United Kingdom Employment Appeal Tribunal

EMPLOYMENT APPEAL TRIBUNAL

		Rolls Building Fetter Lane, London, EC4A 1NL
		4 June 2024

SUMMARY

Race Discrimination

The Employment Tribunal did not err in law in holding that the burden of proof had shifted to the respondent to disprove discrimination and that it had failed to do so.

HIS HONOUR JUDGE JAMES TAYLER

Introduction

1. This is an appeal against the judgment of the Employment Tribunal sitting at Leicester from 23-27 January 2023; Employment Judge Ahmed sitting with members. The Judgment was sent to the parties on 22 March 2023. The parties are referred to as the claimant and respondent as they were before the Employment Tribunal.

Outline facts

2. The claimant has an MA in Social Work. She started work for Leicestershire County Council on 5 June 1989. She worked for Leicestershire County Council, with some time out for studies, until 1997, when her employment was transferred to the respondent, Leicester City Council.
3. The claimant has considerable experience of social work and management.
4. In 2005, the claimant was appointed as Head of Service for Locality West, one of the Service Areas within the Adult Social Care and Safeguarding Division.
5. The claimant had responsibility for a number of teams, each of which had a Team Leader.
6. Prior to the events that gave rise to the claim, the claimant had not been the subject of any disciplinary proceedings or performance measures.
7. The claimant described herself as a British National of Indian origin.
8. The Employment Tribunal held that there were poor relations between Contact and Response, another Service Area within the Adult Social Care and Safeguarding Division, and the claimant's Service Area, Locality West, which often extended to conflict between team leaders.
9. Ruth Lake was the Director of Adult Social Care and Safeguarding. She had to deal with the conflict between these two Service Areas.
10. A number of staff members who remain employed by the respondent were referred to by initials in the Judgment. I am not aware of any compelling reason why this derogation from the open justice principle was considered necessary, but have little choice but to adopt the same approach.
11. HM, described in the Judgment as white British, was the Head of Service for Contact and Response. In May 2018 HM swore during a telephone call with Ms Lake, such that she could be heard in the open plan office. Ms Lake discussed the incident informally with HM, who accepted that her conduct was inappropriate and unprofessional. Ms Lake took no further action.
12. On Friday 16 November 2018, HM sent an email stating that from the next Tuesday members of the public who contacted her division, Contact and Response, would have to be directed to Locality West. The email was received badly by staff in Locality West. The plan was described by one recipient as "completely unfeasible". Another complained about a lack of "professional dialogue/handover". The claimant was on holiday when she responded. The claimant alleged that HM was guilty of harassment or bullying. Ms Lake took no action about the way in which HM had dealt with the handover.
13. On 8 January 2019, the claimant attended a supervision meeting with Ms Lake, in which she raised the tone of communications from HM. The claimant accused Ms Lake of having an unconscious bias against black and ethnic minority Heads of Service and alleged that when Ms Lake came to Bosworth House, where the Heads of Service were based, she always chose to sit with white colleagues. Ms Lake asked the claimant if she was accusing her of being racist. The claimant said that Ms Lake "needed to reflect on that, but that was how she felt".
14. Ms Lake spoke with the team leaders in Locality West and found that they were in the process of lodging a collective grievance against HM. When the grievance was submitted the team leaders said they felt "devalued and denigrated by the oppressive and inequitable treatment we have found ourselves subjected to on two occasions in a year".
15. On 14 February 2020, a session was held by JD, a Principal Social Worker in Social Care and Education. One of the attendees, JR, sent an email after the session suggesting that she had felt humiliated and denigrated by JD. JR replied that the content and tone of the email was unacceptable and she had raised it with Ms Lake. Ms Lake offered mediation between JD and JR. She did not instigate a formal disciplinary process.
16. SCC, a team leader, was Acting Head of Service for Learning Disability Services from November/December 2020 onwards. The Employment Tribunal described her as white British. On 13 December 2020, SCC sent an email to Ms Lake stating that she no longer wanted to be mentored by the claimant and that she had concerns about Locality West. The concerns raised were described by the Employment Tribunal as at a "fairly low level"; involving awkward interactions between the claimant and her subordinates. It was suggested that the claimant had been rude in a telephone call.
17. On 16 December 2020, AE sent an e-mail to Ms Lake in which she raised a number of concerns about the working relationship between Locality West and Contact and Response. She did not complain specifically about the claimant.
18. On 20 December 2020, Ms Lake contacted HR to discuss her "concerns" about the claimant.
19. On 4 January 2021, there was an angry e-mail exchange between JR, SR and AE about a case in which safeguarding alerts apparently had not been actioned. The claimant was copied in as Head of Service. She saw the email on her return from holiday on 5 January 2021. The claimant thought that the dispute was a difference of opinion about when time alerts should be entered on the system. She decided to seek the opinion of a Principal Social Worker.
20. On 7 January 2021, AE sent an email to Ms Lake complaining about the claimant and "her team". AE wrote "The West locality have a long reputation of intimidating and unhelpful behaviour long before I or SR joined the authority and I cannot understand why they are still able to behave in the way they do". A little later AE suggested that she would be complaining against the claimant, asserting victimisation. Subsequently, AE tendered her resignation alleging that the claimant was "vindictive and unprofessional".
21. The Employment Tribunal recorded that Ms Lake gave evidence that she spoke with HR in January 2021 to discuss what to do and that "both she and HR agreed that a disciplinary investigation was appropriate". Ms Lake decided to temporarily transfer the Claimant from her role as Head of Service". On 12 January 2021, the claimant was informed of her temporary transfer while a disciplinary investigation took place.
22. Between 26 January and 5 February 2021, Ms Lake interviewed nine witnesses. SR was not interviewed.
23. On 9 February 2021, Ms Lake sent a letter inviting the claimant to a disciplinary investigation meeting on 19 February 2021. The allegations were vaguely phrased:

"Failure to behave in accordance with agreed management/leadership standards, both ASC specific and as set out in the code of conduct and professional social work standards;

Failure to ensure that people you directly manage behave in accordance with agreed standards;

That these failures have created an environment that is detrimental to individuals and to the delivery of core functions, which could impact on people who need our support"

24. An investigation meeting was held with the claimant by Teams on 19 February 2021. The meeting was not concluded. A follow-up meeting was arranged for 24 February 2021.
25. On 22 February 2021, the claimant sent an email to a large number of recipients, including the Mayor and all elected Councillors, raising whistleblowing issues and asserting race discrimination against "BAME" staff by Ms Lake.
26. The claimant was absent from work from 23 February 2021 to 25 March 2021, suffering from work related stress. The meeting that had been arranged for 24 February 2021was postponed.
27. At some point between 24 February and 4 March 2021, Martin Samuels, Strategic Director of Social Care and Education, decided that the investigation into the claimant's conduct would be taken over by Caroline Tote, Strategic Director of Social Care and Education.
28. On 17 March 2021, Ms Tote watched the Teams interviews that had been undertaken as part of the investigation. Ms Tote wrote to SR requesting an interview, but that did not prove possible because SR was recovering from an illness.
29. On 26 March 2021, the claimant was invited to a reconvened investigation meeting with Ms Tote to be held on 22 April. At that meeting the claimant said:

I hope you can understand that I have been interviewed for 3 hours now and I still don't know what I have failed in. What code of conduct have I failed in? What professional standard have I failed in? Why can't somebody tell me what I have done wrong?

30. The Employment Tribunal held that there was "no substantive reply from Ms Tote to that question" despite the fact that she had by then reviewed all of the interviews.
31. A further meeting was arranged by Ms Tote for 7 May 2021 at which the claimant was to be told that there was no case to answer and the process was being brought to an end. The claimant did not attend the meeting. The outcome of the investigation was confirmed in writing the same day.
32. The claimant presented her ET1 Claim Form to the Employment Tribunal on 7 May 2021.
33. The Employment Tribunal referred to information provided after a subject access, which was supplemented at the hearing:

The only other Head of Service to be the subject of a disciplinary investigation was JSB who is of Asian origin. The only other person of a comparable grade to the Claimant against whom a disciplinary investigation was commissioned by Ms Lake was KR, who is also of Asian origin. Ms Lake has not commissioned any disciplinary investigations against any white employees of a comparable status.

The Issues in the Employment Tribunal

34. The issues were agreed and included the following factual issues:

Did the Respondent subject the Claimant to the following treatment:

1. Making false allegations against the Claimant on or around 12 January 2021;

2. Transferring the Claimant from her role as Head of Service;

3. Causing a disciplinary investigation to be carried out against the Claimant on or around 12 January 2021;

4. Causing the Claimant to have to attend several disciplinary investigation meetings only for the Claimant to be advised that there was no case to answer;

5. Not considering lesser and more proportionate means of dealing with any alleged work or conduct issues by the Claimant such as mediation on or around 12 January 2021.

The Employment Tribunal's direction as to the law

35. The Employment Tribunal directed itself to relevant provisions of the Equality Act 2010 ("EQA"):

13 Direct discrimination

(1) A person (A) discriminates against another (B) if, because of a protected characteristic, A treats B less favourably than A treats or would treat others.

23 Comparison by reference to circumstances

(1) On a comparison of cases for the purposes of section 13, 14, 19 or 19A there must be no material difference between the circumstances relating to each case.

136 Burden of proof

(1) This section applies to any proceedings relating to a contravention of this Act.

(2) If there are facts from which the court could decide, in the absence of any other explanation, that a person (A) contravened the provision concerned, the court must hold that the contravention occurred.

(3) But subsection (2) does not apply if A shows that A did not contravene the provision.

36. The Employment Tribunal cited Igen Ltd (Formerly Leeds Careers Guidance) and Others v Wong [2005] EWCA Civ 142, [2005] ICR 931 and Madarassy v Nomura International plc [2007] EWCA Civ 33; [2007] ICR 867:

63. In Madarassy v Nomura International Plc [2007] IRLR 246, the Court of Appeal (approving Igen) made it clear that the burden does not shift to the employer simply on the Claimant establishing a difference in status (for example a difference in race) and a difference in treatment. Such would only indicate a possibility of discrimination, not that there was in fact discrimination. "Could conclude" must mean that a reasonable Tribunal could properly conclude from all the evidence before it. Thus the first stage of the two-stage process envisaged by section 136 EA 2010 is to consider whether the Tribunal could properly conclude from the facts whether discrimination is a possible explanation for the treatment. At the second stage, once the Tribunal is satisfied that the Claimant has proved facts from which an inference of discrimination can be drawn, the Respondent must provide a non-discriminatory explanation for its treatment of the Claimant. If, on a balance of probabilities, the Respondent is not able to show that discrimination was not the reason for the treatment the Claimant must succeed. If the Respondent discharges the burden by proving, for example, that a non-discriminatory reason for the treatment exists, then the claim must fail.

The conclusion of the Employment Tribunal

37. The Employment Tribunal decided that the burden of proof had passed to the respondent. The primary reason was the disparity in treatment between the claimant and other white employees who might also have been subject to formal disciplinary investigation but were dealt with informally ("the disparity finding"):

69. Having considered the evidence we are satisfied that the Claimant has

established a prima facie case, that is to say the Claimant has proved facts from which an inference of discrimination might be drawn and the burden thus shifts to the Respondent to establish a non-discriminatory reason for the treatment.

70. We arrive at that conclusion for the reasons below but in essence it is because in a number of comparable situations where a disciplinary investigation might reasonably have been instigated Ms Lake chose not to do so when it involved employees of a different race to that of the Claimant. Instead her normal approach was to offer mediation or to deal with it informally by discussion. In the case of the Claimant however she decided to take much more drastic action and she did so after she had been accused of unconscious racial bias. In particular we take into account the following:

70.1 HM had admitted to swearing. It was undoubtedly inappropriate conduct and was viewed as such by Ms Lake at the time because she took the time to go and see HM personally.

70.2 HM sent an email which caused such consternation that it led to a collective grievance yet no disciplinary investigation was commenced against HM. Whilst the final outcome as to the distribution of work was agreed by Ms Lake the manner in which it was done clearly offended others to the extent that they felt a collective grievance was necessary. There was nothing equally serious against the Claimant.

70.3 JR made strong allegations against JD that she said chose to offer mediation rather than instigate an investigation.

70.4 Despite victimisation of SR being a central allegation against Mrs Parmar (by AE) Ms Lake never interviewed SR which suggests she could not have thought there was any substance in the allegation. We do not consider the reason for not interviewing SR was that Ms Lake had simply not got round to it. Ms Lake had completed all of the nine proposed interviews by 5 February 2021. SR was never on the list of people to be interviewed. The Claimant was interviewed after the others which suggests that Ms Lake had decided to complete all her interviews until she got to the Claimant. It would make no sense to leave a meeting with SR after the Claimant.

70.5 The issue involving AE and the matter which HM thought was JR and not the Claimant. Mrs Parmar had acted properly in seeking advice from JD.

38. The Employment Tribunal also noted ("the treatment of Asian employees finding"):

70.6 The only employees that Ms Lake has ever disciplined are of Asian ethnicity. … [emphasis added]

39. The Employment Tribunal considered an apparent concession made by the claimant ("the concession"):

71. Mr Livingston argues that the allegations against the Claimant were serious and it was right and proper that there should be an investigation. The fact that nothing came of them in the end was irrelevant it is only when the allegations were properly investigated that it could fairly be concluded that no action was necessary. He also points out that in cross examination the Claimant accepted that there were serious concerns or that these were serious allegations .

72. We will deal briefly with the concession. In fairness to the Claimant her concession was subject to the proviso that they were only serious "if they were correct" which she did not believe they were

73. The Claimant clearly did not accept that these were serious allegations at the relevant time because she had consistently complained that there was no substance to them and she did not understand what she was supposed to have done wrong. Furthermore, Ms Lake never informed the Claimant of the allegations by AE and SCC. The Claimant only discovered them during the disclosure process in this case. She was therefore hardly in a position to admit they were serious if she did not know anything about them. [emphasis added]

40. The Employment Tribunal analysed whether there was a proper basis for the claimant being subject to a formal disciplinary investigation ("the unfairness finding"):

74. The reality is that there was nothing of substance to start a disciplinary investigation. There is nothing to suggest that Senior Managers such as Ms Lake routinely initiate such investigations. Ms Lake must have appreciated there was nothing of substance because the wording of the allegations calling Mrs Parmar to an investigation did not set out any identifiable acts of misconduct. Ms Tote decided quite properly to discontinue the investigation once she considered the evidence. There is nothing to suggest Ms Lake would have discontinued it if she had not been removed from the process by Mr Samuels.

75. The e-mail sent by AE on 16 December raised general concerns about Locality West and not about the Claimant in particular. The two specific complaints were about another employee who was said to have spoken in an inappropriate or overly harsh way and who challenged a simple administrative request.

41. The Employment Tribunal returned to the disparity point:

76. We conclude on the evidence that when it came to assessing the merits of behaviour allegations against white employees such as HM, AE and JR, Ms Lake was slow to move to formal measures. In the case of the Claimant she moved fairly speedily to investigation and suspension for something which was either at the same or lower level of alleged misconduct. We are satisfied that race played a part in her decisions. There is no other credible explanation. [emphasis added]

42. Another matter that the Employment Tribunal took into account, to a limited extent, was failings in disclosure by the respondent ("the disclosure finding"):

77. We also draw adverse inferences from the failure by the Respondent to disclose relevant evidence. There was a conscious decision by the Respondent (or their legal team) not to disclose highly relevant evidence. It was clearly relevant because it was on the basis of such evidence that Ms Tote decided there was no substance in the allegations and decided as a consequence to terminate the investigation.

78. In fact it is not only the recordings of interviews that the Respondent has failed to disclose. Ms Lake said she made notes of witness interviews and may have shared these with HR. She may have typed them and saved them as a file. None of these notes or files have been disclosed. Miss Tote said she took notes from the investigation meetings which she kept for 6 months. The Claimant began proceedings on 7 May so the need to preserve such notes would have been obvious from the outset. Ms Tote's notes have not been disclosed. Ms Tote also interviewed OC and at least some interviews were recorded. Those recordings have not been disclosed. [emphasis added]

43. The Employment Tribunal considered the respondent's explanations ("the explanations")

79. We have considered the potential non-discriminatory explanations relied on by the Respondent. They are as follows:

79.1 That the disciplinary procedure was appropriate in order to speak to witnesses and gather evidence.

This argument is without substance. There was nothing preventing Ms Lake from making informal enquiries as she often did. Not every enquiry requires a formal investigation and suspension so that witnesses can be interviewed. Ms Lake was quite content to ask SCC to set out her concerns in writing.

79. 2 That Ms Lake spoke to HR and it was agreed that this was the appropriate action to take.

Ms Lake cannot hide behind HR's actions or advice. It was her managerial decision to instigate a disciplinary investigation and temporary suspension.

79.3 The disciplinary process was preferable to any potential grievance procedure because it did not rely on complainants to raise grievances.

This was never offered as a rationale at the time and in any event has no merit. It is somewhat bizarre to suggest that because no-one had lodged a grievance that this somehow made a disciplinary process appropriate.

79.4 That when Ms Lake had raised issues with the Claimant in the past they had not been acknowledged as valid concerns.

This is an argument that appears to have been developed in the course of Ms Lake's evidence. In cross-examination Ms Lake could only give three examples where the Claimant had failed to acknowledge past issues. One was in October 2020 when Ms Lake told the Claimant to think carefully about her tone of emails as they could come across as 'curt and dismissive'. The email in question was not included in the bundle so we were not able to assess it fully. Clearly it was not seen as significant otherwise it would not have been omitted from the bundle. In any event the Claimant had replied to say that she was happy to discuss this at the next 1:1 meeting and Ms Lake had written back to say: 'That's fine - I know we won't see eye to eye on the tone thing'. At the next meeting the issue was not even raised.

The second example relates to a discussion at a supervision meeting on 8 January in which Ms Lake had noted that she had seen emails from Locality West managers, including the Claimant, that she perceived to be 'rude and aggressive'. Those emails have not been included in the bundle either so the same observation applies.

The third relates to a supervision meeting on 3 March 2020 where it was noted that there were:

"wider concerns expressed by people who did not wish to formalise their concerns. . . . about how working with some colleagues in West makes them feel - anxious, berated, attacked... BP reflected with RL on the individuals involved and their style/approaches/strengths/areas that require support or challenge and how RL could develop the relationships with TLs, JR in particular."

This concerns discussions some 10 months before the Claimant was subjected to an investigation. It was therefore largely historical. Ms Lake could scarcely have been thinking of this in January 2021 when she was deciding whether to start a disciplinary investigation.

79.5 That given the potential misconduct an investigation was appropriate.

There was no potential misconduct in reality. The allegations have never been particularised.

79.6 That at the time Ms Lake had received statements and emails from SCC, AE and a conversation with AE about AE's intention to resign

The suggestion of AE resigning was clearly not serious nor do we find Ms Lake treated it as such. The matters relating to AE and SCC could not have been seen as serious because they were never communicated to the Claimant.

79.7 That there was potential targeting of SR and allegations of victimisation.

If Ms Lake had genuinely believed SR was being targeted it was at least necessary to interview him. SR had not complained of victimisation himself. [emphasis added]

44. The Employment Tribunal concluded that the respondent had not discharged the burden of proof:

80. We are satisfied in all of the circumstances the Respondent has not established, on a balance of probabilities, a non-discriminatory explanation for the treatment of the Claimant. [emphasis added]

45. The Employment Tribunal concluded that the claimant had been subject to race discrimination ("the discrimination finding"):

81. We are also satisfied that (in the same or similar circumstances involving a white employee or at any rate one who was not Asian) Ms Lake would not have initiated a disciplinary investigation or suspended an employee from their role as Head of Service. We are satisfied the Claimant has been treated less favourably because of her race. [emphasis added]

46. The Employment Tribunal summarised its conclusions on the agreed factual issues:

Issue 1

Making false allegations against the Claimant on or around 12 January 2021.

83. We do not find that the allegations were manufactured or fabricated. This allegation, as framed, is therefore dismissed.

Issues 2 and 3

Transferring the Claimant from her role as Head of Service and causing a disciplinary investigation to be carried out against the Claimant on or around 12 January 2021.

84. We are satisfied that the Claimant was treated less favourably than a hypothetical white comparator would have been (that is to say someone who was white British) in the same or similar circumstances for the reasons given above. These two allegations are therefore upheld.

Issue 4

Causing the Claimant to have to attend several disciplinary investigation meetings only for the Claimant to be advised that there was no case to answer.

85. The allegation is in relation to the disciplinary investigation interview conducted by Ms Lake on 19 February 2021, the first meeting with Ms Tote on 22 April 2021 and the meeting with Ms Tote on 7 May 2021.

86. The allegation in relation to the meetings on 19 February and 22 April are upheld for the reasons given above. Although Ms Tote took the latter meeting, and no allegation of discrimination is made against her, the decision to hold that meeting was really that of Ms Lake, the meeting having been postponed from its original intended date of 24 February when the Claimant fell ill. The second meeting was effectively a continuation of the first.

87. The allegation in relation to the meeting on 7 May is not upheld. The purpose of that meeting was simply to inform the Claimant that the investigation was to be discontinued. It is possible that the Claimant could have been informed prior to that date of discontinuance but there is nothing discriminatory about that. In any event it was not a 'disciplinary investigation' meeting.

Issue 5

Not considering lesser and more proportionate means of dealing with any alleged work or conduct issues by the Claimant such as mediation on or around 12 January 2021.

88. This allegation also succeeds for the reasons given above. [emphasis added]

The Appeal

47. The appeal challenges the decision of the Employment Tribunal on 11 grounds, including multiple perversity and reasons challenges. Section 3.8.1 of the EAT Practice Direction 2023 provides:

An error of law should be easy to identify in a few words. The experience of the Judges of the EAT over many years is that short and focussed grounds of appeal are usually more persuasive than a long one and, in general, the more grounds raised the more it suggests that none is a good one.

48. After considering the appropriate approach to such appeals, I will deal with each ground.

The test on appeal

49. Pursuant to section 21 of the Employment Tribunals Act 1996, an appeal lies to the Employment Appeal Tribunal ("EAT") only on a question of law.
50. In British Telecommunications Plc v Sheridan [1990] IRLR 27, the Master of the Rolls held:

"34. … Any court with the experience of the members of the Employment Appeal Tribunal, and in particular that of the industrial members, will in the nature of things from time to time find themselves disagreeing with or having grave doubts about the decisions of Industrial Tribunals. When that happens, they should proceed with great care. To start with, they do not have the benefit of seeing and hearing the witnesses, but, quite apart from that, Parliament has given the Employment Appeal Tribunal only a limited role. Its jurisdiction is limited to a consideration of questions of law.

35. On all questions of fact, the Industrial Tribunal is the final and only judge, and to that extent it is like an industrial jury. The Employment Appeal Tribunal can indeed interfere if it is satisfied that the Tribunal has misdirected itself as to the applicable law, or if there is no evidence to support a particular finding of fact, since the absence of evidence to support a finding of fact has always been regarded as a pure question of law. It can also interfere if the decision is perverse, in the sense explained by Lord Justice May in Neale v Hereford & Worcester County Council [1986] ICR 471 at 483." [emphasis added]

51. In Brent London Borough Council v Fuller [2011] ICR 806, Mummery LJ, held, at paragraphs 28 to 30:

"28. The appellate body, whether the Employment Appeal Tribunal or this court, must be on its guard against making the very same legal error as the tribunal stands accused of making. An error will occur if the appellate body substitutes its own subjective response to the employee's conduct. The appellate body will slip into a similar sort of error if it substitutes its own view of the reasonable employer's response for the view formed by the tribunal without committing error of law or reaching a perverse decision on that point.

29. Other danger zones are present in most appeals against tribunal decisions. As an appeal lies only on a question of law, the difference between legal questions and findings of fact and inferences is crucial. Appellate bodies learn more from experience than from precept or instruction how to spot the difference between a real question of law and a challenge to primary findings of fact dressed up as law.

30. Another teaching of experience is that, as with other tribunals and courts, there are occasions when a correct self-direction of law is stated by the tribunal, but then overlooked or misapplied at the point of decision. The tribunal judgment must be read carefully to see if it has in fact correctly applied the law which it said was applicable. The reading of an employment tribunal decision must not, however, be so fussy that it produces pernickety critiques. Over-analysis of the reasoning process; being hypercritical of the way in which the decision is written; focusing too much on particular passages or turns of phrase to the neglect of the decision read in the round: those are all appellate weaknesses to avoid." [emphasis added]

52. The correct approach to be adopted by the EAT was reiterated by the Court of Appeal in DPP Law Ltd v Greenberg [2021] IRLR 1016:

"57. The following principles, which I take to be well established by the authorities, govern the approach of an appellate tribunal or court to the reasons given by an employment tribunal:

(1) The decision of an employment tribunal must be read fairly and as a whole, without focusing merely on individual phrases or passages in isolation, and without being hypercritical. In Brent v Fuller [2011] ICR 806, Mummery LJ said at p. 813:

"The reading of an employment tribunal decision must not, however, be so fussy that it produces pernickety critiques. Over-analysis of the reasoning process; being hypercritical of the way in which a decision is written; focussing too much on particular passages or turns of phrase to the neglect of the decision read in the round: those are all appellate weaknesses to avoid."

This reflects a similar approach to arbitration awards under challenge: see the cases summarised by Teare J in Pace Shipping Co Ltd v Churchgate Nigeria Ltd (The "PACE") [2010] 1 Lloyds' Reports 183 at paragraph 15, including the oft-cited dictum of Bingham J in Zermalt Holdings SA v Nu-Life Upholstery Repairs Ltd [1985] 2 EGLR 14 that the courts do not approach awards "with a meticulous legal eye endeavouring to pick holes, inconsistencies and faults in awards with the object of upsetting or frustrating the process of arbitration". This approach has been referred to as the benevolent reading of awards, and applies equally to the benevolent reading of employment tribunal decisions.

(2) A tribunal is not required to identify all the evidence relied on in reaching its conclusions of fact. To impose such a requirement would put an intolerable burden on any fact finder. Nor is it required to express every step of its reasoning in any greater degree of detail than that necessary to be Meek compliant (Meek v Birmingham City Council [1987] IRLR 250). Expression of the findings and reasoning in terms which are as simple, clear and concise as possible is to be encouraged. In Meek, Bingham LJ quoted with approval what Donaldson LJ had said in UCATT v Brain [1981] I.C.R 542 at 551:

"Industrial tribunals' reasons are not intended to include a comprehensive and detailed analysis of the case, either in terms of fact or in law … their purpose remains what it has always been, which is to tell the parties in broad terms why they lose or, as the case may be, win. I think it would be a thousand pities if these reasons began to be subjected to a detailed analysis and appeals were to be brought based upon any such analysis. This, to my mind, is to misuse the purpose for which the reasons are given."

(3) It follows from (2) that it is not legitimate for an appellate court or tribunal to reason that a failure by an employment tribunal to refer to evidence means that it did not exist, or that a failure to refer to it means that it was not taken into account in reaching the conclusions expressed in the decision. What is out of sight in the language of the decision is not to be presumed to be non-existent or out of mind. As Waite J expressed it in RSPB v Croucher [1984] ICR 604 at 609-610:

"We have to remind ourselves also of the important principle that decisions are not to be scrutinised closely word by word, line by line, and that for clarity's and brevity's sake industrial tribunals are not to be expected to set out every factor and every piece of evidence that has weighed with them before reaching their decision; and it is for us to recall that what is out of sight in the language of a decision is not to be presumed necessarily to have been out of mind. It is our duty to assume in an industrial tribunal's favour that all the relevant evidence and all the relevant factors were in their minds, whether express reference to that appears in their final decision or not; and that has been well-established by the decisions of the Court of Appeal in Retarded Children's Aid Society Ltd v Day [1978] ICR 437 and in the recent decision in Varndell v Kearney & Trecker Marwin Ltd [1983] ICR 683."

58. Moreover, where a tribunal has correctly stated the legal principles to be applied, an appellate tribunal or court should, in my view, be slow to conclude that it has not applied those principles, and should generally do so only where it is clear from the language used that a different principle has been applied to the facts found. Tribunals sometimes make errors, having stated the principles correctly but slipping up in their application, as the case law demonstrates; but if the correct principles were in the tribunal's mind, as demonstrated by their being identified in the express terms of the decision, the tribunal can be expected to have been seeking faithfully to apply them, and to have done so unless the contrary is clear from the language of its decision. This presumption ought to be all the stronger where, as in the present case, the decision is by an experienced specialist tribunal applying very familiar principles whose application forms a significant part of its day to day judicial workload."

53. Perversity may be established where there is no evidence to support a finding of fact: see Piggott Brothers & Co Ltd v Jackson [1992] ICR 85 at 92D-G. However, where it is asserted that, on the basis of conflicting evidence, a perverse decision has been reached, the test to establish perversity is especially high, in that it must be concluded that no reasonable tribunal could have arrived at the decision reached: see Yeboah v Crofton [2002] IRLR 635 at [93]:

"Such an appeal ought only to succeed where an overwhelming case is made out that the employment tribunal reached a decision which no reasonable tribunal, on a proper appreciation of the evidence and the law, would have reached'. Even in cases where the Appeal Tribunal has 'grave doubts' about the decision of the Employment Tribunal it must proceed with 'great care.'"

54. The relevant authorities about the extent to which an Employment Tribunal is required to give reasons were recently summarised by Cavanagh J in Frame v The Governing Body of Llangiwg Primary School UKEAT/0320/19/AT:

47. The relevant principles can be summarised as follows:

(1) The duty to give reasons is a duty to give sufficient reasons so that the parties can understand why they had won or lost and so that the Appellate Tribunal/Court can understand why the Judge had reached the decision which s/he had reached;

(2) The scope of the obligation to give reasons depends on the nature of the case;

(3) There is no duty on a Judge, in giving his or her reasons, to deal with every argument presented by counsel in support of his case:

(4) The Judge must identify and record those matters which were critical to his decision. It is not possible to provide a template for this process. It need not involve a lengthy judgment;

(5) The judgment must have a coherent structure. The judgment must explain how the Judge got from his or her findings of fact to his or her conclusions;

(6) When giving reasons a Judge will often need to refer to a piece of evidence or to a submission which s/he has accepted or rejected. Provided that the reference is clear, it may be unnecessary to detail, or even summarise, the evidence or submission in question; and

(7) It is not acceptable to use a fine-tooth comb to comb through a set of reasons for hints of error or fragments of mistake, and try to assemble them into a case for oversetting the decision. Nor is it appropriate to use a similar process to try to save a patently deficient decision."

The Grounds of Appeal

55. Multiple grounds of appeal, inevitably with a myriad of sub- grounds, are an invitation not to see the wood for the trees. Before considering the grounds it is worth standing back for a moment to see the judgment as a whole; concentrating on the forest not the trees. The primary reason for the burden of proof shifting to the respondent was the disparity finding; that Ms Lake had not disciplined employees of other ethnicity than that of the claimant in similar circumstances. That was linked to the unfairness finding; that the claimant had been treated unfairly while those of a different race with who she compared her treatment had not: and the treatment of Asian employees finding; that the comparable employees Ms Lake had previously subject to disciplinary action were Asian. A further minor factor was the disclosure finding; that there were material shortcomings in the respondent's disclosure. The burden having shifted to the respondent, the explanations were rejected as proving that discrimination had not occurred. On the face of it, this was a paradigm application of the principles in Igen, although the respondent would have us believe otherwise.

Ground 1: Misapplication of the law - Failure to consider the burden of proof in relation to each allegation separately

56. The respondent relies on Essex County Council v Jarrett [2015] UKEAT 0045/15/041. In that case the claimant made 28 allegations of race discrimination. There were a number of perpetrators of acts, such as harassment, that appeared to be unrelated to the main complaint that the claimant should have been appointed to a post as head of employment. Langstaff J stated at paragraph 32:

We are of the view that it was an error to take a blanket approach, and to regard the burden of proof as shifted on all allegations. Although it may create a difficult task for some Tribunals who are faced with a very significant number of allegations to analyse each by asking in respect of each separately whether less favourable treatment by reference to an actual or hypothetical comparator has been established, whether detriment exists and whether in the light of those findings and others the burden of proof might be shifted, in practice we think that those problems can be overstated. Although each allegation must of course be looked at separately, since it forms a separate complaint, the fact that there are other allegations that are thought individually to be substantiated will give force to the suggestion in respect of the next allegation that it too is to be substantiated, and the next beyond that, such that there is such evidence as justifies the reversal of the burden of proof in respect of each of those allegations. Each allegation if substantiated necessarily has an effect upon each other allegation, just as each allegation if not substantiated may equally inform a decision on another though it will not be decisive of it. If the Tribunal had made it clear that it was adopting some such approach, we would have found it easier to accept it, but it did not. Its adoption of a blanket approach was unreasoned.

57. I do not accept that Jarrett is authority for the proposition that in every case in which there are a number of allegations the Employment Tribunal will necessarily err in law if the application of s136 EQA is not considered separately for each allegation. The decision that it was an error of law to adopt a blanket approach must be seen in the context of the case in which there were multiple allegations against a number of individuals that were not apparently linked.
58. I find support for this commonsense approach, in the decision in of Supperstone J in Commissioner of Police of the Metropolis v Maxwell [2013] EqLR 680, EAT in which he said of a case in which there were 120 allegations of direct discrimination and harassment:

There was no obligation on the tribunal to refer to the two-stage process again in relation to each and every complaint or indeed to adopt the two-stage process in relation to each and every complaint if the tribunal considered it was not appropriate to do so.

59. Similarly, I do not consider that Maxwell is authority for the proposition that it will never be an error of law to consider the burden of proof together for all of the allegations. There may be some circumstances in which a blanket approach is inappropriate, but others in which it is permissible. All depends on the facts of the case, including the nature and number of allegations and whether there are a number of alleged perpetrators. Save in the clearest of cases the EAT should be slow to decide that the Employment Tribunal, that was best placed to make that assessment, got it wrong.
60. This case is quite different to Jarrett. The key allegations related to the conduct of Ms Lake and might have been described compendiously as initiating and progressing an unwarranted formal disciplinary investigation. In the circumstances of this case, I can see no error of law in the Employment Tribunal considering the evidence overall in deciding whether the burden of proof had shifted so that the respondent was required to prove that Ms Lake's actions were not discriminatory.
61. It is asserted that, in the summary section, the Employment Tribunal did not expressly state that the treatment comprising issues 1-3 was less favourable treatment because of race. While that was not spelt out in the brief summary section, it is clear, from reading the conclusions of the Employment Tribunal as a whole, that was the decision.

Ground 2: Use of a mere difference in treatment to shift the burden of proof

62. This ground as advanced conflates two different points: that it is insufficient for a claimant to establish (1) a mere difference in status and difference of treatment and/or (2) mere unreasonable treatment, to shift the burden of proof.
63. The respondent relies on the statement of Mummery LJ in in Madarassy v Nomura International plc [2007] ICR 867:

The bare facts of a difference in status and a difference in treatment only indicate a possibility of discrimination. They are not, without more, sufficient material from which a tribunal "could conclude" that, on the balance of probabilities, the respondent had committed an unlawful act of discrimination.

64. It is important to note that Mummery LJ referred to the "bare facts of a difference in status and a difference in treatment". i.e. absent any other factors such as the extent to which the situations are similar and the specific nature of the treatment. The comparative exercise in discrimination claims was recently considered by Cavanagh J in Martin v St Francis Xavier Sixth Form College Board of Governors [2024] EAT 22, [2024] IRLR 472:

54. Section 13 of the EqA requires that two matters be established for there to be a finding of direct discrimination. The first is that there has been treatment of the claimant (A) which is less favourable than the treatment that was meted out, or would have been meted out, to a comparator (B). The second is that the less favourable treatment was on the ground of the protected characteristic. Whilst it is open to an Employment Tribunal to go straight to the second question, the "reason why" question (see Shamoon v Chief Constable of the Royal Ulster Constabulary [2003] UKHL 11; [2003] ICR 337, at paragraph 8, per Lord Nicholls of Birkenhead) it is common for a Tribunal to consider the "comparator" question, as the Tribunal did in the present case.

55. There are three potential types of comparator: an actual (or statutory) comparator, an evidential comparator, and a hypothetical (statutory) comparator.

56. An actual comparator exists when there is no material difference between the circumstances relating to the claimant's case and the comparator's case. Express statutory provision is made for such a comparator in section 23(1) of the EqA, which states,

"(1) On a comparison of cases for the purposes of section 13…. there must be no material difference between the circumstances relating to each case."

57. A comparison with an actual comparator (also known as a statutory comparator) may support or undermine a claimant's case.

58. However, it is clear that, even where the circumstances of a proposed comparator are not materially the same as those of the claimant, a Tribunal may take account of the way in which the respondent treated that person if there are some relevant similarities between their circumstances. A Tribunal may be assisted by seeing how unidentical, though not wholly dissimilar, comparators had been treated in relation to other individual cases. See Chief Constable of West Yorkshire Police v Vento [2001] IRLR 124 (EAT), at paragraph 7, per Lindsay J. Such comparators are known as evidential comparators.

59. Furthermore, a Tribunal may consider whether it is assisted by considering how a hypothetical comparator in a similar (i.e. not materially different) position to the claimant, but who does not have the protected characteristic, would have been treated. Such a hypothetical comparator will be a statutory comparator, for the purposes of section 23.

60. It should be borne in mind, however, that the purpose of a Tribunal's consideration of comparators is to use it as an evidential tool to see whether an inference of discrimination is justified. It is not an end in itself. This was made clear by Lord Scott of Foscote in Shamoon, in the course of a very helpful summary of the law relating to comparators, at paragraphs 107-110:

"107. There has been, in my opinion, some confusion about the part to be played by comparators in the reaching of a conclusion as to whether a case of article 3(1) discrimination - or for that matter a case of discrimination under section 1(1) of the Sex Discrimination Act 1975 , or under section 1(1) of the Race Relations Act 1976 , or under the comparable provision in any other anti-discrimination legislation-has been made out. Comparators come into play in two distinct and separate respects.

108. First, the statutory definition of what constitutes discrimination involves a comparison: "treats that other less favourably than he treats or would treat other persons". The comparison is between the treatment of the victim on the one hand and of a comparator on the other hand. The comparator may be actual ("treats") or may be hypothetical ("or would treat") but "must be such that the relevant circumstances in the one case are the same, or not materially different, in the other" (see article 7). If there is any material difference between the circumstances of the victim and the circumstances of the comparator, the statutory definition is not being applied. It is possible that, in a particular case, an actual comparator capable of constituting the statutory comparator can be found. But in most cases a suitable actual comparator will not be available and a hypothetical comparator will have to constitute the statutory comparator. In Khan's case [2001] ICR 1065 one of the questions was as to the circumstances that should be attributed to the statutory hypothetical comparator. It is important, in my opinion, to recognise that article 7 is describing the attributes that the article 3(1) comparator must possess.

109. But, secondly, comparators have a quite separate evidential role to play. Article 7 has nothing to do with this role. It is neither prescribing nor limiting the evidential comparators that may be adduced by either party. The victim who complains of discrimination must satisfy the fact-finding tribunal that, on a balance of probabilities, he or she has suffered discrimination falling within the statutory definition. This may be done by placing before the tribunal evidential material from which an inference can be drawn that the victim was treated less favourably than he or she would have been treated if he or she had not been a member of the protected class. Comparators, which for this purpose are bound to be actual comparators, may of course constitute such evidential material. But they are no more than tools which may or may not justify an inference of discrimination on the relevant prohibited ground, e g sex. The usefulness of the tool will, in any particular case, depend upon the extent to which the circumstances relating to the comparator are the same as the circumstances relating to the victim. The more significant the difference or differences the less cogent will be the case for drawing the requisite inference. But the fact that a particular chosen comparator cannot, because of material differences, qualify as the statutory comparator, e g, under article 7 , by no means disqualifies it from an evidential role. It may, in conjunction with other material, justify the tribunal in drawing the inference that the victim was treated less favourably than she would have been treated if she had been the article 7 comparator.

110. In summary, the comparator required for the purpose of the statutory definition of discrimination must be a comparator in the same position in all material respects as the victim save only that he, or she, is not a member of the protected class. But the comparators that can be of evidential value, sometimes determinative of the case, are not so circumscribed. Their evidential value will, however, be variable and will inevitably be weakened by material differences between the circumstances relating to them and the circumstances of the victim." Emphasis added

61. We mention in passing that, at paragraph 109 of the passage cited above, Lord Scott was using the phrase "actual comparator" in a different way from the way in which the Tribunal in the present case, and we, have been using it. Lord Scott used the phrase "actual comparator" to mean a real person, as opposed to a wholly hypothetical person, whether that real person satisfied the statutory test in s23 or was an evidential comparator. The Tribunal and the Appeal Tribunal have been using the phase "actual comparator" to mean a real person who is a statutory comparator for the purposes of s23, and the phrase "evidential comparator" to mean a real person who is not a statutory comparator. Nothing rests on this difference of terminology. We should add that a hypothetical comparator will be a statutory comparator, because of the words "or may treat" in section 23 (see, Shamoon, at paragraph 108): when looking at a hypothetical comparators, a tribunal will be looking at someone who is in the identical circumstances as the claimant, but who does not possess the protected characteristic.

62. For practical purposes, it may not make a great deal of difference as to whether a comparator is an actual comparator or an evidential comparator. In Watt (formerly Carter) v Ahsan [2007] UKHL 51; [2008] 1 AC 296, Lord Hoffman, with whom all of the other Lords agreed, said:

"37. It is probably uncommon to find a real person who qualifies under section 3(4) as a statutory comparator. Lord Rodger's example at para 139 of Shamoon of the two employees with similar disciplinary records who are found drinking together in working time has a factual simplicity which may be rare in ordinary life. At any rate, the question of whether the differences between the circumstances of the complainant and those of the putative statutory comparator are "materially different" is often likely to be disputed. In most cases, however, it will be unnecessary for the tribunal to resolve this dispute because it should be able, by treating the putative comparator as an evidential comparator, and having due regard to the alleged differences in circumstances and other evidence, to form a view on how the employer would have treated a hypothetical person who was a true statutory comparator. If the tribunal is able to conclude that the respondent would have treated such a person more favourably on racial grounds, it would be well advised to avoid deciding whether any actual person was a statutory comparator."

63. The question, in direct discrimination cases, as to whether the situations of the claimant, on the one hand, and the proposed comparator, whether actual or evidential, on the other, are comparable is a question of fact and degree: Hewage v Grampian Health Board [2012] UKSC 37; [2012] ICR 1034. The Supreme Court upheld the view of the Inner House of the Court of Session, restoring the decision of the Employment Tribunal, that unless the Employment Tribunal's judgment could be said to be absurd or perverse it was not for the Appeal Tribunal to impose its own judgment on the point. To like effect, in Kalu v Brighton & Sussex University Hospitals NHS Trust (UKEAT/0609/12), Langstaff P said, at paragraph 24, that the identification of a comparator is a question of fact.

64. In order for a comparator to be an actual or statutory comparator, is not necessary that the circumstances are the same in every particular. In Vento, above, Lindsay J said, at paragraph 12:

"…. it is all too easy to become nit-picking and pedantic in the approach to comparators. It is not required that a minutely exact actual comparator has to be found."

65. In Kalu, at paragraph 24, Langstaff P said, "The purpose of making the comparison … needs to be understood before a comparator may properly be identified." In our judgment, this is of central importance. Whether a point of difference has any significance or not depends on the nature of the less favourable treatment about which complaint is made. So, for example, if the complaint is about the claimant not being selected for a job, whilst the comparator was selected, the fact that the claimant and comparator have similar academic qualifications may well be relevant if the job required developed intellectual skills, but it is not relevant if the job requires solely manual labour or (to use one of Langstaff P's examples) is to model clothing.

65. Cavanagh J approved my analysis of the interrelationship between the use of comparators and the shifting burden of proof, while noting that I did not state that "in every case in which the claimant has been treated less favourably than an actual (statutory) comparator, the burden of proof will shift":

67. The interrelationship between the use of comparators and the shifting burden of proof has recently helpfully been considered by the EAT (HHJ Tayler) in Virgin Active v Hughes [2023] EAT 130; [2024] IRLR 4, at paragraphs 61-69,

"61. In many direct discrimination claims the claimant does not rely on a comparison between his treatment and that of another person. The claimant relies on other types of evidence from which it is contended that an inference of discrimination should be drawn, the comparison being with how the claimant would have been treated had he had some other protected characteristic.

62. In other cases, the claimant compares his treatment with that of one or more other people. There are two ways in which such a comparison may be relevant. If there are no material differences between the circumstances of the claimant and the person with whom the comparison is made (the person is usually referred to as an actual comparator), this provides significant evidence that there could have been discrimination. However, because there must be no material difference in circumstances between a claimant and a comparator for the purpose of section 23 EQA it is rare that a claimant can point to an actual comparator. The second situation in which a comparison with the treatment of another person may provide evidence of discrimination is where the circumstances are similar, but not sufficiently alike for the person to be an actual comparator. The treatment of such a person may provide evidence that supports the drawing of an inference of discrimination, sometimes by helping to consider how a hypothetical person whose circumstances did not materially differ to those of the claimant would have been treated (generally referred to as a hypothetical comparator). Evidence of the treatment of a person whose circumstances materially differ to those of the claimant is inherently less persuasive than that of a person whose circumstances do not materially differ to those of the claimant. That distinction is not always sufficiently considered when applying the burden of proof provisions in section 136 EQA: …

63. Probably the most regularly quoted passage concerning section 136 EQA is from the judgment of Mummery LJ in Madarassy v Nomura International plc [2007] ICR 867 at paragraph 56:

"56. The court in Igen Ltd v Wong [2005] ICR 931 expressly rejected the argument that it was sufficient for the complainant simply to prove facts from which the tribunal could conclude that the respondent "could have" committed an unlawful act of discrimination. The bare facts of a difference in status and a difference in treatment only indicate a possibility of discrimination. They are not, without more, sufficient material from which a tribunal "could conclude" that, on the balance of probabilities, the respondent had committed an unlawful act of discrimination."

64. It is worth noting that in Madarassy the Employment Tribunal did not analyse the treatment of the claimant in comparison to actual comparators. Ms Madarassy's claim was not analysed on the basis that there were men who were actual comparators, but that the scoring of men in a redundancy exercise could help establish how a hypothetical comparator would have been treated.

65. Where there is an actual comparator, it might be said that there is more than the bare fact of a difference of status and a difference of treatment. In Laing v Manchester City Council and another [2006] ICR 1519 Elias J noted:

"65. In our view, if one considers the burden of proof provision in the context of what a claimant needs to establish in a discrimination claim, what it envisages is that the onus lies on the employee to show potentially less favourable treatment from which an inference of discrimination could properly be drawn. Typically this will involve identifying an actual comparator treated differently or, in the absence of such a comparator, a hypothetical one who would have been treated more favourably. That involves a consideration of all material facts (as opposed to any explanation)."

66. Laing was approved by the Court of Appeal in Madarassy, which itself was approved by the Supreme Court in Hewage v Grampian Health Board [2012] UKSC 37; [2012] ICR 1034 and Efobi v Royal Mail Group Ltd [2017] UKEAT 0203/16, [2018] ICR 359 ; see the analysis of Underhill LJ in Base Childrenswear Limited v Nadia Otshudi [2019] EWCA Civ 1648 at paragraphs 16-18 .

67. If anything more is required to shift the burden of proof when there is an actual comparator it will be less than would be the case if a claimant compares his treatment with a person whose circumstances are similar, but materially different, so that there is not an actual comparator.

68. For example, if two people who differ in a protected characteristic attend a job interview and one is appointed but the other is not, that, of itself, would not be enough to shift the burden of proof, but if they scored the same marks in the assessment, so there is an actual comparator, the difference of treatment would seem to call out for an explanation. As Elias J noted in Laing at paragraph 73:

"As I said in Network Rail Infrastructure Ltd v GriffithsHenry (unreported) 23 May 2006, para 17, it may be legitimate to infer that a black person may have been discriminated against on grounds of race if he is equally qualified for a post which is given to a white person and there are only two candidates, but not necessarily legitimate to do so if there are many candidates and a substantial number of other white persons are also rejected."

69. Accordingly, where a claimant compares his treatment with that of another person, it is important to consider whether that other person is an actual comparator or not. To do this the Employment Tribunal must consider whether there are material differences between the claimant and the person with whom the claimant compares his treatment. The greater the differences between their situations the less likely it is that the difference of treatment suggests discrimination."

68. We respectfully agree with this analysis, but it is important to emphasise that HHJ Tayler did not say that in every case in which the claimant has been treated less favourably than an actual (statutory) comparator, the burden of proof will shift. It is more likely that the burden of proof will shift, but it does not follow that in every case the burden of proof will shift. The Tribunal must apply the statutory test as set out in section 136(1), when deciding, in a particular case, whether the burden of proof has passed to the respondent.

66. I have quoted these extensive passages to emphasise that comparing the treatment of a claimant with that of another person is a subtle business. The analysis is highly context specific. Where such a comparison is made, as part of an analysis of a range of relevant factors, it is not valid to pick apart small components of the comparative analysis, and to trot out the well-worn phrase that there is nothing more than a mere difference of status and treatment, while ignoring all of the other relevant findings of the tribunal that contributed to the overall analysis.
67. The respondent also contends that reliance should not have been placed on mere unfair treatment. The respondent relies on another well-known passage that is frequently quoted out of context. I will set out a little more of the passage from Glasgow City Council v Zafar [1997] 1 WLR 1659 than is usually quoted (the passage often quoted in italics, the context in bold):

The reasoning of the industrial tribunal on this issue is wholly defective. The Act of 1976 requires it to be shown that the complainant has been treated by the person against whom the discrimination is alleged less favourably than that person treats or would have treated another. In deciding that issue, the conduct of a hypothetical reasonable employer is irrelevant. The alleged discriminator may or may not be a reasonable employer. If he is not a reasonable employer he might well have treated another employee in just the same unsatisfactory way as he treated the complainant in which case he would not have treated the complainant "less favourably" for the purposes of the Act of 1976. The fact that, for the purposes of the law of unfair dismissal, an employer has acted unreasonably casts no light whatsoever on the question whether he has treated the employee "less favourably" for the purposes of the Act of 1976.

I cannot improve on the reasoning of Lord Morison, delivering the opinion of the court, who expressed the position as follows, 1997 S.L.T. 281, 284:

"The requirement necessary to establish less favourable treatment which is laid down by section 1(1) of the Act of 1976 is not one of less favourable treatment than that which would have been accorded by a reasonable employer in the same circumstances, but of less favourable treatment than that which had been or would have been accorded by the same employer in the same circumstances. It cannot be inferred, let alone presumed, only from the fact that an employer has acted unreasonably towards one employee, that he would have acted reasonably if he had been dealing with another in the same circumstances."

68. The passage refers to circumstances in which the only evidence is of unfair treatment, not to where there is unfair treatment of a person with one protected characteristic but not of a person with a different protected characteristic or, indeed, where there is reason to believe that the employer would normally act reasonably.
69. The respondent also relied on the statement Elias J (P) held stated in Bahl v The Law Society [2003] IRLR 640:

"The significance of the fact that the treatment is unreasonable is that a tribunal will more readily in practice reject the explanation given than it would if the treatment were reasonable. In short, it goes to credibility. If the tribunal does not accept the reason given by the alleged discriminator, it may be open to it to infer discrimination But it will depend upon why it has rejected the reason that he has given, and whether the primary facts it finds provide another and cogent explanation for the conduct. Persons who have not in fact discriminated on the proscribed grounds may nonetheless sometimes give a false reason for the behaviour. They may rightly consider, for example, that the true reason casts them in a less favourable light, perhaps because it discloses incompetence or insensitivity. If the findings of the tribunal suggest that there is such an explanation, then the fact that the alleged discriminator has been less than frank in the witness box when giving evidence will provide little, if any, evidence to support a finding of unlawful discrimination itself."

70. Elias J stated that the fact that treatment is unfair may go to credibility. If there is no reason to believe that the respondent would generally act unfairly that may be relevant to shifting the burden of proof.
71. The extracts from the authorities relied upon by the respondent must be seen in their proper context. It is also important not to salami slice a judgment into multiple components, each of which is individually assessed, out of context, against the criteria of whether there is no more than a difference in status and difference of treatment and/or no more than an allegation of unreasonable conduct. If there are multiple examples of different treatment between those of different status, and of unfair treatment, it is unlikely to be a case where it can be said that there is mere difference of status and treatment and/or mere unfair treatment.
72. The primary argument in Ground 2 of the grounds of appeal is set out at paragraphs 6 and 7 of Mr Allen's skeleton argument:

6. At ¶70 [13-14], the ET set out that the "essence" of the reason why they concluded that the burden of proof shifted was that Ms Lake decided to take much more drastic action against the Claimant than she did in a number of comparable situations where a disciplinary investigation might reasonably have been instigated (and which involved employees of a different race to that of the Claimant). Similar reasoning can be found in the ET's analysis at ¶76 [15].

7. This was an error of law in that the core reason for finding that there was a prima facie case of discrimination was a difference in treatment between the Claimant and others.

73. The core reasoning of the Employment Tribunal is the opposite of a mere difference in status and difference of treatment. A number of employees of different race to the claimant have not been subject of formal disciplinary proceedings in circumstances similar to those in which the claimant was. The similarity of the circumstances, and the fact that a number of employees of different race have been treated more favourably, obviously establishes more than a mere difference of treatment and status. If what the Employment Tribunal found is not evidence that could support a claim of race discrimination it is hard to imagine what is. It was the totality of the evidence that resulted in the shift in the burden of proof.
74. The respondent makes a number of submissions about findings of the Employment Tribunal that might be asserted to be "something more" than a mere difference in status and difference of treatment (I set them out below followed by my analysis):

¶70 … notes that the treatment of the Claimant by Ms Lake came after Ms Lake had been accused of unconscious racial bias (see also ¶19 …). It cannot be said that such an allegation, made two years prior to the treatment complained of, amounted to a fact from which race discrimination can be inferred. There was no contemporaneous finding that Ms Lake did in fact show unconscious racial bias and the ET made no such finding.

75. This was a very minor element of the reasoning of the Employment Tribunal. While I accept that the finding was of little, if any, assistance in the analysis in the absence of a complaint of victimisation, I do not accept that the reference to this point, in amongst all of the other reasons for the shift in the burden of proof, undermines the reasoning of the Employment Tribunal to a material extent.

¶70.1-70.3 … simply set out the difference in treatment between the way in which the Claimant was treated and the way in which HM was treated (twice) and JR was treated. This is, at most, a mere difference in treatment, and is not something from which race discrimination can be inferred.

76. It simply cannot be said that this is "a mere difference in treatment" it is a difference of treatment of a person of a different race in similar circumstances. It was also just one component of a multi-faceted decision.

¶70.4… sets out the ET's disbelief of the Respondent's asserted reason for Ms Lake not interviewing SR. As discussed below under Ground 5, this was an error of law in itself. Regardless:

i) the ET did not find that Ms Lake was untruthful in this regard (she was not asked about this in cross-examination and therefore did not provide evidence on this topic) – therefore this cannot be said to amount to a fact from which race discrimination by Ms Lake can be inferred.

ii) Even if it had found that this was a false reason given by an alleged discriminator, this provides little, if any, evidence to support a finding of unlawful discrimination itself (see Bahl).

77. This argument again separates out one minor component in the overall analysis that the burden of proof had shifted to the respondent. I do not consider that the Employment Tribunal erred in law in taking into account the fact that Ms Lake did not provide a credible explanation for her failure to interview SR. Even if there was an error in taking this factor into account, it is clear that the burden would have shifted to the respondent in any event. This point is considered further below as it was also raised as a freestanding ground of appeal.

70.5 … sets out that the ET thinks that the Claimant acted properly. At most, this amounts to a finding that the decision to investigate her was unreasonable. Unreasonable treatment is not sufficient as a basis for an inference of discrimination (see Zafar, Bowler, and Efobi).

78. This argument once again ignores the context; the Employment Tribunal found that the claimant was subject to formal disciplinary investigation when others of a different race were not in similar circumstances; this is obviously not a mere difference in status and difference of treatment.

¶70.6 …, in which the ET noted that the "only employees that Ms Lake has ever disciplined are of Asian ethnicity" is not a fact from which an inference of discrimination can be drawn.

i) This does not reflect a sufficiently large number of cases (one or two individuals other than the Claimant since 2017) for it to even come close to amounting to "'statistical' evidence that may tend to show a discernible pattern of treatment by the employer to the Claimant's racial group from which a Tribunal might infer unlawful discrimination" (see Home Office v Kuranchie).

ii) As below at Ground 6, this mischaracterises the ethnic/national origin relied upon by the Claimant as part of this claim.

79. There is no proper basis for the contention that only treatment of other employees that involves a sample of sufficient size that it could be statistically relevant can be taken into account in claims of discrimination. Were that the case comparative evidence could rarely be relied on. Kuranchie refers to statistical evidence about the treatment by "the employer" of those of a claimant's racial group: i.e. evidence that may involve the treatment of a number of employees by a number of individual discriminators who work for the employer; rather than treatment afforded by one employee, such as Ms Lake. The fact that Ms Lake subjected one or more Asian employees to formal disciplinary proceedings, whereas she had not disciplined white employees in similar circumstances, was clearly a factor that the Employment Tribunal was entitled to take into account.

¶71-74 … sets out the ET's view on whether the Claimant should be held to her concession that the allegations were serious and then sets out the ET's view that there was nothing of substance to commence a disciplinary. At most, this is a finding that the explanation for certain treatment was "inadequate, unreasonable or unjustified" (see Bowler) and does not by itself amount to a fact from which race discrimination can be inferred.

80. The question of whether the allegations were serious is not, in reality, something that the claimant could make a relevant "concession" about. What the claimant thought about the allegation was not the issue. It was open to the Employment Tribunal to consider whether Ms Lake genuinely thought that the allegations were sufficiently serious for the claimant to be called to a formal disciplinary investigation meeting. The Employment Tribunal permissibly held that "Ms Lake must have appreciated there was nothing of substance because the wording of the allegations calling Mrs Parmar to an investigation did not set out any identifiable acts of misconduct." This clearly was a factor that the Employment Tribunal was entitled to take into account as part of its overall analysis.

15. Finally, at ¶76, the ET reiterated that Ms Lake's treatment of the Claimant was different to her treatment of HM, AE and JR … . This is, again, a mere difference in treatment without 'something more'. The ET then concludes that they are satisfied that race played a part in Ms Lake's conclusions as "There is no other credible explanation". This is an error of law in itself, compounding the ET's erroneous approach to the issue:

a. It suggests that the ET has analysed and rejected the Respondent's explanations for the Claimant's treatment at this stage of the burden of proof test, which would be erroneous.

b. Even if it was open to the ET to conclude at this stage that there was no other credible explanation for the Claimant's treatment, it is not a fact from which race discrimination can be inferred. It is a considerable leap from a finding that the Respondent's explanation was inadequate, to the positive finding that race was a reason.

c. It involves an implied finding that race playing a part in Ms Lake's decision was a credible explanation. However, there is no analysis or explanation of why that was a credible explanation for how Ms Lake acted.

81. For the reasons set out above the Employment Tribunal was entitled to compare the treatment of the claimant to HM, AE and JR. There obviously was more than a mere difference of treatment and status. This was part of the overall analysis. The criticism of the Employment Tribunal for stating that there was "no other credible explanation" is pernickety in the extreme. The Employment Tribunal may have referred to its conclusion slightly out of place, but it is clear that it did consider the respondent's explanations and rejected them. The overall analysis of the Employment Tribunal shows why they considered that a "credible explanation" of Ms Lake's conduct was that race played a part in her decision.

Ground 3: Failure to provide adequate reasons regarding the facts from which it concluded it could infer that treatment was because of race

82. The respondent asserts that:

16. As an overarching point, the ET singularly fails to identify what the "facts" are, from which it could conclude that Ms Lake's treatment of the Claimant was because of her race.

19. In this case, the ET failed to identify the fact or feature capable of supporting an inference of discrimination; and has certainly failed to set out the primary facts from which any inference of discrimination is drawn 'with the utmost clarity'. As a result, the Respondent has had to guess what those features might have been and, as set out above (at ¶14-15 of this skeleton argument), none of these actually amount to a fact or feature capable of supporting an inference of discrimination.

20. The ET therefore failed to provide adequate reasons for its decision.

83. Were the reasons of the Employment Tribunal so lacking, it is hard to understand how the respondent managed to find 11 grounds of appeal to challenge them. The extracts of the Employment Tribunal set out above clearly show the facts found by the Employment Tribunal and how it concluded that the burden of proof had shifted to the respondent. This ground adds nothing to the previous ground.

Ground 4: Misapplication of the law - Erroneous approach to comparators

84. This ground contends that the Employment Tribunal conflated "comparable situations" with an actual or hypothetical comparator". On a fair reading of the Judgment as a whole I consider it is clear that the Employment Tribunal considered that there were evidential comparators who assisted in the process of drawing inferences.
85. The respondent contends that in every case in which there is no actual comparator the Employment Tribunal must expressly construct a hypothetical comparator. Mr Allen relied on a passage from an unreported decision of the EAT, The Chief Constable of Cambridgeshire Constabulary v McLachlan [2003] EAT/0562/02/RN in which, after having rejected the proposed actual comparators, Rimer J stated:

13 We infer, therefore, that there was no helpful actual comparator. This being so, we consider that the Tribunal had a duty to construct a hypothetical comparator against which they could test Mrs McLachlan's case. Their failure to do so involved an error of law because it goes to the manner in which the Tribunal is to approach the case (see Balamoody -v- The United Kingdom Central Council for Nursing, Midwifery and Health Visiting [2001] EWCA Civ 2097 [2002] IRLR 288 at paragraphs 51 to 60 in the judgment of Lord Justice Ward). The obvious hypothetical comparator for the purposes of the exercise before the Tribunal in this particular case was a male part-timer with small children to care for and who would, or might, therefore, be as much of a clock-watcher as Mrs McLachlan, and had disobeyed an order so that he could go home on time in order to look after his child.

86. It is clear that Rimer J was referring to the particular facts of that case, which were somewhat complex. He did not set out a general principle that in every case in which there is not an actual comparator the Employment Tribunal must set out the characteristics of a hypothetical comparator in elaborate detail. That would be wholly incompatible with the approach to comparators set out in the decisions referred to in Martin, particularly Hewage. As Cavanagh J makes clear in Martin the purpose of a Tribunal's consideration of comparators is to use it as an evidential tool to see whether an inference of discrimination is justified. A one size fits all analysis is not required.
87. The Employment Tribunal clearly considered the hypothetical treatment of a white employee in comparable circumstances:

81. We are also satisfied that (in the same or similar circumstances involving a white employee or at any rate one who was not Asian) Ms Lake would not have initiated a disciplinary investigation or suspended an employee from their role as Head of Service. We are satisfied the Claimant has been treated less favourably because of her race.

88. When the Employment Tribunal summarised its decision on Issues 2 and 3 it stated:

84. We are satisfied that the Claimant was treated less favourably than a hypothetical white comparator would have been (that is to say someone who was white British) in the same or similar circumstances for the reasons given above. These two allegations are therefore upheld.

89. The fact that the Employment Tribunal did not repeat that terminology when it summarised its conclusion on Issues 4 and 5 does not mean that it did not have the comparison in mind

Ground 5: Findings of fact for which there was no evidence in support

90. The respondent contends:

At ¶70.4 …, the ET concluded that Ms Lake never interviewed SR "which suggests she could not have thought there was any substance in the allegation…We do not consider the reason for not interviewing SR was that Ms Lake had simply not got round to it". Similarly, at ¶79.7 [16], when analysing the Respondent's explanation for treatment in relation to the potential targeting of SR and allegations of victimisation, the ET relied on the fact that Ms Lake did not interview SR and uses it to impliedly conclude that Ms Lake did not genuinely believe SR was being targeted.

91. This is a repetition of one of the sub-grounds of Ground 2. For the reasons set out above, I reject it. It is not always an error of law for an Employment Tribunal to make a finding rejecting the evidence of a witness who has not been specifically challenged about the matter in cross-examination. It is always a matter of substantive fairness. Ms Lake knew there was an issue about why she did not interview SR and gave evidence about her reasoning. I do not consider that the Employment Tribunal was bound to accept her explanation even if it was not the subject of cross-examination. In any event, even if there was an error in this regard, I do not consider that this was a major factor in the decision of the Employment Tribunal that the burden of proof had shifted to the respondent. It would have shifted absent this finding. The core reasoning for the shift in the burden of proof was the disparity finding, linked with the unfairness finding.

Ground 6: Misapplication of the law/material error - Erroneously describing the Claimant's ethnic/national origin

92. It is correct that the claimant described herself of as being a British national of Indian origin. I do not consider there was any error of law in the Employment Tribunal seeing this as being a subset of being of Asian ethnicity. In any event, when considering evidential comparators the Employment Tribunal is not limited to considering only those who describe their race in precisely the same terms as the claimant.

Ground 7: Misapplication of the law - Erroneous approach to drawing adverse inferences from a failure to disclose relevant evidence

93. The Respondent relies on D'Silva v NATFHE [2008] IRLR 412, in which Underhill J stated:

"…we have observed a tendency in discrimination cases for Respondents' failures in answering a questionnaire, or otherwise in providing information or documents, to be relied on by Claimants, and even sometimes by Tribunals, as automatically raising a presumption of discrimination. That is not the correct approach. Although failures of this kind are specified at item (7) of the "Barton guidelines" as endorsed in Igen Ltd. v. Wong [2005] ICR 931 (see at p. 957 B) as matters from which an inference can be drawn, that is only "in appropriate cases"; and the drawing of inferences from such failures — as indeed from anything else — is not a tick-box exercise. It is necessary in each case to consider whether in the particular circumstances of that case the failure in question is capable of constituting evidence supporting the inference that the respondent acted discriminatorily in the manner alleged; and if so whether in the light of any explanation supplied it does in fact justify that inference. There will be many cases where it should be clear from the start, or soon becomes evident, that any alleged failure of this kind, however reprehensible, can have no bearing on the reason why the Respondents did the act complained of, which in cases of direct discrimination is what the Tribunal has to decide. In such cases time and money should not be spent pursuing the point."

94. Underhill J, warned against a failure in providing information or documentation "automatically raising a presumption of discrimination". That was not the case in this judgment. It was a minor factor, amongst many other, that resulted in the burden shifting. I do not consider that the Employment Tribunal erred in law in taking it into account, to the limited extent that it did so. The burden would have shifted absent this component of the analysis.

Ground 8: Misapplication of the law – whether there was a non-discriminatory explanation for the treatment of the Claimant

95. This is a dressed up perversity ground. It is clear on a fair reading of the judgment that the Employment Tribunal did consider whether there was a non-discriminatory reason for the treatment of the claimant in submitting her to formal disciplinary investigation and found as a fact that the respondent had failed to establish that was the case. The respondent comes nowhere near to surmounting the high threshold for establishing perversity.

Ground 9: Failure to take into account relevant factors – Respondent's explanations for treatment

96. The Employment Tribunal clearly did consider the respondent's explanation for the treatment of the claimant, but rejected it. The Employment Tribunal did not accept that the real reason for the instigation of a formal disciplinary investigation was the claimant's conduct. The Employment Tribunal did not disregard the concerns that had been raised, but concluded that if the claimant had been white they would have been dealt with informally.

Ground 10: Conclusion for which there is no evidential basis/which is perverse

97. This is yet another retread of the argument that it was perverse of the Employment Tribunal not to accept that the allegations against the claimant required a formal disciplinary investigation. I do not consider the respondent comes close to establishing perversity.

Ground 11: Conclusion for which there is no evidential basis/perversity/insufficient reasons (re issue 4)

98. The respondent seeks to challenge the decision that the invitation of the claimant to disciplinary investigatory meetings was an act of race discrimination. The respondent comes nowhere near to establishing perversity. The reasoning of the Employment Tribunal was more than adequate to explain why the burden of proof shifted to the respondent and why the respondent failed to discharge the burden upon it.

Overall

99. I regret that this judgment has ended up being so long; I have not had the time to make it any shorter. If I had, I might merely have said that this is just the type of appeal that the Court of Appeal warned against in Greenberg.