R Godfrey v Natwest Markets Plc

The issue raised by this appeal relates to the (actual or constructive) knowledge of disability required for the purposes of sections 13 and 15 Equality Act 2010 (“EqA”); specifically, as to whether the Employment Tribunal (“ET”) properly considered this question by reference to the relevant factual features of the disability in issue, rather than its diagnosis.

In giving this judgment, I refer to the parties as the claimant and respondent, as below. This is my judgment on the claimant’s appeal against the decision of the London Central ET (Employment Judge Norris, sitting with Dr V Weerasinghe and Mr P Secher, over four days in November 2021), sent to the parties on 21 December 2021. The claimant had brought claims of disability discrimination under sections 13 and 15 of the EqA; the disability on which he relied was Asperger’s syndrome, an autism spectrum condition. The ET dismissed the claims after finding (unanimously) that the respondent did not have actual knowledge of the claimant’s disability and (by a majority) that it did not have constructive knowledge of that disability.

The claimant has appealed against the ET’s decision and, after a hearing under rule 3(10) Employment Appeal Tribunal Rules 1993 (as amended), on 9 March 2023, His Honour Judge Auerbach permitted the appeal to proceed on one ground only:

Both parties were represented by counsel before the ET, albeit Mr Milsom did not then appear. At the outset of the hearing, Mr Milsom sought to widen the scope of the appeal to argue that the ET had erred in its application of the burden of proof. After reviewing the respondent’s note from the rule 3(10) hearing, however, Mr Milsom accepted that this point had previously been considered, with permission being refused, and the appeal was thus limited to the one ground I have set out above.

The claimant’s claims before the ET were of: (1) direct discrimination because of disability, under section 13 EqA, and/or (2) discrimination by way of unfavourable treatment because of something arising in consequence of the claimant’s disability, under section 15.

In a case of direct disability discrimination, the respondent must have treated the claimant less favourably because of disability. That implies that the respondent knew of the relevant disability, although that need only be knowledge, whether actual or constructive, of the facts that made the claimant a disabled person under the EqA; as the Court of Appeal (per Rimer LJ) explained in Gallop v Newport City Council [2013] EWCA Civ 1583, [2014] IRLR 211 (then considering the relevant provisions of the legacy statute, the Disability Discrimination Act 1995 (“the DDA”)):

In Gallop, it was held that the ET had erred in finding that Newport Council had been entitled to deny relevant knowledge by relying simply on its unquestioning adoption of the unreasoned opinion of its occupational health advisers that Mr Gallop was not a disabled person. As Rimer LJ observed:

In advancing his submissions in support of the appeal, it was Mr Milsom’s position that this case is more appropriately to be understood as one of unfavourable treatment because of something arising in consequence of the claimant’s disability for the purposes of section 15(1) EqA. In this regard, section 15(2) provides that the protection:

It is common ground that the approach that an ET is to adopt when considering the defence thus provided by section 15(2) is the same as that which applies (perGallop) when considering knowledge for the purposes of section 13. As the EAT explained in A Ltd v Z [2020] ICR 199:

The legal principles, derived from the relevant case-law and guidance, that will inform the ET’s approach were summarised at paragraph 23 A Ltd v Z, as follows:

The definition of disability is now to be found at section 6 EqA, which (relevantly) provides:

It is thus the adverse effect of the relevant impairment that is to be the subject of the ET’s assessment, requiring it to focus not upon what the claimant can do but upon that which he cannot; Aderemi v London and South Eastern Railway Ltd [2013] ICR 591 EAT paragraph 15. As for the requirement that the adverse effect be “substantial”, that means only that it is “more than minor or trivial”; section 212(1) EqA. Unless an effect can be classified as trivial or insubstantial, it will necessarily be “substantial”; Aderemi paragraph 14. Moreover, given that the protection is intended to enable employees to advance in their employment, in considering the effect on “normal day-to-day” activities, these are to be taken to encompass the activities that are relevant to participation in professional life; Paterson v Commissioner of the Police of the Metropolis [2007] ICR 1522 EAT paragraphs 66-67.

Given that the unfavourable treatment must be because of something arising in consequence of the claimant’s disability, the question of knowledge, whether under section 13 or 15(2) EqA, is directed at the putative discriminator (see, by analogy, CLFIS (UK) Lt v Reynolds [2015] ICR 1011 CA, as applied in Gallop v Newport City Council (No. 2) [2016] IRLR 395 EAT). Where, however, an employer’s agent or employee (such as an occupational health adviser or human resources officer) knows of the relevant disability, paragraph 5.17 of the Equality and Human Rights Commission Employment Statutory Code of Practice suggests that:

The determination of what a respondent knew, or ought reasonably to have known, will depend on the findings of fact made by the ET. As such, provided the decision reached reveals no error of legal principle, takes into account all relevant factors, and is not influenced by that which is irrelevant, there is no scope for interference on the part of the EAT, unless the conclusion can properly be described as perverse; that is, where an overwhelming case is made out that no reasonable ET, on a proper appreciation of the evidence and the law, could have reached that conclusion; Yeboah v Crofton [2002] IRLR 634 CA.

As for the approach I am to adopt to the ET’s reasoning, as emphasised in DPP Law Ltd v Greenberg [2021] EWCA Civ 672, the decision is to be read holistically, without being hypercritical, and:

That said, where an ET has fallen into error, the EAT’s role is not to strive to uphold a decision where the reasoning reveals a fundamental error of approach; as Sedley LJ observed in Anya v University of Oxford [2001] ICR 847 CA:

By his ET claim, presented on 31 December 2018, the claimant complained that he had suffered disability discrimination under the EqA due to alleged refusals by the respondent to appoint him, or allow him the opportunity to be considered for appointment, to various vacancies. Those vacancies were said to have arisen (relevantly for present purposes) between March 2017 and January 2019. The claimant had previously been employed by the respondent, then called Royal Bank of Scotland plc, from 11 September 2006 until he resigned on 31 January 2011. The disability relied on by the claimant was the autistic spectrum condition known as Asperger’s syndrome; although a lifelong condition, the claimant did not, however, have a diagnosis until 2018, some years after he had left the respondent’s employment. As for the “something arising” for section 15 EqA purposes, this was said to be the claimant’s need for quiet and space, and/or the fact that he “would not engage in conversation or social interactions in the same way as others” (ET, paragraph 11).

The claimant’s claims were resisted by the respondent, which put in issue the question whether he was, at any material time, disabled for the purposes of the EqA. That question was considered by the ET (EJ Snelson) at a hearing on 4-5 August 2020, where the ET received oral evidence from the claimant together with statements from two of his friends and reports from two psychiatrists, Dr Craig and Dr Smith. Dr Smith’s report, dated 5 September 2019, recorded a detailed history taken from the claimant, and:

On the evidence, the Snelson ET was:

The issue of disability having been resolved, directions were given for the final liability hearing, which was ultimately listed to take place on 23-26 November 2021. One of the issues identified for determination at the final hearing was the question of the respondent’s knowledge of the claimant’s disability, and, at a further preliminary hearing before the ET (EJ Spencer), on 1 February 2021, it was recorded:

At the hearing in November 2021, the ET heard evidence from the claimant and from two witnesses on his behalf: Mr Hammacher (a friend of the claimant since 2003, who also happens to be medically qualified, although he has not practised for many years) and Mr Horne (who had worked closely with the claimant at Credit Agricole, where the claimant was employed after he left the respondent); the claimant also relied on a statement from his former line manager at the respondent, Mr Welzel. For its part, the respondent called one witness, Mr Muscatt, who had worked on the same trading desk as the claimant between 2008 and 2011, and who the claimant contended had been a block to his being considered for appointment in the period 2017-2019.

Having reminded itself of the relevant statutory provisions and the guidance in A Ltd v Z, the ET considered the evidence it had received. For the claimant, it was urged that the ET ought to take into account his demeanour during the hearing. The ET disagreed, finding:

In his witness statement, the claimant had referred to repeated experiences of cruel language and comments, with some colleagues having described him as:

In any event, it was only in June 2017 that the claimant undertook a psychiatric evaluation at the behest of his then employer (Aviva), after he had had a breakdown, receiving a preliminary diagnosis of Asperger’s syndrome in 2018, with a full diagnosis in September that year. It was the claimant’s case that, although he had only known of the impact of his condition since his diagnosis, those who sat around him at work would have been fully aware of his communication and social interaction problems on a daily basis.

Correctly directing itself that it was for the respondent to show it did not have the requisite knowledge, rather than for the claimant to show that it did (ET, paragraph 20), the ET nevertheless considered that the claimant needed to “give some context”. Considering the claim form, it noted that the claimant had referred to the following incidents during his employment with the respondent, where:

It is apparent that the ET sought to elicit further explanation from the claimant and he had then also referred to another incident when he was working for the respondent, when he had had an “awkward” conversation with a colleague who had said to him:

For the respondent, however, it was Mr Muscatt’s evidence that, during his earlier period of employment, the claimant “did not stand out” (ET, paragraph 31); that while he had found the claimant to be “slightly arrogant” (which he did not see as a necessarily bad trait), he did not feel he was different from anyone else. Mr Muscatt had been under the impression that the claimant had either been dismissed or asked to leave on agreed terms because he failed to adhere to internal trading limits (ET, paragraph 36); this was not, however, the respondent’s position. Considering it was not required “to make alternative substantive findings”, the ET did not proceed to determine why (or even if) Mr Muscatt might not have wished the claimant to be re-employed by the respondent, but observed that he had appeared to have the mistaken belief that the claimant had previously been required to leave his employment (ET, paragraph 67). Otherwise, the ET noted:

Disagreeing with Mr Muscatt’s characterisation as to how he would have come across, the claimant relied on the evidence of Mr Hammacher, a friend from university days, who explained that the claimant:

In any event, while acknowledging that the claimant could cope very well in a structured team environment and on a one-to-one, or one-to-two basis, it was Mr Hammacher’s evidence that the claimant became withdrawn and quiet in group social situations. Although Mr Hammacher had only ever seen the claimant in social, rather than professional, settings, he made clear that he would be surprised if the claimant’s unusual behaviour did not transfer to the workplace.

As for the statement provided by Mr Welzel (the claimant’s manager during his employment with the respondent), the ET recorded that this was very short (less than one page), and, although signed, contained five sentences in square brackets. While Mr Welzel had referred to being aware of the claimant’s “communication difficulties”, no further context or explanation was provided, and the ET concluded:

In seeking to determine what the respondent had known (or ought reasonably to have known) about the way the claimant had been affected by his impairment, the ET also had regard to the documentary evidence available from the time of his previous employment, between 2006 to 2011. Noting that this included a reference to Mr Muscatt referring to the claimant as an “irritating (plastic) spoon”, the ET found, however, that this was an expression Mr Muscatt sometimes used, drawing attention to his own class background, further recording that the claimant appeared to accept expression was unrelated to his disability (ET, paragraph 39).

The ET also recorded an exchange in April 2010, involving some older employees, referring to the claimant being late into work after a night out with some junior colleagues, during which the claimant was referred to as “a TAD strange”, and “odd”. The ET did not, however, view this as significant:

Considering all the evidence before it relating to the claimant’s previous period of employment with the respondent, and referring to further evidence given by the claimant, the ET concluded as follows:

As for the evidence of how people within the respondent had spoken about the claimant after he had left, the ET recorded that, in November 2012, a Mr Rad (employed by the respondent between November 2008 and May 2014 as head of a team) had told a recruitment agent that the claimant was “a no-go!” and had subsequently (in March 2013) said that “jobs are hard to find for guys like him”. Reflecting on the fact that Mr Rad had been prepared to make a number of unguarded comments relating to others, and was said to have behaved in an openly offensive manner to a colleague with a physical impairment, the ET concluded that, had the claimant’s “manner, or the impact(s) to which his condition gives rise”, been noticeable to Mr Rad, he would not have held back from referring to this; he had, however, not done so (ET, paragraph 48).

The ET also noted that Mr Muscatt had a Bloomberg chat with someone from another organisation during which Mr Muscatt had referred to both the claimant and a Mr Stevens in the following terms:

Considering other evidence relating to how the claimant’s behaviour might have been perceived in a workplace context, the ET referred to the evidence of Mr Horne, who had worked with the claimant at Credit Agricole from January 2011, immediately after the claimant had left his employment with the respondent. Mr Horne had described the claimant having “mood swings, being non-communicative, offhand and abrupt during the early part of the day, suddenly returning to “normal” later on” (ET, paragraph 50). The ET noted, however, that the circumstances in which the claimant was working at Credit Agricole were very different (and more stressful) to those he had experienced at the respondent, finding that the behaviour described by Mr Horne would not have been replicated at the respondent, and concluding:

The claimant had also provided a supplemental witness statement that included a reference from a line manager review from 2018 stating that he had “a tendency to not inform seniors of what he is doing” (it seems no further detail was given). The ET did not find this of assistance:

In addition, it was part of the claimant’s case that, in or around late 2018, he had in fact specifically told one of the respondent’s senior traders, Mr Balax, of his diagnosis, and he considered that Mr Balax would have told Mr Muscatt so that they would both have had actual knowledge of his disability from that time. Mr Muscatt denied that any such information had been communicated to him, and his evidence was accepted by the ET. As for what Mr Balax had been told, the ET did not find the claimant’s evidence reliable:

Although the ET’s reasoning reveals some differences between the majority and minority in the approach taken to certain aspects of the evidence, on the question of actual knowledge, the conclusion was unanimous:

Considering, however, what the respondent might reasonably have been expected to know (its constructive knowledge), the ET panel was divided. Asking:

… The Claimant has had a number of jobs, for all of which we may safely assume he has passed at least one interview. His behaviours apparently do not normally manifest during those interviews … to cause employers to repel him. He also told us that on one occasion he had got through to a “final round” of interviews but had become nervous and not presented well, so he did not secure the role. That is entirely in keeping with behaviour for somebody who does not share the Claimant’s disability.”

… the Claimant would have been resistant to an assessment. Until he had his breakdown, the Claimant did not seek medical advice. This was notwithstanding, on his own evidence, contemplating in 2015/2016 a possible medical explanation for the adverse treatment he felt he was receiving and despite, he said in cross-examination, every recruitment agency he worked with during the relevant time suggesting that he should undergo a psychiatric evaluation. …”

The ET minority took a different view. Referring to the findings of EJ Snelson (based on the report of Dr Smith) and to the evidence of Messrs Hammacher, Horne and Welzel, the minority concluded that:

For the claimant, Mr Milsom emphasises that the ET was required to ask not whether the respondent had actual or constructive knowledge of his diagnosis but of the relevant facts. The claim - best understood under section 15 EqA - was that the respondent had subjected the claimant to unfavourable treatment by reason of something arising in consequence of his disability (relevantly, the fact that the claimant “wouldn’t engage in conversation or social interactions in the same way as others”). EJ Snelson had previously found it clear that the claimant’s condition had adversely affected “his interaction and communication with others (in the workplace and elsewhere) to a significant extent”; in subsequently determining the question of knowledge, the ET had, however, concluded that there was no factor that would have prompted the respondent to take “more account” of the claimant’s differences and to “realise they might be the result of a mental impairment or autistic spectrum disorder” (ET, paragraph 52). That, however, was the wrong test and demonstrated that the ET had focused on whether the respondent had actual or constructive knowledge of the claimant’s diagnosis and not on the salient ingredients of section 6 EqA status. This, moreover, was not a solitary slip; the ET’s reasoning demonstrated that it had (notwithstanding any prior legal direction) applied the wrong test throughout (see paragraphs 44, 45, 54, 66 and 68), such that any attempt to rescue the decision would fall into the trap deprecated by Sedley LJ in Anya.

Having initially sought to expand the appeal to include an argument relating to the burden of proof, but having to accept that was a potential ground of appeal dismissed at the rule 3(10) stage, Mr Milsom’s submissions also referenced parts of the evidence relied on by the claimant before the ET in support of a more general perversity challenge. Acknowledging, however, that the appeal had similarly not been permitted to proceed on that basis, these were ultimately matters relied on as supporting the claimant’s submission that any remission (assuming the appeal to be upheld) must be to a differently constituted ET.

For the respondent, Mr Zovidavi submits that, the ET having reminded itself of the test laid down in A Ltd v Z, there was no basis for inferring it had done other than to then apply that guidance; in particular, it had taken care to consider various features of the claimant’s presentation, behaviour and conduct, and had had regard to all the evidence that might assist in determining how those matters might have appeared to the respondent at the material time. Acknowledging that the wording (e.g. at paragraph 54 of the ET’s reasoning) went “a little further than it ought”, Mr Zovidavi argues that that alone could not detract from the considered reasoning set out elsewhere in the decision, which evidently focused on the respondent’s knowledge of the factual features of the claimant’s disability rather than knowledge of the disability itself; to the extent the ET referred to the claimant’s diagnosis, that was plainly just a shorthand, getting to conclusion of exercise. Mr Zovidavi further contends that it was of note that there was no evidence before the ET that went to knowledge of impact on day-to-day activities, or long-term nature, or of any impairment being more than merely transient.

In any event, the ET majority had found that there was no trigger event during the claimant’s employment that would have caused the respondent’s managers to make a referral to occupational health; that was a finding of fact against which there was no challenge. Yet further, the majority had found that the claimant would have been resistant to any proposal of a referral to occupational health; that again was a finding of fact, against which there was no challenge; by analogy with A Ltd v Z, the answer to the question what the respondent might reasonably have been expected to know if such a referral had been proposed would be “nothing”.

In addressing this appeal, I have followed Mr Milsom’s steer and focused on section 15 EqA (although it is not suggested that there would be any difference if I were instead to consider this as a case of direct discrimination under section 13 EqA), which provides that it will be an act of discrimination if a respondent treats a disabled person unfavourably because of something arising in consequence of that person’s disability and the respondent cannot show the treatment to be a proportionate means of achieving a legitimate aim. By section 15(2), however, a respondent is provided with a further potential defence to such a claim if it can show that it did not know, and could not reasonably have been expected to know, that the complainant had the disability in question. Whether or not a respondent is able to make good such a defence must always be a question of fact, to be determined on the evidence before the ET; it will not be open to the EAT to look behind the decision of the ET unless it is apparent that it has applied the wrong test, has failed to have regard to a relevant factor, or has taken account of that which is irrelevant, or has reached a decision that is properly to be said to be perverse. On the particular ground permitted to proceed in the present appeal, the challenge is put solely on the basis that the ET erred by failing to apply the correct legal test.

In this case, it is common ground that the legal test the ET was to apply was as set out in A Ltd v Z: in seeking to establish that it did not know of the claimant’s disability, it would not be sufficient for the respondent to show that it did not know (and could not reasonably have been expected to know) the specific diagnosis; rather, it would need to show that it did not know (and could not reasonably have been expected to know) that the claimant (i) had a physical or mental impairment, that (ii) had a substantial and long-term adverse effect on his ability to carry out normal day-to-day activities. It is equally not in dispute that the ET expressly set out the guidance provided in A Ltd v Z; in accordance with DPP Law v Greenberg, I therefore start with the premise that, having reminded itself of the correct legal test, the ET should be assumed to have applied that test unless the contrary is clear from the language of its decision.

The ET was concerned with allegedly discriminatory conduct by the respondent in the period March 2017 to January 2019. Although the claimant had suggested that the relevant decision maker, Mr Muscatt, must have known of his diagnosis in or around 2018 (having been told by Mr Balax), the ET unanimously rejected that possibility. Given the limited scope of this appeal, the claimant’s case thus has to proceed on the basis of the ET’s rejection of any direct knowledge of his diagnosis on the part of any relevant decision maker within the respondent through information that the claimant considers he imparted to Mr Balax.

That was not, however, the end of the task for the ET, and its analysis did not stop at that point. Consistent with the legal test it was required to apply, the ET carried out a detailed assessment of all the evidence available to see what would (or should) have been known by the respondent about how the claimant might have been affected by his disability. That its analysis reached back to 2006, notwithstanding that the claimant had no diagnosis of his condition until 2018, supports the view that the ET understood that it was concerned with the respondent’s knowledge (actual or constructive) of the facts constituting the claimant’s disability rather than the particular label that was to be attached to that disability upon its eventual diagnosis.

Thus, in carrying out its assessment of the evidence, the ET looked back to what was (or ought reasonably to have been) apparent to the respondent during the claimant’s earlier period of employment from 11 September 2006 until he resigned on 31 January 2011, including within the scope of its review the wider picture provided by friends of the claimant, who could attest to how his disability affected him in other social or workplace settings. In his skeleton argument in support of the appeal, Mr Milsom listed the various aspects of that evidence that he suggested supported a conclusion that the respondent was (or ought reasonably to have been) aware of facts consonant with the claimant’s diagnosis. Many of the matters included within that list were, however, expressly found by the ET to be irrelevant (for example, Mr Muschatt’s use of the word “spoon”, which had an entirely different context (as the claimant appeared to accept); or his reference to “space cadets”, which the ET accepted referred to young people who enjoyed getting drunk), or simply neutral (Mr Muschatt’s description of the claimant as “slightly arrogant”). As any possible perversity challenge in this regard was dismissed at the rule 3(10) stage, I am bound to consider this appeal on the basis of the material that the ET did not reject, which essentially came down to the following:

Although it is also right to note that the ET did not reject the evidence of Mr Hammacher or Mr Horne, it nevertheless unanimously concluded that their respective experiences of the claimant’s behaviour did not undermine the respondent’s evidence (essentially given through Mr Muscatt, but as corroborated by other material) that there was nothing that had caused it to have actual knowledge of the claimant’s disability. More generally, as I have recorded, the ET unanimously accepted the respondent’s case that the claimant’s behaviour, as manifest during his earlier employment from 2006 to 2011, was not such as to mean it had actual knowledge that he was disabled, albeit the ET was split when asking whether the respondent had also demonstrated there was equally nothing that ought reasonably to have caused it to have had that knowledge.

Accepting that the claimant cannot go behind the findings of fact made at first instance, Mr Milsom contends that the conclusions reached are nevertheless rendered unsafe because the ET continually focused on the claimant’s diagnosis rather than the relevant factual questions, namely whether the respondent had known, or ought reasonably to have known, that he suffered a mental impairment which had a more than minor or trivial, long-term impact on his ability to carry out normal day-to-day activities. In support of this submission, Mr Milsom picks out various references to the diagnosis of the claimant’s disability within the ET’s reasoning, as follows:

The reference at paragraph 66 does not assist Mr Milsom’s argument, as the ET was there concerned with the claimant’s evidence (not found to be reliable) that he had expressly told Mr Balax of his diagnosis. At first sight, however, the other passages relied on might seem to evince a focus on the specific disability in issue (Asperger’s syndrome), rather than the question the ET had to answer, which concerned the respondent’s knowledge (actual or constructive) of the relevant factual constituents of that disability. That said, such an approach would be counter to the ET’s self-direction as to the legal test it was required to apply, and it is the respondent’s submission that the use of a convenient short-hand (rather than continually referring to whether it knew (or ought to have known) that the claimant suffered a mental impairment, which had a more than minor or trivial long-term adverse effect on his ability to carry out normal day-to-day activities) should not cause me to infer that the ET lost sight of relevant test which it had earlier set out.

Given the limited evidence of impairment that the ET found to be reliable in this case (summarised at sub-paragraphs 50(1)-(6) above), it is tempting to adopt the approach urged by the respondent, and to read the passages relied on by Mr Milsom as effectively using the diagnosis of the claimant’s disability as a shorthand for a finding that the requisite constituent facts - the existence of a mental impairment, with a non-trivial and long-term impact on the claimant’s ability to carry out normal day-to-day tasks - were not, and could not reasonably have been, known to the respondent. In support of that approach, I remind myself that, not only did the ET correctly direct itself to the relevant legal test, it then carried out a wide-ranging assessment of the evidence relating to how the claimant would have presented during his earlier period of employment with the respondent. Thus the scope of the enquiry was certainly consistent with a focus on the manifestation (if any) of an impairment, as demonstrated (for example) by the following parts of the ET’s reasoning:

The difficulty remains, however, that the reasoning at no stage expressly answers the specific questions the ET was required to determine, namely: whether the respondent had demonstrated that it did not know, and could not reasonably have known, that the claimant suffered a mental impairment, with a non-trivial and long-term impact on the claimant’s ability to carry out normal day-to-day tasks. The question is whether that, taken together with the passages relied on by Mr Milsom, renders the ET’s decision unsafe?

In relation to actual knowledge, it seems to me that the answer must be no. On the evidence before it (and taking into account the matters cited in the claimant’s claim form), the ET permissibly concluded that:

As for what the respondent ought reasonably to have known, as Mr Zovidavi acknowledged in his skeleton argument, parts of the ET majority’s reasoning (for example, paragraph 54) might be read as suggesting that it was setting a higher test than that required under section 15(2) EqA. The question was not whether a reasonable assessment of what was known, along with the making of any enquiries that might reasonably have been expected, would have led the respondent to an understanding that the claimant had an autistic spectrum disorder; the issue for the ET was whether the respondent had shown that, even with such an assessment and any enquiry, it could not reasonably have known that the claimant was suffering a mental impairment with the relevant effects.

Understandably, Mr Milsom places emphasis on the conclusion reached by the minority member of the panel, who considered that the claimant’s persistent deficits would have been sufficiently apparent to the respondent that it would have been reasonable for it to have known that he suffered some form of mental impairment, such that it would also have been reasonable for it to have investigated matters further. Acknowledging that, during the period of his employment with the respondent (and, more generally, prior to 2018) the claimant would himself “not have known that there was anything amiss” the ET minority concluded that if the respondent had then “initiated an investigation”, the proper diagnosis of that impairment would have resulted (as it had in 2018).

The ET majority did not, however, share the same view, concluding that it would not have been incumbent upon the respondent to make further enquiry when there was little or no basis for it to do so. Given the very limited evidence of any manifestation of the claimant’s condition during his employment with the respondent, that, it seems to me, might well have been a conclusion open to the ET majority in this case. The difficulty, however, is that the language used in the majority’s reasoning suggests that it reached that decision by imposing a requirement that the respondent had been put on notice that the claimant might have suffered from an autistic spectrum disorder. Contrary to the legal test it had earlier referenced, the ET majority’s reasoning at (for example) paragraphs 44 and 54 of the decision is focused on whether the respondent might reasonably have been put on notice of a particular medical diagnosis, rather than the question whether it might reasonably have been alerted to the need to make further enquiry about, more generally, the possible effects of some mental impairment. Acknowledging the respondent’s argument (that the reference to autistic spectrum disorder might have been intended as a convenient shorthand), but also being alive to Sedley LJ’s warning in Anya, the language used by the ET majority does not enable me to be satisfied that it in fact applied the correct legal test. On this point, therefore, I would agree that the decision of the majority is rendered unsafe.

As Mr Zovidavi submits, however, that is not necessarily fatal to the ET majority’s decision on constructive knowledge. Notwithstanding that it had found there was no basis for the respondent to have reasonably considered any investigation was required, the majority went on to consider the counterfactual, asking itself what would have happened if the respondent had, nonetheless, sought to make further enquiry. In this regard, on the evidence before it, the ET majority was satisfied that the claimant would have resisted any suggestion that he should undergo an assessment (ET, paragraph 56). That was a finding of fact that was open to the ET majority, and against which there is no extant challenge. On that basis, as Mr Zovidavi observes, any attempted investigation on the part of the respondent into the cause of the claimant’s behaviours - whether any aspect of his conduct might be due to a mental impairment that gave rise to a more than trivial, long-term adverse effect on his ability to undertake normal day-to-day activities - would have taken matters no further. As such, the ET majority was entitled to conclude that, in its later dealings with the claimant in 2017-2019, not only did the respondent not have any direct knowledge of the claimant’s disability but that it also could not reasonably have been expected to have such knowledge.

For the reasons provided, I therefore dismiss this appeal.