Strasbourg Observers

European Court of Justice keeps the door to religious discrimination in the private workplace opened. The European Court of Human Rights could close it.

March 27, 2017

By Saïla Ouald-Chaib and Valeska David

On 14 March 2017, the European Court of Justice issued two judgments, in the cases of Achbita and Bougnaoui concerning the manifestation of beliefs in the private workplace. From the perspective of inclusion and human rights law, the judgments are very disappointing. They basically legitimize and even provide a recipe for discrimination of employees on the basis of their religious or other convictions.

The two cases were long-awaited. Last year we published a series on this blog discussing the divergent opinions of the advocate-generals Kokott and Sharpston in both cases. The outcome of the cases was, because of these conflicting opinions, very difficult to predict. The general tendency, however, was to hope that the ECJ would follow Sharpston’s view instead of Kokott’s. In the end, the ECJ followed both, probably in an attempt to find a compromise. In this post we will shed light on how this attempt ultimately compromised the ECJ’s legal reasoning and the effectiveness of the Employment Equality Directive for Muslim women. Following an introduction to the facts of the cases, we will briefly explain why the ECJ’s judgments on neutrality policies in private companies are problematic from several perspectives. Firstly, we will discuss the approach towards direct discrimination, especially in the case of Achbita. Secondly, we will reflect on the legitimate aim invoked.  Thirdly, we will critically look at the proportionality analysis. Finally we will explore the role the European Court of Human Rights could play in the future when it concerns the right to manifest one’s religion in the private workplace. For this, we will look at the ECJ’s judgments from the perspective of the Strasbourg Court’s judgment in Eweida v. The United Kingdom.

Facts and outcomes

Both cases concern Muslim women employees who work(ed) in private companies. In the first case, Ms. Achbita worked already for three years for the Belgian branch of G4S before she decided to wear an Islamic headscarf, as a consequence of which she was dismissed. G4S argued that this was not in accordance with their company neutrality policy, which they introduced only after Ms. Achbita made clear that she wanted to wear a headscarf. In the second case, Ms. Bougnaoui was already wearing a headscarf at the time she was hired as an engineer by the French company “Micropole”.  After complaints of customers, however, she was asked to remove her headscarf which she refused and as a consequence faced dismissal as well.

In the case of Ms. Bougnaoui, the European Court of Justice finds a direct discrimination on the ground of religion, since the employer’s decision was not based on a general neutrality policy of the company. The customers’ wish not “to have the services of that employer provided by a worker wearing an Islamic headscarfcould not be considered as a ‘genuine and determining occupational requirement’. (Bougnaoui, §41) In the case of Ms. Achbita, however, the Court finds no discrimination. It considers it legitimate to dismiss an employee when a company has a policy of political, religious and philosophical neutrality in relation with both public and private sector customers.


Equal discrimination is still discrimination

In Achbita v. Belgium, the ECJ doesn’t find that direct discrimination is at stake since:

“In the present case, the internal rule at issue in the main proceedings refers to the wearing of visible signs of political, philosophical or religious beliefs and therefore covers any manifestation of such beliefs without distinction. The rule must, therefore, 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.” (Achbita §  30)

At first sight this reasoning could be considered reasonable. However, it starts from a wrong comparison. It compares the different manifestations of religious, political or philosophical beliefs targeted by the policy with each other, while the true comparator should be the ‘other’ employee who doesn’t manifest any belief. Comparing the treatment of different beliefs with each other would be similar to comparing the treatment of employees with different disabilities. It is not because an employer would discriminate all employees with a different disability in the same way that such a policy would be less discriminatory. The same counts for manifestations of beliefs. In the context of religious beliefs, the comparator should therefore be other employees who are secular or who are non-visibly religious. It is undeniable that with these neutrality policies, employees who visibly manifest their beliefs face a less favourable treatment than employees who don’t.

Discrimination in the name of the freedom to conduct business?

The ECJ starts in Achbita from the observation that “the desire to display, in relations with both public and private sector customers, a policy of political, philosophical or religious neutrality must be considered legitimate.”(§ 37) It continues that

 “[a]n employer’s wish to project an image of neutrality towards customers relates to the freedom to conduct a business that is recognised in Article 16 of the Charter and is, in principle, legitimate, notably where the employer involves in its pursuit of that aim only those workers who are required to come into contact with the employer’s customers.”(§ 38)

This part of the reasoning has been widely criticised. Not only does the ECJ blindly accept neutrality policies in private companies to be legitimate, it also considers neutrality policies which prohibit employees to manifest their religion to be part of the right to conduct a business. The ECJ is more critical in the case of Bougnaoui, where it does not accept that customers’ wishes trump employees’ right to manifest their religion. In the case of Achbita, however, such a critical stance is missing. At the same time, the critical approach in the former case is made insignificant. Indeed, ultimately, even though the ECJ does not accept in Bougnaoui that individual customers’ wishes are used as a basis for dismissing employees, it does accept, in Achbita, that employers put a policy in place which doesn’t allow manifestations of beliefs because of possible non neutral perceptions this might engender with customers. In other words, through its judgment in Achbita, the rationale of not offending customers serves as a basis of discrimination on a more structural level, as long as it is included in a general neutrality policy. Now, even if the company’s neutrality policy would be accepted as a legitimate aim, a serious proportionality analysis should be undertaken. The ECJ however, fails to do so, as will be explained hereafter. In fact, the freedom to conduct a business dominates the Court’s reasoning in the case of Achbita, while the employee’s interests and her fundamental human right to manifest her religion are totally left out of the picture.

Covered discrimination is still discrimination

As the ECJ also recognizes, a difference in treatment can only be justified if it pursues a legitimate aim and if the means of achieving that aim are appropriate and necessary. The question at stake should therefore be whether dismissing an employee for wearing a headscarf at the workplace is an appropriate and necessary means for pursuing a neutral image of the company towards customers. The ECJ finds that it does. At least, as long it concerns employees with a customer contact job and as long as the policy is applied in a consistent manner. The ECJ continues its proportionality analysis by reflecting on the question whether or not “it would have been possible for G4S, faced with such a refusal, to offer her a post not involving any visual contact with these customers, instead of dismissing her”. (§ 43)

The ECJ seems to apply a less restrictive alternative reasoning here, where it considers whether or not a less restrictive alternative to the measure at stake is available. Concretely, ‘offering’ a back office job with no customer contact would be a reasonable alternative for the ECJ. This reasoning is shocking, to say the least. In fact, it shows a total disconnect of the Luxemburg judge from reality. A first practical question would be: how many workplaces offer jobs with no customer contact whatsoever? Maybe call centre jobs and jobs with night shifts? What about the increasing number of Muslim women and other employees manifesting their beliefs, who aspire to professions that require contact with the public such as lawyers, doctors, professors, social workers or engineers? In practice, this ‘compromise’ would in any case mean a serious curtailment of people’s right to access the job market. A second question would be, what does an “image of neutrality” of a company mean?  Towards whom is a company neutral when it decides to hide its visibly religious employees for the sake of their corporate image? Should the main concern of an employer not be that his/her employees have the capacities and skills to serve the customer in a professional and neutral way? Moreover, when neutrality is assessed form the perspective of customers it is often forgotten that customers are also diverse. What kind of message is given to customers with the same background as the employees who are discriminated because of these kind of policies? In an increasing diverse society, with an increasing number of people with high potential and from different visible and non-visible faiths this kind of reasoning is lacking awareness of reality.

Finally, the main problematic aspect of this reasoning is how the ECJ easily accepts the reasoning that employees who are treated differently would not be discriminated against, as long as you ‘give them’ the ‘opportunity’ to have a back office job. It encourages employers to hide diversity. To hide people who are visibly religious in the ‘closet’ of the company.[1] Kenji Yoshino calls this phenomenon “covering”.[2] Imagine the same rationale would be applied to other discrimination grounds. Imagine customers might feel uncomfortable in front of an employee who has a disability. Would it be generally accepted that a European Court says: just make sure you give these employees an alternative in a back office and we won’t consider it discrimination. Or imagine a company doesn’t allow employees who are pregnant to carry out a job with customer contact, how would the reaction be? And where are the limits to these kinds of legitimations of discrimination? Would an employer be allowed to require an employee with the name Mohammad to use another name in his contact with customers, since he might give the impression to belong to a faith some customers might feel uncomfortable with? (see also an earlier post) The covering of discrimination is still discrimination. Also when it is enshrined in a general company policy and also when it is applied in a consistent manner. The opposite message the ECJ gives in these cases is shockingly at odds with the principle of equality enshrined in the EU Charter of Fundamental Rights and in the European Convention on Human Rights.

Contrasting Luxembourg to Strasbourg: Eweida v. UK

Many of us are now left with the question: what role is there for the Strasbourg Court to play? What difference, if any, can the European Court of Human Rights make? The Court’s record in protecting Muslim women’s rights is, let’s admit it, far from satisfactory. Several posts on this blog have criticised the Court’s case law on veil bans. (see here, here and here) However, when it comes to the non-discriminatory exercise of freedom of religion by an employee working in the private sector, the legal picture offered by Strasbourg is different. When the Court dealt with this question in the case of Eweida and Others v. The United Kingdom, the bar was set high to justify such interferences. (The case was discussed here) Interestingly enough, the ECJ’s references Eweida in support of its ruling. But, as we explain here, the reference seems to be on the wrong side of the arguments.

To start with, a preliminary remark on neutrality is in order. As we just discussed, in Achbita the enforcement of a neutrality policy by a private undertaking does not unsettle the ECJ. Restricting religious freedom for the sake of “private neutrality” is taken for granted. Well, in the Strasbourg case law, the neutrality argument does not hold in any context.[3] The ECtHR’s acceptance of restrictions to religious manifestations in the name of neutrality has a narrow scope. Such restrictions pertain to public institutions, mostly in the educational field, which are subject to a principle of neutrality with constitutional rank. Therefore, in the context of the ECtHR, it is hard to imagine that a wish to project an image of neutrality by a private undertaking could attain similar levels of protection.

Now, let’s take a closer look at Eweida. In this case, a woman employed by a private company in a face-to-face customer service was suspended on account of wearing a cross. The Court found a violation of the applicant’s religious freedom. And in doing so, it left little room to private companies’ restrictions on the religious freedom of their staff for the sake of the company’s corporate image. The judgment provided at least three important grounds that run against the interpretation adopted in Achbita. These are: (1) the difference between a human right and a business interest; (2) the seriousness of the restriction for the applicant; and (3) the lack of evidence on the harm inflicted to the company. We shall briefly examine each of these points:

(1) Human right vs. business interest: While the Court in Eweida accepted as legitimate the private companies’ wish to project a certain corporative image, it made it clear that this does not stand on a same footing with the right to manifest one’s religion. For the ECJ, an employer’s wish to project an image of neutrality is covered by the freedom to conduct a business (Article 16 Charter). This reasoning doesn’t have a place at the ECtHR. Freedom of religion, like the prohibition of discrimination, is a human right protected by the ECHR (Article 9); the private companies’ interest to project an image is not. The Court thus concluded in Eweida that too much weight was accorded to this interest (§ 94).

(2) Proportionality requires weighting what is at stake for the applicant: Ascertaining this aspect in Eweida, the Strasbourg Court acknowledged “the value [of religious manifestation] to an individual who has made religion a central tenet of his or her life to be able to communicate that belief to others” (§ 94). Moreover, in its necessity test, the Court also considered the socio-economic harm faced by Ms. Eweida and by other applicant manifesting religion in a private workplace. That is, it weighed  the possibility of changing job and the seriousness of losing one’s job (§§ 83 and 109).

(3) Need for evidence of the alleged threat: Of course, balancing presupposes a careful examination of what is at stake at the side of the employer too. That is why the ECtHR in Eweida went on to examine the damage experienced by the company. But the Court didn’t find “any negative impact on British Airways’ brand or image.” There was “no evidence of any real encroachment on the interests of others” (§§ 94- 95).

Astonishingly, the ECJ frames its proportionality inquiry in the following terms: “it must be determined whether the prohibition is limited to what is strictly necessary.” That is “whether the prohibition […] covers only G4S workers who interact with customers” (§ 42) The ECJ simply jumps to the last questions one could possibly ask. What about the initial crucial questions of proportionality that the ECtHR asked in Eweida? Namely, what was at stake for the applicant and what weight does this have in the balancing? How did her religious manifestation threaten the company’s interests and what evidence was provided to that effect?

Finally, Eweida contains a few other noteworthy elements: First, in Eweida the applicant declined British airways’ offer to take a post that didn’t involve visual contact with customers. While this offer, according to the ECtHR, could have mitigated the interference, it did not make the restriction proportionate (§§ 94-95). Thus, “hiding the employee” is not a human rights solution. Moreover, in Eweida the Court rightly found a violation even though it had not been established that the company’s uniform policy had put Christians generally at disadvantage (§§ 14-16). We could therefore envision that the actual disadvantage that such policies impose on Muslim women who wear a headscarf would only add weight against those policies.

Strasbourg to the rescue?

The ECJ judgments leave a dangerous door open: it allows private companies to impose bans on the wearing of religious signs in the workplace, in particular for employees with contact jobs. In practice, this kind of regulations particularly affects Muslim women like the applicants in the present cases. In a context of rising Islamophobia where for Muslim women, restrictions are becoming the rule and freedom the exception  – a context which is completely disregarded in these judgments –  this new development in the European case law opens the door to even more exclusion. Nevertheless, this open door is not the only available one. Although these cases give employers a legitimation to discriminate against certain employees, they don’t instill in them an obligation to do so. In fact, employers still have the ultimate power to do the right thing and to choose for inclusion. Moreover, as we discussed in this post, the judgment in Eweida v. UK contains a promising reasoning for employees who are visibly religious. At the same time we know that cases involving Muslim women were until now not very successful in Strasbourg. We are therefore left today with the question whether the ECtHR would close the door the ECJ opened in case it would be confronted with the same question from an employee who wants to wear a headscarf in the workplace. We would like to believe that it would.

[1] See also, Brems, Eva, Analysis: European Court of Justice Allows Bans on Religious Dress in the Workplace at

[2] Yoshino, Kenji. Covering: The hidden assault on our civil rights (Random House Trade Paperbacks, 2007).

[3] On the limitations of the neutrality argument and of other grounds to restrict the manifestation of religion, see also, Third Party Intervention by Human Rights Centre of Ghent University in  Lachiri v. Belgium, Application No. 3413/09, Available at

Print Friendly, PDF & Email

Related posts

Leave a Reply

Your email address will not be published. Required fields are marked *