Religious Neutrality, Laïcité and Colorblindness: A Comparative Analysis

Introduction

The concept of religious neutrality is on the rise in Europe. Neutrality has become a concept central to European law, both at the level of the European Union (EU) and at the level of the Council of Europe. It also infuses a variety of national legal regimes based on secularism.

Under the European Convention of Human Rights, religious neutrality has indeed assumed center stage in Article 9 (freedom of religion) jurisprudence. The Strasbourg Court has explicitly referred to European States’ duty of neutrality and impartiality with respect to religions since a 2000 Grand Chamber ruling,1 and it has elaborated on that jurisprudence in later cases. For instance, when it upheld France’s ban on the niqab, it insisted on “the State’s role as the neutral and impartial organiser of the exercise of various religions, faiths and beliefs . . . [that] is conducive to public order, religious harmony and tolerance in a democratic society.”2 It further insisted that the State’s duty of neutrality and impartiality is “incompatible with any power on the State’s part to assess the legitimacy of religious beliefs or the ways in which [they] are expressed.”3

Such affirmation of an obligation of State neutrality is quite remarkable in the context of the Council of Europe where a wide variety of state-religion arrangements prevail across forty-seven Member States. Some countries such as Turkey and France certainly have regimes of strict separation in which the concept of State neutrality might be unsurprising, but others have systems of established or official churches (this is the case in the United Kingdom, Norway, and Greece, for instance), while others organize milder regimes of separation that allow various forms of collaboration between State and religious authorities. In Italy for instance, religious marriage automatically translates as civil union; other countries (such as Germany, Belgium, or Spain) can also be read into this third category.4

Because it captures and governs a wide and differing range of church-state arrangements, the concept of neutrality that stems from European Court of Human Rights (ECHR) case law is quite broad and variegated. Julie Ringelheim has established that it oscillates between three main understandings: one refers to neutrality as “absence of coercion,” another defines neutrality by the “absence of preference,” and a third one extends to forms of “exclusion of religion from the public sphere.”5 The third understanding is however conceptually quite distinct from the former two, especially since it possibly extends to private individuals; it refers to a religious neutrality of the regulated (society) rather than to the religious neutrality of the regulator (the State), and is illustrated, for instance, by the various rulings upholding the French and Belgian “burqa bans.”6

A principle of religious neutrality is also emerging at the level of the European Union. The Council of Europe and the European Union (EU) certainly have different raisons d’être; while the former rests essentially on a human rights paradigm, the latter was built as a space for economic integration and only later took human rights on board. A concept of non-discrimination was, however, always important in the shaping and developing of a European internal market. As anti-discrimination law was considerably strengthened by the adoption of important directives in 2000,7 the EU started producing a legal discourse on religious discrimination. It strikingly converges with that of the ECHR, especially in the ways that it immediately installed a concept of religious neutrality at its very center—from the very first judicial interpretations of “religion” under EU anti-discrimination law. The Court of Justice of the European Union (CJEU) was indeed given its first opportunity to interpret the scope of “religion” under directive 2000/78 as two preliminary references emanating from Belgian and French high courts led the court to clarify whether instances of female Muslim workers having been fired due to their refusal to remove their headscarf in the workplace amounted to religious discrimination. As it was given its first opportunity to interpret the scope of “religion” under directive 2000/78,8 the court chose a wide definition of religion, insisting that it included both forum externum and forum internum dimensions.9 It then established the legitimacy of religious neutrality policies in the workplace, finding that internal corporate neutrality policies banning the expression of all convictions or beliefs in the workplace (religious, but also philosophical or political) were legitimate in principle as they stemmed from the right to conduct a business that is guaranteed by Article 16 of the EU Charter of Fundamental Rights (EUCFR).10

The court ruled that such policies did not amount to direct discrimination, nor did they constitute indirect discrimination as long as they were applied in a consistent and appropriate manner and remained proportionate to the objective they serve. It also ruled that customer preferences could not qualify as “genuine and determining occupational requirement[s]” susceptible of justifying a different treatment based on religion.11 While these rulings have triggered much criticism from the perspective of anti-discrimination law (for they are thought to severely weaken its operation), they also testify to the elevation of the legitimacy of neutrality policies in the workplace, which is a significant shift if not departure from the previous state of legal affairs (and indeed, one that is said to considerably lower the standard of human rights protection that prevailed until then in a number of EU Member States).12

Interestingly, as both the full veil cases in Strasbourg and the hijab in the workplace ones in Luxembourg show, the notion of religious neutrality has expanded from the public sphere to the private sphere. If initially, it merely required neutrality from the State (its institutions, its legal rules and, at most, its embodiments such as buildings and civil servants), it now tends to be used in a very distinct sense, one that requires religious neutrality from individual people. Rules and policies of neutrality in the workplace can thus be upheld, and there are several other social spaces in which private individuals can be subjected to rules of religious neutrality. This is the case, in several countries, for pupils in schools,13 students in universities,14 and sometimes laypeople on the street15 or in public spaces generally—not to mention the rules of neutrality that weigh on those individuals who can be said to represent or embody public authority and may thus be subjected to such neutrality rules: such as school teachers,16 university professors,17 and also wider groups of civil servants, including nurses.18

However, this increasing elevation of neutrality as a cornerstone of European legal responses to the issues raised by religion in contemporary societies is suspected to play out adversely for specific religious minorities—in particular, Muslim minorities. Some authors argue that the rise of neutrality conceals a strong dimension of Islamophobia “concealed in the principled garb of secularism.”19 Mathias Möschel has unearthed a “European way of colorblindness” and recalled that it emerged, historically, in reaction to the persecution of Jews; he has also established the ways in which, nowadays, “the issue no longer is one of racism but one of religion and secularism.”20 Analyzing ECHR law, Samuel Moyn speaks of a “European devotion to a neutral state”21 and underlines the ways in which ECHR case law on freedom of religion seems to systematically play out against Islam: “[o]ne case can be an honest mistake, but an almost unbroken trend demands some other interpretation.”22

This adverse impact of European neutrality is, in part, implicit. It results from the failure to see the extent to which social norms, as well as a number of the legal rules that actuate them, are deeply embedded in historical and religious structures23 that have become problematic as religious pluralism has increased in contemporary European societies. Questions such as the “predominant rule of Sunday rest or official annual calendars” are a case in point24 of “nonchalant attitudes towards Christian symbols,”25 but there are many others, from the labeling of nativity crèches as “not exclusively religious” signs,26 to the acceptance of the presence of crucifixes in public school classrooms.27 But the impact can also be explicit.

This is the case, for instance, when legislation explicitly pits specific religious practices (veiling, avoidance of physical contact, request of exemptions from the general norm) as radical, fundamentalist, or separationist and thus inadmissible on the grounds that they undermine the conditions of “living together”28 that are necessary for a society to hold together.29 This is also the case when an employer’s decision to ban the expression of any conviction by all employees is deemed legitimate as an expression of his or her wish to “project an image of neutrality towards customers [that] relates to the freedom to conduct a business that is recognized in Article 16 of the [EUCFR].”30

It is indeed hardly disputable that all the restrictions to religious freedom that have been upheld by European Courts (both the ECHR and the CJEU) over the years either originate in cases involving Islamic religious practice or signs that disproportionately impact the practice and visibility of Islamic faith (and often both). In Strasbourg, this is certainly the case for the burqa bans that were upheld,31 as well as for most of the cases involving the wearing of religious garb—on the street, in schools, in the courtroom, and elsewhere, including the workplace—with the notable exception of the Eweida v. United Kingdom rulings in which the Strasbourg court chose rare Catholic cases for strengthening its standard of review.32 In Luxembourg, it is also significant that in two cases involving female workers who had been fired for refusing to remove their headscarf, the court ruled on the admissibility of neutrality policies within the anti-discrimination law framework.33

These exclusionary dimensions of the European legal regime of religious neutrality are increasingly read as a shortcoming if not a failure of both the human rights framework and that of anti-discrimination law. Both of them tend to be read as incapable of adapting to the contemporary conditions of pluralism that feature in European polities. Further, this legal regime of neutrality is also read as resting on an exclusionary ideology—one that recalls Said’s Orientalism and pits “the West” against an “Other” that is almost always portrayed throughout to Muslim practices (of veiling, slaughtering, learning, etc.) and couches this opposition into neutral terms:34 secularism, “living together,” and neutrality (indeed), but also laïcité, or “republican” traditions—including, in a highly problematic fashion, gender equality.35 Much of the literature that critiques the rise of European neutrality in these terms has paid close attention to the French example that is said to be both emblematic of the ways in which a polity may organize around the central value of laïcité and inspirational for the wider European model. Authors such as Joan W. Scott,36 John Bowen,37 or Mayanthi Fernando,38 for instance, have looked at the French regime of laïcité in that way, explaining the extent to which it has remained profoundly shaped by the colonial encounter between republican values and Islam and continues to hesitate between outright forms of rejection and discrimination on the one hand and commandments and injunctions to assimilate on the other hand.39 Joseph Weiler has similarly read the CJEU Achbita ruling as one “following the French tradition.”40

This Paper suggests that these dimensions of the current debate over the public expression of religion in Europe echo some aspects of the debate over racial equality in the United States. As it explores the ways in which European debates over religious neutrality and American debates over racial equality respond one to another, it reflects on the extent to which some of the critiques that have been voiced with respect to the notion of colorblindness, which has played an important role in shaping the constitutional debate over race in the United States,41 may illuminate the role that the legal principle of neutrality might be acquiring in contemporary European law. In particular, the Paper reflects on possible parallels that can be drawn between colorblindness as a cause for the failure of U.S. law to redress (or even address) enduring patterns of racial subordination as well as the mere ideological mask of white supremacy on the one hand,42 and the exclusionary dimension of contemporary iterations of religious neutrality in Europe on the other hand.

The Paper unfolds in three parts. The first Part is essentially a caveat to the rest of the demonstration, for indeed there seems to be one significant dimension in which the attempted parallel reading of colorblindness and neutrality does not prove operational. In U.S. law, colorblindness has consistently led to the constitutional doctrine of anti-classification. Subsequently, its reach has been contained to explicit racial categories: only when and where there is an explicit legal recognition of a race is the Equal Protection Clause triggered—and the standard of strict scrutiny43 set in motion—by operators of judicial review. Neutrality, by contrast, is hardly an anti-classification device: as it exists and is increasingly elevated in European law, it neither necessarily entails nor requires a juridical prohibition on religious classifications. The first Part of the Paper thus addresses this significant caveat to the comparability of the two legal notions and explains why and how it nonetheless purports to compare the theoretical and scholarly analyses they have triggered (rather than the actual constitutional concepts themselves). By reflecting on the ways in which some of the critical work on colorblindness in American law may illuminate aspects of the current European debate over legal responses to religious pluralism, it engages in a meta-comparison. The second Part of the Paper moves to explore the similar patterns of legal reasoning that judges deploy when mobilizing either colorblindness (in the United States) or neutrality (in Europe). It first insists in particular on the formalistic and a-teleological dimension of the judicial reasoning that both these concepts trigger. It then underlines the importance it reserves to a variety of notions (from the public/private divide to the State/federal one) that hinders the full-fledged application of equality and anti-discrimination law and systematically contains (rather than expands) it. In the third Part, the Paper underlines the similar outcomes of the judicial interpretation of colorblindness and neutrality. Racial and religious othering, including with troubling concessions to overtones of purity, seem indeed to infuse the case law on both sides of the Atlantic.

I. A Caveat to Comparability

There is nothing natural or self-evident to legal comparisons: legal systems and rules are essentially a social product and, therefore, there is no particular reason why they should even be comparable. Institutional setups as well as substantial principles may well differ so radically from place to place that the actual operation of comparison of similar concepts across different legal systems might well be pointless—let alone that of different concepts. Furthermore, it could well be argued that rules and principles relating to the legal treatment of the issue of racial inequality in the United States on the one hand and to that of religion (and the public manifestation thereof) in Europe on the other hand provide with particularly ill-suited candidates for comparative work, for the issues of race and of religion have had vastly differing histories—legal and political—across the Atlantic. In light of such immense differences, what, then could possibly be compared in the contemporary legal responses to issues of racial and religious equality in the United States and Europe? This Paper posits that while comparing these actual legal concepts might indeed be impossible or problematic, focusing on the analyses they have triggered is enlightening.44 In other words, it engages in a work of meta-comparison rather than comparison itself. It seeks not to compare (American) “colorblindness” and (European) “neutrality” but rather, to take stock of what decades of critical scholarly work on colorblindness in the United States have brought about that might enlighten current legal developments and challenges that European law is facing when confronted with issues of religious diversity.

As a result, this Paper does not dispute nor is it impeded by the fact that the two concepts of colorblindness and neutrality are hardly conceptual or functional equivalents. To the extent that colorblindness is generally associated with the famous excerpt of Justice Harlan’s dissent in the infamous Plessy v. Ferguson case ruled by the Supreme Court in 1896 (“[o]ur Constitution is color-blind, and neither knows nor tolerates classes among [its] citizens”),45 it could probably be affirmed that European law, by and large, operates under a somewhat similar principle of religious-blindness entailing that individuals are not to be subjected to unequal treatment because of their religious beliefs. Legally though, while colorblindness has provided with a historic articulation of what would later be coined an anti-classification requirement under the Equal Protection Clause, there is no such anti-classification dimension to “religion-blindness” (if at all) in European law.

Quite to the contrary, neutrality as it exists in contemporary European law cannot be defined as a ban or prohibition on religious recognition. This is of course self-evident in a number of European countries where there is an official or recognized church: by definition then, one religion enjoys special status—and forms a particular class. At times, this structural element percolates deeply into the legal system and individuals may be defined by their religious affiliation—be it on their identity documents46 or for the sake of labor law and the determination of their entitlements to holidays.47 Across Europe, however, although religious discrimination is generally prohibited by domestic constitutional rules and/or EU legislation,48 not all constitutions have explicit anti-classification provisions. While Italy49 and France50 do prohibit all forms of religious classification, the Belgian Constitution for instance remains silent on the subject.51 Basically, then, in Europe, neutrality cannot really be read as a classificatory (or not classificatory) legal principle. It rather commands that states, as regulators of constitutional principles of equality, pluralism, and tolerance, ensure that all religions are treated equally—including when the starting point is not equal (i.e., when one church enjoys special status).52

It is thus in full awareness of the strong differences between colorblindness and religious neutrality that this Paper engages in an exercise in meta-comparison that seeks to bring some of the main lessons and conclusions of decades of scholarly work on racial equality in the United States into the European legal debates over religious pluralism. In particular, I pay attention here to the ways in which the strongly anti-classificatory dimension of the American constitutional doctrine of colorblindness is generally understood to fail to address and redress the subordination dimension of racial equality. Regardless of the actual ebb and flow of acceptance of and backlash against the anti-subordination dimension of the actual judicial interpretation of the Equal Protection Clause in U.S. law, I wish to take stock of the theoretical arguments and demonstrations pertaining to the shortcomings of a solely anti-classification approach to discrimination in order to suggest that they would be immensely beneficial to contemporary European legal debates over religious pluralism. I thus examine the theoretical critique of the pattern of legal reasoning that the doctrine of colorblindness, as read through a solely anti-classification lens, has generated in American constitutional law in order to highlight a number of parallels with the current operation of European anti-discrimination law with respect to religion (Part III). I then turn to the critique of the outcomes of the judicial interpretation of colorblindness and again suggest that parallels can be fruitfully drawn between the constitutional production and maintenance of white supremacy in the United States and the exclusionary dimension of religious neutrality in Europe (Part IV).

II. Modes of Reasoning

In terms of how American (racial) colorblindness and European (religious) neutrality can be usefully reflected upon simultaneously, I wish to suggest that two features of the specific modes of legal reasoning that are triggered by both these principles are worth unpacking. The first is formalism. In constitutional debates over both race in the United States and religious neutrality in Europe, concerns are expressed in relation to the triumph of overly formalistic modes of reasoning that lead to core provisions such as the Equal Protection Clause or anti-discrimination law being interpreted in a way that potentially completely detaches them from their purpose.53 As a consequence, it is not uncommon to encounter actual dishonest or cynical moments in legal interpretation. The second feature is the limited reach of equality and anti-discrimination law that ensues: as legal actors promote a formalistic reading of the provisions of reference, they seem to seize every opportunity to turn unrelated legal principles into shields to the full application of equality law.54

A. Formalistic Legal Reasoning, Anti-Classification as Symmetry, and the Limited Reach of Anti-Discrimination Law

In terms of patterns of judicial reasoning, the study of contemporary legal debates over religious pluralism in Europe provides a strong sense that European legal actors are strangely stuck in formalistically anti-classificatory interpretations of anti-discrimination law, which a brief examination of their American counterparts’ experiences could help them avoid: anti-classification is probably not enough for achieving equality. Indeed, although symmetry and formalism are not necessary consequences of the constitutional concept of colorblindness,55 important strands of legal scholarship have established that they actually killed much of the egalitarian promise that they were deemed to convey.

To be sure, colorblindness as a foremost anti-classification device did allow for some progress in terms of racial equality in American law, to the extent that it contributed to the dismantling of the Jim Crow era’s explicit de jure segregationist laws and policies. This actually explains why the civil rights movement has a history of cherishing and mobilizing the semantics of colorblindness. It also explains why the 1954  Brown v. Board of Education56Brown v. Bd. of Educ. (Brown I), 347 U.S. 483 (1954). ruling is generally associated with the constitutional affirmation of the inadmissibility of de jure racial segregation (as well as with, more generally, that of racial classification), although the ruling itself only applies to the issue of segregation in public schools and remains essentially silent about its exact rationale57 (in particular, it does not clarify whether the harm to children is caused by the classification itself or by the subordination that ensues). Whether or not a full-fledged anti-classification normative program was actually present in the Supreme Court’s case law as early as 1954, what is clear is that its radicalness led the Court to only very progressively unfold it.58 The Court admitted that “additional time might be necessary to carry out the ruling in an effective manner,”59 only weakly requiring that this be done with “all deliberate speed”60 in order to preserve the Court’s own legitimacy and/or to prevent (or contain) backlash.61 Consequently, and regardless of the importance of the Brown ruling, it took another decade for the radical notion of colorblindness to actually become law through the adoption of the Civil Rights Act, which forbade discrimination on the basis of race.

Despite its (progressive) role in dismantling de jure racial segregation,62 colorblindness proved to have severe limitations with respect to the struggle for racial equality. First, it was intrinsically unable to tackle, challenge, and question the many proxies for race that lied at the core of racially unequal laws and policies—and, therefore, to address the more deeply entrenched forms of racism and racial inequality. Zoning decisions,63 recruitment policies, criminal sentences64—many practices and policies relied on an entrenched racial bias without explicitly relying on racial categories and have thus remained essentially out of reach of the constitutional doctrine of anti-classification. Second, conservative forces progressively succeeded in imposing a symmetrical structure to judicial interpretations of colorblindness, thus turning anti-classification into the exact opposite of anti-subordination.65

As Reva B. Siegel has shown, although colorblindness and anti-classification could have been interpreted separately (and, in fact, had been interpreted separately until the end of the 1960s),66 post-Brown courts progressively started striking down all race-conscious remedies on the basis that they were forms of classification that always ran counter the anti-classification principle—one that no longer allowed for distinctions to be made according to whether the classification at stake was pernicious or benign.67 This soon led, in the words of Derrick Bell, to a situation where “a state law or policy designed to increase minority participation in the railway construction industry receives the same judicial scrutiny as a state law requiring black railway passengers to sit in the rear of each car.”68 Neil Gotanda further explains that this occurred as strict scrutiny started being triggered not only when ruling on the constitutionality of restrictions that curtailed the civil rights of minorities (e.g., Brown),69 but also more generally when deciding any kind of racial classification, regardless of its historical and social context70 (e.g., City of Richmond v. J.A. Croson Company).71

A third reason why colorblindness progressively appeared to be more of a foe than a friend to actors mobilizing in favor of racial equality was that the focus on racial classification it entailed was found by many to actually divert the attention from the real cause of injustice and/or the preferred aims of the combat for justice. Authors and actors associated with the critical race theory movement were instrumental in this evolution of the debate. Derrick Bell, for instance, wrote influential pieces underlining the ambiguities of the post-Brown embrace of colorblindness as a matter of constitutional law and of integration as a matter of educational policy by the civil rights movement.72 He denounced the growth of the objective of integration and the ways in which it completely sidelined and obfuscated the more complex and imperative need of Black communities for quality education.73 For all these reasons and more, colorblindness can be seen as a tool of racial inequality rather than an impediment thereto.74

From a legal standpoint, there are two main illustrations of these shortcomings, both of which show the extent to which American constitutional racial equality law has retained a deeply formalistic reasoning structure. First, they have led to the containment of the doctrine of anti-classification’s reach to explicit legal racial categories. American constitutional law is replete with examples in which the anti-classification doctrine was allowed to strike down explicit racial classifications, such as rules that restricted Blacks and other people of color’s access to schools, restaurants, public transportation, fountains, and the like. But it remained essentially toothless toward more structural impediments to their full access to equality, such as education, fairly diverse neighborhoods, and employment opportunities.

Emblematic in that respect is Justice O’Connor’s concurring opinion in Hernandez v. New York, in which she insists that juror dismissals that led to the exclusion of all and only Latinos were not to be found unconstitutional because, although they “may have acted like strikes based on race, . . . they were not based on race. No matter how closely tied or significantly correlated to race the explanation for a peremptory strike may be, the strike does not implicate the Equal Protection Clause unless it is based on race.”75 Neil Gotanda has written an illuminating account of one of the conceptual reasons for this specific shortcoming of the doctrine of anti-classification in which he demonstrates that, throughout American constitutional law, the word “race” refers to differing and indeed diverging meanings. He suggests that contemporary Supreme Court cases no longer use the concept of race that permeated rulings such as Brown and Bakke but rather a different concept of race that he coins “formal-race,” in which racial classification entertains no connection to social reality.76 This latter concept sees racial categories a merely formal categories that do not entail or imply any substantive meaning.77 Technically, this has meant that the colorblindness/anti-classification’s very hermeneutics of the Equal Protection clause have prevented the doctrine from redressing unintentional forms of discrimination. This proved to be especially true when unintentional forms of discrimination resulted from facially neutral laws and policies.78 As long as these laws and policies do not explicitly operate on the basis of suspect (racial) categories, the fact that they may result in excluding disproportionate numbers of one particular racial group has remained essentially irrelevant to the assessment of their constitutionality.79

A second shortcoming of formalistic judicial reasoning on racial equality relates to the fact that, because its very definition began and ended over the notion of distinct classes and categories, colorblindness was progressively construed as an impediment to active or remedial measures of racial inequality. Because the mere existence of a category may trigger the colorblind/anti-classification reasoning, judicial reasoning progressively expunged the raison d’être of that category from constitutional reasoning on its possible acceptability. As motive was progressively exfiltrated from the constitutional assessment of the admissibility of racial categories, constitutional law deprived itself of the cognitive tools necessary to distinguish between benevolent and invidious forms of classification. As a result, race-conscious mechanisms seeking to increase or favor minorities’ access to rights and services were placed at risk of unconstitutionality because of the concept of colorblindness.80 Justices on the bench developed theories of the moral and legal equivalence of all racial categories,81 be they noxious or remedial. Efforts to promote racial integration in schools,82 or to strip away discriminatory bias in government employment,83 were thus put in jeopardy as insufficiently narrowly tailored to their aim.

In the aggregate, these shortcomings have infused much of constitutional reasoning in the field of racial equality in the United States with strictly symmetrical and a-teleological readings of equality. These characteristics are hallmarks of formalistic legal reasoning when it comes to reasoning in the field of equality and anti-discrimination law. In her work on substantive equality, Sandra Fredman insists that to actually redress persistent forms of disadvantage, a distributive dimension must be incorporated into the concept of equality.84 As a consequence, legal reasoning ought to be asymmetric: legal rules and mechanisms that rely on (racial) categories to pursue the aim of redressing structural or historical forms of disadvantage cannot be subjected to the same forms and standards of scrutiny as those that classify for segregationist and discriminatory purposes.85 Compensatory or preferential treatment is actually a matter of realism that is negatively revealed by the fact that much of the constitutional construction of the admissibility, if not legitimacy of, segregation proceeded from a blatant denial of reality—one that at times hid sheer dishonesty.

Among many examples, the notion of privilege or favoritism that infuses much of the constitutional debate over race-conscious remedies of inequality in the United States is highly illustrative of such dishonesty.86 In the Plessy case, Justice Brown famously insisted that there was a fundamental “fallacy” in Homer Plessy’s argument: “the assumption that the enforced separation of the two races stamps the colored race with a badge of inferiority.”87 Justice Brown further insisted that such a notion was “not by reason of anything found in the act, but solely because the colored race [chose] to put that construction upon it.”88 Earlier in the 1883 Civil Rights Cases, when the Supreme Court struck down the first set of federal civil rights laws that sought to protect the rights of Black people to access public facilities on an equal basis, Justice Bradley’s rhetorical question was equally highly problematic. Justice Bradley asked, “But what has [segregation] to do with the question of slavery?” to claim the inapplicability of the Fourteenth Amendment to the case at stake.89 He went on to start infusing the notion that once slavery had been abolished, Blacks were not to request or expect preferential legal treatment, famously insisting that “[w]hen a man has emerged from slavery, and by the aid of beneficent legislation has shaken off the inseparable concomitants of that state, there must be some stage in the progress of his elevation when he takes the rank of a mere citizen, and ceases to be the special favorite of the laws . . . .”90 That a piece of civil legislation granting Blacks access to public accommodations on equal footing with white people could be read as turning the former into “the special favorite of the laws” is ironic, indeed.91 It is ironic in that it equates ending unfavorable treatment with actually enacting preferential treatment.92 It is also cynical, as it expresses a notion of unfair and undue preference purportedly enjoyed by Blacks that found political traction in the following years as Jim Crow legislation proliferated between 1890 and 1910 after the invalidation of the federal civil rights laws.93

However problematic, this rhetoric of privilege and favoritism endures to this day. As the Supreme Court’s famous Bakke ruling started to considerably weaken affirmative action policies taken by universities,94 it did so on the grounds of a notion of undue privilege of minority students, which, in turn, resulted in oppression and disadvantage of the majority. In the words of Justice Powell, an admissions policy that reserved sixteen seats out of one-hundred for disadvantaged minority applicants indeed risked leading to a situation where “innocent persons . . . . [were] asked to endure [deprivation as] the price of membership in the dominant majority . . . .”95 Hence, Justice Powell’s offering of “diversity” was a compromise96 “that would allow limited voluntary race-conscious efforts at desegregation to continue, in a social form that would preserve the Constitution as a domain of neutral principles.”97 In the words of Justice Mosk of the Supreme Court of California:

That whites suffer a grievous disadvantage by reason of their exclusion from the University on racial grounds is abundantly clear. The fact that they are not also invidiously discriminated against in the sense that a stigma is cast upon them because of their race, as is often the circumstance when the discriminatory conduct is directed against a minority, does not justify the conclusion that race is a suspect classification only if the consequences of the classification are detrimental to minorities.98

This tendency has only strengthened in more recent times. Summing up the past few decades of constitutional attacks on affirmative action, George Lipsitz writes that “[a]nti[-]discrimination law becomes portrayed as race discrimination. Measures designed to secure rights for Blacks are caricatured as making Black people into special favorites of the law. Whites asked to obey the law, conversely, are represented as victims of reverse racism.”99

Formalistic legal reasoning that rests on symmetrical and a-teleological forms of reasoning and consequently fails to address (and redress) unintentional forms of discrimination, actively prevents the adoption of positive action measures and does not hesitate to misrepresent the lived realities of those who claim they remain unprotected is a pattern that can also be found in European law. The Achbita case, decided by the CJEU in 2017, illustrates the neutralization of anti-discrimination law that can be caused by symmetrical and a-teleological judicial modes of reasoning. When asked whether an internal neutrality policy that prohibited all personnel from expressing any kind of convictions (religious, political, or philosophical convictions, for example) amounted to discrimination, the court first deployed a symmetrical mode of reasoning to determine that such a policy did not constitute direct discrimination.100 Since the disputed policy prohibited employees from expressing any convictions in the workplace, the court ruled that it treated “all workers of the undertaking in the same way by requiring them, in a general and undifferentiated way, inter alia, to dress neutrally, which precludes the wearing of such signs.”101

Such symmetrical mode of reasoning verifying that all convictions and all workers are treated equally is, however, debatable with respect to the very goals of anti-discrimination law. It amounts to saying that, in terms of discrimination, as long as all members of a group are discriminated against for expressing their convictions in the workplace—as in Achbita—none are discriminated against. Such reasoning also rests on an arguably flawed choice of comparator. In a claim of religious discrimination, as was the case in Achbita, the applicant’s situation should be assessed in comparison to situations of people expressing no beliefs rather than to situations of people expressing other kinds of beliefs.

These are some of the reasons why the Achbita ruling has been critically received by anti-discrimination law scholars, many of whom have underlined that it provides prejudiced and overcautious employers with a blueprint to avoid dealing with religious pluralism in the workplace and, consequently, discriminate on the basis of religion (or, for that matter, on the basis of philosophical or political beliefs). All they need to do now is implement an overall ban on the expression of all convictions,102 which will not qualify as a form of discrimination.103 This normative result of the rulings is also at odds with the project of a pluralist society that infuses the general program of anti-discrimination law in the first place,104 thus echoing the loss of purpose and a-teleological interpretation of the ways in which anti-discrimination law is supposed to govern the workplace. The wording of the ruling certainly testifies to the fact that, had the company prohibited the expressions of belief corresponding to one particular faith, or even those of religious convictions (as opposed to other types of convictions), the court’s symmetrical reasoning might have led to the finding of an instance of discrimination, as not all workers (and/or not all convictions) would have been treated similarly badly.

This case is also a clear illustration of the ways in which formalistic and symmetrical modes of reasoning with respect to equality and non-discrimination law hinder their ability to address indirect forms of discrimination—ones that do not take the form of direct classification. Anti-discrimination scholars who had tackled the issue before these 2017 CJEU rulings had mostly considered that neutral bans in the workplace would likely constitute “indirect discrimination on the basis of religion, as they seem to disproportionately affect non-Christian faiths, which are more often acquainted with dress codes . . . .”105 This also explains the dominant—if not unanimous—critical tone of scholarly commentary on the 2017 rulings.106

The court unfolded a highly formalistic reasoning, finding that as long as it (1) pursues a legitimate aim, (2) is appropriate and pursued in a consistent and systematic manner, and (3) is limited to what is strictly necessary, an internal neutrality rule should not lead to indirect forms of discrimination.107 It lists these three conditions as allowing to rule out an instance of indirect discrimination, but gives no guidance or assessment as to what they truly refer to. For instance: would a supermarket that caters to religious clients by vending religiously vetted foods be “consistent” if they required (religious) neutrality from their personnel? Or: as it rules that as long as a neutrality policy “only covers those employees who interact with customers” it can be deemed necessary, it fails to explain what “interaction” refers to—not to mention the front-office–back-office situation it risks generating.108 Could a janitor wearing a uniform (and therefore, identifiable as an employee) thus be subjected to such a rule on the grounds that he or she might come across customers, or would that only apply to employees whose job it is to actually speak with customers? All these questions remain unanswered as the court only provided the keywords of an anti-discriminatory checklist but refrained from giving their meaning—thus hindering their actual applicability and effectiveness.

European human rights law does not seem to fare much better in this respect, as it has consistently failed to find breaches of the European Convention on Human Rights Article 14 non-discrimination clause in cases involving religious freedom when plaintiffs were Muslim.109 Its rulings on the burqa bans against France and Belgium epitomize this failure: in all four cases, the applicants were Muslim women who claimed that the bans (both of them phrased in facially neutral terms) amounted to indirect forms of discrimination they experienced as Muslim women. Even though the court did express its concern vis-à-vis growing tendencies of Islamophobia in European societies that weigh on certain categories of the population especially,110 it declined to read the bans as discriminatory on the grounds that they pursued a legitimate aim—one that it famously designed for the occasion, i.e., the conditions of “living together.”111 Although these rulings upholding the burqa bans are emblematic, there are many other examples in which the court has declined to find facially neutral laws and policies discriminatory. In that, the ECHR’s case law on religious freedom illustrates that the absence of classification is enough for rejecting a claim of discrimination. The court simply does not look beyond that static conclusion. It does not engage in the more dynamic assessment in terms of (anti-)subordination.

Overall, it seems that European legal standards on religious discrimination are not significantly higher than those defined by the United States Supreme Court that fail to grasp the racially unequal disparate impact of a number of facially neutral legal rules.

B. Shielding Discrimination from Equality and Anti-Discrimination Law

It is particularly striking when American constitutional law and European human rights or anti-discrimination law are compared, to see how prompt judicial actors are to invoke a number of legal principles that actually operate as shields preventing the affirmation and application of the Equality Clause (U.S. law or of anti-discrimination law). It is thus not uncommon for courts to refer either to a particular understanding of the public/private divide or to modes of articulation of state/federal (U.S. law) or domestic/transnational (European law) in order to contain and eventually refuse the application of equality and anti-discrimination law.112

1. Public/Private Divide as a Limit on Equality Law

To what extent do anti-discrimination norms regulate private relations? The affirmation of anti-discrimination law in the field of employment has marked a major upheaval in terms of limiting private individuals’ contractual freedom and autonomy; and in the United States, this is one of the main reasons why the Civil Rights Act has appeared to many as revolutionary as it regulates private action.113 Consequently, the paradigm of equality and non-discrimination certainly does apply to a wide host of private relations (employment, housing, provision of goods and service, etc.);114 but as it is, the broader theoretical question of its legitimacy to apply legally to these private relations remains somewhat disputed. With respect to the questions examined here, one question that remains unsettled is the extent to which—or the circumstances under which—(if at all) private organizations (associations, corporations, unions, clubs, etc.) can discriminate on the basis of religion when they make employment or membership decisions? This is certainly a vexing and complex question.

Philosophically, it was strikingly captured early on in the development of the legal paradigm of non-discrimination. In her controversial intervention à propos the Little Rock events,115 Hannah Arendt suggested that “enforced integration is no better than enforced segregation,”116 and subsequently sketched a distinction between the political sphere in which strict equality should be enforced, and the social sphere that she defined as a “hybrid . . . between the political and the private,”117 in which, to the contrary, discrimination ought to be the governing principle. Strikingly, this Arendtian notion that the law should acknowledge its limitations and refuse to impose equality, tolerance, and pluralism in a number of social spaces finds strong echoes in equality and anti-discrimination law—both in the United States and in Europe.

In the United States, references to the public/private divide have been read as commanding limits to the actual scope and application of the Equal Protection Clause since its earliest stages. For many authors, constitutional law should be read as prescribing that discrimination is unconstitutional only in the realm of state action (the constitutional prohibition on discrimination applies only to governmental action), and that discrimination in the private sphere remains outside the scope of the Fourteenth Amendment (which does not empower Congress to pass laws about equality unless they are under the competence of Article I of the original Constitution—usually the Commerce Clause). Hence the highly sensitive issue of the determination of the limits of state action.118 Hence, also, the importance of a number of constitutionally protected rights that shape and indeed restrain the scope of anti-discrimination law. Neil Gotanda explains that “[u]nder the color-blind mode of constitutional analysis, freedom of contract, freedom of association and speech, and free exercise of religion protect certain racially based acts when made in the private sphere.”119 This has appeared in broad light in a number of controversial rulings limiting the scope and applicability of equality and anti-discrimination law, such as in the Boys Scouts of America v. Dale case,120 where the Supreme Court held that the organization had the right to exclude homosexuals from membership based on freedom of association and despite the anti-discrimination law requirement requiring equal treatment in public accommodations.121

The invalidation of the civil rights legislation in 1883 rested precisely on the notion that the Fourteenth Amendment did not empower Congress to regulate private conduct and private rights.122 The core of the constitutional challenge in that case was not so much directed towards the merits of the 1875 Civil Rights Acts that prohibited racial discrimination in public accommodations, but rather towards whether Congress had the authority to pass the Act in the first place. As he argued that the Equal Protection Clause could not be a valid ground for the said Act since the former’s purpose was to abolish slavery whereas the latter intended to tackle segregation,123 Justice Bradley insisted on the difference between legal and social equality. While legal equality might have become a constitutional requirement, he insisted that social equality was a matter of private choice that ought to remain beyond the reach of governmental intrusion.

Later, the ruling in Plessy also rests on some understanding of a private sphere that ought to remain protected from governmental intrusion in the name of equality. Justice Brown’s opinion is thus based on a distinction between political and social equality—only the former being acceptable, while the latter is manifestly out of reach of the law:

The object of the amendment was undoubtedly to enforce the absolute equality of the two races before the law, but, in the nature of things, it could not have been intended to abolish distinctions based upon color, or to enforce social, as distinguished from political, equality, or a commingling of the two races upon terms unsatisfactory to either.124

Even Justice Harlan’s articulation of colorblindness did not fundamentally challenge the consequences of this public/private distinction on the containment of equality law. Scholars that have thoroughly analyzed his opinions converge in identifying two among them that contrast with his otherwise consistent colorblind creed. One of them pertained to a state law punishing interracial adultery more severely than same-race adultery, and the other concerned a decision by a local school board to close a school attended by Black students for financial reasons, while maintaining another school that catered to white students. Linda Przybyszewski reads these two opinions as explainable precisely by the fact that Harlan viewed both the issues of schooling and marriage as too intimate to fall under the scope of equality law: “Although Harlan moved public accommodations out of the category of social rights and into that of civil rights, he shied away from doing the same with the more intimate contact of schooling, which might result in friendships and more, and marriage, in which racial identity would indeed be lost.”125

In fact, this is also the reason put forth by numerous scholars who account for the belated granting of certiorari by the Supreme Court to a case challenging anti-miscegenation laws. The Court in fact declined to accept one such case right shortly after ruling on Brown. Says Andrew Kull: “Naim v. Naim was an embarrassment only because it was out of the question, in 1955, that the Court uphold the constitutionality of a law prohibiting interracial marriage”126—but also that it strike it down. As is well known, it is not before 1967 that the Supreme Court struck down such statutory prohibitions on interracial marriage.127

Such modes of reasoning are, again, echoed in recent developments of European law pertaining to religious pluralism. In the Achbita case for instance, the CJEU chose to rely exclusively on the employers’ right to conduct a business elevated by Article 16 of the EU Charter of Fundamental Rights (CFREU),128 as it ruled that an employer’s choice to implement an internal neutrality policy prohibiting employees from expressing any kind of belief in the workplace was legitimate in principle. This mode of reasoning not only chooses to rely on a right of corporations to determine their image (akin to what the U.S. Supreme Court could refer to as “freedom of expressive association”), it also failed to make any reference to other rights elevated by the CFREU such as workers’ freedom of religion129 or the right to fair working conditions.130

As it chose to only refer to the right to conduct a business, the court effectively positioned a shield between the company and anti-discrimination law on the basis of the respect that is legally due to private employers’ choices. Such protection of the private choice via a reference to employers’ and companies’ fundamental rights within the framework of anti-discrimination law effectively insulates the workplace from any robust understanding—let alone application—of anti-discrimination law. In other words, this is a clear example of the ways in which the very framing of the issue as a “private” issue, whose resolution ought to be determined by the legitimacy of the autonomous choice of an employer to display an image of neutrality, precludes the full application of anti-discrimination law. It is all the more true that the sister ruling to Achbita delivered by the court on the very same day131 only reinforces this interpretation. The combination of the two rulings does indeed read as follows: although they ought not to restrict their employees’ freedom of religion on the grounds of their customers’ prejudiced preferences, employers (and corporations) can legitimately anticipate such prejudices by internalizing neutrality policies—thereby avoiding any “risk” of interfering with their employees’ expression of beliefs and convictions.132 In other words, the CJEU has effectively given companies a blueprint that protects them from claims of religious discrimination from the moment they adopt an internal neutrality policy. The result of these important interventions is thus one that delineates “the workplace” as a specific social space that can well be shielded from any robust intervention of anti-discrimination law with respect to religion. The fact that this particular social space is by definition a private one, and that the precise legal artifact that effectively shields it away from the operation, and application of anti-discrimination law is the right to conduct a business are further elements that confirm that, in Europe too, the public–private divide is a central element of the limitation of the scope of anti-discrimination law.2.

2. Local Autonomy as a Limit on Equality Law

There is another cross-cutting legal concept that similarly plays out to restrict and contain the reach of equality and anti-discrimination law principles: that of local autonomy. Both American and European law swarm with cases in which judges have used that concept in order to hold back equality and non-discrimination clauses.

In the United States, the long constitutional history of segregationist legal rules is replete with rulings whereby federal judges, including Supreme Court Justices, have declined to strike them down on the grounds that the issue at stake (be it the regulation of public transportation, educational policy, or other such issues) and related matters belong to the various states. This mode of reasoning is of course structural to American law, given the federal nature of the legal system. It remains noteworthy that it plays out massively on the issues that this Paper contemplates, even though equal protection requirements have been held to apply to the federal government.133 It is also worth noting that “[d]e jure segregation was primarily enacted as a matter of state law[,]”134—hence the particular acuteness of the issue of whether federal constitutional law constrained and even forbade legal segregation at the state level.

Again, going back to those opinions written by Justice Harlan that do not match the vibrant colorblind creed he expressed in his famous dissent over the Plessy case,135 several scholars consider that one of reasons why Harlan did not oppose the decisions of a local board of education to close a school that catered to Black students for financial motives, while maintaining the school attended by white students, had to do with the fact that schooling was, at any rate, a matter belonging to the respective states.136 This rationale remains essential throughout much of the constitutional struggles over segregation.

Here again, similar patterns are observable in European law and in particular, in the case law of the ECHR whose religious freedom jurisprudence preserves a significant role of domestic determinations. Here again, there is a structural element to this: the doctrine of the national margin of appreciation has been an instrumental part of the Strasbourg case law since its early years,137 and classically allows the court to loosen the hold of European law over domestic determinations and defer to them.138 For instance, while the Convention-protected right to private and family life requires that some form of union be legalized and accessible for same-sex couples, states typically remain free to restrict marriage itself to heterosexual unions.139 Similarly, although States remain free to consider surrogacy arrangements legal or not, it stems from the right to private life that a child who is legally born out of such an arrangement abroad is entitled to the legal recognition of their parents (genetic and intentional).140

However, as it has been noted, the ECHR’s Article 9 (religious freedom) seems to be one of the Convention articles whose judicial interpretation carves out a particularly important space for states’ margin of appreciation.141 In fact, not rarely does the court explicitly refer to the States’ “wide”142 margin of appreciation in matters pertaining to religious freedom. As a result, States have been found to legitimately: require that people appear bareheaded on their drivers’ license or passport photographs;143 or prevent schoolteachers,144 university students,145 and school students,146 as well as public nurses147 from wearing a hijab or full-face veils.148 Each time, the court has found that the States’ decision did not exceed their margin of appreciation.

Although the ECHR legal system is emblematic of the ways in which the concept of local autonomy, under the guise of the national margin of appreciation, operates as a shield to the full application of freedom of religion and the non-discrimination principle, similar observations can be made with respect to EU anti-discrimination law. As several authors had underlined well before the CJEU was given opportunities to interpret “religion” under directive 2000/78, the EU legal order was bound to award some discretion to Member States under anti-discrimination law—if only because of the necessity to respect and allow the significantly divergent domestic approaches to anti-discrimination law in general.149 Mark Bell, for instance, evoked the court’s commitment to “moral subsidiarity” in the actual operation of EU anti-discrimination law as an important element in accounting for the failure of transnational anti-discrimination legal mechanisms to impose a binding rule of recognition of same-sex marriages.150 Similar considerations were bound to infuse the reflection on religious non-discrimination if only because of the considerable normative diversity there is on the particular topic throughout EU Member States. In the Netherlands, for example, it has long been deemed unthinkable to refuse to employ a Muslim woman as a public school teacher just because she wears a headscarf,151 whereas in several other countries, such as France,152 or, to a different extent, Germany,153 public school teachers are or can routinely be prevented from any expression of religion.

In sum, some of the lines of criticism that have been addressed to the judicial response to racial inequality and segregation in the United States do seem to fruitfully enlighten some aspects of the legal response that issues of religious pluralism are receiving in Europe today. In particular, the somewhat cynical and purposeless formalistic interpretation of equality and non-discrimination clauses as well as the readiness of judicial authorities to mobilize various forms of cross-cutting legal concepts (such as the public–private divide or the federal/transnational–local one) similarly result in weakening the strength of these on both sides of the Atlantic. If such similarities have to do with judges’ modes of reasoning, it is noteworthy that further ones can be identified in the outcomes that they produce.

III. Outcomes

Here again, the starting point of this attempt in meta-comparative analysis lies in the critical theoretical work that has been produced by various strands of American legal scholarship, including critical race theory, with respect to the concept of colorblindness. It is contended here that some of the insights offered by this literature may fruitfully be transposed to the contemporary developments of European law in the face of religious pluralism—especially as they confirm the centrality of the concept of neutrality. There are two main lines of analysis that I thus wish to borrow from critical legal scholarship vis-à-vis the role of colorblindness in the face of racial inequality. First, the affirmation that despite leads to the contrary, colorblindness does not prevent nor eradicate racial othering.154 Second, the notion that colorblindness conveys notions of racial purity that are or have been instrumental in the preservation of racialized forms of privilege and supremacy. It will be argued here that European neutrality too can be analyzed along those lines, as it contributes to the othering of Muslim minorities and also rests on some versions of purity and “sanitization” of a number of social spaces.155

A. The Othering Intrinsic to Colorblindness and Religious Neutrality

Harlan’s famous dissent in the Plessy case is instrumental to the demonstration that colorblindness is not an antonym to racial othering. Harlan may well have argued that the Constitution “neither knows nor tolerates classes among its citizens.”156 The colorblind norm that he subsequently elevated can be read, following Ian Haney López’s work on “race” as a “relational category,” i.e., one that is not absolute but rather relative.157 In that sense, “colorblindness” as a constitutional norm could still be accommodated with the empirical (social, moral, etc.) belief that races did exist—and ought not to be coerced into mingling. As a consequence, Harlan’s point might have been a legalistic one (the Constitution ought not to know or tolerate racial classes) but certainly did not extend beyond the legal sphere of reasoning. Nothing in Harlan’s dissent is incompatible with the notion that races do exist to the extent that they organize social life. He merely prescribes what importance and role they should (or should not) be given as far as legal taxonomy. At any rate, other segments of his famous dissent famously support the view that he himself was hardly convinced or committed to the principle he articulated.

To a certain extent, this ambiguity on which Harlan’s dissent lies also has infused much of the later debates on sex and gender. As discrimination on the basis of sex was gradually prohibited in the United States and Europe as in many other parts of the world, a number of judges and legal authorities were led to affirming that any classification on the basis of sex was to be struck down. A robust anti-classification approach of sex discrimination does not, however, dismantle the very existence of the sexes. Neither the admission of women into Virginia Military Institute (United States)158 nor the inclusion of men in post-natal benefits (EU)159—to take but isolated examples—have fragilized or destabilized in any way the classification of people as either men or women that thus continues to govern American and European law. In fact, the very existence of the sexual binary was only later challenged as the issue of the rights of the LGBTQIA emerged.160 And even then, it can be argued that only the claim by intersex people that an additional category of sex be recognized in order to account for their specific “neither-nor” identity really jeopardized a binary approach to sex. By contrast, indeed, the rights of transgender people to change their legal sex for instance (or to marry, give birth, etc.) do not question the existence of the two categories of male and female; they only imply that people be allowed to circulate from one to another. The claims of intersex people are more radical in that sense, for they question and potentially dismantle the binary itself.161 To a certain extent, one might consider that the real challenge for the sexual binary is the intersex—and potentially then, that the real challenge for racial categories are not the existence of races but that of mixed-race individuals.162 As the detour via the concept of sex shows, a sex-blind approach to legal norms does not necessarily entail the erasure of sexual classes; similarly, a colorblind reading of the Constitution does not necessarily entail that no racial classes exist.

In fact, other excerpts of the famous dissent by Harlan confirm this notion that his colorblind creed was perfectly compatible with racial distinctions and othering. His words relating to the “Chinese race”163 are telling in this respect, as he describes it as “a race so different from our own that we do not permit those belonging to it to become citizens of the United States.”164 He further insists that: “[I]n the eye of the law, there is in this country no superior, dominant, ruling class of citizens,” even though:

[T]he white race deems itself to be the dominant race in this country. And so it is, in prestige, in achievements, in education, in wealth and in power. So, I doubt not, it will continue to be for all time, if it remains true to its great heritage and holds fast to the principles of constitutional liberty.165

As stems from these excerpts of Harlan’s famous articulation of colorblindness, this concept does not do away or erase racial categories. Time certainly went by since Harlan famously dissented. It is, however, unclear that all derogatory ways of not only distinguishing races but also inferiorizing some would have become a thing of the past. Contemporary constitutional law as it applies, for instance, to American Indians, appears to be a case in point.166

With this failure of colorblindness to dismantle racial categories in mind, it is fruitful to mobilize Ian Haney López’s research on early American immigration and naturalization law—and in particular, the ways in which he establishes the relational nature of racial categories that it produced and heavily relied on.167 López establishes that the legal construction of whiteness was largely the result of labeling candidates to naturalization as “non-whites.” The two categories were largely interdependent; the explicit production of the one (non-whites, in López’s study) simultaneously generated the other.

This goes for both the categories themselves and for the features that were simultaneously associated with them: the production of a largely encompassing category of non-whites in naturalization law168 not only also results in the shaping of the opposite category of whites, it also generates a discourse of denigration and validation, inferiority and superiority:169 “because races are constructed diacritically, celebrating Whiteness arguably requires the denigration of Blackness.”170 In other words, the legal construction of racial categories is part and parcel of processes of otherization: as one category is defined and assigned a number of features, another is also oppositionally produced and assigned features. In that, to the extent that colorblindness does not challenge the existence of racial categories, it also does not prevent, halt, or hinder the othering processes that lies at the very principle of racial categories.

Here again, the parallel reading of European developments in the field of religious neutrality requires some effort, for neutrality does not say much about religious classifications. There is, however, considerable traction to the notion that contemporary developments in European law that rest on this concept of neutrality do a lot of work in producing religiously othered communities in Europe; a conclusion that holds especially true for Muslim minorities. As I have established here above, European neutrality has been interpreted over the past two decades so as to allow, if not legitimize, a host of legal restrictions on religious freedom such as policies of neutrality in the workplace, bans on the niqab, bans on the expression of religious beliefs in public schools, etc. Oftentimes, the reasons put forth by domestic and transnational legal actors alike for justifying or upholding these restrictions have either intended to or resulted in “othering” the people and communities who claimed the freedom to engage in the said practices.

This has operated throughout the pitting of a “true meaning” of a particular practice associated with Islam such as, say, veiling against values deemed central to European/Western polities such as: gender equality, secularism, or other more generic and unspecified “republican values.” This, of course, raises a number of issues. One is epistemological and semiotic in nature, for the very possibility of interpreting a sign or practice such as the Islamic veil (or any other religious practice) and associating it with one stable meaning can be doubted.171 Another is legal: as most European legal orders affirm a version of the doctrine of separation between churches and state,172 it follows that in principle, it is not for secular institutions to interpret, label, or qualify religious beliefs and practices. Despite these serious arguments, however, “the veil” is often given a meaning in European legal discourses that is then pitted against either gender equality, or secularism, or yet other European legal values. Famously in its inadmissibility decision in Dahlab v. Switzerland, the ECHR indicated that the circumstances of the applicant (a converted schoolteacher whose students were aged between four and eight and thus deemed “easily influenced”) led to a situation where:

[I]t cannot be denied outright that the wearing of a headscarf might have some kind of proselytising effect, seeing that it appears to be imposed on women by a precept which is laid down in the Koran and which . . . is hard to square with the principle of gender equality. It therefore appears difficult to reconcile the wearing of an Islamic headscarf with the message of tolerance, respect for others and, above all, equality and non-discrimination that all teachers in a democratic society must convey to their pupils.173

In Ebrahimian v. France, the court accepted the notion that State neutrality can validly ground restrictions on private individuals’ religious freedoms.174 In Achbita,175 the CJEU upheld internal neutrality policies based on employers’ right to conduct a business and therefore shape and choose an image of neutrality for their business as valid grounds for restrictions to religious freedom in the workplace. In some countries, yet, other values are opposed to religious freedom. In France, the concept of “essential Republican values” 176 has significantly become quite central to acquisition of nationality cases. There are a number of cases in which judges have considered that women wearing the niqab,177 men whose wives wear the niqab,178 or women refusing to shake hands with representatives of the State during a naturalization ceremony on religious grounds179 engage in “radical” practices of religion that may lead to their being denied access to French nationality on the grounds that such radical practices run counter to “essential values of the Republic.”180

Such oppositions between a host of Islamic practices and Western values have thus become significantly present in legal and judicial discourse. They have fed into a process of othering in which legal rules have played an important role. As I have argued elsewhere,181 the very framing of the issue of the niqab as a public policy issue was emblematic of such othering processes; the many steps that led to the 2010 Act that effectively prohibits the concealment of the face in public (and therefore, the wearing of the niqab)182 consistently “othered” radically pious Muslim women. The parliamentary reports that were written throughout the course of the legislative process are very revealing in that perspective.183 For instance, rather than explaining that they sought to tackle the wearing of the niqab as an emerging practice hic et nunc, they described it as a practice imported from the Middle East.184 The very choice of words is illuminating of the process of othering that was at stake in that it was the burqa that was consistently referred to rather than the niqab185—although the burqa was (and remains) virtually unseen on French soil. The choice to refer to one rather than the other creates overtones connoting to the Taliban regime in Afghanistan where women were forced into wearing the burqa from the mid-1990s onwards.186 Another illustration of the many forms of othering that the 2010 Act simultaneously produced and reproduced can be found in the sanction mechanism that it includes.187 The main penalty is financial: women who fail to conform to the prohibition to conceal their faces in public can be subjected to a 150-euros fine.188 But they can also be sentenced to enrollment in a so-called “citizenship internship” (stage de citoyenneté). Such training sessions were initially created in 2004 as an alternative to jail sentences applicable for mild criminal convictions.189 The fact that they were extended to women wearing the niqab is quite interesting in terms of what it reveals from the legislator’s prejudices: if women wearing the niqab are to be trained as citizens, this either means or implies that they are not regarded as citizens in the first place—or minimally, that they are regarded as bad or failed citizens in need of training. Again, a strong othering process is thus at stake here.

This othering process is not merely religion-based. As the secularist norms that disseminate throughout Europe are increasingly questioned in terms of their potentially oppressive or discriminatory impact, some underline the racialization of religion processes that are at stake. Joan W. Scott for instance has described some iterations of secularism as:

[A] mask for the political domination of “others,” a form of ethnocentrism or crypto-Christianity . . . . Its claim to universalism . . . has justified the exclusion or marginalization of those from non-European cultures . . . whose systems of belief do not separate public and private in the same way . . . .190

Because of the dominant focus on Islam whenever issues of religious pluralism are raised, and because Muslim groups throughout Europe are mainly of foreign origin, religious minorities are readily associated with racial categories—by those who discriminate as well as, potentially, by the legal norms that purport to redress such harms. Looking at European law, Titia Loenen establishes that “banning religious symbols in public schools is not really just about guaranteeing the denominational neutrality of schools, but about how the majority society perceives and reacts to the position of its immigrant minority groups of non-Western descent.”191 This is even clearer in countries such as the United Kingdom where some of the domestic legal protection against religious discrimination operates through the Race Relations Acts,192 as the famous JFS case193 has shown (an Orthodox Jewish school was found to have discriminated on the basis of race when denying the application of a child whose mother, having converted under the auspices of a non-Orthodox synagogue, was not recognized as Jewish).194

But such processes also operate at a deeper level. The first paragraphs of Advocate General Julianne Kokott’s opinion in the Achbita case are troublingly illuminating in this respect; she opens her remarks delivered in May 2016, arguably at the peak of what Europe was then experiencing as a “migration crisis” with the following words:

There is no need to highlight here the social sensitivity inherent in this issue, particularly in the current political and social context in which Europe is confronted with an arguably unprecedented influx of third-country migrants and the question of how best to integrate persons from a migrant background is the subject of intense debate in all quarters.195

The mere fact that she felt it relevant or useful to contextualize a discrimination in the workplace case against the “migration crisis” speaks for itself. There is little doubt that what this reveals is that Muslim minorities in Europe, even when they are integrated and work, are apprehended as coming from migrant backgrounds and in need of (more) integration. And there is little doubt that, had the case not involved a veiled Muslim worker, Advocate General Kokott would not have chosen such elements of contextualization in her opinion. At any rate, those words effectively “other” Ms. Samira Achbita and the likes of her—a point powerfully made, among others, by Joseph Weiler.196

As has been noted by Mathias Möschel, this opinion is also interesting in that it is underwritten with a process of “white-washing of intersectionality.”197 Although the issue at stake was the dismissal of a female Muslim employee wearing a headscarf, Kokott failed to see the relevance of the concept of intersectional discrimination; rather, she chose to contextualize the case with respect to the issues of migration and security at the European level. Only one month later, however, the same Advocate General was to mobilize the concept and the literature on intersectionality as it offered its reading of a case involving pension rights of elderly gay men, thus proving that she both knows the concept and retains it as relevant in concrete judicial work.198 Her failure to mobilize it in the Achbita case only henceforth acquires more relief.

B. Overtones of Purity Conveyed by Colorblindness and Religious Neutrality

The previous Section has thus established that the legal racial categories that remained acceptable and relevant for a long period of modern American constitutional law were not weakened or challenged by the concept of colorblindness. Further, I now wish to insist that because racial categories were themselves a “metaphor . . . of purity and contamination,”199 colorblindness also fed into notions of purity. The biological overtones of social and legal discourses on race are certainly metaphorical, for race is a social construct. The one blood drop rule according to which any trace of African ancestry makes one Black,200 for instance, essentially served to define who is white and why one is indeed. In his thorough analysis of the “pre-requisite cases” and early American naturalization laws,201 Ian Haney López has powerfully established that these determinations actually relied on common knowledge more than on science, even in the judicial discourse; judges make determinations about immigrants’ race (white or non-white) based on “common knowledge.”202 Historically, “what people believe” (and, be it the case, an explicit rejection of science) was thus the real foreground of the legal construction of race.

These racial categories were further based on the notion that their purity ought to be maintained. This is clearly expressed by the famous Supreme Court ruling striking down anti-miscegenation laws: as Justice Earl Warren wrote for a unanimous court in Loving v. Virginia, he explained that state bans on interracial marriage were “designed to maintain White Supremacy.”203 Anti-miscegenation laws were indeed quintessentially emblematic of the dimension of purity that infused much of American segregationist law and policies. The desire to make human reproduction across racial boundaries illegal proceeded from a desire to maintain racial purity; it followed that: races should not miscegenate, and white and Black blood must not be mixed for racial purity (and in fact, the purity of the white race) to be preserved.204 Neil Gotanda captures this powerfully: “White is unblemished and pure, so one drop of ancestral Black blood renders one Black. Black ancestry is a contaminant that overwhelms white ancestry.”205

Because the existence of distinct and hierarchized racial categories could well be accommodated within the paradigm of colorblindness, so could these notions of racial purity and white supremacy. Important works in the field of critical race theory have indeed presented colorblindness as an actual vehicle of white supremacy. In its more contemporary iterations, colorblindness certainly does not explicitly endorse any notion of purity or supremacy, for indeed, the one “moral triumph” the civil rights movement has enjoyed lies precisely in the dismantling of the legitimacy of white supremacy.206 At the same time, however, it appears that “the dominance of Whites across the range of social, political, and economic spheres continues . . . .”207 Hence numerous authors make the correlation between the successes of colorblind ideology in constitutional interpretation and the preservation of the racially unequal status quo.

Again, this Paper wishes to suggest that some articulations of religious neutrality in Europe also have overtones of purity. Some legal responses to the issues of religious pluralism can indeed be read through a lens of purity—albeit a social more than biological concept of purity.

One example can be taken from a recent ruling by a French administrative appellate court. A young female Muslim adult enrolled in a vocational training session that was to take place inside a building that was otherwise occupied by a public high school. Since the Act of 15 March 2004 (2004 Act), students of public elementary, middle, and high schools are prohibited from manifesting their religious beliefs.208 Based on that prohibition, the school principal denied access to the school to the female adult who wished to attend her vocational training session, and she challenged his decision in court. The court eventually upheld the principal’s decision, arguing that the simultaneous presence within the one building of students who are under a prohibition to express their religious beliefs and one young adult wearing a hijab could indeed cause trouble and threaten the public order internal to the school.209 Two elements seem particularly noteworthy in this reasoning.210

First, what is at stake is the “simultaneous presence” in the school of “students” subjected to the 2004 ban and “one young intern.”211 This line of reasoning strikingly conveys a sanitary paradigm that implies that there should be no contact, no mixing of any sort, between high school students and any number of persons (including a single person) wearing a visible religious sign such as the hijab. It also implies that schools are to be religion-free zones and that their purity as such could be jeopardized by the presence of one person wearing the hijab.

Second, this ruling conveys a spatial interpretation of the 2004 Act that produces the norm of schools as religious-free zones. Technically speaking, the Act reads as one that should apply to students qua students—not to schools as buildings or spaces. It is true that, in France, as civil servants in general are subjected to a requirement of religious neutrality and are thus prohibited from expressing their religious (or other) beliefs in the workplace, most people who work in public schools are religiously neutral; this is true for teachers and professors, but also principals and janitors.212 To that extent, as it also subjects the students to the same kind of prohibition, the 2004 Act can be read as banning religion from public schools in general; and indeed, as a number of legal and social actors immediately sought to apply the Act in an expansive manner, one of their rationales was to interpret it as applying not so much to students as users of the public service of education, but rather to schools as spaces that ought to remain religion-free.

A number of schools thus sought to amend their internal rules so as to insert requirements of neutrality applicable to the wider educational community that gravitates around public schools, including external collaborators and, above all, parents; in several instances, it became required from parents who had offered to take part in school activities (typically, by accompanying class outings to museums, cinemas, field days, etc.) that they subject themselves to a rule of religious neutrality. But litigation ensued, and although courts have addressed this issue in many different voices,213 it was established by the Conseil d’Etat in its advisory capacity that the 2004 Act could not be read as a valid legal ground for subjecting parents to such an obligation of religious neutrality.214 The Act only subjects students to such an obligation. The tendency to read it spatially as regulating schools as spaces rather than children as students persisted nonetheless, as illustrated by a recent ruling by the Lyon appellate administrative court that referred to “a requirement of religious neutrality imposed on parents . . . that only applies when they participate in activities that take place inside classrooms and during which they exercise functions comparable to those of teachers.”215 These examples illustrate the ways in which the 2004 Act has partially been hermeneutically pulled towards a notion of schools as pure, cleansed spaces; as harbors that are safe in that they are protected from any expression of religious belief—thus characterizing some interpretations of laïcité (neutrality) as related to notions of purity.

Other examples can be found where the overtones of purity are read into biology (blood and citizenship) rather than space. In Germany, for instance, a first instance judgment in an early case pertaining to schoolteachers’ right to wear a headscarf in the workplace is interesting in this respect. Although the plaintiff eventually lost her case, the initial judgment considered a number of elements, among which the fact that the plaintiff was a German-born woman who had been educated in the Lutheran faith was taken as an indication that no radical or fundamentalist intentions should be ascribed to her choice to cover her head.216

Conclusion

This Paper has argued that some of the analyses of the limits and indeed adverse effects of the ideology of colorblindness in American constitutional law on racial equality are worth transposing in the contemporary European debates on religious pluralism. The Paper’s starting point is the strong and parallel concerns that while colorblindness might have entrenched rather than corrected racial inequality in the United States, some versions of the contemporary elevation of “religious neutrality” in Europe may be producing similarly inegalitarian outcomes. After clarifying the proposed level of analysis (that of a meta-comparison), the Paper first compared American and European judicial reasoning on issues of race and religion, and then examined outcomes. It confirmed that many similarities could be drawn, as judges across the Atlantic rely on similarly formalistic and a-teleological patterns of reasoning that privilege private ordering and judicial deference to local choices. As a result, processes of othering and overtones of anxieties over the preservation of certain ideas of “purity” that have been documented to infuse American race law can also be found in European religion law.

As the demonstration unfolds, the Paper confirms a familiar trope of comparative methodology: comparing a given model with another renews the observations and analyses one makes about the former. In that sense, the comparison of contemporary European law as it confronts issues related to religious diversity with ongoing North American debates over colorblindness and racial equality gives European scholars an opportunity to ask questions about the European case that might otherwise not be asked—especially as they encounter myths and taboos around the issues of race, inequality, and purity that have been positioned more conspicuously in the U.S. debate. This is not to say that this conversation in the United States is easy or soothed in any way, even in its academic version. But it does take place, whereas in many European countries, it is made all the more difficult that many actors continue to toy with the project of deleting the word “race” from their foundational texts and legislation,217 and actively combat the elevation of the word “Islamophobia,” including in human rights bodies and institutions.218 European countries also react strongly and adversely against the move of international human rights towards new paradigms in the field of religious discrimination, as shown both by the strong criticism after the United Nations Human Rights Committee found that some disproportionate infringements on religious freedom amounted to intersectional discrimination219 and by the consolidation of a significant divergence between the UN approach and that of the European courts of justice.220 In other words, the comparison with the American constitutional debate operates as an enabler, a facilitator for opening an uncomfortable conversation.221

As uncomfortable as it may be, this conversation is much needed. In the United States, as well as in Europe today, overt instances of racial or religious discrimination are becoming rarer. “Colored only” signposts, as yellow stars attached to coats and jackets, are hopefully a thing we now only see in cinematographic representations of the past or read of in literature. By all accounts, though, racism and prejudice have not disappeared. White supremacy (in both its racial and religious dimensions) endures and requires that we tackle and confront its unconscious dimension.222 This can only be done if and when we unearth and confront the problematic exclusionary dimensions of even those constitutional principles that we cherish most for their historical and structural centrality to our polities.

 


* Professor of Law, Université Paris Nanterre and Senior Member, Institut Universitaire de France, Deputy Director, Centre de théorie et analyse du droit (UMR 7074) – CREDOF. Email to: stephanie.hennette-vauchez@parisnanterre.fr. This Paper was presented at the IACL Conference in honor of Norman Dorsen that was held at Cardozo Law School March 31–April 1, 2019. The author wishes to thank the organizers for the invitation to be part of that event, as well as the participants for useful feedback and questions. Preliminary aspects of the Paper were also presented at the international conference on Critical Race Theory that was organized by Magali Bessone and Isabelle Aubert at the Philosophy Department of Université Paris 1 Panthéon Sorbonne in January 2019. Here again, rich discussions and exchanges were instrumental to the development of the present Paper. I particularly want to acknowledge early discussions with Daniel Sabbagh for encouraging me to move forward with this project, as well as very useful comments on an earlier version by Eric Christiansen and Mathias Möschel. This Paper owes a lot to these three colleagues, even though all imperfections remain mine.