Hostname: page-component-cd9895bd7-mkpzs Total loading time: 0 Render date: 2024-12-26T18:05:21.311Z Has data issue: false hasContentIssue false

When can religious employers discriminate? The scope of the religious ethos exemption in EU law

Published online by Cambridge University Press:  06 April 2022

Martijn van den Brink*
Affiliation:
Jacques Delors Centre, Hertie School of Governance, Berlin, Germany
*
Corresponding author. E-mail: vandenbrink@delorscentre.eu
Rights & Permissions [Opens in a new window]

Abstract

When are religious employers exempt from the prohibition of discrimination (i.e., when can they discriminate against non-adherents)? The European Union (EU) Equality Framework Directive exempts religious employers from the prohibition of religious discrimination, but the scope of the religious ethos exemption is disputed and its interpretation by the Court of Justice of the European Union (CJEU) in Egenberger and IR v JQ has been criticised for being ultra vires and for disrespecting the constitutional identities of the EU Member States. This article clarifies the religious ethos exemption, by examining the underlying legal and normative issues that determine its scope. It shows that the scope of the exemption depends not just on the Framework Directive but also on the relationship between EU law and national constitutional law and that between EU law and international law. Thus, this article not only provides clarity regarding the religious ethos exemption, but also uses these judgements as an opportunity to revisit these related constitutional issues, and in particular the role of the CJEU and EU legislature in defining the place of national constitutional identity in EU law.

Type
Core analysis
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2022. Published by Cambridge University Press

1. Introduction

When can religious employers discriminate against non-adherents? Can the Church discriminate against non-Christians in the appointment of clergy? Can an Islamic school dismiss teachers who do not observe the core principles of the Islamic faith? Can a Christian hospital refuse to employ qualified doctors who are not members of the Church? European Union (EU) Member States think differently about such questions. Some grant religious employers broad exemptions from the prohibition of discrimination;Footnote 1 others have narrowly circumscribed the right of such organisations to discriminate against non-adherents.Footnote 2 However, the scope of the religious ethos exemption is no longer determined by national law alone. The EU Equal Treatment Framework Directive provides a legal framework for combating discrimination on the grounds of religion and belief, disability, age and sexual orientation, in the area of employment and occupation.Footnote 3 The Framework Directive also lists several exemptions from the prohibition of discrimination, including a religious ethos exemption that religious employers can use to justify discrimination on the grounds of religion and belief. The scope of this exemption is disputed, however, and there are different interpretations as to when religious employers are exempt from the prohibition of discrimination under EU law (see below). To clarify the scope of this exemption, this article studies the relevant legal provisions and principles. As we will see, its scope depends not only on the Framework Directive but also on our response to deeper questions of EU constitutional law. In attempting to shed light on the religious ethos exemption, this article also seeks to advance the debate on related constitutional questions and controversies.

Tensions over the exemption have mounted following two rulings by the Court of Justice of the European Union (CJEU) in the cases Egenberger and IR v JQ.Footnote 4 The Egenberger case concerned a dispute between the Protestant Church and an applicant for a job involving the task to draw up a report on the UN Racial Discrimination Convention. The applicant had not been invited for an interview because she was not a member of the Church. The IR v JQ case involved a conflict between IR (a Catholic non-profit organisation carrying out the work of Caritas) and a doctor who used to work for IR. He had been dismissed for entering into a marriage that was invalid under canon law. In both cases, the Bundesarbeitsgericht (German Federal Labour Court) asked the CJEU to clarify the conditions under which religious employers may discriminate against non-adherents. It also asked whether principles of national constitutional law could be invoked to exempt compliance with these conditions. The CJEU ruled that the scope of the religious ethos exemption must be narrowly construed and that principles of national constitutional law cannot exempt compliance with the conditions set out in the Framework Directive. These decisions led the Bundesarbeitsgericht to construe the internal autonomy of religious organisations more narrowly than before by the Bundesverfassungsgericht as a matter of German constitutional law,Footnote 5 to the dismay of many experts of German constitutional and church law.Footnote 6 The defendant in the Egenberger case, the Protestant Church, subsequently lodged a constitutional complaint with the Bundesverfassungsgericht, alleging that the CJEU exceeded the limits of EU competence and violated the constitutional identity of Germany.Footnote 7

To those not familiar with these disputes and the issues they raise, let me provide some essential legal and societal background. The Framework Directive was adopted in 2000, shortly after the Treaty of Amsterdam expanded the EU’s competence to enact legislation to combat discrimination. Until then, EU non-discrimination law only prohibited discrimination on grounds of nationality and gender, as a corollary to the EU’s ambition to establish an internal market among the Member States.Footnote 8 The Framework Directive is in part a continuation of this goal of creating a level playing field for companies, regardless of which domestic market they are active in, but it also serves as a tool to deliver social policies beyond the internal market. According to its 11th Recital, it contributes to ‘the attainment of a high level of employment and social protection, raising the standard of living and the quality of life, economic and social cohesion and solidarity, and the free movement of persons’. According to the CJEU, moreover, the Directive is ‘a specific expression … of the general prohibition of discrimination laid down in Article 21 of the Charter’.Footnote 9 So far, the vast majority of litigation in relation to the Directive has concerned age discrimination. It took a while before the CJEU was finally confronted with religious discrimination, and in terms of numbers, this protected ground has generated the fewest court cases.Footnote 10 These few cases have, however, generated some of the most controversial judgements.Footnote 11

The question in Egenberger and IR v JQ was essentially when employers can justify religious discrimination as a legitimate occupational requirement. Article 4 of the Framework Directive provides two similar, yet distinct, occupational requirement exceptions that can be used to justify discrimination. Article 4(1) lays down the general occupational requirement exception:

Member States may provide that a difference of treatment which is based on a characteristic related to any of the grounds referred to in Article 1 shall not constitute discrimination where, by reason of the nature of the particular occupational activities concerned or of the context in which they are carried out, such a characteristic constitutes a genuine and determining occupational requirement, provided that the objective is legitimate and the requirement is proportionate.

This exception covers any of the grounds protected by the Directive, not just religion or belief, and is in principle uncontroversial. Some jobs are such that differentiation on the basis of a characteristic related to a protected ground is hard to avoid. Atypical example is the choice of a modelling agency for a female model to advertise women’s clothing.Footnote 12 The exception can also be invoked to justify discrimination on the grounds of religion or belief. The same modelling agency can reject someone who insists on wearing the Islamic headscarf to model in a shampoo commercial. As we shall see, Article 4(1) can also be used by employers with an ethos based on religion or belief to exempt specific employment practices from the prohibition of discrimination.

More controversial is the religious occupational requirement exception in Article 4(2) of the Framework Directive, specifically for churches and other employers with an ethos based on religion or belief:

Member States may maintain national legislation in force at the date of adoption of this Directive or provide for future legislation incorporating national practices existing at the date of adoption of this Directive pursuant to which, in the case of occupational activities within churches and other public or private organisations the ethos of which is based on religion or belief, a difference of treatment based on a person’s religion or belief shall not constitute discrimination where, by reason of the nature of these activities or of the context in which they are carried out, a person’s religion or belief constitute a genuine, legitimate and justified occupational requirement, having regard to the organisation’s ethos. This difference of treatment shall be implemented taking account of Member States’ constitutional provisions and principles, as well as the general principles of Community law, and should not justify discrimination on another ground.

Provided that its provisions are otherwise complied with, this Directive will thus not prejudice the right of churches and other public or private organisations, the ethos of which is based on religion or belief, acting in conformity with national constitutions and laws, to require individuals working for them to act in good faith and with loyalty to the organisation’s ethos.

This provision has rightly been criticised by Ellis and Watson as ‘possibly one of the most opaque to be found on any statute book’, an example of legal compromise at its worst – sloppily worded and even apparently contradictory.Footnote 13

There are serious disagreements regarding the meaning of Article 4(2) and its added value relative to Article 4(1), and the provisions are often interpreted through a national lens. The prevailing view in the English literature seems to be that Article 4(2) ‘adds nothing’ to Article 4(1).Footnote 14 For example, Rivers has said that the difference between both exceptions is ‘impossible to grasp’.Footnote 15 But ask a German constitutional lawyer, and we will most likely hear a very different view; namely, that Article 4(2) contains a much broader exception to the non-discrimination duty. The fact that this provision refers to national constitutional law (twice!) tells them that it may be interpreted in accordance with national constitutional law, including constitutional principles that grant religious employers a broader exemption from the prohibition of discrimination than Article 4(1) of the Framework Directive. Furthermore, they often draw attention to Article 17(1)of the Treaty on the Functioning of the European Union (TFEU) in support of this view, which provides that ‘the Union respects and does not prejudice the status under national law of churches and religious associations or communities’.Footnote 16

The practical significance of such interpretative disagreements becomes clear when we examine more closely the scope of the internal autonomy of religious organisations under German law. Article 140 of the German Constitution in conjunction with Article 137(3) of the Weimar Constitution provides that ‘every religious community administrates its own affairs without interference of state or community’. Thanks to a broad interpretation by the Bundesverfassungsgericht, religious organisations enjoyed sweeping exemptions from the application of secular labour law: they are allowed to discriminate against non-adherents in all their employment activities.Footnote 17 In this context, it is essential to understand that these organisations play a crucial role within the German welfare state: they run hospitals, kindergartens, nursery homes, and the like – the two main churches combined are Germany’s second-largest employer after the state, employing around 1.5 million people.Footnote 18 And all these employees may be required to be members of the Church and act in accordance with its religious doctrines, and they may be dismissed for misconduct. This is why a medical doctor working for a Catholic hospital could be dismissed for entering into a marriage invalid under canon law, a situation unheard of in most other Member States.

To place the German law on the regulation of religious employers in its wider European context, some EU Member States provide an exemption that is similar in scope.Footnote 19 For instance, the Cypriot constitution provides for the full autonomy of the established religious organisations,Footnote 20 and Austrian law allows church-run breweries, lumber mills and hotels to recruit staff on the basis of their religious beliefs.Footnote 21 Whether the discriminatory practices of religious organisations in those countries penetrate society as deeply as in Germany is not always clear, however, as they may not play as large a role in the provision of social welfare. What is clear is that religious employers in other Member States enjoy a narrower exemption from the prohibition of discrimination. According to the Spanish Constitutional Court, religious employers may discriminate against employees only if their employment is closely linked to the employer’s ethos.Footnote 22 We find a similar standard in the case law of Dutch courts.Footnote 23 Finally, some Member States, such as Sweden and France, do not provide for a separate exemption for employers with an ethos based on religion and belief in their domestic law.Footnote 24

However, Egenberger and IR v JQ are interesting not only because they once again raise the question of how the EU should deal with moral diversity and demonstrate what far-reaching legal and social implications EU law may have in this regard. Broader lessons can be learned from these judgements for EU law – lessons that must be understood to determine the scope of the religious ethos exemption and to assess the criticism that has been levelled at these judgements. Most importantly, is it the case that the judgements are ultra vires and did the CJEU fail to observe its duty to respect the constitutional identities of the Member States? As should be clear by now, the scope of the exemption depends not just on Article 4 of the Framework Directive, but also on other aspects of EU constitutional law; in particular, on the relationship between EU law and national constitutional law and that between EU law and international law. Adiscussion of the relevant legal principles should help to clarify the scope of the religious ethos exemption and resolve existing disagreements, or at least to clarify the reasons that explain why the CJEU reached different conclusions than some of its critics had liked. Following a discussion of the normative rationale of the principle of religious autonomy in section 2, the three elements that condition the scope of the exemption will be examined in turn. Section 3 defines the relationship between Articles 4(1) and 4(2) of the Framework Directive, section 4 considers to what extent the religious ethos exemption is conditioned by national constitutional law, and section 5 assesses to what degree the influence of EU law on the internal autonomy of religious employers is constrained by international law.

2. The normative rationale for religious autonomy

Barring an appropriate justification, a person’s religious beliefs cannot normally be invoked to treat that person less favourably than other persons. What then is the justification for exempting religious employers from the obligations of EU non-discrimination law? What values does the principle of religious autonomy promote that the law seeks to protect? To ascertain the purpose behind the religious ethos exemption, we need to know the deeper values underpinning this principle. This is all the more important as the exemption is clearly at odds with the central purpose of EU non-discrimination law: it protects religious groups whose behaviour may inflict expressive harm on non-adherents and restrict their socio-economic opportunities. This section explains that the normative rationale for religious autonomy must be found in the value of religious freedom, not of religious organisations but of their individual members. This points to a deeper normative and legal tension at the heart of Article 4 of the Framework Directive, between the right to religious freedom and the right to be free from religious discrimination. Tensions arise particularly as more value is placed on the right to religious freedom, or put another way, as a broad scope of the autonomy of religious organisations is considered necessary for the protection of this right.

Intuitively, one might think that the right of religious organisations to have their internal autonomy respected is a right they enjoy because there is something valuable and worthy of protection about these organisations, as such. According to Rivers, the foundational principle behind the law on organised religions is that of religious autonomy, by which he means ‘the power of a community for self-government under its own law’.Footnote 25 However, it seems incorrect to think that the value of religious autonomy resides in the protection it affords to religious organisations qua organisations. Instead, EU non-discrimination law protects the internal autonomy of religious organisations in order to protect the individual autonomy of their members.Footnote 26 Individual autonomy is widely regarded as one of the cornerstones of liberal society, which encompasses the capacity of individuals to choose from an adequate range of valuable options without coercion or manipulation.Footnote 27 Decisional autonomy in relation to religion seems integral to the realisation of individual autonomy, for the simple reason that religion is a valuable option to many persons. In this respect, religion is like other valuable options such as the freedom to enter into social relationships with others.Footnote 28 And just as liberal societies must value the autonomy of individuals in social matters, they must, as Calderwood Norton observes, ‘value autonomy in relation to religious matters too’.Footnote 29 That is, they must respect and guarantee individuals’ freedom to choose their religious beliefs and to engage in the attendant religious practices and rituals.Footnote 30

The performative dimension of religion varies greatly from one religion to another, but religion often has a communal dimension. Religious organisations serve as a place for collective religious practice and prayer and allow individual believers to observe and pursue their deeply held religious beliefs. Individual believers thus have an autonomy-related interest in being able to participate in the services and ceremonies of their religious community. They also have an autonomy-related interest in their religious community being able to uphold its religious principles. After all, as Laborde points out, ‘a religious association that is unable to insist on adherence to its own religious tenets as a condition of membership is unable to be a religious association’.Footnote 31 Such an organisation would also be unable to provide its members with a place to practise and observe their religious beliefs. The right to religious autonomy is therefore a right rooted in the interests of the members of the organisation.Footnote 32 That is, religious organisations in liberal societies enjoy a right to religious autonomy – including a prima facie right to discriminate against non-adherents and to enforce sanctions against members employees who refuse to abide by their religious principles – to protect the joint interest of their individual members to live by their deepest commitments.Footnote 33

However, while this provides a principled justification for exempting religious employers from certain obligations under EU non-discrimination law, the question is not just whether, but also in respect of which employment activities discrimination on the grounds of religion and belief should be permissible. Underlying this question is a fundamental tension in the relationship between two fundamental rights for the protection of religion: the right to freedom of religion and the right to be free from religious discrimination. These rights often complement each other in protecting religion, but they serve distinct interests that can be incompatible.Footnote 34 As Khaitan and Norton Calderwood have explained, the right to freedom of religion is best understood as protecting our individual autonomy in religious affairs, whereas the right against religious discrimination is best understood as protecting us against the disadvantages that may result from membership of a religious group.Footnote 35 Non-discrimination law is centrally concerned with preventing differentiation between persons based on their membership of salient social groups.Footnote 36 However, the exercise of the right to freedom of religion by individual adherents, or by them collectively as part of a religious organisation, may interfere with the right to be free from discrimination and impose specific disadvantages on certain social groups – on non-adherents but also on women or sexual minorities.Footnote 37 The question then is how to balance these competing rights: when to restrict religious autonomy and when to accept discrimination?

There is a relatively straightforward answer to this question from a liberal democratic perspective. Access to important opportunities should not depend on religious affiliation, just as it should not depend on gender, race or sexual orientation. The state has a moral obligation to protect its citizens from discrimination on the basis of such personal characteristics and to guarantee equality of opportunity in, among others, the labour market.Footnote 38 Exceptions to the prohibition of discrimination must therefore be both adequately justified and narrowly circumscribed. The protection of the right to religious freedom may justify an exception for religious organisations, but the exception must not go beyond what is necessary to protect an individual adherent’s freedom to live by her deepest commitments. In a liberal society, the right of religious organisations to discriminate on religious grounds in employment and occupation should therefore be limited to employees who carry out religious functions. The closer the occupational activities are to an organisation’s religious ethos, the stronger the claim to an exemption from the prohibition of discrimination. For an exception to be justified according to liberal democratic principles, these activities will most likely need to involve the teaching and promotion of the organisation’s religious ethos. What is certain, however, is that the internal autonomy of religious organisations and their right to discriminate on the grounds of religious belief will need to be strictly delimited if they are to be compatible with such principles.Footnote 39

From the EU’s point of view, the answer to the question regarding how to strike a balance between the right to religious autonomy and the right not to be discriminated against is decidedly less straightforward. Some Member States have, for historical or other reasons, drawn the scope of the principle of religious autonomy far more broadly than would be permissible under liberal democratic premises. Although the EU should also be judged on how far its policies adhere to liberal principles such as individual autonomy and equality, many also believe that it is consistent with or even required by liberal principles that the EU should accommodate national cultures and identities,Footnote 40 including, in that case, national conceptions of the place of religion in society.Footnote 41 Besides, the EU is not obliged or empowered to right every wrong at the national level. This is not the place to delve deeply into such matters; suffice it to say that a degree of respect by the EU for national constitutional principles and democratically legitimated norms seems consistent with liberal principles.

In any case, the Framework Directive falls somewhere midway between protecting liberal norms of non-discrimination and equal opportunity and observing national constitutional conceptions on the appropriate position of religion. On the one hand, the Directive aims to reduce the disadvantages faced by individuals on the basis of socially salient personal characteristics, including religion and belief. To that aim, it provides that the Member States may allow employers with an ethos based on religion and belief to discriminate on these grounds, but only on the conditions set out in Article 4. These conditions do not completely exempt religious employers from the prohibition of discrimination or allow them to determine for themselves when religion is a legitimate condition of employment.Footnote 42 Some of the conditions are strict, in fact, and, on the face of it, demand a close connection between the occupational activity and the religious ethos of the organisation for discrimination based on religion and belief to be justified. On the other hand, Article 4(2) also refers to national constitutional principles, which suggests that it values the right of Member States to determine the position of religion in their society and allows for a wider conception of the principle of religious autonomy than can be justified from a strictly liberal democratic point of view. Of course, this does not answer how wide the scope of the religious ethos exemption in EU law is. For that, we have to examine more closely Article 4 of the Framework Directive and the place of this provision in the overall scheme of EU law (i.e., its interaction with national constitutional law and international law).

3. Article 4 of the Framework Directive

This section seeks to ascertain the meaning of Article 4 of the Framework Directive; that is, the two occupational requirement exceptions that religious employers may invoke to justify religious discrimination in employment and occupation: the general occupational requirement exception in Article 4(1) and the religious occupational requirement exception in Article 4(2). This section will discuss what constitutes a reasonable interpretation of the conditions set out in these provisions, which must be met to justify the use of religion and belief as an occupational requirement. The references to national constitutional law in Article 4(2) will not be discussed yet, since the relationship between EU law and national constitutional law is the subject of analysis in the following section.

It might be thought that employers with an ethos based on religion and belief will only be interested in the exception in Article 4(2), given that this provision seems to be broader in scope than the exception in Article 4(1) and is specifically designed to protect their internal autonomy. However, Article 4(2) provides that ‘Member States may maintain national legislation in force … or provide for future legislation incorporating national practices existing at the date of adoption of this Directive’. Acondition for its application is therefore that Member States have incorporated the exception into their domestic legislation, which not all have done.Footnote 43 The religious ethos exemption of employers established in countries that have not transposed Article 4(2) will be conditioned exclusively by Article 4(1) – again, subject to the condition that it is enshrined in national legislation. Let me therefore discuss both exceptions in turn.

It is apparent from the conditions set out in Article 4(1) that the exception set out therein is narrow in scope. It provides that differences in treatment based on a protected personal characteristic are justified if the difference constitutes a ‘genuine and determining’ occupational requirement that pursues a ‘legitimate objective that is proportionate’. Recital 23 of the Framework Directive supports this view and states that the exception applies in ‘very limited circumstances’. Likewise, and in accordance with legislative intent, the CJEU has decided that the exception must be ‘interpreted strictly’;Footnote 44 that the occupational requirement must be ‘objectively dictated by the nature of the occupational activities concerned or of the context in which they are carried out’.Footnote 45 Advocate General Sharpston summarised the case law on the exception as follows: ‘the derogation must be limited to matters which are absolutely necessary in order to undertake the professional activity in question.’Footnote 46 Although Article 4(1) has not, to date, been applied in disputes involving religious employers wishing to discriminate against (potential) employees on the basis of their religious beliefs, the above means that such discrimination is most likely only permissible when sharing the religious beliefs of the organisation is strictly necessary for the exercise of the occupational activity in question. As Vickers put it, most likely only ‘in the case of those employed in religious service, whose job involves teaching or promoting the religion, or being involved in religious observance’.Footnote 47 In other words, Article 4(1) does not authorise the broad type of exemptions that religious organisations enjoy in Member States such as Germany or Austria.

Whether such exemptions are permissible under Article 4(2), instead, depends on the meaning of that provision. It provides that a person’ religion or belief is a legitimate ground for discrimination if it constitutes ‘a genuine, legitimate and justified occupational requirement’. The CJEU has interpreted the terms ‘genuine, legitimate and justified’ as follows: The term genuine means that ‘professing the religion or belief on which the ethos of the church or organisation is founded must appear necessary because of the importance of the occupational activity in question for the manifestation of that ethos or the exercise by the church or organisation of its right of autonomy’.Footnote 48 The term legitimate means that ‘the requirement of professing the religion or belief on which the ethos of the church or organisation is founded is not used to pursue an aim that has no connection with that ethos or with the exercise by the church or organisation of its right of autonomy’.Footnote 49 The term justified means that the organisation must be capable of showing ‘that the supposed risk of causing harm to its ethos or to its right of autonomy is probable and substantial, so that imposing such a requirement is indeed necessary’.Footnote 50 The interpretation of these terms bears a striking resemblance to the CJEU’s understanding of the principle of proportionality: to be lawful, the occupational requirement must in essence be appropriate and necessary.Footnote 51

Although this interpretation seems reasonable, it leaves plenty of uncertainty as to its application in concrete and specific cases. When do religion and belief constitute a ‘genuine, legitimate and justified’ occupational requirement in accordance with Article 4(2), and how exactly does the meaning of these criteria differ from that of the terms ‘genuine’ and ‘determining’ in Article 4(1)? The crucial difference between the two provisions seems to lie in the condition in Article 4(1) that the occupational requirement must be ‘determining’. As explained above, it follows from this condition that having a particular religion or belief must be strictly necessary for the performance of an employment activity. It would thus appear that no such requirement of strict necessity is imposed by Article 4(2). But what does this mean in practical terms?

The following example can help to spell out the difference between Articles 4(1) and 4(2) more clearly. Nowhere is the role of religious organisations in the provision of public services more controversial than in the area of education.Footnote 52 Asignificant proportion of schools in many Member States have a religious ethos, which can lead them to discriminate against teachers on religious grounds in decisions on their employment or dismissal. Membership of a particular religious community may be a condition of employment, and the violation of the religious principles of this community reason for dismissal, even if sharing these principles is not strictly necessary to undertake the teaching job in question. It is one thing for a faith-based school to expect a religion teacher to share its ethos but quite another to make religious membership a condition of employment for mathematics or physics teachers. As far as my understanding of physics and mathematics goes, being religious is not a ‘determining’ requirement to be able to teach these subjects – it is not strictly necessary to be able to teach principles of mathematics or physics. If this is correct, faith-based schools may use Article 4(1) to justify the expectation that religion teachers share their religious ethos, but they cannot invoke this provision to require that physics or maths teachers do so. However, a Member State may permit them to rely on Article 4(2) to justify the use of religion as an occupational requirement for all teaching positions. After all, it seems plausible that the use of religion as an occupational requirement in respect of all teachers at faith-based schools is a genuine, legitimate and justified – genuine because the religiosity of teachers is important for the school to manifest its ethos; legitimate because such a requirement pursues an aim connected to its religious ethos; and justified because it can prevent probable and substantial harm to its ethos. Such harm may result from the fact that schools are otherwise unable to provide their pupils with the desired religious environment. Thus, faith-based schools for which religion is an occupational requirement are likely to meet the three conditions set out in Article 4(2), at least as the CJEU understood them.

As this example suggests, the religious occupational requirement exception in Article 4(2) is wider in scope than the general occupational requirement exception in Article 4(1). Nonetheless, the conditions set out in the former do limit the internal autonomy of religious organisations significantly. In particular, organisations that discriminate against individuals on the basis of their religious beliefs will not always meet the condition in Article 4(2) that the occupational requirement is ‘justified’. This is because many types of employment within religious organisations can be performed by non-adherents without probable or substantial harm to the organisation’s ethos – think of positions as medical specialists, janitors or legal advisers. For instance, it is highly unlikely that a medical specialist responsible for treating ailing patients will harm the ethos of a Catholic hospital substantially if he is not married according to the principles of canon law. Areligious employer that requires such employees to share and act in accordance with its religious ethos will be acting contrary to the conditions in Article 4(2) that have been discussed so far.

An exception is employers with a religious ethos which, as the UK Employment Tribunal once said, ‘permeates … the work, and daily life, and activities in the workplace’.Footnote 53 The employer in question, the Leprosy Mission, began and ended formal meetings with prayer and began each working day with half an hour of prayer and gospel reading. With such employers, all employment activities can be deemed to be covered by the exemption in Article 4(2), as it would do probable and substantial harm to their ethos if they were required to employ non-adherents. In contrast, the occupational activities of employers where religion does not permeate every aspect of the workplace do not automatically benefit from the protection that Article 4(2) can provide. In this respect, the Bundesarbeitsgericht seems to have ruled correctly, following IR v JQ, that religion cannot be a requirement for employment as a surgeon in a hospital, and following Egenberger, that being a member of the Church cannot be a requirement for employment as a legal expert with the responsibility to draft a report on the UN Convention on the Elimination of All Forms of Racial Discrimination.Footnote 54 In neither case was religion a genuine, legitimate and justified occupational requirement, necessary to prevent probable and substantial damage to the employer’s religious ethos.

4. Respect for religious organisations under national constitutional law

As we have seen, the scope of the religious ethos exemption is determined not only by Article 4 of the Framework Directive but also by this provision’s relationship to the various provisions of EU law that demand respect for national constitutional values. At least three different provisions suggest that principles of national constitutional law must be considered in the application of the exemption. First, Article 4(2) of the Framework Directive itself: the last sentence of its first paragraph provides that it ‘shall be implemented taking account of Member States’ constitutional provisions and principles’, and its second paragraph states that:

Provided that its provisions are otherwise complied with, this Directive shall thus not prejudice the right of churches and other public or private organisations, the ethos of which is based on religion or belief, acting in conformity with national constitutions and laws, to require individuals working for them to act in good faith and with loyalty to the organisation’s ethos.Footnote 55

In addition, the Treaties also place the EU under an obligation to take national constitutional law into account. According to Article 4(2) of the Treaty on European Union (TEU) (nota bene: a different Article 4(2)), EU institutions must ‘respect the equality of Member States before the Treaties as well as their national identities, inherent in their fundamental structures, political and constitutional’.Footnote 56 Finally, Article 17(1) TFEU requires that the EU ‘respects and does not prejudice the status under national law of churches and religious associations or communities in the Member States’, a requirement that seems to find more specific expression in Article 4(2) of the Directive.

Seemingly in contravention of these provisions, the CJEU declined to accommodate principles of German constitutional law in Egenberger and IR v JQ and decided to curtail the autonomy hitherto enjoyed by religious organisations under the German constitution. The fact that these judgements did not assign greater weight to the German constitutional principle of religious autonomy law has probably been the main source of frustration and criticism.Footnote 57 This section will address such criticism and examine the extent to which the scope of the religious ethos exemption should be conditioned by principles of domestic constitutional law. Did Egenberger and IR v JQ give insufficient weight to such principles? As we shall see, this is a terribly complex question – far more complicated than critics of the judgements have realised – the answer to which depends on certain underlying assumptions concerning Treaty interpretation and the optimal relationship between the CJEU and the EU legislature. We will further see that the CJEU can and probably should be criticised for showing a lack of respect for principles of national constitutional law concerning the status of religious organisations, regardless of our assumptions on these underlying issues. On the other hand, the extent to which this criticism is justified depends heavily on our assumptions about, in particular, the degree of weight that should be assigned to the choices of the legislature. If the application of a principle of judicial deference to legislation was appropriate, the judgements are not manifestly flawed; if not, they violate Article 4(2) TEU or Article 17(1) TFEU (section 4.A). If, however, yielding to and enforcing the criteria set out in Article 4(2) of the Framework Directive was appropriate, the CJEU only failed to take due account of principles of national constitutional law in its interpretation of this provision (section 4.B).

A. Articles 4(2) TEU and 17(1) TFEU

Articles 4(2) TEU and 17(1) TFEU impose an obligation on the EU to respect fundamental norms of national law – the former to respect the constitutional identities of the Member States; the latter to respect norms of national law governing the status of churches and other religious associations. It is not entirely clear how the two provisions relate to each other, but the prevailing view, defended by several Advocates General, is that Article 17 TFEU ‘gives specific effect to and complements the more general requirement enshrined in Article 4(2) TEU on respect for the national identity of the Member States’.Footnote 58 The CJEU has not made its position explicit but seems to hold the same view. After all, it concentrated exclusively on Article 17(1) TFEU in Egenberger and IR v JQ, even though Article 4(2) TEU also seemed relevant. Moreover, as we shall see below, both provisions are subject to the same principles of interpretation. This interpretation of Article 17(1) TFEU as a concretisation of Article 4(2) TEU is, in my view, reasonable and will therefore be followed in the following analysis.

Before addressing the criticism that the CJEU should have accommodated German constitutional law, it is useful to consider what was said about Article 17(1) TFEU in Egenberger and IR v JQ. In essence, the CJEU held that the provision does not affect the interpretation of the Framework Directive:

Article 17 TFEU expresses the neutrality of the European Union towards the organisation by the Member States of their relations with churches and religious associations and communities; that article is not such as to exempt compliance with the criteria set out in Article 4(2) of Directive 2000/78 from effective judicial review.Footnote 59

In support of this conclusion that Member States must comply with the criteria set out in Article 4(2) of the Framework Directive despite Article 17 TFEU, the CJEU observed that the latter provision had been considered during the legislative process leading to the adoption of the Directive:

The wording of Article 17 TFEU corresponds, in essence, to that of Declaration No 11 on the status of churches and non-confessional organisations, annexed to the Final Act of the Treaty of Amsterdam. The fact that Declaration No 11 is expressly mentioned in recital 24 of Directive 2000/78 shows that the EU legislature must have taken that declaration into account when adopting the directive.Footnote 60

Thus, because the EU legislature had considered the requirements of Article 17 TFEU, the CJEU adhered to a principle of judicial deference to legislation and imposed the conditions set out in Article 4(2) of the Framework Directive on religious employers.

Two objections have been raised to the position that Article 17(1) TFEU does not affect the interpretation of Article 4(2) of the Framework Directive. First, by restricting the autonomy of religious associations, the CJEU is said to have exceeded the limits of EU competence as determined by Article 17(1) TFEU.Footnote 61 This objection is not convincing. While the status of churches and other religious associations is indeed a national competence, it is settled case law that Member States must exercise their competences in conformity with EU law.Footnote 62 Recall, in this context, the Kreil judgement, in which the CJEU held that the organisation of the armed forces – obviously a national competence – must be exercised with due regard to EU non-discrimination law.Footnote 63 In the same vein, it ruled in Parris that marital status falls within the competence of the Member States, but that this competence must be exercised in conformity with EU non-discrimination law.Footnote 64 Egenberger and IR v JQ bear striking resemblances – the status of religious associations is a national competence, but one that must be exercised in accordance with the criteria set out in Article 4(2) of the Framework Directive. In this respect, the CJEU’s reasoning was predictable and rests on hardly contestable principles of interpretation. As de Witte has rightly reminded us, the ‘obligations contained in an international Treaty surely restrict the exercise of state competences, without those competences themselves being transferred to the international level’.Footnote 65 It should not surprise, in other words, that matters falling within the ‘competence of the Union may have a religious dimension’.Footnote 66

The second objection is more powerful and occupies the remainder of this section: it contends that the Framework Directive should have been interpreted in accordance with Articles 17(1) TFEU and 4(2) TEU, with a view to ensuring its compatibility with EU primary law. This objection is rooted in established case law pursuant to which ‘all Community acts must be interpreted in accordance with primary law as a whole’ in order not to affect their validity.Footnote 67 Instead, the CJEU held in Egenberger and IR v JQ that Article 17 TFEU does not ‘exempt compliance with the criteria set out in Article 4(2)’. Critics therefore claim that by failing to interpret the Framework Directive in accordance with EU primary law, the CJEU violated the EU’s constitutional limits.Footnote 68 Let me explain why the issue is not as straightforward as they suggest.

Most would agree that the EU should tread carefully when its decisions risk encroaching on principles of national constitutional law, provided that these principles respect the fundamental values that form the foundation of the EU legal order as set out in Article 2 TEU.Footnote 69 However, as is generally accepted too, the obligation under Articles 17(1) TFEU and 4(2) TEU to respect the constitutional identities of the Member States – including provisions of national constitutional law governing the status of religious associations and communities – is conditional rather than absolute; it does not attribute automatic precedence to the constitutional principles of the Member States, but rather requires that a balance is struck between principles of national constitutional law and competing standards of EU law.Footnote 70 Therefore, it is not sufficient for critics of Egenberger and IR v JQ to show that these judgements affect the status of religious associations under national constitutional law. Rather, they must demonstrate that an improper balance was struck between the German constitutional principle of religious autonomy and the prohibition of discrimination under EU law.

What might seem to support such a position is that the appeal to the protection of principles of national constitutional law within the scope of application of EU primary law was brushed aside rather hastily and without further justification in Egenberger and IR v JQ. It used to be the case that the principle of proportionality was the ‘common denominator for all national identity claims’.Footnote 71 As such, Member States could cite national constitutional law to justify a derogation from EU law, provided that the derogation is ‘based on objective considerations and is proportionate to the legitimate objective of the national provisions’.Footnote 72 In Egenberger and IR v JQ, however, the CJEU defined the place of national constitutional identity in EU law differently, not through the application of the principle of proportionality but by adherence to the principle of judicial deference to the EU legislature. It did not weigh principles of national constitutional law against competing norms of Union law but simply held that Article 17(1) TFEU could not ‘exempt compliance with the criteria set out in Article 4(2) of Directive 2000/78’. National constitutional law was not accommodated because the EU legislature had enacted the competing norm of EU law, a norm which, moreover, was meant to protect the national autonomy of the Member States albeit within the conditions set by EU non-discrimination law.

The application of a principle of judicial deference to legislation as the instrument for settling national identity claims may have fed into scepticism about Egenberger and IR v JQ. As a means of determining whether national constitutional law must be respected, this principle seems rather blunt in comparison with the principle of proportionality. In contrast to the principle of judicial deference, the application of the proportionality principle would allow for a more exacting review of EU law, whereby all factors relevant to determining whether it impermissibly infringes national constitutional law can be considered.Footnote 73 It should be noted, however, that Egenberger and IR v JQ are not isolated cases: the CJEU has also favoured deference to legislation in other judgements where national constitutional law was relied on to justify derogations from the application of EU law. In Melloni, the CJEU refused to accommodate the right to a fair trial in Spanish constitutional law because the contested norm of EU law had been adopted by the EU legislature – it effected ‘a harmonisation of the conditions of execution of a European arrest warrant in the event of a conviction rendered in absentia, which reflects the consensus reached by all the Member States’.Footnote 74 In M.A.S., on the other hand, it decided that the principle of legality under Italian constitutional law warranted a derogation from EU law, because ‘the limitation rules applicable to criminal proceedings relating to VAT had not been harmonised by the EU legislature’.Footnote 75 Thus, whether priority is accorded to principles of national constitutional law depends, according to the most recent case law, not only on the substance but also on the source of the contested norm of EU law, on the EU institution that issued the norm.

It is also worth observing that these judgements seem to be part of a more general trend in the case law towards deference to the EU legislature. The case law on the free movement of persons is illustrative in this regard. First, until a few years ago, the legislative conditions under which EU citizens could claim equal access to social assistance were frequently disregarded; they were interpreted in accordance with EU primary law to expand the conditions for obtaining social assistance set out therein.Footnote 76 In recent years, however, the CJEU has followed the criteria set out in the Citizenship Directive more closely – a decision motivated by the fact that ‘the principle of non-discrimination, laid down generally in Article 18 TFEU, is given more specific expression in Article 24 of Directive 2004/38’.Footnote 77 In other words, its decision not to interpret legislative provisions in accordance with EU primary law was based on the legislature having taken primary law – the principle of non-discrimination – into account and having established more precise conditions for its application. Second, that the CJEU seems more prepared these days to accept the constraints set out in legislation is also clear from the case law conditioning the exportability of social security benefits. In earlier case law, the CJEU at times ignored legislative provisions prohibiting their exportability by interpreting these provisions in light of principles of EU primary law.Footnote 78 By contrast, in more recent case law it has enforced legislation more strictly, as another approach ‘would ultimately undermine the very fabric of the system which Regulation 1408/71 sought to establish’.Footnote 79

Although some of the above judgements have attracted serious criticism, it is hardly surprising that the CJEU exercises judicial deference and self-restraint, two judicial virtues that every court should display when applying the law.Footnote 80 It is sometimes assumed that adjudication by the CJEU must, as the case law of the European Court of Human Rights (ECtHR), ‘start from a position of deference’ to national institutions.Footnote 81 Yet such an assumption of equivalence ignores the fact that the CJEU and ECtHR are very different courts, embedded in an entirely different institutional context. Crucially, the CJEU exercises authority not just vertically vis-à-vis national institutions, but also horizontally relative to other EU institutions, including the legislative process. Moreover, the EU can only effectively realise its objectives through a joint course of action that binds all Member States, and the most effective way of determining this course of action is through the legislative harmonisation of national standards. It would thus be very hard for the EU to realise its objectives if the CJEU were to start from a position of deference to national institutions each time they act contrary to the choices of the EU legislature. On the contrary, for reasons of institutional legitimacy and institutional capacity, it is justified to adopt a position of deference to the EU legislature.Footnote 82

It is not, however, my intention to reopen this debate, for the more specific question that interests me here is whether deference to the EU legislature is also virtuous when its decisions encroach on the constitutional principles of the Member States – clearly a very controversial approach in a very controversial area of EU law. As Iwill explain, the use of a principle of judicial deference to legislation seems, under specific conditions at least, an appropriate way of deciding national identity claims. This should also explain why the CJEU’s decision in Egenberger and IR v JQ to adhere to a principle of deference seems justifiable.

EU law can outweigh national constitutional identities for various reasons. For example, it is generally accepted that Member States cannot rely on their national identities to justify conduct that violates the fundamental values enshrined in Article 2 TEU – human dignity, freedom, democracy, equality, the rule of law and respect for human rights.Footnote 83 But even when national constitutional principles are in accordance with these fundamental values, there may be valid reasons for according precedence to EU law. For instance, another reason for not automatically giving priority to a Member State’s constitutional choices is that they might harm the citizens of other Member States or run counter to the collective interest of the EU as a whole. Melloni provides a good example that illustrates this point: the EU could not effectively tackle cross-border crime if Member States could refuse to extradite their nationals as a matter of national constitutional law. Coman does so too: the EU would not be able to ensure the free movement of all EU citizens if respect must be owed to constitutional norms that refuse to recognise marriage between same-sex couples.Footnote 84 Other examples could be given, but the point should be clear: the EU’s capacity to take into account and defend interests that are not represented in national decision-making processes would be undermined if Member State constitutional identities would automatically outweigh other transnational interests.Footnote 85 This is not to say that EU law must automatically prevail; it merely shows why a fair balance must be found between the competing interests pursued by the EU as a whole and by the Member States individually.

As Claes has observed, finding this balance requires that ‘all available channels for communication and conversation are used’.Footnote 86 Logically speaking, this includes the EU legislative process.Footnote 87 That is, it seems only natural that the CJEU should take legislative decisions into account when assessing whether EU law should accommodate national constitutional law. The EU legislature provides a forum, however imperfect, in which the interests of the individual national peoples and of the citizens of the European Union are represented, and where a compromise can be found between the different and sometimes conflicting national and European societal goods. Member States can pursue their individual interests and defend their own fundamental social choices within the legislative process, but not unilaterally, prejudicing the citizens of other Member States or the interests of the Union as a whole.Footnote 88 It forces Member States to negotiate their interests within the constraints imposed by the supranational environment in which the legislative process is embedded. The legislative process may in many respects be imperfect, but the alternatives do not seem to offer a fairer representation of the interests involved in European integration.Footnote 89 This is why legislature may be considered, at least prima facie, to provide a fair basis for defining the place of national constitutional identity in EU law, and why it is appropriate for the CJEU to assign significant weight to legislative decisions in cases where national constitutional law is in danger of being affected.

Yet, the crucial question seems to be not whether judicial deference to legislative acts that encroach on a Member State’s constitutional identity is ever justified, but under which conditions it is. Afull examination of this question is beyond this article’s remit; Iwill just note that, if ever it is justified, it will be under the two conditions set out by the CJEU in its case law. In Melloni, it motivated its decision to exercise deference by pointing out that the legislative act reflected the consensus reached by all Member States.Footnote 90 In Egenberger, it did so on the basis that the legislature had taken into account national constitutional law (i.e., the status of churches and other religious organisations under national law).Footnote 91 Especially when both conditions are met – the legislature has taken into account national constitutional norms and the legislative action has been agreed upon by all Member States affected – deference to the choices of the legislature seems reasonable and appropriate. Such circumstances at least warrant a heightened degree of judicial deference. This, of course, leaves open the question of what degree of deference is appropriate if one of these two conditions has not been met, but this question is not relevant to our assessment of the scope of the religious ethos exemption in EU law. The EU legislature had considered the status under national law of churches and other religious associations and its act, the Framework Directive, reflects the consensus of all Member States. This gives us reason to believe that the CJEU’s use of a principle of judicial deference to legislation as an instrument for deciding national identity claims in Egenberger and IR v JQ was reasonable.

B. National constitutional law in the Framework Directive

It may seem as if adhering to a principle of judicial deference to legislation will be damaging to national constitutional identities, but this does not need to be the case. First, as we saw in the M.A.S. judgement, adherence to this principle means that the CJEU will be more inclined to respect national constitutional law when there is no harmonising legislation in place. Moreover, if the CJEU is committed to this principle (i.e., to respecting the constraints set out in legislation), it should yield to principles of national constitutional law where legislation so provides. This seems to follow from settled case law, according to which provisions of EU law that make ‘no express reference to the law of the Member States for the purpose of determining its meaning and scope must normally be given an autonomous and uniform interpretation’.Footnote 92 Inversely, Member States should be entitled to interpret EU law in light of their national law where legislation makes express reference to it.Footnote 93 It would be inconsistent and unprincipled if the CJEU were to apply only those legislative criteria that limit the authority of the Member States and to disregard legislative provisions that leave room for national difference.

There lies the main problem with the CJEU’s reasoning in Egenberger and IR v JQ. It held that Article 17(1) does not exempt compliance with the criteria set out in Article 4(2) of the Framework Directive, but then applied those criteria selectively. It ignored the part of this provision that refers to national constitutional law. As we have seen, Article 4(2) not just provides that it applies on the condition that religion or belief constitute ‘a genuine, legitimate and justified occupational requirement’, but refers to national constitutional law twice – the first paragraph states that it ‘shall be implemented taking account of Member States’ constitutional provisions and principles’; the second paragraph that:

Provided that its provisions are otherwise complied with, this Directive shall thus not prejudice the right of churches and other public or private organisations, the ethos of which is based on religion or belief, acting in conformity with national constitutions and laws, to require individuals working for them to act in good faith and with loyalty to the organisation’s ethos.

And while the criteria encroaching on national constitutional law – genuine, legitimate and justified – were strictly adhered to, the CJEU said nothing about provisions of national constitutional law despite the references thereto in Article 4(2). The judgements are therefore vulnerable to the criticism that they applied the conditions set out in this provision partially and selectively. In view of the broad meaning accorded to the principle of religious autonomy in German constitutional law, the CJEU should have clarified what, within the meaning of Article 4(2), it means to take national constitutional law into account.Footnote 94

Having said that, it is not evident that the CJEU should have reached a different conclusion in Egenberger and IR v JQ if it had interpreted Article 4(2) by reference to national constitutional law. First, the judgements respect the second paragraph of Article 4(2). According to Greiner, this paragraph must be read as lex specialis to the first paragraph, supporting the autonomy of the Member States in determining the internal autonomy of the established churches and their institutions.Footnote 95 This interpretation is incorrect: the second paragraph has no value at all and is merely stating the obvious. It says that the Directive will not prejudice the right of churches and other employers with an ethos based on religion and belief, ‘provided that its provisions are otherwise complied with’. But that goes without saying: provisions of EU law do not prejudice anything provided they are otherwise complied with. So, contrary to what might appear at first sight, the second paragraph of Article 4(2) does not exempt religious employers from EU non-discrimination law, nor does it authorise them to define their own sphere of autonomy or to determine independently when religion is an appropriate occupational requirement.Footnote 96 It does not alter the meaning of the Directive at all, so does not need to be considered in determining which matters are within the internal autonomy of religious organisations.

This is different with regard to the statement in the first paragraph that provisions of national constitutional law shall, in the interpretation of Article 4(2), be taken into account. The CJEU failed to address this criterion. More specifically, it failed to clarify how its decision took account of the German principle of religious autonomy and how national courts may take into account national constitutional law. But while this omission exhibits the sort of incoherence that critics of Egenberger and IR v JQ may rightly draw attention to, taking into account principles of national constitutional law is not the same, of course, as preserving or protecting such principles. In this regard, it must be noted that it would be impossible to fully respect national constitutional law while ensuring that the use of religion is a genuine, legitimate, and justified occupational requirement. It just so happens that certain national constitutional courts – including the German Constitutional Court – take such a broad view of the principle of religious autonomy that employers can use religion as an occupational requirement insituations where it is not genuine, legitimate or justified. Article 4(2) requires that provisions of national constitutional law be considered, not necessarily that they are complied with; the latter interpretation cannot be reconciled with the other conditions set out in that provision.

That poses the question of what it means and requires to take national constitutional provisions and principles into account. First, the CJEU must at least show that it is conscious of what is at stake (i.e., that it is aware of the fact that its decisions may affect fundamental norms of domestic law). The dissatisfaction with Egenberger and IR v JQ is undoubtedly partly due to the CJEU just ignoring the status of churches and other religious organisations in German constitutional law. Second, it must explain how it takes into account principles of domestic constitutional law, and it must offer sound reasons for decisions that do not accommodate such principles. Why does it consider deference to the legislature’s choices justified and why were principles of national constitutional law not upheld even though legislation requires these principles to be taken into account? We may expect the CJEU to have considered such questions, but also to give clear and considered answers thereto. Finally, it seems appropriate to interpret Article 4(2) as giving national constitutional principles on the status of religious organisations the benefit of the doubt when it is unclear whether the other criteria set out in this provision are met. However, surely there is an important difference between providing some leeway to national constitutional law and complete deference thereto. Egenberger and IR v JQ are not sufficiently motivated as concerns the EU’s obligation to respect national constitutional law, but it does not seem unreasonable that Article 4(2) was interpreted as restricting the principle of religious autonomy under German constitutional law.

5. The legal status of concordats under EU law

In addition to Article 4 of the Framework Directive and the interaction between EU law and national constitutional law, the third determinant of the scope of the religious ethos exemption in EU law is the relationship between EU law and international law. This third determinant was not at issue in Egenberger and IR v JQ, but must nonetheless be considered if we are to determine when religious organisations may discriminate against non-adherents. This is because the legal status of religious organisations is partly governed by legal agreements between the Member States and these organisations. Such agreements regulate matters as diverse as the provision of pastoral care in the army and prison, the imposition and collection of church taxes, and the involvement of religious organisations in providing social welfare. Of these agreements, those concluded with the Catholic Church are in a way unique: such ‘concordats’ are concluded with the Holy See and thus have treaty status under international law.Footnote 97 Catholic organisations that may not receive the protection they desire under EU law, via Article 4(2) of the Framework Directive or national constitutional identity, may therefore wish to invoke international law to protect their internal autonomy. Of course, these concordats are binding only on the parties that have signed them – the Member States or their respective regions and the Holy See – but due to the ‘triangular status’ between national, international, and EU law, their legal status within the national legal orders depends on the relationship between EU law and international law.Footnote 98 Thus, insofar as concordats regulate activities falling within the scope of the Framework Directive, the Directive’s application to the employment practices of Catholic employers will depend on the position of international law in relation to EU law.Footnote 99 This is why it is necessary to consider this relationship.

Around a dozen Member States have signed multiple agreements with the Holy See. Many of these concordats are not of interest to us, however, for the simple reason that they do not concern employment and occupation. For instance, many concordats deal with the civil status of marriages contracted under Canon law or the financing of the Catholic Church and their activities through state taxes. Moreover, concordats dealing with employment often regulate only specific employment activities that already are exempt by Article 4(2) of the Framework Directive from the prohibition of discrimination – for example, concordats regulating religious education in Catholic schools. So, only where concordats grant Catholic employers privileges that are contrary to the provisions of the Framework Directive, the question of the relationship between EU law and international law is pertinent.

The relevant Treaty provision in this regard is Article 351 TFEU, on the status of prior agreements of the member states with third countries:

The rights and obligations arising from agreements concluded before 1 January 1958 or, for acceding States, before the date of their accession, between one or more Member States on the one hand, and one or more third countries on the other, shall not be affected by the provisions of the Treaties.

To the extent that such agreements are not compatible with the Treaties, the Member State or States concerned will take all appropriate steps to eliminate the incompatibilities established. Member States will, where necessary, assist each other to this end and will, where appropriate, adopt a common attitude.

In applying the agreements referred to in the first paragraph, Member States will take into account the fact that the advantages accorded under the Treaties by each Member State form an integral part of the establishment of the Union and are thereby inseparably linked with the creation of common institutions, the conferring of powers upon them and the granting of the same advantages by all the other Member States.

This section studies the interpretation of this provision by the CJEU in order to determine whether and under what conditions concordats concluded by the Member States exempt Catholic employers from the obligations under EU non-discrimination law.

According to the CJEU, Article 351 TFEU allows ‘the Member State concerned to respect the rights of non-member countries under a prior agreement and to perform its obligations thereunder’.Footnote 100 EU institutions cannot ‘impede the performance of the obligations of Member States which stem from a prior agreement’.Footnote 101 To this end, it ruled in Minne that secondary legislation ‘cannot apply to the extent to which [incompatible] national provisions were adopted in order to ensure the performance by the Member State of obligations arising under an international agreement concluded with non-member countries’.Footnote 102 In effect, therefore, Article 351 TFEU ‘allows a derogation from the principle of primacy of EU law’,Footnote 103 which also applies, of course, to agreements concluded by Member States with the Holy See. To be clear, concordats do not bind the EU as regard to the third country in question. One condition for the EU to be bound by an international agreement to which it is not a party is that all Member States are parties to that agreement,Footnote 104 which is not the case as far as concordats are concerned. Thus, the fact that some Member States have concluded agreements with the Holy See on the rights and privileges of the Catholic Church does not prevent the EU from adopting non-discrimination legislation bearing on the Catholic Church. Such legislation just does not abrogate the obligations of Member States under a prior agreement with the Holy See, so the Framework Directive cannot limit the application of provisions of national law adopted to ensure that the obligations arising under a concordat are met.

As Article 351 TFEU states clearly, it applies only to international agreements concluded with third countries ‘before 1 January 1958 or, for acceding States, before the date of their accession’. Concordats not concluded before that date do not justify non-compliance, therefore, with the obligations under EU non-discrimination law. For instance, Portugal and Slovakia concluded their concordats mid-May 2014,Footnote 105 after their date of accession – in the case of Slovakia, only two weeks after its accession – and will thus not exempt these countries from compliance with the Framework Directive.Footnote 106 On the other hand, most Member States that have signed an agreement with the Holy See did so prior to 1958 or their accession to the EU. For example, the 1933 Reichskonkordat concluded between the Holy See and the then-emerging Nazi regime is still in force,Footnote 107 as is the concordat signed that same year by the fascist regime of Austria. These agreements grant extensive rights and privileges to the Catholic Church and would allow Germany and Austria to invoke a derogation from EU non-discrimination law if it would affect the rights the Church enjoys under these agreements. To illustrate this point, it might be that the defendant in IR v JQ – Caritas, a humanitarian and social welfare organisation under the control of the Catholic Church – could invoke Article 351 TFEU to claim an exemption from EU non-discrimination, but not the defendant in Egenberger, the Protestant Church. After all, agreements signed with the Protestant Church have no treaty status under international law.

Yet, contrary to what Article 351 TFEU may seem to suggest, prior agreements concluded with third countries do not enjoy unconditional primacy over EU law. Based on a contextual interpretation of Article 351 TFEU,Footnote 108 the CJEU found in Kadi that it ‘may in no circumstances permit any challenge to the principles that form part of the very foundations of the [EU] legal order’.Footnote 109 Hence, Member States cannot oppose the application of EU non-discrimination law on the basis of concordats that violate the fundamental values laid down in Article 2 TEU,Footnote 110 including basic human rights and core principles of liberal democracy. However, it is not clear that concordats have this effect. Although equality is among the foundational principles listed in Article 2 TEU, and concordats may be a contributing factor to discrimination by institutions under the control of the Catholic Church, it is probably not the case that every restriction on the principle of equality automatically contravenes Article 2 TEU. Indeed, it is unlikely that Member State will automatically infringe the values enshrined in Article 2 TEU by not giving full effect to the principle of equality. Ultimately, it will be for the CJEU to determine whether Article 351 TFEU can be invoked to uphold the rights that the Catholic Church derives from prior agreements concluded by the Member States with the Holy See, but at first glance, it seems unlikely that Article 2 TEU would prevent this.

But the protection afforded by Article 351 TFEU to prior agreements is limited by another obligation: Member States must renegotiate commitments with third countries that are incompatible with EU law. The second paragraph of Article 351 TFEU provides that, ‘to the extent that such agreements are not compatible with the Treaties, the Member State or States concerned shall take all appropriate steps to eliminate the incompatibilities established’. It follows from the case law of the CJEU that this is a strict obligation: a failure to comply with the obligation cannot be justified by reference to extraordinary difficulties in renegotiating the agreement with a third country. If a Member State is not in a position ‘to adjust an agreement, it must denounce the agreement’.Footnote 111 Therefore, those Member States whose obligations towards the Holy See are, following the judgements in Egenberger and IR v JQ, incompatible with Article 4(2) of the Framework Directive will be required to renegotiate these commitments and to eliminate the incompatibilities with this provision. In the event that this proves impossible, they will be required to denounce their concordats to ensure the full effectiveness of EU non-discrimination law. In renegotiating their commitments, Member States will, according to Article 351 TFEU, ‘where necessary, assist each other to this end and shall, where appropriate, adopt a common attitude’. The Commission may be tasked with the responsibility ‘to take any steps which may facilitate mutual assistance between the Member States concerned and their adoption of a common attitude’.Footnote 112 For now, Article 351 TFEU will exempt Catholic organisations from the prohibition of discrimination on grounds of religion and belief in so far as their right to discriminate against non-adherents is governed by an agreement with the Holy See, but prior international commitments that have an impact on the scope of the religious ethos exemption should not create a permanent derogation from EU non-discrimination law.

6. Conclusion

This article has sought to clarify the scope of the exemption of employers with an ethos based on religion and belief from the prohibition of discrimination on grounds of religion and belief in EU law. In doing so, it has attempted to shed light on several questions of EU constitutional law. The judgements in the Egenberger and IR v JQ cases show that the relationship between EU and national constitutional law remains a delicate issue, although these judgements have also raised questions about the limits of EU competence. The criticism that the CJEU exceeded these limits turned out to be incorrect. The status of churches and other religious associations is indeed within the competence of the Member States, but it should not be controversial that this competence must be exercised in accordance with EU non-discrimination. However, whether the CJEU assigned sufficient weight to principles of national constitutional law in its decisions is a more delicate question, whose answer depends on certain prior assumptions such as on the proper domain of legislative authority. As explained, we have good reasons to think that the application of a principle of judicial deference to legislation can be justified where legislation may affect the constitutional identities of the Member States, especially under specific conditions that are met by the Framework Directive. But the judgements in Egenberger and IR v JQ also show the importance of further discussion on the proper place of national constitutional identity in EU law and on the responsibilities of the legislature in defining that place.

The scope of the religious ethos exemption depends in total on three factors: in addition to the interaction between EU law and national constitutional law, it depends on the interpretation of Article 4 of the Framework Directive and on the interaction between EU and international law.

There is no single correct interpretation of these factors and thus of the scope of the exemption, but the latter two factors are significantly less controversial than the relationship between EU law and national constitutional law. What seems clear and relatively uncontroversial is that the occupational requirement exception in Article 4(1) of the Framework Directive is narrow in scope: it allows discrimination against non-adherents only when sharing the employer’s religious ethos is strictly necessary for the performance of a function, for example, when it involves teaching or promoting religious beliefs. What also seems fairly uncontroversial is that derogations from EU non-discrimination law caused by prior international agreements with the Holy See cannot be permanent; they must be eliminated by renegotiating the agreements. Disagreement on the precise scope of the religious ethos exemption will relate primarily to the meaning of Article 4(2) and, in that connection, the respect that must be shown to principles of national constitutional law. It is to be hoped that the CJEU will find the opportunity to find more satisfactory answers to the questions this provision raises in future cases.

Acknowledgements

The author wishes to thank Ronan McCrea, Lucy Vickers and Cécile Laborde for discussion of earlier drafts.

Funding statement

This work was supported by the British Academy. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1 Examples are given below.

2 As is the case in the Netherlands (see, Netherlands Institute of Human Rights, Judgement in Case 2015-68, decided on 9 June 2015 and Opinion in Case 2012-84, decided 4 May 2012) and it seems also in Belgium (see, Constitutional Court, Judgement no 39/2009, of 11 March 2009) and Spain (European Equality Law Network, ‘Country Report on Non-Discrimination: Spain’ (2020) 46-48. Available at: <https://www.equalitylaw.eu/downloads/5227-spain-country-report-non-discrimination-2020-1-56-mb> (last accessed 12 February 2022).

3 Council Directive 2000/78/EC of 27 November 2000 establishing a general framework for equal treatment in employment and occupation (OJ L303/16).

4 Case C-414/16 Egenberger, ECLI:EU:C:2018:257; Case C-68/17 IR v JQ, ECLI:EU:C:2018:696.

5 BAG, 8 AZR 501/14 (25 October 2018); BAG, 2 AZR 746/14 (20 February 2019). See in addition the decision of the Karlsruhe Labour Court prohibiting the use of religion as a condition for a secretarial function within the Church. ArbG Karlsruhe, 1 Ca 171/19 (18 September 2020).

6 For criticism, Hans Michael Heinig, ‘Why Egenberger Could be Next’, available at: <https://verfassungsblog.de/why-egenberger-could-be-next/> (last accessed 12 February 2022); Peter Unruh, ‘Im Spannungsfeld von Antidiskriminierung und kirchlicher Selbstbestimmung – Zur Einordnung und Kommentierung der neuen religionsrechtlichen Tendenzen des EuGH’ in Diakonie Deutschland (ed), Evangelische Identität und Pluralität Perspektiven für die Gestaltung von Kirche und Diakonie in einer pluraler werdenden Welt (2018); Gregor Thusing and Regina Mathy, ‘Das deutsche kirchliche Arbeitsrecht vor dem EuGH – Tendenz- oder Transzendenzschutz?’ in Hermann Reichold (ed), Tendenz- statt Transzendenzschutz in der Dienstgemeinschaft? Aktuelle Anstöße zur Loyalitätsfrage durch den Europäischen Gerichtshof (Verlag Friedrich Pustet 2019).

7 Frankfurter Allgemeine Zeitung, ‘Das erste Karlsruher Nein? (2 May 2019); Heiko Sauer, ‘Kirchliche Selbstbestimmung und deutsche Verfassungsidentität Überlegungen zum Fall Egenberger’ <https://verfassungsblog.de/kirchliche-selbstbestimmung-und-deutsche-verfassungsidentitaet-ueberlegungen-zum-fall-egenberger/> (last accessed 10 May 2021).

8 Mark Bell, ‘The Principle of Equal Treatment: Widening and Deepening’ in PP Craig and GDe Búrca (eds), The Evolution of EU law (2nd ed, Oxford University Press 2011).

9 Case C-507/18, Associazione Avvocatura per Idiritti LGBTI, ECLI:EU:C:2020:289, para 38.

10 For these statistics, Raphaële Xenidis, ‘The Polysemy of Anti-Discrimination Law: The Interpretation Architecture of the Framework Employment Directive at the Court of Justice’ 58 (2021) Common Market Law Review 1649, 1652–5.

11 See also the headscarf cases: Case C-157/15 Achbita, ECLI:EU:C:2017:203; Case C-188/15 Bougnaoui, ECLI:EU:C:2017:204; Joined Cases C-804/18 and Case C-341/19, IX v Wabe and MH Müller Handels, ECLI:EU:C:2021:594.

12 See, Evelyn Ellis and Philippa Watson, EU Anti-Discrimination Law (2nd edn, Oxford University Press 2012) 382.

13 Ibid 394.

14 Julian Rivers, The Law of Organized Religions: Between Establishment and Secularism (Oxford University Press 2010) 133. See also, Jane Calderwood Norton, Freedom of Religious Organizations (1st ed, Oxford University Press 2016) 79; Ellis and Watson, EU Anti-Discrimination Law 395. But see, for a different perspective, Ronan McCrea, Religion and the Public Order of the European Union (Oxford University Press 2014) 166–7.

15 Rivers, The Law of Organized Religions 133.

16 See, for example, Stefan Greiner, ‘Kirchliche Loyalitätsobliegenheiten nach dem “IR”-Urteil des EuGH’ (2018) Neue Zeitschrift für Arbeitsrecht, 1289–94; Claus Dieter Classen, ‘Das kirchliche Arbeitsrecht unter europäischem Druck – Anmerkungen zu den Urteilen des EuGH (jeweils GK) vom 17 April 2018 in der Rs. C-414/16 (Egenberger) und vom 11 September 2018 in der Rs. C-68/17 (IR)’ (2018) Europarecht 752–67.

17 BVerfG 70, 138 – Loyalitätspflicht (4 June 1985); BVerfG 2 BvR 661, 12 (22 October 2014). See, for further discussion, Gerhard Robbers, Church Autonomy in the European Court of Human Rights – Recent Developments in Germany, 26 JL & Religion 281 (2010).

18 Josef Hien, ‘The Return of Religion? The Paradox of Faith-Based Welfare Provision in a Secular Age’ [2014] MPifG Discussion Paper 14/9.

19 For instance, the European Commission issued a reasoned opinion to Ireland in 2008 for its broad interpretation of the exemption. See further, Amy Dunne, ‘Tracing the Scope of Religious Exemptions under National and EU Law: Section 37(1) of the Irish Employment Equality Acts 1998–2011 and Ireland’s Obligations Under the EU Framework Directive on Employment and Occupation, Directive 2000/78/EC’ 31 (2015) Utrecht Journal of International and European Law 33.

20 European Equality Law Network, ‘Country Report on Non-Discrimination: Cyprus’ (2021) 65. Available at <https://www.equalitylaw.eu/downloads/5529-cyprus-country-report-non-discrimination-2021-1-91-mb> (last accessed 12 February 2022).

21 European Equality Law Network, ‘Country Report on Non-Discrimination: Austria’ (2021) 42–3. Available at <https://www.equalitylaw.eu/downloads/5474-austria-country-report-non-discrimination-2021-1-41-mb> (last accessed 12 February 2022).

22 European Equality Law Network, ‘Country Report on Non-Discrimination: Spain’ (2021) 50–1. Available at <https://www.equalitylaw.eu/downloads/5479-spain-country-report-non-discrimination-2021-1-56-mb> (last accessed 12 February 2022).

23 European Equality Law Network, ‘Country Report on Non-Discrimination: The Netherlands’ (2021) 50. Available at <https://www.equalitylaw.eu/downloads/5518-netherlands-country-report-non-discrimination-2021-1-44-mb> (last accessed 12 February 2022).

24 European Equality Law Network, ‘Country Report on Non-Discrimination: Sweden’ (2021) 49. Available at <https://www.equalitylaw.eu/downloads/5493-sweden-country-report-non-discrimination-2021-1-61-mb> (last accessed 12 February 2022); European Equality Law Network, ‘Country Report on Non-Discrimination: France’ (2021) 70. Available at <https://www.equalitylaw.eu/downloads/5530-france-country-report-non-discrimination-2021-pdf-1-75-mb> (last accessed 12 February 2022).

25 Rivers, The Law of Organized Religions 334.

26 Calderwood Norton, Freedom of Religious Organizations. See also, Rivers, The Law of Organized Religions 334.

27 Which are two of the three components of autonomy on the account of Joseph Raz, The Morality of Freedom (Clarendon Press 1988) 372–8.

28 Tarunabh Khaitan and Jane Calderwood Norton, ‘The Right to Freedom of Religion and the Right against Religious Discrimination: Theoretical Distinctions’ 17 (2019) International Journal of Constitutional Law 1125, 1137–41.

29 Calderwood Norton, Freedom of Religious Organizations 16.

30 Of course, the right to manifest a belief can be limited. See, in this regard, Eweida and others App no 48420/10 (ECtHR, 15 January 2013) para 80.

31 Cécile Laborde, Liberalism’s Religion (Harvard University Press 2017) 179.

32 See, on the relation between collective group rights and the rights of individual members, Raz, Freedom of Religious Organizations 208.

33 Khaitan and Calderwood Norton, ‘The Right to Freedom of Religion and the Right against Religious Discrimination’ 1141.

34 See, in particular, Khaitan and Calderwood Norton, ‘The Right to Freedom of Religion and the Right against Religious Discrimination’; Tarunabh Khaitan and Jane Calderwood Norton, ‘Religion in Human Rights Law: ANormative Restatement’ 18 (2020) International Journal of Constitutional Law 111. See, for other assessments of the relationship between the two rights, Ilias Trispiotis, ‘Religious Freedom and Religious Antidiscrimination’ 82 (2019) The Modern Law Review 864; Ronan McCrea, ‘Squaring the Circle: Can an Egalitarian and Individualistic Conception of Freedom of Religion or Belief Co-Exist with the Notion of Indirect Discrimination?’ in Hugh Collins and Tarunabh Khaitan (eds), Foundations of Indirect Discrimination Law (Hart Publishing, an imprint of Bloomsbury Publishing Plc 2018).

35 Khaitan and Calderwood Norton, The Right to Freedom of Religion and the Right against Religious Discrimination’.

36 For a prominent account of socially salient group membership, Kasper Lippert-Rasmussen, Born Free and Equal? APhilosophical Inquiry into the Nature of Discrimination (Oxford University Press 2014) chapter 1.

37 At an individual level, we are of course familiar with clashes between the right to freedom of religion and the right to non-discrimination in disputes over the refusal of marriage registrars to celebrate same-sex weddings and of employees to shake hands with their colleagues.

38 For explorations of the moral justifications for non-discrimination law, see, among others, Tarunabh Khaitan, A Theory of Discrimination Law (Oxford University Press 2016); Sophia Moreau, Faces of Inequality: ATheory of Wrongful Discrimination (Oxford University Press 2020); Lippert-Rasmussen, Born Free and Equal?.

39 Laborde, Liberalism’s Religion 178–90.

40 For a defence of this position, Elke Cloots, National Identity in EU Law (Oxford University Press 2015) 89–94. She draws extensively on liberal scholars like, Will Kymlicka, Multicultural Citizenship: ALiberal Theory of Minority Rights (Clarendon Press 1995); Yael Tamir, Liberal Nationalism (3rd printing and 1st paperback printing, with new preface) (Princeton University Press 1995).

41 For discussion of that issue, Daniel Augenstein, ‘Religious Pluralism and National Constitutional Traditions in Europe’ in Camil Ungureanu and Lorenzo Zucca (eds), Law, State and Religion in the New Europe: Debates and Dilemmas (2012).

42 As affirmed by Egenberger, paras 42–69.

43 Including Sweden and France.

44 Case C-447/09 Prigge and Others, EU:C:2011:573, para 72; Case C-416/13 Vital Pérez, EU:C:2014:2371, para 47. For a more extensive discussion of the case law, see Ellis and Watson, EU Anti-Discrimination Law chapter 9; Justyna Maliszewska-Nienartowicz, ‘Genuine and Determining Occupational Requirement as an Exception to the Prohibition of Discrimination in EU Law’ in Thomas Giegerich (ed), The European Union as Protector and Promoter of Equality (Springer 2020); Sara Iglesias Sanchez, ‘The Concept of “Genuine and Determining Occupational Requirements” in EU Equality Law: ACritical Approach’ in Giegerich, The European Union as Protector and Promoter of Equality.

45 Bougnaoui para 40.

46 Case C-188/15 Bougnaoui, ECLI:EU:C:2016:553, Opinion of AG Sharpston, para 96.

47 Lucy Vickers, Religious Freedom, Religious Discrimination and the Workplace (2nd ed, Hart Publishing 2016) 145. See also, McCrea, Religion and the Public Order of the European Union 162.

48 Egenberger para 65; IR v JQ para 51.

49 Egenberger para 66; IR v JQ para 52.

50 Egenberger para 67; IR v JQ para 53.

51 See, in particular, IR v JQ para 54. In Egenberger para 68, by contrast, the CJEU mentioned the proportionality requirement as an additional and separate requirement, but this was confused. After all, the CJEU understood the terms genuine, legitimate, and justified to ensure that the occupational requirement is appropriate and necessary, thus proportionate.

52 For detailed discussion, Rivers, The Law of Organized Religions chapter 8.

53 UK Employment Tribunal, Mohammed v Leprosy Mission [2009] Case no 2303459/09.

54 BAG, 8 AZR 501/14 (25 October 2018); BAG, 2 AZR 746/14 (20 February 2019). ArbG Karlsruhe, 1 Ca 171/19 (18 September 2020).

55 Italics mine.

56 Italics mine.

57 See section 4.A.

58 Case C-74/16 Congregación de Escuelas Pías Provincia Betania, ECLI:EU:C:2017:135, Opinion of AG Kokott, para 31; Case C-414/16 Egenberger, EU:C:2017:851, Opinion of AG Tanchev, para 95; Case C-193/17 Cresco Investigation, ECLI:EU:C:2018:614, Opinion of AG Bobek, para 23.

59 IR v JQ para 48; Egenberger para 58 (italics mine).

60 IR v JQ para 48; Egenberger para 57.

61 For example, Greiner, ‘Kirchliche Loyalitätsobliegenheiten nach dem “IR”-Urteil des EuGH’; Classen, ‘Das kirchliche Arbeitsrecht unter europäischem Druck’.

62 See, for example, Case C-267/06 Maruko, EU:C:2008:179, para 59; Case C-443/15 David LParris, ECLI:EU:C:2016:897, para 58. For further discussion, Bruno de Witte, ‘Exclusive Member State Competences – Is There Such a Thing?’ in Sacha Garben and Inge Govaere (eds), The Division of Competences Between the EU and the Member States: Reflections on the Past, the Present and the Future (Hart Publishing 2017) 61–2.

63 Case C-285/98, Kreil, ECLI:EU:C:2000:2, para 16. See also, David LParris.

64 David LParris paras 57–8.

65 de Witte, ‘Exclusive Member State Competences’ 62.

66 Norman Doe, Law and Religion in Europe: AComparative Introduction (Oxford University Press 2011) 243. See also, Case C-414/16 Egenberger, ECLI:EU:C:2017:851, Opinion of AG Tanchev, para 98.

67 Joined cases C-402/07 and C-432/07 Sturgeon and Others, ECLI:EU:C:2009:716, para 48.

68 See Heinig, ‘Why Egenberger Could be Next’; Unruh, ‘Im Spannungsfeld von Antidiskriminierung und kirchlicher Selbstbestimmung’; Thusing and Mathy, ‘Das deutsche kirchliche Arbeitsrecht vor dem EuGH’.

69 See, for example, Cloots, National Identity in EU Law; Armin Von Bogdandy and Stephan Schill, ‘Overcoming Absolute Primacy: Respect for National Identity Under the Lisbon Treaty’ 48 (2011) Common Market Law Review 1417; Gerhard van der Schyff, ‘The Constitutional Relationship between the European Union and its Member States the Role of National Identity in Article 4(2) TEU’ 37 (2012) European Law Review 563.

70 Case C-213/07 Michaniki, Opinion of AG Maduro, ECLI:EU:C:2008:544, para 33; Von Bogdandy and Schill, ‘Overcoming Absolute Primacy’ 1441; Monica Claes, ‘National Identity: Trump Card or Up for Negotiation?’ in Alejandro Saiz Arnaiz and Carina Alcoberro Llivina (eds), National Constitutional Identity and European Integration (Intersentia 2013); van der Schyff, ‘The Constitutional Relationship between the European Union and its Member States’.

71 Ana Bobić, ‘Constitutional Pluralism is Not Dead: An Analysis of Interactions Between Constitutional Courts of Member States and the European Court of Justice’ 18 (2017) German Law Journal 1395, 1409.

72 Case C-391/09 Runevič-Vardyn and Wardyn, ECLI:EU:C:2011:291, para 83; Case C-208/09 Sayn-Wittgenstein, ECLI:EU:C:2010:806 81; Case C-438/14 Bogendorff, ECLI:EU:C:2016:401, para 48.

73 Which may explain why many favour the use of the principle of proportionality, Von Bogdandy and Schill (n 71) 1441; van der Schyff, ‘The Constitutional Relationship between the European Union and its Member States’ 579; François-Xavier Millet, ‘The Respect for National Constitutional Identity in the European Legal Space: An Approach to Federalism as Constitutionalism’ in Loïc Azoulai (ed), The Question of Competence in the European Union (1st ed, Oxford University Press 2014) 263; Theodore Konstadinides, ‘Dealing with Parallel Universes: Antinomies of Sovereignty and the Protection of National Identity in European Judicial Discourse’ 34 (2015) Yearbook of European Law 127.

74 Case C-399/11 Melloni, ECLI:EU:C:2013:107, para 62 (italics mine). See also, Case C-399/11, Melloni, ECLI:EU:C:2012:600, Opinion of AG Bot, para 126.

75 Case C-42/17, M.A.S. and M.B., ECLI:EU:C:2017:936, para 44 (italics mine). For a good analysis of both decisions, Clara Rauchegger, ‘National Constitutional Rights and the Primacy of EU Law: M.A.S.’ 55 (2018) Common Market Law Review 1521.

76 Among them, Joined cases C-22/08 and C-23/08, Vatsouras and Koupatantze, ECLI:EU:C:2009:344; Case C-413/99 Baumbast, ECLI:EU:C:2002:493; Case C-184/99 Grzelczyk, ECLI:EU:C:2001:458.

77 Case C-333/13 Dano, ECLI:EU:C:2014:2358, para 62. See also, Case C-67/14 Alimanovic, ECLI:EU:C:2015:597; Case C-299/14 García-Nieto, ECLI:EU:C:2016:114. For extensive discussion of these developments, Niamh Nic Shuibhne, ‘Limits Rising, Duties Ascending: The Changing Legal Shape of Union Citizenship’ 52 (2015) Common Market Law Review 889; Daniel Thym, ‘The Elusive Limits of Solidarity: Residence Rights of and Social Benefits for Economically Inactive Union Citizens’ 52 (2015) Common Market Law Review 17.

78 Case C-287/05 Hendrix, ECLI:EU:C:2007:494, paras 52; Case C-406/04 De Cuyper, ECLI:EU:C:2006:491, para 39; Case C-228/07 Petersen, ECLI:EU:C:2008:494, para 52.

79 Case C-211/08, Commission v Spain, ECLI:EU:C:2010:340, para 79. See also, Case C-208/07 von Chamier-Glisczinski, ECLI:EU:C:2009:455, paras 64–5; Case C-345/09 van Delft, ECLI:EU:C:2010:610. For excellent analysis, Herwig Verschueren, ‘The EU Social Security Co-Ordination System: AClose Interplay between the EU Legislature and Judiciary’ in Philip Syrpis (ed), The Judiciary, the Legislature and the EU Internal Market (Cambridge University Press 2012); Nicolas Rennuy, ‘The Emergence of a Parallel System of Social Security Coordination’ 50 (2013) Common Market Law Review 1221.

80 Jan Zglinski, Europe’s Passive Virtues: Deference to National Authorities in EU Free Movement Law (Oxford University Press 2020).

81 Janneke Gerards, ‘Pluralism, Deference and the Margin of Appreciation Doctrine’ 17 (2011) European Law Journal 80, 115. See also, Zglinski (n 82) 159.

82 Jeff King, Judging Social Rights (Cambridge University Press 2012) 130, including the literature referred to. For discussion of arguments in favour of judicial self-restraint in the interpretation of EU legislative decisions, Martijn van den Brink, ‘The European Union’s Demoicratic Legislature’ International Journal of Constitutional Law (forthcoming); Gareth Davies, ‘Legislative Control of the European Court of Justice’ 51 (2014) Common Market Law Review 1579; Phil Syrpis, ‘The Relationship Between Primary and Secondary Law in the EU’ 52 (2015) Common Market Law Review 461.

83 For example, Armin von Bogdandy and others, ‘Guest Editorial: APotential Constitutional Moment for the European Rule of Law – The Importance of Red Lines’ 55 (2018) Common Market Law Review 983; Armin von Bogdandy and Luke Dimitrios Spieker, ‘Countering the Judicial Silencing of Critics: Article 2 TEU Values, Reverse Solange, and the Responsibilities of National Judges’ 15 (2019) European Constitutional Law Review 391; Christian Calliess and Anita Schnettger, ‘The Protection of Constitutional Identity in a Europe of Multilevel Constitutionalism’ in Christian Calliess and Gerhard van der Schyff (eds), Constitutional Identity in a Europe of Multilevel Constitutionalism (2019) 365–7.

84 Case C-673/17 Coman and Others, ECLI:EU:C:2018:385.

85 Floris de Witte, ‘Sex, Drugs & EU Law: The Recognition of Moral and Ethical Diversity in EU Law’ 50 (2013) Common Market Law Review 1545.

86 Claes, ‘National Identity’ 123.

87 See also, Cloots, National Identity in EU Law 196; MDobbs, ‘Sovereignty, Article 4(2) TEU and the Respect of National Identities: Swinging the Balance of Power in Favour of the Member States?’ 33 (2014) Yearbook of European Law 298, 323.

88 Francis Cheneval, Sandra Lavenex and Frank Schimmelfennig, ‘Demoi-Cracy in the European Union: Principles, Institutions, Policies’ 22 (2015) Journal of European Public Policy 1; Francis Cheneval and Kalypso Nicolaidis, ‘The Social Construction of Demoicracy in the European Union’ 16 (2017) European Journal of Political Theory 235; van den Brink, ‘The European Union’s Demoicratic Legislature’.

89 On comparing imperfect alternatives, Neil KKomesar, Imperfect Alternatives: Choosing Institutions in Law, Economics and Public Policy (University of Chicago Press 1997); Adrian Vermeule, Judging under Uncertainty: An Institutional Theory of Legal Interpretation (Harvard University Press 2006).

90 Case C-399/11 Melloni, ECLI:EU:C:2013:107, para 62 (italics mine). See also, Case C-399/11, Melloni, ECLI:EU:C:2012:600, Opinion of AG Bot, para 126.

91 Egenberger para 57.

92 Case C-195/06 Österreichischer Rundfunk, ECLI:EU:C:2007:613, para. 24; C-66/08 Kozłowski, ECLI:EU:C:2008:437, para 42; Case C-400/10 PPU McB, ECLI:EU:C:2010:582, para 41. See for further discussion, Cloots, National Identity in EU Law 337–40.

93 Which seems to have been the approach followed in Case C-51/15, Remondis, EU:C:2016:985, paras 40–1.

94 Given that AG Tanchev reflected on the issue in his Opinion in Egenberger paras 63–4, the CJEU must have known about this.

95 Greiner, ‘Kirchliche Loyalitätsobliegenheiten nach dem “IR”-Urteil des EuGH’; Stefan Greiner, ‘Konsequenzen aus der EuGH-/BAG-Rechtsprechung zur Kirchenmitgliedschaft als Einstellungs- bzw. Kündigungskriterium’ in Hermann Reichold (ed), Kirchliches Arbeitsrecht auf neuen Wegen: Reformbedarf im Recht der Loyalitätsobliegenheiten und in der Pflege (LIT Verlag 2020) 17.

96 See also, IR v JQ para 46.

97 That the status of the Holy See amounts to statehood will be assumed to be correct but is not universally accepted. John RMorss, ‘The International Legal Status of the Vatican/Holy See Complex’ 26 (2015) European Journal of International Law 927.

98 Katja SZiegler, ‘The Relationship between EU Law and International Law’ in Dennis MPatterson and Anna Södersten (eds), A Companion to European Union Law and International Law (John Wiley & Sons, Inc 2016) 43.

99 See also, Peter MHuber, ‘Konkordate und Kirchenverträge unter Europeänisierungsdruck?’ (2008) Archiv für katholisches Kirchenrecht 411.

100 Case C-264/09 European Commission v Slovak Republic, ECLI:EU:C:2011:580, para 41; Case C-84/98 European Commission v Portuguese Republic, ECLI:EU:C:2000:359, para 53; Case C-812/79 Burgoa, ECLI:EU:C:1980:231, para 8.

101 Burgoa (n 102) para 9.

102 Case C-13/93 Minne, ECLI:EU:C:1994:39, para 19. See also, Case C-158/91 Levy, ECLI:EU:C:1993:332, para 22.

103 Allan Rosas, ‘The Status in EU Law of International Agreements Concluded by EU Member States’ 34 (2011) Fordham International Law Journal 1304, 1321.

104 Case C-135/10 Società Consortile Fonografici, ECLI:EU:C:2012:140, para 41; Case C-188/07 Commune de Mesquer, ECLI:EU:C:2008:359, para. 85. For further discussion, Robert Schütze, Foreign Affairs and the EU Constitution: Selected Essays (2014) 109–16; Manuel Kellerbauer, Marcus Klamert and Jonathan Tomkin (eds), The EU Treaties and the Charter of Fundamental Rights: ACommentary (Oxford University Press 2019) 2065–79.

105 For a complete list see, <https://www.iuscangreg.it/accordi_santa_sede.php> (last accessed 14 April 2021).

106 That is also the case when the new agreement is a renegotiation of the old agreement. ‘The Member States are prevented not only from contracting new international commitments but also from maintaining such commitments in force if they infringe Community law’. Case C-467/98 European Commission v Kingdom of Denmark, ECLI:EU:C:2002:625, para 39.

107 For excellent further reading, Ernst-Wolfgang Böckenförde, Mirjam Künkler and Tine Stein, Religion, Law, and Democracy: Selected Writings Vol II (Oxford University Press 2020) chapter 2. See also, Frank JCoppa (ed), Controversial Concordats: The Vatican’s Relations with Napoleon, Mussolini, and Hitler (Catholic University of America Press 1999).

108 On contextual interpretation, Case C-283/81 CILFIT, ECLI:EU:C:1982:335, para 20. See further, Anthony Arnull, The European Union and Its Court of Justice (2nd ed, Oxford University Press 2006) 608.

109 Joined cases C-402/05 Pand C-415/05 PKadi and Al Barakaat, ECLI:EU:C:2008:461, para 304. For discussion, Gráinne De Búrca, ‘The European Court of Justice and the International Legal Order after Kadi’ 51 (2010) Harvard International Law Journal 1; NTürküler Isiksel, ‘Fundamental Rights in the EU after Kadi and Al Barakaat’ 16 (2010) European Law Journal 27.

110 Pieter Jan Kuijper and others, The Law of EU External Relations: Cases, Materials, and Commentary on the EU as an International Legal Actor (2nd ed, Oxford University Press 2015) 799–800; Ziegler, ‘The Relationship between EU Law and International Law’ 49.

111 Case C-170/98 European Commission v Kingdom of Belgium, ECLI:EU:C:1999:411, para 42.

112 Case C-205/06 European Commission v Republic of Austria, ECLI:EU:C:2009:118, para 44; Case C-294/06 European Commission v Republic of Sweden, ECLI:EU:C:2009:119, para 44.