336

Chapter 10  Pas de Deux: Deference and Non-Deference in Action

otherwise defined in the Canadian Human Rights Act) could not be interpreted to extend
protection to a same-sex couple. The majority, as well as two judges dissenting on the result,
adopted a correctness standard of review.54 (Here we encounter the not-uncommon phe-
nomenon of judges agreeing on a correctness standard and then disagreeing on what the
“right answer” is upon applying that standard.) L’Heureux-Dubé J. wrote a dissenting judg-
ment adopting a patent unreasonableness standard.
The concurring majority judgments in Mossop express two variants of what I have de-
scribed as a positivist approach to statutory interpretation. Lamer C.J. (Sopinka and Iaco-
bucci JJ. concurring) grounds his analysis in legislative intent. This analysis is framed by a
gesture toward the parallel universe of normative jurisprudence available only to claimants
invoking the Charter.55 That is, given that Mr. Mossop had opted to base his claim exclu-
sively in arguments from statutory interpretation, Lamer C.J. suggests that the Court is
bound by contextual indicia that “family status” was not intended to encompass same-sex
relationships. Here he emphasizes the absence of “sexual orientation” from the statute’s pro-
hibited grounds of discrimination at the time the proceedings arose, even in the face of a
recommendation by the Canadian Human Rights Commission that it be added. The failure
of Parliament to act on the commission’s recommendation, he reasons, amounted to its
“refusal” to do so. Lamer C.J. states the principle driving his analysis as follows:
Absent a Charter challenge of its constitutionality, when Parliamentary intent is clear, courts
and administrative tribunals are not empowered to do anything else but to apply the law. If
there is some ambiguity as to its meaning or scope, then the courts should, using the usual rules
of interpretation, seek out the purpose of the legislation and if more than one reasonable inter-
pretation consistent with that purpose is available, that which is more in conformity with the
Charter should prevail.56

In this case, however, Lamer C.J. determined that the legislative intent was clear, so that no
recourse to the Charter as an interpretive aid could be made.
In his concurring judgment, La Forest J. (writing for himself and Iacobucci J.)57 also fo-
cuses on legislative intent. However, his judgment is more insistently fixed on the statutory
text and, specifically, the word “family.” He describes his approach as consistent with “the
ordinary rules of statutory interpretation,” which demand that one give “to the words used in
a statute their usual and ordinary sense having regard to their context and to the purpose of
the statute.”58 In applying this principle, however, La Forest J. emphasizes not the clear pur-
pose of the Canadian Human Rights Act—that of eradicating discrimination—but the “usual
and ordinary sense” of the word “family.” His conclusion is that the “dominant” and thus

54 Audrey Macklin discusses the rationale behind selection of a correctness standard in Mossop in Chapter 9,
section V, Pragmatic and Functional Redux: Pushpanathan v. Canada (under heading V.B, Expertise; see also
heading V.D, The Nature of the Problem).
55 Canadian Charter of Rights and Freedoms, part I of the Constitution Act, 1982, being Schedule B to the Can-
ada Act 1982 (U.K.), 1982, c. 11 [the Charter].
56 Mossop, supra note 46 at 581-82.
57 The fact that Iacobucci J. concurs with the opinions of both Lamer J. and La Forest J. on the correct approach
to statutory interpretation alerts us to the compatibility of these approaches.
58 Supra note 46 at 585.
III.  The Standards of Review in Theory and Practice 337

ordinary sense of the word, or that which represents the consensus of the Canadian public,
is the “traditional family.” This is the meaning that therefore must be understood to have been
Parliament’s intent both when the statute was enacted and at the time the dispute arose.
L’Heureux-Dubé J. is alone in arguing that a patent unreasonableness standard (at the
time, the only alternative to correctness review) should apply. However, her disposition of
the matter is endorsed in the separate dissenting reasons of McLachlin and Cory JJ., who
determine that a correctness standard is in order. This seems to indicate that for McLachlin
and Cory JJ., the judgment of L’Heureux-Dubé J. represents the best account of the “right
answer” to this interpretive dispute.59 Thus, while our objective in this section is to consider
correctness review, it is worth contrasting the approach taken by L’Heureux-Dubé J. to that
of the majority.
The judgment of L’Heureux-Dubé J. is not the superficial gloss endorsed by some early
on as the proper model of patent unreasonableness review.60 And yet it distinguishes itself
from the approach of the majority in two key ways. First, it takes an emphatically normative
approach to statutory interpretation. L’Heureux-Dubé J. expresses the distance between her
approach and that of the majority as follows:
Even if Parliament had in mind a specific idea of the scope of “family status,” in the absence of
a definition in the Act which embodies this scope, concepts of equality and liberty which ap-
pear in human rights documents are not bounded by the precise understanding of those who
drafted them. Human rights codes are documents that embody fundamental principles, but
which permit the understanding and application of those principles to change over time. These
codes leave ample scope for interpretation by those charged with that task.61

The point here is that, at least with human rights statutes, it is essential to orient inter-
pretive judgment with reference to human rights principles (in this case, equality or non-­
discrimination). These are asserted in this passage not as statute-specific but as “fundamental.”
As such, they may be curtailed only by the most explicit expressions of legislative intent.
Thus it seems that there is, if not a right answer in this case (and perhaps, for L’Heureux-
Dubé J., there is), then at least a right and a wrong way of going about resolving the question.
Both the reliance of La Forest J. on an “unexamined consensus” and the reliance of ­Lamer C.J.
on legislative intent as inferred from legislative history come in for criticism as insupport-
ably prioritizing the will of an ostensibly unitary majority over the human rights principles
engaged by this instance of interpretation. In contrast, L’Heureux-Dubé J.’s judgment reflects
a commitment to a model of the rule of law known as “common-law constitutionalism.”62

59 The implications of this are the subject of careful analysis in Dyzenhaus, “Constituting the Rule of Law,” supra
note 5 at 464-68.
60 See the judgment of Wilson J in National Corn Growers, supra note 13.
61 Mossop, supra note 46 at 621.
62 See the discussion of normative approaches to statutory interpretation in section II, above. For an illuminat-
ing account of historical and contemporary features of common-law constitutionalism, see M.D. Walters,
“ ‘Common Public Law in the Age of Legislation’: David Mullan and the Unwritten Constitution” in Inside
and Outside, supra note 19 at 421, and “The Common Law Constitution in Canada: Return of Lex Non
Scripta as Fundamental Law” (2001) 51 U.T.L.J. 91.
338 Chapter 10  Pas de Deux: Deference and Non-Deference in Action

Under this model, statutes are understood not as closed systems but as requiring interpret-
ation in light of the animating principles and values of the wider social and legal tradition.
The second major difference between L’Heureux-Dubé J.’s judgment and that of the ma-
jority (and here she also differs from her fellow dissenters) is, of course, her selection of a
patent unreasonableness standard. Thus, while the tradition of common-law constitutional-
ism tends to place the burden of identifying and prioritizing fundamental values on the
shoulders of judges alone, L’Heureux-Dubé J. indicates through her adoption of deference
that the tribunal has a legitimate role in this enterprise. This is confirmed in her method of
reasoning, which throughout draws on and amplifies the evidence and argument accepted
by the tribunal. That is, the decision of the tribunal is drawn on not simply as a base of fact
or policy separable from legal analysis, but as an exercise in statutory interpretation, which
is itself a model of the interdependency of facts and values in legal judgment. As such,
L’Heureux-Dubé J. demonstrates both a respect for Parliament’s will to assign the tribunal
the role of administering this statute, and a commitment to the idea (again, cast as a prin-
ciple of law) that in administering the statute, the tribunal’s task is to transcend narrow
constructions of parliamentary will.
In agreeing with L’Heureux-Dubé J.’s disposition of the case, McLachlin and Cory JJ. in-
dicate their support for the idea that the statutory text must be read in light of social context
(and thus changing social conceptions of family), and with particular sensitivity to the ways
that human rights principles inflect and are inflected by that text and context. Yet according
to L’Heureux-Dubé J., this sensitivity is best served by careful attention to the reasons of the
tribunal. In light of the fact that the other dissenters select a correctness standard, a question
that arises is whether they are to be understood to say that the “right answer” would have
been secured absent the tribunal’s reasons or, alternatively, that in this case, correctness re-
view is compatible with—even in some measure dependent on—attentiveness to tribunal
reasoning.63
Since Mossop and, moreover, the introduction of review for reasonableness (as an al-
ternative to deference attuned only to “patent” unreasonableness), the Court has indicated
an increased willingness to accept that deference to human rights tribunals on matters in-
volving interpretation of human rights statutes may, at least in some circumstances, be war-
ranted—for instance, where the matter is “fact-intensive” or goes to tribunal processes.64 At
the same time, it has shown some willingness to defer to decisions affecting human rights
even where the administrative decision-maker is not a human rights tribunal.65 This trend
receives further support with Doré,66 which, as discussed in the next section, states that the

63 See Dyzenhaus, “Constituting the Rule of Law,” supra note 5 at 467-68.

64 Pushpanathan v. Canada (Minister of Citizenship and Immigration), [1998] 1 S.C.R. 982 at para. 45 [Push-
panathan]. More recent consideration of this question is found in Canada (Canadian Human Rights Com-
mission) v. Canada (Attorney General), 2011 SCC 53, [2011] 3 S.C.R. 471 [Mowat] at paras. 19-27. See Audrey
Macklin’s discussion of the issue in Chapter 9, section V, Pragmatic and Functional Redux: Pushpanathan v.
Canada (under heading V.B, Expertise) and section VI, Dunsmuir: And Then There Were Two (under head-
ing VI.D, What Is a Question of Central Importance to the Legal System as a Whole (and Outside the
­Decision-Maker’s Area of Expertise)?).
65 For example, see Council of Canadians with Disabilities v. VIA Rail Canada Inc., 2007 SCC 15, [2007] 1 S.C.R.
650. The decision is characterized not as law-interpretive but rather as a highly discretionary application of
legal standards to the facts.
66 Supra note 4.
III.  The Standards of Review in Theory and Practice 339

exercise of adjudicative discretion involving the balancing of Charter values with other legal
values should generally attract deference.
It is not at all clear, however, that a decision like the one engaged in Pushpanathan v.
Canada (Minister of Citizenship and Immigration)67 would attract deference after Doré. In
Pushpanathan, the Court applied a correctness standard of review to a decision of the Con-
vention Refugee Determination Division of the Immigration and Refugee Board, involving
interpretation of a provision of the Immigration Act excluding from refugee status persons
who have “been guilty of acts contrary to the purposes and principles of the United Na-
tions.” The language had been incorporated from a UN convention. Here, the central matter
in dispute was constructed not as an exercise of discretion (despite the necessary operation
of discretion or of value-laden judgment in interpreting the provision), but rather as inter-
pretation of a “general legal principle”—and the board was found not to have applicable
experience or expertise. On applying a correctness standard, both the majority and dissent-
ing judgments of the Supreme Court of Canada undertook an independent analysis of the
interpretive problem. That is not to say that either took a positivistic, as opposed to norma-
tive, approach. Indeed, the selection and application of correctness review in this case is
rooted less in the thesis that the question gives rise to one right answer than the thesis that
judges hold a unique institutional capacity to adjudicate general legal principles of broad
importance, in particular those affecting fundamental human rights.68 Notably, the majority
and dissent arrived at diverging interpretations based on diverging opinions about the
proper balancing of competing values and objectives.
Correctness review has also continued to be applied in ways that suggest a more positiv-
istic understanding of the enterprise of statutory interpretation. We may ask how the
­Dunsmuir majority’s statement that a tribunal engaged in interpretation of its enabling stat-
ute will typically attract deference 69 would affect the analysis in the pre-Dunsmuir case,
Barrie Public Utilities v. Canadian Cable Television Assn.70 There, Gonthier J. for the major-
ity characterized the phrase “the supporting structure of a transmission line” as a matter of
“pure statutory interpretation” 71 attracting correctness review. He then proceeded to iden-
tify the plain meaning of this phrase. Yet the contrary interpretation (favoured by the Can-
adian Radio-television and Telecommunications Commission) arguably reflected cogent
policy goals consistent with its mandate. Bastarache J., in dissent, criticized the majority’s
approach in comments that drew on the reasons of L’Heureux-Dubé J. for a unanimous
Court in Domtar Inc. v. Quebec (Commission d’appel en matière de lésions professionnelles):72
Substituting one’s opinion for that of an administrative tribunal in order to develop one’s own
interpretation of a legislative provision eliminates its decision-making autonomy and special
expertise. Since such intervention occurs in circumstances where the legislature has determined

67 Supra note 64.

68 I discuss these judgments further in “A Fine Romance?,” supra note 4 at 231-32.
69 “Deference will usually result where a tribunal is interpreting its own statute or statutes closely connected to
its function, with which it will have particular familiarity” (Dunsmuir, supra note 1 at para. 54). See also the
majority judgment in Alberta (Information and Privacy Commissioner), supra note 2.
70 2003 SCC 28, [2003] 1 S.C.R. 476 [Barrie Utilities].
71 Ibid. at para. 16.
72 [1993] 2 S.C.R. 756 [Domtar].
340 Chapter 10  Pas de Deux: Deference and Non-Deference in Action

that the administrative tribunal is the one in the best position to rule on the disputed decision,
it risks, at the same time, thwarting the original intention of the legislature. For the purposes of
judicial review, statutory interpretation has ceased to be a necessarily “exact” science and this
Court has, again recently, confirmed the rule of curial deference set forth for the first time in
Canadian Union of Public Employees, Local 963 v. New Brunswick Liquor Corp.73

If statutory interpretation is not an “exact” science, then on what grounds can courts
continue to apply correctness review to disputes of this sort, as such giving no credence or
weight to tribunal interpretations? Again, as with Bibeault, we may ask whether or how Bar-
rie Utilities might have been differently decided after Dunsmuir.
The post-Dunsmuir judgment of the Supreme Court in Northrop Grumman Overseas
Services Corp. v. Canada (Attorney General)74 offers a final example of a dispute about statu-
tory interpretation that was resolved on a correctness standard. There the question was
whether the Canadian International Trade Tribunal (CITT) had authority to hear a com-
plaint brought by a non-Canadian corporation, Northrop Grumman Overseas. The corpor-
ation wished to argue that Public Works and Government Services Canada had not evalu-
ated its bid for a procurement contract in accordance with the Agreement on Internal Trade
(AIT). The statute governing the tribunal’s authority, the Canadian International Trade Tri-
bunal Act,75 contemplated complaints from “potential suppliers” of procurement contracts
under the AIT where certain threshold conditions were met. The question was: did those
with standing before the CITT include non-Canadian corporations like Northrop Grum-
man? The CITT departed from its own precedents to decide that it could hear the complaint.
On its analysis of the text and purposes of the CITT Act and Regulations in addition to the
AIT, it determined that there was no basis on which to conclude that the “potential suppli-
ers” who may make complaints under the AIT were restricted to Canadian suppliers.76
Both the Federal Court of Appeal77 and the Supreme Court of Canada78 applied a cor-
rectness standard to the CITT’s decision, in light of past jurisprudence of the federal court
and, more generally, a determination that the matter went to the tribunal’s jurisdiction.79

73 Barrie Utilities, supra note 70 at para. 128 (per Bastarache J.), quoting Domtar, supra note 72 at 775.
74 2009 SCC 50, [2009] 3 S.C.R. 309.
75 R.S.C. 1985, c. 47 (4th Supp.).
76 Northrop Grumman Overseas Corp. v. Department of Public Works and Government Services (12 September
2007), PR-2007-008, online: Canadian International Trade Tribunal <http://www.citt-tcce.gc.ca/procure/
Determin/archive_pr2h08r_e.asp>.
77 Canada (Attorney General) v. Northrop Grumman Overseas Services Corp., 2008 FCA 187, [2009] 1 F.C.R. 688.
78 Supra note 74.
79 Ibid. at para. 10. Rothstein J., writing for the Court, justifies selection of a correctness standard, first on the
basis that recent case law from the Federal Court establishes that “a CITT decision on whether something falls
within its jurisdiction” will attract correctness review. This is followed by the more categorical statement: “The
issue on this appeal is jurisdictional in that it goes to whether the CITT can hear a complaint initiated by a
non-Canadian supplier under the AIT. Accordingly, the standard of review is correctness.” See David Mullan’s
critique of the courts’ reliance on identification of a jurisdictional question to justify correctness review in this
case, rather than identifying the concern for consistency as determinative (given that different panels of the
tribunal had come to conflicting conclusions): D. Mullan, “Consistent Decision-Making,” supra note 43 at 11.
III.  The Standards of Review in Theory and Practice 341

Both courts concluded that the tribunal’s decision should be quashed. They supported this
result by way of close analysis of provisions of the AIT and the associated statutory instru-
ments. In particular, the reviewing courts determined that article 502(1) of the AIT, which
states that the AIT applies to procurement contracts “within Canada,” specifically contem-
plates contracts between governmental parties and suppliers with a base of operations in
Canada. This interpretation was attentive to an express purposes clause in which the section
addressing procurements is described as seeking “to establish a framework that will ensure
equal access to procurement for all Canadian suppliers.”80 Justice Rothstein, writing for the
Supreme Court, additionally took account of the consequences for Canada’s international
trading relationships if foreign corporations were to gain rights under the AIT despite ex-
clusion of those same rights from the international trade agreements negotiated with their
home nations.81
One may be persuaded by the careful construction of the text of the CITT and AIT under-
taken by the Federal Court of Appeal and Supreme Court of Canada, which is supported by
analysis of the adverse consequences for Canada’s international trading interests if the tribun-
al’s interpretation were to be accepted. However, comparison of the way correctness review
was conducted by the Court of Appeal and Supreme Court of Canada is arguably instruc-
tive. The Court of Appeal extensively details the tribunal’s reasoning prior to launching into
an examination of the AIT and relevant statutory instruments.82 Subsequently, in disposing
of the case, it specifically addresses the aspect of the tribunal’s reasoning that, it holds, was
in error—characterizing that error as a “failure to consider” a key element of the statutory
scheme.83 In this, the judgment of the Court of Appeal has some appearance of entering into
a dialogue with the tribunal. In contrast, the Supreme Court references the arguments
brought by the disappointed party without noting the reasoning of the tribunal in any detail.
Moreover, neither court directly addresses the tribunal’s purposive thesis that the economic
objectives of the Act may be best satisfied by recognizing the standing of non-Canadian
suppliers.84 To simply enter into an independent analysis in the absence of direct engagement
with the tribunal’s reasoning is, of course, strictly in accord with the law on correctness re-
view. Moreover, such an approach may be particularly appropriate to the Supreme Court’s
decision in this case, given that it adopts the reasoning of the Court of Appeal in significant
part. However, the question to consider in comparing these judgments is whether the Court
of Appeal takes the insights of the tribunal more seriously in its analysis—and, moreover,
whether such an approach (which, again, is in tension with the basic tenets of correctness
review) is to be preferred, precisely because it enables the parties, the wider public, and the
tribunal itself to evaluate whether and how the decision to quash was justified.
Examination of how correctness review was applied in the above cases reveals tensions
between a positivist approach to statutory interpretation, which looks to the text (or some-
times text and context) as a closed system indicative of a determinate legislative intent, and

80 Northrop Grumman, supra note 74 at paras. 22-24.

81 Ibid. at paras. 41-42.
82 Supra note 77 at paras. 13-24 (FCA).
83 Ibid. at paras. 79-81.
84 See the tribunal judgment, supra note 76 at para. 30.
342 Chapter 10  Pas de Deux: Deference and Non-Deference in Action

a normative approach, which views problems of statutory interpretation in light of back-

ground assessments not only of social facts but also of competing value-laden purposes.
Arguably the latter approach, taken seriously, begins to erode the idea that courts need not
give any weight or respect to the justificatory efforts of tribunals, even on matters tradition-
ally reserved for correctness review. This proposition is further supported by the observa-
tion that it may be difficult, if not impossible, ever to achieve a surgical separation of fact
and law or policy and law.
The bedrock of correctness review is the concept of jurisdiction: the idea that adminis-
trative decision-makers do not enjoy unlimited authority and, moreover, do not enjoy au-
thority (or final authority) on questions going to the scope of their mandate. Further, as we
have seen, the correctness standard of review reflects the rule-of-law concern for stability in
law, particularly in matters of general legal (including constitutional) significance. Yet for all
that, the standard sits uneasily with the aspiration of integrating the work of administrative
tribunals more fully into the constitutional order. For signals that this is indeed an aspira-
tion that is proper to the modern jurisprudence on the standards of review, we must now
turn to the alternative to correctness review: review for reasonableness.

B. Review for Reasonableness

Let us consider, then, what it means to express deference on review. More fully stated, the
question pursued in the jurisprudence examined in this section is whether or how the im-
peratives of deference and supervision may be integrated where judges are tasked with re-
viewing the substantive legality of administrative decisions. This is the question at the heart
of reasonableness review, one that has driven successive transformations in this area of law
over the past three decades as courts have struggled to strike a principled balance between
these imperatives.

1. Enduring Questions from the Pre-Dunsmuir Case Law

To begin with the pre-Dunsmuir deferential standards—patent unreasonableness and rea-
sonableness simpliciter 85—we know now that we must view with suspicion the case law that
sought to distinguish between these standards by reference to the depth of probing or mag-
nitude of error allowable under each. Such reasoning came in for authoritative criticism in
Dunsmuir, as both conceptually empty and practically unhelpful. However, the pre-Dun-
smuir jurisprudence may nonetheless assist in alerting us to two main areas of controversy
about deference on substantive review that remain alive today: (1) controversies about the
method or conduct of judicial reasoning on review that is most consistent with deference,
and (2) controversies about the substantive indicia of reasonableness or unreasonableness.

a.  The Conduct of Deferential Review

The majority in Dunsmuir endorses, as fundamental to the law on deference, the statement
from the 1979 decision in CUPE that statutory language may accommodate more than one

85 For a more expansive examination of these standards, see “A Fine Romance?,” supra note 4 at 233-47.
III.  The Standards of Review in Theory and Practice 343

reasonable interpretation. That thesis is essential to a further proposition that has driven the
last few decades of jurisprudence on the standards of review—namely, that there may be
good reasons for courts to defer to tribunals’ interpretations of law where those interpreta-
tions fall within the ambit of reasonableness. The question is: how are the limits of reason-
ableness to be discerned, consistent with deference?
Cases decided after CUPE at times manifested tensions in the impulse toward deference,
as courts ostensibly deployed the standard of patent unreasonableness yet engaged in what
appeared to be a search for the “right answer” to the matter in dispute.86 Nonetheless, a basic
principle of deference that prevailed both in the law on patent unreasonableness and rea-
sonableness simpliciter was that the reviewing judge must not measure the decision against
his or her sense of the “correct” decision.87 That, however, raised the question of how to
adjudge the legality of the decision under review, if not in light of the judge’s opinion of the
correct decision.
One response, which persisted until the shift to a single reasonableness standard in Dun-
smuir, was to fix on the “depth of probing” allowable on patent unreasonableness as opposed
to reasonableness review. This metaphor flourished despite the fact that, even before the rise
of the reasonableness simpliciter standard in the mid-1990s, the case law had rejected the
idea that deference meant refraining from careful engagement with administrative rea-
sons.88 That is, it was generally agreed that courts need not restrict their inquiries to, for
example, the basic defensibility of the decision-maker’s construction of the statute, absent
consideration of how the law was applied in the case at hand. Still, the jurisprudence re-
mained uncertain on how exactly to express deference, if not by “not looking” at the deci-
sion or aspects of it or by refusing to test it against the full set of statutory and evidentiary
materials on the record.
A methodological alternative to independently seeking the “right answer” on reasonable-
ness review—one that anticipates the developments we will see in Dunsmuir—is stated in
Ryan, wherein Iacobucci J. urges judges to “stay close to the reasons” for an administrative
decision, while searching for “a line of analysis within the given reasons that could reason-
ably lead the tribunal from the evidence before it to the conclusion at which it arrived.”89
Here Iacobucci J. revives the principle, reaching back to CUPE, that “a decision may satisfy
the reasonableness standard if it is supported by a tenable explanation even if this explana-
tion is not one that the reviewing court finds compelling.”90 This is bolstered by the further

86 See the examples rehearsed in the judgment of LeBel J. in Toronto (City) v. C.U.P.E., Local 79, 2003 SCC 63,
[2003] 3 S.C.R. 77 at paras. 96-99. The cases discussed include Canada Safeway Ltd. v. RWDSU, Local 454,
[1998] 1 S.C.R. 1079 and Lester (W.W.) (1978) Ltd. v. United Association of Journeymen and Apprentices of the
Plumbing and Pipefitting Industry, Local 740, [1990] 3 S.C.R. 644.
87 Ryan, supra note 39 at para. 50. I discuss the case law supporting this point in more detail in “A Fine Ro-
mance?,” supra note 4 at 233 and 236.
88 See National Corn Growers, supra note 13, and the discussion of that case in “A Fine Romance?,” supra note
4 at 236-38.
89 Supra note 39 at paras. 49 and 55.
90 Ibid. at para. 55.
344 Chapter 10  Pas de Deux: Deference and Non-Deference in Action

statement that judges should assess the “basic adequacy of a reasoned decision,” and so
­refrain from “seiz[ing] on one or more mistakes or elements of the decision which do not
affect the decision as a whole.”91
At the same time, Iacobucci J. in Ryan puts to rest the proposition that with the inception
of the reasonableness standard, further and finer gradations of deference are implicitly
registered on the spectrum of the standards of review—mirrored, perhaps, by distinctions
in the allowable “depth of probing” or “magnitude of error.” The persistence of the contro-
versy was understandable. For if the reasonableness standard was forged in an attempt to
better calibrate reviewing practices to the diversity of administrative contexts, then more
standards would seem to mean even finer calibration. In Ryan, Iacobucci J. suggests that
increasing efforts at fine-tuning when identifying the standard would only distract judges
from the central work of “explaining why the decision was not supported by any reasons
that can bear a somewhat probing examination.”92
Two cases, both of which were decided under the patent unreasonableness standard, il-
lustrate the challenge of adopting a mode of reasoning on review that avoids measuring the
contested decision against the court’s independent determination of the correct answer. The
first (and arguably more successful) is Justice Dickson’s judgment in CUPE.93 The dispute
about statutory interpretation at the heart of the case is discussed by Audrey Macklin in
Chapter 9, and I will not take up the details here.94 Instead I wish to briefly outline the rea-
sons for claiming that Dickson J.’s judgment for the Court is exemplary of deference. What
is exemplary about the judgment, I suggest, is its evaluation of the decision under review,
not simply in light of the statutory purposes, but in light of the tribunal’s own reasoning
about those purposes.95 That is, Dickson J. does not embark on a wholly independent assess-
ment of the statutory scheme or the factors of mandatory relevance to the dispute. Rather,
like L’Heureux-Dubé J. in Mossop, he repeatedly refers to, and may be said to amplify, the
tribunal’s purposive construction of the contested provision of the statute even as he in-
quires into the supportability of its interpretive conclusion. The importance of this approach
becomes clear as Justice Dickson explains that a generalist judge (or one who “draws too
heavily upon private sector experience”) might find the tribunal’s conclusions counterintui-
tive without the scaffolding of the tribunal’s reasons, which canvass the historical trade-offs
informing the regulation of public service labour relations.96 In short, Justice Dickson
grounds his analysis in a careful appraisal of the tribunal’s reasons. And so this judgment
remains a touchstone of what we have termed a constitutional pluralist orientation to the
project of statutory interpretation on review.

91 Ibid. at para. 56. See also Canadian Broadcasting Corp. v. Canada (Labour Relations Board), [1995] 1 S.C.R.
157 at paras. 48-49.
92 Ryan, supra note 39 at para. 46.
93 Supra note 3.
94 See Audrey Macklin, Chapter 9, section III, The Blockbuster: C.U.P.E. v. New Brunswick Liquor
Corporation.
95 I discuss the judgment in more detail in “A Fine Romance?,” supra note 4 at 235-36.
96 CUPE, supra note 3 at 242.
III.  The Standards of Review in Theory and Practice 345

We may contrast this to another CUPE case, decided in 2003: C.U.P.E. v. Ontario (Minis-
ter of Labour)—often referred to as the Retired Judges case.97 The decision of the majority in
this case arguably stands as an example of the interpretive attitude of judicial supremacy,
discussed in section II. According to the majority, the decision of the Ontario Minister of
Labour to appoint only retired judges as the third members of tripartite hospital labour
arbitration panels was patently unreasonable, because that decision failed to reflect con-
sideration of labour relations experience and general acceptability within the labour rela-
tions community. On a purposive and historical analysis of the statutory scheme, these two
factors were identified as having mandatory relevance to the exercise of the minister’s broad
statutory discretion. Yet the majority’s extensive analysis of the statutory scheme, performed
in an effort to discern the limits of the minister’s discretionary powers, is conducted prior
to, and wholly independent of, its formal adoption of a deferential standard or its applica-
tion of that standard to the impugned decision. In other words, the majority independently
ascertains, if not the “correct answer” to the matter in dispute, then the construction of the
statute that is determinative of the result. Is this consistent with deference?98
Bastarache J.’s dissent in Retired Judges directs criticism at the excessive “probing” re-
quired to identify the majority’s factors of mandatory relevance. Bastarache J. also argues
that the majority’s conclusion that the minister failed to consider certain factors is but a
weak cover for its impermissible reassessment of the relative weight accorded by the minis-
ter to these and other factors.99 We inquire further into this form of critique in a moment,
in taking up the substantive indicia of unreasonableness. The main point for now is that the
majority judgment in the Retired Judges case sits uneasily, not only with the traditional bar
on revisiting administrative assessments of the weight accorded factors of relevance to their
decisions, but with the methodological imperative of deference established above—that is,
that judges “stay close” to the reasons of the decision-maker on review. We must acknow-
ledge, however, that the latter imperative is far more likely to be satisfied where the decision
under review is adjudicative or otherwise supported by formal reasons than in circum-
stances like those in Retired Judges, where the reasons were in effect judicially reconstructed
in light of the legislative history and statements from members of the executive, along with
responses given in the cross-examination of the minister’s senior advisor.100

b.  Substantive Indicia of (Un)reasonableness

We have seen that, from its inception in the mid-1990s, the law on reasonableness simpliciter
review urged judges to attend closely to administrative decision-makers’ reasons on review.

97 2003 SCC 29, [2003] 1 S.C.R. 539 [Retired Judges].

98 I discuss the case further in “A Fine Romance?,” supra note 4 at 238-40. For a trenchant critique of the ma-
jority judgment, emphasizing the importance of the legislature’s having left the disputed decision to the
minister’s “opinion,” see G. Huscroft, “Judicial Review from CUPE to CUPE: Less Is Not Always More” in
Inside and Outside, supra note 19 at 296. See also L. Sossin, “Empty Ritual, Mechanical Exercise or the Disci-
pline of Deference? Revisiting the Standard of Review in Administrative Law” (2003) 27 Advocates’ Q. 478,
especially at 504-5.
99 Retired Judges, supra note 97, per Bastarache J. at paras. 35-36.
100 Ibid. at paras. 181-82.
346 Chapter 10  Pas de Deux: Deference and Non-Deference in Action

Indeed, it may be said that what animates the subtle interposition of reasonableness simplic-
iter between correctness and patent unreasonableness review is a confidence in reason. This
confidence is expressed, on the one hand, in the idea that judges are capable of appreciating
the reasons of administrative decision-makers (though they must work to do so, by turning
their mind to the statutory and wider institutional rationales for deference) and, on the other
hand, in the idea that administrative decision-makers can communicate the reasonableness
of their decisions to reviewing judges (though they too must work at this, by articulating
their decisions in a manner that is sensitive to the demands of public justification).
This is confirmed in statements from Iacobucci J. in Canada (Director of Investigation
and Research) v. Southam101 on the links between deference, expertise, and reason-giving.
He writes: “In the final result, the standard of reasonableness simply instructs reviewing
courts to accord considerable weight to the views of tribunals about matters with respect to
which they have significant expertise.”102 Yet to this Iacobucci J. appends an extended cita-
tion to the effect that unless experts are able to explain “to a fair-minded but less well-­
informed observer, the reasons for their conclusions,” then “they are not very expert” and
no deference is commanded.103 That is, the expert status of a tribunal will not serve as a
stand-alone justification for deeming a decision reasonable, absent reasons verifying the
application of expertise to the matter at hand.
But apart from the imperative of “staying close” to administrative reasons on review,
what guidance does the pre-Dunsmuir case law provide on the substantive indicia of reason-
ableness, or more pointedly, unreasonableness? Again we may draw on Iacobucci J.’s judg-
ment in Southam, where he states:
[A] court reviewing a conclusion on the reasonableness standard must look to see whether any
reasons support it. The defect, if there is one, could presumably be in the evidentiary founda-
tion itself or the logical process by which conclusions are sought to be drawn from it. An ex-
ample of the former kind of defect would be an assumption that had no basis in the evi-
dence … . An example of the latter kind would be a contradiction in the premises or an invalid
inference.104

This description of reasonableness simpliciter review plays up a concern to identify deci-

sions that have no basis in evidence or that fail to adhere to basic principles of practical
reasoning or logic. Yet it arguably leaves out the sort of cases that have proven most chal-
lenging and controversial. That is, the case law on both patent unreasonableness and review
for reasonableness simpliciter (and more recently, Dunsmuir reasonableness) has brought
the challenges of deference into clearest focus where what is in issue is the decision-maker’s
construction of the values or ends to be advanced under the enabling statute. Just how the
substantive limits of legality are to be identified in such cases is a central problem for rea-
sonableness review.

101 [1997] 1 S.C.R. 748 [Southam].

102 Ibid. at para. 62.
103 Ibid., citing R.P. Kerans, Standards of Review Employed by Appellate Courts (Edmonton: Juriliber, 1994) at 17.
104 Southam, supra note 101, at para. 56.
III.  The Standards of Review in Theory and Practice 347

In Ryan, Iacobucci J. emphasizes (as Dickson J. had in CUPE) that, “[u]nlike a review for
correctness, there will often be no single right answer to the questions”105 attracting review
for reasonableness. Here he directly engages problems of statutory construction hinging on
the prioritization of competing statutory purposes:
For example, when a decision must be taken according to a set of objectives that exist in tension
with each other, there may be no particular trade-off that is superior to all others. Even if there
could be, notionally, a single best answer, it is not the court’s role to seek this out when deciding
if the decision was unreasonable.106

This statement affirms the principle, long accepted in the case law (but, as we will see,
increasingly subject to erosion or exceptions), that deferential review must avoid second-
guessing administrators with respect to the weight or priority they assign to competing
statutory purposes.107 That is, traditionally, deference has been understood to mean that
reviewing judges are prohibited from revisiting the relative weight placed by administrative
decision-makers on the competing factors of relevance to their decisions—although they
may insist, as a matter of law, that the decision-maker consider all factors of mandatory rel-
evance (typically identified through a purposive construction of the statutory scheme). In
this, “failing to consider a relevant factor” has taken on the character of an objective legal
defect, while judgments about the weight to be accorded competing values, interests, or
other considerations relevant to the statutory mandate have been deemed matters not of law
but of discretion or policy, rightly falling to the administration. In Southam, Iacobucci J.’s
refusal to revisit the weight placed by the tribunal on competing elements of a multi-­
factored economic analysis stands as a compelling illustration of how the prohibition may
play out in practice.108 But is the prohibition on revisiting the weight accorded by adminis-
trative decision-makers to the factors relevant to their decisions consistent with your intu-
itions about the substantive qualities of reasonableness in law or, for that matter, with the
developing law on substantive reasonableness?
The prohibition was significantly called into question in Baker. There, the majority judg-
ment of L’Heureux-Dubé J. exposed the potential for conflict between this imperative of
deference as traditionally understood and the commitment to ensure that administrative
decisions and decision-makers are fully integrated into the work of advancing the rule of
law. You will recall that, on applying a reasonableness standard to the minister’s (or rather,
the minister’s delegate’s) exercise of discretion in that case, L’Heureux-Dubé J. determined
that the notes of a junior officer, deemed to be the reasons for the decision, were inconsis-
tent with “the values underlying the grant of discretion.”109 Moreover, L’Heureux-Dubé J.
indicated that the values that underlie or set reasonable limits to discretionary powers issue

105 Supra note 39 at para. 51.

106 Ibid.
107 The most recent incursion on this principle, as discussed further below, is Doré, supra note 4.
108 See my discussion of Southam (supra note 101) in “A Fine Romance?,” supra note 4 at 242-43.
109 Baker, supra note 6 at para. 65. The legal and factual background to the case, along with a more complete
analysis, is provided by Geneviève Cartier in Chapter 11, Administrative Discretion: Between Exercising
Power and Conducting Dialogue.
348 Chapter 10  Pas de Deux: Deference and Non-Deference in Action

not only from a decision-maker’s enabling legislation and associated regulations, but also
from instruments of soft law (for example, departmental policies and guidelines), the com-
mon law (“the principles of administrative law”), the Constitution (“the principles of the rule
of law” and “the principles of the Charter”), international law, and even what are tantaliz-
ingly (and, depending on one’s perspective, perhaps worryingly) held out as the “fundamen-
tal values of Canadian society.”110 L’Heureux-Dubé J. concluded that the sources relevant to
the case at hand (the statute, an international convention ratified but not incorporated into
domestic legislation, and ministerial guidelines) established that “the rights, interests, and
needs of children and special attention to childhood are important values that should be
considered in reasonably interpreting the ‘humanitarian’ and ‘compassionate’ considera-
tions that guide the exercise of the discretion.”111 Because the officer’s notes failed to reflect
that the decision-maker was “alive, alert, or sensitive to the interests of Ms. Baker’s chil-
dren,”112 the decision failed to meet the standard of reasonableness simpliciter.
Following the ruling in Baker, commentators and courts alike were concerned to settle
the matter of whether this decision was a straightforward example of the tradition of vitiat-
ing a discretionary decision for failure to consider a relevant factor at all (the children’s in-
terests). Or was it a departure from the traditional approach, amounting to a revisiting of
the weight placed on a particular factor (those interests)? The comments of L’Heureux-
Dubé J. left some uncertainty. At one point she wrote that the officer’s notes were “com-
pletely dismissive of the interests of Ms. Baker’s children.” But in the next sentence, she
wrote, “the failure to give serious weight and consideration to the interests of the children
constitutes an unreasonable exercise of the discretion.”113 Further uncertainty arose from
L’Heureux-Dubé J.’s suggestion that, depending on the circumstances, a discretionary deci-
sion-maker may be accorded deference in connection with the factors he or she deems
relevant to a given decision, and not merely in connection with the weight he or she gives
to mandatory relevant factors.114 This represents an even more radical departure from the
traditional approach to substantive review (even on a deferential standard)—antithetical to
the constitutional traditionalist’s view of the mandate of courts to patrol the limits of admin-
istrative jurisdiction.
In Suresh v. Canada (Minister of Citizenship and Immigration),115 the Court responded to
some of these uncertainties. Baker, it stated, “does not authorize courts reviewing decisions
on the discretionary end of the spectrum to engage in a new weighing process, but draws

110 Baker, supra note 6 at paras. 56 and 67.

111 Ibid. at para. 73.
112 Ibid.
113 Ibid. at para. 65. See the discussion of how these statements might be reconciled, in D. Mullan, “Deference from
Baker to Suresh and Beyond—Interpreting the Conflicting Signals” in D. Dyzenhaus, ed., The Unity of Public
Law (Portland, OR: Hart Publishing, 2004) 21 at 31-37 [The Unity of Public Law]. Among the more convincing
interpretations is that the factor in question (the children’s interests) read in light of the statute, the applicable
ministerial guidelines, and international law, necessarily carries with it “elements of weight or degree.”
114 Baker, supra note 6 at para. 56: “In fact, deferential standards of review may give substantial leeway to the
discretionary decision-maker in determining the ‘proper purposes’ or ‘relevant considerations’ involved in
making a given determination.”
115 2002 SCC 1, [2002] 1 S.C.R. 3 [Suresh].
III.  The Standards of Review in Theory and Practice 349

on an established line of cases concerning the failure of ministerial delegates to consider and
weigh implied limitations or patently relevant factors.”116 Here the possibility that certain
factors may be deemed by courts to demand prioritization (or “serious weight and consider-
ation”) in light of statutory or other legal sources in particular cases is left open. The Court
in Suresh does not address the proposition advanced by L’Heureux-Dubé J. in Baker that
some discretionary decision-makers may receive deference in relation to the factors they
deem to be relevant to their decisions.
The multiple, complex messages in Baker—endorsing, on the one hand, the idea that
deference may even be directed at decision-makers’ assessments of statutory purposes and
so to their assessments of the considerations that are of mandatory relevance to their deci-
sions, and on the other, the idea that those decision-makers must conform with the “values
underlying the grant of discretion,” even necessarily placing significant weight on one or
another of these where the reviewing judge deems that to be required by the statutory text
or context—express a tension that is basic to reasonableness review on the romantic ac-
count. That is, this tension is inherent in the idea of “deference as respect” advanced by
David Dyzenhaus and explicitly endorsed by L’Heureux-Dubé J.117 Such respect places con-
fidence in the capacity of decision-makers to discern the limits of their legal mandates in a
nuanced, context-sensitive fashion, while at the same time insisting that this capacity be
demonstrated in their decisions. In other words, deference under this standard carries with
it an expectation that decision-makers can and will identify, and evince appropriate sensi-
tivity to, the values (including the significant individual interests) that should inform the
exercise of their statutory powers. But, we may ask: is this really so different from correct-
ness review? Is this just an old-time assertion of the courts’ “jurisdiction”—and so of judges
(historically prioritizing individual rights over the public purposes driving administrative
mandates) having the last word on matters given to tribunals to decide? If not, how exactly
is it different?
On the romantic account of substantive review, the decisions of judges on review must
also be publicly justified. To that end, would it not be better that judges’ background assess-
ments of weight, where these assessments differ from those of administrative decision-
makers, be explicitly stated and subjected to public scrutiny—thus alleviating the possibility
of their covert operation? On turning to post-Dunsmuir developments, we will see that,
with Doré,118 the Supreme Court has unleashed a new set of possibilities for revisiting the
weight accorded to competing legal values on reasonableness review—at least, where the
decision is constructed as an exercise of adjudicative discretion and where Charter values
are understood to be engaged.
In sum, although the logic of reasonableness simpliciter was drawn out of the law on pat-
ent unreasonableness—a line of case law ostensibly tolerant of administrative decisions that
were unreasonable, though not “patently” so—the cases elaborating the new standard intro-
duced new expectations of judicial attentiveness to administrative reasoning, along with

116 Ibid. at para. 37.

117 “Deference as respect requires not submission but a respectful attention to the reasons offered or which could
be offered in support of a decision”: Baker, supra note 6 at para. 65, citing “The Politics of Deference,” supra
note 4 at 286.
118 Supra note 4.
350 Chapter 10  Pas de Deux: Deference and Non-Deference in Action

new expectations that administrative decision-makers publicly justify their decisions. Argu-
ably, these central features of reasonableness review begin to erode the logic underpinning
the formal non-engagement with tribunal reasons that characterizes correctness review.119

2. Reasonableness Post-Dunsmuir
a.  Dunsmuir Reasonableness in Theory
As Audrey Macklin has explained in Chapter 9, Dunsmuir’s collapsing of patent unreason-
ableness and reasonableness simpliciter into a single standard of deferential review marks an
effort to simplify this area of law and, at the same time, to set it on a more principled foun-
dation. In place of the prior case law’s shaky gestures to depth of probing and magnitude of
error, the Dunsmuir majority makes tentative attempts to offer clearer guidance about what
it means to express deference on review. Yet the majority does not pull the essential features
of the revised reasonableness standard out of thin air, but rather builds on the foundations
of the prior case law on deference, including Justice Iacobucci’s descriptions of reasonable-
ness review in Southam120 and Ryan121 and Justice Dickson’s judgment in CUPE.122 Those
judgments centred on the idea that judges applying a reasonableness standard should
closely attend to administrative reasoning, and that the decision should stand unless it “can-
not be rationally supported by the relevant legislation”123 or the evidence.
The question is whether Dunsmuir or the subsequent case law adds anything to the law
on deferential review that might ease the historical tensions—at times wild mood swings—
affecting this area of law. I refer to the aforementioned instability in the case law applying a
deferential standard as between attitudes of judicial supremacy (setting strict limits of legal-
ity within which administrative reasoning is closely hedged) and attitudes of judicial abdica-
tion (for example, refusing to peer “too deeply” into the reasoning or evidentiary record, or
to revisit administrative assessments of the relative weight of competing factors, including
statutory objectives or legal values). Ultimately, we may be forced to conclude that the jury
is still out on the success of Dunsmuir and the subsequent case law in negotiating these ex-
tremes. At the same time, read in light of the romantic’s concern that the law on substantive
review should help coordinate the work of judges and administrative decision-makers in a
“culture of justification,” the principles of reasonableness review asserted in and after
­Dunsmuir are arguably a step in the right direction.

119 See D. Dyzenhaus, “David Mullan’s Rule of (Common) Law” in Inside and Outside, supra note 19 at 474-75.
Dyzenhaus argues that reasonableness simpliciter “shears the correctness standard off the continuum of the
standards of review.” Also see “Constituting the Rule of Law,” supra note 5 at 495.
120 Supra note 101.
121 Supra note 39.
122 Supra note 3.
123 CUPE, ibid. at 237, cited in Dunsmuir, supra note 1 at para. 35.