How do you solve a problem like manifestation of belief?

Anyone with the slightest interest in employment law would need to have been living under a rock over the last few years to have missed the series of cases that have come before the Employment Tribunal and appellate courts concerning disputes arising from employees’ manifestations of their protected religious and philosophical beliefs. The likes of Page[1], Forstater[2], Miller[3], and Higgs[4] have required the judges involved to earn their corn, presumably given them a few sleepless nights, and in the process exposed how ill-equipped the law in this country currently is to deal with these issues.

The Supreme Court’s recent decision to refuse permission to appeal in Higgs has cemented for the foreseeable future a status quo which, in this author’s opinion, represents a deeply problematic fudge. But there’s no question that manifestation of belief cases throw up unique problems to which there are no easy solutions. How, then, did we get here, and how might those problems best be resolved?

The unique difficulty of manifestation of belief

The reason that manifestation of belief has become such a problem for the courts is an inevitable result of the way that ‘religion or belief’ has come to be defined within s.13 Equality Act (‘EqA’) 2010, read with s.4. As noted by Underhill LJ in the Court of Appeal in Higgs (at §54), the decision of the Court of Justice of the European Communities (‘CJEU’) in Bougnaoui[5] authoritatively established that, in respect of legislation which implemented the Framework Directive (such as the EqA), protection against direct discrimination on the grounds of religion or belief must be interpreted as encompassing protection both for holding a protected belief and outwardly manifesting that belief.

While this aligns the interpretation of the protected characteristic of ‘religion or belief’ in the context of direct discrimination with the definition contained in Article 9(1) of the European Convention on Human Rights (‘ECHR’), it also sets it apart from the approach taken to direct discrimination in the context of all the other protected characteristics in s.4 EqA. By way of a contrasting example, things that arise in consequence of – or in other words are manifestations of – a person’s disability, although protected under s.15, are not treated as part of the protected characteristic of disability for the purposes of s.13.

That the expansive definition given to ‘religion or belief’ in the context of s.13 should lead to problems is inevitable. Since direct discrimination cannot be justified, this definition on its face means that any less favourable treatment found to have been because of an employee’s manifestation of a protected belief will be unlawful. However, as Underhill LJ pointed out in Higgs (at §92):

‘Direct discrimination in manifestation cases is (uniquely) different from discrimination on the ground of other protected characteristics (and indeed from simple belief discrimination) because it is based […] not on the possession of the characteristic as such but on overt conduct, which thus has the potential to impact on the interests of society and the rights and freedoms of others. That distinction may be said to put it in a special category which requires a more flexible approach.’

While manifestations of protected beliefs are plainly liable to cause offence, the real problem is that there is nothing preventing a manifestation of belief from amounting to harassment of another; the two are not in any sense mutually exclusive. If employers are not able to sanction such manifestations, their ability to protect their employees from harassment is seriously undermined.

Unless employers are to be prevented from taking any action at all against employees where they have manifested their protected beliefs in an inappropriate way, the courts therefore either have to find a way to allow for manifestation of belief discrimination to be objectively justified or put absolute faith in the ability of tribunals of fact to find a separable reason for the less favourable treatment in appropriate cases.

If the fundamental issue with the first of these options is that on the face of it direct discrimination cannot be justified, the difficulties with the second are slightly more nuanced. There will be times where purporting to find a reason separable from the manifestation seems very much like splitting hairs. If the manifestation is harassment of another by virtue of a particular cocktail of its content and the context in which it was delivered, it may well be artificial to pretend that the manifestation itself was not a significant factor in the less favourable treatment. But the employer surely must retain the ability to sanction the harasser.

In rarer cases, there will also be situations where the causal connection between the manifestation of belief and the less favourable treatment is unarguable, but where it must still be permissible for the employer to take the impugned step. If, for example, an employee attends work wearing an item of religious clothing that poses a health and safety risk, and an employer that does not have a dress code policy requires them to remove it, it is hard to see how this could either be analysed through the lens of indirect discrimination, or how the reason for the less treatment could be distinguished from the manifestation of belief (see Underhill LJ in Higgs at §85). But it surely must be lawful for the employer to require the employee to remove the item.

Unless and until Parliament sees fit to amend the EqA, the courts are therefore stuck between a rock and a hard place on the issue of manifestation of belief discrimination, either required to manufacture an objective justification defence or rely on what may be somewhat artificial findings of separability to resolve problem cases. That may make it understandable that the solution that the courts have settled on is a flawed one, but it does not mean that a better alternative is not possible.

Page – Underhill LJ sets out his stall

Notwithstanding the potential flaws in this approach highlighted above, until the Court of Appeal decision in Page, the way that the courts had generally sought to navigate tricky manifestation of belief cases was through relying on separability. They had therefore avoided the issue of whether less favourable treatment because of an employee’s manifestation of their belief could be justified.

However, in Page, Underhill LJ saw fit to innovate, setting down a rule of law that defined when it was open to a court to find separability in the context of manifestation of belief by reference to the justifiability of the employer’s conduct.

Page concerned a Christian non-executive director of an NHS trust who was dismissed for repeatedly making controversial statements about the morality of homosexuality and same-sex adoption on TV. His statements were found to have arisen from protected beliefs, but the Court of Appeal upheld the ET’s dismissal of his claim.

The claimant submitted that, in every case of religion or belief discrimination, a tribunal should start by deciding whether there had been a breach of the claimant’s rights under Article 9 and Article 10 ECHR and analyse the EqA claim through that lens. Although Underhill LJ stated that he did not consider this to be correct, he did hold (at §67) that protections under the ECHR and the EqA ‘must be intended to be coextensive’. He then commented (at §68) that, in the context of religion or belief discrimination:

‘…case law has recognised a distinction between (1) the case where the reason is the fact that the claimant holds and/or manifests the protected belief, and (2) the case where the reason is that the claimant had manifested that belief in some particular way to which objection could justifiably be taken. In the latter case it is the objectionable manifestation of the belief, and not the belief itself, which is treated as the reason for the act complained of. Of course, if the consequences are not such as to justify the act complained of, they cannot sensibly be treated as separate from an objection to the belief itself’ [emphasis added].

At a stroke, Underhill LJ thus did away with the old orthodoxy of relying on separability in tricky manifestation of belief cases and replaced it with an objective justification test. What Underhill LJ was suggesting was that even if the employer could identify a reason for the treatment separate from the manifestation of belief itself, it would still need to be able to show that the treatment was justified under Article 9(2) ECHR, or it would nevertheless be an act of direct religion or belief discrimination. This was highly unorthodox and a major departure from the previous position, yet Underhill LJ did not explain the jurisprudential basis for his decision anywhere in his judgment.

However, Underhill LJ was then presented with a further chance to explain himself when another manifestation of belief case – Higgs - came before the Court of Appeal in 2024.

Higgs – Underhill LJ doubles down

The claimant in Higgs was a Christian employee of a school who was dismissed for posts that she shared on her personal Facebook page. The posts commented on government policy on what schoolchildren should be taught about homosexuality and gender identities. The language used was inflammatory and included claims that ‘our children’ were being ‘brainwashed’ and that the ‘transgender roster’ and the ‘LBGT [sic.] crowd’ were ‘promoting mental illness’.

The ET dismissed the claimant’s direct discrimination and harassment complaints, holding that she was dismissed because the language used in her posts could lead a reader to conclude that she held homophobic and transphobic views, and such views did not qualify for protection under the EqA. However, in the EAT[6], Eady J held that the ET had failed to (1) engage with the question whether the posts were in fact a manifestation of protected beliefs held by the claimant and (2) apply the objective justification test established by the Court of Appeal in Page. Eady J explained the analysis that needed to be undertaken in respect of justification as being the application of the Bank Mellat[7] test familiar to human rights law (§92).  

Eady J remitted the matter to the ET, but the claimant then appealed the decision to remit on the basis that Eady J should have gone further and held for herself that the claim succeeded. This brought the issue before the Court of Appeal and Underhill LJ once more.

The Court of Appeal allowed the claimant’s appeal and Underhill LJ took the opportunity to explain his decision in Page with both hands. He put forward two alternative bases for introducing an objective element into s.13(1) EqA.

Firstly, Underhill LJ stated that it was necessary to read down s.13(1) EqA under s.3 Human Rights Act (‘HRA’) 1998 to render it compatible with the Convention (§81-§84). He seemed to appreciate that in doing so he was pushing the envelope: he prefaced his comments by acknowledging that ‘on the face of it, the protection conferred by section 13 is perfectly compatible with employees’ Convention rights’ (§82) and stating that he ‘initially thought that it followed that section 3 has no application in this case’ (§83). But on ‘further consideration’, he now thought ‘a more sophisticated argument’ could be made. 

Underhill LJ then went on to suggest that, because the CJEU in Bougnaoui had held that legislation implementing the Framework Directive needed to be construed as protecting manifestation of belief, the courts were obliged to interpret s.13 EqA so as to achieve that result (§83(2)-(3)). And because the CJEU had given achieving consistency with the Convention as the basis for its conclusion in Bougnaoui, the appropriate domestic tool was s.3 HRA. Since reading down under s.3 is only permissible to the extent necessary to achieve compatibility, and since Article 9(2) ECHR provided only qualified protection, it was necessary to interpret s.13 EqA as containing an objective justification defence in respect of manifestation of belief (§83(4)).

Secondly, Underhill LJ stated that inserting an objective justification defence into s.13 EqA where less favourable treatment was said to be because of manifestation of belief was in accordance with the legislative purpose of the EqA (§85). Underhill LJ said that Parliament could not be taken to have intended that employers ‘be obliged to tolerate any conduct at all by an employee which constituted a manifestation of belief, whatever form it took and whatever the circumstances’ and that this was ‘reinforced by the fact that the drafters of the Convention thought it necessary to qualify the right to manifest a belief’. It would therefore be necessary to imply on objective justification defence into s.13(1) EqA even if s.3 did not apply.

As to how that objective justification defence should operate, Underhill LJ said that he had considered possible alternatives to his approach in Page – in which he treated it as going to causation – but still considered the ratio of Page preferable (§97). Underhill LJ also confirmed that in applying this test it would be ‘best practice’ to consider each of the Bank Mellat steps (§75) and that it could therefore be equated with the test under Article 9(2) ECHR (§76).

Underhill LJ’s judgment in Higgs therefore explained the purported jurisprudential basis for his decision in Page very clearly. But whether it did so persuasively is another matter. In this author’s opinion, there remain considerable problems with his approach.

The problems with Higgs and the new status quo

The issues with the Court of Appeal’s decision in Higgs can be placed in two categories: those that go to its jurisprudential basis, and those that go to its impact.

Starting with the former, notwithstanding that Underhill LJ devoted a significant portion of his judgment to explaining them, it is hard to see how either of the bases on which he relies for introducing an objective justification test into s.13(1) EqA stand up to scrutiny.

Underhill LJ’s attempt to explain how s.3 HRA applied (at §83) was impressively creative, but he did not address how his reasoning overcame the fundamental point that s.13 EqA is not incompatible with the Convention. As already noted, Underhill LJ acknowledged that s.13 appeared ‘on the face of it’ to be ‘perfectly compatible’ and even observed that the protection that s.13 provided against direct manifestation of belief discrimination ‘may, because it is unqualified, go further than the Convention requires, and so place additional obligations on the employer’ (§82). But merely acknowledging this point before setting out his rationale is no substitute for adequately addressing it, which Underhill LJ conspicuously did not do. 

Even harder to understand, however, is Underhill LJ’s suggestion that, s.3 HRA aside, it is necessary to read an objective justification defence into s.13(1) EqA in the context of manifestation of belief as an aspect of construing it purposively. It is hard to argue with Underhill LJ’s statement that Parliament would not have intended that employers were ‘obliged to tolerate any conduct at all by an employee which constituted a manifestation of belief’ (§85). But for Underhill LJ to use this as a basis for reading into s.13(1) EqA an objective justification defence that simply is not there is not a purposive interpretation of the law but impermissible judicial legislation.

As for the implications of Underhill LJ’s approach, it is surely undesirable that findings of direct discrimination are likely to turn on assessments of the proportionality of the employer’s actions. The corollary of Underhill LJ’s approach would appear to be that, even where a protected belief is manifested in such a way as to amount to harassment of another, if the sanction that the employer chooses to impose in respect of such harassment is deemed to have been heavy-handed by a tribunal, the employer will be guilty of direct religion or belief discrimination.

By way of an example, an employer faced with a situation whereby a devoutly religious employee repeatedly (and pointedly) brings up their opposition to same-sex marriage in the presence of a gay colleague, and which is attempting in good faith to manage the situation as best it can, is placed in an extremely invidious position. Should it seek to impose any sanction, Page and Higgs effectively close the door on a tribunal being able to find that this was, as a matter of fact, simply because of a separable reason. The proportionality of the employer’s response is guaranteed to require investigation, and if the tribunal disagrees with the employer’s judgment call, then that will be direct discrimination. This surely cannot be right, not least when such findings can in some contexts carry serious regulatory consequences. Employers may well feel obliged to tread on eggshells, and this may in some cases mean that their employees are not afforded adequate protection from conduct which amounts to harassment.

Given Underhill LJ’s concern about employers potentially being ‘obliged to tolerate any conduct…which constituted a manifestation of belief’ (§85), it is ironic that his approach seems likely to make employers more fearful of imposing sanctions on employees who manifest their beliefs objectionably than they were when cases turned on separability.

However, notwithstanding the apparent flaws in Underhill LJ’s reasoning in Higgs, the Supreme Court recently refused permission to appeal. As such, Underhill LJ’s approach seems likely to remain the status quo for the foreseeable future. 

Amending the EqA – a better alternative?

No shortage of alternatives to Underhill LJ’s approach have been proposed both by the advocates in Higgs and in subsequent commentary on the case. These suggestions have ranged from the straightforward (such as a return to relying on separability) to the sweeping (such as amending the EqA).

While there is obvious appeal in a simple solution which cuts through the complexity, in this author’s opinion, none of the more straightforward options proposed adequately navigate the unique problems presented by manifestation of belief cases. Instead, there is a very strong case for amending the EqA, specifically by introducing to it a new form of discrimination which expressly provides that manifestation of belief discrimination can be justified.

For all that the jurisprudential basis for Underhill LJ’s decision in Higgs does not withstand scrutiny, it is plainly desirable for the law to recognise that it is possible for manifestation of belief discrimination to be justified. As already discussed, the old orthodoxy of relying on separability will not assist in every case where an employer has a legitimate basis for taking issue with a manifestation of belief. Moreover, even if there is currently no incompatibility with the Convention, it surely is preferable for the EqA to reflect Article 9(2) as closely as possible.

The real problem with Higgs is not the suggestion that detrimental treatment caused by manifestations of belief can be justified but the violence that has to be done to s.13(1) and to orthodox causation principles to get there. And the root cause of Underhill LJ’s need to be so creative to shoehorn his justification test into the EqA is that the EqA currently deals with the unique complexities thrown up by the protected characteristic of ‘religion or belief’ so poorly.

By treating the definition of ‘religion or belief’ in s.13 (read with s.4) as encompassing both the holding of and manifesting of a protected belief, the EqA on its face affords no room for these two aspects of a belief to be treated in the differing ways that they clearly should be to maintain consistency with Article 9.

The law would surely be clearer, more certain, and more accessible if the approach that the EqA took to ‘religion or belief’ as a protected characteristic was more similar to the approach taken to disability. If discrimination on the ground of manifesting a belief should be possible to justify, but discrimination on the ground of holding a belief should not, the logical way to facilitate this is to confine the definition of ‘religion or belief’ in all contexts to the holding of a belief only, and to introduce into the EqA a s.15 equivalent that deals with unfavourable treatment because of ‘something that is a manifestation of a protected belief’ (or words to that effect). There appears to be no reason why the ‘proportionate means of achieving a legitimate aim’ test could not then be used to assess justification so as to maintain consistency with the other provisions of the EqA, and so in substance this new form of discrimination could be extremely similar to s.15.

If the above form of discrimination was added to the EqA, there would also be no reason why ordinary causation principles should be suspended in the context of manifestation of belief. It should remain open to a court or tribunal to find that, as a matter of fact, the reason for impugned treatment was something separate from either the holding of or manifestation of a belief, just as it is open to a court or tribunal to find that impugned treatment was neither caused by a claimant’s disability nor something arising from it. Plainly, judges should scrutinise cases where employers rely on reasons separate to the holding or manifestation of a belief extremely carefully. They should be entitled to infer from the fact of a sanction being disproportionate that the proffered reason may not be the real reason, and it would be appropriate for them to be slow to find that the reason is separable where a protected belief has been manifested. But such a finding should not be impossible, and if a judge is satisfied that as a matter of fact the reason was separable, the impugned treatment should not require justifying. 

Some may suggest that introducing an entirely new form of discrimination into the EqA is a heavy-handed response to the problems presented by Underhill LJ’s judgment in Higgs. It certainly seems unlikely to happen any time soon, not least since the Supreme Court’s decision to refuse permission to appeal. However, without a s.15 equivalent for manifestation of belief addressing the issue clearly and comprehensively, it seems inevitable that manifestation of belief cases will continue to trouble the appellate courts regularly in the years to come. Parliament would therefore be doing employees, employers, lawyers, and judges a significant favour if it stepped into the arena and tried to resolve the problem of manifestation of belief once and for all.

[1] Page v NHS Trust Development Authority [2021] EWCA Civ 255

[2] Forstater v CGD Europe [2022] ICR 1

[3] Miller v University of Bristol (ET Case No: 1400780/2022)

[4] Higgs v Farmor’s School [2025] EWCA Civ 109

[5] Bougnaoui v Micropole SA (C-188/15), [2018] ICR 139

[6] Higgs v Farmor’s School [2023] ICR 875

[7] Bank Mellat v HM Treasury (No. 2) [2013] UKSC 39