The Unwritten Constitution

and the Rule of Law

. Dyzenhaus *
David

l. INTRODUCTION
This article addresses the themes of the conference panel for which
it was prepared: "Constitutional Interpretation, Extrapolation and Inter-
polation." It does so not from the perspective of a bill of rights, but from
the perspective of the common law of judicial review. My argument in
brief is that even in the absence of a written constitution or bill of rights,
judges have to engage in interpretation of constitutional values, if they
are to make sense of the fact that we aspire to live under the regime of
the rule of law.
I believe the terrain of such values - the terrain of the unwritten
constitution - to be important to the theme of the legitimacy of these
modes of interpretation for the following reason. If interpretation of
constitutional ~alues were an activity which judges could not avoid,
then it might seem that what they are doing when they interpret is also
legitimate. However, one should not confuse necessity with legitimacy.
That one is forced to do something might excuse without justifying.
Suppose that a theory holds that illegitimate activity occurs when judges
impose values on a statute, in the sense that the values they claim to find
in the statute are not the values explicitly stated by the legislature. Sup-
pose further that the reason this activity is illegitimate is that the legisla-
ture represents the people and should have a monopoly on the creation
of binding values. On this theory, one should shun bills of rights because

"' Professor of Law and Philosophy, University of Toronto. Thanks to Grant Huscroft
for comments on the first drapt and to my research assistants Erika Eineigel and Umut Oszu
for both their help and comments. I also thank Will Waluchow, who commented on the
second draft of the paper at the seminar which followed the Conference at which the paper
was presented, and Genevieve Cartier for several conversations about the topics dealt with in
this article.
384 Constitutionalism in the Charter Era

they give judges a licence to impose their sense of value on statutes. A

bill of rights is, that is, a political mistake. Moreover, it remains a mis-
take even when the decision to have a bill of rights is taken by the legis-
lature. There is a puzzle here similar to the puzzle about why one cannot
validly consent to enslavement. But just as we know the answer to the
puzzle about consent, so we should know the answer the theory has to
give to what we might think of as constitutional enslavement. The Iegis·
lature has enslaved itself- severely curtailed its liberty - by commit-
ting itself to prior value constraints. More accurately, it has committed
itself to rule by judges. Judges will not only have very different views
about the values required, but will be tempted to use the values as a
vehicle for importing values that were never contemplated by the legis-
lature when it enacted the statute. Judges might have no choice but to
interpret the bill of rights when it is clearly relevant, but that they have
no choice does not make their activity legitimate.
So my argument is that the necessity of judicial interpretation of
constitutional values does not depend on the presence of a bill of rights.
Even in the absence of such a bill, judges must - if they are to be faith-
ful to their duty to uphold the rule of law - interpret the positive law of
a legal order in light of their understanding of unwritten constitutional
values. It follows that the objection to the legitimacy of constitutional
interpretation and to its more creative sprouts - extrapolation and in-
terpolation - is an objection to the rule of law. A choice to govern
through legal order involves a commitment to abide by the values of the
rule of law, the unwritten constitution of legal order.
My foil is the legal theory of United States Supreme Court Justice
Antonin Scalia, whose position on constitutional interpretation is also
the foil for Laurence Tribe's and Michael Dorfs article, where they
deploy the terms constitutional interpretation, extrapolation, and inter-
polation to describe the reasoning in a dissent by Harlan J. 1 But I will
argue that Tribe and Dorf underestimate the challenge of Scalia J. 's
position, a challenge best taken up on the terrain of the unwritten consti-
tution.

1
Tribe and Dorf, "Levels of Generality in the Definition of Rights" (1990) 57 U. Chi-
cago L. Rev. 1057, referring to Poe v. Ullman, 367 U.S. 497, at 522 (1961).
 The Unwritten Constitution 385
and the Rule o.fLaw

II. INTERPRETATION: A NECESSARY BUT ILLEGITIMATE ENTERPRISE

Tribe and Dorf respond to a suggestion by Scalia J. about how to se-

lect the appropriate level of generality in detennining a right. His
method examines "the most specific level at which a relevant tradition
protecting, or denying protection to, the asserted right can be identi-
fied"2 in order, they say, to achieve value neutrality and thus avoid the
arbitrariness of judicial value judgments. But Tribe and Dorf argue that
Scalia J. 's method turns out to be a disguise for the importation of val-
ues and they advocate a different method, one that requires.value judg-
ments but avoids arbitrariness. They advance
a theory of constitutional interpretation that takes as its point of departure
Justice Harlan's observation in Poe that the search for unenumerated
rights should proceed by interpolation and extrapolation from the
enumerated rights. . . . [W]e argue that a typical judicial opinion
distinguishes between essential and non-essential facts, and that by paying
attention to such distinctions, judges trained in the method of the common
law can generalize from prior cases without merely imposing their own
3
values.

According to the~
Justice Harlan was engaged in a process of interpolation and extra-
polation. From a set of specific liberties that the Bill of Rights explicitly
protects, he inferred unifying principles at a higher level of abstraction,
focusing at times upon rights instrumentally required if one is to enjoy
those specified, and at times upon rights logically presupposed if those
4
specified are to make sense.

In order to avoid the intricacies of American constitutional debate, I

want to restate their challenge to Scalia J. more abstractly. Judicial in-
terpretation of a constitutional document is necessarily a value-laden
enterprise. Hence, claims to value neutrality must disguise judicial im-
portation of value. And since that importation happens in a subterranean
fashion it is more likely than other methods to be arbitrary in the sense
that the outcome will depend not on the law but on the judge's political

2
Michael H. v. Gerald D., 491 U.S. 110, at 127 (1989).
3
Tribe and Dorf, supra, note 1, at 1059 {emphasis omitted).
4
Id., at 1068.
386 Constitutiona/ism in the Charter Era

preferences. For these other methods acknowledge the value-laden na-

ture of interpretation by explicitly engaging in it. Their structure weeds
out arbitrariness by anchoring all interpretations firmly in the text of the
constitution and by accepting the onus of justifying any principles that
are not enumerated in the text as principles necessary to make sense of
or achieve the purposes of the enumerated principles.
Necessity does not, of course, mean anything like logical necessity
since one has to recognize that judges who work within this structure
will on occasion, even often, disagree with each other about how the
onus is best discharged. So their conclusions will depend in a non-trivial
way on their individual understandings of how to make the best moral
sense of the constitutional text, that is, on their individual moral sensi-
bilities. But to recognize this fact is not to concede that the conclusions
are arbitrary. The kind of objectivity that is available in constitutional
interpretation is not to be equated with value neutrality, but with adher-
ence to the rigours of the explicit, properly structured interpretative
process.
If my more abstract account of the challenge is accurate then it is
one whose substance I by and large accept. But my concern is that it
does not quite confront Scalia J. 's position because it requires that he
share an important assumption that I think the coherence of his legal
theory requires him to reject. This is the assumption of the legitimacy of
judicial interpretation, however understood, of the written constitution.
It might seem strange to attribute to a well-known judge of the
United States Supreme Court a rejection of the legitimacy of constitu-
tional interpretation. Consider, however, the central elements of Scalia
J. 's description of his legal theory in A Matter ofInterpretation. 5 Justice
Scalia' s target in that book is the methodology of common law interpre-
tation, which requires that judges try to reach a conclusion they think
represents the best resolution of the case. He claims not to object to that
methodology when it is applied to the common law. But he objects
strongly on democratic grounds to its application to statutes, since such
application involves substituting the judicial sense of what is appropri-
ate for what the legislature as a matter of fact intended to convey in the
text of the statute. And he objects even more strongly when that same
methodology is applied to interpreting the constitution. What we have

5
Scalia, A Matter ofInterpretation: Federal Courts and the Law (1997).
 The Unwritten Constitution 387
and the Rule ofLaw

then is "the common law returned, but infinitely more powerful than
what the old common law ever pretended to be, for now it trumps even
the statutes of democratic legislatures."6 In particular, he objects to the
idea of the "Living Constitution," the idea that the Constitution should
be regarded as a document whose meaning will evolve to conform to
new understandings of the rights and principles that require protection.
Not only do we get through this idea judicial legislation of appropriate
moral standards, but it contradicts the "whole purpose" of a constitution
- "to prevent change."7
It is significant that Scalia J. does not make this point just about the
United States Constitution, but about constitutions in general. It is not
only the case that the judges of many other supreme courts are much
more open than the judges of his Court to the idea of a living constitu-
tion, but in fact their constitutions often enjoin them to be open in this
way, even requiring them to interpret their domestic documents in the
light of evolving standards of international law. That Scalia J. is willing
to make this point tells us that his textualism is not so much about the
words in a text. Nor is it even about what one might infer about the
intentions of the drafters or the expectations of their audience, for ex-
ample, that judges charged with interpreting the text of the South Afri-
can constitution are under a duty to update their understanding of what
the constitution requires in the light of evolving international law.
Rather, his textualism is a political stance about how to interpret consti-
tutional texts, whatever the texts say, in order to try to preserve to the
greatest extent possible the integrity of his theory of legal order.
His theory is within the tradition I call democratic legal positivism.
According to this tradition, we live in an era where not only are statutes
the primary source of law but also, because ours is the democratic era
and because statutes represent the judgment of the people, statutes are
the only legitimate source of law. Nevertheless, in the face of this reality
judges wedded to common law methodology under the guise of inter-
preting the law persist in imposing their understandings of what law
ought to be on statutes, thus usurping legislative authority.
These positivists combine a democratic argument that it is for the
legislature alone to decide what values should have legal force with a

6
Id., at 38.
7
Jd., at4047, at40.
388 Conslitutionalism in the Charter Era

positivist argument that the most fundamental value of the rule of law is
certainty. On their model of law, values and norms have legal force only
when they can be identified as law by hard or factual tests. Such tests
require judges to avoid reliance on their own moral judgment, thus help-
ing to ensure that legal values and norms are only those values and
norms that have been explicitly incorporated by the legislature into the
law of the land.
Positivist judges who find that their duty to apply the law includes
the duty to interpret a bill of rights will experience some dissonance
because they desire to avoid the kind of moral deliberation required by
this duty. They cope with the dissonance by confining the scope of their
interpretations to various proxies for factual legislative intention -
what the founding fathers in fact had in mind, what their immediate
audience would have taken them to have in mind, and so on. These
interpretative techniques are rife with well-known problems. But the
problems are serious only if one regards the techniques as genuine. at-
tempts to legitimate constitutional interpretation. If instead they are seen
as techniques or holding actions, designed to limit the scope of an ille-
gitimate activity in which judges have no choice but to engage, then the
techniques are a lot more plausible.8 It follows that any interpretative
activity that goes beyond these techniques is even more illegitimate.
Extrapolating from the constitutional text is more illegitimate because it
departs by definition from textualism, while interpolation - adding
words to the text-is even worse.9

8
See the evocative title of Scalia, "Originalism: The Lesser Evil" (1989) 57 U. Cin·
cinnati L. Rev. 849, though he does not depict his position as I do here. In his presentation at
the conference, Scalia J. poured scorn on those who do not hold his position because he
alleged that they are committed to believing that the drafters of the Constitution intended that
the Constitution mean whatever judges in the future want it to mean. That belief is of course
absurd, but it is not more absurd than Scalia J.'s belief that the drafters intended the Constitu-
tion to be interpreted as they wanted it interpreted. As countless critics of Scalia J.'s kind of
position have pointed out, no one who drafts a bill of rights can, as a matter of logic, intend
the bill to be governed by her particular understanding of the rights, otherwise she would
enact that understanding. The fact that judges and academics persist in holding this equally
absurd position deserves an explanation. And I believe that the explanation might well lie in
the fact that the position, while illogical, is required for judges of a positivist bent to continue
to work in a legal order much of whose structure is resistant to their views.
9
See Scalia, supra, note 5, at 147, where he concedes that if one does not take it as a
given that judges should enforce a constitution against the legislature then his "argument
ceases to bave force as a justification for [his) mode of interpretation [and] becomes an
 The Unwritten Constitution 389
and the Rule ofLaw

Textualism is then not a way of legitimating an approach to consti-

tutional interpretation but a compromise positivist judges make in order
to prevent a bad situation from getting worse. It is much the same ap-
proach that is advocated for interpreting the text of ordinary statutes
when they incorporate open-textured value terms like fairness or rea-
sonableness. These terms should not be treated by judges as invitations
to engage in deliberation about their meaning, but as landmines the
judges should try to defuse by confining their scope to the extent possi-
ble.10 Similarly, the common law is to be treated as far as possible as a
system of determinate rules whose content does not form a backdrop for
interpretation of general law, but rather are rules that apply only within
particular areas of private law. Thus positivist judges will try hard to
"hedge,, themselves in by "announcing rules" in their judgments. 11
The compromises positivist judges make are forced on them by the
fact that the Benthamite dream of a completely codified legal order in
which all law is positive law with a determinate content was never real-
ized. They are thus forced to try to make the institutions of the legal
order in which they find themselves conform as far as possible to their
understanding of law and the rule of law. 12 At best, such judges will
have a profound ambivalence to the common law, something nowhere
better illustrated than in A Matter ofInterpretation. Thus, while Scalia J.
is willing to have the writ of the common law run in private law, he is
hostile to the idea that the common law should form an interpretative
backdrop to the interpretation of statutes. Here he exemplifies the classic

argument directed to the overall inconsistency of the evolutionists: Why, given what they
believe the Bill of Rights is, would they want judges to be its ultimate interpreters?"
10
This point explains why, for example, it makes sense to describe Viscount
Maugham's judgment in Liversidge v. Anderson (1941), [1942] A.C. 206 as a positivist one,
despite the fact that he interpreted the requirement that the minister had to have "reasonable
cause to believe" that someone was a threat to national security and so should be detained as
really meaning if the minister were "satisfied that ...." And that "interpolatory" move had the
result, in his view, that the Court could not require the minister to justify the detention to it.
See my "Intimations of Legality Amid the Clash of Arms," International Constitutional Law
Journal [forthcoming).
11
Scalia, "The RuJe of Law as a Rule of RuJes" (1989) 56 U. Chicago L. Rev. 1175, at
1180.
12
See Dyzenhaus, "The Genealogy of Legal Positivism" (2004) 24 Oxford J. Legal
Stud. 39. This article provides the bridge from my argument in this chapter to the kind of
positivism articulated by HLA Hart and Joseph Raz. In the absence of that bridge, my argu-
ment has to be confined to positivist judges.
. 390 Conslitutionalism in the Charter Era

hostility of positivism to the common law tradition. He also displays the

concern that common law judges will say that they are simply deciding
in accordance with the reason of the law in order to bootstrap them-
selves into an increasingly powerful position in their relationship with
the legislature.
This shows that positivists cannot be ambivalent about one area of
constitutional law in common law jurisdictions, the common law of
judicial review of administrative action, where - to adopt a classic
formulation - the ')ustice of the common law will supply the omission
of the legislature." 13 To them, the common law of judicial review seems
like the paradigm of judicial usurpation of an alleged legislative monop-
'oly on law-making power.
One way of trying to legitimate the common law of judicial review
is to point out that there has to be an omission - a legislative silence or
gap for judges to fill. Since the legislature has the power either to pre-
empt such judicial creativity by being explicit from the outset or to react
by way of amending legislation, it is legitimate for judges to assume that
initial silence or the failure to amend are signs of tacit legislative con-
sent. In my view, this is not a very convincing explanation for the le-
gitimacy of judicial review. If the idea that one can discern actual
legislative intent in statutory text is problematic, it seems doubly prob-
lematic to infer actual intent from silence. I will come back to the issue
of legitimacy below. For the moment, I want to note only that judicial
review might seem less objectionable, albeit still illegitimate, to positiv-
ists if the legislature in the absence of a written constitution seems able
to either pre-empt or override the judges. That same fact might also
make the common law of judicial review seem not really part of consti-
tutional law, if one equates constitutional law with law that constrains
legislatures.
I will argue below that this equation is wrong. The fact that a legis-
lature has the power to override a constitutional principle does not show
that the principle is not constitutional. And there is perhaps a kind of
appreciation of this point in the unwillingness of most democratic legal

13
Cooper v. Wandsworlh Board of Works (1863), 143 E.R 414, at 420 (Eng. C.P.).
 The Unwritten Constitution 391
and the Rule oflaw

positivists to take any comfort from the potential for legislative over-
ride.14 My argument starts with a sketch of judicial review in Canada.

Ill. THE COMMON LAW OF JUDICIAL REVIEW

Consider a recent decision of the Canadian Supreme Court,

C. U.P.E. v. Ontario (Minister of Labour). ' 5 This case concerned the
interpretation of an Ontario statute that replaced the right of certain
unions to strike and of employers to lock out with a system of compul-
sory arbitration - the Hospital Labour Disputes Arbitration Act
16
(HLDAA). In simplified outline, the system of arbitration that had
evolved was one in which arbitrators were drawn from a mutually ac-
ceptable list of experienced people. When the parties were unable to
agree on an outcome, the arbitrator would be appointed to settle the
matter.
A Conservative provincial government came to power in 1995 on a
platform that included massive restructuring of the public service. It did
not amend the statute but decided to replace the list of mutually acceptw
able arbitrators with one composed of retired judges. This practice was
challenged in the courts and the Ontario Court of Appeal upheld the
union's argument that the minister, in changing the process for making
appointments, had violated the principles of natural justice. Justice Bin-
nie for the majority of the Supreme Court also found for the union but
on different grounds - the interpretation of the statute, in particular
section 6(5) which gave the minister the power to appoint as arbitrator
"a person who is, in the opinion of the Minister, qualified to act." In
Binnie J. 's view, once the statute was understood in its context, one had
to see that the minister had a "narrow role" . . . "to substitute for the
parties in naming a third arbitrator in case of their disagreement." 17 In
that context, he found that there were implied limits on the minister's
discretion - the statute required him to ~'select arbitrators from candi-
dates who were qualified not only by their impartiality, but by their

14
The exception is Goldsworthy, "Judicial Review, Legislative Override, and Democ-
racy" in Campbell, Goldsworthy, and Stone, eds., Protecting Human Rights: Instruments and
Institutions (2003) 263.
15
[2003} 1 S.C.R. 539 [hereinafter "CUPE '].
16
R.S.O. 1990, c. H.14.
17
CUPE, supra, note 15, at para. 50.
392 Constitutionalism in the Charter Era

expertise and general acceptance in the labour relations community." 18

He thus found that when the minister selected candidates without any
regard to these limits, he was acting in a way that was manifestly or
patently unreasonable, and thus did not meet the appropriate standard
for review of such decisions.
One of Binnie J.'s principal points of reference is Roncarelli v. Du-
p/essis,19 a Supreme Court decision from the 1950s, and in particular the
reasoning of Rand J., who authored several judgments which are re-
garded as relying on the idea of unwritten constitution or implied bill of
rights. In Roncarelli the issue arose because Roncarelli had the liquor
licence for his restaurant revoked as a punishment for standing bail for
Jehovah's Witnesses, whose missionary activities were bitterly opposed
by the government of the province. In order to find that the premier of
Quebec had acted illegally in instigating the revocation, Justice Rand
had to overcome the obstacle that the official in charge seemed a "law
unto himself,'' as one of the dissenting judges put it. This judge held the
view that the only controls on the official were the controls that the
legislature had explicitly prescribed and there all that the legislature had
done was give the official a subjectively phrased discretion.20
Two paragraphs of Rand J. 's judgment are worth quoting in full on
this point. He said:
In public regulation of this sort there is no such thing as absolute and
untrammelled "discretion", that is that action can be taken on any ground
or for any reason that can be suggested to the mind of the administrator;
no legislative Act can, without express language, be taken to contemplate
an unlimited arbitrary power exercisable for any purpose~ however
capricious or irrelevant, regardless of the nature or purpose of the statute.
Fraud and corruption in the Commission may not be mentioned in such
statutes but they are always implied as exceptions.

"Discretion" necessarily implies good faith in discharging public duty;

there is always a perspective within which a statute is intended to operate;
and any clear departure from its lines or objects is just as objectionable as
fraud or corruption. Could an applicant be refused a permit because he had
been born in another province, or because of the colour of his hair? The
ordinary language of the legislature cannot be so distorted. . .. That, in the

18 •,J
l«., at para. 49.
19
[1959] S.C.R. 121 [hereinafter "Roncare/U"].
20
Id, at 166-67, per Cartwright J. (dissenting).
 The Unwritten Constitution 393
and the Rule ofLaw

presence of expanding administrative regulation of economic activities,

such a step and its consequences are to be suffered by the victim without
recourse or remedy~ that an administration according to law is to be
superseded by action dictated by and according to the arbitrary likes,
dislikes and irrelevant purposes of public officers acting beyond their
duty, would signalize the beginning of disintegration of the rule of law as
21
a fundamental postulate of our constitutional structure.

Justice Rand is reasoning that it is a constitutional requirement that

administrative officials act in accordance with the rule of law, whether
or not there is a written constitution that imposes that requirement.
Moreover, for him that requirement makes sense only if the rule of law
is understood quite substantively, as including values that inform the
perspective within which judges are to understand the statute which
delegates authority to the administration.22
Now in CUPE the government lawyers did not claim that the minis-
ter was a law unto himself- that he had an "absolute and untrammeled
discretion." Rather their claim was that the minister had legitimate rea-
sons for his conduct which were closely associated with the purpose of
the statute, including the chronic delay and cost associated with arbitra-
tions and the fact that he was drawing on "people who had spent their
professional lives as neutrals. "23
Against this, the union's lawyers pointed out that in other legisla-
tion, enacted after the Ontario Court of Appeal's decision in this case,
the government had put into the statute a provision that gave the millis-
ter the authority to appoint an arbitrator who "has no previous experi-
ence as an arbitrator," "has not previously been or is not recognized as a
person mutually acceptable to both trade unions and employers," and "is
not a member of a class of persons which has been or is recognized as
comprising individuals who are mutually acceptable to both trade unions
and employers." 24 While the minister said this subsequent legislation was

21
Id., at 140-42.
22
I deal with Roncarelli in detail in an article that takes forward the argument of this
chapter: "The Deep Structure of Roncarel/i v. Duplessis" (the Ivan C. Rand Lecture Novem-
ber 2003) U.N.B.L.J. [forthcoming].
23
CUPE, supra, note 15, at paras. 90-92.
24
Back to School Act (I'oronto and Windsor), 2001, S.O. 2001, c. 1, s. 11(4). I have
heard that one of the factums presented to the Supreme Court called this provision the
"orangutan" provision, on the basis that it permitted the government to appoint an orangutan
as an arbitrator.
394 Constitutionalism in the Charter Era

irrelevant, the union argued that it "manifests an explicit legislative

intent to exclude the otherwise crucially relevant criteria of expertise
and general acceptability." The union's view was that the "new legisla-
tion shows the HWAA as the Minister would like it to be, but is not." 25
Justice Binnie did not comment directly on these claims. But he did
say that "courts generally infer'' that "the legislature intended the statu-
tory decision maker to function within the established principles and
constraints of administrative law."26 And he said that when it came to
the adjudication of "interest" disputes, that is, the sorts of disputes at
stake under the HLDAA, "it is particularly important to insist on clear
and unequivocal legislative language before finding a legislative intent
to oust the requirement of impartiality either expressly or by necessary
implication. ''27
In making these two claims, Binnie J. is recognizing that for a
common law judge it takes what I call a substantive privative clause if
the legislature wishes to oust judicial review on the basis of values that
judges consider to be legal fundamentals. As I will now show, the dif-
ference between the substantive privative clause and a general privative
clause is that the latter seeks to undercut the formal aspect of judicial
independence while the former seeks to remove the substantive basis for
which they have independence.
A general privative clause is a provision in a statute that purports to
oust the jurisdiction of the courts to review the decisions of an adminis-
trative tribunal. The general privative clause is directed at the formal
place judges have in the legal order - formal in that without an inde-
pendent institution there would be no check on the legal limits of legis-
lative delegations of authority. Since the 1960s and 1970s, courts in the
common law world have held that a privative clause is largely redundant.

is CUPE, at paras. 81-82.

26 J.d ., at para. 99.
27
Id., at para. 121. Note that these remarks occur in the section of his judgment where
he deals with procedural fairness and the fact that the minister, who has an interest in the
outcome of the arbitration, appoints the arbitrator. But the remarks clearly apply to the part of
his judgment that deals with substantive review since they tell us what it would take for
Binnie J. to find the substantive exercise precluded by the statute. That is, the legislature
would have to go further than stating explicitly that the minister has the power of appoint-
ment. It would have to, as in the Back to School Act, state that that power could be exercised
unconstrained by factors that conditioned an impartial exercise of the power, given that it was
the minister's to exercise.
 The Unwritten Constitution 395
and the Rule ofLaw

They argued that a legislature cannot seriously intend that a tribunal be

delegated legally unlimited authority and thus that judges are constitu-
tionally entitled to review decisions that trench on those limits.28
Courts often concluded that the legislature was sending the same
sort of hands-off message by couching grants of discretion subjectively,
that is, using the kind of "if satisfied that" language found in the grant of
discretion to appoint arbitrators in the HLDAA. 29 Such a grant of author-
ity, even one which says that the "decision will be within the sole and
unfettered discretion of the minister," might seek to deliver the same
general hands-off message to judges, but it does not do so by saying "do
not review." Rather it can be interpreted to say that the court is not enti-
tled to enquire into the grounds of the decision. Thus courts did not find
this kind of message as constitutionally obnoxious as the general priva-
tive clause, as they claimed they could still check the legality of the
decision, at least to see whether the official had acted in bad faith.
This claim was always rather hollow in the absence of other re-
quirements. But it has been long recognized that in general those subject
to a decision should be given a hearing at which they can learn the basis

28
In Canada, the courts had a constitutional peg on which to hang their argument, in
s. 96 of the Constitution Act, 1867 [(U.K.), 30 & 31 Viet., c. 3, reprinted in R.S.C. 1985, App.
II., No. 5], which resetves the power of appointment of superior court judges to the federal
government. My own view of s. 96 is that it is not only a slim peg, but that if it had not
existed the course of Canadian constitutional history in respect of jurisprudence on judicial
independence might have been little different. Without it, the Supreme Court would have
simply asserted an authority on the basis of its inherent common law jurisdiction to maintain
the rule of law. Put differently, s. 96 is just a convenient basis for the judges' sense that they
are the ultimate guardians of the principles of the unwritten constitution. Note that the House
of Lords seems to have asserted an even more extensive, constitutionally protected review
authority against the United Kingdom Parliament without ever referring to the Act Qf Settle·
ment [(U.K.), 12 & 13 Will. 3, c. 2]. The more extensive review arises from the gloss placed
by the House of Lords in R. v. Lord President of the Privy Council, ex parte Page, [ 1993]
A.C. 682 onAnisminic Ltd. v. Foreign Compensation Commission, [1969] 2 A.C. 147. While
the latter could be interpreted as holding that a preclusive or privative clause could not tota11y
exclude judicial review, the former has been interpreted as emptying such clauses of all
meaning.
29
Indeed, the HLDAA contained a double whammy as s. 7 was a privative clause pre·
eluding any application for judicial review once an arbitrator had been appointed, so that it
would be "presumed conclusively that the board has been established in accordance with this
Act." Both the majority and the dissent took the combination of privative clause and subjec·
tive conferral of discretion to indicate that the standard of review was the least intrusive
patent unreasonableness standard. See the dissent in CUPE, supra, note 15, at para. 17 and
Binnie J., at para. 151.
396 Constitutionalism in the Charter Era

of the official's view as to how the matter should be disposed and an

opportunity to contest that basis. That requirement - the audi alteram
partem rule - did not seem to judges to negate the delegation of discre-
tionary authority to the official, since the entire substance of the deci~
sion was left to the official. In other words, the requirement was firmly
on the process side of the distinction between process and substance.
But if all that requirement imposes is a show of presenting a case
and listening to the subject of the decision, the claim about legality still
looks hollow. Unsurprisingly, judges developed more intrusive tests for
legality, though they tried to frame these negatively, as having to do
with factors officials could not take into account, ''irrelevant factors," or
"improper purposes," thus seeking to avoid the impression that they
were telling the official how to decide. Indeed, these negative tests stay
very close to the idea of bad faith, as is illustrated by Roncarelli. If the
official acted for such and such a reason, his decision would be illegal
because acting on that reason disclosed bad faith or something close
to it.
However, in the absence of a duty to give reasons for the decision,
and, moreover reasons that go some distance to showing how the offi-
cial justified his decision in the light of all the factors, the claim still
seems hollow. So it was a kind of logical development of the view that
the requirements of legality are not hollow, that judges came to articu-
late a general duty to give reasons, and tests for the adequacy of those
reasons which permit the judge to ask whether the reasons provide a
reasonable justification of the decision. And with these developments
comes a sense that in order for some decisions to receive an adequate
justification, there are not only factors that the decision maker must not
take into account, but also factors that he must take into account -
considerations that are not only relevant, but mandatory.
Thus in a decision that is Justice Binnie' s other principal point of
reference, Baker v. Canada (Minister of Citizenship & Immigration),30
the Supreme Court decided that an immigration official who had to
decide whether a deportation order should be stayed on "humanitarian
and compassionate" grounds had to give substantial weight to the inter-
ests of the deportee's children. Moreover, it is in this decision that the
highest court of a common law country has for the first time articulated

30
[1999] 2 S.C.R. 817 [hereinafter "Baker"].
 The Unwritten Constitution 397
and the Rule ofLaw

a general duty to give reasons, which will then be assessed on the ap-
propriate standard. So the "perspective within which a statute is in-
tended to operate,, is constituted not only by the statute. As Madame
Justice L'Heureux-Dube put it in Baker, discretion must still be "exer-
cised in accordance with the boundaries imposed in the statute, the prin-
ciples of the rule of law, the principles of administrative law, the
fundamental values of Canadian society, and the principles of the Char-
ter."31 In making these claims she consciously elided what had hitherto
been seen as two quite distinct categories of review - review of official
interpretations of the law and review of discretionary decisions - argu-
ing that the latter required officials to interpret the law of their statutory
mandate in much the same fashion as the former.
In sum, the flip side of the constitutional disability of the legislature
to preclude judicial review and thus judges' formal independence in the
constitutional order, is that judges will and should use their independ-
ence in the service of the fundamental values of that legal order. The
substantive privative clause does not then prohibit review. Rather, it
seeks explicitly to remove grounds for review, for example, by saying
that the courts may not review even if a decision is unreasonable, bi-
ased, or made in bad faith. Only the latter kind of message serves to
exclude the constitutional rule of law approach that in my view under-
pins Binnie J. 's reasoning. The general privative clause or subjectively
couched grant of discretion has to be reinterpreted, so that there is an
independent check on the executive to ensure that it acts within the
limits of its authority. But those limits are not formal; they are substan-
tive in that they are set by fundamental or constitutional legal values.
Judicial independence is not then a good in itself but instrumental in
that it serves those values. If there is no written bill of rights in place
that protects those values, then the supreme legislature will have the
authority to override those values using a substantive privative clause -
one which makes it clear that the officials to whom authority is dele-
gated may act outside of the controls of the rule of law. In doing this,
the legislature deprives the judges of the substance of their constitution-
ally unassailable independence. But while they cannot exercise their
independence, this deprivation does not make it futile. They can still use
their independence to point out in their judgments the lawlessness of the

31
Id., at para. 56.
 398 Constitutionalism in the Charter Era

regime the government has used, even abused, the legislative form to
create. And that means that the government bears the costs of political
accountability.
Now the obvious problem for this view of judicial independence is
that it might well seem that we get rule by judges in place of the rule by
the delegates of the people in accordance with the intention of the legis-
lature. Judges will simply substitute their decisions for the delegates. In
Canada, judges deny that they are in the substitution game. They have
developed a sophisticated jurisprudence for substantive review accord-
ing to which administrative decisions will be reviewed on a continuum
of standards, ranging from the most intrusive correctness standard or
substitution on the merits, through "reasonableness simpliciter,'' to the
- least intrusive standard of patent or manifest unreasonableness.
The denial might, however, seem unconvincing the more sophisti-
cated the jurisprudence of deference is. In particular, the idea of a stan-
dard in between correctness and patent unreasonableness seems to put
judges on dangerous ground. The legitimacy of adjudication in this area
rests on a distinction between review and appeal. It might seem that one
can preserve this distinction by establishing a set of limited questions
which require the correctness standard while using the absurdity stan-
dard for all other questions. But reasonableness, the standard which the
Court said was appropriate in Baker given the importance of the deci-
sion to Baker's future, requires judges to test the adequacy of the rea-
sons for the decision and thus seems to require them to evaluate those
reasons. This is a quite different process from seeing whether there are
no reasons for the decision other than very bad reasons, the manifest
unreasonableness or absurdity standard, and might seem hardly different
from the correctness standard. However, without something more intru-
sive than the patent unreasonableness standard, judges risk appearing
concerned only with upholding the formal aspect of their independent
role in the legal order.
As a result, judges who are minded to uphold rather than to pay
mere lip service to the rule of law sometimes experience severe ten-
sions. For example, the Supreme Court of Canada is now rather preoc-
cupied with the idea that whatever judges do, they should not "reweigh''
the factors officials have to take into account in order to demonstrate
that their decisions are reasonable. Weight is, however, just a metaphor
for a proper inquiry into the balance of reasons. It became part of the
Canadian discussion because in Baker the majority was clearly influ-
 The Unwritten Constitution 399
and the Rule ofLaw

enced by the fact that Canada had ratified, though had not incorporated
by legislation, the Convention of the Rights of the Child, 32 which in
Article 3 requires that in administrative decisions affecting children, the
"best interests" of the children be "a primary consideration.''
The partial dissent to this decision claimed to object only to this as-
pect of the majority's reasoning and put the objection on classic dualist
or positivist grounds - if the Charter33 is not directly involved, parlia-
ment is the sole source of legal value. Thus the dissent claimed not to
object to the majority's holding that the statute itself, as well as ministe-
rial regulations, required that the children's interests be given "substan-
tial weight," nor that judges should check to ensure that officials had
been "alert, alive and sensitive to" the issue of whether appropriate
weight had been given.34
But in the Court below - the Federal Court of Appeal - which
upheld the decision to deport, Strayer J. was clear that the most that a
judge can do is check whether a relevant factor like the children's inter-
ests has been taken into account. For a court to evaluate how that factor
was taken into account is to reweigh, which is illegitimate. 35 Since
Baker, the Supreme Court has retreated from its position expressed there
and has adopted the view, more like that of the Federal Court of Appeal,
that judges must never evaluate the way that relevant factors figure in
the official's reasoning. They can check that the right reasons were
taken into account, but may not go into the balance of reasons, which is
to say, reweigh the reasons. It is hardly an accident that this apparent
retreat from Baker took place in the first major decision in the national
security area given by the Supreme Court after September 11, 2001,
Suresh v. Canada (Minister of Citizenship and lmmigration). 36
In his judgment in CUPE, Binnie J. found that the appropriate stan-
dard of review was patent unreasonableness and that, following Suresh,
the Court was not entitled to reweigh the factors the minister had to take

32
U.N. G.A. Res. 44125 (1989).
33
Canadian Charter of Rights and Freedoms, Part I of the Constitution Act, 1982, be-
ing Schedule B to the Canada Act 1982 (U.K.), 1982, c. 11.
34
Baker, supra, note 30, at para. 75,per L'Heureux-Dube J.
35
[1997] 2 F.C. 127, at paras. 12, 31.
36
[2002] I S.C.R. 3. For comment, see Mullan, ..Deference from Baker to Suresh and
Beyond - Interpreting the Conflicting Signals" in Dyzenhaus, ed., The Unity of Public Law
(2004), at 21.
400 Constitutiona/ism in the Charter Era

into account. However, he also said that while the Court could not re-
weigh the factors, it was
entitled to have regard to the importance of the factors that have been
excluded altogether from consideration. Not every relevant factor
excluded by the Minister from his consideration will be fatal under the
patent unreasonableness standard. The problem here . . . is that the
Minister expressly excluded factors that were not onlr relevant but went
3
straight to the heart of the HLDAA legislative scheme.

In my view, Binnie J. 's claim that he was not involved in the illicit
practice of reweighing has to be treated with some skepticism. In CUPE
he used language taken directly from Baker to describe what had gone
wrong: the Court had "look[ed] in vain for some indication in the record
that the Minister was alive to these labour relations requirements; "38 In
its place, said Binnie J., there was an "active disclaimer of any such
requirement, by the Minister's senior advisor charged with the search
for retired judges, who made it clear in his cross-examination the Minis-
ter's rejection of both expertise and broad acceptability as qualifica-
tions."39 But if Binnie J. were right that there was an active disclaimer,
then this means that the minister in some sense weighed the require-
ments against his political objectives and found them less weighty.
Similarly, in both the Federal Court Trial Division and the Federal
Court of Appeal in Baker, the judges found that the immigration offi-
cials had weighed the children's interests because they had taken into
account that Baker had children. They thus seemed to think that the
officials had taken Baker's Canadian-born children's interests into ac-
count because the officials thought of the existence of the children as a
kind of aggravating circumstance or reason to get rid of Baker. This is
of course a rather Pickwickian sense of taking interests into account.
Since the children had interests and since the children were considered,
therefore their interests must have been considered too. And it can be
justified only by the thought that for the judges to decide how the inter-
ests should be considered involves the illicit practice of assigning the
interests weight.

37
CUPE, supra, note 15, at para. 176.
38
Id, at para. 181 (emphasis added).
3!>
Id., at para. 182.
 The Unwritten Constitution 401
and the Rule ofLaw

I suspect that the dissent in CUPE was largely fuelled by the same
stance that drove the Federal Court in Baker: as long as relevant factors
are taken into account, how they are to be taken into account is up to the
official. 40 It seems to follow that the only way to understand Binnie J. 's
judgment is that he assigned certain factors a weight and found this
weight not manifested in the balance of reasons. So much seems to be
recognized in his statement that the correctness standard has something
in common with the patent unreasonableness standard. "A correctness
approach means that there is only one proper answer. A patently unrea-
sonable one means that there could have been many appropriate an-
swers, but not the one reached by the decision maker.,,41
What one sees then is that judges experience certain tensions as they
seek to give content to their constitutional standing as the independent
guardians of the rule of law. On the one hand, if they fail to give the rule
of law substantive content, they will appear to be more concerned with
upholding their sense of role than with doing the job that explains why
they should have that role. On the other hand, as they give the rule of
law content, so they run the risk of appearing to usurp the legislative
role, whether this be by their claim to discern fundamental legal values
that stand free of any legislative texts and which control the legislature
or by their use of the same interpretative process to find controls on the
authority of the administrative delegates of the legislature.
So my point is not to deny that judges will face serious tensions in
this area. It is not insignificant that in CUPE three judges of the Su-
preme Court, including the Chief Justice, disagreed with the majority on
the basis that the majority's decision was too interventionist. Indeed,
this section of my article is intended in part to illustrate that such ten-
sions are inevitable. But this should be no comfort to the positivists,
who also face tensions, as is illustrated by returning to Scalia J.
In a well known article, Scalia J. comments on the American doc-
trine of deference to administrative officials, the "Chevron doctrine/'

40
See the dissent, id., paras. 35-36, relying heavily on Suresh. Indeed, the dissent in
CUPE tracks another feature of the Federal Court in Baker - that is, some doubt is expressed
about whether factors that are not expressly stated in the statute to be relevant can be consid-
ered so.
41
Id., at para. 164.
402 Constitutionalism in the Charter Era

which consists of two steps.42 To begin with, the court must determine
whether Congress had a ',;clear" and "unambiguously expressed" intent
when enacting the statute in question. If the court finds that Congress
did have such an intent, that is "the end of the matter" and the court has
no authority to modify or interfere with the interpretation or implemen-
tation of the statute. However, if no such intent can be discovered, the
court must determine whether the administrative agency came to its
decision on the basis of a "permissible construction of the statute.'''0
Scalia supports the Chevron doctrine - the introduction of "an across-
the-board presumption that, in the case of ambiguity, agency discretion
is meant. " 44 But he does not do so on grounds to do with agency exper-
tise, nor with the separation of powers and the inappropriateness of
judges deciding policy issues. In respect of.expertise, he says that if it
were true that officials were better situated to determine the pwpose of
legislation than judges, this would constitute "a good practical reason
for accepting the agency's view, but hardly a valid theoretical justifica-
tion for doing so." In respect of separation of powers, he argues that the
courts are constantly in the business of determining policy, especially
when it comes to working out what is the intention or range of permissi-
ble intentions that can be attributed to a statute, so that this task cannot
be reserved to the administration.45
Instead, his approval of Chevron is based on the rise of the modem
administrative state. The kind of statute-by-statute assessment that was
common prior to Chevron was becoming increasingly difficult to im-
plement given the complexity of present-day administrative decision
making. Jn addition, he contends that in the majority of cases Congress
does not have a "clear" intention and it does not mean to provide an
agency with discretionary powers. Instead, it simply fails to consider the
matter. Because of this, Chevron is "unquestionably better" than that
which preceded it. Not only does Congress now know that statutory
ambiguities will be resolved by agencies rather than courts, but these
agencies will be able to deal with them with sufficient flexibility to
ensure that their decisions are not "eternal" or "immutable." Indeed, he

42
Scalia, ••Judicial Deference to Administrative Intepretations of Law'' (1989) Duke
L.J. 511, analyzing Chevron, U.S.A., Inc. v. NRDC, 467 U.S. 837 (1984).
43
Here I rely on the quotations from Scalia, id., at 511-12.
44
Id, at 516.
45
Id., at 514-16.
 The Unwritten Constitution 403
and the Rule ofLaw

argues that one of the great benefits of Chevron is that it accords agen-
cies the space to alter their interpretations and approaches in the light of
changing conditions.46
Justice Scalia' s view of the proper role of agencies is very much the
Bentharnite picture of appropriate adjudication. Officials who are
charged with interpreting the law have wide discretion about how to
apply the law and wide discretion when it comes to interpreting the law
when the content of the law is indeterminate or ambiguous. But when it
comes to the second activity, the officials' decisions are not to have any
precedential force, lest these come to be regarded as a constraint on the
discretion of officials in the future.
However, Scalia J. still has to make sense of his own role, qua
judge. Here it is w01th quoting at some length the link he draws between
one's "method" of interpreting statutory and constitutional documents
and one's definition of"clear" in the first step of Chevron.

In my experience, there is a fairly close correlation between the degree to

which a person is (for want of a better word) a "strict constructionist" of
statutes, and the degree to which that person favors Chevron and is willing
to give it broad scope. The reason is obvious. One who finds more often
(as I do) that the meaning of a statute is apparent from its text and from its
relationship with other laws, thereby finds less often that the triggering
requirement for Chevron deference exists. It is thus relatively rare that
Chevron will require me to accept an interpretation which, though
reasonable, I would not personally adopt. Contrariwise, one who abhors a
"plain meaning" rule, and is willing to permit the apparent meaning of a
statute to be impeached by the legislative history, will more frequently
find agency-liberating ambiguity, and will discern a much broader range
of "reasonable" interpretation that the agency may adopt and to which the
courts must pay deference. The frequency with which Chevron will
require that judge to accept an interpretation he thinks is wrong is
infinitely greater.47

Justice Scalia's positivism thus draws him to the view that his tests
for statutory meaning are likely to come up with a plain meaning of the
statute and that once that meaning has been determined there is no rea-
son for the judge to defer. Since, as he argues elsewhere in the same
article, it is rare that a judge, whatever his interpretative approach, will

46
Jd.,at516-17.
47 ',/
Ju., at 521.
 404 Constitutionalism in the Charter Era

find that on his approach there is in fact "equipoise'' between conflicting

interpretations, one can infer that generally Scalia J. will find no reason
for deference.48 However, I think we can assume he would be required
to find that "if satisfied thaf' delegations of discretion are unambiguous
delegations of authority to officials, which the legislature did not intend
to have constrained other than by those constraints explicitly stated in
the statute. The tension he encounters then arises out of his view of the
rule of law as the rule of a system of statute-based rules with determi-
nate content. It arises because that view requires, on the one hand, that
when the statute imposes constraints, these rigidly constrain officials in
accordance with the judges' understanding of the correct interpretation
of the law. On the other hand, it also requires that when that kind of
constraint does not exist, officials are accorded a more or less free-
wheeling discretion - they are a law unto themselves.
Justice Binnie's approach contests both aspects of this view and
. does so moreover in a way that is not best described as the product of
the mindset of one who "abhors a 'plain meaning' rule, and is willing to
permit the apparent meaning of a statute to be impeached by the legisla-
tive history,, and who is thus prone "more frequently [to] find agency-
liberating ambiguity." The issue is not abhorrence of the idea that there
is plain meaning, but rather adoption of a regulative assumption that
Parliament, the executive, and judges are committed to a rule of law
project which is about the realization of fundamental constitutional
values, whether written or unwritten. Judges should thus try to find that
legislation is legislation that seeks to achieve its particular objectives in
the light of a wider legal project. Thus legislative meaning is not a top-
down communication - a "one way projection of authority,n as Lon
Fuller described the positivist view. Rather, as Fuller preferred to put it,
law is the product of a relation of reciprocity between ruler and ruled. 49
A corollary of the view of law as a product of a value-based rule of law
project is that no particular institution in legal order has a monopoly on
the best understanding of law and that is why judges have reason to
defer to administrative interpretations of the law of the particular admin-
istrative mandate. But they should defer only if the officials do a rea-
sonable job of justifying their interpretation of the law or decision.

48
Id., at 520.
49
Fuller, The Morality ofLaw (1969, revised ed.), at 207.
 The Unwritten Constitution 405
and the Rule o.flaw

This idea of deference, which I have termed "deference as respect"

in order to distinguish it from the kind of submissive deference to legis-
lative command which is the positivist understanding of deference, was
adopted by the Supreme Court of Canada in Baker as an appropriate
way to characterize the relationship between judges and administrative
officials. 50 Justice Scalia is well aware of this kind of idea, though he
describes it in rather harsh terms as "mealy-mouthed" deference, which
does "not necessarily mean anything more than considering those views
with attentiveness and profound respect, before we reject them."51 And
he goes on to say that if one were to try to give more force to this idea of
deference, if those views would be binding if they were judged reason-
able, the result would be a "striking abdication of judicial responsibil-
ity."s2
Justice Scalia's claim increases the tension just identified in his po-
sition when a judge is faced with a general privative clause. The priva-
tive clause is an unambiguous statutory command not to review
administrative interpretations of the law. In the face of such a clause, a
positivist judge like Scalia J. is in a dilemma. Either he can interpret the
command in accordance with his textualist approach and say that this
official is a law unto himself, even when it comes to interpreting the
law, or he can say this particular command of the legislature can not be
taken seriously.
In fact, judges of a positivist inclination have chosen the second op-
tion, along with their more common law inclined colleagues. The differ-
ence between the two groups is that the positivist judges then find that
they can review on a correctness standard. In contrast, the others try to
make sense of the clause as an indication of the degree of deference the
legislature intended. For they are committed to showing that the legisla-
ture and the administration are involved in an endeavour to maintain the
rule of law. In other words, the positivist judge thinks that the only kind
of deference that is worth the name is submissive deference, submitting
to the command of the legislature. When the command contradicts his
understanding of his role as a judge, he finds that he can ignore that

50
Baker, supra, note 29, at para. 65, referring to Dyzenhaus, "The Politics of Defer-
ence: Judicial Review and Democracy" in Taggart, ed., The Province of Administrative Law
(1997Ji 279.
1
Scalia, supra, note 39, at 514.
52
Scalia, supra, note 39, at 513·14.
406 Constitutiona/ism in the Charter Era

command altogether. A substantive privative clause, however, presents

no problem to him at all All such a clause does is make explicit what
judges took subjective grants of discretion to command implicitly -
that official authority is unconstrained by the values of the rule oflaw. 53
At least, the substantive privative clause should create no tension
from the theoretical vantage point of a positivist judge. But qua judge,
positivists have never been able to say that discretion is entirely free-
wheeling. They have always said that at least discretion must be exer-
cised in good faith, etc. And I suspect that an explicit announcement by
the legislature that an official might act in bad faith, or with bias, or
unreasonably, disquiets all judges. But what that disquiet shows is that
positivist judges fmd that a commitment to the rule of law demands
more of them than their legal theory pennits. The requirement - under-
stood as a requirement of legality - that decisions not be in bad faith,
etc. is the thin end of the wedge of the common law's conception of the
.~le of law. As my description of the path that led to Baker indicates, if
such requirements are to be worth anything, if they are to provide stan-
dards to which officials are genuinely accountable, judges will find
themselves engaged in the kind of value based review which positivist
judges shun. Indeed, as I will now argue, it is not clear that this stance of
shunning value based review is an authentic option within a common
law legal order, or, for that matter, any order governed by the rule of
law.54

IV. THE VERY IDEA OF LEGAL ORDER

Jeremy Waldron has recently argued that the idea of the rule of law
should be understood as an instance of what Gallie called an "essentially
contested concept."55 Waldron takes the key to essential contestability to

53
To create a tension for such a judge the statute would have to both command officials
to decide in accordance with a set of explicitly stated rule of law values and command judges
to avoid checking to see if the officials had so decided.
54
Will Waluchow objected that I tend to slide between ''the legal order" and "legal order''
in my argument. He is right but the slide is deliberate. It is intended to reflect the fact, on my
Fullerian view, that any particular legal order aspires necessarily to be an ideal legal order.
55
Jeremy Waldron, "Is the Rule of Law an Essentially Contested Concept (in Flor-
ida)?" (2002) 21 Law & Philosophy 137, at 148-53, referring to Gallie, "Essentially Con-
tested Concepts" (1955-56) 56 Proceedings of the Aristotlean Society 167.
 The Unwritten Constitution 407
and the Rule oflaw

reside in a combination of normativity and complexity such that debate

over the idea requires participation in the contes~ one where the partici-
pants try to capture the elusive ideal to which the rule of law aspires.
That the idea is essentially contested should not, he thinks, lead to pes-
simism about participation. Even if participation cannot achieve agree-
ment, participants in the contest can still have the hope that the
"contestation - and the sense of the underlying ideal at subsequent
stages - will be the better for one's intervention. " 56
Waldron detects three levels of complexity when it comes to debate
about the normative ideal of the rule of law, which is, he takes it, to
have law in charge rather than men. 57 There is the level at which the list
of criteria or principles of the rule of law are set out, for example,
Fuller's eight principles of what he called the "internal morality of law."
Here Waldron does not think that much of a contest emerges. Contro-
versy really begins at the second level, when the list is related to the
theorist's particular conception of how the list amounts to the law being
in charge, how it responds to the challenge to "make law rule, rather
than men. " 511 The third level comes about because at issue in the contest
is not only what the ideal of the rule of law consists in, for example, the
positivist ideal of the rule of rules or the common law idea of the rule of
principles. The contest is also about the values the rule of law serves, for
example, liberty or fairness or human dignity, or even about whether the
rule of law serves values at all, since some argue that the rule of law is
merely the instrument of whatever values the powerful desire it to pro-
mote. 59
I want to bring out a feature of essential contestability that is, I
think, implicit in this account. Assume that there are only two groups of
judges competing within the practice of the rule of law: the democratic
legal positivists and the common lawyers. They will debate their differ-
ent conceptions of the rule of law. But their debate will not take place
only or even mainly at a theoretical level. It is a deeply practical debate
both because it is focused on the resolution of particular disputes and
because the resolutions have consequences for future practice. The prac-

56
Jd., at 153.
51
id., at 153-58.
58
Id., at 157.
59
Id., at 158-59.
408 Constitutionalism in the Charter Era

tice of law and the content of the rule of law are at stake in this contest.
Law will be different to a significant extent depending on which group
has the upper hand.
However, if this practical debate is nicely captured by the idea of
essential contestability, it must be the case that at best one group can
dominate for a time - there is no complete victory. Put differently,
there must be something about law itself that will resist complete cap-
ture of the practice by either group. The democratic legal positivists will
find that they cannot reduce law to the rule of rules - principles and
values are at least in some very minimal way necessary to make sense of
legal order. Similarly, the common lawyers will find that not all statutes
are susceptible to being understood in the light of the values they take to
be the constitutional values of legal order.
My description of the evolution of the common law of judicial re-
view supports this feature of debate about the rule of law. In favour of
the common law position is that positivist judges suppose that there are
values of legality that control official decision making even in the ab-
sence of a written constitution and explicit statutory command. In fa-
vour of the positivist position is that in the absence of a written
constitution common law judges will submit to an explicit legislative
command to exclude the values of legality, except in the case where
they (and positivist judges) will refuse to understand literally a general
privative clause.
If this is right, then I think there is at least one modest victory, that
is, at the level of theory. We seem to have established that for a gov-
ernment to rule through the medium of law is for it to submit to rule by
the values of legality, as these are interpreted by judges. It follows that
one view of the rule of law, the kind of positivism that says that the rule
of law is a mere instrument of those who happen to be in power, is
wrong.
It is, I think, very important in thls regard to notice that Thomas
Hobbes, the reputed founder of this kind of positivism, was in fact far
from holding this view. As I have argued elsewhere,60 the tension in
Hobbes's position is evidence of the fact that someone who takes the
idea of the rule of law seriously will find it difficult to stick with the

60
Dyzenhaus, "Hobbes and the Legitimacy of Law" (2001) 20 Law & Philosophy 461.
 The Unwritten Constitution 409
and the Rule ofLaw

instrumentalist accounts of the rule of law which contemporary legal

positivists support.
There is, of course, no doubting Hobbes's hostility to the common
law tradition and his desire to deprive judges of their authority to issue
judgments whose force extended beyond the particular case. But he also
suggests that a judge insults the sovereign lawmaker and is in derelic-
tion of duty when the judge attributes an intention to the lawmaker to
cause inequity. He clearly regards it as the judicial duty to bend over
backwards to find a reasonable interpretation of legislation, one that will
bring about equity rather than inequity. The basis for such interpretation
is, he says, the judicial sense of natural law. A careful read of his cata-
logue of the laws of nature shows that he took them to include many of
the principles that the common law of judicial review claims to be the
basis of the rule of law. Further, whenjudges engage in this kind of rule
of law interpretation, Hobbes does not understand them as imposing the
controls against the sovereign; rather, they are completing the exercise
of sovereignty. Hobbes generally regards the legislature, the administra-
tion, and the judiciary as involved in a partnership whose aim is to make
concrete the laws of nature in the civil law to which individuals within
the scope of the sovereign's authority are subject.61 And it is only this
idea of partnership that can support his claim about the role of equity in
interpretation of the law.
Hobbes then wants the common law's understanding of the rule of
law without the common law. Moreover, he gives judges a significant
role in upholding the rule of law, even to the extent of placing them
under a duty to read legislation against an interpretative backdrop of the
values of the rule of law. And the duty is not grounded in the tacit con-
sent of the sovereign, in the sovereign's after the fact failure to override
the judge's interpretation by statute. Rather it is grounded in the values
that make legal order a legitimate project that both the sovereign as
supreme judge and his staff of officials are bound to bring to realization.
It is true that a sovereign has the power to override his judges and can
thus enact positive law that contradicts the requirements of the laws of
nature. But I think that an explicit element of Fuller's legal theory is

61 While there is overwhelming textual evidence to support this claim, it is most strikM
ingly advanced in the Introduction to Leviathan. See Hobbes, Leviathan, MacPherson, ed.
(1985), at 81-83.
410 Constitutionalism in the Charter Era

implicit in Hobbes - the more the sovereign engages in this exercise,

the more his claim to govern through law is put under strain and the less
his power will appear legitimate.62
The idea of strain must be emphasized here because of a tendency in
legal and constitutional theory to suppose that if there are fundamental
or constitutional values of legality, there must be an ultimate guardian of
these values - judges who have the authority to overrule the legisla-
ture. But it is surely a sufficient test for the fundamental nature of such a
value that a ruler who deliberately flouts it puts in doubt his claim to
rule through law. That nearly all democratic legal positivists object to
judges relying on interpretations of the fundamental values of the un-
written constitution for review shows, I thlnk, an appreciation of this
point. These positivists do not take comfort from the fact that the legis-
lature can and has always been able to overrule judges' understanding of
the substance of the rule of common law values. They rightly under-
stand that a concession that there are any intrinsic values of legality that
judges legitimately interpret opens the door to expansion of the list
through both extrapolation and interpolation.63
·Hobbes was not of course a democratic legal positivist. It was Ben-
tham who attempted to expurgate legal order of values so as to clear the
lines of legal communication between a democratic assembly and the
people. It is Bentham, not Hobbes, who puts forward the instrumental
view of the rule of law, more recently articulated in a depoliticized ver-
sion by Joseph Raz: the only internal constraints on law are those that
make law more effective in communicating the judgment of the com-
mander to his subjects. 64 Raz's attempt to tum Fuller's list of principles
of legality into criteria of efficacy or instrumentality follows HLA
Harf s account of the rule of recognition in trying to make sense of the
idea that the commander is not uncommanded - he too is subject to
law - while avoiding any claim that might be the basis for an argument
about the intrinsic values of legal order.

62
For innovative exploration of this idea, see Evan Fox-Decent, ..Sovereignty's Prom-
ise: The State as Fiduciary" (Ph.D. diss., University ofToronto, 2003).
63
In "Aspiring to the Rule of Law," Campbell et al., supra, note 14, at 195, I argue that
Goldsworthy's attempt in "Judicial Review, Legislative Override, and Democracy' to move
away from this position undermines democratic legal positivism.
64
Raz, "The Rule of Law and its Virtue" in Raz, ed., The Authority ofLaw: Essays on
Law and Mora/tty (1979) 210.
 The Unwritten Constitution 411
and the Rule ofLaw

But this avoidance makes sense only if it is put on explicitly politi-

cal, Benthamite grounds about the best way to design a legal order,
which is to make it not only statute based but free of the common law.
In the absence of these political grounds, the three levels of debate about
the rule of law which Waldron identifies are detached from each other
with the result that there is no satisfactory engagement with the question
of how law is best understood to be "in charge." In other words, the
instrumental view of the rule of law is a radical reform proposal for
legal order. The values it serves have to do with a particular conception
of another essentially contested concept - democracy. And it serves
those values by requiring that judges understand law as free of values
other than those the legislature has seen fit to inject into law.
In the context of this article, there are two interesting results from
the observation that democratic legal positivism is best understood as a
radical reform proposal. The first is noted, though not in these terms, by
Mary Ann Glendon in her comment on Scalia J.'s paper in A Matter of
Interpretation. She points out that in the civil law systems Scalia J.
admires there is a pull to common law reasoning in order to make sense
of the idea of the rule of law in the era of the administrative state. 65 If
this pull is inevitable, it would be evidence of the fact that legal order is
inherently resistant to reform along Benthamite lines. It is difficult, if
not impossible, for judges to make sense of their commitment to legality
or the rule of law other than by thinking of it in terms of some set of
intrinsic legal values that gives content to the idea that the law is in
charge.
The second result is for positivist judges in common law legal or-
ders that have not been reformed in this way, even though these orders

65
See Glendon, "Comment," in A Matter of Interpretation, supra, note 5, at 95, 102-
103. This pull to a common law methodology might well be evidence of a flaw in the dream
of a completely codified system, detected by Gerald Postema in Bentham and the Common
Law Tradition (1986) 453-59. Postema argues that Bentham's requirement of publicity for
judges' reasons for decision will focus the attention of those subject to the law as well as of
other judges on the reasons. Once public reasons for decisions are focused upon, and not the
fact that the issue has been resolved, those subject to the law will come to expect that the
reasons articulated for a decision on one problem of interpretation will fofluence judges who
decide similar cases. Judges will then have to take into account the utility served by not
disappointing expectations about how cases will be decided. Hence, an informal doctrine of
precedent will arise, even if, as Bentham wanted, judgments are deprived by the constitution
of precedential force.
412 Constitutionalism in the Charter Era

have been greatly changed by the statutory regimes that created the
administrative state. Such judges find themselves working within a
space structured by the common law tradition. In that space, they have
to cope, however reluctantly, with more than the fact that their judg-
ments have legal force beyond the case; they also have to cope with the
fact of precedents that set out a vision of the rule of law not as the rule
of rules but as the rule of fundamental values. As Scalia J.'s legal theory
shows, the tendency of such judges will be to try to positivise their role.
In so doing, they adopt a role that was designed for a legal order other
than the one in which they work. They thus fail to do their duty as
judges to uphold the rule of law, a duty which is the rationale of their
constitutionally guaranteed independence.
The disagreement between these two approaches is at base one
about the force of practices of reason giving and it gives rise, as I have
argued elsewhere, to two different conceptions of legal culture - the
culture of reflection and the culture of justification.66 Reflection does not
mean here reflective thought but reflection as mirroring. Legislation
should reflect the preferences of the people and judges must interpret
legislation so as to live up to this conception - law with the detenni-
nate content the people's representatives in fact intended it to have.
In contrast, within the culture of justification law is legitimate when
its content is determined through a process of reasoned justification,
which takes into account the fundamental legal or constitutional values
of a society. Whether a decision about the law deserves respect thus
depends not on whether one would have given that decision oneself, but
on whether the person who gave it is able to provide an adequate justi£-
cation. To a large extent, any talk about deference in terms of the degree
of divergence between what one thinks is correct and the actual decision
is inappropriate, perhaps more suited to positivism. The issue is ade-
quate justification not divergence. Of course, judges and others will
differ about whether a justification is adequate or not, as did the judges
in CUPE. But the claim that such differences are antithetical to the rule
of law stems from a positivist mindset which refuses to countenance the
space of legality that exists between the certainty of the rule of rules and
the arbitrary rule of men.

66
Dyzenhaus, ''The Justice of the Common Law: Judges, Democracy and the Limits of
the Rule of Law" in Saunders and Le Roy, eds., The Rule ofLaw (2003) 21.