse a finding that a tribunals reasoning process is inadequately revealed
with disagreement over the conclusions reached by the tribunal on the evidence before it.
It is true that the breach of a duty of procedural fairness is an error in law. Where there are no
reasons in circumstances where they are required, there is nothing to review. But where, as
here, there are reasons, there is no such breach. Any challenge to the reasoning/result of the
decision should therefore be made within the reasonableness analysis.200

VIII. Review of Standard of Review: I Laughed, I Cried,

IStood on My Chair and
The tensions lying at the heart of jurisprudence about the standard of review have not
changed and will not go away. In its recent jurisprudence, the majority of the Supreme
Court of Canada has staked out a position that, in principle, inclines toward deference. It
has told and retold the story about why and when courts ought to defer to the decisions of

197 See e.g. Clifford v. Ontario Municipal Employees Retirement System, 2009 ONCA 670 (2009), 93 Admin. L.R.
(4th) 131; Vancouver International Airport Authority v. Public Service Alliance of Canada, 2010 FCA 158,
[2011] 4 F.C.R. 425.
198 2011 SCC 62, [2011] 3 S.C.R. 708.
199 Ibid. at para. 14.
200 Ibid. at paras. 21-22.

320

Chapter 9 Standard of Review: Back to the Future?

administrative decision-makers. Each major iteration reveals shifts in emphasis; additional

nuances; glosses on past recitations; and attempts to reconcile, distinguish, or conceal apparent anomalies. On rare occasions, we get a new plot twist: from two standards of review
to three, then back to two (but not the same two); from the formalism of preliminary or
collateral question, to multifactor balancing, to a defeasible rule (or maybe not).
The job of discerning the appropriate standard of review became simpler after Dunsmuir,
and for this students and practitioners of administrative law should feel relieved. But they
should also attend to the prediction of Binnie J. in Dunsmuir. By streamlining the standardof-review analysis and winnowing deference down to a single standard of reasonableness,
the Court has not resolved the challenge of operationalizing deference in all its multifarious
contexts. Rather, it has exported the task to the next stage of judicial reviewnamely, the
application of correctness or reasonableness review to actual decisions. Thus, the conclusion
of this chapter is really a prologue to Chapter 10 where, post-Dunsmuir, the action unfolds.
SUGGESTED ADDITIONAL READINGS
BOOKS AND ARTICLES

Canadian
Heckman, G., Substantive Review in Appellate Courts Since Dunsmuir (2009) 47 Osgoode Hall L.J. 751.
Jacobs, L., Developments in Administrative Law: The 2007-2008 TermThe Impact of
Dunsmuir (2008) 43 Sup. Ct. L. Rev. (2d) 1-34.
Lahey, B., & D. Ginn, After the Revolution: Being Pragmatic and Functional in Canadas
Trial Courts and Courts of Appeal (2002) 25 Dal. L.J. 259.
MacLauchlan, H.W., Transforming Administrative Law: The Didactic Role of the Supreme
Court of Canada (2001) 80 Can. Bar Rev. 281.
McLachlin, The Honourable Justice B., The Roles of Administrative Tribunals and Courts
in Maintaining the Rule of Law (1999) 12 Can. J. Admin. L. & Prac. 171.
Mullan, D., Dunsmuir v. New Brunswick, Standard of Review and Procedural Fairness:
Lets Try Again! (2008) 21 Can. J. Admin. L. & Prac. 117.
Sossin, L., & C. Flood, The Contextual Turn: Iacobuccis Legacy and the Standard of Review in Administrative Law (2007) 57 U.T.L.J. 581.
I nter national
Dyzenhaus, D., The Politics of Deference: Judicial Review and Democracy in
M. Taggart, ed., The Province of Administrative Law (Oxford: Hart Publishing, 1997).