Achbita v G4S: Religious Equality Squeezed between Profit and Prejudice

By Gareth Davies

And below: Bougnaoui v Micropole: Mildly Surreal Thoughts on Competence and Clothes (particularly when worn by women)

The two cases were decided on the same day by the Grand Chamber of the Court of Justice. Although they both concern essentially the same fact set – a firm wishing to dismiss an employee who insists on wearing an Islamic headscarf – the questions referred were different, and the substantive discussion is found in Achbita. Bougnaoui, briefly noted at the end of this blog, addresses just one, odd, point: the Court confirmed that the fact that a Muslim woman wears a headscarf does not make her incapable of doing her work. That is little comfort though – since Achbita decided that she can probably be dismissed anyway.


In Achbita v G4S the Court of Justice was asked whether a private firm could prohibit the wearing of Islamic headscarves by employees who dealt with customers, or whether this violated the ban on religious discrimination in the workplace, found in Directive 2000/78. The claimant, Ms Achbita, worked as a receptionist for G4S in Belgium. When she began wearing a headscarf she was warned that it was against company policy, which disallowed all religious, political or philosophical signs in the workplace. When she continued, she was dismissed.

The Court found that under the right circumstances a company might be entitled to have a policy of this sort. One condition was that the policy must be in writing – in the interests of certainty and clarity. Another condition was that it must apply without distinction to all beliefs.

This latter condition may raise a few questions, most obviously whether it is necessary to include philosophical and political beliefs. Could a policy legitimately cover just religion? Given that there is no ban on discrimination on political grounds, it seems plausible that a narrowly religious policy is possible. Still, most organisations will choose for a wider rule, covering any belief that might upset anyone, but leading to possible definitional issues: what is a manifestation of a particular philosophy? Are hipster beards prohibited at G4S? Pinstriped suits? Could a skirt below the knee, in some contexts, be seen as ‘Christian’?

These are nevertheless relatively marginal issues. The substantive question in the case is whether a rule of this sort is in fact discriminatory. That involves two sub-questions: does it disadvantage those of some beliefs more than others, and if so, is it justified? The Court’s answer to both of these was remarkable for its complete misunderstanding of the law and of logic, and its total lack of respect for the idea of religious equality, which it subordinates both to the economic interests of employers and the prejudices of customers.

Do neutrality rules affect everyone equally?

Regarding the impact of the rule, the Court said that it was for the national court to decide if it affected some beliefs more than others, but it was not inconceivable that they might conclude that it did (paragraph 34).

In an essay this might be enjoyed as dry humour. In a judgment, it is bizarre and irresponsible. Obviously the rule affects some beliefs more than others, because only some beliefs are considered by their adherants to require the wearing of certain clothes. Strict Jews, Sikhs, and Muslims are hit, Christians are not. It is of course coincidental that these rules are being considered on an overwhelmingly Christian continent. If the Court wants to have one of its periodic fits of deference and refer the issue of disparate impact to the national court it is entitled to do so, but it should preferably give clear guidance, or none, but not be actively misleading. It is, contrary to the implication in the judgment, inconceivable that the rule does not affect some beliefs more than others.

Nevertheless, within these few muddled words some interesting social issues are to be found. What exactly is the difference between the impact of a neutral-clothing rule on Muslim women and others? Many people might feel the desire to wear a badge making a political statement, or hang a cross around their neck, or in some other way announce their world view. If a Muslim woman wears a headscarf out of choice, rather than out of a sense of religious compulsion – perhaps in order to show that she is a Muslim or because it makes her feel comfortable – then her position is not so different, and there is some argument that she does not experience a different impact. She, like the Christian who must take off their cross, is prevented from making a statement, but neither of them are forced to violate the requirements of their belief by doing so.

On the other hand, if a Muslim woman believes that Islam requires her to wear a headscarf, then there is a fundamental difference between the way the rule affects her and others: she is not simply prevented from making an announcement, or exercising a preference, like the communist or Catholic, but actually prevented from following her religion. Since at least some Muslim women take this view of the headscarf, that is why there is obviously a disparate impact. There is a clear difference from not being allowed to announce your belief, and not being allowed to follow it.

What makes the situation complex however is that wearing a headscarf is not always seen in such black-and-white terms. Some Muslim women emphasise that it is a choice, that perhaps at a certain point in their lives they feel ready for it: they do not see it as an imposition by their religion, but a voluntary way of showing their adherence. Indeed, there is a feminist view of the headscarf which presents it as potentially empowering: it is a woman’s expression of her right to choose which parts of her body are seen, that choice being personal, and cultural, but precisely not dictated by wider society or by men. If women, and men, are allowed to choose when to reveal their bodies, then they should be allowed to choose when they reveal their hair, and active personal choices of this type are to be celebrated, whereas attempts by institutions to impose standardised norms are repressive – particularly when those norms are historically based, and so inevitably patriarchal in their origins.

This reframing of the headscarf has legal consequences however. If the headscarf is presented as an empowering assertion of identity, and if this is put forward as a reason to defend women’s choice to wear it, then that very vision undermines its legal protection. For the law protects religion, not culture or choice – it protects what we feel obliged to do, not what we are used to or what we want. The more the headscarf becomes personal, cultural, political, and assertive, the harder it is to claim that banning it violates religious beliefs, rather than simply violating personal preference.

What is ‘neutrality’ and does it deserve legal protection?

The law says that if a rule affects some beliefs more than others, it must be justified. Despite its apparent doubts about the first part of this, the Court did go on to consider justification. Here it found that the desire of a company to present a neutral face to customers must be seen as a legitimate goal (paragraph 37). Now, in a more academic context this simplistic idea of neutrality would be hard to take seriously – if we see a customer-service company in Paris or Brussels which does not employ any women wearing headscarves then we probably know something about that company, and neutrality is not a very good term to describe what it is that we know – but perhaps it is unrealistic, and even undesirable, to expect courts to engage with ideas on this level of sophistication. Nevertheless, there is a certain tension in finding first that a rule affects Muslims more than others, and then that the rule is an expression of neutrality regarding belief.

The greater difficulty, however, is in unpacking what neutrality is supposed to mean. Is it the desire of the company to make its customers believe that it, the company, has no views on political or religious questions? This would be hard to fit with donations to political causes, or contentious statements by company chairmen. It is also rather hard to believe that anyone thinks that the beliefs of an employee, as expressed in their headscarf or beard or badge, reflect the official position of the company.

The desire to present a neutral face to customers may instead mean the desire to reassure customers that they will not be treated worse because of their own views. The classic critique of judges in headscarves is that non-Muslims will think they will not be judged fairly, and perhaps some customers, if they become aware of the personal views of the person serving them, will become nervous about favouritism.

Writing from an academic environment in the Netherlands this seems rather surreal. Most religious people do not think that their belief requires them to do illicit favours for their co-religionists, and indeed many may feel a religious obligation to be fair and honest. In many situations in Europe it would require a high degree of paranoia to worry about whether the engineer, nurse, waitress or postal worker will do her job properly because she is Muslim and the customer is not. However, in some places and contexts it might seem more plausible to think that whether your family is local or not, or perhaps whether you are known to be right or left politically, might affect whether you get a job, how the shopkeepers and the people in the market treat you, and what sort of treatment you get in the town hall. From this clientilist perspective, suspicion of a culturally alien service-provider would make more sense.

Yet if the customer assumes that those of other beliefs cannot be trusted, is this really a reason to hide those beliefs? For one thing, there is a kind of stupidity to it: merely because people cannot display their beliefs, does not mean that they do not have them. Removing signs does nothing to prevent religious favouritism if it exists. One might plausibly argue that if belief is a threat to fairness, then it is better to know who believes, rather than have it hidden. The sheer lack of any objective basis is ultimately fatal to this view of neutrality. One cannot justify a policy which addresses a threat which has not been shown to exist, and does nothing to remove that threat anyway.

Rather it seems that neutrality is really much simpler. It is not based on any genuine or objective threat to the interests of the customer, but simply on the fact that some customers may not like certain beliefs and if employees visibly have those beliefs the customers may shop elsewhere. The Court refers to the right to pursue a business, and the goal of neutrality is to protect the economic interest of businesses in not offending or disturbing their customers.

The problem is that when customers dislike the idea of being served by a Muslim or other religious person, without having any concrete reason for this, then the appropriate word is prejudice. A policy of neutrality is a policy of respecting the prejudices of customers in the interest of the business.

That is understandable. We can feel sorry for the shopkeeper in a backward village who knows that if he employs the woman in a headscarf he will lose half his customers – or perhaps we should say, if we believe G4S, the security company with backward clients. It is a difficult situation.

However, as a matter of law, it is far from obvious that prejudice can serve as a justification for otherwise discriminatory rules, even indirectly. If G4S had customers who did not like black people, would it be justified in employing only white receptionists? The answer is negative, and if there was any doubt we have Firma Feryn to confirm it. Legally the situation is different here, because that would be direct discrimination, which cannot be justified, although the difference is more a matter of legal artefact than deep logic. However, what about a firm which refused to employ telephone operators who had any trace of an ethnic minority accent, because customers did not like it. Would that customer racism be a good justification?

Presumably the Court thinks it would. However, if discrimination on grounds of X is prohibited, and a rule tends to exclude people with characteristic X, but the rule is justified by the fact that some people involved do not like X, this rather hollows out the very prohibition on discrimination. Is it not incompatible with the idea of religious equality that an exclusionary rule be justified by a dislike of religion? Perhaps if that dislike was truly of all religions equally, the justification might make sense: yes, it is a prejudice, but not a discriminatory one, and it is discrimination law that applies here. However, as a matter of fact, it is perfectly well known to everyone, including the judges on the Court, that the rule in this case reflected a specific discomfort with Muslims. The Court’s view is analogous to finding that a dislike of homosexuality could justify indirect discrimination against gay people, or a dislike of old people could justify indirect discrimination on grounds of age. That seems little more than a principled rejection of the very premise and purpose of the law.

That’s all very well, the business person may cry, but I cannot help the prejudices of my customers. Am I to go bankrupt in the name of your equality then? A narrowly legal response is that where one right – the right to pursue a business – conflicts with another right – religious equality – these have to be balanced, just as is the case with privacy and free speech. The most plausible standpoint is that this balancing has been done by the legislature in adopting the directive and indeed the right to pursue a business has been subjected to basic constraining norms of equality, just as it is subjected to burdens of taxation, social costs, and so on. An alternative standpoint is that the two rights must be balanced in the individual situation – perhaps taking account of the demonstrable impact on the business of allowing Muslim women to be employed.

What seems to be completely unprincipled and indefensible is to treat the right to pursue a business as evidently taking priority. Yet that is what the Court does in saying that ‘neutrality’ is a justification for restricting religious equality. It says that the economic interest in pandering to customer prejudice demarcates the limits to religious equality. So much for human rights.

Deference, or harmonisation?

Athough the judgment emphasises the role of the national court in ultimately deciding the case, the discretion left to them is distinctly asymmetrical. The Court ruled explicitly that the goal of ‘neutrality’ towards customers ‘must be considered legitimate’ (paragraph 37) and that prohibiting signs of belief ‘must be considered strictly necessary’ for achieving this (paragraph 42).

That begs the question: what about Member States who see things differently? For there may well be Member States where the legislature or the courts take the view that pandering to prejudice is not a legitimate goal: that customer prejudice does not provide a legitimate reason for indirectly discriminatory rules. They might well not allow rules such as those in Achbita. Do they still have this freedom?

Apparently not: Achbita seems to elevate the interest of a business in neutrality towards its customers to the status of a right, and to one that takes precedence over religious equality. It does not just allow Member States to permit employers to exclude religious clothing, but it appears to have denied them the option not to. This is not a judgment that accepts national diversity on equality – rather, it imposes a harmonised uber-secularist view which Spaventa notes that even the French and Belgian governments did not support.

It is one thing to defer to prejudice, and another thing to impose it. Will any states rebel against this aspect of Achbita, or at least read the judgment differently? They might give more thought to what neutrality means and challenge the view that it entails banning religious clothes, or they might question whether respect for prejudice can justify discrimination. It is to be hoped that some do, and a challenge to the judgment could be justified, either using the argument that the Court exceeds its jurisdiction in ruling so definitively on factual questions, or using constitutional values. Is there perhaps a state whose constitutional identity puts equality above profit?

The social consequences

The consequences of the judgment may be more nuanced. It will certainly encourage dress codes in the short term, and some Muslim women will be forced out of their work, as was the case with Ms Achbita. However, some of these may go to more tolerant companies, and if there are some customers who do not want Muslims, there may be others, perhaps even more, who do not want companies who reject them. The real breakthroughs in tolerance and integration do not come in supreme courts, but in public awareness, and by refusing to protect rights in law the Court hands the buck back to the public, and in some ways promotes social progress. Still, it will be a bumpy road: in leaving religious equality to the market, the Court abandons it to inevitable ups and downs and local variations.

In the meantime, some Muslim women may choose to take off their headscarves rather than lose their job. There will be non-Muslim observers who embrace this as a sign that the law is working to promote integration and empancipation, yet surely they should struggle to square this perception with the sense of violation and oppression that those women will feel. Emancipation by violence is hard to defend – in fact the entire judgment is a rejection of that idea when applied to states.

Which is not to say that it never works, for anyone: some of the Muslim women who decide to uncover may feel secretly relieved. Some may have an excuse to escape social or family pressures, while others may be confirming to themselves that they wore it by choice, not obligation, and that confirmation will be liberating. The relationship with the headscarf is by no means monolithic.

For those who wear it because they feel it protects them from the male gaze, there is also the paradox that in forcing them to remove it the employer implicitly accepts a greater responsibility to protect them. If a company claims that such modesty is unnecessary then it surely accepts the burden of ensuring that no male employee, or customer, makes a woman feel uncomfortable. The law may not enforce this – or perhaps a little via harassment and equality law – but the institutional reality is that if women feel vulnerable, employers will have to address this, and economic reality will ultimately make them want to. Public support for women without headscarves is greater than for those who wear them: by making a woman remove her headscarf an employer violates her rights, but empowers her socially. There is nothing to be proud of in this trade-off, but it is real.


The Court probably did not want this case. In its repeated insistence that the national court must decide, and in its remarkable openness to all possible – and impossible – views of the facts, it was probably saying ‘as long as you comply formally with the rules, do whatever you like: just don’t send us any more like this’. Given the strength of feeling in some states, realpolitik may provide some justification for that approach. The law does not.

Nevertheless, the judgment has a certain cathartic quality, and it remains to be seen what its real effects are. As an intervention in society and politics it is, as suggested above, violent, but complex and intriguing. As a contribution to legal thinking it is incoherent and unprincipled, but nevertheless perhaps worthwhile: for sometimes it takes an explicit statement of the orthodoxy to reveal quite how hollow it is.


Bougnaoui: Mildly Surreal Thoughts on Competence and Clothes (particularly when worn by women)


Bougnaoui was decided on the same day as Achbita, also by the Grand Chamber. It was referred by the French Cour de cassation, and concerned analytically similar facts to Achbita: a Muslim woman working as a design engineer in Paris began wearing a headscarf, and some customers to whom she was sent on assignment complained: apparently it upset their employees. Her employer, Micropole, asked her to stop wearing it and she refused, and was dismissed.

However, the question referred in Bougnaoui was much narrower than in Achbita. The French court asked simply whether a customer’s desire to have IT services provided by someone who did not wear a headscarf could mean that not wearing it was a genuine and determining occupational requirement.

At first glance, the idea that absence of religious clothing could be a genuine occupational requirement for an engineer is rather surreal, and so it is not surprising that the Court said that it could not. For those who doubt whether it is possible to program computers or perform other technical work while wearing particular clothes, Bougnauoi stands as an authority that it is.

However, the doubt was of course not connected to the effects of headwear on technical abilities. Really, the question was whether the ability to satisfy the cultural expectations and prejudices of customers should be seen as a genuine occupational requirement – as part of the job. That question is rather more slippery. After all, many, even most, businesses expect their staff to wear suits, look businesslike, and generally conform to non-functional stereotypes, not because there is any demonstrable link between clothes and the skills of a banker, lawyer, or consultant (or judge), but because ‘customers expect it’. The ability and willingness to pander to stereotypes is in fact widely seen as part of the ability to do a job. It is then but a small step to the question in Bougnaoui. The case holds up a mirror to the society that we have created and accept.

Nevertheless, the Court found in favour of Ms Bougnaoui. EU law does not cover discrimination on grounds of style – except perhaps where there is a sexist element to it, as in a requirement that women wear make-up – but if it did, the case would provide conceptual support to those who would like to compel their employers to drop dress codes, and allow them to wear jeans to work. As has often occurred in the history of discrimination law, one group’s fight for equality may end up helping others too.