Essay Volume 66:1

Lost in Translation? Bill 21, International Human Rights, and the Margin of Appreciation

The adoption of Bill 21, which bans religious symbols for civil servants in Quebec, has stirred considerable debate politically and constitutionally in the province and in the rest of Canada. Neglected, however, has been a more in-depth analysis of how international human rights law often serves as an implicit frame of reference for many of the debates surrounding Bill 21. This essay focuses, in particular, on the invocation of the case law of the European Court of Human Rights, which seems to have validated bans of religious symbols in various contexts. It gives an overview of that jurisprudence and specifies the parameters within which it operates, emphasizing the complexity of translating a supranational case law into a domestic debate. It argues that whilst Quebec is less alone in banning religious symbols than is sometimes argued, the European case law needs to be handled carefully. In particular, the essay emphasizes the importance of the so-called “margin of appreciation” as heavily impacting the outcome in those cases. Although the margin suggests that there is national leeway in adopting bans based on certain national traditions and specificities, it hardly opens the door to all bans. Rather, the margin emphasizes the significance of divergences between states parties on an issue and respect for procedural safeguards. The essay concludes with some thoughts on how importing human rights arguments out of context can be perilous, but also about how the margin itself may be problematic.

L’adoption de la Loi 21 interdisant le port de symboles religieux pour les fonctionnaires du service public au Québec a créé une controverse politique et constitutionnelle considérable dans la province et dans le reste du Canada. Néanmoins, la manière dont la dimension internationale, notamment en droits humains, a souvent agi comme schème de référence implicite dans les débats, a été négligée. Cet essai étudie notamment l’invocation de la jurisprudence de la Cour européenne des droits de l’homme, laquelle a validé des interdictions du port de symboles religieux dans plusieurs contextes. En donnant un aperçu de cette jurisprudence, il s’agira d’identifier les paramètres à l’intérieur desquels elle procède et de mettre l’accent sur la difficulté à transposer des décisions supranationales dans un débat interne. Si le Québec n’est en effet pas seul à banir les symboles religieux, les décisions européennes doivent être comprises dans leur contexte spécifique. L’essai met l’accent sur l’importance de ladite « marge d’appréciation » comme ayant un effet considérable sur l’issue de ces jugements. Bien que la marge suggère un espace de manœuvre dans l’adoption d’interdictions fondées sur certaines traditions et spécificités nationales, celle-ci n’est pas une porte ouverte à toutes les interdictions. Elle place au cœur de la réflexion l’existence de divergences entre États parties ainsi que le respect de certaines garanties procédurales. Il s’agira, en définitive, de bien évaluer les dangers d’importer des arguments tirés des droits humains hors contexte, mais aussi de souligner certains problèmes liés à la marge d’appréciation elle-même.

* Professor and William Dawson Scholar, Faculty of Law, McGill University, and Co-Director, Centre for Human Rights and Legal Pluralism. I am grateful to insights and comments from Johanne Poirier and Stéphanie Hennette-Vauchez on previous versions of this essay.

Table of Contents


I.  Transnational Influences on the Bill 21 Debate

A.  Deprovincializing Bill 21?

B.  Some Preliminary Caveats

II. Speaking of the Margin of Appreciation and Its Dialects

A.  The Substantive Margin and the European “Consensus”

B.  The Procedural Turn in the Margin of Appreciation



The adoption of Bill 21, the (in)famous Loi sur la laïcité de l’État,[1] which bans the wearing of religious symbols in the public service in Quebec, and the fracas it has caused are an invitation to ponder the limits of religious freedom, the nature of majority rule, or the definition of secularism. Rather than address these issues on their own terms, however, this essay seeks to bring attention to the various national, regional, and international legal frames of reference within which they arise. It aims to show how the way these frames are used can be highly determinative of how one addresses these underlying issues.

If nothing else, Québécois and Canadians should care about international law on this matter because it is highly likely, in the event that the law survives constitutional review in Canada, that it will be further challenged internationally. Indeed, the domestic constitutionality of a law has never been, per se, an argument for its international legal validity.[2] However, the point is that the influence of international law is already visible in the public debate surrounding Bill 21 and extends much further than the simple question of whether the proposed law “violates” international law or not. In Quebec, the transnational influence of ideas about rights has occurred most spectacularly through the importation of mostly French views of laïcité that have been foregrounded with much vigour in the public debate.[3]

But the existence of another national model that seems inspired by the same principle as Bill 21 does not provide conclusive evidence of the law’s validity in Quebec—especially since the French model has itself been criticized as problematic.[4] This is why some of the more sophisticated defenders of Bill 21 have not failed to point out that the French model and others like it that have banned various forms of religious symbols in the public sphere have been challenged before the European Court of Human Rights (ECtHR), only to see states win.[5] Superficially at least, as will be discussed in more detail in this essay, this seems like a remarkable validation of the case for Bill 21. But the basis on which such invocations of international case law are made needs to be clarified. In this essay, I suggest that the fact that international human rights bodies, notably the ECtHR, have occasionally allowed religious symbol bans does, in fact, less work than is assumed by proponents of Bill 21—although it also probably does more work than opponents concede. My more general contention will be that the import of international human rights law into the Quebec debate, even assuming a best-case scenario in which that law is applicable and binding in Canada (which, in the case of the European Convention on Human Rights, it clearly is not), would require significantly more nuance than has been displayed in the debate so far.

That debate, in particular, has been marred by inadequate “translation” of what that a priori sympathetic case law means in its own context, let alone what it might mean in the Canadian context. I suggest that the ECtHR case law is the legal equivalent of a classic linguistic faux ami: bearing an uncanny resemblance to something familiar but in fact not (quite) the same thing. Even the meaning assigned to such religious symbols as the “veil” is heavily constrained by the contexts within which the issues surrounding it arise: there are as many conceptions of the veil—imagined, feared, and fantasized as it may be—as there are jurisdictions or purposes for its regulation. The broader point is that if we are to have international and inter-regional dialogue about rights—a process which in principle should be encouraged—then we must invest significantly in an understanding of how each vernacular of rights functions on its own terms, lest we mistake our grappling with rudiments of a foreign language for proficiency in it.[6]

In Part I, I highlight how the debate on banning the veil in Quebec has been both enriched and subtly misled by references to international law, particularly European human rights law, showing how a number of cases—taken individually and at face value—seem to reinforce the argument for the legality of limited bans of religious symbols. In Part II, however, I contextualize the outcomes of these European decisions as part of the ECtHR’s “margin of appreciation” reasoning. This renders significantly more complex the import of arguments taken from the European context to the situation of Quebec and Canada. Whatever international human rights case law tells us about the validity of banning the hijab or the niqab in certain contexts must be deeply adjusted for local specificities that do not necessarily apply in Canada. I conclude with a conceptualization of international human rights law as a body of law which allows us to sharpen our sense of the stakes of domestic and localized struggles by seeing them, as it were, “from outside” rather than as a ready-made prescription for complex legal and political dilemmas.

I.  Transnational Influences on the Bill 21 Debate

Shortly after its election in 2018, the provincial Coalition Avenir Québec (CAQ) government proposed Bill 21, the so-called Loi sur la laïcité de l’État.[7] The law was enacted in June 2019.[8] It proclaims Quebec’s secularism and requires the government’s neutrality in relation to religions. Most notoriously, it requires certain categories of civil servants, notably those exercising coercive authority such as police officers, judges, and prison guards, to refrain from wearing any visible religious symbols. All civil servants are prohibited from covering their faces when providing public services.[9] The law comes on the heels of more than a decade of political debates about the “reasonable accommodation” of religious minorities,[10] including one influential commission,[11] several failed legislative proposals,[12] as well as one adopted law.[13]

In Canada, this is largely a Quebec-based debate. It is one that is deeply embedded in the province’s rapport with the rest of the country and its sense of identity, although it has had echoes in other provinces as well.[14] It has affected a myriad of constituencies and raises questions about freedom of religion, discrimination against religious minorities, freedom of expression, equality, liberty, and security of the person. The bill has significant—although far from universal—support in Quebec. It has provoked an often fierce reaction both within and outside Quebec from religious minorities who stand to be affected by it and others who see it as an odious encroachment on basic liberties.[15] It has led to so far unsuccessful litigation to overturn it in Quebec.[16]

But the law, or at least some of the principles behind it, have also been defended passionately by secular-minded liberals and some feminists.[17] One of the main arguments in the “secularist” camp has been that the law is far less anomalous than it is often portrayed as being and that it is, in fact, inspired by the French approach to laïcité from which it derives a certain legitimacy. I leave aside the vast and distinct debate about whether France’s approach can be imported into the Canadian context wholesale and on what basis (not to mention the contested character of laïcité and its evolution in France itself). However, those who invoke the French case must concede that it is a mere domestic precedent, with little legal traction by itself.

Indeed, faced with a barrage of opposition, it was not long before pro-Bill 21 commentators picked up on a vast body of a priori sympathetic international human rights case law.[18] The unique advantage of that body of law is that it shows the French model as far less of an outlier than it is sometimes described as being. The ECtHR, in particular, had found that the banning of the burqa in France or Belgium was not illegal, nor was the banning of the hijab in Turkish universities or Swiss kindergartens. As one commentator had put it much earlier in the debate on “reasonable accommodation”:

Le Québec se mettrait donc au banc de l’Occident et de l’ONU? C’est simplement faux. Voici la vérité, … Elle est bien différente de ce que prétendent MM Bouchard, Taylor et tous ceux qui voudraient culpabiliser les Québécois de tenir, chez eux, un débat qui a cours dans plusieurs autres démocraties avancées, dans le respect des normes internationales.[19]

In the same spirit, a letter addressed to the CAQ leader by twenty-five Québécois lawyers supporting Bill 21 noted that

[c]e même objectif de laïcité, de nombreux autres États dans le monde l’ont inclus dans leur droit. Le Québec ne serait certes pas le premier État à adopter des règles de droit pour restreindre le port de symboles religieux par ses fonctionnaires dans l’exercice de leurs fonctions. À l’échelle internationale, de telles mesures ont été soumises dans de nombreux cas à un processus de contrôle judiciaire, tant devant les tribunaux nationaux que devant la Cour européenne des droits de l’homme, et ont été jugées valides. … L’expérience européenne en atteste, l’interdiction de symboles religieux chez les fonctionnaires de l’État n’est pas un crime contre l’humanité et peut s’avérer une mesure législative tout à fait légitime et nécessaire.[20]

Many others have pointed to the European case law as showing Bill 21 in a much less negative and isolated light.[21] This train of thought eventually made its way into the reasoning of Justice Mainville in his separate opinion in a Quebec Court of Appeal decision on an application for a provisional stay of the bill. As he put it, “The European Court of Human Rights had numerous opportunities to consider the issue and concluded that this type of legislation does not infringe the freedoms of thought, conscience and religion set out in the European Convention on Human Rights.”[22] Subsequently, the Crown commissioned a number of expert reports on the question of banning religious symbols in the public service, including one which squarely focused on the treatment of the issue in the European context. Co-authored by Professors Marthe Fatin-Rouge Stefanini and Patrick Taillon, it provided a remarkably detailed and nuanced presentation of what remains an ill-understood legal reality in Canada, in ways that seemed sympathetic to Bill 21.[23]

But what exactly is the importance of the Strasbourg case law in Canada? Leaving the hyperbole of political commentary aside, those in Quebec who draw on the European case law are of course not misinformed about its legal applicability. They evidently do not claim, either in the court of public opinion or, presumably, in actual courts, that the European case law is actually binding before Canadian courts. In effect, any notion that international human rights law, let alone European human rights law, would be directly helpful in the Canadian debate was set aside in no uncertain terms by Judge Marc-André Blanchard of the Quebec Superior Court.[24] This was a fortiori the case given the outsized role that that the invocation of the section 33 “notwithstanding” clause had in that debate.[25]

The point made by those who invoke European precedents is a broader and subtler one. Thinking about the international dimension of rights debates has often been impoverished, in Canada and elsewhere, by a simplistic, “command and control” model that emphasizes the vertical relationship of international law to domestic law. Many have noted that this opposition is a descriptively inaccurate and unhelpful reflection of the practice of both political and judicial actors in Canada.[26] The reality is that international human rights law instruments are as likely to be invoked in general and public debates about the search for the best human rights policy rather than simply as arguments with which to “win” courtroom battles. At any rate, the discussion already antipates what might be, whatever the limitations of Canadian dualism, international contestations of Bill 21 (for example, before the Human Rights Committee) if it continues to be found to pass constitutional muster by Canadian courts, for example as a result of the invocation of the notwithstanding clause.

In effect, the existence of a congenial international human rights case law on bans of religious symbols already seems to have served a variety of subtle political functions domestically. The invocation of these European precedents has tended to smuggle the French model of laïcité back into Quebec,[27] except this time as one validated by the Strasbourg court: transnational and supranational influences thus combine harmoniously to mutually reinforce each other. It has deprovincialized the Bill 21 approach, presenting it as part of a much larger movement and compensating for Quebec’s intra-Canadian isolation by drawing on an extra-Canadian sense of community; it has provided a kind of totemic retort to opponents of Bill 21 who are eager to invoke international human rights arguments, turning the table on them and, as it were, outperforming them in their imagined internationalism; and it has helped repolicitize complex human rights debates as involving a healthy dose of democratic indeterminacy. It could even show proponents of Bill 21 in their best light as dutiful servants of what international human rights law, in fact, requires.

A.  Deprovincializing Bill 21?

The basic point that a range of attitudes toward the regulation of religious symbols (including bans) has been found compatible with international human rights law[28] is, to a significant extent, correct and ought to be acknowledged as such. It frames the debate as one between different traditions of liberalism rather than between liberalism and illiberalism.[29] In fact, it is probably true that more international human rights bodies have found religious symbol bans to be compatible with human rights than the opposite. While the more sophisticated commentators have noted that European courts have not condoned blanket bans of religious symbols (with the notable exception of the niqab), they rightly point out that the European case law at least points to the possibility of a variety of intermediary, carefully calibrated bans.[30] This is precisely, as it happens, the kind of ban promoted by Bill 21, which only targets civil servants in positions of authority.

I can only give a cursory overview of that complex European jurisprudence here, but its context and broad outline are as follows. The European Convention on Human Rights (European Convention) was adopted in 1950 and has been in force since 1953. It applies to the forty-seven members of the Council of Europe, a wide variety of countries located on the European landmass, including states as diverse as France, the UK, Russia, or Turkey. The ECtHR, which hears challenges from nationals who claim that their European Convention rights have been violated, has jurisdiction over more than 800 million individuals—one-eighth of humanity.[31] Although on some level the ECtHR is only a regional court and one regional court among others, this expansiveness explains why the Court is seen as a primus inter pares: the single most efficient, binding, and productive supranational human rights jurisdiction in the world.[32]

For most of its history, the Court hardly heard any case involving the wearing of religious symbols and secularism. The adoption of bans on religious symbols is a late development, often precipitated by the emergence of significant Muslim minorities in a context of immigration to Western European states and resulting anxieties in host nations. The one exception is Turkey, which is a country with a ninety-nine per cent Muslim majority, but where a rigid doctrine of separation of religion and state took hold a century ago under the leadership of Mustafa Kemal Atatürk.[33] As will become clear in the next section, it is highly meaningful that the question arose with particular intensity in some countries and not in others. It is also pertinent that the issue crystallized around various forms of the Islamic veil (hijab, niqab, etc.) rather than other religious symbols.

The ECtHR has, time and time again, validated bans of different types of veils. Moreover, those court decisions were often unanimous or adopted by sweeping majorities. For example, in the leading Leyla Şahin v. Turkey decision in 2005, the Grand Chamber found no violation of article 9 (freedom of religion) of the European Convention by sixteen votes to one and unanimously found no violation of articles 8 (right to privacy), 10 (freedom of expression) and 14 (non-discrimination) as a result of a university student being prevented from taking exams because she wore the Islamic headscarf.[34] Similarly, the Grand Chamber in S.A.S. v. France in 2014 found that France’s prohibition of the niqab did not violate articles 8 and 9 by fifteen votes to two, and unanimously held that there had been no violation of articles 10 and 14.[35] In 2008, a Chamber of the Court (5th Section) unanimously found no violation of article 9 in Dogru v. France and Kervanci v. France after two high school students were excluded from physical education for having worn a headscarf.[36] Another Chamber (Second Section) unanimously decided in 2017, in Belcacemi and Oussar v. Belgium and Dakir v. Belgium, that Belgium’s niqab ban did not violate articles 8, 9, and 14.[37] Finally and perhaps most notably for our purposes, in the 2015 case of Ebrahimian v. France, the Court found no violation of article 9 after a hospital social worker’s contract was not renewed because she would not remove her headscarf.[38] Adding to this list of decisions on the merits are a series of cases where complaints about bans were dismissed at a preliminary stage as manifestly ill-founded and inadmissible.[39]

The litany of cases finding bans not to be in violation of human rights obligations is therefore long and remarkably consistent, lending at least superficial credence to the idea that such bans may well be compatible with human rights. Proponents of Bill 21 who invoke European precedents are also on solid ground in one other key respect: much of the practice of human rights law—certainly in Europe but also in Canada and elsewhere internationally—involves sophisticated discussions about how to limit and weigh human rights.[40] Rather than simply proclaiming the contours of rights and then checking whether they have been violated, most decisions end up routinely spelling out ways in which states might legitimately restrict the conditions under which rights are exercised for a range of reasons linked to public order, morality, or the rights of others. This parallels Canadian constitutional law and its emphasis on the conditions for admissible limitations on Charter rights, including that the limitations be reasonable and demonstrably justified in a free and democratic society.[41]

The ECtHR, it is true, sometimes does its best to not concede that point too explicitly. It has argued, for example, that it seeks to achieve “a just balance between the protection of the general interest of the Community and the respect due to fundamental human rights while attaching particular importance to the latter.”[42] Be that as it may, limitations are typically the core focus of human rights adjudication. Hence the overwhelmingly negative tone of much of the human rights jurisprudence that comes across particularly clearly in the religious symbol case law: the (interesting) question is not whether you can wear the hijab (the default opening position being typically—but often deceptively—liberal-sounding), but when the state can prohibit you from doing so. At any rate, proponents of the ban in Quebec can point not only to a specific position emanating from the European case law, but also to a broader ethos in which the central stake of human rights is an exercise of weighing individuals’ rights against collective prerogatives.

Moreover, given the importance of social values in this weighing exercise, there is certainly no reason to think that the most individual-friendly, liberal position is always the one that will be chosen. This is particularly the case when it comes to issues involving strong state preferences, which supranational rights protections have struggled to offset.[43] It is even more so in a context where the European system has been strongly contested by a number of states (including the UK, Russia, and Turkey) and where the Court’s jurisdictional legitimacy sometimes hangs by a thread.[44] The Court, in short, may be tempted to not rock the boat too much by adopting judgments that manifestly run against the grain of (what are felt to be) a particular society’s heartfelt values or prejudices. Indeed, the Court has at times seemed to embrace whatever justification states threw at it when it came to curtailing the rights of religious minorities. These have included the aim of protecting the rights and freedoms of others and public order, even in cases where it was not clear how either might be threatened, and even, when that failed, a heretofore unheard-of ground such as “vivre ensemble.”[45] The ban of the hijab has also been upheld as conducive to gender equality in a context where much evidence points to the contrary—or at least to the complexity of the trajectories of those who actually wear the veil,[46] and where projects of secularism have not themselves necessarily been notable historically for their attention to gender equality.[47]

What is remarkable about the plasticity of human rights discourse, then, is that the limitation of the freedom to wear religious symbols can be framed as a measure to protect human rights. This saves the Court the discomfort of being seen as prioritizing state prerogatives over individual liberties and instead allows it to present itself as merely engaged in a complex trade-off between different rights or the rights of different groups. At times, the ECtHR comes dangerously close to suggesting that a state must (as opposed to merely can) ban religious symbols in order to maintain its neutrality vis-à-vis religions and thus guarantee freedom of religion.[48] Laïcité, then—at best originally one mode among others of organizing relations between state and religion—seems to have become a human rights value in itself,[49] and to have been reified as an impromptu ground for limiting individual rights.

The Court has also tiptoed repeatedly around any suggestion that discriminatory animus was involved in bans. It has done so by suggesting that since the bans applied to all religions, they could not be discriminatory.[50] This is despite ample contextual evidence that bans of religious symbols not only disproportionately affect Muslim women, but were often more or less explicitly designed to do so in the public debate. Note that this contrasts with the ECtHR’s alertness to discriminatory outcomes when it comes to wearing the veil itself, which is portrayed as negatively affecting women[51]—not to mention a certain obliviousness by the Court to potentially discriminatory outcomes when the historically dominant Christian religion is concerned.[52] This fits with a tendency among some in Quebec to see discrimination selectively: Bill 21 is unimpeachable because of its neutral application to all religions (no discrimination) and its positive attitude to fighting gender oppression (anti-discrimination), whereas the veil is clearly problematic because it symbolizes the oppression of women (discrimination).

At any rate, whatever one thinks of these judgments, they do help establish a fairly powerful prima facie case for the notion that the Quebec position is not an outlier internationally. In fact, as we have seen, some may be tempted to argue that this position is fully in line with evolving human rights “best practices.” In particular, the ECtHR has provided a sort of broad conceptual apparatus for all bans, resorting to similar arguments in a range of cases. That one is not entirely outside the boundaries of human rights respectability globally may not be much of a human rights argument for Bill 21, but it certainly does no harm to be able to portray one’s position as part of a distinct secular sensibility that European courts have signaled does not necessarily offend the rights canon.

B.  Some Preliminary Caveats

Nonetheless, in the rest of this essay I want to suggest that there is often something distinctly ill-informed about invocations of the ECtHR’s case law by proponents of Bill 21, which points to a deeper misunderstanding about the nature of that case law and how it might be imported into the Canadian public and legal debate. On a very preliminary level, at least three aspects of that case law must be underlined. They will help to carefully circumscribe those decisions to their rightful place, and to caution against believing that it provides an uncontentious basis for the adoption of a law in Quebec.

First, all of the areas where the limiting of the veil has been found not to be a violation of human rights were quite specific. It is not within the scope of this essay to engage in a comparative political analysis of why certain polities are more or less concerned with certain symbols, in certain places, and as worn by certain persons. Suffice it to say that quite different issues are involved in each case and that we have reason to think that the justifications for banning the veil in one context do not translate easily, even mutatis mutandis, to another. In other words, not all bans of religious symbols provide authority to impose them specifically in relation to civil servants. The one instance where a near total ban has been allowed is in the very specific case of the niqab. For the rest, the majority of the European case law is heavily focused on religious symbols worn in institutions of education, from kindergarten to university. Moreover, this is still an area of jurisprudence that is evidently quite torn by the fact that it is required to arbitrate between rights, and between the rights of certain groups of individuals and the prerogatives of the state. It is not, by any means, a licence to curtail the wearing of religious symbols in all contexts.

The case law on bans specifically targeting the wearing of religious symbols by civil servants is, as it happens, both scarce and contradictory. This is so even though one might think that justifications in this case are more readily available given the sensitivity of being in a position of authority and representing the state. The ECtHR’s Ebrahimian judgment is, as has already been noted, quite clear in allowing a ban. But it is also worth noting that some attempts to introduce a prohibition on religious symbols in public services have been struck down domestically as incompatible with the European Convention. The Belgian Council of State was asked to give an opinion about a draft bill that would have required “civil servants [to] refrain, during the exercise of their function, from showing any distinctive philosophical, religious, community or partisan expression.”[53] It found that “the proposal does not justify enough” its broad application to any civil servant “independently of the nature of his or her function and the fact that he or she exercises it in contact with the public or not.”[54] The German Constitutional Court also found that a Muslim teacher could not be forbidden from wearing a headscarf in a public school.[55] The fact that the headscarf ban was found incompatible with the European Convention in some countries does not mean, as we will see in more detail below, that no ban can be adopted. But nor does a finding that a ban is legal in some countries mean that it is legal for alleven under the auspices of the same international human rights instrument.

Second, it should be noted that these ECtHR decisions and a variety of domestic ones applying the European Convention have been roundly criticized both internationally[56] and from within the countries involved.[57] The part of the Şahin judgment that found no violation of freedom of religion was adopted by the Grand Chamber by a majority of sixteen to one, but the separate dissenting opinion by Judge Tulkens was neither anecdotal nor based on mere technicalities. It involved a wide-ranging and eloquent critique of the majority as engaged in: a simplification of the socio-cultural meaning of the hijab; a minimization of the ripple effects, notably in terms of education and access to employment, for those affected; a tendency to paternalistically assume that Muslim women are alienated by their religion; the reification of purported state values whilst minimizing their contentious and artificial character; the failure to appreciate the feared threats associated with religious symbol in concreto, and the tendency, instead, to take at their word the generalities adduced by states.[58] This is not to deny that the Şahin judgment is authoritative, but to point out the contentiousness of its reasoning, from the point of view of both religious freedom and women’s rights. Moreover, there has long been a concern that the ECtHR’s relative enthusiasm for allowing bans in cases of Islamic religious symbols must be contrasted with its unwillingness to find that the presence of the crucifix in certain state buildings, for example, violates the rights of non-believers. This may lead to suspicions that the whole issue is one redolent of religious and maybe even ethnic discrimination[59] (with eerie similarities to the Quebec debate, as it happens).

In short, many of the arguments that one hears in opposition to the banning of religious symbols in Europe perhaps predictably echo those made in Quebec and in the rest of Canada against Bill 21. They certainly belie any notion that either Europe, or indeed particular countries within it (France and Turkey come to mind), are of a single mind on these issues. Of course, it could be that the critics are simply wrong, but one must at least recognize that these are highly controversial matters that have not been settled once and for all on the basis of a few faraway judgments. This is especially true given the evolutive nature of the ECtHR’s own jurisprudence and its understanding of the European Convention as a “living instrument.”[60] If one takes into account national judgments interpreting the European Convention, there has certainly been, as we will see, significant dynamism to that case law. It must also be seen in light of significant pushback from leading civil society organizations in Europe[61] and a range of commentators.[62] In other words, the point about the ECtHR case law is well taken to a degree, but it needs to be contextualized in the inherent contradictions and dynamism of that case law.

Third, it is important to understand what a judgment by the ECtHR means and what it does not mean. The findings in Dahlab and Şahin are that Switzerland and Turkey did not violate their obligations under the European Convention. However, this “violations framework” sets the bar quite low.[63] It indicates that the relevant states are not fundamentally in contempt of their obligations, but it does not necessarily applaud them either. The Court agreed that Turkey could ban the wearing of the hijab in university, not that it should (and even less, as we will see shortly, that others should). This is a dimension worth bearing in mind when thinking through the transnational import of such seemingly sympathetic case law. The practices at stake are not illegal and invalid, but they could still be found significantly wanting from a rights point of view. In other words, simply because a practice is not a violation of human rights according to certain courts at a certain point in time does not make it perfectly commendable from a broader human rights perspective.

Note also that there is what one might call a liberal asymmetry involved in the curtailing of rights such as the freedom to manifest one’s religious belief: there are almost no cases of individuals complaining that their rights are violated as a result of others being allowed to wear religious symbols.[64] In fact, such cases would be highly likely to fail in the face of strong liberal assumptions in favour of the freedom to wear such symbols. By contrast, there are, as we have seen, many cases where individuals have complained that their rights were being violated by virtue of having been denied the ability to manifest certain religious beliefs. Although many such cases were lost, they testify to the vigour of rights opposition to bans and have at least forced states to be on the defensive, even before the most sympathetic of benches. Banning and not banning religious symbols are thus hardly on a par as human rights options, and only the former attracts considerable and almost instinctive human rights suspicion.

Finally, it bears mentioning that the attention to the international human rights case law by supporters of Bill 21, whilst not indefensible, is quite selective. This could be a clue that something is amiss. In particular, very few of the commentators who so enthusiastically cite the ECtHR simultaneously point to the case law of the United Nations Human Rights Committee (HRC) on the question. Although the HRC might be found to be a less authoritative source than the ECtHR since it is not a court, it monitors the International Covenant on Civil and Political Rights (ICCPR). The ICCPR is a universal and not just regional human rights instrument and, more importantly, one to which Canada is a party and is therefore bound. It is particularly striking that the reports commissioned by the Crown on the issue discuss the European Convention (to which Canada is not and could not be a party) at length, but largely omit references to the ICCPR.[65]

This may not be entirely coincidental. As it happens, the HRC has been much more willing to find bans of the veil to be a violation of freedom of religion. In the Hudoyberganova v. Uzbekistan case in particular, the HRC had to render a view on whether the freedom of thought, conscience, and religion of Raihon Hudoyberganova, a young Uzbek student, had been violated as a result of her being repeatedly expelled from university for wearing the hijab.[66] The HRC found that it had, although admittedly this was largely because Uzbekistan had failed to justify the measure. The HRC also explicitly said that it did not want to prejudge “the right of a State party to limit expressions of religion and belief in the context of article 18 of the Covenant and duly taking into account the specifics of the context.”[67] Perhaps more to the point, in the Seyma Türkan v. Turkey case, the HRC found that banning a university student from wearing a wig on the basis that it was a substitute for the veil was a violation of her freedom of religion and a form of discrimination.[68] Finally, in 2018, after two women complained that they had been fined for wearing the niqab, the HRC issued two decisions against France that took a very clear stance against its ban of the niqab.[69] Again, although seemingly leaving all of its options open, the HRC found that, in the absence of a better justification by France (which had come prepared and was well represented), the law disproportionately harmed the petitioners’ right to manifest their religious beliefs and therefore their freedom of religion.

It is true that none of these cases tested the banning of religious symbols for certain civil servants specifically. But what is notable is that all conclusions of the HRC were at odds with similar case law by the ECtHR, suggesting an approach that is much more liberal in principle toward freedom of religion and therefore potentially more likely to be unfavourable to even a public service-specific ban. There is a clear rift on the issue between the ECtHR and the HRC, which indicates that the question is much more open than a singular focus on the Strasbourg case law might suggest. In invoking the notorious European precedents, therefore, all that partisans of Bill 21 can do is make a basic point that banning the veil is not entirely beyond the pale from the point of view of one of the most respected systems of regional human rights protection. A duty of candour, however, would require them to simultaneously acknowledge the limited scope of these decisions, their considerable contentiousness generally, and the existence of a sizeable body of contrary case law by at least one other international human rights body of particular relevance to Canada.

II. Speaking of the Margin of Appreciation and Its Dialects

Let us, however, assume a best-case scenario for Bill 21, one in which the European case law is hypothetically applicable to Quebec by analogy. I also leave aside the crucial question, which is beyond the scope of this essay, of whether Quebec or Canada would, hypothetically, be the appropriate site to evaluate the margin of appreciation, except to note in passing that it would more plausibly be the latter given that it is Canada as such that is a party to the relevant international human rights instruments (and that the margin of appreciation would presumably be appreciated differently if the unit of reference were Canada rather than Quebec).[70] Let us assume, however, that Quebec is the appropriate unit of reference for the purposes of assessing the margin.

Surely, then, at least the European case law would stand a very good chance of buttressing the argument in favour of Bill 21? After all, as we have seen, the ECtHR has made it clear that protecting the religious neutrality of the state is definitely a legitimate aim in a democratic society and even agreed that bans may be a necessary and proportional measure to achieve that aim. In this section, I want to suggest that even in such a scenario—one that stacks the deck in favour of proponents of Bill 21—the European case law hardly provides the foolproof case for it that is hinted at.

It bears mentioning the obvious, namely that the ECtHR is not legislating European policy on the matter. Its decisions are generally authoritative and binding, but, even within the European system, only for the parties to any given case. Moreover, the case law on the veil emerged from countries that sought to forcefully limit its display based on a heightened sensitivity to the display of (certain) religious symbols in the public sphere. This is of course not surprising, but it can account for a certain outcome bias in the case law of the ECtHR depending on what lands on the court’’s docket. For a handful of cases where the Court found that the limitations were legal, there are a great many situations where no decision was forthcoming simply because there was no ban in place and therefore nothing to challenge. This further limits the existing decisions’ general precedential and jurisprudential value as themselves involving relatively exceptional bans, even within the confines of the European human rights system. At the risk of sounding trivial, all cases involving bans before the ECtHR emerged from the subset of states that, for a host of complex and quite specific reasons, sought to adopt such bans in the first place.

That in itself might be thought to be of limited interest except that, as it turns out, there is a deeper and more crucial reason why only the parties to a case are bound, and why the limited palette of states involved in religious symbol bans is relevant. That reason is that the European decisions were based, perhaps first and foremost, on what is known as the “margin of appreciation,” notably in evaluating limitations to rights (and, a contrario, the HRC’s much more ban-skeptic decisions can be understood as based on its rejection of the margin of appreciation). The margin of appreciation has become absolutely central to the case law of the European human rights system and indeed to that system’s self-understanding. It was originally based on a French development by the Conseil d’État but has since emerged in most civil law jurisdictions and is widely considered to go to the core of the European human rights system’s economy.[71] It is what explains how the ECtHR, starting from very similar premises as the HRC (with virtually identical provisions and largely comparable cases), has ended up in a very different place, at least in its decisions concerning France, Turkey, and Switzerland. In assessing whether a right has been properly limited, the margin of appreciation offers a kind of ongoing, country-specific compromise between the singular ambitions of the European Convention and the need to adapt to the local realities of its forty-seven member states.

Essentially, the margin of appreciation doctrine posits that states should have some latitude, when appreciating the necessity and proportionality of rights limitations, based on their national specificities and priorities. Perhaps unsurprisingly, the first decision of the Court setting out the margin of appreciation arose in the national security field. This was an area where the European Commission on Human Rights felt that what was involved was a “delicate problem of appreciating complex factors and of balancing conflicting considerations of the public interest” and so felt it should defer to “the Government’s appreciation.”[72] But what may have been initially a concession to the particular circumstances of national security by a court eager in its early years to not be seen to meddle too much with states, gradually became normalized and extended to a much broader range of governance issues.[73]

The foundation for the rule, then, is a form of subsidiarity and supranational judicial self-restraint. The Court “cannot disregard those legal and factual features which characterise the life of the society in the State which … has to answer for the measure in dispute.”[74] Indeed, the Court “cannot assume the role of the competent national authorities, for it would thereby lose sight of the subsidiary nature of the international machinery of collective enforcement established by the Convention.”[75] The Court went out of its way to bow to national specificities in the Handyside decision where it famously held that

it is not possible to find in the domestic law of the various Contracting States a uniform European conception of morals. The view taken by their respective laws … varies from time to time and from place to place … By reason of their direct and continuous contact with the vital forces of their countries, State authorities are in principle in a better position than the international judge to give an opinion on the exact content of these requirements as well as on the “necessity” of a “restriction” or “penalty” intended to meet them.[76]

This is of course quite appealing as a limiting and contextualizing factor in international human rights adjudication. The margin defers simultaneously to two realities. First, that countries are indeed quite different, that human rights are broad abstractions that need to be given meaning in particular circumstances, and that the Court is ill-suited to second guess complex arbitrages by governments or even domestic court decisions. If this is true in the European system, then it ought to be a fortiori true on a global level, despite the HRC’s reticence.[77] Second and closely related, that states, in addition to being different, also have distinct valid domestic reasons to opt for certain understandings of how rights should operate. These reasons are linked not only to their closer knowledge of local realities, but also to the specificity of their histories and legal systems and, of course, to democratic rule. Note that this is not a question of merely opposing arbitrary “political” preferences to “human rights”: international human rights law itself is deeply sympathetic to ideas about collective self-determination and thus to the ability of states to maintain a certain distinctiveness from each other through democratic governance, on the basis of their distinct traditions.[78]

As it happens, the margin has featured prominently in the European religious freedom case law, given that “[w]here questions concerning the relationship between State and religions are at stake, on which opinion in a democratic society may reasonably differ widely, the role of the national decision-making body must be given special importance.”[79] I bracket here the important question of whether the margin of appreciation is itself a legitimate human rights tool or the graveyard of human rights ambitions (the latter possibility made painfully obvious by, precisely, hijab case law, as some scholars have suggested).[80] I also leave aside that the HRC, in not embracing the margin despite presiding over a far more diverse international reality, has shown that, quite strikingly, two leading human rights regimes can rely on dramatically different devices and philosophies of interpretation.[81] These tensions suggest worthwhile and important debates.[82] But let us accept, for the sake of argument, that the margin’s status in international law is well entrenched (or at least that it is a useful metaphor to think about the global destiny of human rights) and that this is what one has to contend with if one seeks guidance in European legal developments for Quebec and Canada.

What I want to first point out is that, at the very least, those who invoke the European decisions without incorporating the margin of appreciation can be suspected of engaging in a deep fallacy. By focusing solely on the outcome, they ignore or omit that this outcome is, in fact, less important than the complex reasoning that riddles it with all kinds of tenuous assumptions and limitations. The ECtHR never said “banning the veil from the public sphere in this way is never a violation of human rights”; it merely said that it was not a violation of human rights in Turkey and in France, because of these countries’ characteristics, as a result of the margin of appreciation. The outcome, in other words, does not stand on its own, but is inseparable from the particular legal politics and legitimization strategies expressed in the European human rights project. Those who miss (or omit) that dimension risk fundamentally mistaking what is at stake.

It may well be, assuming the recognition of the European case law’s indebtedness to this sort of context-bound reasoning, that the margin still holds a certain promise as a way of defending Bill 21. Professors Fatin-Rouge Stefanini and Taillon, for example, make much of the fact that the European case law has proved very supportive of bans in the case of states with a distinctive tradition of secularism, whilst the report of Yvan Lamonde made much of, precisely, Quebec’s own distinctive tradition of secularism.[83] This might appeal to those for whom the debate on laïcité was always closely linked to an identitarian promotion of “les valeurs” of Quebec:[84] the idea that one can be part of a common, overarching human rights project yet differ markedly in how one understands and applies it domestically; the idea that the local has a certain pride of place over the international; and the idea that outside commentators and adjudicators should, as it were, know their place and exercise a certain prudence. If the French can have their laïcité prohibiting schoolgirls from wearing the hijab in the playground whilst the British can have turban-wearing police officers, both with the blessing of Strasbourg, then new vistas seem to open for a vision of a decentralized and pluralistic international order[85]—one that does not get in the way of what are imagined to be strongly felt majoritarian and culturally specific priorities in the name of imposing “human rights” from above. Certainly, if nothing else, the margin does stand for a certain flexibility in the application of human rights.

Having said that, incorporating margin of appreciation reasoning in the Canadian or Québécois context remains a fraught exercise intellectually. This is not just because the contours of the margin of appreciation, even in Europe, are heavily contested and evolving, and nowhere more so than on the divisive question of religious symbols in the public sphere.[86] The margin is an intellectual cottage industry unto itself, having given rise to more writing than any other concept in the European human rights edifice.[87] This is unsurprising given the difference it has often made, but one should be alert to how contentious it is. Its proper parameters are the object of constant practical and theoretical struggles and, more often than not, the central stake of major decisions involving (as they often do in the European context) not textbook violations of human rights (say, torture), but complex and contested issues of social regulation (freedom of the press and national security, end-of-life decisions, same-sex marriage and adoption, etc.).

In other words, the exercise can never be as simple as applying an uncontroversial margin of appreciation to Quebec. It is one thing to note the occasional permissiveness of the European regime; it is another to argue that Quebec can unproblematically avail itself of that permissiveness. This is especially important given the abundance of concerns that laws such as Bill 21 are textbook instances of majoritarian violations of minority rights. This fear is not allayed by the fact that the law was adopted—and may only have been saved constitutionally before the courts so far—through a highly contentious invocation of section 33. The margin of appreciation, for all its pluralistic dimensions, cannot be a simple blank cheque for any given society to engage in its political preferences on the basis that human rights mean whatever they are made to mean in that particular society.

As the ECtHR has put it frequently, including in religious garment cases, the margin “goes hand in hand with a European supervision embracing both the law and the decisions applying it.”[88] The margin of appreciation is thus also a specific discipline, however flexible, which imposes constraints on those who invoke it. It is not, in particular, a self-standing mode of evaluating rights compliance, but a variable to consider the appropriateness of specific limitations to rights. Specifically, it is by now accepted that the ECtHR has hesitated between one conception of the margin (call it the substantive conception) and another (call it procedural), arguably transitioning to the latter notably in its reasoning on religious symbols. In the following subsections, I examine how the Bill 21 argument would fare under both of these conceptions, explaining how they came about and how they might play out in the Canadian context.

A.  The Substantive Margin and the European “Consensus”

I refer to the “substantive” conception of the margin here as the more ambitious version of its use—one that aims to reconcile the values of the European Convention (the “European standard”) on the one hand and those of particular state parties on the other, through an extensive review of where domestic legal systems stand in relation to certain European human rights obligations. The Court has designed a whole conceptual apparatus to guide its reasoning in assessing the scope of the margin of appreciation in specific cases. In practice, the Court must find whether the limitation proposed in the context of a particular country is one where the margin of appreciation can have more or less incidence. This involves at least a cursory comparative analysis of the laws of Council of Europe member states to determine whether the area is one where there is consensus or dissensus. As a general rule, the Court has emphasized that

[w]here … there is no consensus within the member States of the Council of Europe either as to the relative importance of the interest at stake or as to how best to protect it, the margin will be wider. This is particularly so where the case raises complex issues and choices of social strategy … There will also usually be a wide margin accorded if the State is required to strike a balance between competing private and public interests or Convention rights.[89]

As it happens, within the European context itself, the ECtHR found that

it was … not possible to discern throughout Europe a uniform conception of the significance of religion in society and … the meaning or impact of the public expression of a religious belief would differ according to time and context. … [R]ules in this sphere would consequently vary from one country to another according to national traditions and the requirements imposed by the need to protect the rights and freedoms of others and to maintain public order.[90]

This certainly reinforces the argument of supporters of Bill 21. Is there consensus among members of the international community about the relative importance of freedom of religion, the place of wearing religious symbols within it, or the role of the state in regulating displays of religious symbols in public spaces or functions perceived as sensitive? This is even more unlikely on a global level than it is in Europe. There is certainly a consensus that freedom of religion should be protected, and that is typically understood to include expressing one’s religious beliefs, including through visible religious symbols, in the public sphere in ways that do not encroach on public order or the rights and freedoms of others. But this is just a general formula, and the devil is in the details. There is deep disagreement about the “necessity” part of the limitations test as to whether bans protect public order, enhance equality between the sexes, or encourage freedom of religion. This much emerges from a comparison of how various European countries have dealt with the issue, based on often strikingly different traditions of the relationship between the state, society, and religion. As a result, and “in case of doubt” as it were, it is clear that the margin of appreciation has been found to be a particularly apposite mechanism by the ECtHR in its hijab cases. The Court has acknowledged the many reasons that states would have to ban the veil and has not particularly problematized their motivations for doing so.

But there are at least two significant limitations to the Court finding for states in this way. First, the margin of appreciation works best where there is a real, incommensurable split between member states on an issue. This does not necessarily mean a 50/50 split, but it means that each side on a debate is fairly robustly represented among member states. It is least operative, by contrast, when a number of states have gradually been put into the minority, allowing the ECtHR to step in and enforce an emerging consensus. What was previously one option among others then increasingly starts looking like it is the marginal approach. Thus the margin, for better or for worse, typically acts as a very rough majoritarian system. The Court has never been so happy to enforce “European supervision” as when facing some unsuspecting laggard (e.g., the Isle of Man, which was the last territory in Europe to allow corporal punishment in schools by the 1980s),[91] and has been at its most coy when confronted with some deep societal rift between and indeed within European states (e.g., euthanasia).[92] Thus, disagreement about what rights mean may not be sufficient if one can argue that there is a clear trend in one or the other direction.

In that respect, it is at least interesting to visualize where Quebec lies globally. Is restricting the display of religious symbols in the way proposed an outlier’s position, is it somewhere in the middle, or is it completely orthodox? This would require more comparative study than is possible here, but it is worth noting that, whilst there has certainly been a trend in many countries toward public bans of the niqab specifically, there are very few states that ban religious symbols by agents of the public service in Europe and even globally.[93] In fact, as the ECtHR itself put it in one of the few cases involving a ban in the public service:

With regard to the margin of appreciation left to the State in the present case, the Court notes that a majority of the Council of Europe member States do not regulate the wearing of religious clothing or symbols in the workplace, including for civil servants … and that only five States (out of twenty-six), one of them France, have been identified as prohibiting completely the wearing of religious signs by civil servants.[94]

The distinct minority that does so is, relatively speaking, even smaller globally (assuming that it is ultimately the ICCPR that is the correct instrument) than it is at the European level. Again, adjusting for the distortion that the cases came from the few countries where individuals had reason to complain, the majority of states do not ban religious symbols for civil servants or, for that matter, for any other workers. This suggests that even though the European case law superficially defers to some bans in the civil service, this may not be as helpful to proponents of Bill 21 as appears at first. It is perhaps even directly and distinctly unhelpful in that it suggests that similarly situated states, even states that purportedly share Quebec’s broad secular sensibility, have not thought that this was a particularly pressing issue.[95]

Second, the argument in Strasbourg is not simply or principally that there is disagreement in Europe in general about the issue, leading to a finding that, henceforth, states have relatively wide discretion when it comes to religious symbols. Rather, states who emphasize that diversity have to show that the arguments for banning the veil specifically work for them; in other words, that they are genuinely, for any number of reasons specific to their circumstances, one of those states that ought to be able to impose such a ban. One could not simply invoke, for example, a general notion that wearing the hijab has occasionally been associated with Islamic fundamentalism (assuming that this argument has any merit) if that was not actually the case in one’s specific society. Moreover, one could not simply invoke European-wide dissensus to justify a country-specific position if one could not point to a certain legal or political tradition on the issue.

In cases where the ban was found not to be a violation, the ECtHR was careful to indicate that measures banning the veil were necessary in a particular context.[96] Although, for example, it has “approved strict implementation of the principles of secularism … and neutrality,” it has done so “where this involved a fundamental principle of the State, as in France.”[97] In other words, whilst it is true that the margin of appreciation generally recommends pluralism and flexibility, the argument must still be made that this is opportune in the particular case at stake.

This is not necessarily a huge hurdle, but it does point to a requirement of justifiability on a country basis as well as, more deeply, a sort of requirement of genuineness. It is one thing for France with its 200-year tradition of laïcité, long vigorously enforced primarily against the Catholic church, or for Turkey with its 100-year tradition of militant republicanism to invoke secularism in defence of a ban. One may of course disagree that secularism need lead along that path, but one cannot deny that, if any state in Europe is going to be able to make that point successfully, it is those established champions of secularism. At least no one can pretend that individuals there were taken entirely by surprise, or that secularism was suddenly brought up in a populist moment. The ECtHR, in fact, has shown itself willing to give quite different meanings to the idea of state neutrality in religious matters depending on particular states’ construction of it—a sign that legal and political traditions have some weight in framing how each country relates to rights.[98]

Indeed, deep political equilibria may be at stake that implicate human rights, but also connect to bigger issues: political stability, a sense of national identity, or the ability to democratically self-determine. For states in which a rigid notion of secularism has become part of a broader entrenched constitutional narrative, to be told by an international court that they cannot ban the veil is destined to be felt more intensely as an unwarranted imposition of a single human rights standard to a complex issue. By contrast, the pro-ban argument is at least given an “air of plausibility” in societies where it can be portrayed not only as grounded in political tradition, but as a continuing response to specific political or security challenges. In that respect, Turkey (pre-Erdogan) was able to effectively challenge the Strasbourg judges to second-guess its sovereign discretion when it came to the link between a student wearing the hijab at Istanbul University and Islamic fundamentalism in Anatolia.[99] Again, we have plenty of reasons to be wary of such arguments in terms of human rights—but they at least delineate factual grounds on which such policies can then be challenged. (For instance, whether there really is a link between hijab-wearing and fundamentalism.) More importantly, however, these arguments ground bans in the constraints that a specific context imposes on rights.

It would probably be an entirely different matter if the UK, with its (hard-won) tradition of respecting religious minorities, its multiculturalism, and its keen embrace of a wide range of manifestations of religious freedom in the public sphere were to suddenly reimagine itself as a hardline secular state that could not thrive except by banning persons from wearing religious symbols in the public service (even, perhaps, as it continued to adhere to the notion of Anglicanism as a state religion). The UK would, in such a configuration, not even be able to make the argument that it was merely defending some deeply felt political specificity. All the red flags would be raised and the ban would appear for what it probably would be: a complete turning of the country’s back to a political tradition that has long shaped British understandings of rights. The UK’s advocates in Strasbourg would have to work overtime to assert some pressing motive (presumably something dramatic like national security or an awakening to the threat that certain religious symbols pose for sexual equality) that did not suddenly betray British political and constitutional tradition. In short, it would be very difficult for the UK to act à contre-emploi as it were because it would proceed from a very fragile and suspicious basis in doing so. States’ traditions and circumstances can be permissive of certain interpretations of the scope of rights but they can also be limitative of other interpretations.

Note that none of this is meant to solidify the Turkish and French arguments in the absolute, but only in relative terms compared to states that cannot invoke a similarly long tradition of reservation about expressions of religiosity in the public sphere to which the Strasbourg Court might want, in its wisdom, prudence, or weakness, to defer. At times, the Court has been willing to recognize certain bans (notably of the burqa) even in countries such as Belgium, whose tradition of laïcité is considerably thinner than France or Turkey’s, but in ways that were quite specific to the current circumstances faced by those countries. At any rate, this is a remarkable result from the point of view of human rights: it means nothing less than the possibility that freedom of religion in the European human rights system is, in effect, both compatible with the banning of the veil (in certain societies) and not compatible with the banning of the veil (in other societies). It also means, perhaps even more intriguingly, that the European human rights system can be conceptualized, in part, as a system that will both capitulate before states’ strong assertions of constitutional and identity preferences, but that will also hold some states to their own self-proclaimed standards. We may feel as if this is a desultory ambition for human rights, but we should at least acknowledge the remarkable compromise it represents between sovereignty and internationalist ambition: the idea that states should not fall below their own human rights ideals.[100]

But where does this hypothetically leave Quebec, as a province (i) that occasionally models itself on the French approach to secularism, but is also part of a country that has been deeply influenced by the Anglo-American approach to multiculturalism; (ii) that has a tradition of secularism that is relatively entrenched, but is certainly more recent than France and Turkey’s; (iii) whose particular form of secularism, like France and Turkey’s, was traditionally understood to only target the once-dominant religious denomination but is now being turned against minority groups that seem much less of a threat to the province’s way of being; (iv) where until recently secularism was not commonly understood to require civil servants to desist from wearing religious symbols; and (v) which is part of a state which, by most accounts, would stand to be the relevant unit of analysis for the purposes of evaluating the margin?[101] These are no doubt complex issues whose detailed discussion is beyond the ambition of this essay, but they are very much the issues that one would have to deal with to make sense of what the margin of appreciation might mean, hypothetically, in this case.

Note however that in an interlocutory injunction appeal on Bill 21, Justice Mainville, despite positively recording the ECtHR’s position on a range of religious symbol bans, also felt compelled to mention that

[o]f course, there are significant differences between the constitutions of these European countries and the Canadian Constitution, as well as between European values and distinctly Canadian values. Given the particular characteristics of the Canadian constitutional framework and Canada’s history, both of which have long recognized our cultural and linguistic diversity, it may well be that Canadian courts will ultimately have quite a different approach to these issues than the European Court of Human Rights or European tribunals. The approach of the European Court of Human Rights may not be the one that Canadian courts should adopt and its limited vision of Islamic veiling may need to be nuanced, if not

      Among the questions that would need to be raised to evaluate the applicability of the margin in the case of Bill 21 are: of how recent vintage and how opportunistic is the turn to a particular concept of secularism requiring the non-wearing of religious symbols in the public service? To what extent was it activated in reaction to specific minorities rather than the plausible and linear development of an ingrained concept of secularism? How can it be rationally connected to the policy aims it claims to pursue? How does it relate to the province’s political traditions, including as they pertain to pluralism and human rights? And how might the invocation of the margin by Canada play out given that the federal government’s position on Bill 21 is very much at odds with Quebec’s?

Here, clearly the Révolution tranquille and its invention of a specific form of laïcité suggest that Quebec’s most recent orientation is not unprecedented and has roots in the reimagining of the province’s cultural identity since the 1960s.[103] Moreover, Quebec can invoke a range of arguments related to the equality of women that have, rightly or wrongly, also found an understanding ear in Strasbourg. At the same time, one must ask whether the current metamorphosis of the secularism debate, notably under the influence of French ideas about laïcité, really represents a continuation of long-existing trends or an abrupt inflection thereof. The initial reluctance to remove the crucifix hanging in the Quebec National Assembly has, for example, led to a suspicion that the version of secularism that began to take hold in the 1990s really targeted only minorities.[104] Finally, the fact that it is, if anything, Canada that has to answer for its compliance with international human rights obligations, even for legislative activity that falls within the competences of one of its provinces, means that the argument is only really indirectly about Quebec. This raises very significant difficulties that are beyond the scope of this essay.[105]

At any rate, what laïcité means is precisely the question, in a context where one thing European comparisons show us is the highly constructed, contingent, evolving, and often subtly misleading character of the term.[106] The point is that this question cannot be addressed in purely national and self-referential terms that international human rights bodies then ought to take for granted. Instead, it must be problematized from the point of view of international human rights law itself. Whether a ban can avail itself of the margin of appreciation hinges in part on whether one can distinguish between a simple assertion of democratic dominance (to which international human rights law has strong resistance) and the ongoing playing out of political and legal traditions (to which it is more understanding).

International human rights bodies, in their effort not to impose a rigid one-size-fits-all model to quite diverse societies whilst clinging to the idea of a common standard, have occasionally validated bans, but only when they emerged from societies whose practices of secularism had a certain pedigree and could be understood to require certain specific bans. In judgments involving Turkey or France, for example, much space is devoted to exploring the centrality and depth of their respective visions of secularism in those contexts. This helps distinguish from what might be merely opportunistic majoritarian turnabouts with no historical foundation.

It is beyond the scope of this essay to assess whether such an argument is available in Quebec, a society that has a certain tradition of secularism, but not necessarily laïcité in the dogmatic French sense,[107] and where, until the adoption of Bill 21, individuals (and not just Muslim women) could and did wear religious symbols in the civil service. Much attention was rightly devoted to this issue in reports commissioned by the Crown establishing convincingly, if a little one-sidedly, that Quebec’s particular brand of secularism was not a late-hour invention, but part of a tradition inaugurated by the Révolution tranquille, if not much earlier.[108]

Nonetheless, it is also clear that the question of religious symbols has led to a flourish of initiatives, mostly in the last two decades. It has led to a certain radicalization of secular discourse, particularly in relation to one immigrant community’s religious practices. Rather than merely reinforcing the argument for the ban’s historical pedigree, the frequent references in the Crown’s defence of Bill 21 to the fact that it is supported by a majority of “Québécois and québécoises”[109] may only crystallize rights concerns about plebeian appeals. It also begs for a contextualization of the evolution of Quebec in the last decades. The province has largely transitioned from a society defined by its defiant emancipation from clerical rule to a multicultural and multiconfessional environment where that threat has fundamentally receded but where new ones (Islamophobia, antisemitism, structural discrimination) have emerged and become much more central. One largely eluded question, for example, is whether secularism ought to be treated similarly in relation to the central, long-dominant religion (Catholicism) and in relation to minority religions. The point here is not to deny that legislatures may guide the evolution of rights in light of changing circumstances. Rather, one must be able to distinguish between the genuine search for contextualized human rights solutions to complex problems within a properly understood horizon of cultural specificities on the one hand, and plain majoritarian human rights violations on the other.

B.  The Procedural Turn in the Margin of Appreciation

It is not even clear, however, that the just described substantive approach to the margin of appreciation is still the reigning or the most useful one. Although the substantive approach remains influential, it has been argued that the ECtHR has since the early 2000s taken a “procedural turn” in evaluating whether to allow a state’s use of the margin.[110] This is no doubt partly a result of some states’ resistance toward the role of an increasingly forceful “European standard” that was seen as moving Strasbourg away from its underlying commitment to pluralism.

Instead of looking at whether states are outliers in relation to an emerging European consensus, the “process-based” approach of the Court increasingly concerns itself with the quality of the decision-making that led to impugned measures, both legislative and judicial.[111] The main thrust of the relevant cases is that the Court will review the seriousness of the national processes through which a given measure was adopted, including the extent to which rights concerns were raised and whether efforts were made to reasonably limit any negative incidence on such rights. This will lead to heightened scrutiny of parliamentary debates, including the degree to which public consultations were held and constitutional and international commitments were discussed in earnest.

As the ECtHR put it in the Animal Rights Defenders International v. United Kingdom case, where it sought to evaluate a blanket ban on a certain form of expression, “[t]he quality of the parliamentary and judicial review of the necessity of the measure is of particular importance in this respect.”[112] A contrario, in the Hirst v. United Kingdom (No. 2) case on the voting rights of prisoners, the Court noted that “there is no evidence that Parliament had ever sought to weigh the competing interests or to assess the proportionality of a blanket ban on the right of a convicted prisoner to vote.”[113] In other words, the UK was faulted not so much for being at odds with a substantive European consensus on this particular issue, as for basically “not taking rights seriously” by failing to even discuss the relevant dilemmas. Rather than evaluating the margin of appreciation on the basis of some “thick” supranational common standard, it is assessed by a renvoi to states’ own procedural care in limiting rights.

This evaluation extends to the implication of domestic courts in efforts at redressing ensuing human rights violations. Hence the ECtHR will assess, in turn, the degree to which such courts have taken into account European human rights law and jurisprudence, even in the process of asserting local specificities.[114] To the extent that they have not, they can be suspected of having merely engaged in a form of parochialism of which the European human rights system disapproves. To the extent that they have, they at least stand a better chance of being seen to have genuinely sought to reconcile the European rights imperative with local specificities.[115] Although its exact place in the jurisprudence of the ECtHR remains contested,[116] this procedural approach has been defended as felicitously combining attention to the substantive European consensus and parliamentary and judicial processes as manifestations of popular sovereignty.[117]

The procedural turn has arguably been further entrenched by the Copenhagen Declaration in 2018,[118] which was a response to increased hostility from key states (notably the UK) and even threats to pull out of the European Convention.[119] The Declaration reaffirms member states’ commitment to the European Convention, but also emphasizes the “shared responsibility” of the Court and states to implement the European Convention and solemnly affirms the importance of the principle of “subsidiarity.”[120] It notes in particular that

there may be a range of different but legitimate solutions which could each be compatible with the Convention depending on the context. This may be relevant when assessing the proportionality of measures restricting the exercise of rights or freedoms under the Convention. Where a balancing exercise has been undertaken at the national level in conformity with the criteria laid down in the Court’s jurisprudence, the Court has generally indicated that it will not substitute its own assessment for that of the domestic courts, unless there are strong reasons for doing so.[121]

On one level, the procedural turn can be seen as watering down the margin of appreciation (it is, in short, a procedural rather than substantive test). But it can also be understood as “giving teeth” to what is otherwise a cursory and somewhat arbitrary assessment.[122] It does suggest that, contra the substantive model’s inbuilt inertia, states under the procedural model are less likely to be held back by tradition and legal identity, let alone elusive supranational standards, even if changing their fundamental approach to a rights issue will require some serious democratic work. Deferring to political bodies in the international human rights context is often justified by reference to the legitimacy of democratic arrangements and the need to reinforce the ECtHR’s authority by making it more responsive to governmental priorities.[123] It is based on the presumptive legitimacy of the human rights outcomes produced by democracies in areas where, of course, the exact scope of human rights remains contested.[124]

This turn is hardly without its critics. For some, it involves a further capitulation of the Strasbourg organs before the whim of states, and a victory, in the long term, for democratic majorities over suffering minorities. It is also not clear why one would want the ECtHR to have democratic legitimacy in the first place given its role in protecting rights and traditional rights wariness about majoritarian rule.[125] Notwithstanding, there is no doubt that, again, this particular, more decentralized understanding of the margin of appreciation may appear at least at first glance as reinforcing the case for Bill 21. It nods heavily in the direction of parliamentary democracy, albeit one infused by human rights considerations, and thus heralds a redistribution of power away from international tribunals to domestic legislatures which seems very much in line with Quebec’s willingness to make this a question of democracy.[126]

Would Quebec’s legislative process leading to Bill 21 hypothetically satisfy this more minimalist procedural approach to the margin of appreciation? To be sure, Bill 21 was examined democratically, with fairly extensive consultations involved.[127] Rights concerns were not absent, and included a discourse that genuinely presented the banning of the hijab and the niqab in the public service in the language of, for example, sexual and gender equality. The fact that civil servants who wore religious symbols as of the time of passing of the law will be able to “grandfather” their practice goes some way to limit encroachment on at least their rights.[128] Late efforts to remove the régime de faveur, from which Catholicism benefited all the way to the hall of the National Assembly, have at least helped attenuate the whiff of blatant discrimination.[129]

But it may be a mischaracterization to say that the democratic debate was driven by a human rights agenda. The law was introduced first and foremost as part of a complex debate on integration, multiculturalism, and identity that foregrounded Quebec nationhood. Its adoption was oftentimes marked by xenophobic and Islamophobic sentiment. Bouchard and Taylor themselves spoke out about the way in which their report was used to introduce a broader ban than they had originally conceived, and backpeddadled on their earlier compromise suggestion that limiting a ban to persons in position of authority would be acceptable.[130] Not all groups, including notably Muslim women’s groups, were allowed to participate in the process. The legislative process by which it was ultimately adopted included resort to a bâillon (gag) which abruptly cut short the debate. It involved a highly opportunistic amendment of the Quebec Charter of Human Rights and Freedoms.[131] The very invocation of the section 33 notwithstanding clause suggests an alertness to the constitutional vulnerability of the law, and a willingness to legislatively bulldoze one’s way through that vulnerability, which is unlikely to be looked at favourably by international human rights bodies.[132] European states have been found to have exceeded their procedural margin of appreciation for less than that.[133]

One of the lessons of the procedural margin of appreciation is that process matters and that, from a human rights point of view, it cannot be reduced to complying with parliamentary rules. The proceduralism of the margin of appreciation’s nouvelle façon is, if anything, a thick proceduralism, one concerned with the synergies between democracy, the rule of law, and human rights. In the case of Bill 21, however, biases in the debates, the cavalier way in which human rights were addressed, and the emphasis on delivering results to certain voters all mean that Quebec could end up caught by its own words. It could, in short, be found claiming that religious symbols were banned in the name of human rights when, in earlier debates, it had insufficiently sought to justify the measure in human rights terms. The point is that one cannot retroactively impute a motivation to the adoption of a law that parliamentary debates show was not there to begin with.


To recapitulate, European human rights case law, while clearly not binding in Canada, is certainly an interesting proxy to think about what might be an acceptable position under international human rights law on the question of religious symbols, notably in the public service. As it happens, the ECtHR has certainly been understanding of limited bans of religious symbols in a series of cases. As such, that case law has already been imported in the Quebec debate by supporters of Bill 21.

By the same token, if the ECtHR’s case law plausibly serves to deprovincialize Bill 21, one must surely also be alert to the need to re-provincialize the ECtHR case law itself, and the trajectory of the likes of France or Turkey within it. Crucially, the ECtHR has condoned veil bans not in the abstract, but on the basis of the application of the margin of appreciation and in relation to certain countries that can avail themselves of it in particular circumstances. The “margin” is a very specific tool without which such outcomes are incomprehensible. Countries that were successful in beating allegations that bans infringed rights often came to the Court in Strasbourg with the full weight of arguments based on a relatively long term, sustained, and linear commitment to some form of rigid secularism that at least maximized the case for an extensive margin. The ECtHR has, rightly or wrongly, obliged in those specific cases, but their significance, properly assessed, remains limited.

Applying the margin of appreciation hypothetically to Quebec both supports some of the arguments for Bill 21 but also exposes others as precarious. It assumes that one fails to pay attention to the ICCPR, to which Canada is bound, whose supervisory body does not rely on the margin of appreciation,[134] and which has consistently found against bans of the hijab and niqab in particular. It also assumes that one takes Quebec as the framework of choice, whereas international human rights law would clearly command that one evaluate the policy with reference to Canada and Canadian norms. Here, the case for the margin of appreciation would be less strong given Quebec’s relative isolation on the issue within the federation and therefore the dilution, on the Canadian level, of a distinct traditition of laïcité. From an international human rights law point of view, the margin of appreciation clearly does not a priori extend to federated entities.[135]

European human rights law, despite being enthusiastically invoked by supporters of Bill 21, therefore appears as a bit of a faux ami, one whose appeal may prove misleading, and that may backfire. The suspicion is that Quebec is, in fact, when it comes to banning the wearing of religious symbols in the civil service, more isolated internationally than some pundits seem to suggest, although not as alone as others claim. This is not simply a factual but a normative point. Isolation in a pluralist international legal system is hardly a sin, but it is often a subtle sign that one lies on the outer limit of what is considered permissible.

Having brought these European precedents forcefully to bear in the conversation, sometimes to make them say something that they did not quite say, proponents of Bill 21 ought now to find it hard to ignore them merely because they, in fact, lead to a different result than hoped. The teachable moment may be that it is difficult to strenuously deny the ability to express a difference within (by wearing religious symbols in the public service) whilst ardently claiming a difference without (by claiming that Quebec should be allowed to go entirely its own separate way), except through a defense of national identity that one does not even pretend to cloak in human rights language. The fact that Bill 21 has already been found to violate section 23 of the Charter in relation to English school boards in Quebec suggests that we have not seen the last of such “pluralist contestations from below.”[136]

Both supporters and opponents of Bill 21 may, in the end, come to the conclusion that international human rights law simply does not do enough work for them. But perhaps the problem is that the source of international human rights’ significance was always confused. International law radicalizes, from outside as it were and as part of an exploration of what a global commonality of human rights fate entails, the dynamics of identity and distinctiveness that we have reason to think are otherwise very much at the centre of the debate around religious symbols. The question is not “what does freedom of religion mean?” in the absolute, but “what does it mean—internationally—in this or that country?” This suggests that international human rights law is both the guarantor of its own quite abstract standards but also, more crucially, of their proper articulation within particular societies and according to those societies’ specificities.


[1]     Bill 21, An Act respecting the laicity of the State, 1st Sess, 42nd Leg, Quebec, 2019 (assented to 16 June 2019), SQ c 12 [Bill 21].

[2]     See generally Anne Peters, “Supremacy Lost: International Law Meets Domestic Constitutional Law” (2009) 3:3 Vienna Online J on Intl Constitutional L 170.

[3]     See Martin Geoffroy, “La crise des accommodements raisonnables au Québec: de la jurisprudence à l’ingérence” (2008) 65 Canadian Studies 57 at 63.

[4]     See e.g. Maxime St-Hilaire, “Le Québec ne se distingue pas par sa laïcité de longue importation française”, Le Devoir (16 April 2018), online: <> [].

[5]     See e.g. Barbara Kay, “Barbara Kay: What the Anglo Media Misses About Quebec’s Religious Law”, National Post (2 April 2019), online: <> []; Jean-François Lisée, “The Inconvenient Truth About Quebec’s Secularism Law Trudeau Doesn’t Want to Face: It’s Popular”, CBC (9 September 2019), online: <> [].

[6]     For more on translation as a motif in human rights work, see e.g. Julia Ruth-Maria Wetzel, Human Rights in Transnational Business: Translating Human Rights Obligations into Compliance Procedures (Luzern: Springer, 2016).

[7]     See Bill 21, supra note 1.

[8]     See ibid. Despite its enactment, the legislation in question is still commonly referred to as Bill 21, and so will be throughout this essay.

[9]     See ibid, Chapters II–III.

[10]    For a very useful synthesis of the context of that reckoning and its implications, see Jean-François Gaudreault-DesBiens, “Religious Challenges to the Secularized Identity of an Insecure Polity: A Tentative Sociology of Québec’s ‘Reasonable Accommodation’ Debate” in Ralph Grillo et al, eds, Legal Practice and Cultural Diversity (Surrey, UK: Ashgate, 2009) 151.

[11]    See Québec, Consultation Commission on Accommodation Practices Related to Cultural Differences, Building the Future: A Time for Reconciliation (Quebec: Government of Québec, 2008) (Chairs: Gérard Bouchard & Charles Taylor).

[12]    See Bill 94, An Act to establish guidelines governing accommodation requests within the Administration and certain institutions, 1st Sess, 39th Leg, Quebec, 2010; Bill 60, Charter affirming the values of State secularism and religious neutrality and of equality between women and men, and providing a framework for accommodation requests, 1st Sess, 40th Leg, Quebec, 2013.

[13]    See Bill 62, An Act to foster adherence to State religious neutrality and, in particular, to provide a framework for requests for accommodations on religious grounds in certain bodies, 1st Sess, 41st Leg, Quebec, 2017 (assented to 18 October 2017), CQLR c R-26.2.01.

[14]    See Supriya Dwivedi, “Our National Silence on Bill 21”, The Walrus (18 October 2019), online: <> [].

[15]    See, in that respect, the joint challenge to the bill by the Canadian Civil Liberties Association and the National Council of Canadian Muslims (Hak c Procureure générale du Québec, 2019 QCCS 2989 (Application for judicial review and for an interim stay, Plaintiffs)).

[16]    See Hak c Procureure générale du Québec, 2019 QCCA 2145 [Hak QCCA]; Hak c Procureur général du Québec, 2021 QCCS 1466 [Hak 2021 QCCS] (note, however, that the court struck down the application of Bill 21 to English-language school boards and to members of Quebec’s legislature).

[17]    See e.g. “Des enseignantes musulmanes défendent le projet de loi sur la laïcité”, Radio-Canada (17 April 2019), online: <> [].

[18]    See e.g. Denis Hurtubise, “Le port de signes religieux: l’Europe à la rescousse”, HuffPost (20 June 2018), online: <> [].

[19]    Jean-François Lisée, “La Charte, les Québécois et le monde: Une mise au point” (30 September 2013), online (blog): Le blogue de Jean-François Lisée <> []. (“Quebec would be an outcast in the West and at the UN? This is simply untrue. Here is the truth, … And it is very different from what Mr Bouchard, Taylor and all those who try to make the Québécois feel guilty about having, in their home, a debate which is familiar in several other advanced democracies, in compliance with international norms.” [translated by author]).

[20]    “Signes religieux: Des avocats s’adressent à François Legault” (12 October 2018), online: Droit-Inc <> [] [emphasis added]. (“Many other countries across the world share this goal of secularism. Quebec would certainly not be the first country to adopt legal rules to restrict the wearing of religious symbols by civil servants in the exercise of their duties. On an international scale, such measures have been regularly subjected to judicial review before both domestic courts and the European Court of Human Rights, and have been found valid. … The European experience is living proof that the prohibition of religious symbols for civil servants is not a crime against humanity and can turn out to be a perfectly legitimate and necessary measure.” [translated by author]).

[21]    See e.g. Francis Vailles, “Pourquoi l’image des profs doit être laïque”, La Presse (10 May 2019), online: <> [].

[22]    Hak QCCA, supra note 16 at para 142.

[23]    See Marthe Fatin-Rouge Stefanini & Patrick Taillon, “Le droit d’exprimer des convictions par le port de signes religieux en Europe: une diversité d’approches nationales qui coexistent dans un système commun de protection des droits” in Quebec, Secrétariat à l’accès à l’information et à la réforme des institutions démocratiques, La laïcité: le choix du Québec. Regards pluridisciplinaires sur la Loi sur la laïcité de l’État (Quebec: Government of Quebec, 2021) [La laïcité: le choix du Québec] 529.

[24]    See Hak 2021 QCCS, supra note 16 at paras 218–31.

[25]    For a discussion of the invocation of section 33 from an in international perspective, see Frédéric Mégret, “Ban on Religious Symbols in the Public Service: Quebec’s Bill 21 in Global Pluralist Perspective” [manuscript under review, on file with the author] [Mégret, “Ban on Religious Symbols”].

[26]    See Armand de Mestral & Evan Fox-Decent, “Rethinking the Relationship Between International and Domestic Law” (2008) 53:4 McGill LJ 573; Gib van Ert, “Dubious Dualism: The Reception of International Law in Canada” (2010) 44:3 Val U L Rev 927; Stephen J Toope, “Inside and Out: The Stories of International Law and Domestic Law” (2001) 50 UNBLJ 11; René Provost, “Judging in Splendid Isolation” (2008) 56:1 Am J Comp L 125; Hugo Cyr & Armand de Mestral, “International Treaty-Making and Treaty Implementation” in Peter Oliver, Patrick Macklem & Nathalie Des Rosiers, eds, The Oxford Handbook of the Canadian Constitution, (New York: Oxford University Press, 2017) 595; Stéphane Beaulac, “New International Law and Its Doctrine Thinking Outside the ‘Westphalian Box’: Dualism, Legal Interpretation and the Contextual Argument” in Christoffer C Eriksen & Marius Emberland, eds, The New International Law: An Anthology (Leiden: Brill Publishers, 2010) 15.

[27]    See Yves Gaudemet, “La laïcité, forme française de la liberté religieuse” (2015) 148 Administration & Éducation 111 at 115–16.

[28]    For a more detailed exposition of this range of attitudes, see Fatin-Rouge Stefanini & Taillon, supra note 23.

[29]    See Marc-André Turcotte, “Présentation de l’ouvrage” in La laïcité: le choix du Québec, supra note 23, 1 at 28–29.

[30]    See e.g. Hurtubise, supra note 18.

[31]    See European Court of Human Rights, “The ECHR in 50 Questions” (2014) at 3–4, 6, online (pdf): European Court of Human Rights <> [].

[32]    Cf Christof Heyns, David Padilla & Leo Zwaak, “A Schematic Comparison of Regional Human Rights Systems: An Update” (2006) 4:3 Intl JHR 163.

[33]    See Hasan Aydin, “Headscarf (Hijab) Ban in Turkey: The Importance of Veiling” (2010) 6:1 J Multicultural Education 1 at 8–9.

[34]    See Leyla Şahin v Turkey [GC], No 44774/98, [2005] XI ECHR 173 at paras 122–23, 166, 44 EHRR 99 [Şahin].

[35]    SAS v France [GC], No 43835/11, [2014] III ECHR 341 at 382, 60 EHRR 245 [SAS].

[36]    See Dogru v France, No 27058/05, (4 December 2008) at paras 7–8, 78; Kervanci v France, No 31645/04 (4 December 2008) at paras 7–8, 78.

[37]    See Belcacemi and Oussar v Belgium, No 37798/13, (11 July 2017) at 21; Dakir v Belgium, No 4619/12, (11 July 2017) at 15 [Dakir].

[38]   See Ebrahimian v France, No 64846/11, [2015] VIII ECHR 99 at paras 3, 72 [Ebrahimian].

[39]    See e.g. Kurtulmuş v Turkey (dec), No 65500/01, [2006] II ECHR 297 at 310.

[40]    See Richard H Pildes, “Why Rights Are Not Trumps: Social Meanings, Expressive Harms, and Constitutionalism” (1998) 27:2 J Leg Stud 725.

[41]   See Canadian Charter of Rights and Freedoms, s 1, Part 1 of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (UK), 1982, c 11. See also R v Oakes, [1986] 1 SCR 103, 26 DLR (4th) 200.

[42]    Case “Relating to Certain Aspects of the Laws on the Use of Languages in Education in Belgium” v Belgium (Merits) (1968), VI ECHR (Ser A) at para B.5, 1 EHRR 252 [Belgian Linguistics Case (No 2)] [emphasis added].

[43]    See June Edmunds, “The Limits of Post-National Citizenship: European Muslims, Human Rights and the Hijab” (2012) 35:7 Ethnic & Racial Studies 1181 at 1190–95.

[44]    See e.g. Jannika Jahn, “The UK’s Potential Withdrawal from the European Convention on Human Rights: Just a Flash in the Pan or a Real Threat?” (17 December 2014), online (blog): Verfassungsblog <> [].

[45]    SAS, supra note 35 at para 121.

[46]    See generally Eva Brems, ed, The Experiences of Face Veil Wearers in Europe and the Law (Cambridge, UK: Cambridge University Press, 2014).

[47]    See Wendy Brown, “Civilizational Delusions: Secularism, Tolerance, Equality” (2012) 15:2 Theory & Event.

[48]    See e.g. Şahin, supra note 34.

[49]    For a defence of laïcité as the regime most conducive to ensuring freedoms, see Henri Pena-Ruiz, “Culture, cultures, et laïcité” (2006) 1259 Hommes & Migrations 6.

[50]    See Şahin, supra note 34 at para 165; Dakir, supra note 37 at para 65.

[51]    See e.g. Dahlab v Switzerland (dec), No 42393/98, [2001] V ECHR 447 at 463 [Dahlab]; Şahin, supra note 34 at para 111.

[52]    See Rob Lamb, “When Human Rights Have Gone Too Far: Religious Tradition and Equality in Lautsi v. Italy” (2011) 36:3 NCJ Intl L & Com Reg 751.

[53]    Council of State, Legislative Section, Opinion No 44.521/AG (20 May 2008) (Belgium), cited in Erica Howard, Religious Clothing and Symbols in Employment: A Legal Analysis of the Situation in the EU Member States (Brussels: European Commission, 2017)
at 83.

[54]   Ibid.

[55]    See Bundesverfassungsgericht (BVerfG) [Federal Constitutional Court], 24 September 2003, 2 BvR 1436/02 (Germany) at para 30.

[56]    See e.g. Kati Nieminen, “Eroding the Protection Against Discrimination: The Procedural and De-contextualized Approach to S.A.S. v France” (2019) 19:2 Intl J Discrimination & L 69.

[57]    See e.g. Stéphanie Hennette Vauchez & Vincent Valentin, L’affaire Baby Loup ou la Nouvelle Laïcité (Issy-les-Moulineaux: LGDJ, Lextenso, 2014).

[58]    See Şahin, supra note 34 at para 12, Tulkens J, dissenting.

[59]    See Frédéric Mégret, “The Apology of Utopia; Some Thoughts on Koskenniemian Themes, With Particular Emphasis on Massively Institutionalized International Human Rights Law” (2013) 27:2 Temp Intl & Comp LJ 455 at 490.

[60]    George Letsas, “The ECHR as a Living Instrument: Its Meaning and Legitimacy” in Andreas Føllesdal, Birgit Peters & Geir Ulfstein, eds, Constituting Europe: The European Court of Human Rights in a National, European and Global Context (Cambridge, UK: Cambridge University Press, 2013) 106 at 108.

[61]    See e.g. “France: Face-Veil Ruling Undermines Rights” (3 July 2014), online: Human Rights Watch <> [].

[62]    See e.g. Baljit Kooner, “The Veil of Ignorance: A Critical Analysis of the French Ban on Religious Symbols in the Context of the Application of Article 9 of the ECHR” (2008) 12:2 Mountbatten J Leg Stud 23.

[63]    See Dahlab, supra note 51; Şahin, supra note 34.

[64]    As distinct from courts themselves more or less opportunistically bringing up the “rights of others” as a ground to better justify a state’s decision to restrict freedom of religion. It is worth noting, however, that such challenges from majorities may arise in places that have very strong popular views about a certain concept of state neutrality. Notably, see e.g. Grant v Canada (AG) (1995), 125 DLR (4th) 556, 31 CRR (2d) 370, in which a coalition of citizens, including police veterans, challenged the reasonable accommodation of turban-wearing Sikh RCMP officers, and Lautsi and Others v Italy [GC], No 30814/06, [2011] III ECHR 61, 54 EHRR 60, which challenged the presence of a crucifix in an Italian school.

[65]    See La laïcité: le choix du Québec, supra note 23.

[66]    See Hudoyberganova v. Uzbekistan, Communication No 931/2000, UNHRC, 82nd Sess, UN Doc CCPR/C/82/D/931/2000 (2004) at paras 1, 3. [Hudoyberganova].

[67]    Ibid at para 6.2.

[68]    See Seyma Türkan v. Turkey, Communication No 2274/2013, UNHRC, UN Doc CCPR/C/123/D/2274/2013/Rev.1 (2018) at para 7.8.

[69]    See Hebbadj v France, Communication No 2807/2016, UNHRC, UN Doc CCPR/C/123/D/2807/2016 (2018) at paras 7.17 to 8; Yaker v. France, Communication No. 2747/2016, UNHRC, UN Doc CCPR/C/123/D/2747/2016 (2018) at paras 8.12–9.

[70]    For a detailed exploration of this particular issue, see Mégret, “Ban on Religious Symbols”, supra note 25.

[71]    See Howard Charles Yourow, The Margin of Appreciation Doctrine in the Dynamics of European Human Rights Jurisprudence (The Hague: Kluwer Law International, 1996) at 14–15.

[72]    Lawless v Ireland (1961), 1 ECHR (Ser B) at 408.

[73]    See Mikael Rask Madsen, “‘La Cour qui venait du froid’: Les droits de l’homme dans la genèse de l’Europe d’après guerre” (2005) 26 Critique Intl 133 at 144.

[74]    Belgian Linguistics Case (No 2), supra note 42 at para B.10.

[75]    Ibid.

[76]    Handyside v United Kingdom (1976), 24 ECHR (Ser A) 5 at para 48, 1 EHRR 737.

[77]    See Dominic McGoldrick, “A Defence of the Margin of Appreciation and an Argument for its Application by the Human Rights Committee” (2016) 65:1 ICLQ 21.

[78]    See Frédéric Mégret, “Having It Both Ways: International Human Rights Law Cannot Both Be in Decline and Be (That) Problematic for International Law” (2018) 96 Texas L Rev Online 114 at 131.

[79]    Şahin, supra note 34 at para 109.

[80]    See Frédéric Mégret, “Traditions of Human Rights: Who Needs Universal Human Rights?” (7 October 2019), online (blog): Centre for Human Rights & Legal Pluralism <> []. See also Eyal Benvenisti, “Margin of Appreciation, Consensus, and Universal Standards” (1998) 31 NYUJ Intl L & Pol 843.

[81]    Although the HRC does not officially condone the margin of appreciation, this has long been criticized as unrealistic (see Yuval Shany, “All Roads Lead to Strasbourg?: Application of the Margin of Appreciation Doctrine by the European Court of Human Rights and the UN Human Rights Committee” (2018) 9:2 J Intl Dispute Settlement 180 at 189­–90). And in cases involving the veil, the HRC has noted that the ability of the state to regulate it arises “duly taking into account the specifics of the context” which includes the conditions of the country (see Hudoyberganova, supra note 66 at para 6.2).

[82]    For a more detailed exploration, see Mégret, “Ban on Religious Symbols”, supra note 25.

[83]    See Yvan Lamonde, “L’histoire de la laïcité au Québec: l’établissement démocratique de la primauté du civil sur le religieux” in La laïcité: le choix du Québec, supra note 23, 175.

[84]    See Yannick Dufresnes & Gilles Gagné, “La religiosité et la question des ‘valeurs québécoises’” in La laïcité: le choix du Québec, supra note 23, 209.

[85]    See Mireille Delmas-Marty & Marie-Laure Izorche, “Marge nationale d’appréciation et internationalisation du droit: réflexions sur la validité formelle d’un droit commun pluraliste” (2001) 46:4 McGill LJ 923.

[86]    See Asim Jusic, “Damned If It Doesn’t and Damned If It Does: The European Court’s Margin of Appreciation and the Mobilizations Around Religious Symbols” (2018) 39:3 U Pa J Intl L 561 at 565–66.

[87]    See e.g. Andrew Legg, The Margin of Appreciation in International Human Rights Law: Deference and Proportionality (Oxford, UK: Oxford University Press, 2012); Yutaka Arai-Takahashi, The Margin of Appreciation Doctrine and the Principle of Proportionality in the Jurisprudence of the ECHR (Antwerp: Intersentia, 2002); Steven Greer, The Margin of Appreciation: Interpretation and Discretion Under the European Convention on Human Rights (Strasbourg: Council of Europe, 2000).

[88]    Şahin, supra note 34 at para 110.

[89]    Dickson v United Kingdom [GC], No 44362/04, [2007] V ECHR 99 at para 78, 46 EHRR 41.

[90]    SAS, supra note 35 at para 130.

[91]    See Tyrer v United Kingdom (1978), 26 ECHR (Ser A) 2, 2 EHRR 1.

[92]    See Pretty v United Kingdom, No 2346/02, [2002] III ECHR 155, 35 EHRR 1. It is worth noting, however, that the Court is not absolutely consistent with this and that states have occasionally been allowed to get away with practices that seemed very much on the margin of the European consensus (see especially A, B and C v Ireland [GC], No 25579/05, [2010] VI ECHR 185, 53 EHRR 429).

[93]    In the EU context, for example, it has been pointed out that, “[t]here are … not many Member States which have legislation prohibiting the wearing of some or all religious symbols in employment” (see Howard, supra note 53 at 88).

[94]    Ebrahimian, supra note 38 at para 65.

[95]    For similar reasoning in the context of a proclamation of national emergency, see A v Secretary of State for the Home Department, [2004] UKHL 56 at para 96.

[96]    See e.g. SAS, supra note 35 at paras 122, 151–56; Şahin, supra note 34 at paras 118–21.

[97]    Ebrahimian, supra note 38 at para 67.

[98]    See Raffaella Nigro, “The Margin of Appreciation Doctrine and the Case-Law of the European Court of Human Rights on the Islamic Veil” (2010) 11:4 Human Rights Rev 531 at 545–46.

[99]    See Şahin, supra note 34 at para 35.

[100] Indeed, often a clue to human rights violations is that states violate their own constitutional norms, hence the relevance of the insertion of “laïcité” in the Quebec Charter (see Charter of Human Rights and Freedoms, CQLR c C-12, Preamble [Quebec Charter]).

[101] See generally Pierre Hurteau, L’avenir de la laïcité au Québec: Pluralisme religieux et espace public (Paris: L’Harmattan, 2015) (on religion and multiculturalism in Quebec).

[102] Hak QCCA, supra note 16 at para 144.

[103] See Dufresne & Gagné, supra note 84; Lamonde, supra note 83.

[104]   See e.g. Jérôme Lussier, “Crucifix, incohérence et banalité”, L’actualité (28 January 2014), online: <> [].

[105] For more on these difficulties, see Mégret, “Ban on Religious Symbols”, supra note 25.

[106] See e.g Hugues Dumont & Xavier Delgrange, “Le principe de pluralisme face à la question du voile islamique en Belgique” (2008) 68 Dr et soc 75 at 78.

[107] See St-Hilaire, supra note 4.

[108] See Dufresnes & Gagné, supra note 84; Lamonde, supra note 83.

[109] See Turcotte, supra note 29 at 26–27.

[110] See Janneke Gerards & Eva Brems, eds, Procedural Review in European Fundamental Rights Cases (Cambridge, UK: Cambridge University Press, 2017).

[111] See Patricia Popelier & Catherine Van de Heyning, “Subsidiarity Post-Brighton: Procedural Rationality as Answer?” (2017) 30:1 Leiden J Intl L 5 at 9–10.

[112] Animal Rights Defenders International v United Kingdom [GC], No 48876/08, [2013] II ECHR 203 at para 108, 34 BHRC 137.

[113] Hirst v United Kingdom (No 2) [GC], No 74025/01, [2005] IX ECHR 187 at para 79, 42 EHRR 849 [Hirst].

[114] See e.g. Von Hannover v Germany (No 2) [GC], No 40660/08 & No 60641/08, [2012] I ECHR 351 at paras 114–15, 125, 32 BHRC 527.

[115] See Paul Anthony McDermott & Mark William Murphy, “No Revolution: The Impact of the ECHR Act 2003 on Irish Criminal Law” (2008) 30 Dublin ULJ 1 at 3.

[116] See Matthew Saul, “Structuring Evaluations of Parliamentary Processes by the European Court of Human Rights” (2016) 20:8 Intl JHR 1077 at 1090.

[117] See Thomas Kleinlein, “Consensus and Contestability: The ECtHR and the Combined Potential of European Consensus and Procedural Rationality Control” (2017) 28:3 Eur J Intl L 871 at 873.

[118] See Danish Chairmanship of the Committee of Ministers of the Council of Europe, “Copenhagen Declaration”, (April 2018), online (pdf): European Court of Human Rights <> [] [Copenhagen Declaration].

[119] See Ed Bates, “Principled Criticism and a Warning from the ‘UK’ to the ECtHR?” in Marten Breuer, ed, Principled Resistance to ECtHR Judgments — A New Paradigm? (Berlin, Springer, 2019) 193.

[120] Copenhagen Declaration, supra note 118 at paras 1, 6–7.

[121] Ibid at para 28.

[122] See Patricia Popelier & Catherine Van De Heyning, “Procedural Rationality: Giving Teeth to the Proportionality Analysis” (2013) 9:2 European Constitutional L Rev 230 at 230, 232.

[123] See Thomas Kleinlein, “The Procedural Approach of the European Court of Human Rights: Between Subsidiarity and Dynamic Evolution” (2019) 68:1 ICLQ 91 at 109–10.

[124] See Patricia Popelier, “The Court as Regulatory Watchdog: The Procedural Approach in the Case Law of the European Court of Human Rights” in Patricia Popelier, Armen Mazmanyan & Werner Vandenbruwaene, eds, The Role of Constitutional Courts in Multilevel Governance (Cambridge, UK: Intersentia, 2013) 249.

[125] See AIRE Centre et al, “Joint NGO Response to the Draft Copenhagen Declaration” (13 February 2018) at 4–8, online (pdf): Amnesty International <> [].

[126] See Robert Spano, “The Future of the European Court of Human Rights: Subsidiarity, Process-Based Review and the Rule of Law” (2018) 18:3 Human Rights L Rev 473.

[127] See “Consultations particulières et auditions publiques sur le projet de loi n° 21, Loi sur la laïcité de l’État” (May 2019), online: Assemblée nationale du Québec <> [].

[128] This is the so-called “clause d’antériorité” (see Bill 21, supra note 1, art 27).

[129]  See Quebec, National Assembly, Journal des débats, 42-1, No 26 (28 March 2019) at 1854–55 (motion to mandate the Office of the National Assembly to remove the crucifix from the National Assembly Chamber).

[130] See Philip Authier, “Bill 21 Hearings: ‘We Were Very Naive’ About Impact of Report, Charles Taylor Says”, Montreal Gazette (7 May 2019), online: <> []; Philip Authier, “Bill 21 Feeds Intolerance, Gérard Bouchard Tells Hearings”, Montreal Gazette (8 May 2019), online: <> [].

[131] See Quebec Charter, supra note 100 (“Whereas the Québec nation considers State laicity to be of fundamental importance” at Preamble).

[132] For a more detailed discussion, see Mégret, “Ban on Religious Symbols”, supra note 25.

[133] See Hirst, supra note 113 at para 79 (in that case, Parliament was faulted for not having strictly debated the proportionality of a ban on voting rights of prisoners, but it was quite clear that the democratic process had otherwise not been hindered in any way).

[134] Although, in truth, that remains quite contested (see McGoldrick, supra note 77).

[135] For an exploration of that as yet unorthodox argument, see Mégret, “Ban on Religious Symbols”, supra note 25.

[136] See ibid.

Contract Variation and Changed Expectations in this issue

related content