McGill Law Journal ~ Revue de droit de McGill
DEMOCRATIZING COMMON LAW
CONSTITUTIONALISM
Evan Fox-Decent*
Common law constitutionalism is the the-
ory that legal principles such as fairness and
equality reside within the common law, are con-
stitutive of legality, and inform (or should in-
form) statutory interpretation on judicial re-
view. This article looks to Justice Rands judg-
ment in Roncarelli v. Duplessis to develop a
democratic and relational conception of common
law constitutionalism. By democratic the au-
thor means a version of the theory that governs
judicial review but which is available to front-
line decision makers independently of the his-
tory and contemporary practice of review. By
relational the author means a theory that pre-
supposes a trust-like and legally significant re-
lationship between public authorities and the
persons subject to their power.
Under the democratic and relational the-
ory, the legality of administrative action is as-
sessed in light of legal principles constitutive of
the trust-like relationship and without refer-
ence to the separation of powers. These princi-
ples flow from the trust-like nature of the rela-
tionship and the implications of working out
how public authorities can hold discretionary
power over individuals without subjecting them
to domination or instrumentalization.
Le constitutionnalisme de common law est
la thorie selon laquelle les principes juridiques
tels que lquit et lgalit rsident dans la
common law, sont constitutifs de la lgalit et
guident (ou devraient guider) linterprtation
des lois lors du contrle judiciaire. Cet article se
base sur le jugement du juge Rand dans laffaire
Roncarelli c. Duplessis pour dvelopper une
conception dmocratique et relationnelle du
constitutionnalisme de common law. Par d-
mocratique , lauteur entend une version de la
thorie qui gouverne le contrle judiciaire mais
qui est mise la disposition des principaux d-
cideurs indpendamment de lhistoire ou de la
pratique contemporaine du contrle. Par rela-
tionnelle , lauteur entend une thorie qui pr-
suppose une relation quasi-fiduciaire et signifi-
cative dun point de vue juridique entre les au-
torits publiques et les personnes qui sont assu-
jetties leur pouvoir.
Selon la thorie dmocratique et relation-
nelle, la lgalit de laction administrative est
examine la lumire des principes juridiques
constitutifs de la relation quasi-fiduciaire et
sans faire rfrence la sparation des pou-
voirs. Ces principes sont issus non seulement de
la nature quasi-fiduciaire de la relation, mais
aussi de la recherche dune faon pour les auto-
rits publiques de dtenir un pouvoir discr-
tionnaire sur les individus sans toutefois les assu-
jettir la domination ou linstrumentalisation.
* Assistant Professor, Faculty of Law, McGill University. I would like to thank Genevive
Cartier for organizing the commemorative symposium on Roncarelli v. Duplessis at
which this paper was presented, and the symposium participants for stimulating ques-
tions and discussion.
Citation: (2010) 55 McGill L.J. 511 ~ Rfrence : (2010) 55 R.D. McGill 511
Evan Fox-Decent 2010
512 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
Introduction
I.
The Constitution of Discretion
A. The Rights/Privileges Distinction
B. The Separation of Powers and Judicial Activism
C. The Content and Legal Nature of the Obligation
in Roncarelli
D. Normative Independence from Judicial Review
II.
Reframing Common Law Constitutionalism
Conclusion
513
514
521
521
524
525
526
534
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 513
Introduction
Common law constitutionalism is the theory that legal principles such
as fairness and equality reside within the common law, are constitutive of
legality, and inform (or should inform) statutory interpretation on judicial
review.1 Because the principles of the common law are settled through the
gradual accretion of judicial precedents, they are presumed to embody the
deep-seated values of the community. Judges are thus entitled to rely on
those principles when they perform their rule of law duty and read down
statutes to keep the administration in check. Common law constitutional-
ism, in other words, is usually understood as a theory about the rule of
law and the role of judges as the rule of laws guardians.
Justice Rands judgment in Roncarelli v. Duplessis2 has become the
Canadian standard-bearer for the rule of law. His reigning in of Du-
plessiss abuse of discretionary power sits comfortably with the conven-
tional understanding of common law constitutionalism, as depicted above.
Yet we will see below that Justice Rands explicit arguments fail to re-
spond convincingly to Justice Cartwrights dissent. We will also see that
common law constitutionalism is vulnerable to several objections pressed
recently by Thomas Poole.3 The burden of this paper is to show that Jus-
tice Rands judgment presupposes a democratic conception of common law
constitutionalism capable of answering both Justice Cartwrights dissent
and Pooles objections. By democratic I mean a version of the theory that
governs judicial review but which is available to front-line decision mak-
ers independently of the history and contemporary practice of review.
Justice Rands common law constitutionalism, I claim, consists in
more than an aversion to arbitrariness in public decision-making. Implicit
in his judgment is a presumption that arbitrariness has legal significance,
and that presumption relies on viewing public actors and the people sub-
ject to their powers as standing in a legally significant relationship to one
another. In Canadian public law scholarship, the shape of this relation-
ship was first noticed by Roderick Macdonald in 1987. He observed:
When viewed through the eyes of the primary decision-maker,
the act of applying a statutory rule can be understood as a response
to the legislative direction (which may be either explicit or tacit) to
take responsibility for administering the statute by virtue of which
1 See e.g. T.R.S. Allan, Constitutional Justice: A Liberal Theory of the Rule of Law (Ox-
ford: Oxford University Press, 2001) [Allan, Constitutional Justice].
2 [1959] S.C.R. 121, 16 D.L.R. (2d) 689 [Roncarelli cited to S.C.R.]. I assume familiarity
with the facts of the case, as presented in the introduction to this Special Issue:
Genevive Cartier, Lhritage de laffaire Roncarelli v. Duplessis, 19592009 (2010) 55
McGill L.J. 375.
3 Thomas Poole, Constitutional Exceptionalism and the Common Law (2009) 7 Intl J.
Const. L. 247 [Poole, Constitutional Exceptionalism].
514 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
the particular power is exercised. In so responding, the statutory de-
cision-maker acts no differently than a trustee administering a trust
indenture.4
I interpret Macdonald to be saying that legislatively delegated power be-
gets responsibility for statutory interpretation and administration, and
that this responsibility is legally significant (in part at least) because
statutory decision-makers are in a trust-like relationship with the public.
I also interpret him to be saying that this trust-like relationship and its
implications may be apprehended fully through the eyes of the primary
decision-maker,5 and so do not depend on judicial review.
In Part I, I argue that the major elements of Justice Rands reasoning
presuppose Macdonalds relational and democratic conception of public
law. Without the relational and democratic conception in place (or a con-
ception substantially similar to it), Justice Cartwrights approach to dis-
cretion remains a viable alternative defensible on principled grounds. In
Part II, I consider a number of Pooles objections to common law constitu-
tionalism. I suggest that the relational theory offers fresh replies to them,
while taking on board many of Pooles important concerns.
I. The Constitution of Discretion
Discretionary power is a bugbear for liberal legalists. On its face, dis-
cretion stands in tension with the principle of formal equality according to
which like cases are to receive like treatment. If a decision maker has true
discretionary power, it seems she must be free to decide like cases differ-
ently. If she is not so empowered, then it seems more apt to describe her
task as the application of a statutory rule rather than the exercise of dis-
cretionary power. And if she is free to decide like cases differently, then it
appears that the affected individual is subject to the arbitrary will of the
decision makerto dominationrather than to stable and predictable
rules consistent with the rule of law.
Yet discretionary power is a salient and arguably necessary feature of
modern administrative states. To take a common example, consider the
disciplinary power held by self-regulating professional organizations.
When a professional organization imposes a discretionary sanction on a
wayward member, it must weigh the public interest against the interest
of the impeached individual, taking into account the particular circum-
stances of the case. No legislature can foresee the circumstances of every
case, nor determine the best balance of the competing interests. Only the
professional organizations discretionary power enables it to fashion
4 Roderick A. Macdonald, On the Administration of Statutes (1987) 12 Queens L.J. 488
at 493.
5 Ibid.
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 515
remedies appropriate to a given case and its wider context. In this sphere
and others, discretion is necessary to the context-sensitive (and, plausibly,
normatively best) implementation of public policy. Its existence, however,
poses a challenge from the standpoint of the rule of law injunction to treat
like cases in a like manner.
The judgments of Justices Cartwright and Rand in Roncarelli provide
two distinct and credible answers to this challenge. In his article in this
issue, Robert Leckey seeks to unsettle the accepted wisdom that on that
matter the majority adopted the sole credible option.6 We shall see that
much can be said in favour of Justice Cartwrights judgment, and that,
read charitably, it poses several difficult challenges that Justice Rand
never really confronts explicitly. Indeed, the argument that follows de-
pends crucially on the idea that Justice Cartwrights opinion is credible
and defensible in the absence of a relational theory to ground Justice
Rands major assumptions.
Justice Cartwrights reasoning relied heavily on article 35 of the Alco-
holic Liquor Act, which stated that the [Quebec Liquor] Commission may
cancel any permit at its discretion.7 The statutory discretion was not ex-
pressly qualified or limited in any way. Justice Cartwright held that the
Commission enjoyed an unfettered discretion to grant or revoke li-
cences.8 He adopted the argument of Duplessiss counsel, which reflects
the position many judges have taken toward discretionary decision-
making:
Under the Statute, no one has a pre-existing right to obtain a
permit, and the permit being granted under the condition that it
may be cancelled at any time, and no cause of cancellation being
mentioned and no form of procedure being indicated, the cancella-
tion is a discretionary decision of a purely administrative character.9
In other words, because there is no pre-existing right to obtain a per-
mit, nor conditions placed on when it may be cancelled, holding a permit
is a privilege that may be revoked at any time, for any reason, and with-
out the benefit of a hearing. Justice Cartwright thus adopts a
6 Robert Leckey, Complexifying Roncarellis Rule of Law (2010) 55 McGill L.J. 721 at
725. Leckey focuses on Fauteux J.s dissent, and makes a compelling case that Ron-
carellis overarching rule of law message cannot be properly understood without attend-
ing to the sense in which Fauteux J. sought to vindicate a particular understanding of
the rule of law by taking a statutory and procedural bar to the action seriously. As my
purpose here is simply to use Roncarelli as a vehicle to show how discretion can be
brought into a democratic conception of common law constitutionalism, I prescind from
making claims about what Roncarelli is really about, and limit discussion to the judg-
ments of Rand and Cartwright JJ.
7 R.S.Q. 1941, c. 255, s. 35, cited in Roncarelli, supra note 2 at 139.
8 Ibid. at 167.
9 Ibid.
516 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
rights/privileges distinction under which privileges are subject to the
unfettered discretion of the decision maker. And the rights/privileges dis-
tinction, in his view, tracks a deeper law/policy distinction that distin-
guishes courts from administrative tribunals in terms of their institu-
tional roles within legal order. Whereas the legislature has a monopoly on
lawmaking, courts have a monopoly on the interpretation of law, and
therefore a monopoly on the determination of legal rights and obliga-
tions.10 Administrative bodies have a monopoly on implementing statu-
tory policy, but the administrations monopoly on policy is not understood
to involve determinations of law (or at least it is not understood to involve
determinations of law that could carry any weight on judicial review).
Justice Cartwright cited Re Ashby11 to support the judicial/admin-
istrative and law/policy distinctions he sought to defend. Whereas judges
deal in legal rights and liabilities established by statute or long-settled
principles of the common law, administrative tribunals base their deci-
sions on policy and expediency.12 On this view, an administrative tribu-
nal, within its province, is a law unto itself.13 An administrative tribunal
is simply politics writ small, where decisions are based on policy and ex-
pediency, unconstrained by legal norms and standards. So long as the
administrative tribunal acts within its province, it is free to act as a law
unto itself.
Justice Cartwrights answer to the problem posed by discretion, then,
is to deny discretion standing as a legal institution. He views discretion as
a political rather than a legal feature of public administration: to think
that discretion conflicts with the rule of law is to suppose errantly that
exercises of unqualified discretion are susceptible to control by legal prin-
ciples. Since for him discretion is not controlled by legality, and does not
enjoy standing as a legal institution, it is irrelevant to the principle that
like cases must receive like treatment. In short, Justice Cartwrights reply
to the problem of discretion is to say that there is no problem at all, since
discretion is extralegal while the principle of formal equality is quintes-
sentially legal.
10 The monopoly vocabulary is from Paul Craigs characterization of Diceys vision of
unitary democracy, a kind of democracy in which the legislature is the supreme and
only lawmaker, and the legitimacy of the judiciary and the executive issues from their
respective mandates to interpret and implement legislation. See P.P. Craig, Adminis-
trative Law, 5th ed. (London, U.K.: Sweet & Maxwell, 2003) at 4-7.
11 [1934] O.R. 421, 3 D.L.R. 565 (C.A.) [Re Ashby cited to O.R.].
12 D.M. Gordon, Administrative Tribunals and the Courts (1933) 49 L.Q.R. 94 at 106-
107, cited in Re Ashby, supra note 11 at 428, Masten J.A., cited in Roncarelli, supra
note 2 at 167.
13 Ibid.
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 517
Justice Rand took a very different approach. He held that 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.14 Discretion, Justice Rand said, 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.15 He con-
cluded that to permit without recourse the suspension of a licence critical
to ones livelihood according to the arbitrary likes, dislikes and purposes
of public officials acting beyond their public duty would signal the be-
ginning of disintegration of the rule of law as a fundamental postulate of
our constitutional scheme.16
In ascribing an unwritten constitutional status to the rule of law, Jus-
tice Rand went further than saying that in this particular case discretion-
ary power had been used for an improper or unauthorised purpose. He in-
sisted that 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.17 As David Mullan suggests, this aspect of Justice Rands
judgment points to a common law (or common sense) perspective of what
would generally be impermissible under any Act.18 This perspective is of
a piece of common law constitutionalism because it treats the common
law as a repository of principles that constitute the rule of law and control
the interpretation of statutes.19
For Justice Rand, the absence of a statutory requirement limiting
permit revocations to specific causes was of no consequence because, for
him, the grant of any legal power carries with it an obligation to exercise
the power non-arbitrarily. This prohibition against arbitrariness has a
positive and a negative dimension. The positive side is that public powers
must be used exclusively for the purposes for which they are conferred.
14 Roncarelli, supra note 2 at 140.
15 Ibid.
16 Ibid. at 142.
17 Ibid. at 140 [emphasis added].
18 David J. Mullan, Judging the Judgment of Judges: CUPE v. Ontario (Minister of La-
bour), Case Comment, (2003) 10 C.L.E.L.J. 431 at 450 [Mullan, Judging the Judg-
ment].
19 See e.g. Allan, Constitutional Justice, supra note 1; Mark D. Walters, The Common
Law Constitution in Canada: Return of Lex Non Scripta as Fundamental Law (2001)
51 U.T.L.J. 91; David Dyzenhaus, The Constitution of Law: Legality in a Time of Emer-
gency (Cambridge: Cambridge University Press, 2006) [Dyzenhaus, Constitution]; Mary
Liston, Governments in Miniature: The Rule of Law in the Administrative State in
Colleen M. Flood & Lorne Sossin, eds., Administrative Law in Context (Toronto: Emond
Montgomery, 2008) 77.
518 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
This aspect is positive in the sense that it looks to the actual legislation in
play to limit administrative decision-making to action consistent with the
statutes purposes. The negative dimension of the proscription against ar-
bitrariness reflects Mullans common law (or common sense) perspec-
tive:20 even broad grants of discretion do not authorize decision makers to
use public power arbitrarily. Regardless of what the substantive aim of a
particular statute happens to be, the presumption is that the aim is to be
pursued through non-arbitrary exercises of public power. The justification
of this presumption is that the arbitrary use of power, by definition, does
not concern itself with statutory purposes. This side of the prohibition
against arbitrariness is negative in that it does not depend on a particular
legislative scheme or purpose, but rather imposes a default condition on
the exercise of public powers. Putting the positive and negative dimen-
sions together, the following rule of law principle emerges from Justice
Rands judgment: public powers may be used only for the purposes for
which they are conferred, and the purposive exercise of public power pre-
cludes its arbitrary use.
Justice Rands answer to the problem of discretion, then, is to insist
that decision makers are under a legal obligation to use it non-arbitrarily.
This answer does not imply that like cases will necessarily receive like
treatment in the sense that all cases of a certain type will be decided for
(or against) the individual. In a given decision-making context, there may
be more than one non-arbitrary or reasonable way to exercise a particular
power.21 But Justice Rands answer does imply that like cases will receive
like treatment in a wider, purposive sense: every case decided under a
particular statute will be evaluated in light of that statutes purpose, and
no case will be decided on the basis of irrelevant considerations. As a con-
sequence, individuals are not subject to the arbitrary will of decision mak-
ers, but to public law regimes structured by public purposes. Domination
is thereby ruled out because arbitrariness is prohibited.22
20 Mullan, Judging the Judgment, supra note 18 at 450.
21 Legitimate expectations may arise if a public body decides a particular kind of case a
certain way over a period of time, and these expectations would lead to complications I
cannot address here. Put briefly, the conception of public law I attribute to Rand J.
would imply that once an agency follows one among several reasonable policies for a pe-
riod of time, the agency would be under an obligation to justify deviations from that pol-
icy in particular cases. If the agency does otherwise, it fails to treat the prejudiced indi-
vidual with the solicitude she is due under Rand J.s conception of public law. For dis-
cussion of the legitimate expectations doctrine in Canada, see Mount Sinai Hospital
Center v. Quebec (Minister of Health and Social Services), 2001 SCC 41, [2001] 2 S.C.R.
281 at paras. 22-38, Binnie J.
22 For the argument that Rand J.s judgment rests on the republican principle of non-
domination, see David Dyzenhaus, Rands Legal Republicanism (2010) 55 McGill L.J.
491 [Dyzenhaus, Rands Legal Republicanism]. Non-domination plays a central, ani-
mating role within the relational theory defended here. Roughly, non-domination sup-
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 519
Justice Cartwright, however, has a principled reply. The reply is that
Justice Rand, in lionizing the rule of law, has given short shrift to another
unwritten constitutional principle, the separation of powers. And in his
disregard for the separation of powers, the reply goes, Justice Rand has
imposed limits on public power that the legislature alone is entitled to
impose. As Justice Cartwright observes, the Quebec authorities had some
basis in statute for revoking Roncarellis licence, since on its face the Al-
coholic Liquor Act gave the Commission an unqualified discretion to can-
cel permits at any time. While Justice Rand attempts to defend his view
in part on the basis of statutory interpretation, the statute itself is at least
unclear with respect to the breadth of the discretion conferred. The legis-
lature could have limited the grounds of revocation, just as it qualified the
grounds of approval in certain cases, but had declined to do so. Thus, Jus-
tice Rand must resort to arguments related to good faith in the discharge
of public duty and the common law perspective within which legislation is
presumed to operate.
These arguments are unlikely to convince judges sympathetic to Jus-
tice Cartwrights position. For them, the permit is a privilege, and thus it
falls on the privilege side of the rights/privileges distinction they support.
Courts protect rights while administrative bodies extend privileges. The
courts, therefore, have no business interfering with political and discre-
tionary decisions concerning privileges because privileges lack underlying
legal entitlements. Moreover, if courts treat Roncarellis permit as a thing
to which he had a right extinguishable only under certain conditions, they
infringe the separation of powers by in effect rewriting the statute and
imposing legal limits without legislative warrant.
The challenge posed by the rights/privileges distinction is to explain
how an interest in an acquired liquor permit can underlie a right to hold
it that is defeasible only on grounds related to the purposes of the Alco-
holic Liquor Act. The challenge posed by the separation of powers is to ex-
plain how judges can have authority to impose on front-line decision mak-
ers a legal obligation to exercise unqualified discretion on limited grounds.
The deeper and more interesting challenge, however, is to explain why we
should think that decision makers are under a legal obligation at all to
order revocations on only limited grounds.
Perhaps they are under no obligation. Assuming for the sake of argu-
ment that non-arbitrariness is a deep-seated common law value, it does
not follow that decision makers are under an obligation to act in accor-
dance with it; not everything of moral value denotes a moral (much less
legal) duty. Heroic sacrifices are morally valuable, but they are heroic
plies the content of the public law obligation to exercise discretion non-arbitrarily, while
the relational theory explains why front-line decision makers are under an obligation to
so exercise their discretion, and why that obligation is legal in nature.
520 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
precisely because the heros sacrifice is supererogatory. The gap between
value and duty entails that there is no inconsistency in supposing that a
principle of non-arbitrariness (or non-domination) is intrinsic to the com-
mon law but that the legislature alone is authorized to give it legal effect.
And even if the value-duty gap can be bridged, it remains to be seen why
the duty is legal rather than simply moral in nature. As every first-year
student of contract law comes to know, not every morally binding promise
is also legally binding. That which is morally required is not, without
some concomitant legal indicia, legally required.
In sum, a theory is needed that can close the value-duty gap while ex-
plaining the legal nature of the decision makers purported duty. The the-
ory must also explain the authority of judges to impose such a duty, and
why a mere economic interest in a permit can ground it. To make matters
more difficult still, if the theory is to be one of common law constitutional-
ism, it has to call on concepts developed by the courts, and yet if it is to be
democratic in the sense that it is specifiable independently of judicial re-
view, the decision makers obligation should be explicable without refer-
ence to the historical role of judges on review. That is, the theory must
explain how the obligation can be normatively independent of judicial re-
view, while also explaining how judges can have institutional authority to
impose the obligation without the prompt of statute.
The theory I will call on to address these issues is Macdonalds hy-
pothesis that primary decision-makers stand in a trust-like and legal rela-
tionship to the people affected by their decisionsi.e., to the public at
large and to individuals such as Roncarelli who are touched directly by an
exercise of public power. A number of scholars have developed this theory
by drawing parallels between private legal relationships with a trust-like
characterfiduciary relationshipsand the position of public authorities
vis–vis the people subject to them.23 In the remainder of this section I
23 See e.g. Paul Finn, Public Trust and Public Accountability (1994) 3 Griffith L.R. 224;
Paul Finn, The Forgotten Trust: The People and the State in Malcolm Cope, ed., Eq-
uity: Issues and Trends (Sydney: Federation Press, 1995) 131 [Finn, Forgotten Trust];
Paul Finn, A Sovereign People, A Public Trust in P.D. Finn, ed., Essays on Law and
Government: Principles and Values, vol. 1 (North Ryde, N.S.W.: Law Book Company,
1995) 1 [Finn, Sovereign People]; Anthony Mason, The Place of Equity and Equitable
Remedies in the Contemporary Common Law World (1994) 110 L.Q.R. 238; Hon. J.J.
Spigelman, Foundations of Administrative Law: Toward General Principles of Institu-
tional Law (1999) 58:1 Austl. J. of Pub. Admin. 3; Lorne Sossin, Public Fiduciary Ob-
ligations, Political Trusts, and the Equitable Duty of Reasonableness in Administrative
Law (2003) 66 Sask. L. Rev. 129; Evan J. Criddle, Fiduciary Foundations of Adminis-
trative Law (2006) 54 UCLA L. Rev. 117. For some of my reflections on the topic, see
Evan Fox-Decent, The Fiduciary Nature of State Legal Authority (2005) 31 Queens
L.J. 259 (defending the claims made in the subsequent text about the nature of trust-
like or fiduciary relationships) [Fox-Decent, Fiduciary Nature]. See also Evan Fox-
Decent, Fashioning Legal Authority from Power: The Crown-Native Fiduciary Rela-
tionship (2006) 4 N.Z.J.P.I.L. 91; Evan Fox-Decent, Is the Rule of Law Really Indiffer-
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 521
show how the theory can answer the challenges raised by Justice Cart-
wrights dissent and thereby provide a deep structure to both Justice
Rands vision of common law constitutionalism and the place of discretion
within legality.
A. The Rights/Privileges Distinction
Trust-like or fiduciary relations can generate obligations protective of
vulnerable interests with respect to which the beneficiary has no right
apart from the entitlement arising from the fiduciary relationship. In
other words, the fiduciarys obligation and the beneficiarys right are rela-
tionship-specific and may be grounded on practical as well as legal in-
terests.24 So corporations have a right to loyalty from their directors and
officers with respect to business opportunities, but have no right to either
loyalty or those opportunities against the world.25 Similarly, under Justice
Rands vision of public law, Roncarelli had a right to non-arbitrary deci-
sion-making against public authorities, notwithstanding his lack of an in-
dependent right to a liquor permit. Within the relational approach, the
right to security against arbitrariness is explained as the legal conse-
quence of the trust-like relationship between Quebec authorities and Ron-
carelli. The relationship arises from the power delegated to administer
the liquor control regime, and this relationship implies a legal obligation
to administer the regime consistently with the purposes for which the
power is granted. As with other fiduciary relations, the correlative right is
specific to the relationship in question and can attach to Roncarellis prac-
tical interest in the permit. That Roncarelli had no freestanding right to
the permit is irrelevant because all that needs to be defended is a rela-
tionship-specific right to non-arbitrary treatment. Thus, the relational
theory explains and justifies Justice Rands assumption that the
rights/privileges distinction is irrelevant.
B. The Separation of Powers and Judicial Activism
The objection based on the separation of powers trades on the idea
that the legislature alone is entitled to create legal rights and obligations,
and so judges overstep if they impose constraints that lack a legislative
anchor. Judges in the Rand camp can attempt to answer the objection by
falling back on legislative intent; the legislature is said to have intended
its statute to operate in accordance with the relevant limit or constraint.
ent to Human Rights? (2008) 27 Law & Phil. 533; Evan Fox-Decent & Evan J. Criddle,
The Fiduciary Constitution of Human Rights (2009) 15 Legal Theory 301.
24 See e.g. Frame v. Smith, [1987] 2 S.C.R. 99 at 136, 42 D.L.R. (4th) 81. See also Norberg
v. Wynrib, [1992] 2 S.C.R. 226 at 276, 92 D.L.R. (4th) 449.
25 See e.g. Canaero Aero Service v. OMalley (1973), [1974] S.C.R. 592, 40 D.L.R. (3d) 371.
522 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
Justice Rand gestures in this direction when he writes that there is al-
ways a perspective within which a statute is intended to operate.26 One of
the difficulties of this approach, however, is that it places considerable
weight on the forgetfulness or oversight of the legislature. This is espe-
cially problematic with statutes such as the Alcoholic Liquor Act in Ron-
carelli in which some grants of discretionary power are expressly limited
while others are unqualified. In these cases it appears that the legislature
did turn its mind to the issue of legal limits on discretionary power within
the statutory scheme.
The relational theory charts an entirely different course, explaining
why the separation-of-powers worry is misguided, and indeed why the
separation of powers is irrelevant to the assessment of the legality of ad-
ministrative action. The separation of powers is misguided because it pre-
supposes that judges illegitimately make law by imposing obligations on
decision makers without statutory warrant. On the relational theory,
judges are simply recognizing rights and duties intrinsic to and constitu-
tive of the trust-like relationship between public authorities and the af-
fected party. Judges who enforce such rights and duties no more infringe
on the law-making authority of the legislature than judges who uphold
rights and duties arising from contract or tort in private law. Even if such
adjudication is in some sense judicial lawmaking, it is the kind of law-
making that common law judges engage in on a routine basis.
The bolder claim is that the separation of powers is irrelevant to de-
terminations of the legality of administrative actiondeterminations that
must be made by primary decision-makers at first instance as well as
judges on review. The separation of powers is irrelevant because, on the
relational theory, the criteria for assessing the legality of administrative
action are supplied by a limited set of legal principles constitutive of the
underlying trust-like relationship. It is beyond the scope of this paper to
defend the set in more than general terms, but it includes: a prohibition
on fraud and corruption; procedural fairness; formal equality or even-
handedness; solicitude in the sense of taking seriously the legitimate in-
terests and human rights of individuals subject to public power; transpar-
ency; proportionality; reason-giving where important interests are at
stake; and purposiveness in the sense of Justice Rands principle that
public powers must be used exclusively for the purposes for which they
are conferred.27 The violation of any of these principles would offend the
26 Roncarelli, supra note 2 at 140.
27 For discussion of several of these principles and their derivation from a fiduciary theory
of public authority, see Evan Criddle & Evan Fox-Decent, A Fiduciary Theory of Jus
Cogens (2009) 34 Yale J. Intl L. 331 [Criddle & Fox-Decent, Jus Cogens]. For discus-
sion of these principles from the same theoretical perspective in the context of U.S. ad-
ministrative law, see Evan J. Criddle, Fiduciary Administration: Rethinking Popular
Representation in Agency Rulemaking (2010) 88 Texas L. Rev. 441.
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 523
central idea that public decision-makers occupy a trust-like position vis–
vis the people they serve, where every agent subject to public power is
necessarily regarded as an equal co-beneficiary of the rule of law. Each
principle is justified as a necessary and constituent part of the trust-like
relationship.
The trust-like relationship itself is normatively grounded on respect
for the agency of persons subject to irresistible public power. In this trust-
like relationship, the imperative to respect agency reflects the demands of
a Kantian idea of dignity, and crystallizes in two intermediary principles
that connect respect for agency to the more determinate legal principles
set out above. One of these intermediary principles is the idea that per-
sons are to be treated as ends always (the principle of non-
instrumentalization), while the other is the republican idea that individu-
als are not to be subject to arbitrary power (the principle of non-
domination). Non-domination complements non-instrumentalization in
the following way. Whereas non-instrumentalization prohibits public ac-
tors from wrongfully interfering with their subjects, non-domination bars
them from holding arbitrary power that ipso facto would pose a wrongful
threat because it could be exercised wrongfully at any time. In other
words, non-instrumentalization controls the actual exercise of power,
while non-domination controls the threat implicit in the mere possession
of power, and so non-domination controls the terms of its possession. Ron-
carelli suffered domination because Quebec authorities interpreted the
Alcoholic Liquor Act to give themselves arbitrary power. He suffered in-
strumentalization because they wrongfully used that assumed power to
cancel his permit. Legal principles such as fairness, purposiveness, and
reason-giving provide a bulwark against the possession and use of arbi-
trary power, and thereby embody in a more determinate form the re-
quirements of non-instrumentalization and non-domination.
With determinate legal principles in place, we can see why the separa-
tion of powers is irrelevant to the legality of administrative action. The
separation of powers is usually deployed to restrict judicial review on the
grounds that the legislature alone is authorized to make law. These
grounds are formal in the sense that they rely on formal distinctions
made between the role of the legislature and the role of the judiciary, and
between interpreting law and making law. As a result of its formalism,
the separation of powers can offer no substantive guidance to the legiti-
mate scope of judicial review,28 making a contentious retreat to legislative
intent almost inevitable. Now, to be sure, the legal principles set out
above constitute a fairly substantial framework to assess the legality of
administrative action. But the framework is limited to more or less those
28 For more discussion, see David Dyzenhaus, Formalism’s Hollow Victory (2002) N.Z.L.
Rev. 525.
524 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
very principles, all of which are derived from the trust-like relationship
between public authorities and the people. Because the principles arise
from this legal relationship, affirmation of them does not usurp the legis-
lative lawmaking role. So the separation of powers is respected willy-nilly,
but it no longer does any work as an analytical tool on judicial review. In-
stead, the legal principles arising from the trust-like relationship do the
work. It follows that once we have in view the relational theory and its
implications, the separation of powers is irrelevant to determinations of
the legality of administrative action.
C. The Content and Legal Nature of the Obligation in Roncarelli
As noted already, Justice Rand held that Quebec authorities were un-
der an obligation to revoke liquor permits on grounds related exclusively
to the purposes for which the power was conferred. The relational theory
explains the content of the obligation because in all trust-like or fiduciary
relationships, discretionary power is held for limited and other-regarding
purposes (i.e., for purposes that further certain interests of some person or
persons other than the entrusted party), and the power must be exercised
exclusively for the sake of the other-regarding purposes for which it is
conferred. This is precisely the content of the principle of purposiveness
Justice Rand affirmed. Furthermore, the must here is a legal must be-
cause trust-like or fiduciary relationships are legal relationships. They are
legal relationships because, in Kantian and republican terms, their struc-
ture and normative presuppositions guarantee de jure that the parties in
them are subject to neither instrumentalization nor domination. So the
relational theory explains both the content and legal nature of the basic
obligation in Roncarelli. In so doing, the theory also provides a legal
framework for discretion. This framework ensures that individuals are re-
spected as participants in and co-beneficiaries of a regime of public pur-
poses and legal principles; they are not left to the arbitrary choices of oth-
ers. The relational theory thus explains how public law can shield indi-
viduals from domination, and it provides the resources for a remedy
should they suffer an abuse of power and instrumentalization.
Readers might question whether the fiduciary idea is appropriate to
the public law context because the hallmark fiduciary obligation of private
law is a duty of loyalty, whereas in public law the decision maker often
must weigh the interest of the affected individual against the public in-
terest. I have argued elsewhere that in the public setting, where both the
public at large and the affected individual are considered beneficiaries,
the overarching fiduciary duty necessarily becomes a duty of fairness and
reasonableness.29 Within this framework, more concrete obligations such
29 Fox-Decent, Fiduciary Nature, supra note 23 at 264-68. This is also the overarching
duty that applies at private law when fiduciaries must exercise their power over multi-
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 525
as duties of procedural fairness and reason-giving are generally owed to
individuals directly touched by public power. But the decision maker must
also take account of the public interest, as the public too is a beneficiary of
public power.
Bearing in mind both the private duty of loyalty and the public duty of
fairness and reasonableness, a more general formulation of the funda-
mental fiduciary obligation can be adduced: at its most general, the fidu-
ciarys basic obligation is to exercise her other-regarding power exclu-
sively for the purposes for which it is held. Arguably, this general fiduci-
ary obligation tracks precisely Justice Rands principle of purposiveness
because public power is always fiduciary in nature. In practice this means
that although courts on review must not advance the applicants best in-
terests to the exclusion of the public interest (as might seem warranted on
an overly literal application of the fiduciary concept to the public realm),
courts must endeavour to interpret statutory terms and purposes in a way
that is solicitous of the applicants legitimate interests. Courts must so
endeavour because statutory powers, under the relational theory, are au-
thorized on behalf of and for the benefit of every person subject to them,
including the applicant who comes before the court on judicial review.
D. Normative Independence from Judicial Review
While the relational theory serves to delimit the scope of judicial re-
view, it is specifiable independently of review. To put the point starkly, if
judges in the common law world had never engaged in judicial review, the
underlying relationship between public authorities and the people would
still be trust-like, and the approach of judges to trust-like relationships in
other domains could still be plumbed for the purpose of developing a
framework for assessing the legality of administrative action. As a conse-
quence, the relational theory provides a unified account of the legality of
administrative action that is as accessible and relevant to primary deci-
sion-makers as it is to judges on review. By distinguishing legality from
judicial review, the theory democratizes common law constitutionalism by
showing that its favoured principles are not the result of judicial fiat.
Rather, they are the constitutive norms of a shared legal order that all
public bodies are responsible for maintaining on behalf of the people. The
relational theory explains and justifies certain elements of judicial review,
as we have seen in the case of Justice Rands judgment, but it does so
from a critical perspective internal to law that gives no weight to the hier-
archical authority of judges to quash agency determinations.
ple classes of beneficiaries with competing claims. See e.g. Equitable Life Assurance So-
ciety v. Hyman, [2000] 2 W.L.R. 798, [2000] 2 All E.R. 331 (C.A.), affd (2000), [2002] 1
A.C. 408, [2000] 3 W.L.R. 529 (H.L.); Edge v. Pension Ombudsman (1999), [2000] 3
W.L.R. 79, [1999] 4 All E.R. 546 (C.A.).
526 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
II. Reframing Common Law Constitutionalism
I argue now that the relational theory can help common law constitu-
tionalists answer their critics. One of those critics, Thomas Poole, has
marshalled an impressive array of arguments against common law consti-
tutionalism,30 turning his attention more recently to David Dyzenhauss
extension of the theory to the emergency powers context.31 Poole criticizes
Dyzenhaus for failing to specify with a requisite degree of detail, what
those hard-edged values are which derive from the common law and give
substance to the rule of law.32 He acknowledges that Dyzenhaus points to
equality, due process, fairness, reasonableness, and sensitivity to human
rights, but objects that Dyzenhauss catalogue tells us nothing about
where the chosen values come from, save that they are inherent in the
very notion of legality, which, given that this is precisely the subject in
dispute, rather begs the question.33 I will refer to this criticism as the cir-
cularity objection.
Poole further claims that the common law values Dyzenhaus affirms
are indeterminate in that they do not offer a coherent blueprint for judi-
cial decision making.34 He objects as well that Dyzenhaus has failed to
show why common law values should outweigh (always? generally?)
other, countervailing values, such as security or even national self-
preservation.35 I will call this the indeterminacy objection.
Poole also challenges Dyzenhaus and other common law constitution-
alists for relying on a falsely Whiggish history of the common law. Within
the Whiggish account, Lord Cokes readiness to place substantive controls
30 See e.g. Thomas Poole, Dogmatic Liberalism?: T.R.S. Allan and the Common Law Con-
stitution, Book Review of Constitutional Justice: A Liberal Theory of the Rule of Law by
T.R.S. Allan, (2002) 65 Mod. L. Rev. 463; Thomas Poole, Questioning Common Law
Constitutionalism (2005) 25 L.S. 142; Thomas Poole, Legitimacy, Rights and Judicial
Review (2005) 25 Oxford J. Legal Stud. 697. For Canadian public lawyers, Pooles work
will call to mind the provocative scholarship of Harry Arthurs and Allan Hutchinson.
See e.g. H.W. Arthurs, Protection Against Judicial Review (1983) 43 R. du B. 277;
H.W. Arthurs, Rethinking Administrative Law: A Slightly Dicey Business (1979) 17
Osgoode Hall L.J. 1; Allan C. Hutchinson, Judges and Politics: An Essay from Canada
(2004) 24 L.S. 275; Allan C. Hutchinson, Evolution and the Common Law (Cambridge:
Cambridge University Press, 2005).
31 Poole, Constitutional Exceptionalism, supra note 3; Dyzenhaus, Constitution, supra
note 19; David Dyzenhaus, Schmitt v. Dicey: Are States of Emergency Inside or Out-
side the Legal Order? (2006) 27 Cardozo L. Rev. 2005; David Dyzenhaus, The Puzzle
of Martial Law (2009) 59 U.T.L.J. 1.
32 Poole, Constitutional Exceptionalism, supra note 3 at 264 ×..
33 Ibid. ×..
34 Ibid.
35 Ibid. at 265.
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 527
on Parliament in Dr. Bonhams Case36 is cast as the timeless emblem of
the common laws rule of reason, whereas, Poole claims, it should be seen
… as an outlier within this tradition.37 In other words, when common law
constitutionalists are pressed to provide a source for the values they es-
teem, they frequently resort to cherry-picking cases that ill-represent the
broader history of the tradition. Moreover, Poole alleges that common law
constitutionalists mischaracterize the common law as mainly a tradition
of legal principles, when in fact its history has been much more dominated
by rules, procedures, maxims, professional practices, and questions of ju-
risdiction, writs and remedies. I will call this the historical objection.
Underlying all of these objections is scepticism of judicial review, or at
least scepticism of the idea that judges occupy a position of first among
equals when it comes to determining and policing the limits of legality.38
The three objections may be thought to combine in the following way.
To avoid the circularity objection, common law constitutionalists look to
the history of judicial review to show that common law values are deeply
embedded in the tradition and therefore deserve the privileged attention
of judges on review. But if the historical objection is sound (and here, I as-
sume that it is), common law constitutionalists are relying on a skewed
vision of the values that historically animated the common law. Without
an historical argument to underpin the theory, common law constitution-
alists face a dilemma. On the one hand, they can resort to a Dworkinian
approach according to which law aspires to embody the avowed principles
of liberal political morality, such as equality and autonomy. But then the
rule of law collapses into a restatement of liberal political morality and
loses its distinctiveness as an autonomous and guiding ideal of the com-
mon law legal order. On the other hand, common law constitutionalists
can insist that their preferred values are somehow intrinsic to the com-
mon law, but then the circularity objection resurfaces. And it is precisely
this circularity that gives the indeterminacy objection its punch: without a
non-question-begging account of the source and priority of common law
values, it is unclear how they are supposed to guide judges who must
weigh other considerations against them on review.
The relational theory goes to the heart of this critique by explicitly
specifying the source and basis of the principles relevant to assessing the
legality of administrative action. The guiding principles do not arise from
36 (1610), 8 Co. Rep. 113b, 77 E.R. 638 (Common Pleas) [Dr. Bonhams Case]. According to
Lord Coke, when an Act of Parliament is against common right and reason, or repug-
nant, or impossible to be performed, the common law will controul it, and adjudge such
Act to be void (ibid. at 652).
37 Poole, Constitutional Exceptionalism, supra note 3 at 268-69 ×..
38 Ibid. at 263, 269-70.
528 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
a common law ether as an exercise in wish fulfillment.39 Instead, they
constitute the normative dimension of a legally significant trust-like rela-
tionship that exists between public authorities and the people. The prin-
ciples therefore have an intelligible source and a normative framework
capable of explaining how they can give rise to concrete duties in public
law. Put slightly differently, the story of where the chosen values come
from40 is not a bootstrapping insistence that they are inherent to legality,
but a separate relational account in which the idea that principles are in-
trinsic to the common law is presented as the conclusion of the argument
rather than as its major premise or an article of faith.
The guiding principles are intrinsic to the common law because they
play a constitutive role in the underlying common law and trust-like rela-
tionship. Just as the duty of non-interference is a constitutive aspect of
tort relationships, duties such as fairness and solicitude are constitutive
elements of trust-like relationships in which the entrusted party must on
act behalf of multiple parties, as is the case in the public sphere where the
public interest is always at stake. The relational theory thus avoids the cir-
cularity objection by explaining how its guiding principles are intrinsic to
legality; they are intrinsic to legality because they constitute the normative
dimension of the trust-like relationship that makes legal order possible.
With respect to the historical objection, the relational theory can grant
it because, as discussed already, the theory is specifiable without refer-
ence to the historical practice of judicial review. That is, the theorys
framework and principles do not rely on the salience of similar principles
within either the celebrated cases from the early seventeenth century,
such as Dr. Bonhams Case41 and Baggs Case,42 or more contemporary
judgments prized by common law constitutionalists, such as Lord Shaws
dissent in R. v. Halliday,43 Lord Atkins dissent in Liversidge v. Ander-
son,44 and Lord Hoffmans opinion in R. v. Secretary of State for the Home
Department.45
39 Ibid. at 266.
40 Ibid. at 264 ×..
41 Supra note 36.
42 (1615), 11 Co. Rep. 93b, 77 E.R. 1271.
43 [1917] A.C. 260 (holding that the general regulation-making power under the Defence
of the Realm Consolidation Act, 1914 did not authorize the government to make regula-
tions empowering the Secretary of State to detain individuals without trial). For discus-
sion, see Dyzenhaus, Constitution, supra note 19 at 157-60.
44 (1941), [1942] A.C. 206, [1941] 3 All E.R. 338 (finding that a wartime detention regula-
tion permitting detention without trial must be read objectively rather than subjec-
tively). For discussion, see Dyzenhaus, Constitution, supra note 19 at 23-26.
45 (1999), [2000] 2 A.C. 115 at 131, [1999] 3 W.L.R. 328 (Lord Hoffman affirming that
[f]undamental rights cannot be overridden by general or ambiguous words).
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 529
The relational theory does, of course, derive some inspiration from the
evolution of fiduciary and trust concepts that date back to at least the fi-
deicommissum of Roman Law.46 These concepts developed through the
doctrines of utilitas ecclesia (for the advantage of the church) and uses in
the Middle Ages.The use was a form of land trust that entailed the trans-
fer of legal title (enfeoffment) to a person who was to hold the property
(the feoffee to uses) for the benefit of another (the cestui que use).47 Fidu-
ciary concepts were subsequently developed and extended by courts of eq-
uity that supervised trust and quasi-trust relationships.48 While still an
academic, Paul Finn pointed out that through the eighteenth and early
part of the nineteenth century, English judges viewed the position of pub-
lic office holders through a trust-like prism.49 Lord Mansfield character-
ized public offices in 1783 as offices of great public trust and confi-
dence.50 Taking as its point of departure the trusteeship of public office
holders, a body of criminal, civil, and equitable law developed to supervise
the use of public power. But by the latter half of the nineteenth century,
the work of this law of offices was being done by the prerogative writs,
and Diceys principles of parliamentary sovereignty and the rule of law
came to dominate the commonwealth constitutional landscape. The law
of offices passed into disuse and obscurity.
The relational theory deploys this history to make a very modest
point: a trust-like conception of public authority is not inimical to public
law in a common law legal order, since public law was once based largely
on this very conception. In other words, the relevant history confirms the
plausibility of the trust-like conception, and supplies pointers to its nor-
mative framework and principles, but does not pretend to be an account of
46 See e.g. Ernest Vinter, A Treatise on the History and Law of Fiduciary Relationship and
Resulting Trusts, 3d ed. (Cambridge: W. Heffer & Sons, 1955) at 2-9 (setting out the
Roman influence on the development of modern fiduciary law).
47 Avisheh Avini, The Origins of the Modern English Trust Revisited (1996) 70 Tul. L.
Rev. 1139 at 1143 [footnotes omitted].
48 For historical accounts, see Thomas O. Main, Traditional Equity and Contemporary
Procedure (2003) 78 Wash. L. Rev. 429; Timothy S. Haskett, The Medieval English
Court of Chancery (1996) 14 Law & Hist. Rev. 245; Mary Szto, Limited Liability
Company Morality: Fiduciary Duties in Historical Context (2004) 23 Quinnipiac L.
Rev. 61; Sir Frederick Pollock & Frederic William Maitland, The History of English
Law, 2d ed., vols. 1, 2 (Cambridge: Cambridge University Press, 1968). In this paper I
use common law in an expansive sense to refer generally to the law developed by judi-
cial authorities, and not in the more technical and precise historical sense in which
courts of common law were separate from courts of equity, and exercised a separate ju-
risdiction. For present purposes, a common law fiduciary relationship is meant to re-
fer to a relationship that has its origins in equity, but which is recognized by common
law judges today, who exercise concurrent common law and equitable jurisdictions.
49 Finn, Forgotten Trust, supra note 23 at 132-53; Finn, Sovereign People, supra note
23 at 9-14.
50 R. v. Bembridge (1783), 3 Doug. 679 at 679, 99 E.R. 679.
530 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
an ancient constitution that is venerable because it is ancient or a matter
of settled judicial custom.
It bears emphasizing that the history relied upon is not a line of select
cases, but an entire class of legal relations that govern the myriad circum-
stances in which one party is empowered to act on behalf of another, leav-
ing the latter vulnerable to the formers discretionary power. The focus on
a capacious class of legal relations rather than select (liberal) cases takes
the sting out of the cherry-picking complaint. It also bears emphasizing
that the lack of normative reliance on the history of judicial review lets
the relational theory specify a conception of legality that is independent of
review. Common law constitutionalism can therefore be seen in a democ-
ratic and ecumenical light because the relational theory sets out a frame-
work of principles whose content does not rely on the threat or practice of
review.
The indeterminacy objection is thornier and cannot be addressed ade-
quately here. But to the extent that the principles of common law consti-
tutionalism are held to be indeterminate and incapable of providing guid-
ance because they arise from cherry-picked cases or rely on circular rea-
soning, we have seen that the relational theory offers a way forward. By
specifying a normative framework based on a trust-like relationship, the
theory invites inquiry into the sorts of principles that must be upheld to
maintain the integrity of that relationship. Relatively few principles
would enjoy peremptory or absolute status, but public international law
suggests that the set of such principles is far from empty. Here too a fidu-
ciary conception of public authority can explain the peremptory norms of
jus cogens governing genocide, arbitrary killing, torture, slavery, systemic
racial discrimination, prolonged arbitrary detention, corruption, minimal
due process, military aggression, and self-determination.51 Legislation or
administration supportive of such policies would deny from the outset the
idea that public authorities were acting as public servants and trustees of
their people; trustees cannot act as such through policies of deliberate vic-
timization.
In the more ordinary course of events, where important interests are
in play but non-derogable human rights are not at stake, principles of
procedural fairness and solicitude combine to ensure that the affected in-
51 See Criddle & Fox-Decent, Jus Cogens, supra note 27. An anonymous reviewer ob-
serves that jus cogens in international law is a muddle in its own right, and largely ar-
bitrary with no decent explanatory theory other than that something is jus cogens be-
cause people keep saying it is and it eventually sticks as such. In the article cited in this
note (ibid.), Evan Criddle and I attempt to address this concern by showing how the re-
lational and fiduciary theory can yield formal and substantive criteria capable of distin-
guishing genuine jus cogens norms from counterfeits. We have since elaborated on the
theorys implications for human rights more generally: see Evan Fox-Decent & Evan J.
Criddle, The Fiduciary Constitution of Human Rights (2009) 15 Legal Theory 301.
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 531
dividual has a fair opportunity to participate in the decision-making proc-
ess, and that her interests are weighed seriously against contending pub-
lic concerns. Some indeterminacy with regard to outcome is inevitable,
especially in hard cases, but indeterminacy is not the same thing as arbi-
trariness. In so far as the legal principles of the relational theory oust ar-
bitrariness by ousting instrumentalization and domination, it is no dis-
credit to the theory that it cannot predetermine outcomes. Individuals are
entitled to a regime of public law in which they do not suffer domination,
and to fair procedures and reasonable decisions that do not instrumental-
ize them. They are not entitled a priori to decisions that favour them, and
about which reasonable decision-makers may disagree.
Before closing I wish to respond to two questions raised by partici-
pants at this commemorative symposium. The first concerns the practical
implications of the relational theory. If the Supreme Court of Canada
adopted the theory, what difference in practice might it make to judicial
review? Consider first Canada (A.G.) v. Mossop,52 a case in which the ma-
jority of the Supreme Court of Canada quashed a decision of the Canadian
Human Rights Tribunal on the basis that family status, a prohibited
ground of discrimination under the Canadian Human Rights Act,53 could
not be interpreted to protect the partner of a same-sex couple from dis-
criminatory treatment. According to the majority of the Court, the legisla-
ture did not intend to extend protection to same-sex couples because sex-
ual orientation had been left out of the prohibited grounds of the legisla-
tion, and because the usual and ordinary sense of family in Canada,
Justice La Forest stated, is the traditional family.54
Justice LHeureux-Dub, dissenting, insisted that the point of human
rights statutes is to ensure equality, and that there are statute-
independent fundamental principles that must be at the forefront of the
interpretation of human rights codes.55 In other words, she subscribes to a
common law constitution that contains fundamental values expressive of
human rights, such as equality, and these values are to guide statutory
interpretation. With this framework in place, it may appear that the work
the relational theory intends to do has already been done. Yet cases such
as Mossop demonstrate that many judges are reluctant to depart from
their favoured interpretation of the original intent of the legislature, and
others still are reluctant to move beyond their understanding of the
usual and ordinary meaning of statutory words.56 Of course, it would be
52 [1993] 1 S.C.R. 554, 100 D.L.R. (4th) 658 [Mossop cited to S.C.R.].
53 R.S.C. 1985, c. H-6.
54 Mossop, supra note 52 at 580-82, Lamer C.J.C., Sopinka and Iacobucci JJ. concurring;
at 585-86, La Forest and Iacobucci JJ. concurring.
55 Ibid. at 621.
56 See supra note 54 and accompanying text.
532 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
nave to think that any theory of common law constitutionalism might
convert a committed originalist, but the fiduciary idea both strengthens
Justice LHeureux-Dubs argument and exposes the majoritys error. As
public fiduciaries, tribunals and judges are legally required to endeavour
to interpret statutes in a manner solicitous of the well-being of the people
subject to them. This requirement is especially demanding when there are
no significant competing interests, as was the case in Mossop.
Another case that brings to the surface the practical implications of
the relational theory is Dunsmuir v. New Brunswick.57 In this case the
Supreme Court of Canada reversed its prior holding in Indian Head
School Division No. 19 v. Knight58 that at pleasure office holders are en-
titled to procedural safeguards prior to termination. The Court in Duns-
muir held that contract law rather than the public law duty of procedural
fairness governs the parties upon dismissal. The majority writers, Jus-
tices Bastarache and LeBel, found that the Crown as an employer is in
much the same position as a private employer, and so the private law of
contract applies. They insisted that termination with reasonable notice
cannot be qualified as arbitrary.59 But here the Court confuses the terms
of notice accompanying dismissal with the grounds for dismissal. If a per-
son is dismissed for capricious or irrelevant reasons, then the dismissal is
arbitrary in the conventional sense that it is based on irrelevant consid-
erations. No amount of notice can undo this arbitrariness.
Under the relational theory, the Crown is a special kind of employer
because it is a public employer, and as such it stands in a trust-like rela-
tionship to both the employee and the public at large. Procedural safe-
guards are owed to public employees given the importance and vulnerabil-
ity of the employment interest, and the trust-like obligation of public bod-
ies to take seriously the legitimate interests of individuals directly af-
fected by public power. The Crown retains authority to dismiss for policy
as well as performance reasons, but under the fiduciary theory it cannot
terminate employment arbitrarily; if it could, the Crowns relationship to
the employee would be one of domination.60 Procedural safeguards provide
a measure of security against arbitrary dismissal by exposing the grounds
of termination and giving the individual an opportunity to be heard. The
Crowns countervailing (and equally trust-like) duty to provide efficient
administration is given its due under the relational theory because par-
57 2008 SCC 9, [2008] 1 S.C.R. 190, 329 N.B.R. (2d) 1 [Dunsmuir].
58 [1990] 1 S.C.R. 653, 69 D.L.R. (4th) 489 [Knight cited to S.C.R.].
59 Dunsmuir, supra note 57 at para. 104.
60 It is no accident that (non-unionized) employment figures prominently as an exemplar
of domination in Philip Pettits major work on non-domination: Philip Pettit, Republi-
canism: A Theory of Freedom and Government (Oxford: Oxford University Press, 1997)
at 48-49.
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 533
ticipatory rights are restricted to those necessary to let the individual
know and reply to the considerations that may be weighed against her.61
If the individuals credibility is a determining factor, an oral hearing may
be required, but in most cases disclosure of the Crowns case and an op-
portunity to make written submissions will suffice.62
The second question raised at the symposium goes to the relational
theorys normative underpinnings, and the explanation it offers of the le-
gal nature of public law duties. The discussant noted that judges sympa-
thetic to common law constitutionalism impose common law duties be-
cause ultimately they think these duties are just or fair. They view their
role in legal order as one that both entitles and requires them to impose
public law duties that track justice and fairness. But, he continued, if
Pooles circularity and historical objections are sound, then these judges
are really just legislating their convictions from the bench. He acknowl-
edged that the relational theory aims to avoid this complaint and explain
public law duties by drawing an analogy to the private law of fiduciaries.
Yet why should we suppose that when judges impose fiduciary duties of
private law they are doing anything more than insisting on what they
think is just and fair? And why should sensibilities of justice and fairness
at private law enjoy some special privilege over like sensibilities of public
law?
To clarify a preliminary issue: the relational theory does not claim
that there is simply an analogy to be drawn between the circumstances in
which public authorities confront legal subjects and private fiduciary rela-
tionships. The claim, rather, is that public institutions actually stand in a
trust-like or fiduciary relationship to the people subject to their power.
Private fiduciary relations help to reveal the constitutive features of fidu-
ciary relationships, but it is the possession of those actual featuresand
not an analogy to private lawthat makes the relationship between pub-
lic institutions and the people fiduciary.
The constitutive features of fiduciary relationships may be appre-
hended by a brief recital of the conditions that give rise to them. Fiduciary
relationships arise when one party (the fiduciary) holds discretionary
power of an administrative nature over the legal or practical interests of
another party (the beneficiary). This administrative power is other-
regarding, purposive, and institutional; it is held so as to be used on be-
half others, for limited purposes, and within the framework of a legal in-
stitution such as a family or a corporation. The beneficiary is peculiarly
61 The theory thus reflects LHeureux-Dub J.s view that every administrative body is
the master of its own procedure and need not assume the trappings of a court: Knight,
supra note 58 at 685.
62 See e.g. Suresh v. Canada (Minister of Citizenship and Immigration), 2002 SCC 1,
[2002] 1 S.C.R. 3 at paras. 121-27, 208 D.L.R. (4th) 1.
534 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
vulnerable to the fiduciarys power in that she is unable, either as a mat-
ter of fact or law, to exercise the entrusted power. The relationship be-
tween public institutions and the people subject to them possesses these
characteristics. Public institutions hold administrative power that is
other-regarding, purposive, and institutional. Moreover, legal subjects, as
private parties, are not entitled to exercise it. The relationship between
public institutions and legal subjects is therefore fiduciary because it pos-
sesses the constitutive features of fiduciary relations.
A relationship in which the fiduciary has unilateral administrative
power over the beneficiarys interests can be understood as a relationship
mediated by law, however, only if the fiduciary is precluded from exploit-
ing her position to set unilaterally the terms of her relationship with the
beneficiary. The fiduciary principle authorizes the fiduciary to exercise
power on the beneficiarys behalf, but subject to strict limitations arising
from the beneficiarys vulnerability to the fiduciarys power and her in-
trinsic worth as a person.63 This Kantian construal of fiduciary relations
trades on the idea that legal order consists in the ongoing construction of
a regime of secure and equal freedom, a regime marked by an absence of
domination and instrumentalization.64 Because beneficiaries in fiduciary
relations cannot exercise or control the power to which they are subject,
the law controls it, protecting them from unilateralism, and ensuring
their secure and equal freedom. In short, fiduciary duties are legal duties
because they safeguard the individual against domination and instrumen-
talization. Within the relational theory defended here, this is the norma-
tive basis of legality and the explanation of why fiduciary duties are dis-
tinctively legal duties.
Conclusion
Admittedly, the relational theory is an ideal theory that does not co-
here especially well with the way most judges and commentators today
think about the theoretical underpinnings of judicial review. The legisla-
tive-intent (or ultra vires) theory and mainstream understandings of
common law constitutionalism score better on that front. But all three
theories share roughly the same methodologyinference to the best ex-
planationas they attempt to explain common law duties of public law
63 When Pettit initially sets out what it would mean for someone to interfere with some-
one else in a way that does not dominate them, the example he uses (cited by Dyzen-
haus, Rands Legal Republicanism, supra note 22) is a paradigmatic fiduciary rela-
tionship: an agent who possesses a power of attorney over anothers affairs. See Pettit,
supra note 60 at 23.
64 For extended discussion of Kants theory of right (part of which includes fiduciary rela-
tions), and how Kants theory informs his politics, see Arthur Ripstein, Force and Free-
dom: Kants Legal and Political Philosophy (Cambridge, Mass.: Harvard University
Press, 2009).
DEMOCRATIZING COMMON LAW CONSTITUTIONALISM 535
and the appropriate approach to statutory interpretation. I have sug-
gested that the legislative-intent theory is unconvincing, and that com-
mon law constitutionalism, as generally conceived, is hard-pressed to ex-
plain how common law values translate into legal duties. The conven-
tional understanding of common law constitutionalism is also vulnerable
to the objections Poole raises. The relational theory, by contrast, offers
principled replies to Pooles circularity and indeterminacy objections, and
can grant his historical argument. It can do so by explaining common law
duties and principles governing statutory interpretation as constitutive
features of a trust-like relationship between public authorities and the
people. The relational theory also explains the irrelevance of the separa-
tion of powers to assessing the legality of administrative action, while
providing a legal framework for discretion respectful of agency because
consistent with non-instrumentalization and non-domination. The theory
thus supplies a framework for Justice Rands judgment that reconciles
discretion with the rule of law and answers the separation-of-powers ar-
gument that supports Justice Cartwrights dissent.
Just as significantly, however, the relational theory democratizes
common law constitutionalism by taking seriously, and ecumenically, the
actual features of the actual relationship that exists between public au-
thorities and the people. Because this relationship would exist with the
same basic features with or without the possibility of judicial review, it
can underlie a theory of legality that would remain intact if Canada,
through constitutional reform, abolished altogether judicial review of ad-
ministrative action. By revealing the democratic promise of common law
constitutionalism, the relational theory thus reveals the democratic prom-
ise of the rule of law.
