Non-discrimination: Transformative contract law?
From a broad social perspective, non-discrimination law is arguably the area of private law with the most sweeping transformative ambitions. This is reflected in the introduction to the EU’s Framework directive (Directive 2000/78), with its declared aim to
lay down a general framework for combating discrimination on the grounds of religion or belief, disability, age or sexual orientation as regards employment and occupation, with a view to putting into effect in the Member States the principle of equal treatment.
While labour lawyers have been sometimes sceptical about the promises of anti-discrimination rules, others – in particular, contract lawyers – have highlighted the considerable challenge non-discrimination brings to freedom of contract and, in particular, the freedom to choose a contractual partner. In recent years, the framework of non-discrimination has extended well beyond employment relationships, casting contract law disputes as a ground for balancing and establishing rules of coexistence in diverse societies.
The struggle to accommodate such diversity is evident in recent disputes before the European Court of Justice (‘CJEU’), which have brought to the fore the ambiguities and tensions hidden in a “framework” approach to non-discrimination in the Union and of a human-rights based approach to the problem. Is non-discrimination, ultimately, a matter of freedom of religion/thought, or is it a matter of inclusion and equal treatment for underrepresented and quite possibly oppressed groups? The consequences of the failure to confront this question are visible in two cases currently pending before the CJEU, which I will discuss in the rest of this blog post.
In these cases, WABE and Müller Weinhandel, two employers have been challenged by former employees who had been dismissed due to their insistence on wearing a headscarf while performing their work activities – as a day care teacher and shop attendant respectively. WABE, the first employer, had put in place a written neutrality policy which prohibited all displays of religious, political and philosophical affiliation, either via clothing or via communication to the children, parents, or third parties; Müller Weinhandel, on the other hand, declared the existence of a more informally established policy prohibiting the display of “large scale” signs only. In other words, wearing a small cross or other small religious symbol would not be considered to fall under the prohibition. Both cases hail from Germany, where the applicants thought national law should make the employer’s instructions invalid and their dismissal lacking justification.
Prequel: Achbita and its siblings
The German courts seized with the cases thought the answer depended on the implications of the CJEU’s 2017 landmark headscarf cases, Achbita and Bougnaoui. The cases required the CJEU to clarify the application of both direct and indirect discrimination rules to (the impossibility of wearing) headscarves. Direct discrimination directly attaches consequences to one’s religion, ethnicity, sexual preferences, or political and philosophical orientation; it is prohibited in general, with the exception of cases in which it can be justified in light of the specific nature of the work to be carried out. Indirect discrimination, on the other hand, occurs when apparently general rules have a disproportionate effect on certain groups. It is also prohibited, but can be allowed when the measures adopted pursue a legitimate goal and are appropriate and necessary. In Bougnaoui, the CJEU found that it was not open to an employer to directly discriminate against an employee wishing to wear an Islamic headscarf on account of client preferences.
In Achbita, however, the CJEU established that a general policy prohibiting the display of all signs of political, religious, philosophical or political affiliation, aimed at enforcing a company’s image of neutrality, can be allowed. Such prohibition represents a form of indirect discrimination – in so far as it disproportionally affects employees belonging to certain (in particular: religious) groups; however, according to the CJEU, such discrimination can be justified in view of the company’s right to conduct business (article 16 EU Charter of Fundamental Rights), which protects a company’s desire to project an image of neutrality.
The decision in Achbita has perhaps unsurprisingly been met with criticism from a variety of corners; what is more interesting here, however, is its legacy. Achbita, in fact, made it possible for employers who wished to prohibit the wearing of an Islamic headscarf to do so in a relatively clean way – that is, by adopting a neutrality policy either via its works council or, possibly, by unilateral regulation. This was not entirely without strings attached – the CJEU, also relying on previous case law by the European Court of Human Rights, expressly required the policy to be implemented consistently. Besides, some were quick to observe that Achbita concerned a Belgian case, and that Belgium, much like France, had a specific view on secularity and neutrality which may have affected the context of the case. Employers in more traditionally multicultural (former) Member States should not rush to adopt neutrality policies on the assumption that Achbita would cover them.
It seems that employers were in fact quite keen to try them out, if we are to judge on the basis of the two German cases currently before the CJEU. Given the publicity afforded to AG Kokott’s opinion in Achbita, it may not be a coincidence, in fact, that Müller Weinhandel announced a neutrality policy one month after the AG’s decision was published back in 2016. The second employer involved in the cases, WABE, published its policy in 2018, that is, a few months after Achbita. In the case of Müller Weinhandel, the limitation of the prohibition to prominent, large-scale, signs raised new questions. According to the employer, neutrality in this context served “inter alia” the purpose of avoiding conflicts among employees.
The CJEU now has to, in essence, decide on four questions: (i) whether such policies are really correctly framed as direct, rather than indirect, discrimination; (ii) whether the employee’s religious freedom should not be weighted differently (than in Achbita) in the context of the indirect discrimination assessment; (iii) whether German constitutional law can offer a higher degree of protection to the concerned employees than the directive (which is a minimum harmonization measure); (iv) whether a prohibition on large-scale signs only is also covered under the Achbita assessment. We will here discard (iii), which would offer a way out in the concrete cases, but would not redress the problems created by Achbita, and instead concentrate on the other points starting from the latter: what is the problem with prohibiting only large-scale displays?
“Large scale” signs and direct vs indirect discrimination
The prohibition of “large-scale” signs under scrutiny in Müller Weinhandel shows the slippery slope intrinsic in Achbita: namely, neutrality policies are introduced merely as headscarf bans in disguise. Such a ban, outside of a general neutrality policy, would qualify as direct discrimination, and would need to clear a much higher justificatory threshold. Let us, however, be clear: such neutrality policies are not the kind of measure the exceptions to indirect discrimination aim to allow. What are those, then? Imagine a company that requires all employees to work on Friday nights or on Sundays. Such a company schedule may place an extra burden on employees from certain religious groups, or even prevent some of them from working for the company, and thus would constitute indirect discrimination. However, if the company’s business model requires (or profits from) such a schedule, the negative effect needs to be accepted. This is very different from a policy that directly aims to compress employees’ fundamental rights and is somewhat legitimized by the fact that, in doing so, it negatively affects several different groups at the same time. To the extent that, as the AG involved in these cases (unlike, one must say, much scholarship) believes, the CJEU was right to consider such policies under the indirect discrimination rules [para 55 opinion], they must be subject to particularly intense scrutiny.
In light of this, contrary to what the AG concludes, a policy which only prohibits large-scale displays is more, not less, suspicious. It may be the case that it is less restrictive than a prohibition covering all signs, but what kind of neutrality can it project? By effectively treating employees differently on the basis of the type of religious items they sport, the policy seems, in particular, unlikely to prevent tensions and conflicts among employees – the very aim stated by Müller Weinhandel, which the AG somehow forgets to go back to in his opinion. Furthermore, the less general the policy is, the closer it comes to direct – rather than indirect – discrimination. The AG’s opinion that a more targeted policy, restricting itself to objects that are more likely to offend customers [para 74] is ultimately a more innocuous policy, is at best naïf and, in any case, ill-conceived. [I am only indirectly engaging with the opinion, which has been extensively and excellently criticized elsewhere]
Taking proportionality seriously: what goes into the balance?
Now that we have established the somewhat ambiguous position of neutrality policy in the non-discrimination panorama, it becomes clear that saying that any infringement of employees’ fundamental rights they may entail is justified when the company declares the intention to project a neutral image amounts to insufficient scrutiny. In fact, the rule of article 2.2.b(i) is an exception to the prohibition of indirect discrimination and as such must be constructed narrowly. This applies throughout the three classical prongs of the proportionality test, and, in particular, to the balancing between the rights and interests at stake. It has been argued elsewhere that this balancing has been misconstrued by the CJEU in Achbita, by framing the wearing of a headscarf as right to manifest one’s religion rather than to practice it; however, I want to propose here a different reading.
AG Rantos, in his opinion, says that not freedom of religion, but rather the prohibition of discrimination [see para 97] is the founding principle of the Directive and that freedom of religion as such should not play a preponderant role in interpreting it: I want to take this claim seriously and argue that on this ground the cases show that Achbita is untenable precedent and should be revised. In fact, the cases show how Achbita, and the neutrality policies it endorses, represent an exclusionary factor for Muslim women trying to access paid work – access which, research shows, is already complicated for hijab-wearing women in western countries. At least one more similar case (from Belgium) is currently pending before the CJEU. If the self-standing goal of the anti-discrimination framework is to provide underrepresented groups access to employment on equal conditions as members of the majority, a mere desire to project a neutral image, which is based on a subjective declaration by the employer that they wish to pursue such an aim, cannot be enough.
How to fix Achbita and save anti-discrimination law
The good news is that Achbita is very easy to fix: it is enough to recall that the CJEU’s interpretation of the Directive must be read in the context of the interpreted rule being an exception and go back to the missing proportionality test:
- at the very least, the outcome of this test could be left to the Member States’ courts, tasked to assess whether, in a given case, the company’s desire should be given such weight to overcome the employee’s right to equal treatment, access to employment and, possibly, freedom to practise their religion;
- in a slightly better scenario, the CJEU would indicate to courts that this assessment needs to be carried out very carefully, requiring the employer to show why such neutrality is required in their business;
- in any case, the CJEU should at least acknowledge that in abstracto neutrality may also be achieved (in considerably less invasive fashion) by pluralism – that is, by allowing the displaying of various signs of political, religious or philosophical affiliation up to a point which does not undermine overall professionalism;
- to finish off, in no event should the CJEU accept AG Rantos’ submission that prohibiting large-scale signs only is a lesser evil in this case: the German court’s submission that such prohibition may fail the adequacy test offers the easiest way out, and all the CJEU need do is confront the employer’s own argument that the aim is not to appease the customer, but to avoid workplace conflicts – an aim which can hardly be pursued by treating similar situations differently.
While EU non-discrimination may still be a long way from delivering on its transformative promises, the CJEU has a chance to prevent it from instead performing a nefarious transformation – effectively facilitating the exclusion of Muslim women and other minorities from the labour market. Let this chance not be wasted: Achbita must fall!
(Photo: Nicola Fioravanti)