Learning to Live With The Override
Lorraine Eisenstat Weinrib*
In Ford v. A.G. Quebec, the Supreme Court of
Canada had its first occasion to interpret the
override clause found in section 33 of the
Charter. The author argues that by focusing on
purely formal requirements the Court incor-
rectly treated the override clause as an ordi-
nary exercise of legislative power.
Although the override clause has been thought
of as repugnant to a rights protecting docu-
ment, the author submits that it does not nec-
essarily involve a choice between rights
entrenchment and legislative sovereignty.
Rather, the interaction of section 33 with sec-
tion 1 of the Charter results in the creation and
entrenchment of an institutional dialogue
between the courts and the legislature. Without
that dialogue we may experience a progressive
weakening of the judicial commitment to
rights protection.
Dans Ford c. P.G. Quebec, la Cour supreme du
Canada a, pour la premiere fois, interprrt6 l’ar-
tile 33 de la Charte et la clause nonobstant
que l’on y retrouve. La Cour n’a, toutefois,
que tenu compte des conditions explicites de
son application. Ceci, d’apr~s l’auteur, carac-
tdrise incorrectement l’application de la clause
nonobstant comme n’6tant qu’un simple exer-
cice de pouvoir 16gislatif.
Bien que la clause nonobstant ait 6t6 pergue
comme 6tant contraire A l’esprit d’un docu-
ment visant la protection de droits fondamen-
taux, l’auteur soutient qu’elle n’implique pas
nrcessairement un choix entre l’enchassement
de droits et la souverainet6 du pouvoir 16gisla-
tif. Au contraire, l’analyse de l’article 33, sous
rdserve de l’article 1 de la Charte, permet au
16gislateur de dialoguer avec le pouvoir judi-
ciaire et ce, A l’intrrieur du contexte institu-
tionnel.
* Associate Professor of Law and Political Science, University of Toronto. Formerly Deputy
Director of Constitutional Law and Policy, Ministry of the Attorney General of Ontario. The author
represented the Province of Ontario in its intervention in the Supreme Court of Canada in the case
under discussion. This paper was originally prepared for a symposium on the sign law cases in
March of 1989 at the Faculty of Law, McGill University and later presented at a Faculty Workshop
at Faculty of Law, Queen’s University in January of 1990. I would like to thank Rhonda Jansen
for her research assistance as well as Peter Benson, David Dyzenhouse, Stefan Dupr6, Allan
Hutchinson and Rod MacDonald for their interest in and comments upon the paper.
McGill Law Journal 1990
Revue de droit de McGill
McGILL LAW JOURNAL
[Vol. 35
Synopsis
I.
II.
Introduction
Background
A. Quebec’s Use of the Override
B. The Quebec Court of Appeal Ruling
IM. Quebec’s Argument in the Supreme Court of Canada: Section 33 as
the Preservation of Legislative Supremacy
IV. The Supreme Court of Canada Ruling
A. Reasons for Judgment
B. The Judgment’s Critical Perspective
1. The Specificity Issue
2. Omnibus Application
3.
The Retrospectivity Issue
V.
Section 33: An Integrated Instrument of Rights Protection
A. History Revisited: From Debate to Text
B. The Institutional Framework of the Charter: Sections I and 33
VI. Conclusion
*
* *
1990]
LIVING WITH THE OVERRIDE
A contingent and limited power bestowed on any organ of government to shift the
orientation of authority ought to be exercised consciously, deliberately, and visi-
bly. The greater the consequences of the shift, the greater the need for observance
of the principle. Making the agent spell out the nature of what is done focuses his
(or its) attention and at the same time assists critical assessment by others. Both
facets of responsible action –
are
implicated.
careful reflection and the need to respond –
-Albert S. Abel (1976)
The Liberals … have learned the hard way, through last year’s debate about lin-
guistic rights, that arguing for the notwithstanding clause is much more painful on
a case-by-case basis, while the Pquistes got away with their general opting out
of the Charter of Rights.
-Lise Bissonnette (February 1990)
I.
Introduction
The momentous decision of the Supreme Court of Canada in Ford v. A.G.
Quebec’ is the first case in which the Court had occasion to interpret the over-
ride clause.2 The case is well-known for the invalidation of Quebec’s legislation
mandating French only in commercial signs, holding that the freedom to use the
language of one’s choice in the commercial marketplace was constitutionally
protected as freedom of expression.3 Premier Robert Bourassa responded to the
decision by invoking the notwithstanding clause to shield his revised language
policy from Charter review. He thereby precipitated a public outcry on the
incongruity of a legislative override in a rights-protecting instrument. Ironically,
the Court’s discussion of the override power has elicited little comment.
Nonetheless, this part of the Ford judgment can provide the indispensable back-
‘Ford v. A.G. Quebec, [1988] 2 S.C.R. 712, 54 D.L.R. (4th) 577 [hereinafter Ford cited to
S.C.R.].
2The text of s. 33 reads as follows:
(I) Parliament or the legislature of a province may expressly declare in an Act of
Parliament or of the legislature, as the case may be, that the Act or a provision thereof
shall operate notwithstanding a provision included in section 2 or sections 7 to 15 of
this Charter.
(2) An Act or a provision of an Act in respect of which a declaration made under this
section is in effect shall have such operation as it would have but for the provision of
this Charter referred to in the declaration.
(3) A declaration made under subsection (1) shall cease to have effect five years after
it comes into force or on such earlier date as may be specified in the declaration.
(4) Parliament or the legislature of a province may re-enact a declaration made under
subsection (1).
(5) Subsection (3) applies in respect of a re-enactment made under subsection (4).
3The Court found freedom of expression to be infringed under both the Quebec Charter of
Human Rights and Freedoms, R.S.Q. 1977, c. C-12 [hereinafter Quebec Charter] and the Canadian
Charter of Rights and Freedoms, Part I of the Constitution Act, 1982, being Schedule B of the
Canada Act 1982 (U.K.), 1982, c. 11 [hereinafter Charter]. It also found that the impugned leg-
islation could not be justified as a limitation on protected rights under either instrument.
REVUE DE DROIT DE McGILL
[Vol. 35
ground for our national re-consideration of s. 33, by illuminating the unseen pull
of historical presuppositions on our understanding of legal constructs and the
extent to which imported political and legal controversies plague our vision of
the Charter. Perhaps, in studying this aspect of the Ford judgment, we can begin
to appreciate the unique and distinctively Canadian qualities of the Charter’s
particular mode of rights protection.
II. Background
A. Quebec’s Use of the Override
The story begins with the exclusion of Ren6 Levesque’s Parti Qufbecois
government from the discussions culminating decades of effort to reach an
agreement on patriation of the Constitution and adoption of a bill of rights. The
representatives of the other ten governments in November of 1981 forged the
consensus that precipitated the most extensive changes ever made to the
Canadian Constitution, a key element of which was to insert the notwithstand-
ing clause into the Charter system of rights protection.4
In its resentment at both the accord and its product, the Quebec govern-
ment seized on the key component of the compromise, the override power, to
shield the province –
from the Charter’s
entrenchment. It embarked on a policy of blanket override of constitutional
rights, using the override clause much like an opt-out clause.
to the greatest extent possible –
The override clause proved to be a cumbersome instrument to this end,
because it required not only an express declaration of override in each overrid-
ing enactment, but specification of the rights overridden as well. To meet the
specificity requirement, Quebec used what came to be called the “standard over-
ride clause,” making reference to all the rights subject to the override.5 To meet
the express declaration requirement a variety of strategies were devised. By an
omnibus amendment enactment,6 the National Assembly repealed and
re-enacted, with the addition of this standard clause, all pre-Charter legislation.
It then inserted this clause into all post-Charter enactments. 7
4The story is
told, from 1976 on,
in R.J. Romanow, J. Whyte & H. Leeson,
Canada…Notwithstanding: The Making of the Constitution, 1976-1982 (Toronto: Carswell, 1984).
See also K. Banting & R. Simeon, eds, And No One Cheered: Federalism, Democracy and the
Constitution Act (Toronto: Methuen, 1983). See infra note 58.
5The text of the “standard override” stated:
This Act shall operate notwithstanding the provisions of sections 2 and 7 to 15 of the
Constitution Act, 1982 (Schedule B of the Canada Act, chapter 11 in the 1982 volume
of the Act of Parliament of the United Kingdom).
An Act respecting the Constitution Act, 1982, S.Q. 1982, c. 21. Assented to June 23, 1982. (First
reading on May 5, 1982; second reading June 2, 1982, and third reading on June 23, 1982.)
6Ibid.
7Quebec’s override policy therefore had three features. The first was the standard form override
clause making reference to all the Charter sections subject to the override. The second was the
1990]
LIVING WITH THE OVERRIDE
The omnibus enactment came into force on June 23, 1982, three months
after the Charter came into effect. To prevent assertion of any of the Charter
rights in question with respect to that period, the National Assembly purported
to give the override clauses retroactive operation to April 17, 1982, the day that
the Charter came into force!
B. The Quebec Court of Appeal Ruling
The Quebec Court of Appeal invalidated the standard override, not in the
Ford judgment, but in Alliance des Professeurs de Montrial v. A.G. Quebec.9
In three separate sets of reasons, for a four member panel, the Court unani-
mously determined that the standard override failed to satisfy the specificity
requirement set out in section 33.” The reasoning is a pioneering attempt to
interpret an important component of the Charter’s institutional structure with
almost no previous judicial guidance. The Supreme Court of Canada’s failure to
consider this attempt seriously is one of the unexplained features of its later
treatment of the same issues.
In the most lengthy analysis, Jacques J.A. treats the question of interpreting
the features of s. 33 as a question of first impression, building his interpretation
of its text on a theoretical foundation. He begins by invoking the special con-
stitutional role of the courts, in particular the special mode of interpretation
which the Constitution demands. Referring to the writings of theorists of con-
stitutional democracy, such as John Rawls, Friedrich Hayek and Ronald
omnibus legislation by which all existing Quebec legislation was repealed and re-enacted with a
standard override clause. The third feature was the routine addition of the standard override clause
in all new pieces of legislation and all new amendments to existing legislation. I refer to this policy
of override as “comprehensive”. See infra note 58.
‘More precisely, the National Assembly repealed and re-enacted each statute originally enacted
before April 17, 1982 to insert a standard override clause. This amending legislation came into
force on June 23, 1982; the amendment came into force on April 17, 1982. Similarly, each statute
adopted in the period between April 17, 1982 and June 23, 1982 acquired, by the same amending
legislation, a standard override clause. These amendments took effect on the date that the enact-
ment in question came into effect. See sections 1 to 7 of an Act Respecting the Constitution Act,
1982, supra, note 5, reproduced in Ford, supra, note 1 at 727-28.
9[1985] C.A. 376, (1985) 21 D.L.R. (4th) 354 [hereinafter Alliance]. This was a decision of the
Quebec Court of Appeal, from which the Attorney General of Quebec had secured leave to appeal
to the Supreme Court of Canada (Chouinard, Lamer & La Forest JJ.) 30 September 1985. The
appeal was never perfected. Justice Lamer was critical, during oral argument, of Quebec’s failure
to bring this case to the Supreme Court. When counsel for the Attorney General of Quebec noted
that the same issues as to the comprehensive override policy were before the court in the cases at
bar, Justice Lamer pointed out that while the issues were contained in the other cases, the failure
to appeal that particular case meant that the parties to it were not before the Court. The Supreme
Court judgment refers to this background in Ford, supra, note 1 at 736-37.
‘Chief Justice Crete presided at the hearing of argument but did not take part in the judgment.
Kaufman J.A. concurred with the judgment of the court. The separate reasons of-Justices Jacques,
Mayrand and Vallerand are the subject of the discussion that follows.
McGILL LAW JOURNAL
[Vol. 35
Dworkin, he concludes that a judge must read the Constitution of Canada, the
source of all state law-making power, as protecting fundamental freedoms in the
exercise of law-making. Law, in other words, is understood not merely as the
product of majoritarian decision-making, but as a system constraining legisla-
tive preferences to certain values. So understood, legislative supremacy is not
the rule, but the exception to the rule; the Constitution, as per s. 52 of the
Constitution Act 1982, reigns supreme.”
Regarding the text of section 33 as an exception to the Constitution’s fun-
damental values in this way, Jacques J.A. reads into the specificity requirement
of s. 33 a stipulation going beyond simple reference to the section numbers of
the Charter sections overridden. He highlights the textual basis for this view
contrasting the reference to the overriding provision with the reference to the
overridden provision: 2
textes dirogatoires
textes auxquels on diroge
une loi…celle-ci ou
une de ses dispositions.
une disposition donn6e de
l’article 2 ou des articles 7
15. [sic]
He also notes that the French, “indrpendamment d’une disposition donn6e”,
marks the same kind of specificity for the overridden provisions as does the
English. While a legislature is free to specify more than one of the rights or free-
doms subject to the override, mere reference to section numbers of the Charter
does not satisfy the requirement of specificity. Indeed, reference to section num-
bers undermines the very purpose of the stipulated form of specification,
namely, “af’m de confronter bien clairement la loi de derogation et les droits
dont les justiciables sont priv6s. ‘”3 To support this interpretive conclusion,
Jacques J.A. again makes reference to the matrix of foundational values in
which one finds s. 33, namely a constitutional document that stipulates in s. 1
that the character of our society is to be free and democratic.’ Referring again
“Ibid., C.A. at 378, D.L.R. at 358. Section 52 states:
(1) The Constitution of Canada is the supreme law of Canada, and any law that is
inconsistent with the provisions of the Constitution is, to the extent of the inconsist-
ency, of no force or effect.
121bid.,
C.A. at 380, D.L.R. at 361. The translation reads:
overriding provisions
… an Act … the Act
or a provision …
thereof…. [sic]
provisions overridden
… a provision included in
section 2 or sections 7
to 15 of this Charter. [sic]
131bid. The English translation reads: “bring into sharp focus the effect of the overriding provi-
sions and the rights deprived.” C.A. at 380, D.L.R. at 361.
4For a similar understanding of the closing words of s. 1 of the Charter, see Dickson C.J.C.
in R. v. Oakes, [1986] 1 S.C.R. 103 at 136, 26 D.L.R. (4th) 200 at 225 [hereinafter Oakes cited
to S.C.R.]:
1990]
LIVING WITH THE OVERRIDE
to the work of John Rawls, he observes that democracy as a constitutional value
is not simply majority rule, but a system premised on freedom of speech and
assembly, thought and conscience, so that political decisions are reached
through rational discussion rather than at the insistence of special interests.’ 5
Jacques J.A. refers to Re Alberta Legislation for Canadian judicial evocation of
the centrality of reasoned debate as the foundation of our democratic function-
ing. 6 He also points out that the override cannot reach the exercise of demo-
cratic rights guaranteed under s. 3 of the Charter.
Reading the constitutional text in light of these fundamental considera-
tions, Jacques J.A. states,
Ce libre drbat du citoyen sur l’action lgislative et gouvemementale ne peut
s’exercer que si l’inforrnation nrcessaire a 6t6 clairement fournie. Dans l’esp~ce,
il s’agit, d’une part, d’indiquer le droit prrcis opposable A la 16gislature dont on
veut priver le citoyen dans le cas d’une 16gislation particulire; et, d’autre part, de
ddmontrer un rapport entre l’un et l’autre. 17
He concludes that the omnibus and standard clause exercise of the override
power by the National Assembly must fall because “les crit~res d’information
A second contextual element of interpretation of s. 1 is provided by the words “free
and democratic society”. Inclusion of these words as the final standard of justification
for limits on rights and freedoms refers the Court to the very purpose for which the
Charter was originally entrenched in the Constitution: Canadian society is to be free
and democratic. The Court must be guided by the values and principles essential to a
free and democratic society….
15At this crucial stage in Jacques J.A.’s analysis, a quotation from J. Rawls, A Theory of Justice
(Cambridge, Mass.: Belknap Press, 1971) at 225, setting out the need for rational discourse in polit-
ical determinations, is incomplete, as it was in the originally released reasons for judgment. See
C.A. at 380, D.L.R. at 362. The full quotation, with the omitted words italicized, reads as follows:
We may take for granted that a democratic regime presupposes freedom of speech and
assembly, and liberty of thought and conscience. These institutions are not only
required by the first principle of justice but, as Mill argued, they are necessary if polit-
ical affairs are to be conducted in a rational fashion. While rationality is not guaranteed
by these arrangements, in their absence the more reasonable course seems sure to be
rejected in favour of policies sought by special interests. If the public forum is to be
free and open to all, and in continuous session, everyone should be able to make use
of it. All citizens should have the means to be informed about political issues. They
should be in a position to assess how proposals affect their well-being and which pol-
icies advance their conception of the public good. Moreover, they should have a fair
chance to add alternative proposals to the agenda for political discussion.
16[1938] S.C.R. 100 at 133, 146, 2 D.L.R. 81 at 107, 119 in Alliance, supra, note 9, C.A. at 381,
D.L.R. at 363.
‘7Aliance, ibid., C.A. at 382. The D.L.R. translation at 364 reads as follows:
Citizens can only exercise their right to discuss legislative and government action
freely if the necessary information has been clearly provided. This case concerns the
issue, on the one hand, of indicating which specific right, disputed by the Legislature,
the citizen is being deprived of in a given piece of legislation and, on the other hand,
of demonstrating the link between the legislation and the right.
REVUE DE DROIT DE McGILL
[Vol. 35
et de rapport” stipulated by s. 33 are lacking.”8 Since reliance upon s. 33 trans-
forms the legal guarantee of the rights in issue, justiciable under s. 24 of the
Charter, into mere interests at risk to political determinations, courts must mon-
itor the formal strictures set down by the constitutional text strictly.
Mayrand J.A., who concurs with Jacques J.A., begins his short judgment
by commending the National Assembly for its simplification of the task of
applying the override to the entire body of Quebec statute law. But praise soon
dissolves as he moves to align himself with the views of Jacques J.A., condemn-
ing the exercise for failing to conform both “A la lettre et A l’esprit” of s. 33.I9
The Charter’s override, he points out, is to be contrasted with global override
powers, such as those in the Canadian Bill of Rights” and the Quebec Charter’
In its novel specificity requirement, s. 33 is designed to illuminate the rights and
freedoms withdrawn from Charter protection in order to encourage “un examen
6clair6 et s6rieux de la d6rogation propos~e” 2
While s. 33 has this informational function in order to promote informed
debate, Mayrand J.A. makes it clear that the stricture is a matter of form, not
of substance. This formal quality does not, however, invite judicial deference.
‘lbid., C.A. at 382, D.L.R. at 365 “criteria of information and of relationship”. He offers two
examples that illustrate the absence of “linkage”. In the first, Jacques J.A. states that “il n’y a aucun
rapport apparent et aucun n’a 6t6 d~montr”
(“there is no apparent link, and none has been dem-
onstrated”) between the Loi sur la raffinerie de sucre d Quibec, 1982, S.Q. 1982, c. 28 and the
right of freedom of conscience and of religion set out in s. 2(a) of the Charter. Ibid., C.A. at 382,
D.L.R. at 364. His point appears to be that reference to section numbers, perhaps particularly by
the standard clause format, gives no information about the actual operation of the override decla-
ration in its derogation from enjoyment of rights guarantees. The link is not obvious and it is not
provided.
The second example is somewhat different: Jacques J.A. concedes that there may be a link
between the freedom contained in s. 2(a) and the Loi concernant le recensement des 6Iecteurs pour
l’annge 1982, S.Q. 1982, c. 34 but he states that “none has been demonstrated”, ibid. Here again,
it is not entirely clear whether it is the standard clause format or the reference to section numbers
per se that infringes the specificity requirement. The underlying point, however, is clear. One must
read the specificity requirement in light of its purpose, and that purpose in light of the larger ideas
of constitutionalism invoked. The comprehensive application of the override, including the stand-
ard clause declaration and the omnibus and routine application, is inconsistent with both the text
and purpose of s. 33.
‘9Alliance, ibid., C.A. at 383, D.L.R. at 355: to the “letter and spirit”.
20R.S.C. 1970, App. III, s. 2 empowers Parliament to declare that a statute “shall operate not-
withstanding the Canadian Bill of Rights”. It is interesting to note that it is not only the subordi-
nated enactment that is referred to in its entirety in this formula. The overriding enactment itself
is to override in toto because there is no provision for invoking the non obstante clause for only
part.21Supra, note 3. Section 52 states: Sections 9 to 38 prevail over any provision of any subsequent
act which may be inconsistent therewith unless such act expressly states that it applies despite the
Charter.
2AIliance, supra, note 9, C.A. at 383, D.L.R. at 356: “an enlightened and serious examination
of the proposed derogation”.
1990]
LIVING WITH THE OVERRIDE
On the contrary, the specificity requirement precipitates judicial verification of
adherence to the form stipulated because, whether there is an informed and rig-
orous debate on the override declaration or not, a valid exercise of the override
withdraws legislation from judicial review for Charter conformity.’
Vallerand J.A., in a short set of reasons, joins his colleagues in their view
of the primacy of rights protection under the Charter. He notes the precision
with which the text delineates the method for departing from those rights, espe-
cially the specification of the right in question, which, in his view, is designed
to facilitate an informed democratic discussion because citizens have the oppor-
tunity to understand, consider and discuss the deprivation contemplated. The
override declarations under review, because of their omnibus and standard
clause features, fail to come within the power granted because they do not iden-
tify clearly those rights subject to the derogation. The failing, however, is not
one of democratic function, which lies outside judicial review, but consists in
the non-compliance with the Charter’s mandatory requirement of particularized
procedure.
The reasoning of all three judgments shares a commitment to the primacy
of the Constitution, including the Charter, as a framework for the exercise of
political power. Within this framework, the override is understood not as a rem-
nant or preservation of unchecked parliamentary sovereignty; it is rather a
mechanism that can work against values having pre-eminent standing in our
society. While it is triggered by the majority, it does not derive legitimacy for
that reason. Instead, the override offers a carefully formulated exception to the
primacy of rights protection, checked by judicial review as to form, with the
understanding that the form stipulated is conducive to informed and reasoned
debate. It is this possibility of rich democratic debate that gives legitimacy to
s. 33 despite the fact that by its exercise particular rights are removed from the
legal realm of constitutional protection and left to the give and take of the polit-
ical forum.
This reasoning is within the stream of Charter jurisprudence described as
“purposive” by the Supreme Court of Canada. It resonates with that Court’s
descriptions of the interpretive task of judges in its invocation of the
Constitution as the highest order of law, constraining every exercise of power
by the state and enlisting judges as its guardians.24 The sensitivity to the text of
23In failing to meet the formal requirements of specificity set down by the terms of s. 33, the
override clauses are rendered ultra vires and void. Ibid.
24For example, see Hunter v. Southam Inc., [1984] 2 S.C.R. 145 at 156, 11 D.L.R. (4th) 641 at
650, R. v. Big MDrug Mart, [1985] 1 S.C.R. 295 at 344, 18 D.L.R. (4th) 321 at 359-60 [hereinafter
Big M Drug Mart cited to S.C.R.], Oakes, supra, note 14 at 119, for Charter cases and Reference
Re Language Rights under the Manitoba Act, 1870, [1985] 1 S.C.R. 721 at 744-45, 751, 19 D.L.R.
(4th) 1 at 19, 24 [hereinafter Manitoba Language Reference cited to S.C.R.]. Some of the argu-
ments are relatively weak, or rather, underwritten. One is hard pressed to accept, as presented, the
McGILL LAW JOURNAL
[Vol. 35
s. 33 lies within this interpretive stream, although reference to other features of
s. 33, such as the sunset provision that coincides with election frequency and the
requirement of “express” declaration, might have bolstered the interpretation
offered.
This contextual vision of Canadian constitutionalism invokes the ultimate
values of our society and the judiciary as their guardian, such that the constitu-
tional text forwards the idea of free and democratic nationhood. These features
of the Quebec Court of Appeal’s approach carry forward the Supreme Court of
Canada’s stated vision of the Charter and the judicial role predicated by it. As
we shall see, the Supreme Court of Canada’s treatment of the National
Assembly’s omnibus, routine and standard clause use of the override does not.
m. Quebec’s Argument in the Supreme Court of Canada:s Section 33 as
the Preservation of Legislative Supremacy
In Ford, the province of Quebec took issue with the interpretation of s. 33
put forward by the Quebec Court of Appeal in Alliance des Professeurs de
Montr-al, arguing that the notwithstanding clause preserves full legislative sov-
ereignty with respect to those rights subject to it. Counsel for the Attorney
General of Quebec took comfort in both the pre-Charter history of legislative
supremacy in Canada and the crucial role the override played in the final com-
promise leading to adoption of the Charter. So understood, the exercise of the
notwithstanding clause was to be constrained only by political considerations,
free from judicial scrutiny.
Legal support for this view was offered. A parallel to such unreviewable
power was suggested, namely s. 92(10)(c) of the Constitution Act, 1867, under
which Parliament is empowered to declare a work or undertaking to be for the
general advantage of Canada and thus bring it within exclusive federal legisla-
tive jurisdiction.26 The Court’s differentiation between legal and political sanc-
argument that the democratic system ensconced by the right to vote in s. 3 of the Charter (which
is beyond the reach of s. 33) subsumes full exercise of the right to freedom of expression, when
the override itself can nullify the protection of freedom of expression itself. I am indebted to Rod
MacDonald for pointing out that the Quebec Court of Appeal’s approach may derive from a civil-
ian sensitivity to the text of the Charter.
25For the sake of clarity I combine written and oral submissions.
26Section 91(29) of the Constitution Act, 1867 (U.K.), 30 & 31 Vict., c. 3 includes in the cat-
alogue of Parliament’s exclusive powers:
Such Classes of Subjects as are expressly excepted in the Enumeration of the Classes
of Subjects by this Act assigned exclusively to the Legislatures of the Provinces.
Section 92(10)(c) excepts from the provincial catalogue of exclusive powers:
Such Works as, although wholly situate within the Province, are before or after their
Execution declared by the Parliament of Canada to be for the general Advantage of
Canada or for the Advantage of Two or more of the Provinces.
See P. Hogg, Constitutional Law of Canada, 2d ed. (Toronto: Carswell, 1985) at 484 for discussion
of these provisions.
1990]
LIVING WITH THE OVERRIDE
tions, in the Reference Re Resolution to Amend the Constitution,27 also demon-
strated the importance of discerning the appropriate forum for reviewing an
exercise of constitutional powers.’
Quebec argued that the formal requirements prescribed in section 33 dem-
onstrated the unreviewable character of the legislative prerogative in s. 33 by
requiring traditional legislative form without expressly excluding standard form
or omnibus usage. More particularly, the various elements prescribed in the text
of s. 33 imposed merely formal strictures on the exercise of the override, the
only relevant one being the stipulation of those Charter sections which were to
yield. This stipulation did not assure a more informed democratic debate, as the
Quebec Court of Appeal had stated, because the difference between reference
to the Charter text by section number or by recapitulation of the text was a
purely formal distinction. Since the text allowed subordination of sections 2 and
7 through 15 of the Charter, reference to any or all of those section numbers
would suffice. Counsel for the Attorney General of Quebec pointed out that the
very values upon which the Court of Appeal based its holding, namely the rights
instrumental to democratic participation, as set out in s. 2(b) of the Charter,
were subject to the override. Moreover, requiring the override declaration to
specify the right or freedom actually subordinated would place too high a bur-
den on legislatures while empowering courts with a forbidden review power.
IV. The Supreme Court of Canada’s Ruling
A. Reasons for Judgment
The Supreme Court of Canada upheld Quebec’s omnibus, routine and
standard clause override, but invalidated the early retrospective application
from April 17 to June 23, 1982.29 The judgment offers little reasoning for these
conclusions and has attracted minimal public attention or scholarly interest.3
27A.G. Man. v. A.G. Can.; A.G. Can. v. A.G. Nfld; A.G. Que. v. A.G. Can.; A.G. Can. v. A.G.
Que., [1981] 1 S.C.R. 753, 125 D.L.R. (3d) 1 (sub nom. Ref. re Amendment of the Constitution
of Canada) [hereinafter Patriation Reference cited to S.C.R.].
2This was an odd argument, but did not attract any attention. In that case the Supreme Court
of Canada deliberated not only on the legal sanctions to “unilateral” patriation –
i.e., a request
for patriation legislation to the Westminster Parliament without consent of a broad segment of
Canadian political constituencies, but also upon the political propriety of the contemplated course
of action by the federal government. Its ruling that the undertaking did not conform to constitu-
tional convention exemplified an exercise of judicial review of political powers that would seem
to cut against Quebec’s position. While one could argue the absence of a power to invalidate the
product of an act of political impropriety, judicial power to declare the impropriety was inconsist-
ent with the attempt to foreclose judicial review altogether.
29The retrospectivity argument was raised only in the factum of the Attorney General for
Ontario, Intervenor. My notes indicate that counsel for the Attorney General for Quebec did not
make reference to the retrospectivity point in his oral argument.
30See G. Marshall, “Taking Speech Rights for an Override: Free Speech and Commercial
Expression” (1989) Public Law 4.
REVUE DE DROIT DE McGILL
[Vol. 35
Nevertheless, this ruling has far-reaching consequences for the institutional
framework of rights protection in Canada and, more particularly, the current
claims as to the impropriety of an override clause in a rights-protecting consti-
tution. The Court’s determination –
that section 33 contains only formal pre-
requisites to the subordination of Charter rights and, for that reason, tolerates
no judicial review whatsoever –
adopts, without argument, the distinct view
that s. 33 is inimical to the Charter’s guarantees. Later in this paper, I suggest
a different approach, namely that the legislative role under the notwithstanding
clause coheres with Canada’s innovative constitutional system of rights protec-
tion. Before developing that idea, I set out the Supreme Court’s treatment of
Quebec’s use of the override between 1982 and 1985.
At the outset, the Court emphasized that it took no position as to the “con-
stitutional perspective” from which to interpret s. 33. It was not “particularly
relevant or helpful” for purposes of interpretation, we are told, to consider
whether s. 33 continued pre-Charter legislative supremacy, on the one hand, or
marked the culmination of a “fully informed democratic process”, on the other.”1
Instead, the Court emphasized the formal quality of the requirements for exer-
cise of the override and deduced from that formality severe limits upon the judi-
ciary’s ability to review exercise of the override.
Turning first to the question of specification, the Court determined that the
legislature need not relate the subordinating enactment to the particularized sub-
ordinated right or freedom. Such a reading would require “a prima facie justi-
fication of the decision to exercise the override authority rather than merely a
certain formal expression of it”. 2 The judicial role precipitated by a requirement
of justification did not follow from the terms of s. 33. Such a requirement was
also unacceptable because it would exact an impossible burden on legislatures,
which might not be able to determine with any certainty the rights affected.
The “essential” form required by s. 33, the Court determines, is express
declaration. Should a legislature intend to override only part of a provision, the
specification of that part would of course suffice. Reference by section, subsec-
tion or paragraph number is sufficiently express, however, given that it is the
standard form of statutory reference used in legislative drafting for purposes of
amendment and repeal. There is “no reason why more should be required”
because this kind of reference would provide “sufficient indication to those con-
cerned of the relative seriousness of what is proposed”.3 The Court goes on to
say that “express” declaration could not have been “intended” to include a
31Ford, supra, note 1 at 740.
32Ibid.
33Ibid. at 741. This idea that the function of s. 33 is to signal a serious political act is repeated
twice but remains unclear. How can the seriousness of the action be conveyed when the legisla-
ture’s view of the effect and implications are not?
1990]
LIVING WITH THE OVERRIDE
restatement in words of the Charter fights overridden because, in the case of the
standard override clause used by Quebec, such a restatement would encumber
the declaration with “a very long recital indeed”. 4
The Court affirms the omnibus use of the override on similar grounds. The
insertion of standard clause override declarations into a large body of statutes
by a single enactment is valid, the Court reasons, because any other conclusion
would lead the judiciary into review of the permissibility of legislative policy
as to exercise of the override.” Accordingly, no review of the “routine” use of
the override is permissible.
After accepting the standard form declaration and its omnibus and routine
enactment, the Court turns to retroactive application. After noting that s. 33(1)
was “not without ambiguity” as to prospective or retrospective application, the
Court invokes the principle of statutory construction that where a provision can
be read prospectively or retrospectively a judge is to read it prospectively.36 It
is on this basis that the Court strikes down the retroactive application of the
override. While the opinion includes a quotation from a text on statutory inter-
pretation, which notes the law’s aversion to interference with vested rights and
obligations, the judgment gives no account of the principle in the constitutional
context.37
B. The Judgment’s Critical Perspective
1.
The Specificity Issue
The most remarkable feature of the Court’s discussion is the sparseness of
its interpretive technique. Two pages in a lengthy opinion dispose of a major
question of constitutional structure. Aside from briefly noting that the phrases
“shall operate notwithstanding” and “a effet indrpendamment” in s. 33(1) allow
for retroactive or prospective operation, there is no examination of the wording
or design of s. 33. There is no reference to other Charter sections. There is no
consideration of the institutional roles implicit in the Charter’s structure. There
are no references to the modes of interpretation established in other Charter
341bid.
3Slbid. at 743.
361bid. at 744-45.
371bid. The passage quoted is from P.-A. C6tr, The Interpretation of Legislation in Canada
(Cowansville: Yvon Blais, 1984) at 96 which includes the following passage from Wright J.’s dic-
tum in Re Athlumney, [1898] 2 Q.B. 547 at 551-52 (Q.B.):
Perhaps no rule of construction is more firmly established than this –
that a retrospec-
tive operation is not to be given to a statute so as to impair an existing right or obli-
gation, otherwise than as regards matter [sic] of procedure, unless that effect cannot be
avoided without doing violence to the language of the enactment. If the enactment is
expressed in language which is fairly capable of either interpretation, it ought to be
construed as prospective only.
McGILL LAW JOURNAL
[Vol. 35
cases. There is no mention of the academic literature on s. 33.3′ The often con-
clusory analysis reads like routine statutory interpretation.39 One has no sense
that it is a Constitution under interpretation or that the Court sees itself as its
guardian.’ The Court also ignores the opportunity to build upon the distinction
it had earlier expressed between limitations upon fights, justifiable under s. 1,
and exceptions to those rights, available only through political action, either
under s. 33 or by constitutional amendment.”
Especially perplexing is the lack of reference to the Charter text under
examination. The Charter differs from the Canadian Bill of Rights and the
Quebec Charter, for example, in that the entire instrument is not open to over-
ride.42 In contrast to these instruments, s. 33 provides that the override operate
“notwithstanding a provision included in section 2 or sections 7 to 15”. And it
so states twice.43 Moreover, the text also states that the declaration must be made
“expressly” and ceases to have effect after a maximum five year period, subject
to renewal by re-enactment. These terms, individually and collectively, are
restrictive. The Court nonetheless sees no significance in these features and
finds in s. 33 an easily exercised formality.”
38Commentators, for example, have suggested that substantive values are attached to exercise of
the override by arguing that the strictures of s. 1 justification apply. See B. Slattery, “Canadian
Charter of Rights and Freedoms – Override Clause Under Section 33 – Whether Subject to
Judicial Review Under Section 1” (1983) 61 Can. Bar Rev. 391, and D.J. Arbess, “Limitations on
Legislative Override Under the Canadian Charter of Rights and Freedoms: A Matter of Balancing
Values” (1983) 21 Osgoode Hall L.J. 113. This view has not received wide support. It is rejected,
for example, by Hogg, supra, note 26 at 690-91, and D. Gibson, The Law of the Charter: General
Principles (Toronto: Carswell, 1986) at 130-31. For general discussion of the political dimension
of s. 33, see “The Charter and s. 33: Holding Politicians Accountable” (1987) 3 Admin. L.J. 21
and D. Greschner & K. Norman, “The Courts and s. 33” (1987) 12 Queen’s L.J. 153. For the view
that the United States Supreme Court’s current retreat from the activism of the Warren Court dem-
onstrates the utility of s. 33 as a rights-protecting mechanism, see A. Petter, Policy Options, April
1990 at 33.
39Three times the Court states there is “no warrant” for a particular interpretation. See Ford,
supra, note 1 at 740-44. We are told that stating the provision to be overridden in the words of the
Charter “cannot have been intended by the use of the word ‘expressly’…” in the text of s. 33.
4Compare, for example, the detailed discussion leading up to the consideration of a heading in
the Charter’s text in Skapinker v. Law Society of Upper Canada, [1984] 1 S.C.R. 357 at 370-79,
9 D.L.R. (4th) 161 at 171-78 [hereinafter Skapinker] or the consideration of text and theory inform-
ing the articulation of the role of the courts under section 1, in Oakes, supra, note 14 at 135-40.
41A.G. Quebec v. Quebec Association of Protestant School Boards, [1984] 2 S.C.R. 66 at 86, 10
D.L.R. (4th) 321 at 327. This case was, however, referred to elsewhere in the reasons for judgment:
see Ford, supra, note 1 at 771.
9, C.A. at 383, D.L.R. at 356. See text, supra, at notes 20 and 21.
42This point is made by Mayrand J.A., in the Quebec Court of Appeal, see Alliance, supra, note
43Ss. 33(1), (2). See text, supra, note 2.
“The specificity feature of s. 33 can be traced to its announcement at the First Minister’s
Conference on 5 November, 1981. The “Fact Sheet” distributed at that time, repeatedly describes
the overridden provision of the Charter as a “specific provision” and appears to contrast this spe-
cific override technique with that found in other instruments, including the Canadian Bill of Rights.
1990]
LIVING WITH THE OVERRIDE
The Court’s key analytic tool is its characterization of the override as
“formal”. This quality is nowhere defined or elucidated, but one thing is clear:
because the override is “formal” it admits no “substantive review of the legis-
lative policy in exercising the override authority in a particular case.”’45
The spectre of impermissible judicial review pervades the Court’s treat-
ment of the standard clause and omnibus character of Quebec’s use of the over-
ride. The Court appears to assume that a requirement of specificity, i.e., refer-
ence to the actual right or freedom infringed by the enactment in which the
override declaration appears, would necessarily engage the judiciary in substan-
tive review of these declarations. This assumption is unsubstantiated, however.
Specificity in exercising the override can be, indeed must be, a formal matter
entirely divorced from the merits or demerits of using the override in a partic-
ular case. The judicial role in determining whether the specificity requirement
has been satisfied implies no judgment concerning the justification of the leg-
islative policy of override. Judicial review would merely ensure that the over-
ride complies with the stipulated elements of its form. The Court’s failure to see
this distinction is odd not only because it is evident but because it is fundamen-
tal to the appellate court ruling.
It is true that by requiring specificity the Court might eventually find itself
called upon to enunciate the requisite standard of specificity. Would reference
to a specific right be enough? Would there be an evidentiary requirement (e.g.,
an opinion, a court decision, a legislative debate) demonstrating that the legis-
lature had considered the Charter impact of the policy? Even this more active
review could escape the badge of “substance” by looking for some evidence
rather than requiring that the legislature do it right. Moreover, the determination
of the degree of specificity required for overriding rights would depend on the
significance the Court attaches to rights generally, rather than on the merits of
particular legislative policies.
At any rate, wherever one might draw the line on specificity, the compre-
hensive override policy would clearly fall on the impermissible side. Quebec’s
use of the override manifested a broad policy of negating all Charter sections
subject to the override rather than a specific policy in regard to particularized
Charter rights. As the Court had no apparent compunction in pointing out else-
where in the same judgment, the override in issue “appears to have been enacted
See A. Bayefsky, Canada’s Constitution Act 1982 & Amendments: A Documentary History
(Toronto: McGraw-Hill, 1989) at 904-05. Previous discussions had contained a draft of opt-in pro-
visions (for the provincial legislatures) and proposals for an override. Roger Tass6, the Deputy
Minister of Justice of Canada during the formulation of the Charter, appears to take the view that
s. 33 contemplates specific reference to the particular right overridden. See R. Tass6, “Application
of the Canadian Charter of Rights and Freedoms” in G.A. Beaudoin & E. Ratushny, eds, Canadian
Charter of Rights and Freedoms, 2d ed. (Toronto: Carswell, 1989) 65 at 104-06.
45Ford, supra, note I at 740-41.
REVUE DE DROIT DE McGILL
[Vol. 35
as part of the well-established legislative policy and practice at the time of
including the standard override provision in every Quebec statute.”’46 That char-
acterization, which the Court apparently did not regard as an impermissibly
“substantive” determination, would have sufficed to invalidate the comprehen-
sive policy of override, without any need to anticipate where to draw the line
in the future.47
In the Court’s view, a substantive specificity requirement would impose
not only an impermissible task on the courts, but an excessively onerous burden
on legislatures as well. Accordingly, the judgment adduces the difficulty of
accurately specifying the rights affected as an additional reason for rejecting
specificity.” This argument, however, is simply one of cost, the cost of action
and the cost of error. Governments can discover in various ways which rights
are implicated. They can seek legal opinions, precipitate court references or
await determinations in litigation. The Court assumes that s. 33 permits a leg-
islature to subordinate potential rights claims of which it is unaware or uncer-
tain. Moreover, the Court assumes that the state should not bear the risk of error,
even though the invalidation of an erroneous override or the failure to specify
rights affected would mean not merely that the state loses, but that rightholders
win. This cost-benefit analysis is truncated. It lets costs win the day, with no
consideration of the benefits (to the functioning of legislatures and govern-
ments) which informed deliberation might purchase.
The Court overruled the Quebec Court of Appeal’s holding that s. 33
requires that the overriding declaration reproduce the Charter’s formulation of
the overridden rights. On analogy to amendment or repeal of statutes, the Court
decided that reference to the numbers of sections and their component parts suf-
461bido at 736.
4 71n questioning during oral argument, Justice Lamer expressed concern for the problem of
omnibus legislation, e.g., a long statute amending a wide variety of statutes or a very long statute,
such as the Criminal Code. How would a court, on review, be able to ascertain when the override
declaration was sufficiently specific? The Court opted to defer to the legislature’s fulfilment of
“form” requirements in all cases, rather than find unacceptable a policy of clearly non-specific
override.
48in oral argument, several justices expressed concern at the extent of the resources that would
be necessary to effectuate frequent or large-scale overrides. The Court has had no difficulty with
arguments of administrative expediency in other contexts: see Re Singh and Minister of
Employment and Immigration, [1985] 1 S.C.R. 177, 17 D.L.R. (4th) 422 and Reference re Section
94(2) of the Motor Vehicle Act (B.C.), [1985] 2 S.C.R 486,24 D.L.R. (4th) 536. The Court, perhaps
because it did not recognize the parallel between the arguments, did not respond in the override
context in the same way that it did in the rights context, i.e., to conclude that rights cannot be hos-
tage to legislative expediency. The override makes rights vulnerable to a particular design of leg-
islative action, one that exacts a higher cost with increased use. It is to this type of legislative effort
that rights bend, not to a version simplified by concern for cost minimization. I am indebted to Rod
MacDonald for pointing out that the “subtle cautionary vehicles built into s. 33” are analogous to
common law doctrines, e.g., natural justice.
19901
LIVING WITH THE OVERRIDE
fices. In declaring that “[t]here is no reason why more should be required under
s. 33″,49 the Court ignores the cornerstone of its Charter jurisprudence, the dif-
ferences between the statutory and constitutional contexts. The Charter’s guar-
anteed rights and freedoms express the highest values in our legal system, as s.
52 of the Constitution Act, 1982 attests. Marking departure from these rights
and freedoms in the words in which the Constitution expresses them is a way
of acknowledging their normative status. The override is more than a mecha-
nism of legal change. As the judgments in the Quebec Court of Appeal so elo-
quently point out, an override declaration transforms a constitutionally pro-
tected right, justiciable under s. 24 of the Charter, into an interest at risk in the
political arena. To equate exercise of the override power to the modification of
statutes, which are by definition alterable at the hand of the legislature, is to
assume what is in issue. The question is whether legislative supremacy has been
bridled; an analogy to an illustration of legislative supremacy at work cannot
provide the answer.”
The difference between word reference and number reference may appear
to be merely symbolic. Once raised, however, the question cannot be resolved
by invoking considerations of legislative form, because neither kind of refer-
ence is more formal or less substantive than the other. Symbolism, however, is
hardly irrelevant to the interpretation of the Charter, which is now the pre-
eminent constitutive symbol of the Canadian legal order. In fact, the difference
between numbers and words is considerable. The words of a legal instrument
are the bearers of discursive meaning.5 Section numbers, on the other hand, are
49Ford, supra, note 1 at 741.
5The question is begged on a different level as well. The issue is the degree of specificity
required. For Charter sections that contain only one right, reference to the section number is, to
those who are familiar with the text, arguably as informative as reference to the text. But the large
majority of the sections subject to the override contain many rights, so that reference to the section
number, while clear, embraces a number of protected interests. Some of those sections, like 10 and
11, contain subsections; the others do not. Therefore, for a declaration under s. 33 to specify the
particular right in issue, something more than section (or subsection) number reference is required.
An analogy to statutory amendment cannot shake this conclusion because statutory amendment
must by its very nature be as specific as necessary to identify the provisions to be amended. An
appropriate analogy would be amendment of the constitutional text itself: one would not amend
section 11 by reference to section 11 unless one meant to include all of section 11. That type of
specificity is what s. 33, and indeed all statutory amendment, appears to require. Thus the Court’s
analogy works against its conclusion.
51The Court, in the context of the issue of freedom of expression, states: “Language is … inti-
mately related to the form and content of expression.” See Ford, supra, note 1 at 748. In the
Manitoba Language Reference, supra, note 24 at 744, the Court stated:
The importance of language rights is grounded in the essential role that language plays
in human existence, development and dignity. It is through language that we are able
to form concepts; to structure and order the world around us. Language bridges the gap
between isolation and community, allowing humans to delineate the rights and duties
they hold in respect of one another, and thus to live in society.
McGILL LAW JOURNAL
[Vol. 35
the most abstract of abstractions. Numerical reference to overridden rights is a
representational divide that distances the reader from the concrete significance
of the legislative act. The use of numbers neutralizes the enactment process by
insulating legislators from direct appreciation of the specific impingements of
the legislation.52
2. Omnibus Application
The Court’s judgment validated the omnibus override mechanism as well
as the standard form of declaration referable to the widest available range of
Charter rights. A legislature could not only deny all Charter rights subject to
the override for any particular enactment; it could also, by one statute, enact
override clauses into all statutes on the books.53 Because s. 33 sets out merely
formal requirements, distinctions of part or whole, or between comprehensive
and particular enactments, had no significance. The Quebec legislature thus was
not seen to enact overrides against the background of residually subsisting
rights. Rather, its policy, as the Court understood it, was to eliminate all the
Charter rights subject to s. 33. Nonetheless, the Court denied the difference
between overriding Charter rights and overriding, to the extent possible, by a
combination of standard and omnibus declarations, the Charter sections them-
selves.’
The Court dismissed arguments against routine override as “essentially
submissions concerning permissible legislative policy.”55 This comment is
52In the course of argument, Justice Estey asked why a broader use of the override, e.g., the
standard clause used by Quebec referring to all the rights and freedoms subject to the override and
not merely the specified right actually jeopardized, would not be a “clarion call to a higher danger”.
This question, as well as the Court’s ruling, underestimates the normative force of a legislature
enacting –
and debating – not a formulation of numbers and not a standard form clause, but its
decision to enact law inconsistent with: freedom of religion and conscience, the right to free
expression and political participation or the right to life, liberty and security of the person assured
by adherence by the state to the principles of fundamental justice. Justice Estey also asked whether,
in the age of xerox legislation, it was not appropriate for legislatures to automate their override to
the extent possible.
53The omnibus enactment did not even mention the names of the statutes into which the standard
override clause was being inserted. See Alliance, supra, note 9, C.A. at 379, D.L.R. at 360, Jacques
J.A.
54During oral argument, the Court expressed concern at the task of drawing the line between rou-
tine and non-routine use. Would 15 times a year, 20 times a year, 25 times a year be too much?
A similar type of concern had arisen in the Patriation Reference, supra, note 27 on the question
of the requisite degree of provincial content for patriation where the Court determined that the
degree of support for the federal government’s initiative was insufficient. The issue before the
Court in Ford was whether the override in regard to every piece of extant legislation and on an
ongoing basis in regard to every new enactment complied with s. 33. It would have been sufficient
for the Court to indicate that comprehensive application of the override was invalid and to outline
its reasoning.
55Ford, supra, note 1 at 743.
1990]
LIVING WITH THE OVERRIDE
revealing, because the category of “policy” that it invokes is elusive. The issue
here was certainly not “policy” in the usual sense: the arguments in question
were not directed against the legislature’s evaluation of social benefits. Rather,
the National Assembly’s comprehensive use of the override was alleged to be
inconsistent with the formal requirements of explicitness and specificity. To be
sure, the legislation impugned was an expression of political will. The Court
was not being asked, however, to review the social benefits of the Quebec pol-
icy on the override but to determine the proper form of override of Charter
rights in view of the Charter’s language, structure, and purpose. There was no
question of policy independent of form. By characterizing the submissions as
concerned with policy, the Court collapsed the very distinction between form
and substantive review upon which it relied in sustaining Quebec’s blanket and
comprehensive override.
3.
The Retrospectivity Issue
The Court’s aversion to substantive review, where form alone was to reign
supreme, did not carry into the discussion of retroactive override. Here, at least
in the material it quoted, the Court went beyond this understanding of form to
consider the propriety of honouring reliance on existing rights or obligations.
Albeit somewhat cryptically, the normative attributes of law –
if not of rights
protection and constitutionalism –
broke the surface of the argument.
The Court might have said more than it did. As I shall explain presently,
it might have made reference to the rule of law,56 to the variety of unusual fetters
on law-making power in s. 33, and to other sections of the Charter, as well as
of our Constitution generally. The core consideration underlying such an
approach is that an individual is entitled to know the law applicable to any con-
templated action.” Changes to the law after the fact are unfair to those who act
in reliance upon existing legal arrangements. Retrospective overriding of con-
stitutional rights undermines not only that basic understanding of legal fairness
but does so in an area where the stakes are particularly high.
A retrospective override power would empower governments to dissolve
crystallized legal relations having constitutional significance. Such a course of
action might be attractive for several reasons: to shield government against
potential litigation, to escape from-a court ruling as to infringement or remedy,
or to alleviate the need to appeal a court order. Or a government might choose,
56The Supreme Court has stated, in a different context, that the rule of law is not merely set out
as a foundational principle in the text of the Charter’s preamble but is “clearly implicit” in the very
nature of constitutionalism. See Manitoba Language Reference, supra, note 24 at 750.
57See also Patriation Reference, supra, note 27 at 805-06 where it is stated:
The “rule of law” is a highly textured expression … conveying, for example, a sense
of orderliness, of subjection to known legal rules and of executive accountability to
legal authority.
REVUE DE DROIT DE McGILL
[Vol. 35
like Quebec had done on the facts before the Court, to use the override to make
a political statement. 8
Retrospective overriding of constitutional rights, for whatever reason, fore-
closes reliable assessment of the extent of one’s constitutional rights –
the most
important and the most fragile interests in our collective lives. A single use of
the retrospective override would shake confidence in those guarantees subject
to s. 33. Repeated use could undermine reliance on the currency of rights pro-
tection altogether, but at a lower political cost than prospective declarations of
override. Indeed, a government would not have to go so far; credible threats of
invocation of the override retrospectively would suffice.59 Retrospective nullifi-
cation of rights therefore effects a much greater erosion of rights than does pro-
spective action. The question is: does the Charter’s text permit it?
The Charter’s text reflects these basic principles of the rule of law. Indeed,
in the Manitoba Language Reference, the Supreme Court interpreted the spe-
cific reference to the rule of law in the Charter’s preamble to include the idea
that one is to be subjected only to known legal rules.’ In addition, other ele-
ments contained in s. 33 itself, such as the requirements that the declaration be
an express part of an enacted law, not exceed five years, and go through the leg-
islature for renewal, do not seem compatible with a retrospective capability.
Section 33 stipulates that the express statement of override take the form
of legislative enactment.” The Court might have regarded this requirement as
a matter of form only, as it did the specificity requirement, and therefore com-
patible with the substantive defects of retroactive operation of statute law. Yet,
SSQuebec did not seem to have difficulty with the values protected by the Charter or the judicial
roles generated by entrenchment of rights. The policy of omnibus and standard form override coin-
cided with increased protection of rights under the Quebec Charter. Indeed, the language policy
in issue in Ford was held by the Supreme Court to be inimical to the Quebec Charter as well as
the Canadian Charter. The controversy was whether the source of the guarantee of the rights would
be constitutional, i.e., applicable to the whole country at both the federal and provincial levels, or
local to and designed for Quebec at the provincial level. See G. Rdmillard, “Rebuilding the
Relationship: Quebec and its Confederation Partners” (speech at Mont Gabriel, Quebec, 9 May
1986) in Bayefsky, supra, note 44, 944 at 945.
59The possibility of retrospective override would thus render the rights subject to s. 33 less reli-
able, even when the override had not been used. This result would run counter to the supremacy
of the Constitution set down in s. 52 of the Constitution Act, 1982, supra, note 11.
Press, 1969) at 51-63 for a discussion of retroactive law-making.
6Supra, note 24. See L.L. Fuller, The Morality of Law, rev. ed. (New Haven: Yale University
61The override in s. 33 is to be contrasted in this respect with the emergency derogation powers
allowed by, for example, the European Convention for the Protection of Human Rights and
Fundamental Freedoms, 21 September 1970 Eur. T.S. No. 5; Council Eur. Doc. H(79) 4(1979) and
the International Covenant on Civil and Political Rights 1966, Annex to G.A. Res. 2200A, 21 U.N.
GAOR, Supp. (No. 16) 52, U.N. Doc. A/6316 (1966). These derogation powers, which apply to
a restricted list of rights and freedoms, when certain grounds obtain, are exercisable on a temporary
basis by the executive.
1990]
LIVING WITH THE OVERRIDE
it preferred to read s. 33 so as to honour reliance in the currency of constitu-
tional rights. Its reason for seeing substance implicit in form here, but not in the
other part of the argument, is unarticulated.
The requirement of an express declaration of override, which is in addition
to the requirement of enactment, emphasizes the public nature of the declara-
tion, so that all rightholders can learn both that the legislature is considering use
of the override and when and if it has put it in place. Retrospective declarations
would still be public, in the sense of being accessible and intelligible, but would
not inform people of the status of their constitutional rights at any particular
moment.62
The form of legislation stipulated by s. 33 for an override declaration offers
assurances of process that are not compatible with retroactive operation. For
example, the subordination of rights pursuant to s. 33 cannot be effected by del-
egated or executive action.63 Accordingly, rights are not hostage to executive
discretion or to administrative action. Conditioning the override on a legislative
process assures an opportunity for debate in the legislatures and, in our media-
saturated society, news coverage as well. Publicity is important because rights
protect against legislative ignorance or indifference, as well as against inten-
tional oppression.’ If those affected take the opportunity to press their claims
to their representatives, legislators cannot act without knowing the full effect of
62 0ne might argue that many, if not most, individuals exercise their rights and freedoms without
knowing that is what they are doing. Indeed, much Charter adjudication arises in the criminal con-
text where one would not expect reasoned consideration of rights protection to be uppermost in
the mind of the actor. But this criticism misses the point. The Charter establishes the ground rules
of the interaction between the state and the individual and encapsulates, in its grid of rights and
freedoms and institutional roles, the values which we deem essential to the kind of society in which
we wish to live. While the exercise of rights may be unwitting or intuitive, indeed even though
the content of many rights may in the first years of the Charter’s implementation be unknown, we
still adhere to the principle of the rule of law that the courts, by means of legal reasoning, declare
ongoing principles of legal ordering. While it may be the rare case where a person consults the law
books before taking a course of action, the legal culture in which we live provides an intuitive
ambience, which is informed by our ideas of what is fundamental to human dignity in our political
community.
63Legislation embodying an override could, likely, authorize delegated action and rule-making,
for example in emergency situations. By the terms of s. 33, there would be debate before the Act
was passed and re-enactment would be required. The protection afforded can be appreciated by
momentary reflection on the government’s reaction to the F.L.Q. crisis in 1970. Without an over-
ride in place, Charter arguments might have been made against the curtailing of civil liberties and
legal rights. Presumably the government’s action would have been sustained or rejected on the
basis of argument under section I of reasonable limits justifiable in a free and democratic society.
With an override in place, there would have been no opportunity for challenge to the rights sub-
ordinated by it. But any dissatisfaction as to the use of state powers would have a forum when the
time came for renewing the override.
6’Big M Drug Mart, supra, note 24 at 334, and Morgentaler v. R., [1988] 1 S.C.R. 30,44 D.L.R.
(4th) 385.
McGILL LAW JOURNAL
[Vol. 35
their policy on rightholders. If the cause is lost at this point, the rightholders
have at least had the opportunity that democracy offers for participation and the
building of political experience and alliances for another try if only when the
five year duration of the enactment expires. Retrospective use of the override
undercuts these features of our political life.
Furthermore, the five year maximum duration of the declaration, an
unusual fetter on legislative prerogative, gives additional opportunity to those
who oppose use of the override, in general or in particular, to press their case.65
The five year stipulation coincides with election frequency, so that each new
administration must take responsibility for departures from Charter norms, both
its own and those of its predecessors. Retrospective operation might muffle the
public debate apparently envisaged in s. 33. Rights notionally apply to everyone
so that the prospect of a period of time in the future without guaranteed enjoy-
ment of rights casts a general shadow. In contrast, using s. 33 to nullify that
guarantee of rights after the fact may not engage the constituency interested in
rightsholding generally, but only those who stand to lose by the changes in their
own established legal relationships with the state.
Moreover, the retroactive operation of the override is incompatible with the
judicial remedies available under s. 24 of the Charter, a section not subject to
the override. An override declaration pro tanto removes the specified right or
freedom from the catalogue of guaranteed interests. Someone who has litigated
or begun to litigate an alleged infringement however has invoked not only the
right or freedom guaranteed but also the access to the courts afforded by s. 24.
Retrospective nullification of rights or freedoms that have entered the judicial
system would amount to a legislative usurpation of the judiciary’s constitutional
role and drain s. 24 of any substance, despite the fact that s. 24 is not subject
to the override.
The Court’s treatment of retrospectivity has a curious relationship to its
treatment of specificity: the two approaches are inconsistent with respect to the
relevance of substantive values. Although the Court insists in its discussion of
specificity that s. 33 “lays down requirements of form only,”‘
it nonetheless
precludes the retroactive impairment of existing rights. At the level of constitu-
tional analysis, however, the Court’s treatment of the two issues is thoroughly
consistent in its refusal to attach significance to the fact that the override der-
ogates from constitutionally protected rights. The values inherent in the struc-
ture, provisions, and purposes of the Charter as the supreme law of the land
have no bearing on the validity of omnibus and standard clause denials of rights.
Although it applies a substantive rights-protecting consideration in the preclu-
sion of retroactivity, the Court ascribes that consideration to a rule of statutory
65See Rawls, supra, note 15.
“Ford, supra, note 1 at 740.
1990]
LIVING WITH THE OVERRIDE
construction. The Court fails to note the values expressed in the rule and, more
to the point, ignores the possibility that the same value that animates the rule of
statutory construction may be present in the Charter and even in the mecha-
nisms of s. 33.
V.
Section 33 as an Integrated Instrument of Rights Protection
A. History Revisited: From Debate to Text
In upholding Quebec’s comprehensive use of the override, but for its ret-
rospective application, the Supreme Court of Canada drew no connection
between its reading of s. 33 and the institutional roles set down in other sections
of the Charter. Nevertheless, the ruling has far-reaching consequences for our
understanding of the institutional framework of rights protection in Canada and
also for our assessment of the currently controversial notwithstanding clause. In
my view, the ruling cuts against the major premises of the Court’s Charter juris-
prudence. Before turning to these broader considerations, we must review some
overly familiar history in order to make salient the Court’s unstated assump-
tions. The Court’s almost complete validation of Quebec’s use of the override
manifests the generally accepted view that s. 33 is inimical to the Charter’s
guarantees. I would like to suggest a different approach, one that accords the
notwithstanding clause an integrated and positive role in our innovative consti-
tutional system of rights protection.
It is common knowledge that section 33 broke the impasse on entrench-
ment by tempering judicial review of rights claims with a legislative escape.
The landmark decision of the Supreme Court of Canada in the Patriation
Reference67 had sent the first ministers back to the bargaining table one last
time, in November of 1981. The Court, on appeal from litigation initiated by the
hold-out gang of eight in the appellate courts of three provinces, declared the
legal validity of the federal government’s contention that it could secure consti-
tutional amendments from the U.K. Parliament at Westminster to effect patria-
tion and entrenchment of the Charter without provincial consent. However, in
an extraordinary foray beyond the legal issues raised, the Court also stated that
such action would depart from established constitutional convention.6” The first
ministers, sensing a constitutional crisis of unprecedented gravity, made the
most of this judicial reprieve, and agreed to what they could: adoption of judi-
cially monitored rights protection tempered by a legislative override.
We associate this last-minute grafting of the override onto the Charter with
those who objected to rights protection altogether, and therefore understand the
change as repugnant to the rights-protecting project. Those who supported
67Supra, note 27.
68The Court had perhaps recognized that entrenching rights without wider support would vest
judges with the legal power to review rights infringements, but no real legitimacy.
REVUE DE DROIT DE McGILL
[Vol. 35
entrenchment as well as those who opposed it, those who acceded to the terms
of s. 33 for the sake of the deal and those who saw in it a refuge from rights
protection, all viewed the notwithstanding clause as inimical to rights
protection.
In this view, to which the Supreme Court subscribed despite its early dis-
claimer, s. 33 keeps alive the legislative sovereignty that pre-dated the Charter.
The override thus offers the country’s eleven legislatures a safe-haven from the
rule of non-elected and non-representative judges. The only requirement is that
the legislature follow the “form” prescribed.
Despite the superficial clarity of the historical record and the popular
understanding of these events, the idea that the override undermines rights pro-
tection may be wrong.69 The debate preceding entrenchment appeared at the
time to demand a choice between two mutually exclusive alternatives, rights
entrenchment supervised by courts and legislative sovereignty.” But perhaps we
are too close to that prolonged controversy to understand its real contribution to
the Charter’s mode of rights-protection: instead of delineating alternatives, the
controversy may have laid the conceptual foundation for a new institutional
hybrid, a complex arrangement that would harness the strengths of both courts
and legislatures to the project of rights protection.
The notwithstanding clause, I suggest, did not mark a triumph for either
side of the entrenchment debate. It embodies neither a failure to entrench rights
absolutely nor a trap door out of rights protection. Instead, it marks a dialec-
tic. But it is a special sort of compromise for it does not merely juxtapose con-
tradictory elements of opposing positions. Instead, it melds the best of the con-
tending views into something new and better. Accordingly, by looking deeper
than the conventional understanding, one can see in the adoption of the notwith-
standing clause the characteristic Canadian process of building alliances, here
creating a coherent structure of rights protection out of the extremes of legisla-
tive sovereignty and unmitigated judicial review. So read, s. 33 would reflect
something positive about Canada’s commitment to rights protection: not rule by
supercourts at the expense of legislatures, but a complex partnership through
691t is beyond the scope of this paper to delve into the historical evolution of section 33, but I
take the view that this position had deep roots in the growth of Canada’s commitment to rights pro-
tection culminating in the Charter’s final text. I elaborate the historical development of the mode
of rights protection which I here ascribe to the Charter, as a unique and distinctively Canadian
structure, in the context of the international development of models of rights-protection at the
national and international level, in a work in progress entitled “The Structure of Charter Rights”.
70Courts, those who were pro-entrenchment argued, were the appropriate bodies to articulate the
content of rights and determine their infringement in the context of individual claims and devel-
oped fact situations, but not competent to weigh rights in all circumstances against the hurly burly
of other interests. Legislatures, the opposite camp argued, holding the political mandate, the
resources and the ties to the people, could rank all interests asserted in that marketplace of interests,
including rights claims.
1990]
LIVING WITH THE OVERRIDE
institutional dialogue between supercourts and superlegislatures. And once we
become accustomed to its interesting transformation of our law and politics, we
might learn to live with –
this new institutional
arrangement.
and indeed grow attached to –
B. The Institutional Framework of the Charter: Sections I and 33
What would such a heretical view of the notwithstanding clause entail? To
approach section 33 as a coherent element in a coherent structure of rights pro-
tection is to integrate it into the other institutional arrangements set down by the
Charter! ‘ Sections 1, 24, and 32 of the Charter, under the umbrella of s. 52 of
the Constitution Act, 1982, provide that institutional context by stipulating that
rights are to be subject only to reasonable limits, demonstrably justifiable in a
ftee and democratic society; that judicial review of infringement and remedial
relief is available in the courts of law; that the strictures of rights guarantees
apply to all government action; and that the Charter, as well as the rest of the
Constitution, form the supreme law of the land. By examining s. 33 within this
institutional mesh, instead of presupposing its repugnancy to it, one can begin
to see the flawed presuppositions informing the judgment that the Court made,
as well as the contours of the legal argument it did not make.
Section 52 of the Constitution Act, 1982 states that the Constitution of
Canada is the supreme law of Canada, that any inconsistent law is pro tanto of
“no force or effect”, and lists the instruments included within the Constitution.
This supremacy of the written Constitution is not new.72 The statement of its
stature and content, at the time of the adoption of the Charter, however, does
serve a particular purpose: it emphasizes that the Constitution frames law-
making power in Canada and, therefore, that the institution of legislative
supremacy functions under law. While the legislatures are “supreme” in their
71In this paper I briefly outline the institutional coherence of the Charter. In a work in progress
entitled, “The Structure of Charter Rights”, I make full argument for this position. I contend, for
example, that the Charter’s unique institutional framework of rights protection reflects the history
and structure of Canadian constitutionalism and thus, while it includes elements of other systems
of rights protection at the national and international level – both past and present –
can only be
understood as remarkably different from them.
72The British North America Act, 1867 –
now the Constitution Act, 1867, supra, note 26 was
for the most part supreme over any law created by a Canadian institution by virtue of the Colonial
Laws Validity Act, 1865, (U.K.), 28 & 29 Vict. c. 63 which provided that statutes of the Parliament
at Westminster extending to a colony were to prevail over any repugnant colonial laws. The British
North America Act, in s. 129, also precluded change to any such imperial statute. The Statute of
Westminstei; 1931, (U.K.), 1931, 22 Geo. 5, c. 4 preserved this arrangement in regard to the British
North America Act although it did repeal the effect of the Colonial Laws Validity Act generally.
See P. Hogg, Constitutional Law of Canada, supra, note 26 at 37-42, B.L. Strayer, The Canadian
Constitution and the Courts: The Function and Scope of Judicial Review, 2d ed. (Toronto:
Butterworths, 1983) at 1-34 and W. Lederman, “The Independence of the Judiciary” (1956) 34
Can. Bar Rev. 805.
McGILL LAW JOURNAL
[Vol. 35
sphere, i.e., have plenary authority to make laws in exercise of the powers and
within the strictures set down by the Constitution, they do not enjoy sovereignty
in Canada. We live under a system of constitutional supremacy, not under a sys-
tem of legislative supremacy.
The roles of the legislatures and other institutions of government, under the
Charter, are delineated by s. 32.” While the precise application of this provision
to the myriad of activities carried out by and under governmental authority
awaits judicial elucidation, it is clear that legislatures must honour the legal
strictures upon their powers set down by the Charter. These strictures, in the
form of rights and freedoms, shape legislative power in a number of ways.
Some rights and freedoms negate power; other mandate government action. Still
others exact a respect for certain values when and if government chooses to act.
The guardianship of the Constitution is vested in the courts. Under s. 24 of
the Charter, those who enjoy the guarantees of rights and freedoms have
recourse to the courts for determination of infringements and remedial relief.74
The institutional structure of the Charter therefore contemplates the judicial
review of rights infringements by the institutions bound to honour the stipulated
guarantees. It is thus not open to the legislatures to definitively evaluate their
own adherence to the Charter. In addition, it is not for the courts to weigh the
advantages or disadvantages, in political terms, of judicial participation in the
enforcement function. The categories of activism or deference are thus dis-
placed, globally as well as in the context of analysis of particular claims of
infringement, by the Constitution’s clear prescription of judicial role.
That role is infused with content by the mode in which the rights and free-
doms are guaranteed. While the rights and freedoms are “set out” in a variety
of sections of the Charter, they are guaranteed under s. 1, which is, after section
33, the least understood of the Charter’s provisions. The text of s. 1 states:
73Section 32(1) states:
This Charter applies
(a) to the Parliament and government of Canada in respect of all matters within the
authority of Parliament including all matters relating to the Yukon Territory and
Northwest Territories; and
(b) to the legislature and government of each province in respect of all matters within
the authority of the legislature of each province.
74S. 24(1) states:
Anyone whose rights or freedoms, as guaranteed by this Charter, have been infringed
or denied may apply to a court of competent jurisdiction to obtain such remedy as the
court considers appropriate and just in the circumstances.
Section 24(2) vests in the courts the discretion to exclude evidence obtained in contravention of
guaranteed rights or freedoms where to do otherwise would bring the administration of justice into
“disrepute”.
1990]
LIVING WITH THE OVERRIDE
The Canadian Charter of Rights and Freedoms guarantees the rights and freedoms
set out in it subject only to such reasonable limits prescribed by law as can be
demonstrably justified in a free and democratic society.
The guarantee function thus has two levels, one dealing with the rights and the
other with the permissible limits upon the rights.
The main characteristic of the judicial function under s. 1, following from
this two level guarantee, is the integration of the values of right-holding into
each of the two levels. Both rights and their limits express the values underlying
rights, first, because of the nature of rights and, second, because limits on rights
are the subject of justification according to the values inherent in a society that
is both free and democratic.
This two level feature of the guarantee maximizes the judicial quality of
the courts’ monitoring of adherence to rights guarantees by organizing the anal-
ysis of the alleged infringement into separate, sequential stages. By distinguish-
ing the right from the permissible limits, and by articulating the type of analysis
applicable to the consideration of limits (demonstrable justification) as well as
the criterion of justification (the combined values of freedom and democracy),
the courts are empowered to engage in rigorous review of the state’s adherence
to rights protection while constrained to legal categories. The courts are
nowhere empowered to go beyond the traditional bounds of the judicial function
into the legislative waters of balancing or maximizing utility –
that function is
left to the legislatures, under s. 33.75
While under s. 1 the courts exercise these two related and judicially appro-
to articulate the content of the rights and to determine their
priate functions –
permissible limitation, all on legally cognizable grounds –
the structure of
rights protection under the Charter includes a third possibility. To the categories
of rights and limits, s. 33 adds the category of denials of the guaranteed rights
and freedoms by exercise of the override.76 For this category, the designation of
75This view of section 1 is elaborated in the context of a discussion of the Supreme Court’s
understanding of the limitation clause in L.E. Weinrib, “The Supreme Court of Canada and Section
1 of the Charter” (1988) 10 Sup. Ct. L. Rev. 469.
76The distinction between limitation and denial of rights is a distinction of long standing in the
formulation of Canadian rights protection. For example, while Article 1 of the Canadian
Constitutional Charter, 1971 (The Victoria Charter) provides for limitations on the fundamental
freedoms protected (“as justifiable in a democratic society in the interests of public safety, order,
health or morals, of national security, or of the rights and freedoms of others…”), Article 2 states:
No law of the Parliament of Canada or the Legislatures of the Provinces shall abrogate
or abridge any of the fundamental freedoms herein recognized and declared.
Similarly, in the Constitutional Amendment Bill (Bill C-60), s. 25 (1978) provides for limitations
on a more selective range of grounds than the Victoria Charter and then, in s. 23 sets out this
caveat:
REVUE DE DROIT DE McGILL
[Vol. 35
lines.
institutional power also follows along traditional functional
Appropriately, it is the legislatures that the Charter permits to nullify rights and
freedoms in the name of majoritarian or representational values.
This exception to the guarantees afforded under s. 1 –
of the rights or their
justified limits as monitored by courts of law –
is constrained by requirements
of scope, duration, and manner in order to intensify the legislative nature of the
function. The override thus applies only to some rights and freedoms;77 must be
renewed by legislative enactment, as desired, after a maximum duration of five
years; and requires certain attributes of form: an express declaration of override
and specification of the rights overridden. The exceptional quality of the instru-
ment of departure from the norms embodied in the rights and freedoms is indi-
cated both by the heading in the text of the Charter, “Exception where express
declaration”, and by the extraordinary fetter placed on the legal effect and con-
tent of the legislative declaration of override.7″
The design of the Charter is thus not to give legislative powers to courts
or to combine constitutionalization of rights with an inimical power of legisla-
tive supremacy over some of those rights. Indeed, the reciprocity of sections 1
and 33 avoids that very institutional anomaly by providing a structure of rights
protection in which courts can be courts and legislatures, legislatures. This
design cannot be appreciated if one approaches Charter interpretation by distin-
guishing between activism and deference, or their proxies form and substance,
and chooses one or the other alternative as appropriate without legal analysis.
Instead, one must recognize that, here as elsewhere in our legal system, form
and function reflect substantive values, and in a properly designed and inter-
preted instrument, coherence as well.
Accordingly, form and substance find expression in both the judicial role
under s. 1 and the legislative power under s. 33. Section 1 authorizes judicial
review of the content of legislation, and the exercise of power thereunder, for
adherence to the values crystallized as rights and freedoms. It also authorizes
the judiciary, if the state assumes the onus of persuasion, to examine the justi-
fiability of a policy that infringes guarantees. The judiciary cannot enter upon
this second stage of consideration of the permissibility of limits on rights unless
To the end that full effect may be given to the individual rights and freedoms declared
by this Charter, it is hereby further proclaimed that, in Canada, no law shall apply or
have effect so as to abrogate, abridge or derogate from any such right or freedom.
See A. Bayefsky, supra, note 44 at 214, 354-55.
77Namely, ss. 2, 7-15, the rights that do not reflect the structural components of our federalism
and separation of powers.
781n Skapinker, supra, note 40 at 370-77, the Supreme Court of Canada deliberated upon the use
of marginal notes in aid of interpretation of the Charter text and concluded that such notes can be
helpful.
19901
LIVING WITH THE OVERRIDE
the state has prescribed its policy as law.7 9 It is here, as a precondition to jus-
tification of limits on rights, rather than in the exercise of the override power
under s. 33 as determined by the Supreme Court of Canada in the cases at bar,
that the simple form of law-making has operative significance. That very min-
imal form of law is set by the Charter text as a precondition to the task of jus-
tification in a court of law of limits on interests otherwise protected from leg-
islative incursion. In this context, this form is to be sustained as law only as
justified on the criteria essential to a society committed to both freedom and
democracy. It is not evidence of fulfilment of the manner and form stipulated
for exercise of the exceptional departure from those values permitted by the
override.
Form also plays an important role in s. 33. The exceptional quality of the
override suggests a narrow reading.” The terms of s. 33 are conducive to such
a reading because they set down extraordinary conditions for the exercise of
legislative power. The legislature must consider not only the policy to be trans-
formed into law, but also the terms of the override and the specific rights and
freedoms in issue. These conditions do not engage the policy determination that
underlies exercise of the override; they work to intensify the democratic func-
tion that legitimates the legislative process by ensuring that debate, i.e., the
democratic process of law-making, focuses on the subordination of the rights or
freedoms in question. The courts are not to review the merits of that policy. Nor
do they investigate whether the debate took place. Their role, as for other con-
stitutional powers, is to monitor adherence to the power granted and any stric-
tures upon it.
Accordingly, the Charter does not envisage the transformation of our
courts into super-legislatures to oversee the determination by a legislature to
override rights or freedoms. On the contrary, it envisages the transformation of
legislatures into super-legislatures, in the exercise of their powers to make law
consistent with rights protection and, when prescribed conditions are met, to
make law that overrides rights. The transformation of the courts is not into
super- legislatures, therefore, but into super-courts, articulating the content of
rights, justifying their limitation on grounds conducive to freedom and democ-
racy, as well as monitoring strict adherence to the form of legislative override.
79The view that compliance with the term “prescribed by law” in s. 1 requires accessibility and
intelligibility of the legal instrument, as required by the principles of the rule of law, is not incon-
sistent with the position taken in the text. See, R. v. Therens, [1985] 1 S.C.R. 613 at 645, 18 D.L.R.
(4th) 655 at 680-81, Le Dain J., for this view. These stipulations are still matters of form and not
of substance because they can be determined without engaging in an evaluation of the social wel-
fare embodied by the policy.
8lhis rule of statutory interpretation is absent from the Court’s canvass of applicable interpre-
tive guides to the text of s. 33. In contrast, it is the cornerstone of the Quebec Court of Appeal’s
judgment on the same issues in Alliance, supra, note 9. See supra, notes 9-24 and accompanying
text.
McGILL LAW JOURNAL
[Vol. 35
VI. Conclusion
The Charter lays down an intricate set of institutional arrangements. It
intertwines a limited legislative override with the judicial oversight of the guar-
antee and limitation of rights. These elements form an integrated regime of
rights protection in a democratic society. By so easily reading s. 33 as empow-
ering the National Assembly of Quebec to negate constitutional rights whole-
sale, the Supreme Court ignored the implications of its decision for the wider
framework of Canadian constitutionalism.
To avoid reviewing the merits of the Assembly’s action, the Court postu-
lated a choice between a formal and substantive reading of s. 33. This dichot-
omy fails to touch the issue at stake. No one claimed that s. 33 calls for the judi-
cial evaluation either of the social good promoted by the legislature or of the
democratic nature of the legislative process at a particular moment. Rather, s.
33 engages values of constitutional process. Such values are within the judicial
competence no less than the process values of criminal or administrative law.
The Court’s handling of the blanket override is rich in ironies. There is,
first of all, an irony about judicial activism. On the one hand, the Court deferred
to the repudiated policy of a defeated government that used the Charter to reject
the Charter. On the other hand, the Court exercised its special mandate as
guardian of the Constitution to vindicate Charter values in respect to language
rights. It thus boldly struck down duly enacted legislation of extraordinary polit-
ical sensitivity, while simultaneously refusing –
to
explore the values of constitutional process that must govern legislative action.
Second, there is an irony about language rights. In dealing with s. 33, the
Court tells us that the language used in a declaration of override is without sig-
nificance. Nonetheless, in striking down the specific legislation, it holds that
language expresses our individual human dignity and is for that reason pro-
tected under the Charter as a fundamental freedom.
on grounds of deference –
Third, there is an irony about democracy. In purported deference to the
democratic function, the Court eliminates all “substantive” costs to the overri-
de’s exercise. It regards the use of the override simply as the exercise of major-
itarian power. The Court thereby defers to a legislative process devoid of its
legitimating qualities of reasoned and focused debate by the people’s represent-
atives.t
Fourth, there is an irony about the future of the override itself. The Ford
decision and Quebec’s subsequent use of the override in its new language leg-
islation catapulted s. 33 back onto the constitutional agenda. The Court’s treat-
81The best proof against this view is the heated reaction across the country to Quebec’s use of
the override after Ford.
1990]
LIVING WITH THE OVERRIDE
ment of the override, however, denies us a satisfactory appraisal by our highest
court of the place of s. 33 in the Charter’s institutional arrangements. Without
insight into the symbiotic relationship between rights and justified limits artic-
ulated by courts of law, on the one hand, and the temporary, express and specific
override enacted by legislative fiat, on the other, it is difficult to appreciate the
way in which the Charter insulates the courts from the political repercussions
of the rights-honouring function. Consequently, we may not realize that s. 33
frees Canada from the crisis of judicial legitimacy that mars other rights-
protecting systems. If the ultimate result of the Court’s judgment is the repeal
of s. 33, our Constitution will lose the very mechanism that allows the judiciary
to make judgments of the kind the Supreme Court of Canada itself made and
presumably deems appropriate, in regard to the language of expression in the
Ford case.
Our constitutionalism is a seamless web, made up of constitutional text,
convention and judicial pronouncement. The override can be understood as
intensifying and strengthening its patterns of respect for democracy, difference
and individual dignity. The Supreme Court’s interpretation of the override’s pur-
pose, function and strictures purports to strengthen our commitment to democ-
racy. However, its reasoning cuts the connection between democracy and the
other fundamental values for the sake of which we hold democracy dear.
