Eroding the protection against discrimination: The procedural and de-contextualized approach to S.A.S. v France

Abstract

Can human rights law adequately address implicit modes of racism and gender discrimination? This question is discussed in this article through the analysis of the European Court of Human Rights case S.A.S. v. France (2014) concerning the ban on the Islamic full-face veil. The so-called ‘headscarf cases’ have been thoroughly discussed by many scholars, yet they seem to offer an endless source of different points of view. Departing from the previous discussion on the headscarf and full-face veil cases, which have largely concentrated on the questions of personal autonomy, identity and subjectivity, this article approaches S.A.S. v. France from the point of view of discrimination. It is suggested that the Court’s procedural and de-contextualized approach to rights results in eroding the protection against discrimination. Procedural approach refers to the Court’s tendency to emphasize procedural aspects of the Convention rights and not to engage sufficiently with substantive analysis. The de-contextual approach to rights on the other hand refers to lack of sensitivity to empirical information concerning the facts of the case at hand. Together the procedural and de-contextual approaches inadvertently erode the protection against discrimination of vulnerable groups, such as Muslim immigrant women.

Introduction

Can human rights law adequately address implicit modes of racism and gender discrimination? This question is discussed in this article through the analysis of the European Court of Human Rights (ECtHR) case S.A.S. v. France ¹ concerning the law prohibiting for anyone to conceal their face in public places (the so-called full-face veil ban). The article identifies weaknesses in the Court’s jurisprudence concerning discrimination and suggests that despite the Court having announced its commitment to tackle particularly discrimination based on gender and ethnicity, there are multiple factors that impede the Court from addressing discrimination to its full extent (see also Arnardóttir, 2007, p. 38; Arnardóttir, 2014; Möschel, 2012, p. 492).

The headscarf and full-face veil cases in the ECtHR have been thoroughly discussed by legal scholars (see, e.g. Bhandar, 2009; Chaib and Peroni 2014; Evans, 2006; Jackson and Gozdecka, 2011; Marshall, 2008; Scott, 2010), yet they seem to offer an endless source of different points of view. The issue is far from exhausted, as indicated by Lachiri v. Belgium, ² Dakir v. Belgium, ³ Belcacemi and Oussar v. Belgium, ⁴ Hamidović v. Bosnia and Herzegovina ⁵ and the recent cases of Achbita ⁶ and Bougnaoui ⁷ in the Court of Justice of the European Union, as well as the views recently adopted by the Human Rights Committee. ⁸ In the jurisprudence of the ECtHR, S.A.S. v. France is the case the Court refers to in its’ more recent rulings, and therefore remains the most central of the cases concerning the full-face veil and the headscarf. Furthermore, S.A.S. illustrates that the logic of the Court’s argumentation concerning religious symbols cannot be entirely predicted by its previous case law, as the Court’s focus shifts case by case (see Nieminen 2015). While the cases, and the issue of Islamic clothing and religious symbols in general, have raised a lot of attention, much of the legal discussion has centred on the questions of personal autonomy, identity and subjectivity and their relationship to freedom of religion on the one hand, and the competing interests of the society on the other hand. Departing from this approach, this article contemplates the full-face veil ban from the perspective of discrimination based on religion, gender and ethnicity and suggests that the prohibition of discrimination (Article 14 of the European Convention on Human Rights (ECHR)) has unused potential despite often being secondary in the Court’s jurisprudence.

In S.A.S., the applicant argued that the ban, ‘which undoubtedly targeted the burqa, generated discrimination in breach of Article 14 of the Convention on grounds of sex, religion and ethnic origin, to the detriment of Muslim women who (…) wore the full-face veil’ (S.A.S. v. France, para. 80). ⁹ The Court, however, found that the ban did not violate any of the freedoms enshrined in the ECHR. In this article, it is argued that the potentially discriminatory aspects of the full-face veil ban being under-recognized result from the Court’s procedural and de-contextualized approach to rights.

The procedural approach refers to the Court’s tendency to emphasize the procedural aspects of the Convention rights and its willingness to accept the government’s assertions at face value without scrutiny (see also De Jong and Van Rijn van Alkemade, 2012). The procedural approach consists of prioritizing the doctrine of margin of appreciation over substantial analysis of the case and diluting the necessity assessment despite insisting on ‘careful examination’ of necessity (S.A.S. v. France, para. 122). It is suggested that rather than genuinely engaging with the necessity test, the Court draws its decisive arguments in S.A.S. from the discourse of militant democracy (see Macklem, 2012).

The de-contextualized approach approach to rights on the other hand refers to lack of sensitivity to contextual, empirical information concerning the ban, such as its drafting history and the way the ban is applied in France. These characteristics of the Court’s approach inadvertently circumvent the scrutiny of potential discrimination. In conclusion, it is argued that the Court’s procedural and de-contextualized approach leads to weakening of the prohibition of discrimination and particularly of the protection of vulnerable groups, such as women who belong to ethnic and religious minorities. ¹⁰ Before moving on to the analysis, the case of S.A.S. is briefly revised in the next section.

S.A.S. v. France

The applicant in S.A.S. was a French national born in Pakistan and belonged to a Sunni tradition of Islam. In her submission, she described herself as a devout Muslim who wore the burqa or niqab ‘in accordance with her religious faith, culture and personal convictions’ (para. 11). She did not, however, cover her face systematically, and she did not keep the niqab/burqa on occasions such as visiting the doctor, security checks, at the bank or in the airport. She challenged the French law of 2011 prohibiting for anyone to conceal their face in public places, arguing that it violated, among other things, her freedom of religion. The French Government, on the other hand, argued that the ban pursued the legitimate aim of protecting the rights of others by ensuring the respect for ‘minimum set of values of an open and democratic society’ (para. 114).

The Court’s legal assessment can be generally divided in two phases: first, the Court determines whether there has been an interference or a limitation of a freedom enshrined in the ECHR, and second, finding an interference, assesses whether there is a legitimate reason for it. The legitimacy assessment consists of determining whether the interference pursues a legitimate aim, whether it is prescribed by law, and often most importantly, whether it is proportionate and ‘necessary in a democratic society’. This necessity test was the crucial phase of the Court’s assessment in S.A.S. The Court found that the ban indeed limited the applicant’s freedom of religion (Article 9) and the right to private life (Article 8). The Court was not persuaded by the Government’s argumentation that the ban aimed at promoting gender equality and human dignity, but accepted that what the Government had described as ‘respect for the minimum requirements of life in society’ or of ‘living together’ could be linked to the aim of the ‘protection of the rights and freedoms of others’ – an aim which could be affirmed as legitimate by the Court. The Court concluded that it was able to accept that the French Government regarded the full-face veil ‘a barrier raised against others’ and that it breached the ‘right of others to live in a space of socialisation which made living together easier’ (para. 122). Finally, the Court came to the conclusion that the ban was both necessary and proportionate to the aim pursued taking into consideration the margin of appreciation granted for the states to interpret the Convention Articles in the context of their respective circumstances, and thus found no violation of Articles 8 or 9.

Likewise, the Court concluded that the ban was not discriminatory. Article 14 of the Convention prohibits both direct and indirect forms of discrimination in conjunction with the substantive rights guaranteed in the Convention, unless there is an ‘objective and reasonable’ justification for differential treatment. While the question of proof is often crucial in discrimination cases, this was not the case in S.A.S. as the Court readily accepted that ‘there is no doubt that the ban has a significant negative impact on the situation of women who, like the applicant, have chosen to wear the full-face veil’ and that ‘the impugned ban mainly affects Muslim women’ (paras 146, 151, 161). However, the Court found that the measure had an objective and reasonable justification.

In the following sections, the Court’s procedural and de-contextualized approach to discrimination in S.A.S. are discussed in detail. The second section addresses the proceduralization of the necessity and proportionality test as well as the relationship between Article 14 and the margin of appreciation doctrine. The third section discusses the Court’s de-contextualized approach to discrimination, namely the way the Court disregards contextual information indicating that the full-face veil ban reflects rising Islamophobia and the racialization of Muslims in Europe. In addition, the potential and the limitations of intersectionality to overcome the challenges identified in this article are discussed in the fourth section. The fifth section concludes and summarizes the ways in which the procedural and de-contextualized approach erodes protection against discrimination, and how the observations made in this article are relevant beyond the so-called head scarf and full-face veil cases.

The procedural approach to discrimination

Necessity, proportionality and the discourse of militant democracy

After accepting that the full-face veil ban aimed at protecting ‘the rights of others to live in a space of socialisation which made living together easier’, the Court acknowledged that the vagueness of the notion of ‘living together’ resulted a risk of abuse and stated that it ‘must engage in a careful examination of the necessity of the impugned limitation’ (S.A.S. v. France, para. 122). However, precisely due to the flexibility of ‘living together’, assessing its necessity in relation to the full-face veil ban was an impossible task to begin with. It is argued in the following that the Court did not genuinely engage with necessity and proportionality assessment and instead adopted a procedural approach (see also Ferri, 2017a).

An important part of the necessity test is proportionality. According to Arai-Takahashi (2002, p. 189), the proportionality test has two aspects: first, the Court weighs the rights of individual applicants against the general interests of the public; second, proportionality assessment involves evaluating whether there is a reasonable relationship between the means (such as the full-face veil ban) and the aims (such as preserving the space of living together). In its preceding headscarf cases, the Court has assessed, in varying degrees, the proportionality of the headscarf ban in the latter sense. In some of the cases, it is possible to see, if not agree with, a relationship between the ban and the aim of preserving a neutral educational setting and/or preventing ‘proselytising effect’. ¹¹ In cases concerning other religious symbols, an attempt to assess the relationship between the ends and means can likewise be detected. For example, in Eweida and Chaplin v. The United Kingdom, ¹² the Court found no causality between banning the wearing of the Christian cross and the aim of preserving a neutral company image on the one hand but found that a similar ban was necessary for the protection of health and safety on a hospital ward on the other hand. In these cases, even if briefly and inconsistently, the relationship between the ban and the aim was considered. However, in the case of S.A.S., the Court did not discuss the ways in which the blanket ban of full-face veil in public places potentially contributes to preserving the conditions for ‘living together’.

The alleged aim of the ban in S.A.S. is conceptualized in such a broad way that evaluating the means to achieve it is practically impossible – how could the Court go about assessing whether the full-face veil ban is necessary for living together? The Court’s application of the test was thus more of an empty gesture. Instead of genuinely engaging with the question of necessity, the Court accepted the Government’s assertion at face value stating that ‘the State may find it essential to give particular weight (…) to interaction between individuals and may consider this to be adversely affected by the fact that some conceal their faces in public places’ (para. 141). This statement was followed by the Court listing consequences of the ban in order to weigh the individual rights at stake against the competing public interest. On the one hand, the Court acknowledged that since there are only a small number of women who wish to wear the full-face veil, the ban may seem excessive; that it is ‘understandable that the women concerned may perceive the ban as a threat to their identity;’ that a large number of actors found the ban disproportionate. On the other hand, the Court found that while the scope of the ban was broad, it did ‘not affect the freedom to wear in public any garment (…) which does not have the effect of concealing the face’, and that the sanctions are ‘among the lightest that could be envisaged’ (paras 143–154).

The Court simply listed the different points of view mentioned earlier without explicitly considering their weight in the case. Instead what decided the matter in the end was an argument that echoes militant democracy – the idea that democracy should be protected by pre-emptively restricting the very rights and freedoms on which democracy is founded (Loewenstein, 1937; see also Macklem, 2012). After briefly recognizing that the full-face veil in fact contributes to pluralism, the Court accepted that the Government sought to protect the interaction between individuals, and thus, the essential requirements for pluralism, tolerance and broadmindedness ¹³ :

As the applicant pointed out, by prohibiting everyone from wearing clothing designed to conceal the face in public places, the respondent State has to a certain extent restricted the reach of pluralism, since the ban prevents certain women from expressing their personality and their beliefs by wearing the full-face veil in public. However, for their part, the Government indicated that it was a question of responding to a practice that the State deemed incompatible, in French society, with the ground rules of social communication and more broadly the requirements of ‘living together’. From that perspective, the respondent State is seeking to protect a principle of interaction between individuals, which in its view is essential for the expression not only of pluralism, but also of tolerance and broadmindedness without which there is no democratic society (…). It can thus be said that the question whether or not it should be permitted to wear the full-face veil in public places constitutes a choice of society. (para. 153, emphasis added)

In other words, the Government restricted pluralism in order to protect pluralism – logic that is familiar from militant democracy. The case of Refah Partisi (The Welfare Party) and Others v. Turkey ¹⁴ has been earmarked as the case which expanded the Court’s use of the logic of militant democracy, as the Court accepted for the first time that democracy’s self-defence could be used to dissolve a political party which was not explicitly opposed to democracy. This logic was further expanded in Leyla Şahin v. Turkey, ¹⁵ where the Court applied it to the Islamic headscarf. In Leyla Şahin, the Court, referring back to Refah, considered it established that challenging the principle of secularity could be equated with challenging democracy in Turkey. Thus, assuming that the Islamic headscarf signified a challenge to secularism, the Turkish Government was acting within its margin of appreciation when banning the wearing of the headscarf in universities (see Invernizzi Accetti and Zuckerman, 2017, pp. 192–193).

Compared to Refah and Leyla Şahin, in S.A.S, the argument of militant democracy (protecting democracy by restricting democratic freedoms) was implicit. The Court did not argue that the full-face veil per se represented a threat to the democratic order and argued instead that the covering of the face was incompatible with the ‘principle of interaction between individuals’ – which was seen as an essential part of the democratic order – and thus, ultimately, incompatible with democracy itself. ¹⁶ While the Court was careful not to explicitly project ‘undemocratic’ symbolism on the full-face veil, the question of meaning seems to creep back into the Court’s argumentation through the back door. While the Court avoided making universalizing claims of what the full-face veil symbolizes, it accepted that ‘individuals who are present in places open to all may not wish to see practices or attitudes developing there which would fundamentally call into question the possibility of open interpersonal relationships (…)’ (S.A.S. v. France, para. 122, emphasis added). Thus, the problem with the veil was not perceived by the Court simply as a practical difficulty in interacting with someone whose face is concealed apart from the eyes – it was also about alleged ‘attitudes’ against such interpersonal interaction in public and more specifically, the right of others not to be subjected to such ‘attitudes’ incompatible with the democratic order represented by the veil. ¹⁷ This was a point that the dissenting judges recognized: that in the majority opinion, the ban was not based on the alleged symbolic meaning of the full-face veil, but on the way it might be perceived by those whose rights and freedoms the ban allegedly protected. According to the dissenting judges, the fears and uneasiness the veil might trigger in some people

are not so much caused by the veil itself, which – unlike perhaps certain other dress-codes – cannot be perceived as aggressive per se, but by the philosophy that is presumed to be linked to it. Thus, the recurring motives for not tolerating the full-face veil are based on interpretations of its symbolic meaning. (S.A.S. v. France, dissenting opinion, para. 6)

Concerns have been raised that instead of protecting the fundamental democratic processes of the State, arguments sourcing from militant democracy are increasingly being used to protect abstract ideas of national values and identity, the final outcome of which ‘could be not a variety of militant democracy, but a form of militant culture’ (Invernizzi Accetti and Zuckerman, 2017, pp. 192–103; see also Müller 2012). It is one thing to protect the functioning of the democratic institutions, and quite another to try to preserve an abstract idea of how the members of the society go about their daily life. As Invernizzi Accetti and Zuckerman (2017) argue, to appeal to militant democracy

as a justification for banning headscarves in public places effectively implies that today in Europe it has become possible to be treated as an ‘enemy’ of democracy, even if one has no intention of forming a political organization and competing for political power.

It can be argued that the full-face veil ban and the Court’s acceptance of the national authorities mandate to legislate on the way the people are required to interact with each other reflects the recent changes in the way the secularity principle, laïcité, is understood in France. According to Hennette-Vauchez (2017, pp. 286–287, 297), laïcité, which for a long time was understood to generate obligations for the public authorities, is now perceived to place obligations to private individuals, reflecting the expansion of arguments drawing from militant democracy. Vauchez argues that the contemporary French laïcité has illiberal and discriminatory dimensions, which can be traced to increasing anxieties over Islam. The 2004 law prohibiting religious symbols in public schools illustrated the first rupture in the relatively consistent interpretation of laïcité in France, as it was ‘the first legal formalization of obligations of neutrality weighing on private individuals’. (Hennette-Vauchez 2017, 207; see also Daly 2016.) The next section discusses the ways in which the Court’s procedural approach to the margin of appreciation and Article 14 in conjunction with Article 9 circumvent the proper scrutiny of potential discrimination.

Circumventing the scrutiny of discrimination

The Court’s necessity test ended with granting the State a wide margin of appreciation regarding whether wearing the full-face veil in public should be permitted. Regarding the margin of appreciation, the Court has maintained that

[a]s regards Article 9 of the Convention, the State should (…), in principle, be afforded a wide margin of appreciation in deciding whether and to what extent a limitation of the right to manifest one’s religion or beliefs is ‘necessary’. (S.A.S. v. France, para. 129; see also Evans, 2006, p. 52; Fokas, 2015)

However, the Court has also declared that when it comes to particularly sensitive aspects of personal characteristics and core choices, such as gender and ethnic origin, the margin of appreciation granted for the State is narrow; the Court’s stance to gender discrimination is in principle very strict, and regarding racial discrimination, the Court considers that it ‘is a particular affront to human dignity and requires special vigilance and vigorous reaction from the authorities’. ¹⁸ It is thus questionable whether the same grounds can legitimate discrimination based on gender and religion, and whether the same margin of appreciation can be applied to all discrimination grounds.

The major flaw in the Court’s argumentation is conflating the assessment of potential discrimination based on gender and ethnicity with the question of potential violation of freedom of religion using the same necessity test and the same scope of margin of appreciation on both questions. Had the Court genuinely addressed the question of discrimination in S.A.S., it would have examined whether the same reasons justifying the limitation on freedom of religion were sufficient also for justifying discrimination based not only on religion but also on gender and perhaps also on race, and whether the same scope of the margin of appreciation applied to religion, gender and ethnic background. This would be important for the Court to be able to live up to its own aspiration to strictly scrutinize discrimination based on particularly sensitive grounds.

Instead the Court found that ‘while it may be considered that the ban (…) has specific negative effects on the situation of Muslim women (…), this measure has an objective and reasonable justification for reasons indicated previously’ (paras 146, 151, 161, emphasis added). These reasons, however, did not relate to gender (nor ethnicity), but only to religion (see paras 144–159). The Court’s approach is problematic because while Article 14 is not independent in the sense that it must be assessed in conjunction with a substantive right guaranteed in the Convention, the Article has also individual weight: for example, a violation of Article 14 in conjunction with Article 9 does not presume that discrimination is based on religion. Therefore, the Court can find that Article 14 has been violated in conjunction with Article 9, while Article 9 itself has not been violated, and that the discrimination is based on any of the protected grounds covered by Article 14 (see Danisi, 2011, p. 794; De Schutter, 2011, p. 34).

The big picture is that when assessing Article 14 in conjunction with Article 9 the Court does not seem to consider any other potential discrimination grounds beyond religion. ¹⁹ It seems likely that the explicit focus of Article 9 on religion impedes the consideration of other discrimination grounds within its scope. Similarly restricted assessment does not occur in relation to any other Convention Article. For example, the Court has found a violation of Article 14 in conjunction with Articles 3 ²⁰ and 8 ²¹ based on race, gender, sexual orientation and religion. Interestingly, in S.A.S., the Court could have assessed the discriminatory aspects of the full-face veil ban within the scope of Article 8 in conjunction with Article 14, as it accepted that in addition to freedom of religion, the ban interfered the applicant’s right to private life (para. 110). Instead the Court argued that the ban ‘mainly raises an issue with regard to the freedom to manifest one’s religion or beliefs’ and decided to examine the application ‘under both Article 8 and Article 9, but with emphasis on the second of those provisions’ (paras 108, 109). Had the Court decided otherwise, it may have been compelled to address the question of discrimination beyond religion.

The de-contextualized approach

In this section, the potential of contextual information is discussed in relation to discrimination in cases such as S.A.S., where both the drafting history of an apparently neutral measure and its practical consequences point to intentional discrimination, in this case against Muslim (immigrant) women. As mentioned above, the Court readily accepted that the full-face veil ban adversely affected Muslim women who wished to wear the veil. It did not, however, acknowledge the racialized implications of the ban despite the contextual evidence, particularly the drafting history of the ban recited in the judgment, implicating that Islamophobia was not merely an undesirable side product of the legislative process (see para. 149), but instead an integral part of the ban itself. The aim of the following analysis is not so much to assess the doctrinal possibilities to address implicit modes of racism and gender discrimination within the ambit of the ECHR, but rather to illustrate the problems embedded in the European human rights law in general concerning particularly racialized forms of discrimination.

In short, racialization refers to ‘the extension of racial meaning’ to a ‘relationship, social practice or group’ (Omi and Winant, 2015, p. 11). Thus, a religion such as Islam can become racialized if it is (implicitly) attached with racial connotations. As Powell (2013, pp. 146–147) argues, racialization ‘requires us to move far beyond the idea of racism as a psychological condition, an attitude, a prejudice – some event that occurs in the mind of an actor that predisposes the actor to take an action that is racist’. Instead of focusing on individual racism, racialization invites us to acknowledge structural forms of racism embedded, for example, in marginalizing practices and racialized social hierarchies (see Fitzpatrick, 1987, p. 123). Racialization of Islam is not a new phenomenon, but particularly in the United States, the racial trope of a ‘Muslim terrorist’ has recently altered the American understanding of Islam (Gotanda, 2011, p. 185).

While racial discrimination is widely condemned, structural forms of racism are still under-recognized in human rights law. In the United States and Europe alike, issues concerning religion are predominantly treated through clauses of religious freedom, which results in failure to properly address the racialization of Islam (Gotanda, 2011, p. 186; see also Meer, 2008). Even Islamophobia is often discussed without its connection to racialization (Meer, 2013, p. 386). Thus, according to Modood (1990), we ‘need a concept of race and racism that can critique socio-cultural environments which devalue people because of their origin but also because of their membership of cultural minority and, critically, where the two overlap and create double disadvantage’. For the ECtHR, this would mean seeing beyond the right to practice one’s religion under Article 9 and recognizing that discrimination of Muslims cannot always be subsumed under discrimination based on religion. Instead, differential treatment of Muslims may indicate racialized forms of discrimination (Meer, 2013, p. 63; see also Arnardóttir, 2017, p. 157). ²²

The Court’s decision to approach the so-called full-face veil ban as a case of indirect instead of direct discrimination deflects the fact that the ban, despite being worded neutrally, intentionally targeted Muslim women. ²³ The crucial difference between direct and indirect discrimination is that direct discrimination is based on a protected ground such as religion or gender, while indirect discrimination may occur without such a direct causal or intentional relationship between the protected ground and unfavourable treatment – simply the fact that an apparently neutral policy or measure has prejudicial effect on a protected group suffices. ²⁴

It can be argued that despite its neutral language, the law prohibiting for anyone to conceal their face in public places in fact specifically and intentionally targeted Muslim women and that it thus could be approached as direct discrimination. ²⁵ The Court’s argument assumes that because the law prohibiting the covering of the face does not explicitly target the full-face veil, it is not a potential case of direct discrimination. However, generally speaking, the difference between direct and indirect discrimination is not that straightforward. For instance, in the European Union context, direct discrimination may also occur ‘where criterion which appears to be neutral is in reality inextricably linked’ to a prohibited discrimination ground (Maliszewska-Nienartowicz, 2014, pp. 46–47). It could be argued that the neutrality of the law or provision does not in itself necessarily indicate neutrality of the intention, if the intention of the law or provision is directed at a specific group of people and is thus based on a protected discrimination ground, whether or not that ground is explicitly mentioned.

The legislative history of the ban cited in S.A.S. includes a report of a parliamentary commission on ‘the wearing of the full-face veil in national territory’, established by the conference of Presidents of the National Assembly in 2009. According to the report,

the wearing of this clothing existed before the advent of Islam and did not have the nature of a religious precept, but stemmed from a radical affirmation of individuals in search of identity in society and from the action of extremist fundamentalist movements. (para. 16)

Moreover, the report found that ‘the full-face veil represented a denial of fraternity, constituting the negation of contact with others and a flagrant infringement of the French principle of living together’, as well as negation the principles of liberty and equality (para. 16). Furthermore, the Conseil d’État carried out a study on ‘the legal grounds for a ban on the full veil’, of which the report was adopted by the Plenary General Assembly in 2010. The main question was, would it be possible to legally ban the wearing of the full-face veil as such, or should the ban be worded in general terms without a reference to the Islamic garment, in order for the ban to pass Constitutional scrutiny. Furthermore, the National Assembly unanimously adopted a resolution ‘on attachment to respect for Republic values at the time when they are being undermined by the development of radical practices’ (para. 24). These ‘radical practices’ that are incompatible with the values of the Republic, according to the resolution, include the wearing of the full veil. An indication that the ban was drafted a ‘foreign’ subject in mind is also the fact that the ban is sanctioned with a fine which can be accompanied by an obligation to follow a citizenship course – the ‘real’ French citizen, the sanction seems to suggest, does not wear the full-face veil.

Both the drafting history of the full-face veil ban and the ban on religious symbols in public schools, as well as the way they are applied, demonstrate that both bans clearly target Muslim women and girls, and not primarily because of what their clothing might symbolize, but because of what they are perceived to represent as persons. Sanctioning Muslim girls for wearing a bandana or a full-length skirt (see Vauchez, 2017, p. 300), for instance, suggests that the problem is not the clothing per se, but the way the girls are perceived by others. Similarly, the full-face veil is not, as discussed above, perceived only to make interaction between individuals difficult in practical terms, but also to represent an undesirable attitude towards society. ²⁶ Thus, it can be argued that the full-face veil ban is not intended first and foremost to ‘make living together easier’ by making interpersonal communication easier, but by dispersing unwanted personal attitudes associated with Islam. ²⁷

Acknowledging the concerns raised by many of the third-party interveners, the Court stated that a State which chooses to legislate on the full-face veil ‘takes the risk of contributing to the consolidation of the stereotypes which affect certain categories of the population and of encouraging the expressions of intolerance’, but that the role of the Court is not, however, ‘to rule on whether legislation is desirable in such matters’ (S.A.S. v. France, para. 149). The Court was thus not concerned over the fact that the ban was not only accompanied by Islamophobic remarks, but also in itself both reflected and contributed to Islamophobia. Had the Court genuinely examined the ban and its effects in their context, it could not have escaped addressing the Islamophobia of the ban as an integral part of the law itself instead of unfortunate side-product of the legislative process.

Similar contextual information was disregarded by the Court in Abdulaziz, Cabales and Balkandali v. the United Kingdom ²⁸ where the Court found that immigration legislation, which permitted entry to join their spouses for wives but not husbands of permanent residents, violated Article 14 in conjunction of Article 8. The Court accepted that the aim of the different treatment of male and female immigrants to protect the domestic labour market was legitimate, but notwithstanding the margin of appreciation, the differential treatment of genders would require ‘very weighty reasons (…) before [such treatment] could be regarded as compatible with the Convention’ (paras 78–79). Thus, the Court concluded that the rules were discriminatory based on gender. As to the alleged racial discrimination, the Court took a different approach and found no violation of Article 14. According to the applicants, the immigration rules resulted in indirect discrimination based on race, because, as the minority of the Commission had found, ‘the main effect of the rules was to prevent immigration from the New Commonwealth and Pakistan. This was not coincidental: the legislative history showed that the intention was to “lower the number of coloured immigrants”’ (para. 84). The Court, however, concluded that the rules ‘made no distinction on the ground of race and were therefore not discriminatory on that account’ (para. 85).

Despite these examples, the Court engaging with the contextual information of the case at hand would not be unheard of: The Court frequently considers contextual, empirical information when it determines the scope of the margin of appreciation and ‘European consensus’, and occasionally uses contextual evidence, such as statistics, in establishing facts. For example, in D.H. and Others v. The Czech Republic, ²⁹ the Court weighed, among other things, statistical evidence in determining whether the over-representation of Roma children in the so-called special schools for children with special needs amounted to discrimination, as it effectively segregated the Roma children from other pupils and did not provide them with the same quality of education as others. Thus, in principle the Court can, if it chooses, engage with contextual, empirical information in its decision-making process.

Next, whether an intersectional approach to discrimination might be an answer to the problem of under-recognizing the gendered and racialized aspects of Islamophobia will be addressed. Could intersectionality help the Court to recognize the problems embedded in its procedural and de-contextualized approach to the full-face veil and headscarf cases?

Potential of intersectional approach to discrimination?

Intersectionality recognizes that ‘subjectivity is constituted by mutually reinforcing vectors of race, gender, class, and sexuality’ that shape the experience of individuals and cannot be captured in a single prohibited discrimination ground (Nash, 2008). The thrust of intersectionality is that the law is unable to recognize identity as dynamic, fluid and multifaceted, and that identity, as acknowledged by law, is fragmented in this regard (see, e.g. Bhandar, 2009; Conaghan, 2009; Grillo, 2010; Vakulenko, 2017). Crenshaw’s (1989) critique has shown that anti-discrimination law compels Black women to assert either race-based or gender-based discrimination claims, which leads to a significant part of their lived experience to go unacknowledged by the law. The applicant in S.A.S. faced similar difficulty, as her case was only assessed from the perspective of potential religious discrimination, silencing the gendered and racialized aspects of her experience.

While the intersectional approach is not a solidly established part of the Court’s doctrine on discrimination, the Court has shown some sensitivity in this regard in its recent case law. In B.S. v. Spain, ³⁰ the Court acknowledged that treatment which might not amount to discrimination on any of the prohibited discrimination grounds as such may have a disproportionate effect on someone who is in a particularly vulnerable position (Yoshida, 2013, p. 198). In B.S., the Court acknowledged that the intersecting aspects of the applicant’s position as a migrant woman from Nigeria exercising prostitution in Spain made here more vulnerable to discrimination than other sex workers of Spanish and European origin or other African women living and working in Spain. The Court concluded that the domestic authorities thus had an obligation to take all possible measures to investigate whether State officials had acted in a discriminatory manner in mistreating her. Moreover, recently in Carvalho Pinto de Sousa Morais v. Portugal, ³¹ the Court found that the intersecting aspects of gender and age had contributed to the discrimination of the applicant in a case concerning damages for clinical negligence.

Considering that multiple third parties in S.A.S. pointed out that the full-face veil ban is linked to Islamophobic attitudes resonating with intersecting discrimination grounds of gender and ethnicity, the Courts approach to the potentially discriminatory aspects of the ban appear superficial and cursory. Ideally, the Court would have treated the third-party statements as an indication that the seemingly neutral ban might both reflect and fuel Islamophobic and racist attitudes. ³² The Court could have addressed the underlying Islamophobia, for example, within the parameters set by the Convention and its own previous case law. Instead of approaching the case as a potential violation of the State’s negative obligation not to interfere with the applicant’s freedom of religion, it could have framed it as a case of positive obligation to promote pluralism and to protect the rights of religious and ethnic minorities (see Ferri, 2017). Thus, the Court could have addressed the State’s obligation to protect a particularly vulnerable group of people, such as Muslim women possibly belonging to an ethnic minority, whose position in the society cannot be grasped in one category. This would have required acknowledging not only the indirectly discriminatory effect of the ban on Muslim women, but also the fact that despite its neutral language, the ban in fact directly targeted Muslim women.

However, even an intersectional approach seems inadequate to capture structural discrimination – a form of discrimination that is not practiced by a given individual. Had the applicant of S.A.S., a Muslim woman born in Pakistan, been subjected to discrimination by, for example, individual state officials (like in B.S. v. Spain), the Court might have been more sensitive towards her vulnerable position and to the intersecting discrimination grounds of religion, gender and ethnic background. But as the contested disadvantageous measure (the full-face veil ban) was an apparently neutral outcome of a democratic process, the applicant faced a different set of challenges in establishing discrimination: she was unable to persuade the Court to engage with substantive rather than procedural assessment of the facts of the case or to address the contextual information concerning the full-face veil ban. Thus, given the Court’s procedural and de-contextualized approach, intersectionality might not suffice to tackle discrimination.

Approaching discrimination from the perspective of intersecting vulnerable identities may not always be a strong legal argument, if ‘the rights of others’ are, like in S.A.S., construed in a way that do not engage with the vulnerable position of the applicant. In addition, a structural challenge for an intersectional approach to discrimination in the established legal framework is the emphasis on identity. The law does not sufficiently recognize that discrimination is not first and foremost based on the self-identification of those who are discriminated against; rather discrimination derives from the way in which the targeted group is identified by others. In Europe, Muslims – particularly Muslims with an immigrant background – are increasingly identified as a rather unified group and regarded ‘on the basis of their “Otherness”’ (Berry, 2011, p. 444). The human rights law fails to offer adequate tools for contesting xenophobia which is not necessarily targeting any specific ethnic group but manifests instead in a general racist attitude towards ‘the Other’.

Thus, the question of who belongs to a vulnerable group is a difficult one, and not always determined by self-identification, much less by any unambiguous objective criteria. While self-identification is useful in many respects, it does present weaknesses when applied to jurisprudence on discrimination. ³³ Following the established jurisprudence on discrimination, it is not viable to argue, without empirical evidence of the actual people affected by the full-face veil ban, that the ban has a disproportionate effect on a particular ethnic group and may therefore be considered racial discrimination. This may be considered problematic in cases where the discriminatory policy or measure is intended to primarily affect racialized groups regardless of whether or not these groups themselves identify with any particular ethnicity or whether they share protected characteristics (see, e.g. Mehta, 2010). Moreover, while discriminatory practices may superficially be aimed at religion, they may be symptomatic of underlying ethnic intolerance (see, e.g. Bhandar, 2009; Gozdecka et al., 2014; Lentin and Titley, 2010; Mills, 1997; Modood, 1997; Williamson, 2010; cf. Berry, 2011, pp. 441–442).

Regarding the contextual information available in S.A.S., the main problem was that the Court only engaged with religious discrimination, and seemed unable, or perhaps unwilling, to consider the possibility that the full-face veil ban embodies Islamophobic, even racist, connotations and that the racialized assumption of the incompatibility of the full-face veil with the French values are ingrained in the ban.

Conclusion

Discussing the case of S.A.S. v. France as an example, this article has suggested that the reason the Court’s jurisprudence lacks rigour in its approach to discrimination is that it is both procedural and de-contextualized. Procedural approach means that the Court does not always engage with necessity and proportionality test but instead, as it did in S.A.S., accepts that the procedural safeguards of the national decision-making satisfy the requirements stemming from the Convention rights. Together with a wide margin of appreciation granted for the state, this results in a superficial legal analysis.

Regarding the prohibition of discrimination, there are two major flaws in the Court’s approach. Firstly, the Court was unable to recognize that granting the State a wide margin of appreciation regarding expressions of religious conviction results in granting the State a wide margin also regarding unfavourable treatment based on gender and ethnicity, unless the Court genuinely and separately assesses whether the same justification applies to restricting the scope of freedom of religion and discrimination based on gender and race. Secondly, the fact that the Court limits its scrutiny of Article 14 in conjunction with Article 9 to religion, needlessly narrows the scope of Article 14.

Combined with the Court’s general approach to Article 14, the above said is disconcerting: when the Court finds a violation of the substantive right, such as freedom of religion, it frequently passes the complaint of discrimination where it considers that this would involve examining essentially the same complaint (Vakulenko, 2017, p. 191). The Court’s approach to discrimination turns out to be problematic, however, when it also effectively passes the complaint of discrimination when it does not find a violation of the substantive right in question, such as in the case of S.A.S. Overall, Article 14 in conjunction with Article 9 appears to hold little individual significance beyond religious discrimination. This is worrying, especially in light of the Court’s assertion that sensitive discrimination grounds such as race and gender need special protection.

The Court’s de-contextualized approach to discrimination in S.A.S. was identified in this article as a crucial reason for possible discrimination going under-detected. The judgment itself includes a section describing the drafting history of the law prohibiting the concealment of one’s face in public places, which clearly shows that the law was motivated to ban the niqab and burqa, and not designed as a neutral ban on covering of the face in general, despite its apparently neutral formulation. The Court did not, however, consider the connection between the ban and Islamophobia further than noting that it was concerned over ‘certain Islamophobic remarks’ that had marked the debate preceding the adoption of the law. This article has also argued that despite the fact that the Court attempted to avoid in S.A.S. the question of what the full-face symbolizes, it inadvertently re-produced the image of Islam as incompatible with Western democracy, as it acknowledged that the French state could reasonably perceive the full-face veil to be incompatible with local democratic values. This logic resonates with militant democracy.

‘Western’ subjectivity inevitably has racial implications, as it is founded upon the dichotomy of the ‘Western’ ‘us’ and the racialized ‘Other’. As Bhandar (2009) points out, racial communities are always constructions, and the legal distinction, for example, between ethnic, religious, linguistic and other groups is inevitably somewhat arbitrary. In this context, it is possible to see S.A.S. as a case of whiteness and non-whiteness – not in the sense of race or colour, but in the sense of a White citizen subject, whose existence must be protected from non-White influences and practices, such as covering of the face in public. What is needed in order to tackle the multiple forms of racism and xenophobia is recognizing the possibility that racist discrimination might hide behind references to culture, tradition and religion. ³⁴ Moreover, it is also important to address discrimination that is based on perception rather than the actual ethnic background of the person discriminated against or on their self-identification with a particular ethnic group. Ethnicity is, after all, not only a way of classifying oneself, but also others (Lyon, 1997, p. 187).

In conclusion, it can be argued that together the procedural and de-contextualized approach to discrimination result in a weaker protection of vulnerable groups by means of Article 14 than to what the Court aspires. Because the Court’s argumentation in S.A.S. follows the established pattern of assessing discrimination under the Convention, the weaknesses identified in this article have general relevance in the analysis of the Court’s jurisprudence.