Ban on religious symbols in the public service: Quebec’s Bill 21 in a global pluralist perspective

Abstract

Bill 21 is a highly contested law adopted in Quebec that bans certain civil servants from wearing religious symbols in the exercise of their duties. Rather than analyse Bill 21 on its merits, the article treats it as a test case for global legal pluralism, examining how the validity of the law from an international perspective depends on the frames one uses to analyse it. It finds that a basic tension permeates the entire debate between a universalist vision of rights and a vision of rights as anchored in particular political configurations that demand a constant process of adaptation. That tension is visible in the dualist opposition between Canadian and international law; in the role of federalism as a significant mediating factor in the implementation of constitutional and international rights; and in the kind of majoritarian check on rights that manifests itself in the Canadian Charter of Rights’ ‘notwithstanding clause’. Throughout, the article explores how these tensions might be mediated in ways that do not simply oppose international and domestic law but seek to make the most of their interaction.

I. Introduction: Frames of reference

Few Canadian laws have been more controversial from a human rights perspective than Quebec’s Bill 21, the so-called ‘Loi sur la laïcité de l’Etat’. ¹ The law bans the wearing of religious symbols for certain public servants in Quebec, specifically those agents who are repositories of an educational or coercive authority of the state. The law also proclaims Quebec’s secularism and requires the government’s neutrality in relation to religions. ² It is widely seen as targeting Muslim immigrants, although it also affects other minorities. The legislation emerged after a decade of fractious political debates on multiculturalism and secularism in the province, and their relationship to freedom of religion. ³ It has caused considerable opposition and contestation both in Quebec and throughout Canada. The law raises fundamental questions about the limits of religious freedom, the nature of majority rule, the definition of secularism, gender equality and the problem of structural discrimination.

This article takes the perspectives that have inspired both Bill 21 and the opposition to it seriously, in that both will be assumed to give voice to genuinely felt constitutional concerns, regardless (or at least independently) of their political agendas. Rather than address these questions on their own terms (e.g. as questions of freedom, equality, discrimination and so on), however, it seeks to bring attention to the various frames of reference within which those debates arise. It aims to show how the way in which one uses these frames can be highly determinative of the answer given – perhaps far more than the debate’s more obvious and immediate stakes. In particular, the debate over the Bill highlights the complex ways in which a fundamental rights issue can become susceptible to pluralist pressures at various levels of governance. Pluralism is no longer merely a variable of adaptation to local specificities of what is otherwise imagined as a straightforward universalist project: it has become the very heart of that cosmopolitan rights debate, introducing a fundamentally variable geometry in its administration.

The debate about Bill 21 arises on at least three levels. It emerges first on a micro level in person-to-person encounters, and the constant negotiation of what is permissible and acceptable among citizens. It has steered passionate interventions by friend and foe of the Bill. It also occurs more formally in the public sphere, as part of broader political conversations about the interplay of parliamentary democracy, constitutional protections and federal/provincial dynamics. In particular, it raises a number of questions about the operation of the Canadian Charter of Rights and Freedoms, which for all intents and purposes largely circumscribes the debate in Canada at present and has provided the basis of litigation. ⁴ In Quebec, it has powerfully reactivated the dialectics of identity and difference that were once uniquely fixated on issues of language but now are increasingly mediated by the idea that the province’s approach to secularism is distinct from that of the rest of Canada. ⁵ In the process, however, it has also steered pluralistic conversations about Montreal versus the rest of the province, the Anglophone minority versus the Francophone majority and particular institutions (hospitals, universities) versus the government and legislature. This domestic emphasis is probably as it should be, given the centrality of Quebec and Canadian law to the eventual resolution of the issue. Scholars are already well at work in making arguments on both sides of that debate, which has exposed as never before some of the deep ideological divides within Quebec and between Quebec and the rest of Canada. ⁶

At the same time, the debate is also a supranational, international and even global one, even though it is not always very explicitly acknowledged as such. The international human rights project is both a human rights and an internationalist project, based on the idea that such rights ought to be protected through international promulgation and cooperation. Despite Canada’s broad commitment to human rights in the international sphere, this has long involved a complex debate about the relationship of its domestic law to international law in what remains a quite hermetic dualist tradition. Yet there is no doubt that the question of how to deal with religious symbols in the public sphere – particularly various forms of the Islamic veil – has become a truly global debate. From Quebec to Uzbekistan, Turkey to the United States, or France to China, the debate has often been mediated by supranational bodies, most notably the European Court of Human Rights (ECtHR) and the Human Rights Committee (HRC). The Quebec debate has often been influenced by references to the ECtHR’s own judgments and their apparent condoning of quite strict limitations over the wearing of the hijab. ⁷ Sooner or later, a rerouting of that particular controversy through the international sphere seems inevitable, if only because the litigants who oppose Bill 21 are likely to bring the case to the Human Rights Committee if they do not succeed in the Canadian judicial system (indeed, several HRC cases have emanated from Quebec before, some of which will be discussed here).

Finally, the debate is also a transnational one between societies, rather than simply a domestic or international one. To the extent that international law influences the debate, it has often done so in ways that are much more lateral, diffuse and roundabout – and in some ways more interesting – than simply by virtue of international law being binding. ⁸ The debate has been saturated, in particular, by outside influences with more of a ‘comparative’ taint. Quebec, a province deeply marked by the civil law tradition, tends to take its ideas about the relationship between state and religion from France. By contrast, the rest of Canada (ROC) operates within a common law environment and hews much closer to an Anglo-American model of multiculturalism. One of the central, if implicit, stakes of the controversy is Quebec and Canadian societies’ unique understandings of immigration, diversity and religion, specifically (although not exclusively) as they concern Islam. The resistance to Bill 21 has also activated confessional and inter-confessional solidarities with diasporic connotations that testify to the agency of those who stand to be subjected to the law – as manifested, for example, by the leading role of the National Council of Canadian Muslims, or of the World Sikh Organization of Canada, as joint interveners in ongoing litigation.

These domestic, supranational and transnational influences make for a debate that is as complex as it is explosive. They suggest that the discussion about Bill 21, beyond its particulars in Quebec and Canada, is emblematic of much broader contemporary questions involving the proper authority of the domestic and international spheres, of constitutional organization and international obligation and of democracy, that go to the very heart of the modern rights project – itself increasingly attuned to and torn by pluralist pressures. These questions, moreover, are raised in relation to an issue – namely, how freedom of religion intersects with multiculturalism and systemic discrimination – that is among the most sensitive and divisive in contemporary human rights discourse. Finally, these tensions emerge in a country that is nominally open to such developments, but has remained relatively closed off from international human rights influences compared with other liberal democracies. In short, the topic of Bill 21 offers a unique opportunity to examine how a global constitutionalist argument that is sensitive to the exigencies of pluralism might play out in an environment where such issues have received relatively scant attention.

This article, then, examines how normative theories of global legal pluralism ⁹ might help us reconcile the many levels on which Bill 21 is being debated. Where the conventional question in Canada has been, ‘How should international human rights law be treated in Canadian law?’, the article will instead ask the question, ‘How should the idiosyncrasies of Canada be treated under international human rights law?’ But where international law is sometimes understood as a unitary standard imposed on the world’s plurality, the article will emphasize the extent to which international human rights law is itself increasingly called upon to act as an arbiter of pluralism among and within states. The article will thus argue for a vision of ‘nested pluralism’ that emphasizes the extent to which legal pluralism increasingly operates along a vertical axis of cascading spheres of autonomy, in which no single juridico-political entity can claim to stably embody its ethos. Rather, every layer of governance is called upon, fractal-like, to replicate in its midst some of the very debates that have militated for granting it a sphere of autonomy in the first place. Such a framing paradoxically emphasizes the question of pluralism as a necessary first step in concretizing the promise of rights, but also as a slippery slope that makes pinning down the meaning of rights outside highly specific contexts a tenuous exercise.

The article will examine three ways in which the same debate about the relationship between national specificities and universal commitment is being conducted in relation to the Bill 21 ban: first, through an analysis of the tension between Canada’s dualist constitution and its international commitments and the role that the ‘margin of appreciation’ might have in mediating between the two; second, by looking at how international human rights law stands to intersect with Canada’s federalism and how pluralism extends across the international/domestic divide; and third, by examining how the invocation of the ‘notwithstanding clause’ in relation to the Canadian Charter of Rights and Freedoms might play out on the international plane as an attempt to forcefully reinscribe a certain democratic pluralism. Each of these topics speaks to a distinct dimension of the pluralist resistance to an excessively homogenizing human rights global constitutionalism, namely the constitutional, the federal and the parliamentarian.

II. Dualism versus international human rights obligations: A role for a margin of appreciation?

A first way of conceptualizing Bill 21 in relation to Canada’s human rights obligations is a function of the operation of the international system’s own openness to pluralism. ¹⁰ Specifically, the margin of appreciation might be understood to help bring closer a Canadian constitutional culture of rights and Canada’s international human rights obligations. This entails establishing the margin’s global credentials and briefly imagining how it might play out in relation to Bill 21 in the Canadian context.

International human rights law in Canadian law/Canadian law in international human rights law

The debate on global legal pluralism when it comes to human rights is structured by the dialectics of the international and the domestic. How far should constitutional and international law lean towards one another to accommodate specificities? In the Canadian legal system, the profound similarity between the Canadian Charter of Rights and Freedoms and the Quebec Charter of Human Rights and Freedoms on the one hand, and similar international instruments to which Canada is party on the other (most notably, the ICCPR) provides a kind of superficial reassurance that, through domestic debates, it is Canada’s international obligations that are simultaneously being addressed. It is true that international human rights law has featured occasionally in the debate on Bill 21. Specifically, proponents of the legislation often point out that the ECtHR has found similar bans to not be in violation of the ECHR. ¹¹ This has already had echoes in Quebec-based litigation. ¹²

But the impression that international human rights law is present in the Canadian debate can be subtly misleading. The frequently cited ECHR, moreover, is not binding on Canada – despite its obvious similarities with the Charter, which have long prompted suggestions that it should be a source of inspiration. ¹³ Canada has not ratified the Interamerican Convention on Human Rights, and thus does not recognize the jurisdiction of the Interamerican Court. If anything, it is to the ICCPR (to which Canada is a party) that jurists might turn for guidance. The case law of the Human Rights Committee on the issue has hardly had any echo in the Quebec public debate, perhaps because it is not especially sympathetic to limitations on the wearing of religious symbols. That position of relative isolation from international human rights enforcement mechanisms has further reduced institutional pressures to make Canadian law fully compliant with Canada’s international obligations.

To that basic international panorama one must add the absence of direct applicability of the international human rights law in Canadian law. The ICCPR itself does not strictly require its incorporation into domestic law, only that, in fine, Canada comply with its international obligations under it. That said, Canada’s dualism has certainly further militated against such an implementation. Even though the Canadian Charter of Rights and Freedoms is sometimes referred to as implementing Canada’s international human rights obligations and the ICCPR was not absent from constitutional debates, the Charter was adopted for largely domestic Canadian political reasons rather than because the government at the time felt a compulsion to implement the ICCPR into domestic law. ¹⁴ It is hard to characterize the Charter – unlike, for example, the UK Human Rights Act – as an effort to come to terms with international human rights obligations in domestic law terms. In fact, Canada is an example of a country whose constitutional evolution was long dominated by mundane issues of governance – notably how to enable quite different peoples to coexist peacefully – and that came relatively late to a culture of rights. ¹⁵

This evidently problematizes the role of international human rights law, given its relative invisibility in domestic debates. At best, it is received indirectly by Canadian courts in ways that seek to foster interpretive harmony between the Canadian Charter and Canada’s international human rights obligations. Canadian courts may draw on international human rights sources for guidance in case of ambiguity in the Charter on the basis that Canada must be presumed to not have intended to violate its international obligations. ¹⁶ It may be that this will have some relevance as a result of constitutional challenges to Bill 21 and the possibility that the Supreme Court of Canada will draw on international sources to elucidate the status of similar bans on religious symbol. However, Canadian courts’ resistance to the direct invocation of international human rights has already been on full display in judgments on Bill 21, in a context where the law is not particularly ambiguous about what it proposes to do. ¹⁷

Canadian dualism effectively increases the odds that Canada will be found to have violated its international commitments. Notwithstanding, it is clear from an international law perspective that Canada’s constitutional specificities are not a defence to failure to comply with its international obligations. How, then, does one reconcile the obvious fact of international law’s supremacy and the equally obvious fact of the Canadian legal order’s relative imperviousness to it? Is there a role for international law itself to deploy a greater pluralism to accommodate national specificities? Indeed, just as Canadian constitutional law has trouble finding a place for Canada’s international human rights obligations, it is not clear that international human rights law always has a very clear vision of where domestic constitutional debates fit within its overall framing.

International human rights law itself remains quite dependent on the central appeal of national sovereignty and democracy. One question in this context is whether, beyond the incantation that all domestic (including constitutional) law is subject to international human rights law, there ought to be a certain deference to considered constitutional debates as evidencing sui generis but good faith efforts to deal with some of the very themes that international human rights law promotes. The more sophisticated contemporary thinking about the relationship of international human rights and constitutional law has tended to evolve in this direction. Grainne de Burca and Olivier Gerstenberg, for example, argue that the relationship between the two should be conceived, from the perspective of international law itself, ‘outside the context of a strict monism-dualism dichotomy’ and with international adjudication having ‘a persuasive function’ rather than an ‘authoritative’ one, such that the two are ‘contextually competing rule-of-law values rather than … conflicting legal sources vying against one another’. ¹⁸

Here the constitutional concern with insulating Canadian law from Canada’s international obligations connects, as it happens, to a corresponding international interest in managing pluralism within the international legal order’s overarching value system. Even as it is more customary to speak of domestic legal systems as being ‘dualist’, the ECtHR itself has been described as embedded in a ‘dualist’ outlook that shuns any notion that it is a ‘fourth degree of jurisdiction’ whose decisions have direct effect. This leaves space for domestic organs to adapt domestic law to international exigencies, and has been described as part of a shift from a univocal focus on ‘implementation’ to a growing interest in ‘translation’ of international obligations into domestic law. ¹⁹ In other words, the Canadian legal system’s resistance to the direct applicability of international human rights law ought to be seen in light of (and perhaps partly mollified by) an understanding of how international law itself occasionally exhibits significant deference toward constitutional equilibria.

The global potential of the margin of appreciation

Within the ECtHR context, the ‘margin of appreciation’ is the emblematic tool through which this sort of pluralism has traditionally been secured. ²⁰ The margin of appreciation is based on deference to sovereignty, democratic processes, constitutional tradition and local judges. It has tended to become increasingly entrenched over time. It allows for significant national variation where an issue of rights is not settled, on the basis that national authorities are better placed to make the necessary assessments. It is particularly relevant to the evaluation of the permissibility of limitations to rights, notably in areas that are relatively contentious and involve a fragmented international approach. It has been particularly used, as it happens, in the process of elucidating whether various bans on religious symbols or garments are compatible with human rights, both deferring to legal and political traditions on the one hand and seeking to maintain a degree of ‘European control’ over departures from the ‘European consensus’ on the other. ²¹

Could the margin of appreciation be a way of conceptualizing the relationship of Canada’s international human rights obligations to Canadian constitutional law? Of course, the ECHR does not apply to Canada. Furthermore, it is not entirely clear whether a margin of appreciation can even be said to operate at all on a universal level. The margin remains by and large a European mechanism. None of the other regional courts has adopted it explicitly. Perhaps more importantly, the HRC has explicitly said it does not rely on it.

Nevertheless, there is something intuitively implausible about the HRC (operating as it does against the background of the most universal and fragmented system) claiming to not endorse the margin. In the European context, subsidiarity has increasingly been affirmed not simply as a logistical concession to the fact that the ECtHR’s docket is saturated, or even as a more or less idiosyncratic European idea, but instead as a more fundamental principle of human rights adjudication reflecting the proper apportionment of law-determining competencies between the international and the domestic. ²² It is worth pointing out that there has been increasing interest in the margin of appreciation in human rights contexts other than the European. ²³ The failure of the HRC to adopt margin-of-appreciation reasoning has been criticized as unrealistic and unhelpful. ²⁴ The same is true of the United Nations Committee on the Elimination of Racial Discrimination. ²⁵ Indeed, it has been argued that, despite its stated opposition, the HRC effectively implements a series of substitutes to the margin that operate along the same lines. ²⁶

Moreover, there has been an uptick in the reception of margin language in other regional contexts. Although the Inter-American Court of Human Rights has refrained from endorsing the margin of appreciation for reasons that are quite specific to its history and its caseload (particularly the prevalence of grave and systematic human rights violations), several scholars have strongly suggested that it should do so, or else risk rendering itself illegitimate. ²⁷ The more the Court moves away from a caseload of massacres and disappearances, the more it has tended to de facto rely on a certain type of margin of appreciation ²⁸ and to be criticized when it fails to do so. ²⁹ Indeed, some commentators argue that it is already showing signs of considerable deference to national courts. ³⁰ Although Canada is not a party to the Inter-American Convention, it is a member of the Organization of American States and as such is susceptible to the Inter-American Commission on Human Rights’ supervisory jurisdiction in relation to the Inter-American Declaration of Human Rights, which has occasionally, as it happens, hinted at a ‘margin of discretion’ in relation to Canada itself. ³¹ Finally, although not in relation to Canada’s international human rights obligations, the ‘margin of appreciation’ is not unknown to the Canadian judiciary and has appeared on the margins of the Supreme Court’s case law, notably as a way to evaluate rights cases with a foreign element. ³²

Bill 21 through margin reasoning

The consequence of Bill 21 being analysed in terms of the margin of appreciation, if one were willing to acknowledge that possibility, may well ultimately be ambiguous except insofar as it would unmistakably foreground the very negotiation of pluralism internationally. The margin is both a rather indeterminate standard and one that has been contested as leaning potentially too far in the direction of states’ preferences at the expense of a counter-majoritarian culture of rights. ³³

A number of very valuable reports were produced at the Crown’s request in the runup to legal challenges of Bill 21 that readily emphasize, on the basis of a detailed study of the case law of the ECtHR, the great diversity of views existing on the matter and the consequent broad scope of margin of appreciation granted to states parties. ³⁴ This certainly deprovincializes Quebec’s approach and makes it appear as one among many possible approaches to a complex and contested question. Margin analysis could go some way towards alleviating traditional Canadian concerns with parliamentary sovereignty and the unimplemented character of international human rights treaty obligations because of the way it already incorporates its own natural element of deference to such sovereignty. It could thus frame Canada’s international human rights obligations in light of Canada’s own constitutional trajectory in a fundamentally pluralistic way. The ambiguity of the margin, at any rate, would presumably be neither more nor less pronounced in the case of Canada and religious symbols in the public service than it has been in a range of countries and on a range of issues.

Margin-of-appreciation reasoning could significantly influence – and complicate – the conceptual assessment of Bill 21’s conformity with international human rights standards along pluralistic lines in at least three ways. First, margin-of-appreciation reasoning grounds conformity with international human rights instruments in a dialogical exercise that takes into account the historical, political and constitutional specificities of the state involved. The question would not be whether any ban on religious symbols in the civil service violated international human rights standards in the abstract, but whether such a ban would be in violation in Canada. It is thus interesting in itself that, rather than constituting an entirely separate standard of human rights achievement, international human rights law would partially point back to Canada’s own specificities in dealing with the issue, albeit within a horizon of global human rights compliance. In that respect, there are certainly characteristics of Canada’s human rights culture that would count against conformity. The European Court leaned towards the highly specific political secular cultures of Turkey and France in finding that their bans on the hijab were not in violation of the European Convention, partly on account of these cultures. But Canada does not a priori have the same model of strict state/religion separation, so its invocation of such a model might seem opportunistic and a sort of majoritarian about-turn against a minority. ³⁵ In addition, Canada operates against the background of a strong commitment to multiculturalism, which makes the ban all the more evidently a departure from its own tradition. ³⁶

Second, margin of appreciation reasoning introduces a distinct horizontal and ‘communal’ element in the evaluation of departures from international human rights norms. Framing the debate as one involving the margin of appreciation on a global level would require one to assess practices of religious symbol bans on a comparative basis. It is beyond the scope of this article to carry out such a detailed contextualization. All other things being equal, however, one may surmise that if the margin has pushed the ECtHR to defer significantly to states in their assessments of limitations on freedom of religion (because of irreducible conceptions of state-religion relations within the European context), then a fortiori the HRC, proceeding from an even more fragmented universal reality, ought to do the same. One way of looking at the exercise of interpreting the margin of appreciation is that it merely requires states to ‘justify those local practices that deviate from a shared, publicly evolving, cross-community set of standards’. ³⁷ In that context, Canada might find that it is hardly alone globally in trying to regulate religious symbols, but that its focus on such symbols in the civil service specifically is quite peculiar.

One legitimate question might be what the territorial scope for the purposes of assessing this sort of ‘global margin of appreciation’ should be: Canada’s North American vicinity, its hemispheric geography or the entire world? If the geographic scope of the comparison is North America or even the Americas, for example, then Canada would appear to be an outlier; if the experience of all “specially interested” states with religious minorities were the standard, then Canada might be able to point to a greater diversity of approaches and therefore international tolerance for relatively more restrictive attitudes. The choice of geographic framing is therefore key to providing a context-sensitive sense of conformity. The kind of communal evaluation of the margin of appreciation implicit in the European system presumes a pre-existing sense of a quasi-civilizational frame of reference. In that respect, it is interesting that many Canadian commentators on Bill 21 typically turn to the United States or Europe for arguments, whereas Canada’s actual international commitments, if anything, orient the debate in a much broader universal direction.

Third, margin-of-appreciation reasoning strongly relativizes the opposition between an international and a domestic law of human rights by suggesting a common commitment to human rights and the rule of law. It does so in particular through an attention to legislative democracy in elucidating the conditions for a pluralism-sensitive rights culture, which was visible at every turn in the Bill 21 debate. ³⁸ This is sophisticated because international human rights law is indeed mindful of democratic equilibria. Theoretically, it can draw on a register, evidently not unknown in Canada ³⁹ but also part of international human rights legacies, that sees democratic self-determination as prior to rights, or at least as strongly foundational for rights culture. In the European human rights system, the standard for limiting rights is whether such limitation is ‘necessary in a democratic society’: a clear, if indirect, reference to the centrality, all other things being equal, of democracy for rights.

As part of a ‘procedural turn’ in the margin of appreciation, ⁴⁰ the ECtHR has increasingly examined the care with which states have adopted limitations to human rights, in particular the quality and persuasiveness of their justification for curtailing them. This introduces a level of scrutiny for supranational human rights bodies to investigate the depth of seriousness with which states have taken their obligations. In cases where domestic standards themselves have been violated, the onus shifts to the state to prove that it has not simultaneously violated its international obligations. ⁴¹ A strenuous effort to comply with the Canadian Charter of Rights and Freedoms, both in parliament and in the courts, might not be enough to satisfy Canada’s international human rights obligations, but it would probably go some way towards a positive evaluation.

From an international human rights perspective, this would entail an analysis of at least the extent to which our previous analysis of Charter limitations reflected similar international concerns about limiting rights only for the most onerous of reasons in a democratic society. One factor that will count towards a finding of compliance with international human rights obligations is that the state actually took into account its international human rights obligations, or at least their domestic reflection. On that score, the relative difficulty of doing so in Canada (given the unimplemented nature of ICCPR obligations) means that legislative processes and judicial challenges have been largely bereft of references to international human rights law, specifically of the ICCPR. This would (rightly) relativize the notion that much effort was put specifically into complying with the Covenant. After all, how can one be seen as trying to comply with something that one does not even mention?

Even this failure, however, would not necessarily be conclusive if (indirectly and non-explicitly) the Canadian legal system could be seen to have adequately struggled with the underlying issues on its own terms, in ways that reflected international imperatives. If not attention to the ICCPR specifically, then there has certainly been considerable attention given to similarly formulated Charter guarantees. ⁴² Notably, several factors are often mentioned as evidencing a clear concern with the rights ramifications of Bill 21, such as the fact that it only applies to certain civil servants (those with educational or coercive authority) and that it makes an exemption for those recruited under the previous legal regime (essentially ‘grandfathering’ their right to wear religious symbols). By the same token, these concessions do reveal the inherent sensitivity of such ramifications and may simultaneously weaken the argument that Bill 21 is foolproof from a rights perspective (why only certain civil servants and not others?).

III. Human rights Russian dolls: Federal stratification in global perspective

A second way of imaging global legal pluralism in relation to Bill 21 is as moving beyond conventional state-focused margin of appreciation reasoning altogether to problematize the very sites of human rights implementation. So far, this article has proceeded as if Canada were the only (or at least by far the most relevant) entity when it comes to assessing Bill 21 from the standpoint of international human rights law. Moreover, it has imagined that Canada’s internal characteristics – notably its federalism – are, for the purposes of international law, largely irrelevant. In reality, however, evaluating Bill 21 in global pluralist perspective involves a significant complicating layer, namely the fact that the legislation, to all intents and purposes, was adopted by Quebec and not Canada, and as part of a complex federal co-exercise of Canadian sovereignty.

From the perspective of international law, domestic idiosyncrasies such as Canada’s federal structure are traditionally irrelevant when it comes to assessing violations of international obligations. This, however, creates a genuine challenge because of the mismatch between what international law considers as the responsible actor (the state) and how Canadian constitutional law deals with the issue (provinces). From a global pluralist perspective, the risk is that one will not do justice to the internal complexities of Canadian federalism, including as they embody complex and rights-sensitive pluralist compromises, on account of a rigid focus on the state that has very little do per se with the spirit of maximizing rights. Pluralism, by contrast, can help complexify the international legal narrative by seeing human rights as simultaneously implemented on many different levels, each of which generates demands for translation and adaptation.

The state-centric view of international human rights implementation

By and large, international human rights law is predictably heavily focused on the state as the primary site of human rights implementation. This is undoubtedly in part a bias that results from the international law of state responsibility. A violation of international human rights law by a federated entity is certain to engage the responsibility of the (federal) state internationally, even if the latter does not particularly acquiesce to or otherwise endorse the policies in question. ⁴³ Federated entities are not, ultimately, the de jure subject of international human rights obligations even when they are de facto responsible for the impugned conduct. If Bill 21 is to be challenged internationally, it will be challenged as against Canada. ⁴⁴ The centrality of the state in international human rights law is also a consequence of the fact that it is Canada as a state that is a party to the ICCPR under international law, and not its provinces.

Indeed, the relative invisibility of federated entities in international law (if not de facto, at least de jure) is connected to a deeper substantive theme in international law, namely a traditionally proclaimed agnosticism about states’ internal organization beyond conforming to basic criteria of statehood. It has been argued that whatever the importance of the principle of subsidiarity, this principle at present plays mostly between international law and the state rather than between the state and hypothetical federated or sub-state entities:

international human rights law does not currently provide strong support for a requirement of territorial subsidiarity within the state, as a claim for federalism or local government. Human rights law ordinarily takes the political subunits of the state, and the allocation of powers among them, as a given. Local government may be desirable as an opportunity for citizens to exercise the right to political participation, but there are many ways to structure political participation. The principal human rights treaties do not give local governments autonomy rights against regions or states, or require that the larger units refrain from regulating matters that the local governments could address. ⁴⁵

It is true that the Human Rights Committee has occasionally evaluated specifically provincial policies in Canada ⁴⁶ but it has evaluated them – somewhat awkwardly, as we will see – qua Canadian policies. Note, moreover, that in the case of Bill 21 the idea that it is Canada as represented by the federal government that should be accountable and not Quebec is given added plausibility given the current reluctance of the federal government to challenge Bill 21 before the courts. ⁴⁷ Internationally, this reluctance could be interpreted as manifesting a tendency by the Canadian state to acquiesce to and solidarize itself with a provincial policy (on political grounds that are beyond the scope of this article), whatever domestic misgivings it may have in terms of principles.

Yet the fact that it is Canada and not Quebec that would hypothetically have to answer for rights violations does create a potentially problematic split between the entity that has undertaken the impugned action domestically and the entity called upon to answer for it internationally. It means that international bodies are not, under normal circumstances, in direct contact with the actual political constituencies responsible for the violations. Conversations about human rights are more or less felicitously mediated by a federal government that may have had very little hand domestically in their occurrence. Moreover, it inclines international human rights organs towards a one-size-fits-all notion of implementation that treats states generically and fails to take into account their internal specificities – notably, in this case, federalism. International bodies may then, somewhat self-defeatingly, require of the state (here, the Canadian federal government) changes to legislation that are not in its competence to make domestically. And the focus on the sovereign tilts the evaluation of the margin of appreciation towards the state in the international system rather than also extending it to federated entities within broader federal arrangements.

Even from an international human rights perspective, therefore, it cannot be entirely irrelevant that Bill 21 was passed by the province of Quebec within the exercise of its competencies, and not Canada itself. Simply because international human rights law does not mandate federalism (indeed, is largely indifferent to it) does not mean that, faced with federal arrangements, it should not intelligently assess their implications for rights implementation. To ignore this dimension on account of a dogmatic and singular concept of international responsibility would be to fail to comprehend Canada’s constitutional specificities, laden with rights and pluralist considerations as they are.

Indeed, the case for a relative visibility of federal arrangements is reinforced by the fact that, outside the question of responsibility, the preference in international human rights law has always been for all organs of the state to participate in the process of implementation. ⁴⁸ This means that federated entities surely ought to implement the international human rights obligations of the state of which they are part, even as the federal state has an obligation to encourage them to do so. ⁴⁹ As the experience of even otherwise monistic and international law-receptive federal states suggests, this cannot be taken for granted. ⁵⁰ In some cases, federated entities have been at the heart of resistance to international human rights law against an otherwise relatively sympathetic federal state. ⁵¹ However, in other cases federated entities may in fact be relatively more permeable to international human rights obligations and act as convenient relays or even objective allies for them “beyond the state”. ⁵² The fact that Quebec clearly claims to be implementing international human rights law in a range of areas, including Bill 21, only reinforces the sense that it should be treated as a relevant actor internationally in that respect, if not on account of international law at least out of concern for human rights. This sheds light on how Canadian federalism relates to international law.

Federal states in international human rights law: The case of Canada

Canada’s plurinational federalism formula is different from otherwise unitary states’ efforts at decentralization. Although it is the federal state that binds Canada to international agreements, Canadian provinces are not like local authorities exercising a delegated responsibility. Rather, they partake as equals in the exercise of Canadian sovereignty within their sphere of competence, with obvious implications for treaty implementation and perhaps for international scrutiny as well. This means that constitutionally, if not internationally, the federal government’s ability to implement its human rights obligations is strongly mediated by federal relations.

At the same time, whatever the constitutional specificities of Canada, under international law it is Canada that is answerable for its commitment to human rights treaties. In practice, this means that it is traditionally the federal government that will defend its compliance with its treaty commitments internationally (such as the ICCPR in Geneva, for example). Even though Quebec may have a certain capacity to conclude its own international arrangements, ⁵³ it is clearly not a party to the main international human rights treaties. International human rights bodies, for their part, remain unaccustomed to treating federal states differently preferring to leave constitutional idiosyncrasies behind a relative ‘veil of ignorance’.

Yet at least three arguments militate in favour of taking into account states’ federal specificity when assessing compliance with human rights from a pluralist perspective. First, pragmatically and in terms of policy, there is something highly artificial about treating states interchangeably when their domestic constitution is a crucial part of their ability to implement rights. If the goal of international human rights bodies is to promote human rights rather than cater to necessary but dated dogmas about sovereignty, then each state should be taken, to a degree, ‘as it is’. For the HRC to treat Canada effectively as if it were a unitary state when its ability to implement the ICCPR is deeply mediated by federalism is to risk engaging in a highly fictional exercise. It is, in fact, in the very spirit of the margin of appreciation that states’ broad constitutional specificities should be taken into account, including in how they affect prospects for rights.

Second, perhaps federal governance can itself be portrayed as related and even necessary to the imperative of protecting human rights, at least in those states where it has historically been understood in that way. ⁵⁴ International human rights law has exhibited, over time, a certain openness to decentralized federalism understood as a form of autonomy – for example, in relation to the protection of minorities. ⁵⁵ Instead of simply an inconvenient variable that complicates international law’s simplistic image of interchangeable sovereigns, it could be seen as an inherent vehicle for the delivery of rights. Canadian federalism is certainly no stranger to a proto-rights logic, grounded as it was for example in the need to guarantee a distinct people – that is, Quebecers – their own legal system and eventually language rights, but also more generally in the adoption of the Canadian Charter of Rights. One finds echoes of this more rights-principled justification of federalism in the distinct but related debate on subsidiarity. According to Paolo Carozza:

The most comprehensive formulation of subsidiarity is one that does not merely reduce it to localism or devolution, nor to a utilitarian principle of efficiency in the allocation of powers, but rather that regards subsidiarity as a principle of justice that requires larger communities to protect the legitimate autonomy of smaller communities, to provide them with the assistance (subsidium) needed to fulfill their ends, and to coordinate and regulate their activities within the common good of the larger community, of which they are a part and which is also necessary to the flourishing of their individual members. ⁵⁶

There is, it has been suggested, ‘a very deep consonance between the underlying premises regarding persons and communities in the idea of human rights and in the idea of subsidiarity’, ⁵⁷ notably the fact that both, from their respective orientations, ‘challenge radically the highly positivistic and absolutist understanding of state sovereignty that have dominated public international law’. ⁵⁸ Thomas Kleinlein argues that ‘the relationship between federal structure and rights reflects the interdependence of individual and democratic autonomies’. ⁵⁹ In effect, minority protection, cultural rights and self-determination – as concepts that are themselves well entrenched in international human rights law – may already militate for a recognition of more decentralized forms of governance ⁶⁰ and even law. ⁶¹

Third, there is a certain evident analogy between the margin of appreciation in the international realm and the theory and practice of federalism, as instances of pluralism. ⁶² In calling for greater accommodation between global and domestic legal orders, but also between plural preferences within domestic legal orders, Michel Rosenfeld argues in favor of relying on transnational tools such as the margin of appreciation domestically. ⁶³ In Canada specifically, it is worth noting that the constitutional argument for ‘Charter federalism’ seeks to balance the aspiration to live together under a common standard of justice and the inevitable recognition of value pluralism incarnated, however imperfectly, by provincial divisions. ⁶⁴ Jeremy Clarke once argued that the margin of appreciation can be a way of ‘developing a consistent model of rights in a federal context’ in ways that allow ‘provincial governments to build distinctive communities’ while abiding by Canada-wide rights standards. ⁶⁵ Among scholars of federalism, moreover, the margin’s rights-promoting, or at least rights-compatible, logic has been stressed. ⁶⁶

Although thinking about the margin of appreciation mostly in a domestic context, these authors make it possible to imagine a continuum between the pluralism of international human rights obligations and the pluralism of domestic federal governance. Canadian courts have, in fact, occasionally hinted at the relevance of the margin as when a judge noted that ‘a course of action may be demonstrably justified in a free and democratic society without being adopted by every political unit within that society … The Charter is not a tool to make Canada a monolith.’ ⁶⁷ The idea that federalism is an inherently pluralistic arrangement geared towards taking into account different approaches, in particular to state–religion relations, featured strongly in the Bill 21 debate. ⁶⁸ Taking into account the intrinsic value of federalism for rights would entail a renewed international appreciation of how Canada’s constitutional model both enables and constrains Bill 21.

Indeed, the profound similarity of the margin of appreciation as an international and as a domestic device has been underscored in debates about Bill 21 where it has been noted that ‘the national margin of appreciation is also characteristic of a system composed of other systems. In that sense, the analogy with plurinational or strongly decentralized federations is particularly relevant.’ ⁶⁹ This is particularly the case when it comes to the state’s rapport with religion, where ‘a similar diversity of formulas can even exist within federal regimes, the least centralized of which concede a significant margin of maneuver to their federated states when it comes to the civil organization of religion conceded’. ⁷⁰ The examples of Switzerland and Germany are often cited as ones where the various cantons and länders not only have considerable autonomy in terms of how they organize state–religion relations, but also happen to differ significantly from each other based on their histories (marked, for example, by Catholicism or Reformation). ⁷¹ Indeed, this is evocative of Canada’s own reality, ⁷² where the Supreme Court has in the past deferred to Quebec’s distinctiveness, including as it manifests itself in relation to religion. ⁷³

The argument in defence of Bill 21 would still need to be made by Canada on the international plane, but Canada’s place under international human rights law would be seen through its federal character. Specifically, Canada would be speaking for itself – both through its federal government (which ratified the relevant treaty) and one of its provinces (which is in charge in part of implementing it). It might thus present a sort of ‘divided united’ front: divided on the scope of freedom of religion, for example, but united in its understanding that the issue must at any rate be handled while respecting Canadian federal pluralism and taking its specific culture of rights seriously.

Note that this presumes the federal government (bound as it is by the Canadian Charter of Rights and Freedoms, to say nothing of a repository of its own legislative and executive outlook on freedom of religion) would be willing to defend Bill 21 internationally on its own terms. There is an interesting paradox here: the federal government may be called to defend in Geneva policies that it disapproves of in Ottawa. The Canadian government might choose to deliberately lose an international challenge or only defend itself in a very pro forma way. Tactically, this might reinforce the federal government’s hand in remonstrating Quebec for the adoption of Bill 21 while allowing it to pretend that its ‘hands were bound’ by international law, although this would presumably lead to a domestic crisis. Conversely and more plausibly, the federal government could decide that, despite its domestic opposition to Bill 21 in principle, it ought to defend the legislation in earnest as a product of federalism that it is at least required to back on the international stage as a result of federal comity.

It is not absolutely clear how the focus on Canada as a federal state would influence the computation of the margin of appreciation internationally. On the one hand, Canada could present itself before international bodies as split between a majority (the federal government and all other provinces) and a minority (Quebec) on the issue of state–religion relations and the wearing of religious symbols in the public service. This might encourage an understanding that it is appropriate in terms of Canada’s international human rights obligations for Bill 21 to be considered valid in Quebec, even as it would almost certainly be a violation of the same obligations for the federal government or other provinces to adopt a similar law (given their inability to justify the margin of appreciation against the background of the right kind of secular tradition).

On the other hand, considerations of provincial specificity might seem much less urgent when filtered through the demands of governing the whole of the country federally and, in particular, ensuring that the Charter is not violated. Even in areas that clearly fall within provincial prerogatives, it might still be argued that the appropriate unit of reference for the margin of appreciation is Canadian federalism. The argument for a specific, ingrained tradition of rigid secularism would then be much weaker in a federal context characterized by strong respect for freedom of religion and multicultural rights. The almost absolute lack of echo of the Quebec proposals to ban religious symbols outside the province (nine provinces representing roughly three-quarters of the country’s population against one), despite Canada facing broadly the same reality in terms of religious diversity, would make it relatively more difficult for Canada as such, even taking federalism into account, to argue that the banning of the veil was a necessary and proportionate measure.

Federated entities in international human rights law: The case of Quebec

Another more radical and pluralistic avenue, however, would be to imagine international human rights bodies as engaging federated entities directly, effectively minimizing the importance of the central state for the purposes of implementation. The focus on the sovereign in international human rights law may ultimately say more about international human rights law’s embeddedness in international law than about a commitment to human rights. Why, from a human rights point of view, should the state be the appropriate unit of reference to compute rights compliance? From an international law point of view, of course, the answer is simple: because the state is the party to the relevant treaty and perhaps out of a deferral to the axiomatic role of the state as the central subject of that body of law but that is not necessarily conclusive of human rights’ own debates. Human rights are certainly mediated by international law, but they are not necessarily reducible to it.

In addition to the just examined possibility that federal states should at least be taken ‘as they are’, it is not inconceivable that human rights implementation could be understood in a much more decentralized and pluralistic way, based on a more fully articulated concept of subsidiarity that does not stop where the state begins. ⁷⁴ Julie Fraser and David Valeska point out that groups that feel under-represented within the state (as Quebecers arguably do in Canada) may have reservations about how they are portrayed (even benignly) by state entities before international institutions and seek out those institutions directly to have their specificity (further) recognized. ⁷⁵ There is a risk that, in purporting to speak for Quebec internationally, Canada would either only represent Canada and its view of Quebec or reify a ‘Quebec identity’ in matters of religion and secularism in ways that might be problematic for minorities in Quebec.

Moreover, the specificities of Canadian federalism stand to further problematize the debate. Provinces’ prerogatives do not only complicate the position of the federal state itself in its representations to international bodies; they also create dynamics uniquely favourable to at least a certain international visibility of federated entities. Instead of being several degrees removed from the domestic–international interface, provinces – precisely because they exercise Canadian sovereignty – arguably operate in its close vicinity. Indeed, the willingness of provinces (especially, as it happens, Quebec) to engage in their own international representation when it comes to matters that fall within their prerogatives and the relative amenability of the Canadian government to such arrangements ⁷⁶ are factors to be considered. The representation of Quebec in Canadian delegations defending the country’s periodic reports before treaty bodies, and the fact that the latter have seen nothing untoward internationally about this, suggest a certain blurring of the notion of the state as necessarily a unitary actor internationally.

For one thing, this renewed sense of federated entities as direct participants in international human rights implementation ⁷⁷ makes more plausible the notion that they can be sites in their own right for the evaluation of the margin of appreciation. Deferral to the norm-producing authority of federated entities or sub-state actors may be justified a fortiori when these have recognized domestic political competencies that happen to coincide with cultural distinctiveness and have been hammered out as part of constitutional ‘grand bargains’ that themselves clearly reflect rights preoccupations. Cormac Mac Amhlaigh, for example, has made the case for ‘autonomy in the “domestication” of international human rights norms in national minority institutional structures’, following a Human Rights Act-based ruling by the UK Supreme Court involving Scottish criminal law. ⁷⁸ As the UK Joint Committee on Human Rights has noted, not only is there no incompatibility between a ‘British’ Bill of Rights and devolution, but there may even be ‘a positive virtue in the broadly defined rights in the international standards being fleshed out into more concrete norms and standards at the regional, national and sub-national level’. ⁷⁹

Note that if Bill 21 were assessed based on Quebec’s own cultural specificities as a quasi-subject of international human rights law, it would stand a higher chance – perhaps unsurprisingly – of being seen as consonant with the margin: Quebec would at least provide its own standard by which to be judged internationally. ⁸⁰ Although it would on one level appear quite isolated in relation to the ROC, it could point to the fact that a certain aspiration to rigid secularism is more authentic there than anywhere else in Canada. Internally, Canada is a striking example of the kind of scenario that the margin of appreciation was imagined for, namely one where commitment to rights is evident but deep differences over the very meaning of the liberal project prevent the emergence of an identifiable single-best position.

Note, however, that therein lies a danger for supporters of Bill 21 who are intent on maximizing the pluralist angle, namely the risk that the conceptual floor of that pluralist argument will collapse under their feet, revealing ever more aspirations to pluralism. Indeed, there remains the question of why, if this fundamentally decentralizing logic holds, one should even stop, from an international human rights perspective, at existing allocations of political power within Canada? If Quebec is recognized for its margin of appreciation, then why not certain constituencies within Quebec that may also have their own competencies and specificities? In that respect, nested pluralism inevitably opens up interrogations about the appropriate frame of reference, even within federal arrangements. This is all the more so given that the ban on the veil is fundamentally contested within Quebec and that, for all the willingness (such as it may be) to defer to the autonomy of a province, it necessarily creates ‘minorities within the minority’.

As it happens, the contestation of Bill 21 in Quebec, notably by the Anglophone minority, has already led, on constitutional grounds, to a recognition that English school boards should not be bound by the prohibition on religious symbols. The English Montreal School Board (EMSB) had argued that the linguistic rights granted to it under Section 23 of the Charter entitled it to adopt its own policy on the issue. Generously interpreting Section 23, Justice Marc-André Blanchard emphasized the close connection between language and culture, including as it relates to confessional specificities. In that respect, not only have Francophones historically been associated with Catholicism and Anglophones with Protestantism, but the Anglophone community’s attitude to the celebration of religious diversity was presented as more pronounced. ⁸¹ Whether this in turn reifies Anglophone culture in Quebec or whether it expands language rights too broadly, it remains that – constitutionally at least – an argument was found for a kind of ‘pluralism within pluralism’ (leading one commentator to describe Quebec henceforth as a form of constitutional ‘Swiss cheese’). ⁸² Pluralism was used ‘defensively’ to at least circumscribe a space of resistance to Bill 21 within Quebec, a situation that will not satisfy those who would have wanted the law lato sensu to be invalidated but maximizes the possibilities of pluralism in contextualizing rights within particular communities.

It would be interesting to know how international human rights bodies, if they were ever confronted with the validity of such an outcome, would treat it. The HRC, for example, has in the past been reluctant to even consider that Anglophone Quebecers are technically a minority in Canada. When asked in the early 1990s to evaluate whether article 27 on the protection of the cultural rights of minorities had been violated as a result of Quebec’s language laws, the Committee found that:

As to article 27, the Committee observes that this provision refers to minorities in States; this refers, as do all references to the ‘State’ or to ‘States’ in the provisions of the Covenant, to ratifying States. Further, article 50 of the Covenant provides that its provisions extend to all parts of Federal States without any limitations or exceptions. Accordingly, the minorities referred to in article 27 are minorities within such a State, and not minorities within any province. ⁸³

Anglo-Quebecers were therefore not a minority for the purposes of the ICCPR, whereas Francophones would be (even though they are a majority in Quebec). It is interesting how quickly, given the opportunity, the HRC reasserted the whole of Canada as the framing of choice, despite compelling arguments that Anglo-Quebecers are effectively a minority in Quebec even though they are a majority in Canada. In another case, this time on state funding of religious schools, the HRC found that the preferential treatment of Roman Catholic schools in Ontario was not justifiable under the ICCPR, even as a historical compromise was embedded in the Canadian constitution that had been key to the birth of the confederation in 1867. It thereby showed itself to be relatively unresponsive to even historical attempts to organize Canadian pluralism when they clashed with equal rights. ⁸⁴ By the same token, it would be difficult for the HRC to second-guess a delicate solution, based on an interpretation of the Canadian Charter of Rights, that seems to create a complex compromise between a particular minority, a province and Canada. Note also that Justice Marc-André Blanchard’s reasoning, in applying Section 23 of the Canadian Charter of Rights in concreto by taking into account the historical specificities of the Anglophone minority (rather than simply promoting a singular human rights standard), has shown a path that is strikingly evocative of the margin of appreciation.

At any rate, one can imagine that, internationally, one could witness the emergence of a three-tiered system for the margin: one involving the state as such; one focused on federated entities; and a final level of deference to notable sub-state entities (municipalities, school boards, etc.) that can make a distinct claim to their own human rights path. From a human rights perspective, there is certainly no strong reason why pluralism should stop at the state and why the fundamentally decentralizing logic of the margin should not reverberate throughout the organization of political power. This would, however, frame the margin of appreciation less as a tool to merely evaluate the rights conformity of state behavior and more as a modality of evaluating the very rights conduciveness of a variety of pluralistic constellations of power.

IV. Notwithstanding the ‘notwithstanding clause’: On the limits of overriding rights in the name of democracy

In this third section, I turn to a final level of decentralization of human rights interpretation and implementation involving pluralism, one centred on parliamentary sovereignty and the potential role of majoritarian fiat in suspending rights in the name of a claim to difference. What happens when a state or a federated entity is intent on claiming the benefit of ‘pluralist’ arrangements for itself while denying such pluralism to minorities within its midst? To what extent is democracy, in particular as expressed through parliaments, a valid mode of circumscribing the appropriate locus and scope of pluralism?

A rather unique feature of the Canadian Charter of Rights and Freedoms is the ability of parliamentary majorities to ‘override’ certain Charter protections. According to Section 33, the so-called ‘notwithstanding clause’, ‘Parliament or the legislature of a province may expressly declare in an Act of Parliament or of the legislature, as the case may be, that the Act or a provision thereof shall operate notwithstanding a provision included in section 2 or sections 7 to 15 of this Charter.’ Section 33, it should be noted, was introduced at the request of provinces, and so very much operates at the intersection of rights protections and federal governance. A similar provision is included in the Quebec Charter of Human Rights and Freedoms.

Quebec decided to pre-emptively invoke Section 33 in the case of Bill 21. Domestically, this proved prescient as it is essentially on the grounds that its ability to do so is constitutionally uncontested that Bill 21 has so far been found to pass constitutional muster. ⁸⁵ Indeed, more than any other province in Canada, Quebec has made considerable use of and developed its own sensibility to the clause, notably in the wake of the Charte de la langue française, which imposes mandatory use of French in a variety of provincial contexts. The clause is seen as a way of setting aside Charter protections (rather than merely interpreting them) in cases where they potentially clash with broad provincial orientations. ⁸⁶ The question of whether such invocation is legal and whether it pre-empts challenges has already given rise to a lively debate in Canada and Quebec in the context of Bill 21. ⁸⁷

But what of the notwithstanding clause’s status under international rights law, an issue that has been far less frequently discussed, assuming that Bill 21 cannot be justified via some sort of pluralist deference to the margin of appreciation? Clearly international human rights bodies are not bound to recognize the idiosyncrasies of Canadian constitutional law, although they may find on their own grounds that they conform to international human rights instruments, especially if plurally understood.

The status of the ‘notwithstanding clause’ in international human rights law

Compared with the kind of sophisticated arguments to which the question has given rise in Canadian constitutional law, the issue under international law is a priori far simpler and more straightforward. Despite the well-entrenched (albeit politically contentious) status of the notwithstanding clause in Canada, it is worth noting that there is simply no international or even domestic analogue to it. ⁸⁸ Canada made no reservation to the ICCPR when it became party to it that would have signalled a desire to make the clause operative in a sui generis way ⁸⁹ – nor, one might add, is it likely that such a reservation would be compatible with the ‘object and purpose’ of the treaty. Certainly, the constitutional status of the clause would not, per se, provide a defence to violating Canada’s otherwise clear international commitments. ⁹⁰ In fact, the idea that parliamentary majority can in certain cases override human rights flies in the face of the highly prevalent notion that human rights exist precisely as a bulwark against majorities. ⁹¹

The Canadian government has previously had the opportunity, in the Singer case, to defend the compatibility of section 33 with its obligations under the ICCPR. ⁹² As summarized by the HRC:

the Government affirms that the existence of Section 33 per se is not contrary to article 4 of the Covenant, and that the invocation of Section 33 does not necessarily amount to an impermissible derogation under the Covenant: ‘Canada’s obligation is to ensure that Section 33 is never invoked in circumstances which are contrary to international law’ … Thus, a legislative override could never be invoked to permit acts clearly prohibited by international law. Accordingly, the legislative override in Section 33 is said to be compatible with the Covenant.

Indeed, in a later case, the Canadian government argued that:

Canada’s obligation is to ensure that Section 33 is never invoked in circumstances which are contrary to international law. The Supreme Court of Canada has itself stated that ‘Canada’s international human rights obligations should [govern] … the interpretation of the content of the rights guaranteed by the Charter.’ Thus, a legislative override could never be invoked to permit acts clearly prohibited by international law. ⁹³

From a global legal pluralist perspective, this means the Canadian government presents itself as a sort of ultimate guarantor that plurally minded democratic overrides will still comply with its obligations as a state (even interpreted broadly, taking into account a margin of appreciation). This would no doubt be a paradoxical situation if Section 33 had been invoked by the federal government itself (it has never done so, at least so far), but it is more plausible if the federal government is in that posture in relation to provinces, although even that begs the question of how it might intervene to prevent a province from invoking the notwithstanding clause. Constitutional pluralism might thus, hypothetically, defeat the ability of Canada to comply with its international obligations.

Some authors in Quebec have argued that the notwithstanding clause or its invocation do not per se infringe Canada’s obligations under the ICCPR. ⁹⁴ Leaving aside the factual question of whether the notwithstanding clause ‘has never’ or ‘could never’ be invoked in violation of international law, and the considerable ambiguity of the reference to ‘clearly’ prohibited acts, this is a vanishingly subtle argument. It is not clear that it amounts to much except to say that section 33 is not a violation of Canada’s international obligations until it is used to violate such international obligations. This leaves open a theoretically conceivable, but somewhat implausible, scenario where section 33 is used to trump the Canadian Charter of Rights and Freedoms – but in ways that do not simultaneously amount to a violation of the corresponding guarantees under the ICCPR. Clearly, given the contentiousness of Bill 21 and its effects on minorities, and even assuming the ban on religious symbols in the public service is not ‘clearly’ prohibited in international law, this would seem an area where any use of section 33 should and would be closely scrutinized internationally.

Reconciling the ‘notwithstanding clause’ with international human rights law

How might section 33 nonetheless be invoked in ways that are compatible with the ICCPR? International law is not particularly formalistic, so what matters is arguably less how section 33 is understood under Canadian law than what it effectively achieves. Just as the Canadian judiciary occasionally resorts to Canada’s international obligations to interpret domestic law, international human rights bodies might exceptionally be inclined to interpret international law in light of domestic specificities in order to validate a strong sense of pluralism.

Even though section 33 is confusingly labeled a ‘derogation’ in the Canadian Charter, it is emphatically not equivalent to the ICCPR’s derogation clause. For Canada to argue that it was invoking a derogation internationally would require it to show that it was facing a ‘time of public emergency which threatens the life of the nation’. There is nothing in the record of the adoption of Bill 21 that suggests anything remotely resembling a public emergency. In fact, the question of religious symbols in the public sphere in Quebec had been debated on and off for a decade by the time the Bill was adopted. Supporters of the legislation are clear that it is primarily a law adopted to promote secularism, and in particular the ideal of laïcité. Needless to say, the Secretary General of the United Nations, as the official repository of the ICCPR, has not been notified by the Canadian government of any intent to derogate, as would be required.

An alternative might be that section 33, whatever its constitutional status in Canada, could still be analyzed internationally as part of conventional ICCPR limitations. This would, in a sense, ‘normalize’ section 33 internationally, seeing it as objectively limiting rights even though it is technically and domestically suspending them. Such a solution has been commended as one that would make sense of both Canada’s domestic and international human rights commitments in that ultimately it does not grant legislatures a power to free themselves of all the strictures normally associated with limiting rights. ⁹⁵

This would nonetheless be an odd argument to make in the context of Bill 21 – even in bad faith – given how section 33 in the economy of Canadian constitutional law is clearly distinct from section 1, which is generally understood to be the foundation of limitations to rights. It would require the HRC to go out of its way to generously interpret this Canadian oddity as a faux limitation. Even assuming the HRC were so inclined, Canada would still be in the awkward situation of having to defend as a conventional limitation what was plainly not conceived as one in the runup to the adoption of Bill 21. This would deprive Canada of a repertoire of actual and credible justifications of such a limitation: instead of carefully weighed priorities (whether Bill 21 was necessary, proportional, etc.), one would find only the Quebec National Assembly’s blunt willingness to suspend Canadian guarantees through the affirmation of legislative sovereignty.

Finally, international human rights bodies might take Section 33 for what it is – namely, a suspension – and, leaning strongly towards an understanding of Canadian constitutionalism on its own terms, consider that a sui generis third way of modulating rights (outside limitations and derogations) was acceptable internationally. This might paradoxically assist those who, domestically, have sought to argue that the invocation of the notwithstanding clause is dependent on the satisfaction of not only formal but also substantive requirements. ⁹⁶ But it would truly go against the grain of international human rights law, and would constitute a remarkable deference to a particular constitutional tradition despite the clear language of treaty commitments. Although the Canadian government strenuously emphasizes the highly hypothetical compatibility of section 33 with its human rights obligations, the better and simpler view may thus be that proposed by Pearl Eliadis, that ‘no government can “notwithstanding” itself out of international law.’ ⁹⁷

Justifying the ‘notwithstanding clause’ in terms of international human rights law

But what if, instead of awkwardly identifying a space for majoritarian politics within an overarching framework of international human rights obligations that seems set up to limit its impact, such majoritarian politics were themselves more explicitly understood in terms of a global legal pluralist approach to human rights? Indeed, a stronger and more interesting argument might seek to reformulate the use of the notwithstanding clause by drawing on the inherent ambiguities of international human rights discourse and, in particular, its own long-standing even if occasionally hesitant implication in the defence and promotion of democracy. After all, the notwithstanding clause is not external to Canadian rights protections, but rather is to be found in the very instrument purporting to guarantee them. There are three ways in which this deeper engagement with the place of parliamentary democracy within the global human rights edifice itself might be constructed, although each comes with its pitfalls.

First, the invocation of the notwithstanding clause could be seen internationally as a way of protecting some kind of ultimate residual margin of appreciation – a sort of doubling down on Quebec’s specificity, backed and made credible by a unique expression of democratic will. This would recognize that one of the ways in which pluralism is given effect, rather than merely through the weight of inherited tradition or a romanticized conception of culture (especially as seen through an international distance), is through the self-determining impulses of bold legislatures. How can one second-guess that aspiration to pluralism if it is validated by democratic deliberation and should not international human rights law, in turn, defer to such clearly expressed will? That argument would be more convincing if ordinary limitations to rights, both under Canadian constitutional and international law, did not already themselves allow considerable deferral to legislative will. Still, the centrality of democratic deliberation to actualizing rights content was amply flagged during the Bill 21 debate ⁹⁸ – and even acknowledged by the courts. ⁹⁹

Jacques Gosselin once argued in fact that, far from being merely a possibility, the use of the notwithstanding clause may in fact be mandated under international human rights law in order to give effect to cultural diversity. ¹⁰⁰ This may seem like an outlandish argument, but it at least takes seriously the notion that human rights are often caught up in a contradictory impulse mediated by parliamentary democracy: a centralizing aspiration to give universal answers combined with a decentralizing desire to give legal expression to human diversity as an intrinsic part of the human experience. Tactically, it seeks to outflank internationalist human rights supporters by portraying deference to the province’s discretion as, in fact, the ultimate international and human rights move.

Here, however, it is not clear that the ‘notwithstanding’ argument adds much to an argument from the margin of appreciation generally, as we explored it in the preceding sections – except perhaps as a further signalling device, democratically doubling down on specificities that, on their own, might justify the exercise of the margin of appreciation (and in that respect, the Quebec National Assembly is quite clear that it is merely giving legislative expression to what some consider to be a long-held vision of laïcité). But it might also paradoxically draw attention to the fact that a margin of appreciation argument cannot be won on its own grounds because, instead of pointing to a deep and uncontested tradition of secularism in Canada/Quebec whose natural ramification is the banning of religious symbols for persons in position of authority, that change needs to be forcefully legislated against what might be presumed to be the default position under the Charter and in the face of significant opposition. This might reveal the debate to be less about the genuine search for pluralism and more about blatantly illiberal majoritarian will.

At any rate, this will not assuage concerns that the notwithstanding clause is basically anathema to the project of international human rights law, perhaps all the more so given that international human rights law’s structure is already inherently pluralistic. One could argue, in a very pluralist vein, that there ought to be a ‘margin of appreciation’ about whether to have something like the ‘notwithstanding clause’ in the first place. Certainly, the clause is woven into complex Canadian constitutional compromises as a safeguard for the provinces, something to which international human rights bodies might be inclined, all other things being equal, to defer. But the margin of appreciation is typically more restrictive when it comes to features of constitutional governance that might fundamentally alter a culture of rights. It applies to limitations of discreet rights, presumably not to a broad ‘escape valve’ such as the notwithstanding clause.

Indeed, as has been emphasized, states are hardly divided on the issue of whether having a ‘notwithstanding clause’ is a good idea from a human rights point of view: to the best of this author’s knowledge, no state with a constitutional rights instrument has simultaneously seen fit to introduce a caveat that such an instrument can be overridden democratically. Indeed, internationally, such a caveat would seem to defeat the purpose of having constitutional rights guarantees. That is not to say that other states do not resort to parliamentary fiat to effectively violate rights as a matter of fact, merely that they do not do so on the basis that it is within the power of parliaments to override human rights guarantees as a matter of domestic or international law.

A second, subtler human rights argument for the notwithstanding clause would be to frame it as a sort of human rights separation of powers thesis and, in particular, the need to be mindful of the importance of parliaments as opposed to judiciaries, both out of democratic pluralist considerations and for the purposes of elucidating the complex meaning of rights. Again, rather than frontally taking exception with the imagined dictates of human rights law, this would more subtly reformulate the argument for the notwithstanding clause from within the categories of pluralism. It would caution against excessive reliance on judicial review and constitutional interpretation as the primary vehicles for the ultimate validation of rights, emphasizing the role of democratic deliberation instead. This was once Canada’s argument before the HRC, emphasizing the need for ‘a balance between the roles of elected representatives and courts in interpreting rights’:

A system in which the judiciary is given full and final say on all issues of rights adversely impacts on a key tenet of democracy – that is, participation of citizens in a forum of elected and publicly accountable legislatures on questions of social and political justice … The ‘notwithstanding’ clause provides a limited legislative counterweight in a system which otherwise gives judges final say over rights issues. ¹⁰¹

This argument can also avail itself of schools of thought that, even as they are evidently sympathetic to rights, have long been wary of the emphasis on judicial review. ¹⁰² Constitutionally, it would have the backing of those who see legislatures, including provincial ones, as having a significant role in constitutional interpretation and implementation via ‘democratic dialogue’. ¹⁰³

Even though that is arguably not the dominant position in international human rights law, the argument based on separation of powers might inject a certain doubt about the wisdom of second-guessing the Quebec National Assembly from Geneva. Incidentally, we cannot assume that courts are always the most progressive or informed interpreters of rights. It might be necessary, in order to implement international human rights obligations, for parliaments to occasionally step in to overrule the judiciary. As an example, imagine if the Supreme Court of Canada, contrary to what happened but not entirely implausibly, had found that permitting same-sex marriage violated the Charter over the wishes of particular provinces that wished to allow it. ¹⁰⁴ A degree of even pre-emptive ‘legislative review’ ¹⁰⁵ of cases that could be ‘wrongly’ decided may be a legitimate feature of a pluralist rights governance, although to the extent that it removes the ability of courts to scrutinize legislation it is not clear how it really fares in terms of separation of power.

Nonetheless, international human rights law does have a pro-judicial review and rule of law bias, which is invested heavily in the idea of courts as a final bulwark against rights violations by majorities and would presumably not look generously at legislative overrides, save in exceptional circumstances. It will probably seem, a fortiori seen from outside Canada, that the invocation of section 33 is less a kind of benevolent manifestation of rights infused democracy than it is an attempt to stop challenges to the law in their tracks through a unilateral assertion of parliamentary supremacy – precisely the sort of crass majoritarianism that international human rights bodies see it as their mission to counteract. This is especially so given the broad role already conceded to parliaments in framing normal limitations to rights, and the lack of a perceived need, internationally, for a further parliamentary ‘nuclear option’, beyond such a role. As the petitioners in the MacIntyre case against Canada put it:

Section 1 of the Canadian Charter already provides such a balance by subjecting human rights to such reasonable limits prescribed by law which are justified in a free and democratic society. Section 9(1) of the Quebec Charter contains limitations to the same effect. In the authors’ opinion, there is no justification, political expediency apart, for the presence of the ‘notwithstanding’ clauses. ¹⁰⁶

The petitioner in the Singer case before the Human Rights Committee also pointed out that ‘the “notwithstanding” clause of Section 33 is a negation of the rights enshrined in the Charter, as it allows (provincial) legislatures to attack minorities and suspend their rights’. ¹⁰⁷

A third, even more dramatic, way of thinking about how the notwithstanding clause might be rationalized internationally is as a sort of constitutional contre-pouvoir against misguided interpretations of international law by international bodies (instead of, hypothetically, domestic courts). ¹⁰⁸ In an age where the United Nations has been faulted for its human rights record, ¹⁰⁹ there is simply no reason, the argument goes, to assume that the international legal position will always be the most progressive, occasionally requiring domestic courts to step in to resist international law. Such is the legacy of the Kadi case, ¹¹⁰ for example, and in Canada itself, of the Abdelrazik case. ¹¹¹ It might be thought, hypothetically, that if international human rights bodies themselves were to adopt grossly illiberal positions, then it would fall upon states (and, as the case may be, federated entities) to ultimately resist those. The notwithstanding clause, then, might serve as a shield, if it came to that, by which the distinctness of certain state or sub-state entities could be protected against the vagaries of an unevenly and inconsistently liberal international law.

This is not entirely implausible as a theoretical argument about global juridical governance; the problem is that it hardly clear how, in this case, Quebec or Canada might claim the human rights ‘high ground’ on the basis of a law that is likely to be seen as dubious from the point of view of core freedoms. The banning of religious symbols in the public service has at best been seen in Strasbourg as an in extremis justifiable limitation to religious freedom in particular circumstances; it is hardly presented as a model for human rights internationally. ¹¹² In that respect, it is worth stressing that although the ECtHR has found that various bans on religious symbols did not violate the ECHR, this was merely as a form of tolerance of the practices of specific countries (Turkey, France, Switzerland or Belgium). At any rate, the idea that the ‘notwithstanding clause’ would claim exception from both the Canadian Charter of Rights and, indirectly, Canada’s international human rights obligations remains more than a little fictitious and is unlikely to be received well by the Human Rights Committee. It threatens to entirely inverse the logic of international supervision and the well-entrenched notion of the superiority of international (human rights) law.

In the end, therefore, the notwithstanding clause – notwithstanding its strong gesturing domestically at the plural irreducibility of provinces – would seem to so prioritize majoritarian democracy over minority rights as to be hardly compatible with the letter or spirit of international human rights law. This is all the more so given that its pluralism affirming-dimension in the context of Canadian federalism is joined at the hip with a pluralism-denying impact in relation to Quebec religious and Anglophone minorities. That impact is unlikely to be considered generously internationally from the point of view of an international human rights project that is not privy to the constitutional intrigue that led to the adoption of section 33, and therefore more likely to see its negative consequences on rights unfiltered. The better view from an international perspective is that human rights protections should not be secured only at the cost of some radical escape clause that seems to defeat the very purpose of having human rights protections.

V. Conclusion: Mediating the local, the national and the global

The adoption of Bill 21 is an opportunity, beyond the circumstances of Quebec and Canada, to think more generally about the arduous exercise of reconciling international human rights commitments and local practices along pluralistic lines. It radicalizes a number of burning questions at the heart of the international human rights project that implicate its relation to constitutions, federalism and parliamentary democracy. The dispute concerns the very nature of the international human rights project as potentially less a project of ruling through human rights than as a continuous process of global constitutional governance with the question of plurality at its core.

This article has sought to transcend facile dichotomies between the international and the domestic, the federal and the federated, as well as constitutional rights and governance, by instead foregrounding the pluralist framework as one that seeks to reconcile opposites on every level. Rather than thinking in terms of the ‘validity’ of Bill 21 and its compatibility with rights and freedoms – an issue on which the article ultimately took no direct stance – it was suggested that every conversation about the law implicates the negotiation and constitution of relevant frames of reference and the ongoing struggle between them. These go to the heart of a legally pluralist mindset ¹¹³ that is sensitive to the need not only of ‘discerning multiplicity in the world, but also to articulating the reasons why it might be valuable, and when it ceases to be so’. ¹¹⁴

The article has suggested several ways in which one might mediate the tension between sovereignty, federalism and domestic constitution on the one hand and a commitment to global values on the other, even as it has found some more convincing than others: while the notion of the margin of appreciation would make sense of Canada’s specificity within its broader international environment, it is less clear how the margin works in the case of federal states, let alone federated entities. Some of the compromises on which Canadian constitutional rights culture is built (such as the ‘notwithstanding clause’) may appear internationally as though they steer the project too far in the direction of democratically ordained pluralism at the expense of a culture of rights supervision. Yet pluralisms are also crucially dependent on each other: just as it is sometimes argued that the notwithstanding clause would not have been necessary in a system that was more cognizant of something like the margin of appreciation, ¹¹⁵ the more international human rights law defers to national and democratic specificity, the more artificial opposition to it on parliamentary sovereignty grounds may appear.

Having said that, the language of mediation between universal values and local preferences (or universal preferences and local values) also comes with its own pitfalls. Perhaps the first and most obvious is the danger or reification of the nation-state as the proper frame of reference to ascertain the validity of human rights claims, even within the horizon of an otherwise pluralistic margin of appreciation reasoning. Why should the plausibility of a human right claim or limitation be assessed in relation to ‘the United Kingdom’ or ‘Spain’ rather than ‘Scotland’ or ‘Catalunya’? Why Canada but not Quebec? Or for that matter, why the EMSB but not Kanawake? To put it another way, why are international human rights involved in the affirmation of state/national identity at all, and why should they recognize its prima facie claim on our political intuitions? Should it make a difference whether that national identity is itself contested? Can one challenge the reference to Canada without reinscribing its statism by upholding Quebec as an alternate locus of sovereignty? Can one affirm the prerogatives of a distinct federated entity without re-subjecting minorities within the minority in the process? And what of the possibility that, in deferring to a particular society’s ‘traditions’, international human rights bodies will powerfully – and perhaps unwittingly – reinforce majoritarian effects, not to mention unduly stereotype complex traditions?

Clearly and perhaps ironically, the human rights question in this case ends up restating the national question (not which rights but why this state?). While this article took the ‘international’ as a prism through which to consider a domestic human rights issue, the issue might just as well be what ought to be considered domestic and international in the first place, or how far within societies our understanding of human rights pluralism should extend. The fact that Quebec’s claimed specificity when it comes to freedom of religion has already been met by forms of resistance to Bill 21 from the city of Montreal or leading hospitals and universities in Quebec that have announced they do not intend to enforce the ban suggests there is no necessary or logical end to how far one should take the pluralist logic. There is certainly no reason to think that pluralism should end at the more or less arbitrary federal subdivisions of the state, or even at the state itself.

A second problem with the variable geometry implicit in a pluralistic understanding of rights is how it impacts the question of equality. How comfortable should one be with a system of human rights (and not just, say, international law) that concludes that it is fine, based on each country’s socio-legal legacies, for Turkey to ban the hijab in universities but that would most likely find that it was wrong for the United Kingdom to engage on the same path? Or which, by extension, considers that it would be wrong for Canada to ban religious symbols in the public service but not for one of its provinces to do so? What does it mean that the same international human rights order can both countenance a limitation and exclude it? Does this dilute the sacredness of human rights, artificially inflating the importance of territory and political community over the promise of having rights anywhere, merely by virtue of being human? Does it allow human rights to lapse into the utter relativism of space and time? Or was the sacredness/universality of the human rights project the wrong way to frame the issue in the first place – one insufficiently open to the chronic challenge of translating rights into particular legal contexts?

The consequences of adopting a pluralist approach to human rights are rarely spelt out and would need to be better understood from a human rights point of view. The question is whether there is a human rights justification to pluralism, and not simply a sovereignist/nationalist one. As I have suggested, it is not impossible to think of ways in which understandings of rights could be localized on the basis of historical, communitarian and political specificities, notably by appealing to collective and democratic foundations. However, in a context where all sides of the Bill 21 debate frame their arguments in human rights terms, the onus is on supporters of the law to explain the conditions under which a particular type of societal priorities might trump individual rights.

A third problem is with the potential conservatism of rights pluralism. By making each society the standard referent by which to evaluate rights arrangements, is one not at risk of feeding into a certain circularity? Put differently, is the danger not that one will reify an immutable vision of a society (e.g. a Francophone, secular, white Quebec) that can then, having been legitimated internationally, be deployed with a vengeance against minorities? Is the margin of appreciation itself, for example, not a recipe for political stasis, an intimation that political entities that did not record their peculiar conception of religious freedom cannot now change their view and innovate? Do some states get away with apparent violations merely because they have ‘grandfathered’ their more restrictive rights practices? Allowing the banning of the veil because, in effect, ‘this is how things have always been done’ or disallowing the banning of the veil because ‘this is not how we go about this’ is deeply problematic. In many ways, human rights have constructed themselves against tradition and as part of an effort to provide a universal and transcendent standard of political achievement, not as an apology for whatever the powers that be decide is the content of rights.

On balance, this article’s suggestion is that the challenges that emerge from the variable geometry of international human rights law are preferable to those that emerge from imposing an illusory, artificial and hegemonic single standard of rights. Only a pluralist concept of human rights takes seriously the notion that rights take meaning in particular societies even as, paradoxically, these societies do not (or no longer) have the monopoly over their interpretation. Indeed, there is no principled reason why that pluralism should not be extended within and beyond the state, despite international human rights law’s historical focus on the peculiar constellation of power that is sovereignty. Nonetheless, there is a point where too much deference to collective preferences may spell not so much the adaptation of rights to local circumstances as the surrender of rights to majoritarian politics, leading to an unhelpfully reified notion of national identity. That is, ultimately, the stake that the pluralist detour through the Bill 21 debate serves to radicalize in the global constitutionalist project.