The relationship between the EU Charter of Fundamental Rights and Directives in horizontal situations

L.S. Rossi, Judge of the Court of Justice of the European Union

The Kücükdeveci ambiguity: “derivative” horizontal direct effects for directives?

According to Article 6(1) TEU, the Charter of Fundamental Rights of the European Union (hereinafter, the “Charter” or “CFR”) has the same legal value as the Treaty. After the entry into force of Treaty of Lisbon, the question therefore arises as to whether the ECJ case-law on the direct effects of EU primary law provisions, dating back to Van Gend en Loos, may be extended also to the rights contained in the Charter.

According to a well-established jurisprudence of the Court, which has recently been reaffirmed by the ECJ in its Hein ruling (paras 48-52), the precondition for such direct effects to arise is the existence of an actual conflict of EU and national norms. In other words, direct effects may come into play only when it is impossible to interpret national legislation at issue in a manner consistent with the Charter. In this case, the direct effects of (some of) the Charter’s rights could be inferred from a contrario interpretation of Article 52(5) CFR, as opposed to those provisions of the Charter containing principles, which (according to Article 52(5)) may only be invoked in relations with the implementing European Union or national legislation.

Indeed, the Court already clarified that some provisions of the Charter that have a prescriptive content should in principle have direct effect in vertical situations (ie against State organs). This is, for example, the case of the principles of equal treatment and non-discrimination under Articles 20 and 21 CFR (cf. Glatzel C-356/12 para 43 and Milkova C-406/15, paras 55 and 64).

However, the question of whether the Charter’s rights could also have horizontal direct effects (ie against private parties) has, until recently, remained unresolved. An argument in favour of such effects could, after all, be drawn from the preamble of the Charter, which states that the enjoyment of the rights reaffirmed by the CFR “entails responsibilities and duties with regard to other persons, to the human community and to future generations”.

The ECJ seemed initially to acknowledge a sort of “derivative” horizontal direct effect to some provisions of the Charter, mediated by the directives which give them concrete effect, and anchored to the general principles of EU law or to the constitutional traditions common to the Member States (see, for instance, as for the principle of non-discrimination on grounds of age, Dansk Industri, C‑441/14, paras 22 and 27, reaffirming Mangold, C-144/04, paras 75-78 and Kücükdeveci, C‑555/07, paras 50-51).

As AG Bot has pointed out in his Opinion in Bauer (C‑569/16 and C‑570/16, para 75), this has resulted in an ambiguity, in that the Kücükdeveci case-law could have been construed as recognising to directives giving concrete expression to a general principle codified by the Charter the capacity to be invoked in disputes between private parties. This was in open contradiction with the settled ECJ case-law according to which even a clear, precise and unconditional provision of a directive conferring rights or imposing obligations on individuals, while giving rise to an obligation of consistent interpretation on the part of the national court, cannot of itself be relied upon in disputes between private parties, including for the purpose of setting aside conflicting national legislation (see Faccini Dori, C-91/92, paras 20 and 26, Pfeiffer, C-397/01 to C-403/01, para 109 and Dominguez, C‑282/10, para 42).

From AMS to Max-Planck: the horizontal direct effects of the Charter

It was only very recently that the ECJ case-law shed some light on the unclear legal relationship between the rights contained in the Charter and the directives on which those rights are based and by which they are given concrete expression when invoked in horizontal disputes.

The starting point of the Court’s line of reasoning is represented by the Association de mediation sociale (C‑176/12, paras 45-49). The Court ruled out the possibility to infer from the wording of Article 27 of the Charter, alone and in conjunction with directive 2002/14, a directly applicable rule of law capable of being invoked in a dispute in order to disapply a conflicting national provision. That conclusion was reached on the ground that it was “clear from the wording of Article 27 of the Charter” – which subordinates the workers’ right to information and consultation within the undertaking to the cases and the conditions provided for by EU law and national laws and practice – “that, for this article to be fully effective, it must be given more specific expression in European Union or national law”. In so doing, the ECJ also made clear that it is only the primary law provision and not the directive concretizing it which may have horizontal direct effect, so that the latter “cannot confer on [the former] the qualities needed for it to be relied on directly in a dispute between individuals” (cf. AG Bot in Bauer, point 74).

Nevertheless, AMS left open the question of the possible direct effects of other provisions of the Charter that, unlike Article 27 CFR, make no reference to the respect of “the conditions provided for by national laws and practices”.

In a series of judgments inaugurated by Egenberger in April 2018 (C-414/16, paras 76-79) and followed by IR (C-68/17, paras 69-70), Hein (C-385/17, paras 76-78) and Cresco Investigation (C-193/17, paras 76 and 77), the ECJ has then admitted the possibility of relying on certain rights conferred by the Charter in dispute between private parties. The provisions of the Charter concerned were, precisely, the prohibition of all discrimination on grounds of religion or belief under Article 21(1) CFR, as well as the right to effective judicial protection under Article 47 TFEU. These provisions were deemed to be “mandatory as a general principle of EU law” in the same way as the “the various provisions of the founding Treaties prohibiting discrimination on various grounds”, “ and “sufficient in itself to concern on individuals a rights which they may rely on as such in disputes between them in a field covered by EU law”.

A final and decisive step forward has been marked by the Bauer (C-569/16 and C-570/16, paras 84-86) and Max-Planck (C-684/16, paras 73-75) rulings of November 2018, in which the Court has drawn the same conclusions with respect to the right to a period of paid annual leave affirmed by Article 31(2) CFR, which is not only a social right contained in title IV of the Charter on ‘solidarity’ but is also considered by the Court itself as an “essential principle of EU social law”.

In particular, the Court held that “by providing, in mandatory terms, that ‘every worker’ has ‘the right’ ‘to an annual period of paid leave’ without referring in particular in that regard — like, for example, Article 27 of the Charter […] — to the ‘cases’ and ‘conditions provided for by Union law and national laws and practices’, Article 31(2) of the Charter, reflects the essential principle of EU social law from which there may be derogations only in compliance with the strict conditions laid down in Article 52(1) of the Charter and, in particular, the fundamental right to paid annual leave”. This right “is thus, as regards its very existence, both mandatory and unconditional in nature, the unconditional nature not needing to be given concrete expression by the provisions of EU or national law, which are only required to specify the exact duration of annual leave and, where appropriate, certain conditions for the exercise of that right. It follows that that provision is sufficient in itself to confer on workers a right that they may actually rely on in disputes between them and their employer in a field covered by EU law and therefore falling within the scope of the Charter”.

These judgments appear to have developed a general test to be applied to all the rights protected by the Charter, a test similar – albeit with a different wordings – to that initially set out by the same Court for determining the direct effects of the provision of the Treaty (van Gend en Loos, 26/62, p. 13) and then of directives (van Duyn, 41/74, paras 12-13). This test is based on a twofold condition, according to which the provisions of the Charter are liable to have – not only vertical, but also horizontal – direct effects where they are both (i) unconditional in nature, and (ii) mandatory.

The first condition requires the provisions of the Charter to be “self-sufficient” (cf. AG Bot in Bauer, point 80 and Lenaerts), in that they must not need “to be given concrete expression by the provisions of EU or national law”. The Court has nonetheless stated that the secondary law may specify certain characteristics of the right concerned, such as its duration, and lay down “certain conditions for the exercise of that right” (see Max-Planck, para 74 and Bauer, para 85).

It follows that the numerous provisions of the Charter which refer to rights “as provided for in national laws and practice” are, in principle, deprived of such horizontal direct effect, as the Court has made it clear in AMS (paras 44-45) and confirmed in Max-Planck (para 73) and Bauer (para 84). Although the ECJ has made no explicit reference to it, this can be viewed as an expression of Article 52(6) CFR, according to which “full account shall be taken of national laws and practices as specified in this Charter” (as to the “horizontal reservations” of the Charter, cf. my article in GLJ).

Reference to national laws and practices is made not only by a series of provisions of the Charter concerning social rights, including articles 27 (workers' right to information and consultation within the undertaking), 28 (right of collective bargaining and action), 30 (protection in the event of unjustified dismissal), 34 (entitlement to social security and social assistance), 35 (right of access to preventive health care and the right to benefit from medical treatment), 36 (access to services of general economic interest), but also by articles 9 (right to marry and right to found a family), 10(2) (right to conscientious objection), 14 (freedom to found educational establishments) and 16 (freedom to conduct a business).

The first condition of the Max-Planck test a fortiori rules out also those Charters’ provisions which, by laying down principles to be implemented by legislative and executive acts taken by EU institutions and Member States, are judicially cognisable only in the interpretation of such acts and in the ruling on their legality, pursuant to Article 52(5) CFR. According to the Explanations relating to the Charter, this is, in particular, the case of Articles 25, 26 and 37, while other provisions of the Charter, such as Articles 23, 33 and 34, “may contain both elements of a right and of a principle”.

As regards the second condition – the mandatory nature of the provisions of the Charter – it appears not only to refer to the absolute nature of the right at issue, from which there can be no derogation, but also to incorporate the traditional criteria of clarity and precision required for having direct effects (see Garlsson, paras 65-66 and cited case-law, with regards to the (vertical) direct effect of Article 50 CFR). In so doing, the Court could have implicitly ruled out the horizontal direct effects of those “rights recognised by this Charter for which provision is made in the Treaties” which, in light of Article 52(2) CFR, “shall be exercised under the conditions and within the limits defined by those Treaties.”

Finally, the fact that certain Charter’s provisions may be capable of having horizontal direct effects should not overlook the existence of general constraints to legal effect of the Charter, flowing from the ECJ case-law and the rules governing the interpretation and application of the Charter laid down by Articles 51 and 52 CFR.

The Directives as “pull factor” allowing (horizontal) direct effects of the Charter under Article 51 CFR

As far as Article 51 CFR is concerned, it should first be noted that in its latest case-law the Court has addressed the issue of whether, by limiting the applicability of the Charter to EU institutions and the Member States, the first paragraph of this article could preclude the Charter from being invoked in disputes between private parties.

In Max-Planck (paras 76-79) and Bauer (para 87-90) the Court has held that “although Article 51(1) of the Charter states that the provisions thereof are addressed to the institutions, bodies, offices and agencies of the European Union with due regard for the principle of subsidiarity and to the Member States only when they are implementing EU law, Article 51(1) does not, however, address the question whether those individuals may, where appropriate, be directly required to comply with certain provisions of the Charter and cannot, accordingly, be interpreted as meaning that it would systematically preclude such a possibility”. Based on Egenberger, the Court has acknowledged that “the fact that certain provisions of primary law are addressed principally to the Member States does not preclude their application to relations between individuals”. Moreover, as regards, more specifically, Article 31(2) CFR, the Court emphasised that “the right of every worker to paid annual leave entails, by its very nature, a corresponding obligation on the employer”.

Secondly, the Court has acknowledged in Max-Planck that, although Article 51(1) CFR does not “systematically preclude” that private individuals may be directly required to comply with certain provisions of the Charter, this is without prejudice to the precondition for invoking such a horizontal direct effect, that is, that the legal situation shall fall within the scope of the Charter. According to the same Article 51 CFR as interpreted by the settled ECJ case-law, this is the case when the relevant legal situations are governed by EU law and the national legislation falls within the scope of Union law (see Åkerberg Fransson, C‑617/10, paras 19-21 and AGET Iraklis, C‑201/15, paras 62-64), which cannot be extended by the Charter itself.

The recent ECJ case-law shows how directives and the Charter, notably in horizontal situations, mutually benefit from the respective legal effects in this respect.

When certain rights contained in the Charter are based on and materialised by directives, it is by means of the same directives that the specific legal situation falls within the scope of application of EU law and therefore enters into the orbit of the Charter.

The Charter’s rights, in turn, – insofar as they are applicable and fulfil the relevant conditions laid down by the ECJ case-law – may be invoked in disputes between private parties concerning national provisions implementing or derogating from the corresponding secondary law. This applies also to those rights which were founded on international instruments and in the constitutional traditions common to the Member States and are now enshrined in provisions of the Charter which are ultimately based on directives, even if the latter cannot of itself enjoy such direct effect.

In other words, the Charter cannot confer horizontal direct effects to directives, since the latter, by their very nature, are unable to have such effects. But, as was the case in Max-Planck, the existence of a directive can attract a horizontal situation in the scope of the Charter.

In the synergic functioning of Charter and directives, special attention should be paid not to go beyond their respective scope of application.

Indeed, while the fact that directives and the Charter reinforce each other should be welcomed, there is an inherent risk that this two-way legal relationship may lead to a circular reasoning. This risk could materialise in two situations.

Firstly, as evidenced by the Explanations relating to the Charter, certain rights contained therein are “based” or draws on directives, which in turn concretise these rights. This is notably the case of articles 8 (protection of personal data), 11(2) (freedom of expression and information), 23(1) (equality between women and men), 31 (fair and just working conditions), 32 (protection of child labour and protection of young people at work), 33(2) (family and professional life), but also of article 27 (workers’ right to information and consultation within the undertaking), referring to directives 2002/14/EC, 98/59/EC, 2001/23/EC and 94/45/EC among the “considerable Union acquis in the field” and article 30 (protection in the event of unjustified dismissal), which in addition to the European Social Charter also refers to directives 2001/23/EC and 80/987/EEC.

It is therefore clear that while directives can be assessed, interpreted and applied in the light of the Charter, the latter could not extend the scope of the former, in particular where its own scope is defined by those directives.

Secondly, a risk of circular reasoning may occur when the situations in which fundamental rights protected by the Charter are not covered in their entirety by the relevant directives and could therefore fall (partly) outside the scope of EU law. This holds true especially for minimum harmonization directives, where the higher standards of protection granted by the Charter cannot be invoked against conflicting national legislation which goes beyond the scope of application of the relevant directives.

In such cases, it should not be possible to extend the reach of the relevant directives beyond their scope in order to ensure the effectiveness of the fundamental right granted by the Charter (cf., along this line, Opinion of AG Pitruzzella in CCOO, points 36, 39, 51 and 95-96). In fact, applying by analogy the latter to situations which fall outside the scope of the former, could lead to circumvent Article 51(1) CFR. In order to avoid such a circular reasoning, it would seem appropriate to emphasised that the Charter can only apply (and therefore have horizontal direct effect) to situations which fall within the scope of EU law.

It is to be noted, however, that the Court has already pointed out in Milkova (C‑406/15, paras 52-54) that, when Member States are not required by a Directive to maintain or adopt a given measure but have discretion in that regard, the national legislation that may have been adopted falls within the scope of EU law, resulting in the Charter being applicable. The Court will be able to clarify this point in the cases C-609/17, TSN and C-610/17, AKT, currently pending before it.

Article 52(1) CFR: striking the balance between conflicting rights in horizontal situations

With regard to Article 52 CFR, its first paragraph enables Member States to introduce legislative limitations on the exercise of the rights recognised by the Charter, insofar as they respect the essence of those rights (see Bauer, para 84 and Max-Planck, para 73). According to the same provision, when evaluating these limitations in light of the principle of proportionality, national judges are called to assess whether they “are necessary and genuinely meet objectives of general interest recognised by the Union or the need to protect the rights and freedoms of others”.

Protecting the rights of other parties is even more a delicate issue in cases where the Charter is deemed to be applied in horizontal situation, and national judges are therefore called on to strike a balance between competing individual rights accordingly (cf. Egenberger, para 80). The question therefore arises as to which court should be competent to strike such a balance, which, in light of Article 52(1) CFR, could also result in possible limitations of the rights protected by the Charter.

When the different rights or interests at the stake are both covered by the Charter, the ECJ can be regarded as the best placed to balance them (see, among others, Sky Österreich, C‑283/11, paras 46-66 and AGET Iraklis, paras 70-104). In any other cases, it can be argued that it is the national court which is called to apply the allegedly directly effective right who is primarily responsible to strike this balance in concreto, after referring a preliminary question to the Court, where appropriate.

The intervention of the ECJ would be necessary not only to verify whether and to what extent the specific provisions of the Charter may have horizontal direct effects and to determine the obligations stemming from the Charter in the specific circumstances, but also to clarify what balance has been struck by the legislature between the competing interests involved in the directive(s) which gives concrete expression to the provision of the Charter at issue (see, in this sense, Egenberger, para 81).

The latest ECJ judgments in Hein and Cresco Investigation, however, indicate the Court’s intention to steer this balancing exercise itself.

In Hein (paras 51 and 61-62), the ECJ recalled that, in order to interpret national legislation in accordance with the Charter’s rights and before eventually disapplying it, national courts are under an “obligation to change established case-law, where necessary, if it is based on an interpretation of national law that is incompatible with the objectives of a directive”. National courts are also prohibited from protecting the legitimate expectation of the losing private party that the case-law of the highest national courts confirming the lawfulness of the contested national provisions will continue to apply, so as to avoid indirectly limiting the temporal effects of the ECJ interpretation.

As for Cresco Investigation (paras 79-86), in order to clarify the scope of the national courts’ obligation to guarantee individuals the legal protection afforded to employees under Article 21 CFR, the Court held that the referring court must not only set aside any discriminatory national legislation, but also apply to members of the disadvantaged group the same advantages as those enjoyed by persons within the favoured category. Furthermore, until measures reinstating equal treatment have been adopted by the national legislature, employers are under an obligation to ensure equal treatment among their employees and notably to recognize to those employees who are not members of any churches entitlement to a public holiday on Good Friday or the corresponding pay.

In cases like these, it seems likely that the national constitutional courts will enter into a debate with the domestic courts concerning the protection of the different rights at stake as well as the outcome of the balancing exercise, assessing the conflicting rights in light of the respective national Constitution: in the new scenario opened by the Max-Planck ruling, new questions of constitutional nature and value will emerge.

Barnard & Peers: chapter 6, chapter 9, chapter 20

Photo credit: The Dugger Law Firm, PLLC

What is the point of minimum harmonization of fundamental rights? Some further reflections on the Achbita case.

Eleanor Spaventa, Director of the Durham European Law Institute and Professor of European Law, Law School, Durham University

Ronan McCrea has already provided a very thoughtful analysis of the headscarf cases; this contribution seeks to complement that analysis by focusing on two issues arising from the Achbita case: first of all, the structural problems with the ruling of the Court, both in terms of reasoning and for the lack of information provided; secondly, the more general implications of the ruling for fundamental rights protections and the notion of minimum harmonization in the EU context.

It might be recalled that in the Achbita case a Muslim woman was dismissed from her employer for refusing to remove her headscarf, contrary to the employer’s policy of neutrality, which included a ban on wearing religious symbols. The case then centred on the interpretation of the framework discrimination Directive (2000/78) which prohibits, inter alia, discrimination on grounds of religion. The Belgian and French Government (which had a direct interest because of the Bougnaoui case) intervened in favour of the employee, believing that the discrimination at issue was not justified (Achbita opinion, para 63). The Court, following the Opinion of AG Kokott, found that the rules at issue might constitute indirect discrimination; that the employer’s aim to allegedly maintain neutrality was a legitimate aim as it related to its freedom to conduct a business as protected by Article 16 Charter. It then indicated that the policy was proportionate, if applied with some caveats.

The reasoning of the Court – some structural deficiencies

The headscarf cases are of fundamental importance to the European Union and to all of its citizens, not only those who practice a non-dominant religion, and as such have been widely reported even outside of the EU. One might have expected the Court to engage with a more thorough analysis of the parties’ submissions and of the issues at stake. Instead, we have two very short rulings with very little detail. Just to give an important example – in both cases the French and the Belgian governments sided with the claimants, hence drawing a very important conceptual limit to the principle of laïcité which is justified, in this view, because of the very nature of the State and its duty of neutrality, a duty which cannot be extended to private parties (or if so only exceptionally). This important distinction is not discussed in the ruling, not are the views of the governments who would be directly affected by the rulings.

More importantly though, the fact that the arguments of the parties are not recalled has also more general consequences: as it has been noted by Bruno De Witte elsewhere, the fact that no hermeneutic alternative is provided might give the impression that no hermeneutic alternative is in fact possible, as if legal interpretation is simply a matter of discovering the true hidden meaning of a written text. This approach, not uncommon in civil law jurisdiction but more nuanced in constitutional cases, hides the fact that, especially in cases of constitutional significance, there is more than one legitimate interpretative path that could be chosen, which also reflect different policy alternatives. Interpretation then is also a choice between those different paths: a choice which is, of course, constrained by the relevant legal system and one that might be more or less persuasive.  The failure to acknowledge counter-arguments then results in rulings, like the ones here at issue and many others in sensitive areas, which are not only potentially unhelpful, but also close the door to more effective scrutiny of the reasons that lead the Court to follow a given interpretation.

In the same vein, the analysis of the discriminatory nature of these provisions is rather superficial. In particular, there is no thought given to the fact that contractual clauses allegedly protecting a principle of neutrality, might not only have a discriminatory effect against certain individuals, but might have important inter-sectional (or multiple) discriminatory effects. In other words, a rule banning religious symbols might in fact also have a more pronounced effect on people from a certain ethnic background or a certain gender. Equally disappointing, and in this writer’s opinion legally flawed, is the approach taken in relation to the finding of the potentially indirectly discriminatory effects of the rules at issue. Here, the Court requires the national courts to determine whether the ‘apparently neutral obligation [(not to wear religious symbols)] (…) results in fact in persons adhering to a particular religion or belief being put at a particular disadvantage.” (para 34, emphasis added).

There are two issues to be noted here: first of all, the Court remains silent as to what type of evidence of indirect discrimination is required, and by whom. In discrimination cases, burden of proof is crucial. This is recognised by the discrimination directives at EU level, including Directive 2000/78 which provides that if the claimant shows direct or indirect discrimination, then it is for the ‘respondent to prove that there has been no breach of the principle of non-discrimination’ (Article 10(1)). One would have expected then the Court of Justice to instruct the national court to require the defendants to discharge this duty with a certain rigour, also by means of statistical analysis of the effect of such policies on religious minorities. Yet, the Court does not even engage with this question.

Secondly, and not less important, the Court seems to imply that a rule that discriminates all religious people would not be problematic. For instance if, say, Muslims and Orthodox Jews were equally discriminated against, whilst non-religious persons were unaffected, then, based on the dicta of the Court, there would be no discrimination. This interpretation seems restrictive and not supported by the text of the directive (or the Charter) that refers to discrimination on grounds of religion in general. In any event, in discrimination cases it is crucial to identify the comparator, and the Court fails to do so clearly and to support its choice with sound legal arguments. But, beside these very important structural issues, the Achbita ruling raises other more technical as well as general issues, as to the extent to which the Court’s interpretation might affect the Member States’ discretion to provide more extensive protection that that provided for in the Directive.

Minimum harmonization and fundamental rights

Directive 2000/78 is intended only to set minimum standards, so that Member States can, if they so wish, provide for a more extensive protection. Indeed many Member States have done so by extending either the protected categories of people, or the field of application of the legislation, or both. In theory then, the Achbita ruling should not be seen as the last word in relation to the treatment of religious people at work. After all, if Belgium or France or any other country finds the ruling problematic, it can simply pass legislation prohibiting private employers from requiring religious neutrality from its employees, unless of course a specific dress code is necessary to ensure the health and safety of the worker or the public. Viewed in this way, and notwithstanding the structural problems identified above, the ruling seems very sensible: it is agnostic, in that it does not impose either model on Member States, allowing therefore a degree of variation in a very sensitive area, something which, as eloquently discussed in McCrea’s post, might not be a bad thing. After all, this is the same path that has been taken by the European Court of Human Rights.

However, things are slightly more complicated in the European Union context. In particular there is nothing in the ruling to indicate that the Directive sets only minimum standards so that it would be open to those Member States to go further in protecting people holding religious beliefs. And, more crucially, the Court, mirroring the opinion of Advocate General Kokott, refers to the EU Charter of Fundamental Rights when assessing the legitimacy of the justification put forward by the employer. In particular, it finds that the business’s wish to ‘project an image of neutrality (…) relates to the freedom to conduct a business that is recognised in Article 16 of the Charter and is, in principle, legitimate’.

The reference to the Charter, which indirectly frames the question as a clash of fundamental rights, is important because, in the EU context, when the Charter applies it sets the fundamental rights standard. In simpler terms this means that should a Member State wish to provide more extensive protection to ensure that employees are not discriminated on grounds of their religious belief, something that is allowed under Directive 2000/78, it might be prevented from doing so since, pursuant to the Achbita ruling, it would infringe the right to conduct a business as protected by the Charter. In this way, far from leaving the desired flexibility and discretion to the Member States, the Court sets the standard – employers have a fundamental right, albeit with some limitations, to limit the employees’ right not to be discriminated against. One might well ask then, much as it has been remarked in relation to the Alemo Herron case, what is the point of minimum harmonization directives if the upward discretion of the Member States is so curtailed.

Conclusions

The Court of Justice did not have an easy task in the Achbita case: it was pretty much a ‘damned if you do, damned if you don’t’ scenario. For sure, some of us would have liked the balance at issue to be tilted firmly in favour of religious minorities, especially given the growing evidence of attacks and discrimination against, particularly, Muslim women. The Court chose a different path and that is, of course, within its prerogatives. However, the way that path was trodden upon leaves many open questions both in relation to the way the result was achieved, and to the many questions it overlooks. What is most troubling is the implication that the freedom of Member States to provide greater protection towards minorities may, in principle, be constrained by the Court’s interpretation of the freedom to conduct a business.

Barnard & Peers: chapter 9, chapter 20

Photo credit: smallbusiness.co.uk

Faith at work: the CJEU’s headscarf rulings

Ronan McCrea*

* Senior Lecturer, UCL Faculty of Laws and author of Religion and the Public Order of the Euorpean Union (OUP 2010) and Religion et l’ordre juridique de l’Union européenne (Bruylant 2013).

Almost seventeen years on from the adoption of the Framework Directive on discrimination in employment, the Court of Justice has issued its first major decisions in relation to discrimination in employment on grounds of religion. In doing so the Court entered into territory that is extremely fraught in political terms and is therefore highly dangerous for an international court such as the CJEU.

The degree to which multi-faith societies can require individuals to refrain from expressing possibly controversial religious identities or beliefs in shared spaces such as the workplace has never been an easy question. But it has become significantly more difficult in recent years since the question of religion’s role in society has become bound up with highly combustible political issues such as migration, changing norms in relation to gender and sexuality, national identity and even national security.

It must therefore have been with considerable trepidation that the Court of Justice proceeded to give its ruling on two cases that involved challenges by two women who lost their jobs for refusing to remove the Islamic headscarves while at work.

The facts of the two cases were subtly, but importantly, different. In Achbita (Case C-157/15), the Claimant began working as a receptionist at G4S in February 2003 and complied with what was, at the time, an unwritten rule within G4S that workers could not wear visible signs of their political, philosophical or religious beliefs at work. In April 2006 Ms. Achbita informed her employers that she intended to wear an Islamic headscarf at work and was told that she could not do so because this violated G4S’s rule requiring philosophical and religous neutrality in their employees attire. In May 2006 G4S adopted a change to workplace rules making the ban on visible signs of political, philosophical or religous belief a written rule and in June 2006 Ms. Achbita was fired for her insistence on wearing the headscarf at work.

In Bougnaoui (Case C-188/15), the Claimant was informed by a representative of Micropole at an October 2007 student recruitment fair that wearing an Islamic headscarf may pose problems when she was in contact with customers. She began to work at Micropole in February 2008 initially wearing a bandana and then a headscarf. In May 2009, a customer of Micropole’s with whom Ms. Bouganoui had worked, informed her employers that Ms. Bouganoui’s wearing of the headscarf had upset some of their employees and requested that there be “no veil next time”. Ms. Bouganoui refused her employers request to confirm that she would agree not to wear the headscarf on future occasions and was fired in June 2009.

The Belgian and French Courts of Cassation both referred questions relating to the prohibition on discrimination in employment on grounds of religion or belief to the Court of Justice which, given their importance, decided to attribute both cases to the Grand Chamber.

In relation to Achbita the Belgian court asked  whether a ban on a female Muslim employee wearing the headscarf at work should be regarded as direct discrimination when the employer in question bans all employees from wearing any outward sign of political, philosophical or religious beliefs at work. This is potentially important as under the Directive, a directly discriminatory rule can only be justified by a “genuine and determining occupational requirement”. Indirectly discriminatory rules, on the other hand, can be accepted if it is shown that they serve a legitimate aim and are pursued by proportionate and necessary means.

In Bouganoui, the French court asked the Court of Justice whether the wish of an customer not to have services supplied by an employee in an Islamic headscarf could be seen as a genuine and determining occupational requirement under the Directive (seemingly assuming that the restriction in question was directly discriminatory).

Therefore, both claims focused on the issue of direct discrimination. However, in addition to ruling on the issue of direct discrimination, the Court of Justice decided to give significant guidance in relation to the question of justification of bans on religious symbols as indirectIy discriminatory measures.

In both cases, the Court noted that the Directive does not define religion but does refer to the rights contained in the European Convention of Human Rights which include the right to freedom of thought, conscience and religion in Article 9. It also notes the reference to the common constitutional traditions of the Member States which it notes were reaffirmed in the EU Charter of Fundamental Rights which includes a similar right in Article 10. Both of these rights include, the Court found, the right to manifest religious faith in public.

General Bans on Symbols of Opinion and Indirect Discrimination

However, in relation to Ms. Achbita, the Court found that the rule preventing her from wearing her headscarf at work did not amount to direct discrimination as it referred to visible signs of political, philosophical or religious beliefs and thus “covers any manifestation of such beliefs without distinction”. The Court therefore concluded that the rule “must (…) be regarded as treating all workers of the undertaking in the same way, by requiring them, in a general and undifferentiated way, inter alia, to dress neutrally, which precludes the wearing of such signs.”

Having answered the referring Court’s question as to whether the rule in question constituted direct discrimination in the negative, the Court decided to provide additional guidance as to how the national court, which has the authority to decide factual matters, should approach the issue of indirect discrimination. It did so on the basis that it was “not inconceivable”that the referring court might conclude that the rule in question was indirectly discriminatory in that it was “an apparently neutral obligation it encompasses results, in fact, in persons adhering to a particular religion or belief being put at a particular disadvantage” .

The guidance given provides significant illumination as to the approach of the Court to the reconciliation of religious freedom with rules that seek to constrain religious expression in particular contexts in multi-faith societies.

The Court, in common with the European Court of Human Rights in Eweida concluded that in principle, the desire, on the part of an employer to project an image of neutrality “must be considered legitimate”. It bolstered this conclusion with reference to the freedom to conduct a business under Article 16 of the Charter of Fundamental Rights of the EU which it feels weighs in favour of the employers’ rights in this regard “notably” when the rule covers only workers “who are required to come  into contact with the employer’s customers”.

However, the Court stressed that a rule restricting religious symbols or attire can only be seen to be appropriate when it is part of a neutrality policy that “is genuinely pursued in a consistent and systematic manner”. Whether this was the case in relation to Ms. Achbita, was, the CJEU ruled for the national court to decide on the facts.

Despite its emphasis on the role of the national court in applying the Court of Justice’s guidance to the case, the judgment gave a notably strong steer to the national judges concluding that if it were the case that the prohibition covered “only G4S workers who interact with customers (…) the prohibition must be considered strictly necessary for the purpose of achieving the aim pursued” though it did note that the national court must assess wehther it would have been possible to find Ms. Achbita a non-customer-facing role.

The judgment in respect of Ms. Bouganoui took as similar approach. The Court stated that a generally applicable ban on all visible symbols of religious, philosophical or political belief would be indirectly discriminatory and referred explicitly to the guidance given in Achbita for the assessment of the legitimacy, proportionality and necessity of such a ban.

The Court said it was for the national court to decide if Ms. Bouganoui’s dismissal was based on non-compliance with such a general ban. If the decision to dismiss was not based on a general ban but was specific to the headscarf, then it would be necessary to answer the question posed by the national court, namely, whether compliance with a request from a client that the employee refrain from wearing an Islamic headscarf at work could be seen as a “genuine and determining occupational requirement” that could justify a directly discriminatory policy.

On this matter, the Court gave a clear answer. It noted that only in very limited circumstances can characteristic related to religion constitute a genuine and determining occupational requirement. Compliance with a client request such as that made in this case did not meet the Directive’s requirement that a discriminatory rule be justified “by reason of the nature of the particular occupational activities concerned or of the context in which they are carried out”.

Significance of Client Preferences

There are a number of interesting features of the Court’s reasoning in these cases. There may appear to be something of a tension between the two rulings in relation to the role of customer preferences as a basis for the restriction of religious expression on the part of employees. On the one hand, in Achbita, the Court appears to say that the need of the employer to present a neutral image to clients makes it more justifiable to impose a neutrality requirement on employees with customer-facing roles. On the other hand, in Bouganoui, the Court found that compliance with a client’s request for “no veil next time” could not be seen as a “genuine and determining occupational requirement”.

However, to see a conflict here is to misunderstand the Court’s reasoning. In its discussion of the significance of customer-facing roles in Achbita, the Court was focusing on justification of an indirectly discriminatory general ban on all religious, political and philosophical symbols and assessing whether such a general ban would meet the conditions of proportionality, legitimacy and necessity required by the Directive in order to justify such indirectly discriminatory measures.

In contrast, in Bouganoui, the Court was assessing justification of a directly discriminatory criterion, namely the client’s request for “no veil next time”. In other words, the fact that the client in Bouganoui made a request that appeared to target the symbols of a particular faith rather than seeking a general ban on all religious, philosophical and political symbols, made compliance with this request a matter of direct not indirect discrimination. As the test for justification of directly discriminatory measures (“genuine and determining occuptational requirement”) is so much more demanding than that for indirectly discriminatory measures, the reasons for the apparent contrast in outcomes in the two cases becomes clear. Customer preferences may be sufficient to justify an indirectly discriminatory measure but cannot justify a directly discriminatory one.

The focus on the question of whether an employee has a customer-facing role in Achbita is also raises the question of the permissibility of dress-code restrictions for those without customer-facing duties. The Court was clear that interaction with customers was a factor that increased the scope for an employer to require an employee to obey a general and systematic ban on symbols of religious, political or philosophical belief. It is unclear whether this means that it is impermissible to impose such constraints on employees without customer facing roles. Given the controversial nature of many religious and other beliefs, it is conceivable that employers will aim to preserve workplace harmony between employees by imposing bans on symbols of belief at work. Indeed, in the well-known case of Ladele, the objection to accommodating a registrar who refused to carry out same sex civil partnerships came not from any clients but from her fellow employees.

Definition of Religion for Purposes of Discrimination

The fact that the Court relied to a significant degree on the definition of religion in the Article 9 jurisprudence of the European Court of Human Rights did not pose problems in this case. However, there is a certain tension between the right to freedom of religion and belief and the idea of indirect discrimination on grounds of religion. The right to freedom of religion or belief has (rightly) been regarded by the courts as primarily an individual right that allows individuals to choose their beliefs and which does not distinguish between widely shared established beliefs and idiosyncratic or heterodox beliefs and does not favour religious over non-religious forms of belief.

Indirect discrimination on the other hand, has generally involved notions of collective disadvantage and the granting of extra rights to individuals who show they are part of a group facing additional “head-winds” on account of a salient characteristic shared with other members of that identifable group. An individual with a belief shared by no one else may not be able to demonstrate such collective disadvantage (see for example the approach of the English and Welsh Court of Appeal in Eweida to indirect discrimination on grounds of religion in Directive 2000/78). Thus, the individualistic approach to religion that is appropriate in relation to cases focusing on religious freedom may not always be appropriate in relation to questions of indirect discrimination on grounds of religion where group disadvantage will be part of the analysis. Therefore, the CJEU’s approach of relying on the definition of religion used in fundamental rights litigation in relation to anti-discrimination cases may cause problems in the future (see R. McCrea “Singing from the Same Hymn Sheet? What the Differences between the Strasbourg and Luxembourg Courts Tell Us about Religious Freedom, Non-Discrimination and the Secular State” Oxford Journal of Law and Religion (2016) 5(2) 183-210).

Conclusion

Given the political salience of the issues underlying these cases, the fact that the Court of Justice has adopted a cautious approach is not surprising. There is no consensus in Europe about how best to approach the issue of religion’s role in public life in the context of religous and demographic change. Various countries have tried different approaches. Some countries, such as the UK, have felt it best to allow religious expression in a wide range of public contexts. Others, such as France, have taken the opposite approach and have pursued a policy that sees coexistence as best served by a degree of reticence in relation to religious expression in non-private contexts. Each approach has its critics. Many French people see the approach adopted in France as overly restrictive, just as many British people argue that cohesion and coexistence have not been well served by the UK’s approach. Other states like the Netherlands have switched to some degree from one approach to the other.

In addition, the status of religion, and the issue of whether religious belief should be regarded as being “special” in the sense of being entitled to a greater level of protection than is provided to other forms of belief, is a question that has a significant impact on issues of fundamental rights (where religious freedom can come often at the cost of the rights of others) and constitutional law, where the idea of religion as a private and individual matter is woven into the constitutional norms of a number of European states.

Therefore, it is unsurpising that the Court of Justice has been cautious in its approach. As I wrote previously on this blog, regulating religion’s role in public life is difficult because religion is both a matter of identity (akin to race) and a matter of belief (akin to political opinion). If one views religion as a set of beliefs this calls out for treatment that is entirely contrary to the treatment that would be appropriate if religion is regarded as a form of identity. Different approaches are appropriate in different scenarios. In relation to blasphemy or anti-apostacy laws, it is probably best to see religion as a set of beliefs, in relation to the right to receive services in a shop, it is probably best to see it as a form of identity. In relation to the workplace, there are good reasons for both approaches so it is appropriate that a degree of leeway is allowed, provided that restrictions are applied in a fair and balanced manner.

However, the story of the judgments in Achbita and Bouganoui is not a story of simple deference on the part of the European Court. While the Court of Justice has upheld the compatibility of rules prohibiting the wearing of religious symbols at work with the Directive it has, at the same time taken steps to ensure that such rules do not become a means to target adherents to minority or unpopular faiths. Rules restricting religious influence over law and politics or limiting religous expression in public contexts that were enacted in good faith have sometimes become the subject of exploitation in bad faith by some with exlusionary agendas. The Front National in France, for example, has in recent years, discovered a fervent love for laïcité that it did not have before they discovered that it could be used as a stick with which to beat French Muslims.

I have written before (Religion and the Public Order of the European Union, chapters 6 and 7, OUP 2010) about how the legal and political order of the European Union has tended to percieve more readily threats to secular law and politics and to egalitarian values when they come from minority faiths than when they come from culturally-entrenched forms of Christianity. This does not mean that restrictions on religion in politics or public life ought to go, often they serve important goals. But it does mean that it is important to ensure that such restrictions are applied fairly. The Court of Justice has taken an important step in these cases in this regard. By insisting that bans on religious symbols can be justified only if part of a genuinely systematic and generally applicable prohibition on the display of visible symbols of all kinds of religious, philosophical or political belief, the Court has sought to ensure that the often justifiable desire to curtail expression of controversial beliefs in the workplace cannot be used as a means to selectively target unpopular minorities.

This is an approach that will leave some unsatisfied. It gives employers significant control over the appearance of employees. In addition, what is seen as neutral is culturally specific and so compliance with neutral rules will be more difficult for adherents to minority faiths. However, as noted above, such rules can serve important goals and national courts will still be able to assess in each case whether the operation of the rules in question is proportionate. In any event, it is not clear that any other option was open to the Court of Justice. There are good arguments for and against allowing religous symbols at work and what is fair and appropriate in one context may be unfair an oppressive in another. Sweeping away all workplace rules accross 28 Member States that imposed a general ban on visible expressions of belief and opinion would represent a striking degree of certainty for an international court and would have risked placing the Union on a constitutional collision course with at least one major Member State. The solution of recognising the legitimacy of general bans but requiring that such bans avoid targetting specific faiths seems like a reasonable one.

Barnard & Peers: chapter 9, chapter 20

Photo credit: nathuLAW

Another failed opportunity for the effective protection of LGB rights under EU law: Dr David. L. Parris v. Trinity College Dublin and Others

Dr Alina Tryfonidou, Associate Professor in EU Law, University of Reading

Introduction

The recent judgment in the Parris case is another failed opportunity for the ECJ to offer effective protection to LGB persons and same-sex couples under EU law. Despite some signs in recent cases (Asociaţia Accept and Hay) that the EU Court has been taking its role as protector of the rights of LGB persons under EU law more seriously, in its judgment in Parris, like in its decision in Léger (discussed here) which was the last case involving LGB persons that was decided prior to Parris, the Court seems to be treading cautiously around matters that are delicate from the point of view of the Member States. In particular, in this case, the Court seems wary of the danger of being accused of imposing its own views with regards to a matter for which there is – still – great diversity of opinion among the Member States, namely, the recognition of same-sex relationships. The judgment, also, demonstrates the ECJ’s failure to accept the reality of multiple discrimination, as it was ruled that if a measure does not give rise to discrimination on any of the grounds prohibited by Directive 2000/78 – when these grounds are taken in isolation – then it cannot be considered to produce discrimination on the basis of the combination of those two factors.

Legal and Factual Background

The request for a preliminary ruling in the Parris case was referred by the Labour Court (Ireland) hearing an appeal from a decision of the Equality Tribunal (Ireland) in proceedings brought by Dr David L. Parris – a retired academic – against Trinity College Dublin (his former employer), the Higher Education Authority (Ireland), the Department of Public Expenditure and Reform (Ireland) and the Department of Education and Skills (Ireland), arguing that he had been discriminated against by the defendants by reason of his age and sexual orientation. The proceedings concerned the refusal by Trinity College Dublin to accept Dr Parris’s request that on his death, the survivor’s pension provided for by the occupational benefit scheme of which he was a member, should be granted to his civil partner. The refusal was based on the fact that Dr Parris entered into a civil partnership with his male partner only after he had turned 60 and the said occupational scheme provides that survivor’s pension is payable only if the claiming member married or entered into a civil partnership before reaching the age of 60. (Note that the civil partnership was entered into in the UK in 2009, once Dr Parris was over 60, but was only recognised in Ireland from 2011 onwards, when the Irish legislation regarding civil partnerships came into force).

In Ireland, civil partnerships can only be entered into since January 2011, whilst marriage between persons of the same sex has been made available only since November 2015. In addition, the statute which gave same-sex couples the right to enter into a civil partnership, excluded the retrospective recognition of civil partnerships registered in another country, which meant that civil partnerships entered into abroad could be recognised in Ireland only prospectively, from January 2011. Thus, as Dr Parris was born in 1946, he could only enter into a civil partnership or marry his same-sex partner in Ireland after reaching the age of 60; and, similarly, any civil partnership he had entered into in another country, could only be recognised in Ireland after he had reached the age of 60.

This meant that under no circumstances would a person who had Dr Parris’s sexual orientation and age be able to claim a survivor’s benefit for his (same-sex) civil partner or spouse under the contested pension scheme. Or, to put the issue more broadly, LGB persons born before 1 January 1951 are excluded in all instances from claiming a survivor’s benefit for their same-sex civil partner or spouse under the contested pension scheme.

The main question of the referring court was whether the application of a rule in an occupational benefit scheme specifying an age by which its members must marry or enter into a civil partnership for their spouse or civil partner to be entitled to a survivor’s pension, amounts to discrimination on grounds of age and/or sexual orientation, contrary to Directive 2000/78.

The AG Opinion

In her Opinion, Advocate General Kokott firstly noted that the contested rule does not amount to direct discrimination on the ground of sexual orientation, as ‘[t]he mere fact that an employee has not married or entered into a civil partnership before his 60th birthday – whether on account of legal barriers or by choice – is not directly linked to his sexual orientation … Had Dr Parris married a woman after his 60th birthday, for example, she would have been excluded from eligibility for the survivor’s pension in exactly the same way as his current partner under the terms applicable to that pension’. The Advocate General then pointed out that the rule does, however, amount to indirect discrimination on the ground of sexual orientation since ‘the 60-year age limit affects a large number of homosexual employees in Ireland more severely and more deleteriously than their heterosexual colleagues … all homosexual employees in Ireland who were born before 1951 were universally barred from entering into a civil partnership in good time before their 60th birthday because the institution of civil partnership did not exist in that Member State until 2011 and the best option previously available to same-sex couples was to live together as “common-law” partners. It was therefore impossible for legal reasons for that group of people to secure a survivor’s pension for their respective partners under the occupational pension scheme at issue and thus to provide the latter with a form of social protection that their heterosexual colleagues and their spouses were able to take for granted’. The Advocate General also found that there was (unjustified) direct discrimination on the ground of age as ‘employees who do not enter into a marriage or civil partnership until after they have reached their 60th birthday are treated less favourably than employees who do so at a younger age’.

Despite the fact that the Advocate General found that the contested rule can amount to discrimination on the grounds of sexual orientation and age taken separately, her preferred approach was to consider that the rule is discriminatory on the combined grounds of sexual orientation and age: ‘In the present case, particular attention will have to be given to the fact that any discrimination perpetrated against the person concerned is attributable to a combination of two factors, age and sexual orientation. The Court’s judgment will reflect real life only if it duly analyses the combination of those two factors, rather than considering each of the factors of age and sexual orientation in isolation.’ The Advocate General explained that ‘employees such as Dr Parris would, in accordance with Article 2(2)(b) of Directive 2000/78, have to be regarded as being at a particular disadvantage by reason of a combination of their sexual orientation and their age because the terms of the pension scheme have the effect of systematically depriving their surviving partners in particular of a survivor’s pension. It is true that, for all employees, the surviving partner’s eligibility for a survivor’s pension is subject to the (apparently neutral) condition that the couple must have entered into a marriage or civil partnership before the employee’s 60th birthday. In truth, however, this systematically excludes homosexual employees born before 1951 in particular – unlike all other categories of employee – from a survivor’s pension of this kind because those employees would never have been able to satisfy the aforementioned condition even if they had wanted to’.

The Judgment

The Court in its judgment was of the view that the contested rule does not give rise to direct discrimination on the ground of sexual orientation because it does not refer directly to the worker’s sexual orientation. Unlike the Advocate General, however, the Court also found that the contested rule did not give rise to indirect discrimination on this ground either.

The Court began by considering the reason behind Dr Parris’s failure to satisfy the contested rule:

‘on the date on which Mr Parris retired, 31 December 2010, he did not satisfy the conditions laid down by the applicable national rule for his civil partner to be entitled to the survivor’s benefit at issue in the main proceedings, since the civil partnership he had entered into in the United Kingdom was not yet recognised in Ireland, and in any event, even if it had been recognised, it could not have given an entitlement to such a benefit, as it had been entered into after the member’s 60th birthday.’

‘the fact that Mr Parris is unable to satisfy that condition is a consequence, first, of the state of the law existing in Ireland at the time of his 60th birthday, in particular the absence at that time of a law recognising any form of civil partnership of a same-sex couple, and, secondly, of the absence, in the rules governing the survivor’s benefit at issue in the main proceedings, of transitional provisions for homosexual members born before 1951’.

The Court then – referring to Recital 22 of Directive 2000/78 – proceeded to highlight the deference it shows towards Member State laws regarding the regulation of marital status in their territory and, in particular, the legal recognition of same-sex relationships: these are matters with respect to which Member States have maintained their full competence, and, thus, they can regulate them in whichever way they choose, provided that when doing so they comply with their obligations under EU law. The Court then explained that the ‘Member States are thus free to provide or not provide for marriage for persons of the same sex, or an alternative form of legal recognition of their relationship, and, if they do so provide, to lay down the date from which such a marriage or alternative form is to have effect’. From this, the Court concluded that ‘EU law, in particular Directive 2000/78, did not require Ireland to provide before 1 January 2011 for marriage or a form of civil partnership for same-sex couples, nor to give retrospective effect to the Civil Partnership Act and the provisions adopted pursuant to that act, nor, as regards the survivor’s benefit at issue in the main proceedings, to lay down transitional measures for same-sex couples in which the member of the scheme had already reached the age of 60 on the date of entry into force of the act’. Accordingly, in the ECJ’s view, the contested rule did not produce indirect discrimination on grounds of sexual orientation.

The Court, however, found that the contested measure did establish a difference in treatment that was directly based on the criterion of age: ‘such a rule thus treats members who marry or enter into a civil partnership after their 60th birthday less favourably than those who marry or enter into a civil partnership before reaching the age of 60’. This difference in treatment, nonetheless, falls – according to the Court – within the scope of Article 6(2) of the Directive, as it ‘fixes an age for entitlement to an old age benefit’ and, hence, it does not constitute discrimination on grounds of age.

The final issue that the Court had to consider was that of multiple discrimination, i.e. whether the contested rule was capable of creating discrimination as a result of the combined effect of sexual orientation and age, where that rule does not constitute discrimination either on the ground of sexual orientation or on the ground of age taken in isolation. The Court noted:

‘while discrimination may indeed be based on several of the grounds set out in Article 1 of Directive 2000/78, there is, however, no new category of discrimination resulting from the combination of more than one of those grounds, such as sexual orientation and age, that may be found to exist where discrimination on the basis of those grounds taken in isolation has not been established.’

‘Consequently, where a national rule creates neither discrimination on the ground of sexual orientation nor discrimination on the ground of age, that rule cannot produce discrimination on the basis of the combination of those two factors’.

Analysis

The judgment in the Parris case confirms and further highlights two trends that had already been prevalent in previous case-law: a) that the Court is reluctant to intervene in situations which touch on matters that fall to be regulated exclusively at Member State level, especially when such matters involve morality judgements for which there is great diversity of views among the Member States and b) that the Court ignores the reality of multiple discrimination.

a) Sensitive Matters that Fall within Exclusive Member State Competence

As seen earlier, Recital 22 of Directive 2000/78 played an important role in the Court’s conclusion in the case that the contested rule did not amount to (indirect) discrimination on the ground of sexual orientation. The Recital provides that ‘[t]his Directive is without prejudice to national laws on marital status and the benefits dependent thereon’. In relation to this, the Court in its judgment explained that the Member States are free to decide whether to open marriage or registered partnerships to persons of the same sex and if they do so to lay down the date from which such a marriage or alternative form is to have effect. From this it concluded that the refusal of the survivor’s benefit was – simply – a consequence of the application of Irish law concerning same-sex partnerships (and, in particular, the lack of a civil status for such partnerships at the relevant time), and, hence, respecting the competence of the Member States with regards to these matters, EU law (and, in particular, Directive 2000/78) could not apply in order to require Ireland to ‘to provide before 1 January 2011 for marriage or a form of civil partnership for same-sex couples’.

This line of reasoning appears – with respect – to be erroneous. In areas like this (i.e. legal recognition for same-sex relationships) which fall to be regulated by Member States exclusively, the ECJ can still intervene in order to require the said legislation to be applied in a manner which is compliant with EU law. Hence, the application of Directive 2000/78 in this instance would require the removal of discrimination on the (combined) grounds of sexual orientation and age – which, in my view, is the discrimination that was suffered on the facts of the case – which could be achieved by an amendment of the rule (most likely, the requirement would be to permit LGB persons born before 1951 to claim survivor’s benefit for their same-sex partner even if they entered into a civil partnership or marriage after they turned 60), but it wouldn’t require Ireland to recognise such relationships retrospectively, by changing the date from which they have effect (which is a matter that falls to be regulated exclusively by Ireland). In other words, Ireland would remain free to determine how to regulate same-sex relationships – as required by Recital 22 of Directive 2000/78 – but would have to require pension schemes etc which apply to such relationships to make provision for the different legal situation of persons that have such relationships and to take that into account, by providing for an exception to the rule in situations where it is (legally) impossible to satisfy the age condition due to the law in Ireland (as opposed to the personal choice of the couple), which only allowed the legal recognition of same-sex relationships after a certain date.

As noted by Advocate General Kokott in response to the argument of the defendants, the UK Government, and the Commission, that a finding of discrimination based on sexual orientation in this case could have the consequence of conferring de facto retroactive effect on the institution of civil partnership (which would go against Recital 22 of Directive 2000/78), such a finding ‘does not in any way compel the Irish State to change the marital status of an employee such as Dr Parris retroactively’ as ‘Dr Parris and his partner are today recognised by the Irish State as living together as a couple, and they are today claiming – prospective – occupational pension scheme benefits corresponding to their marital status as it stands today. They are not in any way claiming a benefit to which their marital status does not entitle them. They are certainly not claiming such a benefit retroactively. Nor are they seeking a retroactive change to their marital status. Rather, they are simply defending themselves against a term contained in the occupational pension scheme at issue – the 60-year age limit – which was laid down in the past but discriminates against them today.’

Accordingly, by hiding behind Recital 22, the Court seems to be avoiding to intervene in this case, in this way allowing Member States not merely to regulate same-sex relationships and the consequences ensuing from entering them (which is, indeed, a matter that is wholly to be regulated at Member State level), but also to discriminate against LGB persons who – by virtue of a legal disability (i.e. their inability to enter into a marriage or registered partnership in a certain Member State until a certain date) – are differently situated from heterosexual persons who had the (legal) option of entering into a marriage or registered partnership by the required age, but chose not to do so. This approach seems to be in line with the approach followed by the Court with regards to issues involving fundamental societal choices. In particular, in relation to matters which involve deeply held national societal mores or values, the Court and the EU legislature have been very reserved in their approach and have focused on respecting the sovereignty of the Member States, even to the extent of avoiding applying EU law rigorously (e.g. Henn and Darby) or at all (e.g. Grogan). This nonetheless comes at a cost, this being that the rights that individuals derive from EU law are sacrificed at the altar of Member State sovereignty. Should there not be a requirement that the EU, which is a polity that values, inter alia, fundamental human rights and equality, act as an external arbiter of the choices of the Member States with regards to these issues when these choices come into conflict with the rights that individuals derive from the Treaty and secondary legislation? In other words, should the EU not come to the rescue of individuals that derive rights from EU law and require the Member States to ‘think federal’ with regards to these matters, as long as the EU does not impose its own views in relation to them? It is not suggested here that morality and value judgements should now be made at the EU level; as Weiler has noted, there should be ‘fundamental boundaries’ which are ‘designed to guarantee that in certain areas communities […] should be free to make their own social choices without interference from above’.[1] However, Member States should be aware that when regulating these matters they must take into account and cater for the rights that individuals enjoy under EU law, and if they do not, then the ECJ or (in most instances) the national courts as enforcers of EU law, should intervene in order to ensure that Member States comply with their obligations under EU law. 

b) Multiple Discrimination

The other notable feature of the judgment in Parris is the Court’s express rejection of the possibility that multiple discrimination can be prohibited by Directive 2000/78. The Court had, already, been faced with a situation involving discrimination on the combined grounds of sex and sexual orientation in the Léger case, which was decided in 2015. Despite the fact that the Advocate General in that case found that the contested measure did amount to such (multiple) discrimination, the Court in that case simply brushed aside the matter, by focusing on the question of whether the said measure amounted (simply) to discrimination on the ground of sexual orientation.

In this case, however, it was more difficult for the Court to avoid the matter as the possibility of multiple discrimination – in this instance on the combined grounds of sexual orientation and age – was specifically mentioned by the referring court and the third question referred was, exactly, focused on this issue. However, as noted earlier, the Court explicitly pointed out that ‘where a national rule creates neither discrimination on the ground of sexual orientation nor discrimination on the ground of age, that rule cannot produce discrimination on the basis of the combination of those two factors’, thus rejecting the possibility of a finding of multiple discrimination under Directive 2000/78.

It is, indeed, true that multiple discrimination presents challenges that are not faced when an assessment of a single ground of discrimination is made.

One such difficulty is that a multiple discrimination assessment contradicts the classic single-ground model of discrimination law analysis which requires the identification of a single hypothetical comparator who must only have a single characteristic – the one that it is claimed that the discrimination complained of is based – that is different from the person that is treated worse. Yet, at the same time, an analysis which is – artificially – pushed to fit this model by insisting on an examination of the difference in treatment by using a single ground, may be incapable of accurately reflecting the situation that pertains in a certain case. For instance, on the facts in Parris, it would be inaccurate to compare all LGB persons with all heterosexual persons; or all LGB persons who have entered into a marriage or registered partnership with all heterosexual persons who have done so too; or all persons born before 1951 with all persons born after 1951. It was only LGB persons who were born before 1951 that were treated worse than everyone else (i.e. LGB persons born after 1951 and heterosexual persons in general). Accordingly, the difference in treatment complained of was based on the combined grounds of sexual orientation and age and, thus, the failure of the Court to find this misrepresents the reality of discrimination that was suffered on the facts of the case and – at a broader level – contributes to the continued invisibility of the phenomenon of multiple discrimination.

Another difficulty with multiple discrimination and its prohibition under EU law is that there is a hierarchy in the protection from discrimination on various grounds (with race and ethnic origin coming at the top, followed by sex, and then by the Directive 2000/78 grounds) which means that it is difficult to apply a single analysis in a situation where discrimination is suffered on more than one ground simultaneously. This, in fact, is the reason why legal advisors handling cases involving multiple discrimination usually make a strategic decision as to which single ground to choose, taking into account the protection afforded in relation to that ground as well as what is possible and attainable on the facts of the case.

Yet, and despite the above difficulties, it is important that where there is multiple discrimination, that this is reflected in the Court’s analysis. This is because, as noted by the Advocate General in Parris, a finding of multiple discrimination requires that a different approach to justifications is taken, as ‘[t]he combination of two or more of the grounds for a difference of treatment referred to in Article 1 of Directive 2000/78 may also mean that, in the context of the reconciliation of conflicting interests for the purposes of the proportionality test, the interests of the disadvantaged employees carry greater weight, which increases the likelihood of undue prejudice to the persons concerned, thus infringing the requirements of proportionality sensu stricto’.

Despite the fact that the EU legislature and the ECJ seem to ignore the reality of multiple discrimination, there have already been calls by the EU institutions to take this form of discrimination more seriously and to take action in order to increase both the capacity to recognise and identify occurrences of multiple discrimination and awareness of the need to combat them as such. It is important for the institutions, bodies and courts, that apply anti-discrimination law to become aware and able to identify the unique ways in which individuals experience multiple discrimination (see, for instance, Report commissioned by the European Commission ‘Tackling Multiple Discrimination: Practices, policies and laws’ (2007) available at http://ec.europa.eu/social/main.jsp?catId=738&pubId=51). Accordingly, the ECJ should take the opportunity – when it arises again – to rule that EU anti-discrimination law prohibits not merely single-ground but also multiple discrimination and to provide guidance as to how to deal with cases of such discrimination. After all – as stressed by the Advocate General in Parris – ‘it is apparent at several points in the Directive [i.e. Directive 2000/78] that its authors were acutely aware of this issue [i.e. multiple discrimination] and assumed that it could be adequately resolved by recourse to the instruments provided by the Directive’.

Further Reading

N. Bamforth, M. Malik and C. O’Cinneide, Discrimination Law: Theory and Context (Sweet & Maxwell, 2008), Chapter 9

K. Lenaerts, ‘Federalism and the Rule of Law: Perspectives from the European Court of Justice’ (2011) 33 Fordham International Law Journal 1338

A. Tryfonidou, ‘The Federal Implications of the Transformation of the Market Freedoms into Sources of Rights for the Union Citizen’ in D. Kochenov (ed.), Citizenship and Federalism in Europe (Cambridge, CUP, 2016, forthcoming)

A. Tryfonidou, ‘Discrimination on the Grounds of Sexual Orientation and Gender Identity’ in S. Vogenauer and S. Weatherill (eds), General Principles of Law: European and Comparative Perspectives (Oxford, Hart, 2017, forthcoming)

Report: ‘Tackling Multiple Discrimination: Practices, policies and laws’ (2007) available at http://ec.europa.eu/social/main.jsp?catId=738&pubId=51

Report by S. Fredman, ‘Intersectional discrimination in EU gender equality and non-discrimination law’ (May 2016), available at http://ohrh.law.ox.ac.uk/new-report-intersectional-discrimination-in-eu-gender-equality-and-non-discrimination-by-professor-fredman/

Barnard & Peers: chapter 20

Photo credit: cbc.ca

---

[1] J. H. H. Weiler, ‘Fundamental Rights and Fundamental Boundaries: On the Conflict of Standards and Values in the Protection of Human Rights in the European Legal Space’, in J. H. H. Weiler, The Constitution of Europe: ‘Do the New Clothes Have an Emperor?’ And Other Essays on European Integration (Cambridge: Cambridge University Press, 2005), pp. 103–104.