NSL Ltd v P Zaluski

I will refer to the parties as they were in the employment tribunal, as claimant and respondent. At a full merits hearing at London South before EJ Corrigan, Dr N Westwood and Mrs S Dengate the claimant succeeded in complaints of indirect discrimination and harassment related to race. There followed a remedy hearing resulting in awards of compensation for injury to feelings, aggravated damages and interest, and the making of recommendations. The respondent appeals in respect of the decision upholding the successful complaints and the remedy decision.

The respondent was represented at the liability and remedy hearings by (different) counsel. The claimant represented himself at both hearings. The respondent was represented before me by Ms Seymour leading Ms Grace of counsel (neither of whom appeared in the tribunal), and they co-authored its skeleton. For convenience, and with no disrespect, I will refer to the submissions of Ms Seymour. Mr Platts-Mills of counsel appeared for the claimant under the auspices of Advocate.

I draw the following factual summary from the tribunal’s decisions and copies of emails that were before the tribunal that are also in my bundles.

The respondent provides, among other things, parking and transport services. It has 4000 staff across 250 locations nationally. The claimant has been employed by it as a Civil Enforcement Officer since September 2013. He is of Polish nationality and national origin. He works on the respondent’s contract with the London Borough of Wandsworth in respect of enforcement of parking regulations.

The claimant had a period of unauthorised leave in January 2020 when he went to Poland following the death of his mother-in-law. Upon his return no action was taken as it was recorded that he had “a very good excuse this time”. The claimant requested three-weeks’ leave to go to Poland in August 2020, but he was allowed only the first and third weeks. In the event, on account of sickness, he did not return to the UK until November 2020 and there was then a 14-day mandatory quarantine period. Mark Shaw, the Client Account Manager, suspected that the claimant had pre-planned to take the leave anyhow, and referred the matter to disciplinary process. However, the disciplinary manager, having considered the medical evidence, accepted that this was an unplanned absence and considered that it was not a disciplinary issue. The claimant accordingly received no sanction on that occasion.

The tribunal found at [14]:

That last observation referred to the fact that the claimant’s father died on 26 December 2020 and, as an only child, and in view of his mother’s age, he needed to go to Poland to organise the funeral and attend to his father’s affairs. In an email of 15 February 2021 he requested three weeks’ leave, from 21 February to 14 March 2021.

Mr Shaw refused that request, citing operational circumstances and the need to consider that there would be a two-week quarantine period upon the claimant’s return to the UK. He granted annual and compassionate leave covering from 22 February until a return to work on 3 March 2021. He provided flight details and suggested organising the funeral via video link or email. He wrote that any unauthorised absence would be dealt with through disciplinary process. The tribunal observed that this letter was unclear, or ambiguous, as Mr Shaw was aware of the two-week quarantine period upon return to the UK, but also required the claimant to be back at work on 3 March, which was clearly not practical if Mr Shaw intended the claimant to incorporate the quarantine within the period prior to returning on that date.

The tribunal observed at [16] that Mr Shaw “said in evidence that he put in the warning about the consequences of unauthorised absence due to the Claimant’s previous history set out above.”

Following intervention from the claimant’s union representative and another manager, Mr Shaw confirmed in an email of 19 February that the leave granted was extended to 12 March, amended in a second email that day to 15 March. In these emails Mr Shaw expressed concern that the claimant had not yet booked his return flight, given what had happened with the absence from August 2020. Both emails noted that a failure to return to work on time would be an unauthorised absence which could be considered misconduct, and disciplinary action could be taken including dismissal.

The tribunal observed at [17]: “Again this still did not allow the claimant the full time in Poland he had requested.” The tribunal continued at [18]

The tribunal accepted that the claimant believed, having looked into it, that he would be able to be Covid-tested upon arrival at the airport in Poland, and so avoid quarantine on arrival there; but in the event no testing was available, and he had to quarantine for 10 days, monitored by the Polish police. Organising the funeral and his father’s affairs required multiple face-to-face visits, so there was limited work he could do in advance from the UK or until he was out of Polish quarantine.

The claimant emailed Mr Shaw on 25 February 2021 explaining that he was unexpectedly subject to a 10-day Polish quarantine, and that his return flight would be on 24 March, which was the soonest he could book. In an email of 26 February Mr Shaw responded that any absence from 15 March would be unauthorised, could be considered gross misconduct and could result in dismissal. He noted that the information from the claimant suggested that, following 10 days UK quarantine after flying back on 24 March, he would not return until April. He urged the claimant to take all possible steps to return to work on the agreed date. If not, the company would have no other option but to consider disciplinary action. The tribunal observed that the implication of this stance was that the claimant would have to return without attending the funeral or conducting any of the other affairs for which he had travelled and spent time in quarantine.

On 1 March 2021 Mr Shaw emailed providing the claimant with details of return flights on 5 March, to enable him to return to work on 15 March. The tribunal observed that this was even though he would not be out of Polish quarantine until 4 March. Mr Shaw also provided details of later flights back on 12 and 14 March.

On 3 March the claimant replied giving more details of the strict Polish quarantine and the tasks he would have to carry out once it ended, and reiterating that he had always said he needed three weeks in Poland. The tribunal accepted that, once out of quarantine, the first date that the claimant could get for the funeral was 19 March, although he did not tell the respondent that date at that point. The tribunal accepted that he spent the entire trip very busy managing his late father’s affairs.

Mr Shaw replied by an email of 4 March that the claimant was expected to return to work on 15 March and any absence after that date would be unauthorised leave. The tribunal observed that “inevitably the claimant continued to do what he needed to do”. On 22 March the respondent wrote querying why the claimant had not returned to work on 15 March. The claimant replied, giving the date of the funeral, stating that he was returning to the UK on 24 March and would thereafter return to work after completing UK quarantine. He was told that there would be an investigation.

Following the Easter weekend the claimant returned to work on 6 April 2021. Also on 6 April the respondent produced a document which it required staff working on the Wandsworth contract to sign, though many refused. This was at page 140 of the tribunal’s bundle. It said:

Two employees from Goa in India, who had travelled there to visit family around this time, were required to isolate upon arrival there, had delayed return flights, and then had to isolate on their return to the UK. The respondent treated the additional time off that had not been pre-approved as unauthorised absence; and they each received final written warnings on 6 and 12 April 2021.

The claimant was referred to disciplinary process and received a final written warning from a Senior Account Manager, Paul Boxall. He considered the claimant’s case to be akin to that of a Kenyan colleague whose return from a trip to Kenya had been delayed for six months because of pandemic restrictions and who had been dismissed. However, he refrained from dismissing because of what he regarded as the mitigating circumstances of the funeral. Mr Boxall considered that the claimant had failed to keep Mr Shaw sufficiently informed of developments as they unfolded. He also considered that the claimant could have handled matters differently, but was going to do what he did, regardless of the respondent. The tribunal considered that Mr Boxall’s attitude was unreasonable.

The claimant appealed. His appeal was considered by a Service Director, Paolo Orezzi, who upheld the warning “due to the need to treat everyone the same”. He decided that, had the claimant raised issues earlier “other plans could have been made”; but the tribunal considered it clear that the claimant did communicate the issue, and it observed that “Mr Shaw never had the intention of extending leave to cover what the claimant needed.” [40]

The tribunal went on to say this at [41]:

The live complaints that were pursued were of direct discrimination, indirect discrimination and harassment. As to race, it was identified that the claimant relied upon “being from an overseas home country”, to which the tribunal added: “i.e. a foreign national working in the UK”.

In relation to the complaint of indirect discrimination the tribunal observed:

The tribunal set out at [5.4] that it had to answer whether the respondent had the following PCPs (I have silently corrected what both parties’ counsel, and I, agree was a typo):

The tribunal found that the respondent did have those PCPs. It said:

The tribunal added that there was no dispute that the PCPs were applied to the claimant and it was admitted that they were applied to everyone including UK citizens. As to whether the PCPs put persons who shared the claimant’s characteristic at a disadvantage compared with those who did not, because it was more likely that they “would have to travel abroad to deal with family bereavements which in turn would take greater time than those who had to address the same issue in the United Kingdom”, the tribunal concluded as follows.

In answer to: “Did the PCP put the claimant at that disadvantage?” the tribunal said:

As to: “Was the PCP a proportionate means of achieving a legitimate aim?” the tribunal said:

In relation to harassment the first question was whether the respondent did the following: “write two emails on 19 February, one on 26 February and one on 4 March 2021 making threats of disciplinary action”. The next was: “If so, was that unwanted conduct?” The tribunal continued:

The complaints of direct discrimination related to the imposition of a final written warning and the decision to uphold that warning on appeal. This conduct occurred. As to whether it was less favourable treatment because of the claimant’s race, the tribunal said:

Turning to the remedy decision, after a self-direction as to the law, the tribunal made an award for £15,000 for injury to feelings, and gave its reasons. This was a single award in respect of injury to feelings caused by both the indirect discrimination and the harassment. The tribunal continued:

I interpose that both counsel confirmed that £484 was the amount that the claimant had claimed in his schedule of loss in respect of lost earnings on account of attending the tribunal hearings.

The tribunal concluded by addressing the interest calculation and, at [13], making recommendations to the effect that the written warning should be removed, and the claimant treated as having a clean disciplinary record in relation to the previous matters from 2020 “given our findings that the previous two matters were a reason for the race-related harassment in this case.”

As well as having the written skeletons, I heard extensive oral argument. In what follows I describe the grounds of appeal, and summarise what appear to me to have been the most significant arguments. All of the more detailed submissions have been taken into account.

The appeal in relation to the liability decision, as originally framed, was considered by Eady P on paper not to be arguable. However, she considered the appeal against the aggravated damages award to be arguable. At a rule 3(10) hearing, at which Ms Seymour appeared, HHJ Katherine Tucker permitted redrafted grounds, which relate (a) to the decision that the indirect discrimination was not justified, and (b) to the finding, in respect of harassment, that the conduct related to race, to proceed; and there is now also a general challenge to the remedy decision, parasitic upon the liability appeal.

Section 19 Equality Act 2010 provides in part:

Ground A1 contends that the tribunal failed to apply the correct test to the assessment of proportionality in section 19(2)(d). It asserts that, instead of asking itself whether the PCPs were a proportionate means of achieving the legitimate aim, the tribunal asked whether they were “appropriate” in a different sense. It erred by factoring into its consideration at [55] its view of whether the leave granted to the claimant was adequate. The PCPs adopted did not, it is said, have anything to do with the amount of leave granted. The tribunal wrongly confused the proportionality of the PCPs with whether the amount of leave granted to the claimant in his specific situation should have been longer, which was irrelevant to the indirect discrimination complaint and justification.

Ms Seymour referred to Homer v Chief Constable of West Yorkshire Police [2012] UKSC at [22]; [2012] ICR 704 where Lady Hale stated:

As explained there, in this context “appropriate” means rationally connected to the legitimate aim. Ms Seymour submitted that it does not mean “appropriate” in any other sense. She also stressed that the tribunal’s focus must be on whether the legitimate aim justifies the specific PCP relied upon, citing by way of example Fasano v Reckitt Benckiser Group plc [2024] EAT 7 at [14].

Ms Seymour acknowledged that the tribunal started by using the correct language, when it asked at [55] whether the respondent’s requirements were “appropriate and reasonably necessary to the aims”. But she submitted that it then failed to consider whether the PCPs were appropriate, in the correct sense. By focussing on the particular circumstances of the claimant’s need for compassionate leave, and the particular amount of time he was seeking for that purpose, the tribunal had also failed to address whether the actual PCPs were proportionate to the aims.

Ms Seymour noted that (as could be seen from the particulars of the grounds of response) the original PCP relied upon by the claimant was one of limiting the duration of paid or unpaid holiday, which was said to put those in the claimant’s group at a disadvantage, as they were more likely to have relatives living overseas, in respect of whom it would take longer to organise funeral arrangements. But the revised PCPs adopted on the tribunal’s initiative, with the agreement of both parties, made no reference to the amount of leave that would, or might, be granted in a given case. The tribunal had then effectively, and wrongly, approached the test of justification as if they did.

Ground A2 contends that the tribunal erred by applying an inadequate level of analysis to the issue of justification. It concluded at [57] that there were more proportionate means of addressing the situation, by using agency workers, and managing leave entitlement to avoid a build-up of leave at the end of the year. But it failed to set out the necessary balancing exercise between the respondent’s business needs and the impact of the PCPs on employees; and there was no consideration, for example, of what the alternatives suggested by the tribunal would have cost the respondent, or whether it was viable to bring in agency workers to cover unexpected delays in a worker returning, despite the findings about the practicalities of onboarding agency workers at [41].

Ms Seymour relied on the observation of Sedley LJ in Allonby v Accrington & Rossendale College [2001] EWCA Civ 529; [2001] ICR 1189 at [24] that what is required is “an objective balance between the discriminatory effect of the condition and the reasonable needs of the party who applies the condition”. Further, the tribunal needed to consider the extent of impact on the disadvantaged group, both quantitively and qualitatively; and whereas the impact specifically on the claimant may be weighed in the balance, proper attention must be paid to how typical their experience is of the adversely affected group. See, on these points, Ralph Gibson LJ in University of Manchester v Jones [1993] ICR 474 at 497G and 498A. Reliance was also placed on the discussion in the speeches in Hardys & Hansons plc v Lax [2005] EWCA Civ 846; [2005] ICR 1565 at [33] (Pill LJ) and [55] (Thomas LJ) of the need for the tribunal to demonstrate a thorough and critical analysis showing a proper understanding of the employer’s business and the financial and other impacts of the approach being contended for by the employee. Ms Seymour submitted that all of that was lacking in this case.

Ms Seymour also submitted that, as far as it went, the tribunal’s consideration was of alternative ways of dealing with the claimant’s particular situation, not, as it should have been, of alternative ways of furthering the legitimate aims generally.

Ground A3 contends that the tribunal impermissibly speculated that there was a good chance that applying the PCPs would push employees to resign, and risk unfair dismissal claims. This was not based on any evidence or submission from the parties. That had been confirmed by a letter from the tribunal of 7 March 2024 in response to a Burns/Barke request. Ms Seymour submitted that this was simply not the sort of matter of which the tribunal could take judicial notice. The actual evidence was that the claimant did not resign. While the authorities recognise that a specialist tribunal may draw on its knowledge of the field to some extent, fairness dictates that the parties be given the opportunity to make representations about such particular matters on which the tribunal might be minded to rely. Ms Seymour referred to the discussion of the relevant principles in Dobson v North Cumbria Integrated Care NHS Foundation Trust [2021] ICR 1699 (EAT) at [42], [43] and [48].

Mr Platts-Mills noted that this was not (save, possibly, in relation to ground A3), a perversity appeal, and cited well-known authorities to the effect that tribunal decisions should be read fairly, not over-critically, and as a whole; and that where a tribunal has given itself a correct self-direction as to the law, the EAT should be slow to conclude that it has not followed it when reaching its conclusions, unless it is clear from the language used that the wrong approach has been applied.

He also made the following submissions specifically about the law relating to justification of indirect discrimination. The burden of proof is on a respondent to establish justification (see, e.g., MacCulloch v ICI [2008] ICR 1334 (EAT)); whether the defence is made out is a question for the tribunal acting as an industrial jury (Coventry City Council v Nicholls [2009] IRLR 345 (EAT)). For the defence to succeed, the PCP must be both appropriate to the aim, and reasonably necessary to it (Homer at [22]). So, he submitted, this challenge should fail unless the tribunal had erred in respect of both limbs; and/or the EAT need not remit, if the outcome would inevitably be the same.

Mr Platt-Mills noted that in Hardys & Hansons plc v Lax Pill LJ at [33] stressed the respect due to the fact-finding tribunal and that the issue for the appellate court is whether the tribunal has understood and applied the evidence and fairly assessed the employer’s attempts at justification. In that case the tribunal’s critique of the employer’s approach was at times “somewhat thinly supported by analysis or reasoning”; but ultimately the employer had failed because it failed to persuade the tribunal that its decisions were justified: [48] - [49]. So the tribunal’s decision in that case was upheld.

Mr Platts-Mills submitted that in the present case, on a fair reading of [55] to [58] as a whole, the tribunal did consider whether the PCPs were appropriate and reasonably necessary, balancing the discriminatory impact on the disadvantaged group against the respondent’s reasonable needs. The tribunal had not confined itself, or gone astray, by just focussing on the claimant’s position at [55]. That passage as a whole, and the earlier reasoning in relation to group impact, showed that the tribunal had considered the impact on employees and overseas staff in the plural, not just the claimant.

The defence failed in this case because the respondent failed to discharge the burden on it. The tribunal’s reasons were concise but adequate. Their brevity reflected the limited evidence that was put before it by the respondent. It was not contended by the respondent that there was any specific relevant evidence which the tribunal failed to consider, or any earlier finding of fact which it failed to take on board when reaching its conclusions. Further, in the 7 March 2024 reply to the Burns/Barke request the tribunal had observed that there was “no evidence either way that the respondent’s policy/practice in relation to leave would ensure enough enforcement staff on the streets.” The PCPs considered by the tribunal disadvantaged the claimant’s group because they would be less likely to get the leave they needed, as happened in his case. The respondent had been content to defend the modified PCPs by reference to its existing case and evidence on justification.

Mr Platts-Mills submitted that the tribunal properly took judicial notice, or drew on its industrial experience, in what it said in the second part of [56] and/or that this was a proper inference to draw from the primary facts found. Alternatively, this was not a finding of primary fact – but just an observation on a chance, which was itself by way of no more than a reflection of, or on, the fact that the respondent had failed to put forward sufficient evidence to show that the PCPs were appropriate to the positive furtherance of the aims. We do not actually know whether the tribunal had invited submissions on the point. Finally, even if the tribunal had erred in this regard, the outcome should still stand, because the respondent had failed to show reasonable necessity.

Section 26(1) of the 2010 Act provides:

Section 26(4) expands on the approach to effect in section 26(1)(b). However, the challenge by this appeal is specifically, and only, to the finding that the impugned conduct “related to” race.

In Tees Esk and Wear Valley NHS Foundation Trust v Aslam [2020] IRLR 495 the EAT said (after citing from Unite the Union v Nailard [2018] EWCA Civ 1203; [2019] ICR 28):

There are, in overview, two strands to this challenge.

First, it is said that the finding that the conduct in question – which was specifically and only by way of the four emails making threats of disciplinary action – was related to race, was inconsistent with (a) the finding, in the context of the indirect discrimination complaint, that the PCPs, which included treating taking unauthorised leave owing to quarantine as potential misconduct, were applied alike to those of the claimant’s race, and those not; and (b) the rejection of the complaints of direct discrimination in respect of the written warning, on the basis that it was likely that a UK national would, in the same material circumstances, have received the same penalty.

Secondly, it is said neither of the matters referred to in the course of [60] related to the claimant’s nationality, and that it was an error to rely upon the context having been that the claimant was an overseas national needing to travel home in view of a bereavement as showing that the conduct was related to the claimant’s nationality. That was an error of the type that occurred in Nailard. The respondent’s skeleton postulates that if, for example, an employee who was a UK citizen had to travel to Poland because of the death of a parent who was living there, they too would, on the tribunal’s findings, have received the same emails.

Ms Seymour submitted that, while the tribunal referred to Mr Shaw having been “intransigent” in relation to the amount of leave granted to the claimant, the four emails in question post-dated his initial response allowing the claimant leave until 3 March, and the first two confirmed that that period was being extended. At best for the claimant it might be said that the third email (of 26 February) did not grant the 25 February request for an extension; but these emails could not be properly described as intransigent. Further, if it was considered that Mr Shaw’s decisions on the amount of leave to grant, or not, were acts of harassment, that would have required an amendment, or at least for that way of putting the complaint to have been squarely put to him.

Further, she submitted, if the tribunal considered that the relevant passages in the emails, referring to disciplinary action, went beyond the policy which was the subject of the PCPs, and led to the written warning, it had not explained how that was. In so far as the claimant relied upon findings that Mr Shaw was influenced by his view of the absence episodes in 2020, that did not show that his conduct was “related to” race – the connection could not be forged in two steps in the manner of a “something arising from” complaint under section 15 of the 2010 Act. It was not sufficient that the claimant’s nationality had a bearing on the impact on him, of the policy that Mr Shaw was following.

Mr Platts-Mills submitted that the burden of the complaint which was upheld was that the impugned conduct went well beyond the mere application of the PCPs, because it involved not the mere giving of information about the implications of returning late, but repeated threats of disciplinary action, the making of which was driven by a prejudicial view of the claimant. As to why these threats were made, the tribunal referred to Mr Shaw’s own evidence that he had put in the warnings because of the claimant’s previous history, and it found that he had in turn taken a prejudicial view of that history, based in part on the fact that the claimant is Polish, and that it was likely that the claimant being from overseas contributed to his intransigent stance.

The tribunal had therefore not, erroneously, reasoned from the context being that the claimant was Polish, but had made a permissible reasoned finding that the conduct related to his being Polish. The fact that the direct discrimination complaints failed did not point to any error in this regard. Those complaints were about the decisions taken by others, to issue, and uphold a written warning, whereas this complaint was about Mr Shaw’s repeated issuing of threats.

In reply Ms Seymour submitted that this complaint related simply to the putting into operation of the PCPs, which might themselves be characterised as a form of threat, but were also themselves found to be race-neutral. The evidence referred to by the tribunal at [16] did not refer to the claimant’s race. The findings about Mr Shaw’s digging in his heels and intransigence at [60] related to his refusal to increase the amount of leave that had been granted, not the “threats”.

There are two points of challenge to this award. First, it is said that the respondent paying its own staff who it required to attend the hearing could not properly be regarded as aggravating conduct. Ms Seymour cited Zaiwalla & Co v Walia [2002] IRLR 697 (EAT) for the proposition that, while an award of aggravated damages arising from some aspect of the conduct of litigation is doctrinally possible, this would require something exceptional. She also contended that it was irrational to regard paying a perpetrator to attend the hearing as, in itself, condoning oppressive behaviour. Secondly, it is said that an award of this type, which is not punitive but compensatory, requires a finding that the aggravating conduct has exacerbated the injury to feelings. See: HM Prison Service v Salmon [2001] IRLR 425 (EAT). There was no such finding in this case.

Ms Seymour also submitted that this award was prompted by the claimant claiming compensation for his lost earnings in attending the tribunal hearings, something for which the tribunal properly concluded it could not award compensation as such. The tribunal had then wrongly decided instead to make an award of aggravated damages in the very same amount.

Mr Platts-Mills submitted that this case fell into the third category discussed in HM Land Registry v McGlue, UKEAT/0435/11 at [31], being of subsequent conduct such as conducting a trial in an unnecessarily offensive manner, or a serious complaint not been taken seriously, or a failure to apologise. He noted that the present tribunal specifically referred to the claimant having won, and found that Mr Shaw’s attendance at the remedy hearing “does add insult to injury”. He submits that this did not connote that the award was punitive, but that it was made because the tribunal considered that the conduct in question had added to the upset caused by the original harassment.

The challenge in this appeal is solely to the tribunal’s reasoning and conclusion in relation to justification. But, when considering that question, the tribunal had to do so by reference to the particular PCPs which it found were applied, the legitimate aim on which it found the respondent relied, and the particular disadvantage to which it found the PCPs put the claimant’s group and the claimant himself. While there is no challenge to the tribunal’s findings in relation to those other elements, as such, in order to evaluate the challenge to the reasoning on justification, it is necessary first to trace through the tribunal’s reasoning and findings in relation to those other elements.

As to the PCPs, those which the tribunal, in the event, with the agreement of both parties, considered, were set out at [5.4]. The tribunal’s conclusions that these were in fact applied, both to those in the claimant’s racial group and those not in his group, were at [47] – [51]. It is clear, reading these passages as a whole, that the tribunal analysed the PCPs as having two substantive strands in practice. The first was the requirement that any obligation to spend time in quarantine be factored in to any period of authorised leave (of any kind) initially sought and obtained. The second was the requirement to return to work at the end of the period of leave initially authorised, irrespective of changing events arising from the Covid pandemic, that is to say, a policy of refusing to entertain extensions to the originally-authorised leave period arising from pandemic-related events during the course of the leave period itself. I will distinguish those by referring, hereafter, to the “advance” requirement and the “no-extension” requirement. The tribunal found that both these requirements were applied both to the claimant and to his Goan colleagues, as well as to those of UK origin.

The group disadvantage postulated was set out at [5.7]. It was said to arise because non-UK nationals were more likely to have to travel abroad to deal with family bereavements or emergencies, which would take them greater time than it would take to address the same issue in the United Kingdom. I pause to note that the formulation given at [5.7] did not, however, spell out expressly why it was postulated that the need to travel abroad would mean that it would take greater time. Nor did it expressly spell out there why it was postulated that those who needed to take greater time would find themselves, for that reason, placed at a particular disadvantage by the PCPs.

At [52] the tribunal found that the postulated group disadvantage had been shown. It found this by reference to the examples of the claimant and the two colleagues of Goan origin. There it did spell out why it considered that having to travel overseas to deal with a family bereavement or emergency (which was a more likely scenario for those in that group) was found by it to be likely to take significantly longer. This was because “[q]uarantine was obligatory in respect of overseas travel and in some situations, as in these three examples, added between 20-28 days on to such a trip.” It is clear from this that the tribunal was taking into account there both the fact that, at the time, domestic requirements imposed a known mandatory quarantine period for those returning to the UK from abroad, but also that, in some cases, including those of the claimant and his Goan colleagues, a need to quarantine or other pandemic-related events following arrival in the country being visited could result in a further extension to the period originally sought and obtained then being required.

Ms Seymour stressed that, unlike the original formulation that was live prior to the start of the tribunal’s hearing, the PCPs that the tribunal in the event considered, and found applied, did not postulate that the respondent imposed limits or restrictions on the overall period of leave that an individual could take at any one given time.

However, I think it is certainly clear that the tribunal found that there was a limit to the overall amount of uninterrupted leave that the respondent was prepared to authorise in advance; and it was the application of what I have called the “advance” requirement in that factual context, that completed the picture of why it concluded that the application of that requirement gave rise to a particular disadvantage to those who had to travel abroad. That was because the period of UK quarantine would have to be factored into the period of overall leave that they required, and the limit on that overall period would mean that they would then likely have insufficient time to do what they needed to do.

Certainly, the tribunal plainly found that this was so in the claimant’s case. I think it is right, as such, that there is no finding that this was applied (or applied with the same severity) in other cases – the findings in relation to the Goans focus on what happened after they arrived in Goa – but, as I have noted, there is no challenge to the findings of group disadvantage, as such. In respect of what I am calling the “advance” requirement, it is therefore the disadvantage in fact found to be caused, by the interaction of that requirement with the limit on the overall period, which had to be justified.

In relation to what I have called the “no-extension” requirement, however, the disadvantage found appears to have arisen simply from the refusal to extend the original authorised leave, combined with the tribunal’s view that not all pandemic-related developments during the course of the leave could reasonably have been anticipated or catered for in advance. So this requirement put the claimant (who the tribunal plainly considered had reasonably failed to anticipate that he would be required to quarantine on arrival in Poland) in a position where, in the tribunal’s view, he had no choice at that point but to take unauthorised leave. But that conclusion, it seems to me, was not intrinsically or necessarily dependent on the factual element of there being a restriction on the initial amount of leave granted, although it might be postulated that the two were related. This difference, and the difference generally between the two requirements, needs to be kept in mind when considering the challenge to the tribunal’s reasoning in relation to justification.

Having identified the aim, and found that it was legitimate, at [54], at the start of [55] the tribunal then referred to whether the PCPs were a proportionate means of achieving that aim and whether they were “appropriate and reasonably necessary” to those aims. As Ms Seymour acknowledged, the tribunal at the outset of this section therefore posed itself the legally correct questions. Although the tribunal did not cite any authorities, it was plainly using the specific language used in the authorities in relation to the test of proportionality.

The remainder of [55] to [58] then set out the tribunal’s substantive reasoning on justification. It must be read fairly as a whole, rather than by dissecting it sentence by sentence or phrase by phrase. Further, having directed itself correctly, the tribunal should be taken then to have applied its own self-direction in what followed, unless it is clear that it did not; and again what matters is to consider the substance and material content of the tribunal’s reasoning. Mr Platts-Mills also fairly made the basic point that the authorities make clear that, once the application of the PCPs, and group and individual disadvantage have been found, the burden is then on the respondent to show justification.

However, the following particular points emerging repeatedly from the authorities (I cite only some examples) also need to be kept in mind. Firstly, the PCP must be “appropriate” to the aim or aims found to have been legitimately relied upon, which means that it must be rationally connected to that aim or aims, in the sense of being logically capable of furthering them (see, for example, Homer at [20] and [22]).

Secondly, the respondent does not have to show that the application of the PCP or PCPs was necessary to the achievement of the aim, in the sense of there being no alternative way to do so. Rather, the question is whether it is reasonably necessary: Hardys & Hansons plc v Lax at [28]. However, the balancing exercise may therefore include consideration of whether there were reasonable alternatives to the imposition of a discriminatory PCP: Homer at [24]. Further, in the proportionality or balancing exercise, the impact of the PCP on the affected group must be weighed against the importance of the employer’s need. The more serious the disparate impact, the more cogent the justification must be: Hardys & Hansons plc v Lax at [19]; Homer at [20] and [24].

An often-cited phrase, is that of Pill LJ at [33] of Hardys & Hansons plc v Lax, which itself draws in particular on the earlier discussion in the authorities, where he states that “a critical evaluation is required and is required to be demonstrated in the reasoning of the tribunal.” Thomas LJ in that case, at [55] also referred to the need for the tribunal to set out a “thorough and critical” analysis. In light of the particular arguments in the present case, it is important to consider why this is so.

One of the arguments considered in Hardys & Hansons plc v Lax was whether, when deciding upon justification, the “reasonably necessary” test means that the tribunal should apply a margin of discretion akin to a “range of reasonable responses” approach. That proposition was rejected by the Court of Appeal. At [32] Pill LJ said:

Pill LJ’s remarks at [33] follow on from this:

In similar vein Thomas LJ observed at [55]:

Gage LJ gave a concurring speech, specifically agreeing with Pill LJ’s reasons.

It is of course the case that the overall onus is on the respondent to show justification and that, by definition, what the respondent is being called upon to justify is, as such, a form of discriminatory treatment. Further, generally, in the nature of things, the employer should also be in possession of the evidence needed to demonstrate that the defence is made out, including, for example, in relation to potential alternative approaches that it maintains are not reasonably practicable or viable.

Nevertheless, what these passages highlight is that the need for a critical and thorough evaluation is not merely a reflection of the fact that what is being considered is whether a form of discriminatory treatment is shown to be justified. It is, specifically, because, where the claim is of indirect discrimination, the outcome is liable to have wider implications beyond the instant case, because of the lack of any margin of appreciation to be afforded in evaluating the employer’s defence, and, correspondingly, because the tribunal has the task itself of forming its own view of what Pill LJ called the working practices and business considerations involved, and so needs to demonstrate that it has understood and engaged with the evidence it has in relation to them. That in turn explains why, in relation to this issue, it is incumbent on the appellate court to consider critically whether the tribunal has demonstrated that it has discharged that obligation.

In the Hardys & Hansons plc v Lax case itself, while the tribunal’s decision that the defence succeeded was ultimately upheld, these considerations explain why Pill LJ, who considered the tribunal’s criticisms of the employer’s approach to be at times “somewhat thinly supported by analysis or reasoning” did so despite “misgivings”, and “not without hesitation” ([47] – [49]); and Thomas LJ did so “only after some hesitation” ([56]). Every case must of course go on its own facts, and every tribunal decision must be judged on its own contents; but I have borne the underlying themes of this guidance in mind when evaluating the rival submissions on the present appeal.

I turn then indeed to the particular challenges raised by this appeal, starting with ground A1.

First, Ms Seymour is right, as such, that the authorities use the word “appropriate” as a term of art to denote that the PCPs must be rationally connected to the aims, in the sense of being logically capable of furthering them. I agree with her also that the point made by the present tribunal in the main part of [55] beginning: “It was not appropriate …” is not actually addressed to that particular question. However, it seems to me that what the tribunal was concerned with, in substance, in this particular paragraph was the disadvantageous impact of the PCPs. This was not, as such, an irrelevant consideration. Notwithstanding the – dare I say – inappropriate use by the tribunal of the word “appropriate” at this point, I do not think that this alone shows that the tribunal in substance, in this paragraph, was having regard to a legally irrelevant consideration.

The next particular criticism, however, is that the tribunal erred by focussing, when weighing the disadvantageous impact, exclusively, or at least predominantly, on the particular circumstances of the claimant’s case (without, it was contended, considering whether they were typical of the group), rather than properly considering and weighing the group disadvantage.

As to that, in terms purely of the language, the discussion in this passage refers at [55] to the impact on “employees” (plural) and at [56], to the claimant “and his colleagues”, and at [58] to the restriction on “overseas staff’s need to travel” which – despite the placing of the apostrophe – was, I think, meant as a reference to the group.

However, I see force in the criticism that, in substance, a significant part of the tribunal’s reasoning drew on what it found to be the claimant’s particular circumstances. Specifically, at [53] the tribunal had found that the requirements put the claimant at a disadvantage “particularly as the respondent did not approve sufficient leave in the first place”. In addition the reference at [55] to the amount of leave agreed making it impracticable for the employee to accomplish what the compassionate leave was requested for, does seem to suggest that the tribunal was again, at this point, thinking of the claimant’s particular circumstances, rather than the impact of the “advance” PCP – which was not confined to those seeking compassionate leave – on the group generally. Nevertheless, as I have found, the finding of group disadvantage, as such, is not challenged by this appeal.

However, as I have discussed, a further criticism raised by the grounds is that the tribunal erred by introducing the question of whether there was a restriction on the amount of leave that the respondent was prepared to authorise into its consideration of proportionality, when this was not a feature of the PCPs considered as such. For reasons I have explained, I do not think that particular criticism is well-founded in relation to what I have called the advance requirement; but I do see the force of it in relation to the “no-extension” requirement.

In particular, Ms Seymour submitted that, when it came to the consideration of alternatives, the tribunal had focussed again on the claimant’s circumstances – she submitted that “the situation” at [57] referred to his situation. What is clear, I think, is that the tribunal was, in that paragraph, considering “the situation” at the particular time when the claimant (and, it appears, his Goan colleagues) took leave, and in particular the significance of it occurring at the end of the leave year. The tribunal specifically referred to that aspect at [57], and it was clearly more generally, in this paragraph, drawing on its earlier findings at [18] that Mr Shaw had referred at one point to the impact of the increase in absences “combined with the end of the holiday year” and on its earlier discussion of its consideration of alternatives at [41]. It had opened that paragraph by indicating that it was considering whether the respondent could have granted “the extended leave” – which was plainly a reference to the extended leave that the claimant had requested – and the discussion there specifically referred to the situation that had developed “where a lot of staff were using untaken leave in March.”

Overall, I conclude that ground A1 succeeds, on the basis that, to some extent in relation to both the advance and the no-extension requirement, the tribunal focussed too heavily on what happened in the particular circumstances of the claimant’s case, and gave insufficient consideration to the more general group impact.

I turn to ground A2.

In the present case, as has been noted, while the tribunal adopted revised PCPs, it did so with the consent of the respondent, and there was no objection on the basis that it had not had a fair opportunity to marshal the evidence that it would wish to adduce to justify the application of those PCPs as reasonably necessary to its legitimate aims. The issues of the risk of penalties under the contract with Wandsworth, the problem caused by the build-up of absences towards the end of the holiday year, and the logistics of bringing in agency staff were all in play and were considered by the tribunal at [41]. This passage plainly then fed in to the conclusions at [57].

While Ms Seymour submitted that the tribunal had not engaged in a sufficiently critical evaluation of, for example, the potential costs implications of heavier use of agency staff, she did not suggest that there was any evidence on this point, or any other specific evidence, that the respondent had in fact presented, but which the tribunal had failed to consider or properly analyse. Ms Seymour did fairly submit that the tribunal had overstepped the mark, in its reply to the Burns/Barke request, when it observed that there was no evidence either way that the application of the PCPs would be effective in keeping sufficient numbers of officers on the streets – because it had not been asked about that by the EAT. Nevertheless, she did not suggest that there was any specific evidence put before the tribunal about that aspect, but with which it had failed to engage.

Paragraphs [41] and [57] also show the tribunal to a degree engaging with, and evaluating the respondent’s arguments – accepting some, such as there being no possibility of homeworking during quarantine, but rejecting others. However, I think Ms Seymour has some justification in criticising its reasoning as failing to take sufficient account of the long lead times associated with bringing in agency workers. The tribunal did specifically engage with that aspect at [41] in relation to the “advance” requirement, and it seems to me plain that it was its view that the respondent should have treated it as inherent in the claimant’s request that he in practice required three weeks in Poland plus two weeks’ UK quarantine upon his return, and planned accordingly. But I see force in the criticism that the tribunal did not engage with the argument that recruiting an agency worker would not be a viable option, when confronted with an employee who was already on leave unexpectedly seeking an extension to their return date, in considering justification of the no-extension requirement.

Striking the right balance, in light of the discussion in Hardys & Hansons plc v Lax, in terms of the standard to which the EAT should hold the tribunal’s reasoning in this particular area, I do therefore see some force in this strand of ground A2, though I might have hesitated to uphold the appeal in respect of justification of indirect discrimination, had this been the only point which had any purchase.

However, ground A3 makes a different, and more trenchant criticism, of what the tribunal said at [56]. The Burns/Barke letter did confirm that the tribunal did not rely upon any specific evidence that there was a “good chance” that the application of the PCPs would push the claimant and/or his colleagues to leave, with an attendant risk of unfair dismissal claims.

This paragraph was plainly a significant part of the tribunal’s reasoning. Indeed, the language tends to suggest that the tribunal viewed this consideration as having a bearing not merely on whether the PCPs were proportionate but also whether they were appropriate – in the correct legal sense – to the aim of keeping sufficient officers on the streets. That is because the tribunal is saying in substance that the PCPs were liable in fact to undermine that aim, a reading reinforced by its statement that “we do not accept this as a means” for achieving that aim. Certainly, it considered that this was a consideration which weighed against the proportionately of the PCPs by comparison with the other “more proportionate” options that it went on to consider at [57].

The question of whether there was a “good chance” that the respondent’s stance would push the claimant and his colleagues to leave and whether the respondent could have been risking unfair dismissal claims, is not the sort of factual matter of which a tribunal could take judicial notice. Tribunals are sometimes called upon, for example in a counter-factual Polkey exercise, to evaluate the chances that a certain scenario might have unfolded, and when making such judgments may properly draw on their specialist industrial experience. But such exercises must always build on an evidential and factual platform. Further, consideration of justification for indirect discrimination, whether as to the “appropriate” threshold, or as to general proportionality, is highly fact-specific. In the present case the evidence was that the claimant did not resign, and the tribunal did not, as the Burns/Barke reply confirms, base this observation on any evidence that anyone else did. Further, in this case, this consideration plainly materially influenced the tribunal’s conclusion. I conclude that the tribunal erred in relying upon it, and ground A3 succeeds.

Having regard to my overall conclusions in relation to these grounds of challenge, I conclude that the tribunal did err in a material way both in relation to the question of whether both the PCPs were appropriate and in relation to the question of whether both were proportionate. The appeal against the tribunal’s decision on justification is upheld. I cannot, however, say that only one outcome would be possible, and so that question must be remitted to the tribunal for fresh consideration.

The complaint of harassment was that this had occurred by the respondent, in the emails in question “making threats of disciplinary action”. The concept of the making of a threat is an everyday and familiar one. It is not correct – and is contrary to all normal usage – to say that the fact that exceeding the authorised leave period could indeed lead to disciplinary action means in and of itself that what Mr Shaw wrote in the emails could not properly be regarded as a threat or, in any event, as materially no different from the application of the policy itself.

The tribunal was perfectly well placed to recognise a threat when it thought it saw it. It found that the emails did make threats of disciplinary action, and it was entitled to do so. That was particularly so having regard to Mr Shaw’s own evidence that he put the warning in the first email because of the prior history, a history which the tribunal found that he “held against” the claimant and which was the “overriding influence” on the way he treated the claimant in this matter. The tribunal was also entitled to consider that the emails as a whole involved repeated threats of this sort.

As well as finding in terms that the making of the threats was influenced by the negative view that Mr Shaw took of the prior history, the tribunal clearly found, on a fair reading of [60] as a whole, that that view was, itself, in turn, influenced by the claimant’s race. It could be argued that the first sentence, read in isolation, referred merely to a “but for” cause, and that the use of the word “context” appearing in the second sentence could suggest that the tribunal thought the claimant’s overseas nationality was no more than context. But that it is not a tenable reading of the paragraph as a whole. The overall sense of these two sentences, and what follows, is that the claimant’s nationality adversely influenced Mr Shaw’s conduct. The tribunal referred twice to Mr Shaw having a “prejudicial view”, which, in this context, plainly, meant prejudice related to the claimant’s nationality, and it referred not merely to the fact that the claimant had to travel back and forth to Poland for family reasons, but to the fact that the claimant “is Polish”; and it found that it was likely that the claimant being from overseas contributed to Mr Shaw’s stance, which the tribunal viewed as irrational.

I do not agree with Ms Seymour that the tribunal was addressing here only Mr Shaw’s stance on the question of the amount of leave, and not the making of the threats of disciplinary action, which was what the complaint of harassment was about. The making of the threats, and Mr Shaw’s approach to the amount of leave he was prepared to allow were, plainly, in the tribunal’s view, intertwined; and, in any event, in light of its overall findings, it is clear that the tribunal considered that his prejudicial attitude permeated his approach to the whole matter, and that the tribunal considered that the making of what it described as repeated threats was objectionable and oppressive.

While Aslam explains that there must be some feature of the factual matrix identified by the tribunal in support of the conclusion that the impugned conduct relates to race, the present tribunal’s findings that the making of the repeated threats was influenced by Mr Shaw’s view of the episodes the previous year, and that that view was itself based in part on the fact that the claimant is Polish, properly supported the finding that the conduct impugned by this claim was related to race. The tribunal found that the conduct was, in effect, because of race (and therefore certainly related to race), because it found that the claimant’s race materially influenced it, as it was influenced by Mr Shaw’s view of the previous episodes, which was itself influenced by race. The tribunal did not, therefore, commit the error discussed in Nailard, of relying on something that was merely background, context, or a but-for cause, but properly relied upon something that it found was a material cause.

Nor does the fact that the PCPs themselves, relied upon for the purposes of the indirect discrimination claim, were also applied to those not of the claimant’s race, point to any inconsistency in the tribunal’s reasoning. The making of the threats was not merely the same thing as the application of the PCPs. Nor do the tribunal’s conclusions in relation to the direct discrimination complaints point to any inconsistency. Those related to the imposition and upholding of the disciplinary sanction of a warning (not dismissal) by different individuals who were considering disciplinary charges referred to them, and whose motivations were, properly, separately considered by the tribunal. Indeed the tribunal referred at [60] to evidence from Messrs Boxall and Orezzi that they considered that, had the claimant requested an extension it could have been granted (albeit they were wrong to think he had not done so). The tribunal’s conclusions in relation to their decisions in no way precluded the tribunal’s particular findings in relation to the influence which the claimant’s race had on the different conduct of Mr Shaw to which the harassment complaint related.

I conclude that the tribunal did not err in law in finding that the conduct complained of as amounting to harassment did relate to race, which was the only element of its conclusions in respect of that complaint that was challenged by this appeal. The appeal in this regard fails, and so the tribunal’s upholding of that particular complaint stands.

I agree with both strands of the challenge to the tribunal’s award of aggravated damages. As the Zaiwalla decision confirms, there can, doctrinally, be conduct in the course of litigation of such an egregious character that, if it exacerbates the distress caused by the original wrong, it may sound in aggravated damages. But the bar is a high one. Litigation in general, and hearings in particular, are for most litigants an inherently distressing and unpleasant process, and the ordinary incidents of the proper and professional conduct of litigation, including the adversarial process, will not be sufficient to cross the line. The discussion in Zaiwalla at [24] suggests that the conduct would need to be tantamount to victimisation, for which, in the context of litigation, the bar is also high.

In the present case the basis of the tribunal’s award appears to have been the mere presence of Mr Shaw at the remedy hearing, and/or that he was paid. There was no suggestion, or finding, of any inappropriate conduct on his part, while there. Further, I do not think it could properly be said that, because he was not a witness, his attendance was therefore obviously ill-intentioned or could serve no proper purpose. I therefore do not agree with Mr Platts-Mills’ suggestion that this was conduct of a type that would fall into the third category in McGlue. I consider that the tribunal erred in treating it as of a nature that was capable of triggering an award of aggravated damages at all.

Further, and in any event, the tribunal erred because it did not make any finding of fact, based on evidence, that Mr Shaw’s presence had caused the claimant additional distress by way of injury to feelings. While the tribunal considered that his presence “added insult to injury”, that appears to have been the tribunal’s own view; and, dare I say, the metaphor was inapposite, because insult is not the same as injury, and some finding of fact that the claimant suffered some appreciable distress, rather than, say, merely being put out by Mr Shaw’s presence, would have been required.

For all these reasons I uphold the appeal against the award of aggravated damages.

It was common ground that, leaving aside the award of aggravated damages, the main award for injury to feelings was plainly a single composite award in respect of both the successful indirect discrimination claim and the successful harassment claim. That award will therefore need to be revisited once the outcome of the tribunal’s fresh consideration of justification for indirect discrimination is known. On that basis I consequentially allow the appeal in relation to it.

When this decision was circulated in draft under embargo I invited submissions on the implications of it for the two recommendations made by the tribunal in its remedy judgment. Ms Seymour and Ms Grace argued that they should both fall, Mr Platts-Mills that they should both stand.

The first recommendation was for the removal of the final written warning. This cannot properly relate to the upholding of the harassment claim, as Mr Shaw was not party to the warning. The direct discrimination claims failed. As such, it must somehow relate to the upholding of the indirect discrimination claim, and so must be quashed. The second recommendation was that the claimant was to be regarded as having a clear disciplinary record including what the tribunal identified as being the two previous occasions in 2020 which were “held against him on this occasion”. This therefore appears to derive from the upholding of the harassment complaint. But the tribunal found that the claimant was not punished in respect of either of those occasions; and in any event, I do not see how the upholding of the harassment complaint about later treatment could, as such, properly give rise to a recommendation in relation to those earlier matters, which were not the subject of the claim. That recommendation is therefore both otiose and ill-founded, and I will quash it as well.

When this decision was circulated in draft, I also invited further submissions as to whether I should direct that remission in respect of justification and remedy be to the same tribunal (if available) or a fresh tribunal and as to any other direction I may be invited to give.

Ms Seymour and Ms Grace invited me to remit the entire complaint of indirect race discrimination to be decided afresh, referring to my reasoning that the decision on the other elements of that complaint was the foundation for the consideration of justification, to points I made about that part of the decision, and making their own points about it. I decline to so direct. The appeal before me in relation to indirect discrimination was confined solely to the tribunal’s conclusions on justification. There was no appeal in respect of its conclusions on the other elements of the complaint, and it is not necessary to the disposal of this appeal to interfere with them. It is far too late for the respondent to seek to expand the scope of this appeal. Remission will be solely in relation to justification, which must be decided taking the existing findings in relation to PCP, group and individual impact as a given, though the tribunal may, I hope, be assisted by my analysis of them when it considers justification afresh.

The claimant’s counsel contended for remission to the same tribunal, the respondent’s counsel for remission to a different tribunal.

I do not doubt that if I remitted to the same tribunal they would do their best to decide the question of justification, and then revisit remedy, afresh, guided by my decision. But, given the language and strength of the expression of the present tribunal’s views and conclusions at a number of points in this decision, touching not just upon the harassment complaint but also in its reasoning in relation to justification for indirect discrimination, I think it would be too big an ask to expect the existing members to be able to approach the question with an entirely fresh collective mind. This outweighs the potential, and more uncertain, benefits of remission to the same panel, given also that, in any event, much of the heavy lifting in terms of basic fact finding has been done, and will stand.

I will therefore direct remission to a different tribunal panel of the issue of justification for indirect discrimination, hence the outcome of that complaint, and then financial remedy and what if any recommendation to make in respect of harassment, and – if, but of course only if, the justification defence again fails – indirect discrimination.

I will leave all other matters of case management, when the matter returns to it, for these purposes, including whether to permit any further adduction of evidence, to the tribunal.