Australian Journal of Human Rights
Institutional dialogue between courts and legislatures in the definition of fundamental rights: lessons from Canada (and elsewhere)
In recent years, a number of commentators have used the concept of dialogue to capture the relationship between Canadian courts and legislatures in the judicial review of constitutionality.1 As long as dialogue is treated sceptically and is not assumed to describe an ideal conversation, the idea of dialogue can provide a useful perspective on human rights protections generally, for it emphasises the extent to which both courts and legislatures have valuable things to say about rights, directs our attention to the ways in which the two institutions interact, and provides tools for evaluating the remarkably broad range of mechanisms that can be used to protect rights.
This paper reviews a spectrum of means by which such ‘dialogue’ over rights can be structured in a constitutional system. It then focuses in more detail on three specific forms of rights protection and evaluates the institutional balance achieved in each:
• the finding, by the courts, that constitutional provisions dealing with other matters contain implicit rights guarantees, so that rights come to be addressed through the adjudication of provisions having little ostensibly to do with rights (an ‘implied rights’ approach);
• the entrenchment of a Bill of Rights in the Constitution, backed by judicial review, but subject to express derogation by legislative action (the approach contained in s 33 — the ‘notwithstanding clause’ — of the Canadian Charter of Rights and Freedoms); and
• the declaration of rights in an ordinary statute, lacking constitutional status but protected by a requirement that derogation occur in a specified manner and form (the ‘statutory Bill of Rights’ approach).
Each of these mechanisms has been used in Canada. In this paper I will draw primarily on Canadian experience, although I will also discuss the Australian jurisprudence on implied rights and refer to approaches to rights taken in a number of other chiefly Anglo-American jurisdictions.
The notion that rights guarantees are characterised by ‘dialogue’ between courts and legislatures may strike the reader as both counter-intuitive and normatively inappropriate. We generally think of judicial review in much simpler and monological terms. Parliament passes laws, and courts either uphold them or strike them down on the basis of their interpretation of the Constitution. Any sense that political actors influence the courts’ interpretation of the Constitution seems incompatible with judicial independence and the rule of law; we assume that courts should come to their own conclusions as to constitutionality, without reference to what political actors like or dislike. Some legal realists have emphasised that political actors do have an impact on judicial review. But this impact is generally treated by the realists themselves as incompatible with claims of judicial independence.
In this paper, I will deal with principled justifications for legislative participation in the definition of rights only en passant, as part of the evaluation of the balance struck by various rights instruments.2 I should note, however, that institutional dialogue over rights is much more common than is often acknowledged — indeed is, to some extent, universal.
In Canada, the claim that judicial review involves dialogue has special credence because of two distinctive aspects of the Canadian Charter of Rights and Freedoms. First, the Canadian Charter contains an express limitation clause, s 1, which states that its rights and freedoms are ‘subject only to such reasonable limits, prescribed by law, as can be demonstrably justified in a free and democratic society’. This clause recognises not only that rights are subject to limits, but also, implicitly, that government should have the burden of justifying those limits (R v Oakes at 136–137). Second, the Canadian Charter permits legislatures to derogate from some of the rights it enunciates. Section 33 provides that legislatures can insulate a statute from certain forms of Charter review by expressly declaring that the statute shall operate notwithstanding certain sections of the Canadian Charter of Rights and Freedoms. When this occurs, judicial review is excluded. Both these provisions suggest that the legislature may actively participate in the definition of constitutional protections.
Although these clauses are distinctively Canadian, there are functional parallels in virtually all constitutions. Other Bills of Rights contain express limitation clauses (see New Zealand Bill of Rights Act 1990, s 5; Constitution of the Republic of South Africa 1996, s 36; for clauses applicable to specific rights, see International Covenant on Civil and Political Rights, art 19(3); European Convention on Human Rights, art 10(2)). But even if they do not, it is generally conceded that all rights are subject to limits. Those limits may not be conceived as restrictions of an otherwise unlimited right; they may be conceived as aspects of the definition of the right. And there may be no clear understanding that government has the burden of justifying limits; it may simply be assumed that the courts will determine the limits. But these distinctions make little difference for our purposes. The rights are, in any case, subject to limits; and when a statute is subjected to judicial review, any government will seek to justify a measure that it wishes to retain.
The Canadian Charter’s ‘notwithstanding clause’ appears more strikingly original than the limitation clause, but even it has functional parallels elsewhere. Virtually every constitution is subject to amendment by some legislative process, onerous though it may be.3 The distinctive characteristic of the Canadian provision is really the ease with which the rights guarantees may be set aside, and the implicit message that it may be legitimate to do so.
All of this suggests that there is something of broader significance to the notion of dialogue between courts and legislature in judicial review. At the very least, it emphasises that constitutional review is about a complex relationship between legislatures and courts. Judicial review is not simply about laws enacted and then subjected to the guillotine of judicial nullification. There is considerably more potential for — indeed presence of — to and fro between legislatures and courts. Or, to put it another way, the interpretation and enforcement of rights by courts always operates within a zone of tolerance created by the relative difficulty of legislative override (whether the override would occur by ordinary legislative process or by constitutional amendment). The control of the courts over legislative action is always conditional, dependent on the degree of institutional friction within the system. This means that it is possible for framers of a constitution to structure the institutional relationship in a variety of ways to achieve a desired balance in legislative and judicial roles.
One final comment. Throughout this paper I use ‘political’ in contrast to ‘judicial’ to refer to the broad spectrum of public debate, decision making and action outside the courts. Of course, judges are also political actors in a broad sense. Moreover, I accept that there is no sharp distinction between the kinds of reasons employed by judges and those employed in political discourse generally (although there are significant differences in the weight given to particular kinds of reasons). My choice of terminology is merely a matter of convenience. In this paper I am concerned with the manner in which judges enter into dialogue with broader political processes (especially with governments and legislatures) in their administration of rights guarantees. Those broader political processes are wide and various; it is simplest to use the catch-all ‘political’ to describe them.
Clearing the underbrush
Before plunging into the mechanisms for dialogue, it is worth setting out a series of assumptions that underlie my discussion of Bills of Rights.
First, I do not accept a simple ‘checks and balances’ — or libertarian — justification for constitutional review, in which judicial review is defended purely and simply because it limits government, regardless of the grounds of limitation. On the contrary, the ability to participate actively in government to achieve societal goals is a key dimension of freedom. A corollary of this is the vesting of a measure of trust in democratic institutions — a commitment to their efficacy, as the most representative of governmental institutions. All other things being equal, democratic institutions should be permitted to make decisions and carry them into effect. The nullification of statutes through judicial review therefore requires specific justification.
Second, the choice of whether to have judicial review of rights guarantees is not a straightforward choice between having rights or not. Legislatures too are concerned with rights, although sometimes in different ways from courts. Indeed, institutional characteristics mean that both courts and legislatures have strengths and weaknesses in the definition, interpretation, and application of rights, to which I will return below. One can therefore be a strong supporter of rights without supporting judicial review on the basis of a constitutionalised Bill of Rights. Support for judicial review requires an additional premise: a reason why that particular institutional form is appropriate.
I do not want to appear disingenuous about this. I do have concerns with the extent of hegemony of the discourse of rights and about its tendency towards simplification, uniformity of treatment, highly symbolic argumentation, and resistance to compromise, all of which I have explored elsewhere (Webber 1993: 2000a). Scepticism towards or opposition to a constitutionalised Bill of Rights is often founded upon opposition to the peculiar role of rights discourse under such a regime, and this paper is no exception. But for the purposes of this paper, it is sufficient to note that scepticism with respect to judicial review may have little to do with support for or opposition to rights. One can be for rights, but consider other means of protection to be preferable.
Third, constitutional Bills of Rights often serve a variety of ends; the substantive protection of individual rights and freedoms is not their sole function. They have often played an important role, for example, in national consolidation. They have established a common basis of citizenship, affirmed that all citizens are subject to the same governmental authority, or sought to enunciate the fundamental values of the nation. This national symbolic role may align closely with the protection of rights but the two functions are not identical. There can be points at which they diverge. There may, for example, be very good reason on grounds closely linked to human rights concerns for recognising a measure of diversity within the public institutions of the state. It may be perfectly appropriate, for example, for cultural minorities (indigenous peoples; large linguistic minorities in a country such as Canada) to have access to their own institutions (schools; social programs; even governmental structures). Yet those very demands have often been resisted in the language of equality derived from constitutional guarantees, for reasons that have little to do with individual liberty but a great deal to do with a desire for national consolidation (Webber 1993: 230-231; 1994: 141-144 and 234 and following; 1999: 260-277). It is important, then, to be clear on the functions to be served by a Bill of Rights. In this article I will focus entirely on the human rights dimension of Bills of Rights, although noting complications posed by their national symbolic role.
Fourth, it is important to pay attention to the real issues that are likely to be dealt with under a Bill of Rights. Many justifications for Bills of Rights rely upon a parade of horribles; the rights guarantees are justified on the basis of the need to prevent the grossest of human rights violations. But in most societies with strong democratic cultures, Bills of Rights are not used to prevent gross violations. They are deployed at the margins. They define the rights’ outer limits, or rule upon their implications in situations in which judgments of right and wrong are highly complex and disputed. Under the Canadian Charter of Rights and Freedoms, for example, the Supreme Court of Canada struck down the Federal Government’s legislation regulating tobacco advertising on the basis of freedom of expression (RJR — MacDonald v Canada)); it has struck down provisions that prohibited interprovincial law firms (Black v Law Society of Aldebta) or required lawyers to be Canadian citizens (Andrews v Law Society of British Columbia). Granted, it has made some decisions that have had an important impact on the rights of individuals. This is true, for example, of its early judgment striking down Canada’s abortion law on procedural grounds (R v Morgentaler), or its more recent decision extending Alberta’s Individual’s Rights Protection Act to include discrimination on the basis of sexual orientation (Vriend). But in the vast majority of cases there have been strong arguments on each side of the issue; the ultimate decision has been a matter of fine determination.
If the rights culture in a particular country is sufficiently strong that gross violations of rights never come before the courts, then one has to ask whether judges must have the last word. If judicial review does nothing more than substitute the reasoned opinion of a judge for the reasoned opinion of the legislature on a matter of real doubt as to what justice requires, why bother? Of course, the situation is rarely that simple. Legislatures can lose sight of rights considerations in their rush to achieve a social objective. Unpopular minorities can find themselves disadvantaged by measures that individually may not amount to much, but that cumulatively impose a significant and debilitating burden. The point is that there are many ways in which rights protections can be achieved. One may structure the institutional relationship differently depending on the matters likely to come before the courts.
This in turn suggests that different forms of rights protection may be appropriate in different contexts. One may need entrenched Bills of Rights most in societies that lack an established democratic or human rights culture, where the symbolic affirmation of rights serves an important role and political power must be forcefully constrained. They may be less necessary in societies in which human rights already form an important strand in political debate.
Institutional dialogue and the structure of human rights protections
What elements determine the scope for dialogue in the enforcement of human rights protections? They are remarkably diverse. When taken together, they define a broad range of options for the institutionalisation of human rights. It will be useful to address them under three headings: Extent of entrenchment; Strategy of judicial review; and Mechanisms for legislative involvement in the definition of rights.
Extent of entrenchment
We commonly think of constitutional entrenchment as being all or nothing: rights are either enshrined in the Constitution, exempt from legislative tampering, or they are left to the mercy of the legislature. However, entrenchment is a much more relative concept than this suggests. In essence, it is concerned with the degree to which rights provisions have a controlling impact on legislation, and the ease with which the provisions can be changed. There are a wide variety of means by which rights can be pursued, each postulating a different relationship between legislature and courts. Often these are cumulated within one legal system. Canadian law sports a particularly rich selection. Here I canvass the possibilities in order of increasing constraint.
Non-binding declarations of rights
Some rights have no binding force within the domestic legal system. The protections operate by virtue of their moral force alone. Their impact can nevertheless be substantial, inducing legislatures to amend their laws to bring them into conformity.
The most common example occurs in the case of international norms. In countries descended from the British constitutional tradition, international norms created by treaty have no direct force within domestic law. They need to be incorporated into legislation in order to take effect domestically. In formal terms, then, the rights are at the mercy of the domestic legislatures and can exercise no direct constraining effect.
The international instruments can nevertheless have a substantial impact. In Canada, for example, a decision by the UN Human Rights Committee held that the definition of ‘Indian’ in the federal Indian Act contravened art 27 of the International Covenant on Civil and Political Rights, because the Act’s gender specific structure (under which Indian women who married non-Indian men lost their status, but Indian men who married non-Indian women did not) unjustifiably denied Aboriginal women the right to enjoy their culture in community with other members of their group (Lovelace v Canada). The Act was amended, even though the measure had previously been upheld by the Canadian courts under the Canadian Bill of Rights (a statutory Bill of Rights pre-dating the adoption of the Canadian Charter of Rights and Freedoms) and even though a number of First Nations disagreed vehemently with the change (An Act to amend the Indian Act (Canada); A-G Canada v Lavell). Similarly, in Australia, a decision of the Human Rights Committee prompted the Commonwealth Parliament to adopt legislation overriding Tasmania’s criminalisation of sex between consenting homosexual adult men (Toonen v Australia; Human Rights (Sexual Conduct) Act (Cth); Croome v Tasmania). And until the incorporation of the European Convention on Human Rights into British law by the Human Rights Act 1998 (UK), that Convention, too had no direct force within domestic British law (except as an aid to interpretation). It nevertheless had a significant impact, leading to a number of amendments to British law (Kinley 1993).
There are also ‘programmatic rights’ in international law and some national constitutions. These rights are similarly exempt from judicial review. Their implementation is left entirely to the discretion of the legislature (see, for example, International Covenant on Economic, Social and Cultural Rights, art 2(1); Constitution of the Republic of Ireland, art 45; Constitution of India, Pt IV).
Interpretive conventions based on implicit norms
Human rights can also have an impact on the law through interpretive conventions, under which courts strive to interpret legislation in a manner consistent with rights. Here, the rights considerations have no independent constraining effect. They cannot be used to strike down legislation. They simply shape the judge’s interpretation of the law, so that the law is rendered as consistent as possible with human rights norms.
These human rights norms can sometimes be a matter of general principle, with no authoritative legislative expression. One good example is found in the Supreme Court of Canada’s decision in MacKeigan v Hickman. There, the relevant statute provided that a Commission of Inquiry could summon ‘any persons’ to give evidence as a witness. The Court held that this very general language should not be taken to override a principle as important as judicial independence; it therefore held that the statute did not permit a Commission to compel a judge to testify as to their reasons for decision in a particular case. Another example is the requirement of compensation when property is expropriated in Canada and the UK. There is no express guarantee of compensation. The courts simply presume that compensation is to be paid unless the legislature stipulates otherwise (A-G v DeKeyser’s Royal Hotel; Burmah Oil v Lord Advocate; Manitoba Fisheries v The Queen).
Here again the impact can be substantial even though ostensibly one is merely in the realm of statutory interpretation. There is always considerable latitude in interpretation. If courts insist upon an extraordinary degree of clarity before they interpret a statute so that it constrains rights, the protection can be very great indeed. It can amount to a requirement that to impair rights, a legislature must do so explicitly.
Interpretive conventions based on explicit norms
In some cases, legislatures enact express norms, which are then used by courts in their interpretation of other statutes. This method of rights protection works much like that described in the previous section: the courts do not invalidate legislation, but they do interpret it restrictively in order to avoid an impairment of rights. As in the previous section, the effect can be significant; the courts can require a very high degree of clarity before they find that rights have been restricted (see, for example, Winnipeg School Division No 1 v Craton at 156).
The difference is the involvement of the legislature in the specification of the norms. This creates real give and take between courts and legislature. The legislature sets the norms, sometimes enshrining them in a statutory Bill of Rights; the courts take those norms and use them to interpret other laws, construing those laws so that they respect the legislature’s norms. This differs from the situation described in the previous section, in which courts are responsible both for the articulation of the norms and for their use in interpretation.
This form of interpretation was one of the ways in which the statutory, pre-Charter, Canadian Bill of Rights was understood to work — although how well it worked is another question. The weakness of the Canadian Bill of Rights — its lack of constraining effect — was fiercely criticised by human rights advocates (Tarnopolsky 1975). A more successful example is the New Zealand Bill of Rights Act 1990. That act cannot be used to invalidate legislation, yet it has nevertheless had a significant impact on criminal procedure and the common law of defamation (see Keith 2000; Allan 2000). Interpretation is also one strategy used by the UK’s Human Rights Act 1998, which incorporates the European Convention on Human Rights. That Act, however, goes well beyond interpretation (i) to invalidate subordinate legislation; (ii) to bind public authorities; (iii) to permit courts to declare that primary legislation is incompatible with the Convention (although that declaration does not affect the statute’s validity); and (iv) to permit a Minister of the Crown to make amendments to bring the statute into conformity.4
In each of these examples, the human rights norms are enacted by the same legislature whose statutes are then subject to scrutiny. But judicial interpretation sometimes draws on norms articulated by other legislatures than the one scrutinised. Indeed, this is one way in which treaty norms can have an impact on domestic law, even without the incorporation of the treaty into domestic statutes. The courts interpret domestic law so that it is, as far as possible, consistent with international law (see, for example, Minister of State for Immigration and Ethnic Affairs v Teoh).
Rights guarantees protected by a manner and form requirement
In the mechanisms examined thus far, the rights norms do not bind the legislature. At least in theory, the legislature could set them aside, as long as it did so with sufficient clarity to overcome the court’s interpretive presumptions. In Canada, however, legislatures have enacted rights norms, binding those very legislatures, through the imaginative use of manner and form requirements.
To understand how these requirements work, it is important to realise that in the British tradition legislatures generally cannot bind themselves. The doctrine of parliamentary sovereignty requires that at any point, the legislature can change its mind, passing laws that contradict its earlier enactments. Whenever there is inconsistency, the former laws are repealed to the extent of that inconsistency. Among other things, this principle operates as an important bulwark of democracy, for it means that legislators can always repeal the work of their predecessors; a government facing defeat cannot bind its successor.
Manner and form requirements operate as a limited exception to this principle, at least in Canada.5 Although legislatures cannot bind themselves as to substance, Canadian courts have held that legislatures can bind themselves as to process. They can stipulate a particular procedure — a specific ‘manner and form’ — by which enactments must be made. Until repealed, these stipulations (as long as they are genuinely procedural and do not amount to a disguised limit on substance) must be followed by the legislature’s successors (R v Drybones; Ford v Quebec; Reference Re: Canada Assistance Plan at 322–324). These requirements can be used to offer a qualified protection to human rights norms. Instead of attempting to impose the norms directly, the legislature enacts, in the rights instrument, a distinctive procedure by which restrictions must be adopted. Because it is purely procedural, the requirement binds the legislature’s successors.
In Canada, the stipulated requirement is usually that, to set aside the rights guarantees, the legislature must state explicitly that the statute is to apply notwithstanding the rights guarantees. This does not bind as to substance; the legislature can always set aside the protections, by an ordinary majority, as long as it does so explicitly. But it does guard against inadvertent restrictions of rights, and it also plays a very important signalling function: any government seeking to set aside the guarantees must do so explicitly, and that will in turn tend to generate a vigorous public debate, forcing the government to justify its actions. The manner and form requirement serves, in other words, as a trigger to the democratic process, by providing clear notice that a rights issue has been raised.
In Canada’s federal system, statutory Bills of Rights of this kind only bind legislation of the level of government that enacts them.6
This was the principal approach adopted in the Canadian Bill of Rights — the statutory Bill of Rights enacted by the Parliament of Canada in 1960. Although its role in rights protection has largely been overtaken by the Canadian Charter of Rights and Freedoms (a constitutional instrument), the Canadian Bill of Rights remains in force. In 1985, it was used by three judges of the Supreme Court of Canada to strike down the refugee determination process in Canada’s Immigration Act (Singh v Minister of Employment and Immigration; the other three justices came to the same conclusion on the basis of the Charter). Alberta employed a similar approach in the Alberta Bill of Rights of 1972. Quebec did so as well in the Charter of Human Rights and Freedoms, a statutory Bill of Rights enacted in 1975. In 1988, the Supreme Court of Canada relied on the Quebec Charter to strike down a s of Quebec’s language legislation that banned the use of English on commercial signs (Ford v Quebec). The decision had a very significant political impact. The Quebec National Assembly re-adopted the sign law in modified form (An Act to amend the Charter of the French Language, 1988, popularly known as Bill 178), this time protecting it from the Canadian Charter of Rights and Freedoms (but not the Quebec Charter of Human Rights and Freedoms) through the use of a ‘notwithstanding clause’. As I discuss further below, that action provoked vigorous criticism from English speaking Canadians. This reaction contributed to the ultimate defeat of a package of constitutional amendments (the ‘Meech Lake Accord’) supported by Quebec (Webber 1994: 138ff).
The effectiveness of manner and form requirements is therefore well established in Canadian law. They are used to give qualified force to statutory rights guarantees. I return to the value of these instruments in structuring institutional dialogue below.
Constitutionally entrenched guarantees
Of course, when we think about Bills of Rights we usually think of instruments subject to full constitutional entrenchment. Even here, however, there is more variation than might at first appear.
First, the nature of the entrenched norms can be very different, with important consequences for the scope of judicial review. Occasionally, decisions that are (one suspects) primarily based on rights considerations are framed in language that has little to do with rights. This was a familiar phenomenon in Canada prior to the adoption of the Canadian Charter of Rights and Freedoms where, in a few celebrated cases, decisions with a strong human rights element were based on the federal/provincial division of powers.
In some cases, this was bona fide division of powers reasoning, in which the rights concern was genuinely tied to the nature of the particular power in issue. Thus, in Canada, criminal law is a federal matter. The criminal power was interpreted to cover a number of legislative aims that had strong potential to impair individual rights, such as the suppression of sedition or the enforcement of religion (an interpretation with some justification, given the purposes to which the criminal law had historically been put). Provincial laws dealing with these matters were struck down as infringing upon federal authority (Switzman v Elbling; AG Ontario v Hamilton Street Railway; Henry Birks & Sons v Montreal).7
At other times, one had the impression that division of powers reasoning was being distorted in order to attain a rights objective. This may have been the case in McKay v The Queen, for example, in which by-laws enacted under provincial authority to regulate signs were held not to apply to federal election signs.
In any case, although provisions dealing with the division of powers have on occasion served the ends of human rights, their potential has been limited. The protection they afforded was minimalist, for the court had to find a plausible hook within division of powers reasoning on which to hang its decision. In principle the decisions were only concerned with who could infringe rights; if the measures of one level of government were struck down, it was always open to the other level to adopt them. Finally, the simple fact that these decisions relied on surrogate arguments meant that it was impossible to develop a coherent and explicit rights jurisprudence.
A second type of entrenched rights consists of implied rights. These rights are not explicitly set out in the Constitution, but instead are derived from other provisions that ostensibly have little to do with rights. They differ from the division of powers judgments in that the court uses the language of rights in its reasoning, finding that a particular right is implicit in the constitution. The notion of implied rights has had a large hurdle to clear. There is a strong commitment to parliamentary sovereignty in the British tradition. That has generally meant that constitutional restrictions are read narrowly, especially if (as in the case of rights) the effect of the restriction would be to prevent all levels of government from enacting the measure.
There are many ways in which rights might plausibly be implied. The ones that have attained the most currency, however, are founded on constitutional provisions that establish democratic structures of government. The argument is that democratic institutions cannot operate without free political debate. Some protection of freedom of speech must therefore be implied.
Although certain Canadian judges flirted with the implication of rights guarantees prior to adoption of the Charter (and indeed after), this reasoning had not found its way into the reasons for judgment of the majority of the Supreme Court of Canada prior to the adoption of the Canadian Charter of Rights and Freedoms. Instead, the majority had always rested its decision on alternative grounds. Ironically, however, Canadian musings about an ‘implied Bill of Rights’ did contribute to the development of a vigorous implied rights jurisprudence in Australia.8 I will return to this form of protection below.
Third, there are explicit rights guarantees. I need say little about them here for they are by far the best known. I simply note that they too can vary considerably in their constraining effect, depending on how they are drafted.
Indeed, not all express rights are justiciable. So-called ‘programmatic’ rights — often rights to services that would impose onerous financial obligations on government (housing; education; social welfare; environmental protection) — serve merely as directions to government to make those aims a legislative priority. The courts have no role in their enforcement.
Entrenched rights also differ in the extent to which they are insulated from legislative impairment or change. There are at least three different ways in which rights protections can be subject to legislative restriction.
First, all rights are subject to limitations; the very process of definition, through judicial interpretation, involves a delineation of limits. When scrutinising a statute impugned on human rights grounds, the courts will attempt to determine whether a legislature has overstepped those bounds. In this process, political actors too play a role. They can discuss rights considerations at the time of adopting legislation, offering rationalisations that in turn influence the courts.9 They can, through their lawyers, offer further justifications when the matter goes to court. Thus, at least through the exercise of persuasion, political actors can have an impact on the scope of the rights. In Canada, this role is contemplated in s 1 of the Charter.
Second, Charters of Rights can expressly permit legislative derogations. As already mentioned, that is the case in s 33 of the Canadian Charter. I explore s 33 in detail below.
Ultimately, rights guarantees can always be set aside by constitutional amendment. The ease of amendment differs from constitution to constitution. Some constitutional provisions, in some countries, are amendable by ordinary majority. In Canada, this was the case with guarantees for the use of French originally applied to the North West Territories and inherited by the provinces of Alberta and Saskatchewan upon their creation. These provisions had fallen into disuse prior to the provinces’ creation, without being formally repealed. When, in 1988, the Supreme Court found that they were still in force, the two provinces immediately repealed the guarantees and validated previous acts passed in violation of the requirements (R v Mercure; The Language Act (Saskatchewan); Languages Act (Alberta)).
Generally the mode of amendment is considerably more difficult than this. It may be so difficult as to be functionally impossible. Indeed, it has been argued that some provisions of some constitutions are unamendable (although this is very rare). Even those constitutions, however, can be overthrown by revolution.10 It is preferable, then, not to treat entrenchment as though it rendered constitutional provisions untouchable. In the last analysis, all constitutions are open to change. The sanctity of rights protections therefore depends, always, upon the dynamic relations among institutions in the political order. Entrenchment is always a matter of degree.
Strategy of judicial review
The extent of institutional interaction — including the extent to which legislatures participate in the definition of rights — is also affected by the judges’ conception of their task. Their definition of the content of rights can either foster legislative action or foreclose it.
Rights guarantees as minimum guarantees
First, room for legislative action can be preserved by courts approaching constitutional provisions on the basis that they are meant to provide minimum guarantees, not to confer on courts the power to regulate everything having to do with the subject matter of particular rights, such as expression or religion. Such a restrained interpretation has the merit of coinciding with most justifications for judicial review on rights grounds. These justifications tend to focus on rights guarantees as minimum guarantees — as ensuring a basic minimum level of respect for individual freedom and equality.11
Constitutional texts rarely take such a minimalist form, however. They speak in ringing language, proclaiming the need to respect ‘freedom of expression’ or ‘equality before the law’, without hedging those concepts about with qualifications. Constitutional provisions do not speak of a ‘necessary minimum of free expression’ or a ‘fundamental baseline of equality’. Courts, faced with the rights’ broad language, can be tempted to give it full rein. They can start with the abstract concepts of expression and equality and seek to define exhaustively their scope and meaning in a liberal society. The role of judicially enforced rights as minimum guarantees can fall away, the concepts come to be defined in plenary terms, and all legislative restrictions compelled to pass rigorous constitutional standards.
This seems to be what has happened in Canada with freedom of expression. The Supreme Court of Canada defined expression broadly to cover any attempt to convey meaning, from commercial advertising to ‘the imagery of a sexual gadget’ (Irwin Toy at 606–608; R v Butler at 472). All expressions are accorded, at least ostensibly, equal constitutional protection. Thus, the constitution is taken to have entrenched the right of the tobacco company, RJR-MacDonald, to advertise tobacco in the same terms as it protects the citizen’s right to criticise his or her government (RJR-MacDonald; see the critique in Hiebert 1999b: 10–15). Even if commercial expression has social utility (and it undeniably does), what compelling reason is there to subject its regulation to constitutional constraints normally reserved for political or artistic expression?
If the rights are taken as minimal standards, some dimensions of expression can be left to unencumbered legislative control. The constitutional rights can be defined as basic guarantees, not all-encompassing codes of expression.
Judicial deference to legislature or Executive in the limitation of rights
Courts can also leave room for legislative or Executive participation through judicial deference, the courts deferring, on some questions, to the judgment of the legislature or Executive. Under this model, a court does not insist that the legislature conform in all respects to the judge’s own opinion. The court recognises that there can be legitimate differences of view, and allows the legislature some latitude. This kind of deference is common in judicial review of administrative action, where courts have often recognised that administrative tribunals, as the primary decision makers, should be permitted to develop their own interpretations of law. After all, legislatures have conferred decision making power on the tribunals, often because the tribunals possess special expertise. The courts have expressed this deference in a variety of ways, depending on the drafting of the legislative regime. But a common formulation is that the courts should not intervene unless a tribunal’s decision is manifestly unreasonable.
Deference is controversial in the area of rights protections, however. Human rights are generally considered to be supervening norms that should stand beyond all legislative or Executive control. Their very purpose is to constrain government. It makes no sense (on this view) for courts to defer to their judgments. The Supreme Court of Canada has, for example, distinguished between constitutional and legislative norms in describing its deference towards administrative tribunals, holding that it wants to hear what administrative tribunals have to say about constitutional issues, but it will in the end make up its own mind (Cuddy Chicks v Ontario at 129–130).
At the same time, the Court has indicated that it will exercise some deference towards the legislatures in its judgments on limitations of rights, when those limitations involve the evaluation of sociological fact or the balancing of multiple interests, on the grounds that legislatures are equally or better placed to make those judgments. This has compensated, to some extent, for its expansive definition of such rights as freedom of expression. The rights may be defined broadly, but legislatures are given latitude in their limitation.12 Using this deference to compensate for an over-broad interpretation generates tensions, however, for on the one hand it appears to weaken the protection of things that genuinely should be rights, and on the other it continues to treat things that should not be rights as though they were worthy of constitutional protection. In Canada, that tension came to the fore in RJR-MacDonald (at 88 and following), where the Supreme Court drew back from the previous extent of its deference to limitations on advertising.
Nevertheless, significant deference continues. The Court has even upheld legislation that was consciously framed, by Parliament, on the basis of the judgments of dissenting, not majority justices, in a previous Charter decision. In the first decision, the Supreme Court had divided 5-4 on the circumstances in which a person accused with sexual assault must, in the preparation of his defence, have access to the records of rape counsellors who had supported the victim in the aftermath of the alleged assault (R v O’Connor). The dissent had argued for a much stricter scope of disclosure. Parliament clearly preferred the dissenting judgment, and framed legislation that departed in material ways from the majority’s position (An Act to amend the Criminal Code (Production of records in sexual offence proceedings) (Canada)). In a subsequent challenge, the Supreme Court upheld that legislation, on the grounds that Charter decisions always left open a range of permissible responses, and that Parliament was entitled to listen to the voices of parties other than the courts (in this case, those vulnerable to sexual violence) when framing their responses. The Court focused primarily on the seriousness with which Parliament had considered the rights concerns, and specifically invoked the idea of dialogue in justifying that degree of deference (R v Mills, especially at 19–20 and 37 and following).13
The interpretation of rights usually proceeds upon the assumption that one can, in theory, describe precisely what rights require and what they proscribe. It may well be impossible to do so in advance — adjudication tends to elaborate norms case by case, as the court is confronted with the circumstances in which the norms operate — but in principle rights are susceptible of detailed statement. This assumption may not hold true of all guarantees, however. Some rights may be incapable of detailed elaboration. Their normative content may operate at a purely abstract level, compatible with a wide range of detailed instantiations. These norms are called ‘framing norms’, for they provide a framework of entitlement without specifying precisely what that entitlement entails (Webber 2000b: 75–76).
One of the best examples of a framing norm is the indigenous right to self-government, which has been widely discussed but not yet recognised in Canada.14 It is unrealistic to think that self-government could ever be reduced to a highly specific set of requirements enforced by the courts. Its elements remain irreducibly abstract: the ability of indigenous peoples to participate in institutions that govern their communities; the right to have governmental institutions reflect norms determined in the community; the capacity of the community to control matters central to indigenous identity. The precise means by which these objectives could be achieved are as various as the communities themselves. Judicial review of such norms therefore takes a distinctive character. Instead of seeking to render the requirements progressively more explicit, it begins with the legislative measures under review, considers them in the light of the broad standards implicit in the norm, and decides simply on the basis of the broad standards whether they conform or not.
Framing norms are not as uncommon as they might appear. Indeed, to some extent all norms have an element of framework about them, for no matter how explicit they are made through the process of interpretation, they never address a specific case in all its particularity. At the last stage of any judgment, the judge must decide how the case falls in relation to the rule, using much the same approach as that described above. The difference between framing norms and other norms lies in the fact that in the former, one no longer expects to be able to render the norms increasingly concrete. Rather, one accepts that they will remain permanently abstract and one is content to work with that abstraction.15
The recognition of a class of ‘framing norms’ does make a difference. To take one example, in a series of cases in the late 1980s the Supreme Court of Canada held that the Charter’s guarantee of freedom of association included no special protection of the right of workers to bargain collectively. The Court came to this decision even though freedom of association had commonly been used, in international conventions, to refer to workers’ rights of collective action (ILO Convention No 87). The principal reason for the Court’s decision was its concern that the phrase, ‘freedom of association’, could not have been intended to constitutionalise Canada’s highly detailed collective bargaining regime. It noted that other countries had adopted very different forms of workers’ control (Re Public Service Employee Relations Act; PSAC v Canada; RWDSU v Saskatchewan). This reasoning is right as far as it goes. It would indeed be inappropriate to hold that the labour relations regime of the 1980s had been frozen in constitutional stone. But at the same time, the Court’s decision deprived freedom of association of one of its principal contemporary elements. The problem may have lain in the Court’s assumption that Charter rights had to be interpreted in a manner that progressively developed a single, precise, determinate content to the right, and that only that content could have normative force. Instead, freedom of association, insofar as it relates to workers’ collective action, may be best conceived as a framing norm, establishing a principle at a level of abstraction only, leaving wide latitude to legislatures as to how the right might be fulfiled.16
Framing norms have become much more familiar in recent years as a result of the development of the European Union. The European Commission’s directives take this form. They do not stipulate the measures that governments must adopt. They establish broad objectives and leave to governments how those objectives might be achieved.
For the purposes of this article, the chief significance of framing norms is that they presuppose a substantial measure of collaboration — of dialogue — between courts and legislature in the practical definition of rights. The courts set the framework; the legislature determines the means.
Finally, courts may simply find that some norms are not susceptible to judicial review and leave them entirely to the political process. Some norms (such as the commitment to equalisation payments between governments in s 36 of Canada’s Constitution Act 1982) may expressly set aside judicial review; these are the ‘programmatic rights’ discussed previously. But it is also possible that the courts might find that other provisions do not give rise to judicial review, either because a different remedy is specified or because the subject matter of the right is held simply to be inappropriate for judicial determination.17 In either case the definition of the right would be left within the legislative realm.
Mechanisms for legislative and Executive involvement in the definition and enforcement of rights
Finally, there are many mechanisms by which the legislature and Executive can themselves shape the meaning of rights.
Most importantly, the very existence of democratic government and popular participation serves to protect rights. Constitutional lawyers tend to think of rights entirely in terms of constraint — and specifically constraint of the legislature — but without doubt, the most powerful bulwark of rights in a democracy is broad participation by an engaged citizenry in the business of government. That participation contributes to the articulation of rights through the prominent role that rights play in parliamentary debate and Executive decision making. The ability to petition, to question one’s government, and to seek to change governments constitute powerful means of challenging abusive conduct and vindicating rights claims. It is profoundly wrong — indeed dangerously so — to believe that rights must operate only by constraining democratic action.
Moreover, the responsibility to enforce constitutional protections has at times been conferred directly on the legislature and Executive, rather than courts. This is true, for example, of the guarantees in the Canadian Constitution Act 1867 with respect to religious schools. Although schools are normally under provincial jurisdiction, those guarantees provide that an appeal from any act or decision ‘affecting any Right or Privilege of the Protestant or Roman Catholic Minority of the Queen’s Subjects’ lies to the Federal Governor-in-Council (the Federal Cabinet). If the Cabinet’s order is not obeyed, the Federal Parliament is empowered to pass remedial legislation (Constitution Act, 1867, s 93(3) and (4)). The Fourteenth Amendment to the Constitution of the United States (adopted following the Civil War to entrench the equality of the former slaves) also provided for Congressional action to enforce its terms.18 Legislative enforcement is particularly useful when trying to protect the rights of minorities within complex public institutions, where judicial remedies may prove inadequately flexible or incapable of creating detailed remedial administrative regimes.
Legislatures have also pursued rights within their own jurisdictions, enacting declarations of rights and establishing commissions for their enforcement. Undoubtedly, in Canada, the practical impact of these measures on the lives of individuals has far outstripped the effect of the Charter, for they apply to private as well as public action and guard against discrimination in employment, in housing, and in a host of other day to day contexts. The doctrines developed under these regimes have also had a significant impact on the interpretation of the Charter, notably in the concept of discrimination adopted by the courts (see, for example, Andrews at 16–19).
Prior examination and review
Constitutional Bills of Rights also stimulate a measure of self-scrutiny on the part of the legislature and Executive. This helps to ensure that legislation is kept within constitutional bounds. It may also furnish careful discussions of rights, which may in turn influence the courts if the legislation is subsequently challenged.
The self-scrutiny can occur in a myriad of places — in law reform commissions, administrative tribunals, the decisions of ombudsmen, and so on. In Canada, there has also been a structured process for reviewing legislation to ensure compliance with the Charter. In the period immediately following adoption of the Charter, existing legislation was examined and amended to bring it into conformity with the law officers’ understanding of the Charter’s requirements. Indeed, the entry into force of the equality guarantee in the Charter was delayed for three years to allow this process to occur. Now, as part of the regular procedure for drafting new legislation, each law is scrutinised to ensure it conforms (Department of Justice Act (Canada), s 4.1).19
The standards applied internally can be more demanding than those applied by the courts, the law officers taking a more expansive definition of the rights. This can happen haphazardly, as law officers make the wrong guess as to the interpretation that will ultimately be adopted by the courts. But sometimes governments purposely disagree with the courts’ interpretation. A good example occurred in relation to Ontario’s Sunday closing legislation. That law prescribed Sunday as the common day of rest in the retail sector. There were exceptions, notably one designed to accommodate individuals whose religious beliefs specified a day of rest other than Sunday, although this exception was restricted to stores under a certain size. The Supreme Court of Canada upheld the restriction as establishing an acceptable balance between the proprietors’ beliefs and the interests of the employees (R v Edwards Books). Yet, under a new government, the Ontario legislature amended the law so that the religious exemption was no longer limited by the store’s size.20
Arguments in justification
As already noted, the Canadian Charter itself implicitly invites governments to justify their legislation as part of Charter challenges. They sometimes do this in anticipation. They do it in argument before the courts.
Legislatures have also, in substance, re-enacted legislation that has been invalidated by the Supreme Court of Canada, or have enacted legislation that appears at variance with the reasons of the majority of the Court, advancing renewed justification under s 1. They have thereby expressed their disagreement with the Court’s decision and forcefully re-engaged the debate over justification. Of course, by declining to use s 33 and relying instead on s 1, they have left the ultimate decision on validity to the courts. But in at least one such case — the enactment of provisions to limit the circumstances in which an alleged victim’s rape counselling records would be released to the accused (discussed above) — Parliament’s arguments have been successful (R v Mills).21
Government consideration of rights concerns, as part of its exercise in justification, can therefore have an impact, at times dramatic, on the practice of judicial deliberation.
Political pressure on judicial decision
Finally, there is one last set of mechanisms by which governments can influence the interpretation of rights: direct political pressure.
We generally think of these mechanisms as being plainly antagonistic to rights, not contributing to any sort of ‘dialogue’. They often strike us as illegitimate exercises of power, almost certainly impairing the rights of individuals. This caution is wise. Judicial decision making is distinctive in its acute attention to the individual case, so that decisions are made in a manner sensitive to the situation, thereby doing justice (as far as human fallibility permits) to the interests of the individuals concerned. Political argument rarely attends so carefully to the individual case. If legislatures or Executives made all the decisions, we would run the risk of losing sight of individuals’ interests in our rush to achieve a social aim. Our defence of judicial independence and due process is therefore a profoundly important principle.
But the fact remains that mechanisms for political pressure do exist, and they are deployed (among other things) in disputes over the meaning of rights. They include fierce political criticism, pressure on the resources of the courts, the packing of courts, physical intimidation, forced removal, and ultimately constitutional amendment. Some of these mechanisms are legitimate in the sense that they are expressly contemplated by the Constitution — such as powers of judicial appointment, or the requirements for constitutional amendment. Nor are their effects always bad. They are sometimes used, it appears in retrospect, on the side of the angels. This was true, for example, of Franklin Delano Roosevelt’s court packing plan, which may well have influenced the US Supreme Court to overturn its doctrine of substantive due process, thus paving the way for the New Deal legislation (Tribe 2000: 1360–1361). Their contribution, then, is not unremittingly negative, although we generally do well to assume the worst.
But despite the evil or ambivalence of their effects, we should remember them. They bring home the fact that no matter how we insulate the judicial role, the relationship is one of institutional interaction, not complete autonomy. Realising that, we can work to structure the interaction in the most appropriate manner, including establishing protections for judicial autonomy.
Moreover, in a backhanded way, recognising the existence of these forces relativises and clarifies our claims about judicial review. I noted at the beginning of this paper that justifications for judicial review frequently rely on the need to guard against the grossest of abuses. There is a real question, however, whether judicial review is much of a bulwark when a political system decays to such a desperate degree. Even if the courts do stand up (and history shows that they often do not),22 the political process has levers at its command to overcome the courts if the democratic ethic of rights protection is absent. This is not an argument for ignoring the potential for abuse, nor is it an argument against human rights adjudication in normal times. But it does suggest that the remedies to gross abuses lie within a broader social/political approach, and that judicial review needs to be justified in terms of its relative contribution in normal, not abnormal times. This too drives us back upon the analysis of interaction, relationship and balance.
Dialogue in three strategies for human rights protection
There are therefore a wide array of mechanisms to provide for institutional dialogue in the definition and enforcement of rights. Nor is this account exhaustive. The reader will doubtless have added his or her own examples.
In this section, I will examine in more detail three means of rights protection — all employed in Canada, one in Australia — to assess the form of institutional relationship created. The mechanisms are:
• implied rights;
• express constitutional guarantees, subject to legislative derogation; and
• a statutory Bill of Rights, protected by a manner and form requirement.
I will focus primarily on the second and third, citing the first principally by way of contrast.
As mentioned above, this mechanism involves the determination that certain rights are implicit in the Constitution. Commonly, the rights are founded upon constitutional terms that establish democratic institutions (from which rights to political free speech are derived) or establish the courts (from which norms to protect judicial independence are derived). This is certainly the case in Australia, where the High Court has recognised a right of free political expression, working off the provisions establishing representative government in the Australian Constitution (notably ss 7 and 24, which state that members of the Senate and House of Representatives are to be ‘directly chosen by the people’). In the first of these cases, Brennan J (as he then was) summarised the essential argument (Nationwide News at 48–49):
... where a representative democracy is constitutionally entrenched, it carries with it those legal incidents which are essential to the effective maintenance of that form of government. Once it is recognised that a representative democracy is constitutionally prescribed, the freedom of discussion which is essential to sustain it is as firmly entrenched in the Constitution as the system of government which the Constitution expressly ordains.
This right was used to strike down laws that prohibited criticism of members of the Industrial Relations Commission and that limited television advertising in elections (Nationwide News Pty Ltd v Wills; Australian Capital Television Pty Ltd v Commonwealth). It also prompted the High Court to revise the law of defamation insofar as it applies to criticism of public officials (Theophanous v Herald & Weekly; Lange v Australia Broadcasting Corporation).
Other rights too might be implied from the Constitution. In McGinty v Western Australia, the Court considered whether the Western Australian Constitution required rough equality in the determination of electoral boundaries (so that each person’s vote would be of approximately equal value). The Court refused to find such a requirement. The High Court has declined to hold that a general right of equality can be derived from the Federal Constitution (Leeth v Commonwealth; Kruger v Commonwealth). But other principles with strong rights implications have been implied. In both Australia and Canada, the courts have developed constitutional guarantees of the independence of the judiciary, from which additional limitations have been derived: in Canada, judicial review of the jurisdiction of administrative tribunals has been constitutionalised on this basis; in Australia, the principle of the separation of powers has been used to strike down a law designed to keep a named offender in preventative detention beyond the end of his original sentence.23
Implied rights involve the finding that rights have been constitutionally entrenched, but do so in a particularly confined way. They do not directly promote dialogue between legislature and courts — they do not involve any special interaction between court and legislature in the definition of rights — but they do tend to leave substantial latitude to the legislature.
Implied rights are confined by the very fact that they are based on implication. There is no general, unambiguous statement of the rights. The courts must, above all, establish that the rights are entrenched, even though the document makes no reference to them. There is often disagreement over the standard to be applied when finding an implied right: does the right have to be necessary to any reasonable interpretation of the text, or is it sufficient that it be broadly congruent with the Constitution’s principles? There is the problem of distinguishing between, on the one hand, a mere assumption on which a constitutional provision is based and, on the other, a true implication, in which the assumption itself is constitutionally protected (Australian Capital Television at 135, per Mason CJ).24 Moreover, the very fact that an implied right is not express, yet has such a significant impact on legislative authority, encourages judicial restraint. Courts are less likely to fall prey to one common hazard of constitutionalised Charters of Rights — the temptation towards maximal rather than minimal definitions of rights.
Finally, an implied right always depends upon cues in the text, and this limits the range of rights that can be implied. The rights tend to be political, strongly tied to the integrity of public institutions. Hence, they protect political speech (on the basis that the Constitution establishes representative government) but not artistic speech. In Australia, an implied guarantee of equality foundered precisely because the right claimed was general and sweeping, lacking a clear constitutional touchstone (Leeth; Kruger).
Implied rights therefore have a restricted scope, tightly tied to the express provisions of the Constitution. They tend to be focused solely on political freedoms. And because the rights are inferred, because they are derivative, never primary, and because they are the product of complex interpolation, there is always the potential for adjustment into the future. These limitations are virtues if one believes that constitutionally entrenched guarantees should be restricted to truly fundamental principles, dealing with the citizen’s political relationship to the state.
Express rights, subject to derogation
Under implied rights, judicial review has a restricted role simply as a result of the allusive and ambiguous form in which the rights are expressed (or unexpressed). In the second mechanism for rights protection examined here — express rights subject to derogation by the legislature — the rights are stated just as they are in any constitutionalised Bill of Rights, but they are then subject to legislative derogation: the legislature can choose to override them in a manner expressly provided in the constitution.
The Canadian Charter of Rights and Freedoms is the principal example of this strategy. Section 33 of the Charter reads in part:
(1) 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 s 2 or ss 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.
Certain conditions limit the use of the override, although the section itself keeps these to a minimum and the courts have been reluctant to find that more exist by implication. First, the override is only applicable to certain rights (those found in ss 2 and 7–15). Paradoxically, these tend to be the most broadly accepted rights, including freedom of religion, freedom of association, freedom of expression, rights arising in the context of criminal proceedings, and equality. The chief exclusions are rights to vote and, significantly, mobility rights and official languages. The last two exclusions give a clue to the rationale underlying the choice to include or exclude. Mobility rights and language rights were highly controversial among a number of provinces at the time of the Charter’s adoption, the former because of potential interference with provincial policies with respect to economic development and absentee ownership, the latter because of Quebec’s desire to require the use of French in certain contexts. At the same time, the Federal Government was intensely committed to both sets of provisions as ways of requiring both economic integration and bilingualism. The override excluded precisely those rights most likely to be subject to derogation.
Secondly, any invocation of s 33 is subject to a five year sunset clause. At the end of five years, the derogation lapses (although it can be re-enacted indefinitely). This requires that derogations be justified repeatedly, at periodic intervals. Five years was chosen as the period because that coincides with the maximum term of any government.
Superficially, s 33 of the Canadian Charter is similar to the ‘notwithstanding’ clause in the Canadian Bill of Rights, the statutory Bill which I introduced above and will discuss further below. Each clause apparently permits the legislature to derogate, and in each case, derogation must be express in the derogating act itself, in order to trigger a popular debate in which the government will be compelled to justify its actions. Indeed, in this respect the Canadian Bill of Rights served as the model for s 33 of the Canadian Charter. But in another respect, the two instruments are very different. In a statutory Bill of Rights, it is the very ability to derogate that renders the Bill effectual; that capacity transforms the Bill from what would be an invalid constraint on the substance of legislation to one of procedure only. Paradoxically, then, it is the very ability to set the rights aside that renders the statutory instrument binding on later legislation. In the Charter, on the other hand, the rights are constitutionally entrenched; they control, of their own force, all legislation. There, the notwithstanding clause does nothing but allow the legislature to set the rights aside.
In the Charter, s 33 was a compromise between parliamentary sovereignty — the British constitutional doctrine that treats the legislature as the supreme branch of government — and judicial review. At the time of patriation of the Canadian Constitution in the early 1980s, the governments of Saskatchewan and Manitoba opposed the inclusion of an entrenched Charter of Rights precisely because they believed it would shift decision making from the legislature to the courts. Saskatchewan’s left leaning Government cited the US constitutional doctrine of substantive due process (which had impeded the adoption of much social legislation up to the 1930s), as a reason for limiting the powers of judicial review. Those provinces agreed to the Charter only on condition that it include a clause permitting legislatures to derogate from its guarantees. Section 33 was therefore adopted precisely because of misgivings, among some governments, about rights review.25 The section has been seen by many (not least in Canada), as inconsistent with the very idea of a Charter of Rights. Some have argued that it should be removed (see Russell 1991: 301–302; Mandel 1994: 87 and following; Hiebert 1999a: 31–32).
Such a stark opposition is unfortunate, however. There is a justification for the notwithstanding clause — perhaps the best justification — that sees it as establishing an appropriate balance between courts and legislature, through the facilitation of dialogue between those institutions with respect to rights. According to this view, s 33 is not anti-rights; it simply allows the legislature to participate in their interpretation and application (Weiler 1984: 79–92; Russell 1991).
The justification starts from the premise that there can be legitimate differences over the definition of rights. There is reason, then, to provide an outlet through which alternative understandings of rights can be advanced and defended. This is especially true when, under entrenched Charters of Rights, courts’ interpretations can be very difficult to correct (the usual mechanisms being revision by the courts themselves or constitutional amendment).
It also recognises that different types of institutions carry different advantages — and different biases — in rights definition (Webber 1993: 218 and following; 2000a: 137–144). Courts excel at the sober analysis of specific claims in a manner that pays close attention to individuals and that is isolated (though only in relative terms) from broader political concerns. They are especially effective where all considerations relevant to a particular dispute are well defined, so that interests are crystallised and the parties implicated are known and represented. They depend, in their procedure, on parties appearing before them and arguing all relevant considerations. They have great difficulty dealing with issues in which interests are widely distributed, so that no one party has sufficient interest to appear and the only viable procedural approach would involve inquiry and investigation. Courts focus overwhelmingly on protection against government interference, not positive action by government to advance the interests of individuals or groups, for judicial review is conceived (rightly) as a restraint on government action, rather than the assertion of an alternative or complementary governmental authority. The net effect of this, however, is that judicial review tends to privilege private over public action, for only the latter is subjected to Charter scrutiny. Finally, courts do best when the interests are simply aligned — when the dispute takes a bipolar (with two opposed positions) rather than polycentric form (where a multitude of competing interests need to be balanced). The former accords most closely to courts’ adversarial procedures, in which, paradigmatically, two parties contend, marshalling their own sets of evidence. Indeed, because of their adversarial procedure and their bias towards the restraint of government, courts are prone to simplify rights claims so that they conform, as closely as possible, to a bipolar form — the individual pitted against the state — even when the state may simply be serving as a stand-in for the broadly distributed interests of individuals in society at large.
Legislatures have strengths and weaknesses that broadly correspond to those of courts. They are much better at dealing with polycentric issues and with widely distributed interests. They have procedures well adapted to investigation and inquiry. They are able to act positively, deploying resources to attain ends rather than standing back to let the situation be determined by the play of private forces. Those advantages come with disadvantages. Legislatures can respond to a perceived balance of interests — or to a strongly articulated majority interest — in a manner that has insufficient regard for minorities or individuals. Because they work by majority rule, they can weigh all individuals’ self-defined interests equally, when some interests may deserve added weight because of their fundamental character (religious beliefs; freedom of expression; the freedom to associate; mobility rights). In their pursuit of general social interests, they can lose sight of the particular situation of individuals.
Section 33, it might then be argued, achieves an appropriate balance between the benefits of the legislature and those of the courts in the definition of rights. It provides for judicial review on a full array of rights concerns, but it also allows legislatures to have the last word if they disagree fundamentally on how rights have been defined and applied. Even then, rights concerns are highlighted because of the requirement that the legislature be explicit in its intention to override the Constitution’s declaration of rights. This triggers democratic scrutiny, which only the strongest justification is likely to surmount.
This may be the best justification for s 33, but in practice the section has been less than successful at fostering a productive dialogue. It has been used in seventeen situations (although in one the act was never proclaimed in force). The experience has been decidedly mixed. The use of s 33 has been concentrated in few hands. The province of Quebec was responsible for fourteen of the seventeen instances; although other jurisdictions have from time to time vehemently disagreed with Charter decisions, all except Alberta have tended, in recent years, to shy away from s 33 and to use renewed justification under s 1 instead. Moreover, in a great many instances, the use of s 33 has escaped public notice almost entirely. Were it not for the recent work of Tsvi Kahana, thirteen of those instances would have remained in utter obscurity (Kahana 2000: 255–291)26.
Quebec was the first jurisdiction to use the clause. In June 1982, soon after the proclamation of the Charter, it passed a law invoking s 33 with respect to all existing Quebec statutes (An Act Respecting the Constitution Act 1982 (Canada)). It then proceeded to insert a comparable clause in each subsequent act. This was done to protest against the patriation of the Constitution over Quebec’s objections. The Quebec Government, under the indépendantiste Parti Québécois, had participated in the negotiations leading up to patriation, but had ultimately rejected the package agreed to by Ottawa and the other provinces. Quebec took the position that it should be able to veto constitutional reform, given that it was the only province with a French speaking majority. When patriation proceeded regardless, it opted out of the new Charter to the maximum extent possible (Webber 1994: 113 and following).
This use of the notwithstanding clause was highly unusual, focused as it was on the legitimacy of the constitutional order as a whole. The blanket use of the clause was challenged in Ford v Quebec, the plaintiffs arguing that the wording of s 33 contemplated a more precise or explicit derogation. They suggested that a valid derogation should occur only after the courts had ruled on the rights concerns, should identify precisely what rights were to be set aside, or should specify the aspect of the law to be protected from scrutiny. They argued that this interpretation was justified by the signalling function of the notwithstanding clause; if there were no express identification of the rights in question — if the invocation of the clause simply exempted all statutes from rights review — no sensible debate over the issues could be joined. This may have been right, but the Supreme Court saw that such requirements, once established, were likely to become disguised substantive limitations on the ability of the legislature to invoke s 33. It declined to start down that road and instead upheld the blanket invocation of the notwithstanding clause. In so doing, it confirmed that s 33 was a recognition of parliamentary sovereignty within the rights context, one that was not to be hedged about by judicially imposed restrictions.
In the Quebec provincial election of 1985, the federalist Liberal Party defeated the Parti Québécois. The new Government stopped invoking s 33 in every piece of legislation, but it did not repeal previous invocations. Instead, it let them expire as they reached the five-year limit. This exposed, for the first time, Quebec’s legislation to full Charter review. Since then, Quebec has invoked s 33 on thirteen occasions. Twelve concerned highly complex legislative regimes dealing with pensions (raising concerns of gender discrimination), the role of religion in education (freedom of religion), and agricultural development (age discrimination). These twelve instances received no public attention (Kahana 2001: 255–291). The reasons are obscure, but are probably due to the arcane nature of the legislative regimes, the limited legitimacy of the Charter in Quebec (in each case the first use of the clause occurred shortly after the blanket invocations were discontinued), and the fact that in the education acts, the provisions were closely related to constitutionally entrenched guarantees of religious school boards.
In 1988, however, there was a very public and highly contested use of the clause. This occurred in response to the decision in Ford, noted above. That decision had struck down provisions of Quebec’s language law that required that commercial signs and company names be in French only. Ironically, the most important sections were invalidated on the basis of Quebec’s own, statutory, Charter of Human Rights and Freedoms, not the Canadian Charter, for the sign law provisions were protected from the Canadian Charter by a still subsisting notwithstanding clause.
Ford generated a fierce political controversy in Quebec. Most francophone Quebecers might have been willing to relax the province’s language laws, but they did not want relaxation to be dictated by the Supreme Court of Canada, in a manner that removed control from Quebecers’ elected representatives. Quebec therefore re-enacted the sign law provisions, in modified form, this time protecting them against the Canadian Charter through the use of the notwithstanding clause (An Act to Amend the Charter of the French Language 1988 (Quebec)). Here, the clause was used precisely in the manner contemplated in the ‘best justification’ for s 33 offered above; a legislative majority set aside rights review on the basis that freedom of linguistic expression, in commercial signage, was outweighed by other public considerations.
The Quebec Government paid a significant price for its use of the clause. Anglophone members of the government resigned, many Quebec anglophones left the provincial Liberals, and an English speaking protest movement took hold. Most importantly, anger in the rest of Canada over Bill 178 (as the Act was known) contributed to the defeat of the Meech Lake Accord, a set of constitutional amendments designed to address Quebec’s constitutional grievances. The events demonstrated, then, that broad popular debate could be generated by the use of s 33 and that governments could incur heavy costs as a result — even when, as in Quebec, they retained the support of majority opinion within the province. Five years later, once passions within the province had cooled, the Liberal Government again amended the sign law so that it fell within the limits set by the Supreme Court of Canada and allowed the notwithstanding clause to lapse — although by that time, much of the damage had been done (An Act to Amend the Charter of the French Language 1993 (Quebec); Webber 1994: 138 and following).
The notwithstanding clause has only been used on three occasions outside Quebec. The first was in the Yukon Territory in 1982. This provision, never proclaimed in force, was designed to protect aboriginal nominations to the Land Planning Board. It received no public attention and was almost certainly unnecessary given the distinctive treatment of indigenous peoples in the Canadian Constitution (s 39(1) of Land Planning Act 1982 (Yukon); Kahana 2001: 258 and 266). In 1986, Saskatchewan used the clause to protect a statute adopted to end a strike. This was ultimately proven unnecessary when the Supreme Court held, in a separate case, that the right to strike was not protected under the Charter’s guarantee of freedom of association. The invocation of the clause elicited real critical debate, although the Supreme Court’s subsequent decision took some of the force out of that discussion (s 9 of SGEU Dispute Settlement Act (Saskatchewan); Kahana 2001: 269–270; RWDSU). In 2000, the Alberta legislature amended its Marriage Act to define marriage in exclusively heterosexual terms, invoking s 33. Again, there was public opposition, although muted, perhaps because capacity to marry falls within federal jurisdiction and Alberta’s enactment would therefore have been purely symbolic (Marriage Amendment Act 2000 (Alberta); Kahana 2001: 268–269). In 1998 the Alberta Government seriously contemplated the use of s 33 in two other circumstances that would have had a material effect on legal rights. In each case it retreated, in one instance in response to a very strong public reaction.27
Thus, the clause has been invoked very rarely outside Quebec, and even in Quebec most of its uses have been buried in highly complex regimes, in six cases closely related to existing guarantees of religious school boards. Indeed, it is fair to say that s 33 has achieved very little legitimacy outside francophone Quebec, with the partial exception of Alberta. It was fiercely criticised by many rights advocates at the time of the Charter’s adoption. There are still forceful calls for its removal, as well as periodic demands that stricter conditions be placed on its use (see Russell 1991: 301–302; Mandel 1994: 87 and following; Hiebert 1999a: 31–32).28 Legislative intervention under s 33 tends to be seen as plainly and simply anti-rights, not as a means by which legislatures can insist on their own interpretations of rights.
Section 33 has, in the end, made very little difference to the way in which rights are conceived under the Charter, at least in English speaking Canada. The mythology of judicial review on rights grounds remains very strong — the sense that courts protect rights, unproblematically defined, against legislative or Executive oppression. There is little conscious weighing of the strengths and weaknesses of legislatures and courts as interpreters of rights. I suspect that this is due in large measure to the fact that the rights are still enshrined in a constitutional text. Regardless of whether the constitution invites legislative participation — as the Canadian Charter arguably does in s 33 — the very context of the clause suggests that when it is invoked, fundamental rights are set aside. Constitutions carry a powerful symbolic charge. They are perceived to be basic laws, setting out the premises on which society should be governed, with a measure of fixity and permanence that is beyond politics. Against this backdrop, s 33 looks very odd indeed, introducing variability and the will of a transient majority into the very application of the Constitution.29
A statutory Bill of Rights, protected by manner and form requirements
This brings us to the last strategy of rights protection examined here: the declaration of rights in statutory form, protected by a manner and form requirement.
I have already described the nature of this mechanism above. The rights are expressed in an ordinary statute. They can be set aside, however, by the legislature expressly stating that an act applies notwithstanding the Bill of Rights. That ability to set the rights aside transforms the act from a substantive to a procedural limitation, permitting it to bind the legislature for the future.
Statutory Bills of Rights are binding (in Canada), but only in relative terms. In fact, there are two ways in which their effect can be modified or displaced: i) a legislature can stipulate that an act applies notwithstanding the Bill of Rights; or ii) it can amend the Bill of Rights itself, changing the very expression of the rights (and indeed there have been a number of such amendments to provincial Bills of Rights in Canada).
The Canadian Bill of Rights, adopted by the Federal Parliament in 1960, was the first example of this legislative method, although the same approach was adopted in Quebec and Alberta. The record of the Canadian Bill of Rights was not particularly noble. Prior to the adoption of the Charter, it was used to strike down only one legislative provision (R v Drybones). For much of its history, there was some question as to whether it rendered legislation invalid or merely provided standards to be applied when interpreting legislation. It was prone to inconsistent, timid, and downright poor interpretations (Tarnopolsky 1975). This has sometimes been blamed on its statutory, not constitutional, character.30 Although this character may have played some role, one suspects that the problem was more general: the Canadian Bill of Rights was adopted at a time when justiciable rights were uncommon outside the US, and Canada’s legal culture and legal education were ill adapted to rights review. Judicial interpretation of statutory Bills of Rights and of human rights codes (which are applicable to particular types of services and do not purport to bind the legislature) has been much bolder and more sophisticated in recent years, suggesting that it was legal culture rather than the Canadian Bill of Rights’ statutory character that determined that instrument’s weak interpretation. Indeed, the interpretation of the Canadian Bill of Rights itself has improved.
Because of the inadequate interpretation of the Canadian Bill of Rights, and because all of Canada’s statutory Bills of Rights have, since the early 1980s, largely been displaced by the Canadian Charter, it is difficult to draw satisfactory conclusions. I strongly suspect, however, that statutory Bills of Rights have advantages that are worth revisiting, especially for those nations, like Australia, that do not have constitutionalised Bills of Rights and are considering adopting them. In particular I believe that statutory Bills of Rights may provide a better framework for dialogue between legislature and courts.
As argued above, one of the chief obstacles to dialogue under the Canadian Charter of Rights and Freedoms has been the symbolic force of a constitutionalised Charter, which drives a deep gulf between the constitutionally protected rights and all other interests. The rights assume a superordinate importance, resistant to balancing. Any attempt by the legislature or Executive to define rights or determine their application is viewed with extreme scepticism, as an illegitimate attempt to impair fundamental liberties. The ability of legislatures to derogate in s 33 becomes virtually unusable.
Statutory Bills of Rights escape some of these evils. The act itself, as an ordinary statute, carries nothing like the same symbolic charge. It is therefore less susceptible to the view that the rights as defined by the courts are immutable, to be kept exempt from any legislative tampering. On the contrary, the statutory rights are emphatically the product of legislative action. The legislature is a collaborator in their adoption and definition, not an antagonist. It has a manifest claim to share in their continued definition, not only through express derogation but more importantly through the amendment of the statutory Bill itself (which clearly involves redefinition, not just the setting aside of the rights). There is thus a viable avenue through which the particular strengths of legislatures — the weighing of dispersed social benefits; balancing in polycentric issues; the ability to inquire and investigate; an understanding of the positive value of state action in promoting social welfare — can make themselves felt in rights jurisprudence.
Moreover, the statutory form avoids some problems with the interpretation of constitutional rights, precisely because the statutory rights are not set apart in a sacrosanct document. The rights can be considered in relation to other interests without an exaggerated gulf between the two. Moreover, their application is less likely to be distorted by the high symbolism and nationalistic aims that sometimes afflict constitutional instruments.
Yet, because of their relative binding force, the statutory provisions do continue to protect against inadvertent or surreptitious legislative impairment of rights. They do this, not by removing a swath of decisions from political debate, conferring them exclusively on the courts (as constitutional provisions tend to do). Rather, they work by stimulating debate, warning that rights may be in peril and thereby prompting (one hopes) vigorous popular engagement in the issues.
In short, through the adjudication of the rights themselves, statutory Bills of Rights provide an opportunity for the kind of careful, individualised and relatively dispassionate analysis of rights concerns that we expect of the courts. That analysis has real effect: the courts can invalidate legislation, certainly protecting against inadvertent impairments and even providing some protection against purposeful rights infringements. Yet they do this in a way that relies upon provisions that are defined by the legislature, and they expressly depend upon the support of an engaged citizenry for their ultimate efficacy. Thus, they feed the democratic process, rather than detracting from it. Statutory Bills of Rights are therefore most compatible with the view, which I accept, that democratic institutions are an indispensable bulwark against repressive conduct — a bulwark which, once eroded, cannot adequately be replaced by judicial institutions, for the courts themselves ultimately depend upon a measure of popular support for their efficacy.31
The range of institutional mechanisms for rights protection is therefore considerably broader than we generally think. Even in jurisdictions most committed to judicial review, rights protections involve a complex interaction among institutions. When addressing issues of constitutional design, there is a variety of ways in which that interaction can be structured.32
My own preference, in societies with a strong rights culture like Canada or Australia, is for mechanisms that permit a continued role — even a predominant role — for the legislature. There are many reasons for this preference, some of which have been evident in this article and which I have discussed further elsewhere. But one essential one is that in such societies, most rights issues do not involve gross violations but rather difficult issues of judgment. I cannot see why the legislature should have no legitimate part in those decisions.
Section 33 of the Canadian Charter of Rights and Freedoms may be best justified in these terms, but it has not fulfiled that promise, for it still participates in the high symbolism of a constitutionalised Bill of Rights, which forcefully impedes institutional dialogue. Section 33 may have served as a pressure valve at the time of adoption of the Charter and may continue to do so as long as there is extensive dissatisfaction with the Canadian Constitution in Quebec, but it is unlikely to provide a continuing mechanism for dialogue. Surprisingly, the strongest examples of institutional dialogue under the Canadian Charter have occurred within the framework of arguments over justification under s 1.
Statutory Bills of Rights have a better chance of capturing the appropriate balance. There, the courts retain a role, but it is a role focused upon the area where courts undeniably have a strong institutional advantage: the application of general norms to particular situations, in a manner that maximises the chance that the particularity of those situations will not be overlooked or overpowered in the rush to achieve a general object. The use of an ordinary statute does mean that rights are vulnerable to a cynical or repressive legislative initiative. There are institutional impediments to this, through the promotion of political debate and the opportunity for judicial pronouncement. But at the end of the day a determined majority can have its way. The peremptory control of a constitutionalised Bill of Rights may be appropriate in some contexts, then, especially where there are permanent political minorities, or where the weakness of a country’s rights culture makes strong constitutional statement desirable.
But it is important to realise that a capacity to overpower rights does not necessarily mean that rights are in jeopardy. On the contrary, maintaining legislative responsibility can foster a broader and more inclusive rights debate, one that reinforces a concern for rights within the popular milieu generally. Nor should we exaggerate the efficacy of judicial review. It, too, can be swayed by an inflamed public. It, too, can lose its nerve.
In countries with a strong rights culture, questions of rights serve as prompts to reflection and institutional self-restraint — as occasions to confront once again the difficult mediation between individual and society. They serve only very rarely as the occasion for heroic stands against oppression. Structures that are more nuanced than constitutionalised Bills for rights — structures that provide for a genuine dialogue between the courts and the democratic process — may well be most appropriate.
* Professor Jeremy Webber is Canada Research Chair in Law and Society, Faculty of Law, University of Victoria; previously Professor and Dean, Faculty of Law, University of Sydney. This is a version, revised for an Australian audience, of a paper originally published in Sadurski (2002), a collection exploring the comparative experience of constitutional adjudication with specific reference to central and eastern Europe. My thanks to Eric Ghosh for his able research assistance and to Eric, Janet Hiebert, Peter Hogg, Rod Macdonald, Andrew Petter, and Wojciech Sadurski for their trenchant comments on an earlier draft of this paper.
1 See Hogg and Bushell (1997); Hiebert (1999a); Hiebert (1999b); Roach (2001); Hiebert (2002). For a precursor, see Tushnet (1995). Hiebert initially preferred the term ‘conversation’, but her most recent work rejects dialogic metaphors and now speaks of ‘a relational approach’: Hiebert (2002), especially 50–51. For criticisms of the idea of dialogue, see Cameron (2001), who rejects the idea of dialogue, at least within s 1 of the Canadian Charter of Rights and Freedoms (the Charter), because it undermines the supremacy of constitutional interpretation without the procedural sanctions of s 33; and, from opposing positions, Manfredi and Kelly (1999), who argue that Hogg and Bushell exaggerate the extent to which there is true dialogue; and Petter (2003), who challenges the extent to which dialogue theory has content or coherence. There is a rejoinder in Hogg and Thornton (1999). The courts themselves have embraced the language of dialogue (Vriend v Alberta at 438–439 per Cory and Iacobucci JJ; Corbière v Canada at 63 per L’Heureux-Dubé J; R v Mills at 19–20 and 37–38 per McLachlin and Iacobucci JJ; Little Sisters Book and Art at 294 per Iacobucci J; R v Hall at 554 per McLachlin CJC and 581-582 per Iacobucci J; Sauvé v Canada (Chief Electoral Officer) at 597 per McLachlin CJC and 628-631 per Gonthier J. The last two cases were decisions of narrow majorities (5 judges to 4), where the implications of ‘dialogue’ were strongly contested.
Hogg and Bushell are chiefly concerned with a less demanding form of dialogue than that which is the focus of this paper: the ability of legislatures to adapt their legislation to withstand constitutional scrutiny, once a law has been declared invalid. This paper concentrates not on legislatures’ responses to judicial decisions, but on legislatures’ impact on the interpretation and application of the rights themselves.
2 I address some of the issues in Webber (2000a).
3 See, however, art 79 of the Basic Law of the Federal Republic of Germany, which purports to prohibit amendments to substantial sections of the Constitution. The Supreme Court of India has also held that the basic structure and framework of the Indian Constitution is unamendable, even though the express terms of the Constitution impose no such limitation (Bhardwaj 1995: 10–13).
4 Australia’s Racial Discrimination Act 1975 has also been used to interpret later legislation. There was considerable discussion when the Native Title Act 1993 was enacted, and when that act was amended in 1997–98, over whether it should be made expressly subject to the Racial Discrimination Act so that any provisions offending the Racial Discrimination Act would be rendered inoperative. On both occasions, Parliament ultimately settled on language that merely invited the courts to use the Racial Discrimination Act to interpret the Native Title Act (Brennan 1998: 69–71, 73–74 and 85).
5 There is greater doubt as to the binding effect of manner and form requirements in Australia, especially with respect to Commonwealth legislation and non-constitutional enactments of the states. The most substantial discussion is found in Winterton (1980). He concludes that the Commonwealth could bind itself to a requirement like that of the Canadian Bill of Rights — as indeed had been contemplated in the Commonwealth’s Human Rights Bill of 1973 (never passed) and discussed at the time of the adoption and amendment of the Native Title Act 1993 (see previous note).
6 It is possible, depending on the nature of the federal regime, for a statutory Bill of Rights enacted at one level to bind legislation enacted by the other level. This is the case with the Racial Discrimination Act, adopted under the Commonwealth’s external affairs power to incorporate the International Convention on the Elimination of All Forms of Racial Discrimination into Australian law. Because it benefits from the paramountcy of federal legislation over inconsistent state legislation, it binds the States without any need to resort to manner and form requirements (Koowarta v Bjelke-Petersen).
7 The celebrated Australian case, Australian Communist Party v Commonwealth, also used division of powers reasoning in the service of human rights ends. There, the High Court considered the validity of the Communist Party Dissolution Act 1950, which attempted to rely on the defence power or on an implied power to defend governmental institutions to declare the Communist Party unlawful and prohibit Communists from serving in the Commonwealth Government and in certain unions. The Court held that the statute was beyond the scope of both powers. Similarly in Adelaide Company of
Jehovah’s Witnesses, the High Court struck down a wartime regulation purporting to dissolve a Jehovah’s Witnesses organisation on the grounds that it went beyond what was necessary for the prosecution of the war and was therefore beyond the scope of the defence power. This last decision is especially interesting for there is an express guarantee for freedom of religion in the Australian Constitution (s 116), which the Court decided would not have been violated by the regulation. This preference for division of powers reasoning may reflect the fact that at least until recently, judicial review on division of powers grounds had more legitimacy than review on the basis of rights, perhaps because judgments based expressly on rights challenge legislative policy much more directly.
8 Canadian cases often considered to open the door for implied rights include: Re: Alberta Statutes at 133–134 (per Duff CJ); Saumur v City of Quebec at 354 (per Kellock J) and 363 (per Locke J); Switzman v Elbling at 328 (per Abbott J); OPSEU v Ontario at 57 (per Beetz J). The Australian decisions are discussed below.
9 Hogg and Bushell (1997: 101–104) give a number of examples where legislatures or government agencies explicitly state, sometimes in the preamble to an Act, why the measure constitutes a reasonable limitation within the meaning of s 1 of the Charter.
10 For unamendable constitutional provisions, see note 3 above. Dicey (1959: 128–131) argued that the fact that a constitution was unamendable might actually encourage revolution, citing the rapidity with which supposedly unamendable French constitutions had been replaced.
11 This approach to interpretation and that in the next section also allow more room for the kind of dialogue emphasised by Hogg and Bushell (1997) — legislative responses to judicial decisions that meet legislative aims and yet conform to judicial decisions — for more latitude is permitted to legislative action. For an example, see Roach (2001: 271–273). It is puzzling, however, that Hogg and Bushell cite (85–87) situations in which the courts direct how legislation might be redrafted as examples of the potential for dialogue. To the extent that the court dictates how a legislature must respond, the relationship hardly seems one of dialogue but rather one of compliance (as Hogg and Bushell acknowledge, though half-heartedly, at 98). See Hiebert (1999b: 10–15) for discussion of one particular instance, and Manfredi and Kelly (1999) generally.
12 See, for example, Irwin Toy v Quebec at 625–630; McKinney v University of Guelph at 665–673. Hiebert (1999a: 28) rightly notes that the Court has rarely held the legislative objective to be insufficient to justify limiting the right. Judgments have, with very few exceptions, been decided on the proportionality of the measures.
13 For background to R v Mills, see Hiebert (1999b: 16–25; 2002: 107–116). Cameron (2001) criticises the decision in Mills on the basis that it grants the legislature licence to ignore the courts’ interpretation of the Charter, undermining the supremacy of the courts in constitutional review. Moreover, Parliament can do this without paying the institutional price demanded in s 33.
14 See, for example, the form of the right in the failed set of constitutional amendments known in Canada as the Charlottetown Accord, ss 29 and following. An inherent right of self-government was recognised, but its instantiation was to be determined through negotiations supervised by the courts. The Corbière decision evinces a similar spirit. It, too, concerned the design of indigenous governmental institutions (although not under an inherent right). There, a statutory provision denying non-resident band members the right to vote was ruled invalid under the Charter. But the declaration of invalidity was suspended for 18 months in order to permit a replacement provision to be developed.
15 It is important, however, that framing norms are limited to situations in which norms genuinely are incapable of further precision. Otherwise, court decisions can leave legislatures in an invidious position, having little sense of how legislation might be redrafted in order to render it valid.
16 The Court has since moved away from the rigour of the ‘labour trilogy’ to find that the exclusion of agricultural workers from protections against unfair labour practices offends freedom of association (Dunmore v Ontario (Attorney General)). The decision was based on the Court’s ability to review and, if warranted, order the extension of underinclusive legislation designed to protect constitutional rights. The protections at issue concerned the core ability to form organisations without interference, not full rights to bargain collectively. But some aspects of the decision suggest that the Court may be inching towards the recognition of framing norms. The Court distinguishes the general value of freedom of association from the specifics of any statutory regime and notes (at 227) that the collective bargaining statute in question ‘does not simply enhance, but instantiates, the freedom to organise’. That said, there are reasons to be grateful for the non-recognition of collective bargaining under the Charter (Webber 2000a: 141–142).
17 Canada’s denominational schools guarantees (Constitution Act 1867, s 93(3) and (4)) have been analysed as having two dimensions, one enforceable by both the courts and the federal executive and legislature, the other enforceable only by the latter: Tiny Separate School Trustees v The King at 369–370. Canadian courts have rejected the ‘political questions doctrine’, under which certain ostensibly constitutional questions would be deemed non-justiciable: Operation Dismantle v The Queen at 472 per Wilson J.
18 This is also one of the strategies used by the UK’s Human Rights Act 1998, s 10, to get around the limitations imposed by parliamentary sovereignty: Ministers are empowered to amend legislation to bring it into conformity with the European Convention on Human Rights if there are ‘compelling reasons’ to do so.
The power of the Canadian Government to disallow provincial statutes (Constitution Act 1867, s 90) — now in disuse, but used actively during the first decades of Confederation — was also employed in the service of individual rights, especially property rights.
19 For discussions of this process, see Hiebert (1999b: 6–9, 2002: 3–19). For arguments in favour, see Slattery (1987) and Hiebert (1998). For similar requirements, see Canadian Bill of Rights, s 3; New Zealand Bill of Rights Act 1990 (NZ), s 7 and Keith (2000: 731–735); Human Rights Act 1998 (UK), s 19.
20 Hogg and Bushell (1997: 104–105) give another example of the same phenomenon, where the Canadian Parliament revised legislation dealing with the tax deductibility of child support payments, even though the Government had won the previous Charter action.
21 The substantial re-enactment of provisions declared invalid occurred in response to R v Daviault, and was contained in An Act to amend the Criminal Code (Self-induced intoxication) (Canada). For background, see Hiebert (2002: 96-107). There has yet to be a definitive ruling on the constitutionality of that legislation largely because, even if valid, it would only apply in very special circumstances. Those decisions that have considered its validity, all by lower courts, are divided (R v Vickberg; R v Brenton; R v Dunn).
22 The US Supreme Court’s acceptance of the internment of Japanese Americans during World War II is often cited (Korematsu v United States). Institutional mechanisms within the public service can sometimes be more effective. In a seminar at the Faculty of Law, McGill University, based on Simpson (1994), Brian Simpson noted the institutional ethic of the British Home Office in the defence of the liberties of the subject, and the relative strength of this branch of the public service, in contrast to the courts, in the protection of the rights of detainees during World War II.
23 The Canadian decisions tend to use s 96 of the Constitution Act 1867 as the foundation for the protection of judicial independence in Canada, although this has been supplemented by s 11(d) of the Charter. Moreover, judicial independence has recently been recognised as a general unwritten principle (Re Remuneration of Judges). Section 96 is, on its face, merely a power of appointment, although it is the first in a set of provisions that enshrine the Act of Settlement 1701 in Canadian constitutional law. For discussion of the implications based upon it, see Hogg (1992: 184 and following). For the constitutionalisation of judicial review of administrative jurisdiction, see Crevier v A-G Quebec.
For the separation of the judicial power in Australia and the derivation of individual rights from that separation, see Zines (1997: 161–218 and 202–212), Chu Kheng Lim v Minister for Immigration; Kable v Director of Public Prosecutions.
24 A good example is s 92 of the Australian Constitution, which for many years was interpreted as prohibiting a range of government measures, especially nationalisation, that would interfere with free markets. In 1988, however, the High Court ruled that the section only prohibited tariff like trade barriers (Cole v Whitfield). The framers of s 92 may have assumed the existence of private ownership, but they did not intend to set it in constitutional stone.
25 ‘Patriation’ is the term coined in Canada to describe the creation of a domestic amending formula for the Canadian Constitution (prior to 1982, the principal constitutional document was simply a statute of the British Parliament, amendable only by that Parliament). As part of patriation the Constitution was significantly amended, notably by the introduction of the Canadian Charter of Rights and Freedoms. For discussions of the process leading to patriation, see Romanow et al (1984), Webber (1994: 99–120).
26 Kahana (2001: 255–291) describes in detail each of the uses of the clause. I am grateful to Janet Hiebert for bringing the pending appearance of the article to my attention.
27 The two circumstances in which Alberta contemplated but refrained from invoking the clause were a proposed statute to limit the damages payable for the forced sterilisation of certain classes of disabled people under previous provincial legislation, and the province’s response to the decision in Vriend. In the former case, strong public outcry forced the Government to withdraw the Bill. In the latter (which would have reversed the Supreme Court of Canada’s extension of Alberta’s Individual’s Rights Protection Act to ban discrimination on the basis of sexual orientation), the Government itself decided not to use the clause, despite demands from its supporters (see Hiebert 1999a: 30 and 34; 2002: 181–182, 190–191 and 197–198).
The Federal Government also seriously contemplated the use of the notwithstanding clause on at least two occasions, although it quickly rejected the option (Roach 2001: 272; Hiebert 2002: 87).
28 It is indicative of s 33’s shaky legitimacy that following the decisions in Daviault and O’Connor, Parliament did not use s 33 but instead relied on a renewed argument under s 1.
29 There is an argument that s 33 may have had an indirect impact on dialogue by rendering the courts more deferential to vigorous claims of justification under s 1. On this view, s 33 provides support for the legitimacy of a legislative role (and a possible option for a frustrated legislature) that may lead courts to enter more readily into dialogue. (I am indebted to Eric Ghosh for this argument.) It is of course very difficult to assess whether there has been such an effect. I suspect that it is absent or modest, given the existence of vigorous arguments between courts and legislatures over rights definition in all countries, the avenue to express such arguments under s 1 of the Canadian Charter, and s 33’s own perceived illegitimacy.
30 For suggestions that the weakness of the Canadian Bill of Rights was due to its non-constitutional character, see R v Big M Drug Mart at 333 and 341–344 (per Dickson J). However, this is probably best understood as the most obvious point of distinction between the Bill and the Charter for a Court seeking to distance itself from the previous case law. For interpretations subsequent to the Charter, see Singh and
MacBain v Lederman. Keith (2000: 730–731 and 737) makes an argument similar to mine in relation to the New Zealand Bill of Rights Act 1990.
31 For discussion of this ‘democratic deficit’, see Webber (2000a: 141–144). It is important to be clear on the nature of the public support necessary for the courts. It is not support for each and every decision. If this were necessary, there would be no advantage whatever (other than spreading the workload) in having courts separate from the legislature. Judicial independence truly would be a mirage. A studied and institutionalised inattention to popular opinion is a critical dimension in the judicial role. Without it, the courts are disabled from exercising precisely the kind of deliberation we expect of them. There does have to be popular support for them as institutions, exercising independent judgment, however. We cannot afford to neglect the public culture of rights and liberties and the institutional safeguards that sustains them.
32 For a stimulating argument along similar lines, see Macdonald (1993).
Adelaide Company of Jehovah’s Witnesses Inc v Commonwealth  HCA 12; (1943) 67 CLR 116
Australian Capital Television Pty Ltd v Commonwealth  HCA 45; (1992) 177 CLR 106
Australian Communist Party v Commonwealth  HCA 5; (1951) 83 CLR 1
Chu Kheng Lim v Minister for Immigration, Local Government and Ethnic Affairs (1992) 176 CLR 1
Cole v Whitfield  HCA 18; (1988) 165 CLR 360
Croome v Tasmania  HCA 5; (1997) 71 ALJR 430 (HCA)
Kable v Director of Public Prosecutions (NSW)  HCA 24; (1996) 138 ALR 577
Koowarta v Bjelke-Petersen (1982) 153 CLR 168
Kruger v Commonwealth  HCA 27; (1997) 146 ALR 126
Lange v Australian Broadcasting Corporation  HCA 25; (1997) 145 ALR 96
Leeth v Commonwealth  HCA 29; (1992) 174 CLR 455
McGinty v Western Australia  HCA 48; (1996) 186 CLR 140
Minister of State for Immigration and Ethnic Affairs v Teoh  HCA 20; (1995) 183 CLR 273
Nationwide News Pty Ltd v Wills  HCA 46; (1992) 177 CLR 1
Theophanous v Herald & Weekly Times Ltd  HCA 46; (1994) 182 CLR 104
Andrews v Law Society of British Columbia 1989 CanLII 2 (SCC); (1989) 56 DLR (4th) 1 (SCC)
AG Ontario v Hamilton Street Railway  AC 524 (PC)
A-G Canada v Lavell  SCR 1349
Black v Law Society of Alberta  1 SCR 591
Corbière v Canada (MINA) (1999) 173 DLR (4th) 1 (SCC)
Crevier v A-G Quebec  2 SCR 220
Cuddy Chicks v Ontario (LRB) (1991) 81 DLR (4th) 121 (SCC)
Dunmore v Ontario (Attorney General) (2001) 207 DLR (4th) 193 (SCC)
Ford v Quebec (1988) 54 DLR (4th) 577 (SCC)
Henry Birks & Sons v Montreal  SCR 799
Irwin Toy v Quebec (1989) 58 DLR (4th) 577 (SCC)
Little Sisters Book and Art v Canada (2000) 193 DLR (4th) 193 (SCC)
MacBain v Lederman  1 FC 856 (CA)
MacKeigan v Hickman  2 SCR 796
Manitoba Fisheries v The Queen  1 SCR 101
McKay v The Queen  SCR 798
McKinney v University of Guelph (1990) 76 DLR (4th) 545 (SCC)
Operation Dismantle v The Queen  1 SCR 441
OPSEU v Ontario  1 SCR 2
PSAC v Canada  1 SCR 424
R v Big M Drug Mart 1985 CANLII 69;  1 SCR 295
R v Brenton (1999) 180 CLR (4th) 314 (NWT SC)
R v Butler (1992) 89 DLR (4th) 449 (SCC)
R v Daviault  3 SCR 63
R v Drybones  SCR 282
R v Dunn (1999) CRDJ 441 (Ont SCJ)
R v Edwards Books and Art  2 SCR 713
R v Hall (2002) 217 DLR (4th) 536 (SCC)
R v Mercure  1 SCR 234
R v Mills (1999) 180 DLR (4th) 1 (SCC)
R v Morgentaler  1 SCR 30
R v Oakes  1 SCR 103
R v O’Connor  4 SCR 411
R v Vickberg (1998) 54 CRR (2d) 83 (BCSC)
Re Alberta Statutes  SCR 100
Re Canada Assistance Plan (1991) 83 DLR (4th) 297 (SCC)
Re Public Service Employee Relations Act  1 SCR 313
Re Remuneration of Judges  3 SCR 3
RJR-MacDonald Inc v Canada  3 SCR 199
RWDSU v Saskatchewan  1 SCR 460
Saumur v City of Quebec  2 SCR 299
Sauvé v Canada (Chief Electoral Officer) (2002) 218 DLR (4th) 577 (SCC)
Singh v Minister of Employment and Immigration  1 SCR 178
Switzman v Elbling  SCR 285
Tiny Separate School Trustees v The King  AC 363 (PC)
Vriend v Alberta (1998) 156 DLR (4th) 385 (SCC)
Winnipeg School Division No 1 v Craton  2 SCR 150
A-G v DeKeyser’s Royal Hotel  UKHL 1;  AC 508 (HL)
Burmah Oil v Lord Advocate  UKHL 6;  AC 75 (HL)
Korematsu v United States  USSC 43; 323 US 214 (1944)
Communist Party Dissolution Act 1950 (Cth)
Constitution of Australia, enacted by Commonwealth of Australia Constitution Act, 1900 (UK)
Human Rights (Sexual Conduct) Act 1994 (Cth)
Native Title Act 1993 (Cth)
Racial Discrimination Act 1975 (Cth)
Alberta Bill of Rights 1972 (Alberta)
An Act Respecting the Constitution Act 1982 (Quebec)
An Act to amend the Charter of the French Language 1988 (Quebec)
An Act to Amend the Charter of the French Language 1993 (Quebec)
An Act to amend the Criminal Code (Production of records in sexual offence proceedings) 1997 (Canada)
An Act to amend the Criminal Code (Self-induced intoxication) 1995 (Canada)
An Act to amend the Indian Act 1985 (Canada)
Canadian Bill of Rights 1960 (Canada)
Canadian Charter of Rights and Freedoms, Pt 1 of the Constitution Act 1982, being Sch B to the Canada Act 1982 (UK)
Charlottetown Accord, Draft Legal Text (9 October 1992): proposed ss 35.1 and following of the Constitution Act 1982
Charter of Human Rights and Freedoms 1975 (Quebec)
Constitution Act 1867 (UK)
Constitution Act 1982, being Sch B to the Canada Act 1982 (UK)
Department of Justice Act 1985 (Canada)
Individual’s Rights Protection Act 1986 (Alberta)
The Language Act 1988 (Saskatchewan)
Land Planning Act 1982 (Yukon)
Languages Act 1988 (Alberta)
Marriage Amendment Act 2000 (Alberta)
SGEU Dispute Settlement Act 1984–85–86 (Saskatchewan)
Basic Law of the Federal Republic of Germany 1949
Constitution of India 1949
Constitution of the Republic of Ireland 1937
New Zealand legislation
New Zealand Bill of Rights Act 1990
South African legislation
Constitution of the Republic of South Africa 1996
Act of Settlement 1701
Human Rights Act 1998
Constitution of the United States
International legal material
International Covenant on Civil and Political Rights 23 March 1976, 999 UNTS 171
International Covenant on Economic, Social and Cultural Rights 3 January 1976, 993 UNTS 3
Convention for the Protection of Human Rights and Fundamental Freedoms 4 November 1950, ETS No. 5
International Convention on the Elimination of All Forms of Racial Discrimination 4 January 1969, 660 UNTS 211
ILO Convention No. 87 concerning Freedom of Association and Protection of the Right to Organise 4 July 1950, 68 UNTS 17
Lovelace v Canada Communication No 24/1977 (30 July 1981) UN Doc CCPR/C/OP/1 at 83 (1984)
Toonen v Australia Communication No 488/1992 (4 April 1994) UN Doc CCPR/C/50/D/488/1992
Books and articles
Allan J ‘Turning Clark Kent into Superman: The New Zealand Bill of Rights Act 1990’ (2000) 9 Otago Law Review at 613–632
Bhardwaj R (ed) Constitution Amendment in India 6th ed Northern Book Centre, New Delhi 1995
Brennan F The Wik Debate: Its Impact on Aborigines, Pastoralists and Miners UNSW Press, Sydney 1998
Cameron J ‘Dialogue and hierarchy in Charter interpretation: A comment on R v Mills’ (2001) 38 Alberta Law Review at 1051–1068
Dicey A V Introduction to the Study of the Law of the Constitution 10th ed Macmillan, London 1959
Hiebert J ‘A hybrid approach to protect rights? An argument in favour of supplementing Canadian judicial review with Australia’s model of parliamentary scrutiny’ (1998) 26 Federal Law Review at 115–138
Hiebert J ‘Why must a Bill of Rights be a contest of political and judicial wills? The Canadian alternative’ (1999) 10 Public Law Review at 22–36
Hiebert J ‘Wrestling with rights: Judges, parliament and the making of social policy’ (1999) 5(3) Choices (Institute for Research on Public Policy) at 1–36
Hiebert J Charter Conflicts: What is Parliament’s Role? McGill-Queen’s University Press, Montreal 2002
Hogg P Constitutional Law of Canada (3rd ed) Carswell, Scarborough 1992
Hogg P and Bushell A ‘The Charter dialogue between courts and legislatures (or perhaps the Charter Of Rights isn’t such a bad thing after all)’ (1997) 35 Osgoode Hall Law Journal at 75–124
Hogg P and Thornton A ‘Reply to “Six degrees of dialogue”’ (1999) 37 Osgoode Hall Law Journal at 529–536
Kahana T ‘The notwithstanding mechanism and public discussion: Lessons from the ignored practice of section 33 of the Charter’ (2001) 44 Canadian Public Administration at 255–291
Keith K ‘”Concerning change”: The adoption and implementation of the New Zealand Bill of Rights Act 1990’ (2000) 31 Victoria University of Wellington Law Review at 721–746
Kinley D The European Convention on Human Rights: Compliance Without Incorporation Dartmouth, Aldershot 1993
Macdonald R ‘The New Zealand Bill of Rights Act: How far does it or should it stretch?’ (1993) Proceedings of the New Zealand Law Conference
Mandel M The Charter of Rights and the Legalization of Politics in Canada (revised ed) Thompson Educational Publishing, Toronto 1994
Manfredi C and Kelly J ‘Six degrees of dialogue: A response to Hogg and Bushell’ (1999) 37 Osgoode Hall Law Journal at 513–536
Petter A ‘Twenty years of Charter justification: From liberal legalism to dubious dialogue’ (2003) 52 University of New Brunswick Law Journal forthcoming
Roach K The Supreme Court on Trial: Judicial Activism or Democratic Dialogue Irwin Law, Toronto 2001
Romanow R, Whyte J and Leeson H Canada ... Notwithstanding: The Making of the Constitution 1976–1982 Carswell/Methuen, Toronto 1984
Russell P ‘Standing up for notwithstanding’ (1991) 29 Alberta Law Review at 293–309
Sadurski (ed) Constitutional Justice East and West: Democratic Legitimacy and Constitutional Courts in Post-Communist Europe in a Comparative Perspective Kluwer Law International, The Hague 2002
Simpson B In the Highest Degree Odious: Detention Without Trial in Wartime Britain Oxford University Press, Oxford 1994
Slattery B ‘A theory of the Charter’ (1987) 25 Osgoode Hall Law Journal at 701–747
Tarnopolsky W The Canadian Bill of Rights (2nd ed) McClelland and Stewart, Toronto 1975
Tribe L American Constitutional Law (3rd ed) Foundation Press, New York 2000
Tushnet M ‘Policy distortion and democratic debilitation: Comparative illumination of the countermajoritarian difficulty’ (1995) 94 Michigan Law Review at 245–301
Webber J ‘Tales of the unexpected: Intended and unintended consequences of the Canadian Charter of Rights and Freedoms’ (1993) 5 Canterbury Law Review at 207–234
Webber J Reimagining Canada: Language, Culture, Community, and the Canadian Constitution McGill-Queen’s University Press, Montreal 1994
Webber J ‘Constitutional poetry: The tension between symbolic and functional aims in constitutional reform’ (1999) 21 Sydney Law Review at 260–277
Webber J ‘Constitutional reticence’ (2000) 25 Australian Journal of Legal Philosophy at 125–155
Webber J ‘Beyond regret: Mabo’s implications for Australian constitutionalism’ in D Ivison, PPatton and W Sanders (eds) Political theory and the rights of indigenous peoples Cambridge University Press, Cambridge 2000 pp 60–88
Weiler P ‘Rights and judges in a democracy: A new Canadian version’ (1984) 18 University of Michigan Journal of Law Reform at 51–92
Winterton G ‘Can the Commonwealth Parliament enact “manner and form” legislation?’ (1980) 11 Federal Law Review at 167–202
Zines L The High Court and the Constitution (4th ed) Butterworths, Sydney 1997