The Egenberger case neatly illustrates the tensions underpinning the approach of the Court of Justice to the question of religion in the workplace. Cases in this area bring together two areas of law in which the CJEU has taken markedly different approaches. This has left the Court torn between following its generally assertive approach in relation to discrimination in the workplace and its generally deferential approach in relation to questions around religion’s role in society. The result has been an approach that has accorded more leeway to Member States than in other areas of workplace discrimination, while also using discrimination law to set down parameters that place some limits on the choices that Member States can make in how they regulate religion’s place in their societies. While this caution has been heavily criticised, in the context of the inevitable uncertainty produced by rapid and unprecedented religious change in Europe, it is the most prudent and politically sustainable approach for the time being.

Employment discrimination and freedom of religion

In relation to employment discrimination the Court has a long history of dynamic and bold interpretations of EU law. As far back as the 1970s, in cases like Defrenne, the Luxembourg judges interpreted the principle of equal pay for equal work in an adventurous fashion that revolutionized the approach to the equality of men and women at work as well as pushing forward the process of European integration. This adventurousness has persisted into later decades in cases such as Mangold where the Court showed a notable willingness to push the boundaries of interpretation in order to promote the principle of non-discrimination.

In relation to religion, however, a notably different approach has been evident. In the early decades there were few cases of note and religion usually appeared only as an incidental factor on cases that turned on other elements. But even during the past fifteen years when cases more directly focused on religion (religious symbols at work, ritual slaughter, time-off for religious observance, ethos-based discrimination) greater caution has been detectable, even in the context of employment discrimination where elements of the Court’s case law pull it in a more interventionist direction.

The CJEU has been notably keen to give Member States considerable leeway to regulate issues relating to religion’s role in society (including individual and collective religious freedom), particularly when dealing with cases that include elements that touch on issues related to the political hot button of multiculturalism. I will suggest that this caution on the part of the Court of Justice about its ability to use its interpretative powers to identify and impose ideal, Union-wide solutions to the difficult issues that arise in relation to the place of religion in contemporary Europe is the correct approach, for both textual and pragmatic reasons.

Deference to Member States choices

Just how deferential has the Court been? The desire to interpret EU law in order to give Member States leeway to pursue different approaches in contentious areas has certainly been a feature. For example, the Court has permitted bans on the wearing of religious or philosophical symbols at work (Achbita) while also making it clear that Member States are also entitled to facilitate the wearing of such symbols if they so choose (WABE e.v.). It has also declined to interfere with prohibitions on religious slaughter (disregarding the advice of the Advocate General who urged a more interventionist approach).

This caution has been subject to significant criticism. Much of this is understandable. It is undeniable that principles such as secularism or neutrality have been used by those who have exclusionary agendas. It is also the case that for many adherents to faiths, such as Judaism and Islam, that place greater emphasis on worn symbols than neutrality rules can present more of a challenge than they do for most Christians. In addition, given that religion often overlaps with racial and ethnic identities and that in many cases the relevant religious symbols are worn by women, there is the additional factor of potential discrimination on grounds of race and sex as well as religion.

In these circumstances many have expressed disappointment that the Court has not been more protective of religious freedom and freedom from discrimination on grounds of religion or belief. Both Spaventa and Weiler were notably critical of the failure of the Court to engage in a more searching analysis of the proportionality of laws restricting religious symbols and their impact on religious individuals.

Textual reasons for a cautious approach

In contrast, there are powerful reasons pushing against the Court from acting in a more assertive fashion in this area. Textually, Article 17 of the TFEU gives a clear steer to the Court of Justice. The Article states that “The Union respects and does not prejudice the status under national law of churches and religious associations or communities in the Member States”. Thus, the Treaty is clear that there is no single model of relationships between religion and state required by the requirements of European Union membership. There must, of course, be limits to such Member State discretion. The manner in which the accession criteria have been applied indicates that a full-blown theocracy cannot join the Union. But significant diversity is acceptable, something that is unsurprising given that the states who drafted and signed the Lisbon Treaty have arrangements that range from recognition of an official state religion to official separation of religion and state. That does not mean that religion and state relations will be entirely unaffected, as the Court of Justice (para. 58) and German Federal Constitutional Court made clear in Egenberger. What Article 17 requires is that Member States’ right to have different constitutional arrangements in relation to religion be taken into account by the CJEU in interpreting EU law, not that EU law may not in any way affect how Member States deal with religion (para. 246).

Broader reasons for a cautious approach

Beyond the text of the Treaty, there are other powerful reasons for the Court to be modest in its perception of its ability to identify ideal models or universally applicable approaches to religion in the different Member States. Not only, as already noted, have Member States always taken different approaches to these issues. It is also the case that in many Member States, the situation is highly fluid with significant societal changes prompting intense political debate and significant legal and political changes in the regulation of religion’s role in society.

In relation to religion, Europe has undergone a number of major changes in the past half century any one of which would have been sufficient to produce endless unforeseen consequences. After many centuries during which a very large majority of Europeans were believing Christians, levels of religious practice and belief suddenly collapsed in most EU Member States in recent decades. As I have written elsewhere: “for centuries, most Europeans went about their day to day lives believing they were being observed and judged by the Christian God. Most no longer do. The scale of changes that that will bring about can only be imagined.” There has also been a revolution in terms of norms around sex, sexuality and gender which are challenging for most traditional religions.

The decline in Christian belief and practice has been accompanied by an unprecedented growth in non-Christian communities, with Islam being by far the most numerically important of these. In many countries the longstanding contest between Christian and secular influences has now become a multiparty contest with other religions, particularly Islam, playing a notable role. As Shadi Hamid has noted, Islam has its own rich intellectual and historical traditions. While European Christianity has, overall, followed a pattern of declining levels of belief and practice, and eventual embrace of the notion of the secular nature of law and politics, as Hamid points out, there is not reason to suppose that Islam in Europe will follow this path. Indeed recent data from France suggests the opposite is the case.

In short, we do not have a large store of precedents for how changes of the magnitude that Europe is undergoing are successfully managed. Indeed, it is notable how, in recent times, there has been significant instability in the approaches of a number of states with governments switching between more multicultural approaches that take a favourable approach to the maintenance and expression of minority identities (including religious identities) and more integrationist approaches, closer to the French model of discouraging religious expression in certain areas.

In this context, and given that most of the religion-related cases before the Court of Justice involve legislation (Directive 2000/78) that is subject to unanimity in the Council and therefore effectively almost unamendable, it is understandable that caution and providing some leeway to Member States has marked the judges’ approach. In a context of such uncertainty and rapid change it would require a remarkable degree of self-confidence for the members of the Luxembourg court to decide that they had the necessary wisdom to use their powers to interpret EU law in a way that sought to resolve longstanding and fast shifting disputes between those who see religious diversity as best managed through facilitating religious expression and those who take the opposite view and regard coexistence as best served by curtailing such expression in some contexts.

Limits to CJEU deference

Notwithstanding its overall caution, the Court has not been entirely deferential. It has been clear that any bans on religious symbols at work must be comprehensive and avoid targeting the symbols of any particular faith (Achbita (para. 40) and Bougnaoui (para. 32-33)). It has required concrete justification for such restrictions rather than abstract reasons (WABE e.v. (para. 65)). It has also made it clear that compliance with customer preferences cannot be seen as a “genuine and determining occupational requirement” that could justify direct discrimination (Bougnaoui (para. 40)). The Court of Justice has therefore married the granting of considerable leeway to Member States with a degree of supervision that rules out openly discriminatory targeting of particular faiths.

As previously noted, the reluctance to set down meaningful restrictions on Member State autonomy is less pronounced in cases that are more remote from the politically-charged scenarios where issues of religion, integration, and multiculturalism are key features. Thus, in Egenberger, where the issue was the scope that religious employers can be given to engage in ethos-based discrimination, the Court of Justice made its position clear. It held that the previous approach, taken by German law of allowing religious employers to determine for themselves, subject only to plausibility review, whether a particular role needed to be subject to a religious affiliation test, was incompatible with EU law (para. 59). The Court justified this conclusion on the basis that Directive 2000/78, which it held to be a codification of the general principle of non-discrimination, required that religious affiliation tests (as well as requirements of loyalty to the ethos of a religious employer in IR) needed to be shown to be proportionate in the context of the nature of the post in question and its proximity to the religious mission of the religious body.

As this ruling significantly affected the constitutional protection of the self-determination of religious bodies under the German Constitution (the Basic Law), it was notably controversial. Indeed, the controversy extended to calls for the German Constitutional Court to declare the CJEU’s ruling ultra vires. However, the eventual ruling of the Constitutional Court has continued the delicate dance between the maintenance of meaningful protections from discrimination by the Court of Justice with the according of significant leeway to Member States to follow their own path in religious matters. The Karlsruhe judges agreed to meaningfully alter pre-existing approaches to religious autonomy to ensure that there is an objective link between a religious affiliation requirement and the tasks involved in any particular role and that an overall assessment of the proportionality in which religious autonomy rights and employee rights to equal treatment are balanced. At the same time, the Constitutional Court maintained much of the previous approach of German law by upholding the central importance of the religious body’s own perception of the requirements of its ethos. This, as Matthias Mahlmann has noted involves a degree of “pluralism of fundamental rights” which amounts, in effect, to a kind of “margin of appreciation” in the application of EU legal norms in this area.

This is an approach that will disappoint many. Those keen on upholding broad notions of religious autonomy and the ability of religions to constitute communities of the faithful, will be disappointed by the use of EU legal norms to curtail that autonomy. Those who see facilitation of religious expression (or as others see it, adherence to religious norms) as clearly the best path to follow in multicultural societies are also no doubt disappointed that the Court of Justice has not required Member States to adopt this approach. But, in the context of the high levels of change and uncertainty that characterise matters of religion in contemporary Europe, the approach of the Court of Justice of setting wide but meaningful boundaries on Member State autonomy in this area may represent the most politically sustainable and wisest approach for the time being.