Tesco Stores Limited v K Element & Ors

Three linked appeals are before the appeal tribunal in this long running multiple equal pay claim. It is brought by what is now a cohort of over 55,000 mainly female store based hourly paid Tesco employees who are seeking pay parity with their better paid, mainly male colleagues who are operatives based in Tesco’s distribution centres (“DC”) on the basis of their work being of equal value to that of their male colleagues in DCs by s.65(6)(b) Equality Act 2010 (“EqA 2010”) which provides that:

“A [a claimant]’s work is of equal value to B [a comparator]’s work if it is…. equal to B’s work in terms of the demands made on A by reference to factors such as effort, skill and decision-making.”

The issue in all three appeals concerns “Stage 2” in an equal value claim under the EqA 2010 which has the purpose of establishing the facts about the work of the claimant(s) and the comparator(s) on which the Employment Tribunal (“the tribunal”)is then able, at a subsequent hearing, to decide whether a claimant’s work is of equal value to that of their comparator. The Stage 2 procedure was conducted under the Employment Tribunals (Equal Value) Rules of Procedure 2013 rule 6 (“the EV Rules of Procedure 2013”) which was in force at the time (Footnote: ¹), contained in Schedule 3 to The Employment Tribunals (Constitution and Rules of Procedure) Regulations 2013, SI 2013 No.1237 (Footnote: ²) (“ET Procedure Regulations 2013”). Schedule 1 to the ET Procedure Regulations 2013 contains the generic Employment Tribunal Rules of Procedure (“ET Rules of Procedure 2013”) that apply to all proceedings. The ET Rules of Procedure 2013 are modified by the EV Rules of Procedure 2013 in equal value claims.

The EV Rules of Procedure 2013 set out a prescriptive regime to assist a tribunal to conduct the task of analysing the different work of a claimant or claimants and those to whom they compare themselves (their comparators), to decide whether their work is of equal value for the purposes of EqA 2010. At apreliminary “Stage 1 Equal Value Hearing”, the tribunal will decide, amongst other things, whether an independent expert(s) (“IE”) should prepare a report to assist the tribunal on the question of whether the claimant’s work is of equal value to that of the comparator, for the tribunal to then decide the question for itself. An IE is appointed from a panel of independent experts designated as such by the Advisory, Conciliation and Arbitration Service (ACAS) under s.131 EqA 2010.

Directions will also be made for the parties to provide an agreed written statement specifying the job descriptions for the claimant(s) and any comparators; the facts which both parties consider are relevant to the question of whether the claimant’s work is of equal value to that of the comparator; the facts on which they disagree (as to the fact, or as to the relevance, to the questionof whether the claimant’s work is of equal value to that of the comparator); and their reasons for disagreeing (r.4(1)(d) EV Rules of Procedure 2013) (Footnote: ³). Rule 4 contains standard orders for a tribunal to make at a Stage 1 Hearing, which a tribunal may vary, omit or add to, but in this case the standard orders set out in r.4(1)(d) were varied only in that the indicative timescale of the parties having 56 days (8 weeks) to present the information to the tribunal became a period of several years.

At a Stage 2 Equal Value Hearing (“Stage 2 Hearing”) conducted under r.6 EV Rules of Procedure 2013, intended to take place shortly after receipt of the agreed written statement provided under r.4(1)(d), the tribunal shall make a determination of the facts on which the parties cannot agree which relate to the question of whether the claimant’s work is of equal value to that of their comparator. The tribunal’s judgment after the Stage 2 Hearing should then produce a comprehensive set of facts that haveeither been agreed between the parties or determined by the tribunal (referred to and defined as “the facts relating to the question” (the question beingwhether the claimant’s work is of equal value to that of the comparator (r.1(1) EV Rules of Procedure 2013)). Where, as has occurred in this case, an IE has been appointed at the Stage 1 Hearing, the IE will then use the tribunal’s determination of facts and the parties’ agreed facts to prepare a report about the work of the claimant(s) and the comparator(s) before what is referred to as the “final hearing” in the EV Rules of Procedure 2013. Practitioners more usually refer to it as the Stage 3 Hearing, since it is rarely the final hearing in an equal pay case, this case being no exception. Three very experienced IEs have been appointed in this case: Ms Spence, Mr Walls and Mr Dews. At the Stage 3 Hearing a tribunal will consider the IE(s)’ report (if one or more has been appointed) together with the evidence of any party’s instructed expert (if that has been permitted by the tribunal) and the tribunal will then determine whether the claimant has established that her work is of equal value to that of her comparator. In short, the Stage 2 Hearing is about the facts of the work and the Stage 3 Hearing about the value of the work.

The significance of the Stage 2 Hearing is that the only facts on which the tribunal shall rely at the Stage 3 Hearing are those which have been agreed or determined at the Stage 2 Hearing. In a way which has echoes of a criminal court jury direction, the IEs and the tribunal may not rely on any evidence other than the facts agreed or determined at the Stage 2 Hearing (rule 6(2)EV Rules of Procedure 2013). There are two provisos to the rule. Where, as in this case, IEs have been required to prepare a report, the tribunal can order the IEs to assist the tribunal in establishing the facts on which the IE may rely in preparing the report, which may be done either on the tribunal’s own initiative, or on the application of a party (rule 5 EV Rules of Procedure 2013). Also, the IEs may make an application to the tribunal for some or all of the facts relating to the question to be amended, supplemented or omitted, which may be done at any stage of the proceedings (rule 6(3)EV Rules of Procedure 2013).

In this case, where there are so many claimants with a number of different job titles, the claimants have been divided into three tranches. This appeal concerns the Stage 2 Hearings for the first tranche which is a sample group of six claimants who perform a Customer Assistant role in the stores and eight comparators who work in the DCs. The claimants in this group are represented either by Leigh Day or Harcus Parker Employment solicitors. I will refer to the parties as they are in the tribunal. The intention is for two further Stage 2 Hearings to be held in respect of each of the second and third tranches of sample claimants and comparators.

The sample claimants in tranche one cover three specific roles: Customer Assistant Replenishment, Customer Assistant Express and Customer Assistant Nights. Of the eight sample comparators, four work in ‘ambient’ DCs or parts of a DC, distributing either non-chilled/frozen food (stored at ‘ambient’ or background temperatures) and four in ‘fresh’ DCs working in assembly of refrigerated or frozen food (collectively called ‘fresh’).

Tribunal proceedings were first commenced in February 2018 by around 10,000 claimants. Since then the number of claimants has grown. Progress in the case has been slow with a considerable number of preliminary hearings (“PH”) and case management orders (“CMO”) several of which have been appealed to this appeal tribunal (“EAT”) with permission granted in at least two appeals to the Court of Appeal on preliminary points (although there have been no full hearings or judgments from the Court of Appeal in this litigation as yet as far as I am aware). Apart from one appeal discussed below at paragraph 69, none of the other appeals are directly relevant to the issues in these appeals, but are mentioned in order to illustrate the complexity of the procedural history and that the litigation is hard fought.

The relevant chronology in these appeals begins on 3 June 2020 at a PH before EJ Manley when she listed the Stage 1 Equal Value Hearing for a one day hearing on 2 September 2020 and a 15 day hearing commencing on 28 February 2022 for the Stage 2 Hearing. At the 3 June 2020 PH, which took place in the height of the pandemic lockdown, instead of the normal procedure of all correspondence between the parties and the tribunal being routed through HMCTS administrative staff, EJ Manley directed that email communication about the case only for the purposes of progression of the case was, “for the moment”, to be directly between her on her E-judiciary account and the representatives, because of the shortage of staff at the Watford tribunal office at that time. The practice of direct communication between the assigned EJ and the representatives has continued uninterrupted since.

After the Stage 1 Hearing on 2 September 2020, at a further PH on 28 January 2021 the tribunal approved an agreed suggestion by the parties that the IEs prepare a draft outline template for job descriptions and factor headings and the IEs were invited by the EJ to attend the next PH on 7-8 July 2021. At the 7-8 July 2021 hearing the Stage 2 Hearing was vacated and re-listed for a total of 34 days – 25 days with the parties and 9 days for the tribunal to deliberate - over various dates from March – May 2023. Orders were made, largely by agreement for the identification of sample claimants and comparators and a detailed process to be followed for the preparation and exchange of draft equal value job descriptions (“EVJD”). The claimants’ legal teams were to prepare the first draft EVJDs for the sample claimants and the respondent those for the selected comparators, both with the involvement of the IEs. Orders were made that after the first draft EVJDs had been exchanged the claimants and comparators were to be interviewed by the parties, the interviews transcribed, second draft EVJDs were then to be drawn up and exchanged and a timetable set for points of clarification to be raised and answered for incorporation into third draft EVJDs. Thereafter the parties were to identify and record in table format a “Record of Dispute” (“RoD”) in which the parties were required to record all factual matters remaining in dispute with reference to the relevant EVJD, the text in dispute and the parties’ respective positions on amendment, deletion or supplementation. A date was set (10 October 2022) for the partes to have held round table discussions (“RTD”) at which they were to seek to agree as many disputed matters as possible with the IEs in attendance if possible, and to identify the disputed matters that remained live. The RoD was to be updated following the RTDs, standard basis disclosure on the issues remaining in dispute was to follow, with fourth and final EVJDs to be drawn up and the usual preparations (bundles etc) for the tranche 1 Stage 2 Hearing in February 2023.

At a further two day PH on 10-11 November 2021 the parties were recorded as having made considerable progress by EJ Manley but a few matters required orders, such as specific disclosure of personnel files and training records. Other than suggesting that no more than 6 people from each side should attend the interviews of the sample claimants and comparators, the EJ urged the parties to resolve issues around the protocol to be adopted for the interviews and RTMs between themselves. Following the retirement of EJ Manley a little while later, EJ Hyams was nominated as the EJ with conduct of the litigation. He conducted a PH on 19 December 2022 recorded in a Case Management Summary. He made two orders (1) for the respondent to provide a statement of the amount of time each comparator spent on each task which they carried out as part of their work from time to time, and (2) for the respondent to state what training modules it would have expected a staff member in each sample claimant’s position to have undertaken in the relevant period and to provide copies if they had not already been provided.

The fourth and final draft of the EVJDs (sometimes also referred to by the parties as the RoDs) for each of the sample claimants and comparators set out the facts that the parties agreed about and those that they disagreed about and why. It also identified whether the disagreement was about the fact contended for by the other side or the relevance of the fact contended for by the other side, in compliance with rule 4 EV Rules of Procedure 2013. The EVJDs were lengthy and so too were the numbers of areas of dispute within each job description. For example, Mrs Worthington’s EVJD contained 682 numbered paragraphs, some of which contained several numbered sub paragraphs and there were 1,427 numbered rows in the tableof disputed points in the RoD. To give a better picture of the scale of the tribunal’s task, the EVJDs for each of the Leigh Day claimants were between 122 to 175 pages, for each of the ambient comparators the EVJDs were between 207 and 342 pages and for each of the four fresh comparators, there were 3,265 paragraphs containing one or more disputed points.

The layout of the EVJDs was in tabular form:

There had also been an attempt by the parties to agree a points of principle document (“PoP”) with the help of the IEs, to assist with progress at the RTDs, but it had not resulted in agreement.The tribunal appears to have accepted the Leigh Day claimants’ submission that the respondent’s subsequent comments on the PoP suggested an unwillingness to engage positively in the process. The respondent then took a decision to cease participation in the RTDs altogetherafter 5 of the intended 13 RTD meetings had taken place, in advance of the Stage 2 Hearing.

The Stage 2 Hearing commenced over a year later than originally listed by EJ Manley and took place before EJ Hyams and lay members Mr R Clifton and Ms M Harris over 27 dayswith the parties present between 6 March – 24 May 2023. In addition two site visits were conducted on 22 February and 13 April 2023. There were over 30 witnesses called who gave evidence and the hearing bundle that had started at 22,000 pages at the beginning of the hearing expanded on a daily basis. Some of the witness statements were very lengthy, for example the witness statement for one of the respondent’s managers, Matt Diment, who gave evidence about one sample claimant, was 187 pages long. The EVJDs and RoD were also in evidence before the tribunal. During the course of the Stage 2 Hearing the EJ continued to correspond directly with the parties at some length. On Friday 12 May 2023, with just 4 days left of the hearing with the parties present, he informed them in an email that he did not consider it would be necessary in the judgment to record matters that had been agreed, relying on the authority of Lavender J in the High Court equal pay case of Beal v Avery Homes (Nelson) Ltd [2019] EWHC 1415 QB 6 June 2019. He also informed the parties that his initial view was that it would be helpful for the tribunal to state its conclusions in narrative, not tabular, form and not by reference to the RoD created by the parties, again drawing on the authority of Beal. He said that unless it was obvious why the tribunal had determined a particular factual dispute, they would give a brief statement of how they had resolved the conflict of evidence. He also indicated a number of legal issues on which he would welcome submissions.

The ET then spent 17 days in deliberations in which it took them over a week to analyse the evidence relating to the work and working conditions of just one sample claimant, Mrs Worthington. In their judgment promulgated on 12 July 2023 (“the First Judgment”) they concluded that the manner in which the parties had put their evidence and contentions:

“…was so inconsistent with the interests of justice that it was not just to determine the factual disputes as they stood at the end of that hearing.” [1]

The tribunal was concerned that the respondent may not have given full disclosure of what the tribunal considered to be the key documents which gave the best evidence of the work of those doing the jobs of the sample claimants and their comparators, namely the documents of in-person training given to the claimants and comparators and the documents that they were required to read and apply during the relevant period [2] (appeal ground A1G2). It rejected the respondent’s submission that the training documents were relevant only if the sample claimants or their comparators had in fact received the training described in those documents [3]. The tribunal ruled that the respondent had been wrong to submit that the factual enquiry had to be about what the sample claimants and their comparators in fact did on a day-to-day basis. The EVJDs were criticised for having been written:

“on the basis of an assumption that the primary focus …was ..what (1) the claimants and their managers and (2) the comparators and their managers said about the manner in which (respectively) the claimants and their comparators did their jobs” [4]

which the tribunal said was wrong. Later on, in the reasons section of the judgment, it described the respondent’s position as “fundamentally erroneous” [89].

Instead, the factual enquiry had to be about “what the respondent required them to do as their jobs as evidenced by the training which the respondent gave to persons doing those jobs” [3](appeal ground A2G2). Accordingly, the training documents would be the best evidence and should have been the primary focus of the ET’s inquiry into and determination of the facts of the question of whether the claimants and the comparators did work of equal value.

The ET ordered both parties to prepare new written job descriptions for all the claimants and comparators and a new statement of relevant facts – both agreed and disagreed - in accordance with rule 4 EV Rules of Procedure 2013 [7]. There would be a further PH a week later on 19 July 2023 for the precise terms of the orders to be discussed, and if possible agreed.

After setting out its 8 paragraph unanimous reserved judgment in the First Judgment, the tribunal then set out its reasons which are lengthy (58 pages). In paragraphs [11] – [26] the tribunal sets out the substantive law with extracts from the relevant case law which it said it had found illuminating, although it was not always entirely clear how the illumination had assisted or what light had been shed. It is clear that the tribunal was at pains to distinguish between finding facts – which it was required to do under r.6 - and the evaluation of the facts, which would start with the IEs’ report ([24]). As well as reaching a firm conclusion that the training documents “were likely to be the best evidence of what the job in each case entailed” [75], it made a number of what it described as “provisional conclusions on the issues of legal principle” (although elsewhere expressed with more certainty) most of which are challenged in the grounds of appeal and set out more fully below. In summary, it concluded that it was necessary for the tribunal to apply its own critical analysis to the question of what factual disputes required determination [28], rejecting the respondent’s submission that the tribunal would have to decide every dispute, as “if a fact might reasonably be relevant for the purposes of the later assessment of value by the IEs or party experts, that fact must be included in the job description.” (respondent main opening skeleton argument [14]), (appeal ground A1G8). In identifying the conclusions and findings of the tribunal that are under challenge, it is important not to overlook that parts of the First Judgment are not criticised, but since they are not relevant to the grounds of appeal it is not necessary to set them out and lengthen what must inevitably already be a lengthy judgment.

After listing the witnesses the tribunal set out its findings of fact on some “salient aspects of the evidence” [50]-[64] in which it concluded that the claimants were put under time pressures in the same way that the respondent’s DC staff were put under time pressures ([55.1]) and that:

“55.2

the only material difference in this regard between the two sets of staff was the fact that some of the DC staff wore arm-mounted computers or terminals about which we heard much oral evidence, as a result of which it was possible to monitor more effectively their use of time.”

The tribunal made a number of clear findings about the reliability of parts of the respondent’s evidence:

“…..on many occasions during the cross-examination of the respondent’s witnesses it became clear that they had attested to things about which they had no knowledge, such as statistics or percentages about relevant things.” [56]

The tribunal accepted the description in the Leigh Day closing submissions that this was “an exercise in ventriloquism.” It went further and accepted in their entirety, two paragraphs of the Leigh Day closing submissions which it described as entirely apt and set out in full [56]. In summary the tribunal found as a fact that the respondent store manager witnesses had diminished the work (and value) of the work done by the claimants whereas the DC management, had, by contrast, talked up the jobs done by the comparators. The tribunal found that the respondent witnesses frequently had to admit in cross-examination that paragraphs in their witness statements asserting that the EVJDs of the comparators were correct, were not in fact correct. One respondent witness accepted that he was “over-egging” the complexity of a task of a comparator in his witness statement and others admitted that the language used in their witness statements gave a spurious complexity to the tasks; and others admitted that they had little direct knowledge of what the claimants were doing whilst working on shift. The tribunal accepted that these were not occasional failures of coming up to proof, but a consistent pattern, leading to the “exercise in ventriloquism” tag [56]. These were clear findings which the tribunal reached by reference to specific examples of witnesses’ evidence.

The tribunal stated that:

“We concluded that the respondent’s legal representatives had compiled evidence at least in part with a view to supporting the respondent’s intended submissions to this tribunal, and then asked the witnesses who were called to give evidence to us to approve a witness statement (or statements) which was (or were) intended to support those submissions. Those witnesses had then felt obliged if at all possible to agree to those statements, and in at least some cases they had not read them properly before doing so."[57]

It led the tribunal to conclude that it could not accept any of the statistics relied on by the respondent in the way they had sought to introduce them as evidence, which had been through the mouths of those who had no knowledge of the statistics and were also unable to explain them: it was not reliable evidence [61].

The tribunal also identified at least one “salient inconsistency in the respondent’s case” which concerned the cages that contain the goods that end up on the supermarket shelf which travel from DC to store. The cages are filled and packed in the DC by the comparators (and those with their job title), transported to the store and then unpacked by the claimants (and those with their job title) and the goods then put on the shelves. The tribunal noted that the respondent’s evidence was that the cages in the DCs had faults which “sometimes…caused lacerations”, but, once in the store, only rarely had faults which might scratch the user, if the user was not taking proper care [63]. The clear inference is that the salient inconsistency means that the evidence was implausible and it is not believable that cages which were faulty in the DCs somehow became not faulty when they reached the stores, which the tribunal described as “a focus [by the respondent] on the intended submissions of the respondent rather than the facts”.

The tribunal concluded that it was not necessary to record the agreed facts, since these had been agreed by the parties. It rejected the parties’ submission that not recording an agreed fact would mean that the parties would be likely to spend a very long time disputing how those things which were agreed should be put before the IEs. The tribunal concluded that this issue would be resolved by their order for the parties to re-cast their case.

The tribunal considered the respondent’s evidence of the performance management regime in place at the DCs and the application of performance indicators derived from the use of information and communication technology by the comparators to measure the Performance Index (PI) rates, which were then applied to different sets of employees to measure effort and which were determined in part by reference to the physical layout and conditions of the workplace in question [70]. The tribunal found that the PI rates were “not irrelevant” and set out the ways in which they were relevant. The tribunal also concluded that two matters relied on by the respondent were not relevant at a Stage 2 Hearing. The first was “the physical location of the workplace as such, for example whether it was in a peaceful residential neighbourhood or an isolated industrial park”(appeal ground A2G10) and the second was “relationships which the job-holder had with his or her colleagues”(appeal ground A1G7) [74].

The tribunal made a discrete decision on what it called a point of principle in relation to training: that there “was no good reason to make determinations about precisely what training the sample claimants and their comparators had actually received” [87] other than training that would determine whether the employee could do an aspect of the work, such as forklift truck driving [88] (appeal ground A1G5). It is a separate ground of appeal to the general challenge to the tribunal’s approach to training materials in ground of appeal A1G2.

The further PH took place on 20 July 2023 (“the 20 July 2023 PH”) (a day later than intended) for the parties and two of the IEs to discuss and if possible agree, the precise terms of the orders required to furnish the tribunal with the information and documentation it had requested. A Case Management Summary was sent to the parties after the 20 July 2023 PH on 26 July 2023 (“the CMO”). Between 12 – 20 July 2023 the Employment Judge conducted what he described as “a dialogue by email with the parties.” In the email correspondence and during the course of the hearing the parties pushed back on the tribunal’s decision that both sides reformulate the job descriptions from scratch as they said it carried very significant time and cost implications. In a rare moment of unity all the parties objected to the tribunal’s approach in the First Judgment (CMO [45]). The respondent also indicated its intention to appeal the First Judgment. In the CMO the tribunal recorded that after taking over a week to determine the factual disputes relating to the work of the first sample claimant, Mrs Worthington, it had then taken the EJ three further days working on a draft set of reasons to record only the first half of its conclusions in the factual disputes about her work and that it could take three months, if not more, of continuous sitting to determine the thousands of factual disputes between the parties, on the basis of the way in which the parties had presented the case [31].

The outcome of the 20 July 2023 PH and CMO was that no order was made for the respondent to undertake further searchessince the claimants accepted that the respondent had already provided full disclosure of all relevant documents. The tribunal stated that, although by the end of the hearing it was even more sure than it had been at the beginning of the day that the manner in which the parties had presented their cases on the evidence was inconsistent with the interest of justice and that it would be necessary for the parties to recast the EVJDs, in light of the resistance of the parties and the planned appeal by the respondent, the tribunal decided not to press ahead with the course of action ordered in the First Judgment. Instead it adjourned the Stage 2 Hearing to 10 November 2023. It ordered the Leigh Day claimants and the respondent to recast the information required for r.4 EV Rules of Procedure 2013 for a single job task for one sample claimant and a single comparator and from that to provide an estimate of the time and cost of recasting the EVJDs and statement of facts agreed and those in dispute for all the others in the way desired by the tribunal (“the Recast Task”). The Harcus claimants had declined to be involved in the process. The tribunal would then seek the views of the IEs on whether the Recast Task would be of material assistance in an appropriate form with a proportionate level of detail. The IEs had contributed to the discussion at the 20 July 2023 PH and informed the tribunal that the standard and, in their experience, invariable practice was to draft job descriptions and state findings of fact (or disagreements) by reference first and foremost to the things which the parties said about the relevant person’s work within the meaning of s.65(6) EqA 2010 [42]. However, with the large volume of highly detailed and prescriptive training material in this case, the tribunal concluded that the standard practice was not applicable in this case [42].

As recorded in the CMO, the tribunal returned to its reasoning for concluding that new job descriptions should be drawn up:

“11……As stated by EJ Hyams then [a reference to the course of the 20 July 2023 PH], that conclusion [that new job descriptions should be drawn up] was based on our own interpretation of section 65(6) of the EqA 2010, reading it against the background of the purpose of the equal pay legislation as we understood it. That conclusion was fortified in particular by what Underhill P said in paragraph 22 of the EAT’s judgment in Prest v Mouchel Business Services Ltd [2011] ICR 1345 (which we set out in paragraph 26 of the Reasons). That paragraph in our view showed how important it was in general to focus on what work the employer in practice required persons who were in the role of the comparator to do, rather than on the way in which that comparator actually worked. Of course, if the comparator worked otherwise than in accordance with the employer’s requirements, then the question would arise whether that different way of working was at least condoned by the employer, but that did not detract from the implication of the judgment in Prest that what is in issue when determining the work within the meaning of section 65(6) of a comparator is primarily what were the employer’s requirements of the job-holder in that job, and not how the job-holder in practice did that job. We could see no justification at all for a different approach to be applied when determining the work (within the meaning of section 65(6) of the EqA 2010) of a claimant, including, in particular, a representative claimant in the case of a number of claimant employees with the same job title and at least broadly similar tasks to do for the employer. … [11] (appeal ground A1G1)

The CMO stated that the CMO and the First Judgment were the definitive statement of the tribunal’s reasons for the orders set out in the CMO. However as will become apparent in the chronology below, the tribunal returned to the issue several times thereafter.

In response to a question from the Harcus claimants the tribunal confirmed that the written and oral evidence of the witnesses called at the Stage 2 Hearing stands.

The tribunal was dissatisfied with what Leigh Day and the respondent had produced in their Recast Task – in the tribunal’s judgment of 5 July 2024 (“the Second Judgment”) they described the document as being no shorter than the previous work andusing a playing card analogy said:

“In effect, we had simply a shuffled deck of the same cards, but now with some more information about the training materials on which the party’s case could originally have been based.” [35]

Leigh Day considered that recasting all the EVJDs would take over 1,770 hours just to get them into first draft stage, before the additional time that it would take for the parties to seek to agree, amend and prepare documents in a final form for the tribunal. The respondent considered it would take approximately 3,400 hours to recast the comparators’ EVJDs and would cost in the region of £1.3m.

In an email from the EJ to the parties of 9 October 2023 the tribunal changed its position and said it would now not make any further orders. Instead, it would:

“go back to the beginning and create a new set of documents stating our conclusions on (a) the job tasks of each sample claimant and each comparator, and (b) such additional facts as we see as being relevant for each claimant and comparator.” [also set out verbatim in para 42 of the Second Judgment]

In its Second Judgment the tribunal explained that it had decided:

“…that we would have to just do the best that we could with what we had. And that we should just get on with the job without further delay.” [42]

The tribunal planned to produce a new set of documents for each sample claimant and comparator, primarily by reference to the training materials. The EJ indicated that the tribunal thought it would be very helpful to hear from the IEs on the value or otherwise of several of the factual disputes between the parties and matters could be discussed further with the parties and the IEs at the 10 November 2023 hearing and directions were made for exchange of skeleton arguments in advance of the 10 November hearing. The purpose of the 10 November 2023 PH hearing would now be to discuss and agree a timetable for the rest of the hearings and sequencing.

In an email from the EJ to the parties on 23 October 2023 the parties were then informed that the tribunal was currently working on the assumption that they would send out a judgment with determinations of all the factual matters relating to all of the 14 employees’ work.

Claimants’ counsel agreed a short “Joint Note on the Law” which was sent at 4.52pm on 27 October 2023 to the EJ and the respondent about the task of the tribunal at Stage 2, the categories of things that constitute the “work” of a job holder and a few points about relevant factors and sources of information. It was described as being intended to assist the parties and the tribunal. It is explained in the covering email that their note set out what the claimants considered was “a correct statement of a number of important applicable legal principles” and that they would seek to agree the note with the respondent in advance of the 10 November 2023 PH hearing. They asked that if the tribunal should take a different view in relation to any of the principles outlined, they would welcome it being explained at the 10 November PH hearing. The respondent did not agree the note at the time, but Mr Reade KC confirmed before this tribunal that the respondent did now agree that the note was accurate in all respects. The text of the joint note is set out in full below at [81].

Within 30 minutes of receipt of the email EJ Hyams replied to the parties by email to say that the tribunal’s understanding of the law was slightly different and he went on to explain why, repeating much of what had been said in the First Judgment of 12 July 2023 and the CMO. The EJ again referred to Prest and stated:

“…it is, we think important to bear it in mind that the work that is being evaluated here is that of customer assistants and distribution centre workers employed by the respondent, so that the proper focus of the tribunal’s enquiry is what were their jobs, or what was their work, for equal pay purposes, and that is a matter which falls to be determined by the employer, not the employee. Thus, the inquiry is primarily what work the employees in question were required by the employer to do, not how they did it or what they in practice did if it was different from how they were required to do it, unless that which they did was explicitly or implicitly approved by the employer so that it became part of their jobs.

As a result of our understanding of the law, as stated above, and as previously stated exhaustively in our reasons of 12 and 26 July [the First Judgment and the Case Management Summary after the 20 July PH hearing], we believe that the proper primary focus of the experts and the tribunal in regard to particular tasks will be what was required by the respondent of the claimants and their comparators in relation to those tasks, so that if it was best shown by the respondent’s training materials or documents relating to safe ways of working, then those materials or documents will be applicable to all employees doing that work ”

The EJ then attempted to draw a line under the debate by saying that he saw no good purpose in going over old ground by way of further submissions and tetchily wondered on what basis the note had been sent to him. However, the tribunal did not draw a line under it. In a further email to the parties of 9 November 2023, the day before the next PH, the EJ returned to the point and provided 26 numbered paragraphs ([9] – [25]) of explanation and comments on the law relating to the determination of facts at a Stage 2 Hearing. The email also stated that the tribunal anticipated that their deliberations would last until at least April 2024.

At the 10 November 2023 PH the Tribunal explained that it had nearly completed the task it had set itself of stating (1) the results of its review of the training materials in so far as they related to the work of Mrs Worthington, and (2) its conclusions on all the outstanding factual disputes in relation to her work. In doing so they had recast Mrs Worthington’s EVJD which was 130 pages (Footnote: ⁴). They were then planning to record their conclusions on the points in dispute in relation to the other 13 EVJDs. The IEs had been present at the hearing and informed the Tribunal that the length and detail of the EVJDs would be a difficulty for them. They had consistently said in the two years that they had by now already been involved in this case, that in all the other large cases that they had worked on, the Stage 2 findings were incorporated into final EVJDs by the parties. They are then exchanged, amended where agreed between the parties and final versions sent to the IEs in order for them to prepare their report on the question of equal value.

On 13 March 2024 the IEs again expressed their concern regarding the format of the final job information. They made clear that if the raw material they were to use was to consist of the parties’ original job descriptions and an additional, separate document containing the tribunal’s findings, they were concerned that it would require them to interpret those findings and make value judgements as to their specific relevance to the facts provided in the EVJDs, which risked further disputes and frequent and/or lengthy applications to the tribunal for clarification. They were anxious to present their report in as clear and unambiguous fashion as possible and requested further evaluation by the tribunal. They asked if the tribunal could consider ordering their findings to be incorporated into single EVJDs for all claimants and comparators.

The tribunal did not do so but continued with its work in providing a Stage 2 judgment.

The tribunal’s prediction of the time it would take to make its Stage 2 findings was an underestimate by three months. TheSecond Judgment was sent to the parties on 5 July 2024. It is 619 pages long with a total of 2,378 paragraphs of its judgment and eight appendices, written after a further 55 days of deliberation in chambers by the full tribunal, making a total of 109 days of hearing, deliberation and site visits and the EJ had also spent most of his time in writing up the judgment in the period since July 2023. The Second Judgment stood by the points of principle set out in the First Judgment [1]. The tribunal, once again, returned to the question of what work is:

“9.

The most basic principle of all was this: an employee’s job is not what the employee says it is, but what the employer says it is. The employer requires the employee to do work, and the employee is required to do that work. Putting it in a slightly different way, the employer agrees to pay the employee for the work which the employer requires the employee to do. If the employee does not do that work then the employee is not doing his or her job. If the employee does not do the job well, then, albeit to a lesser extent, the employee is also not doing his or her job, i.e. the work which he or she was employed to do. The employee cannot by failing to do his or her job or doing it badly change the nature of the job. Only if the employer agrees to the employee not doing the work in the manner required by the employer will the employee be able credibly to say (both generally and, in our view, for the purposes of a claim for equal pay for work of equal value) that his or her work is now to be done in that different manner.”

The tribunal then again revisited a recognised key early case in this area that it had discussed in the First Judgment, Shields v. E.Coomes (Holdings) Ltd.[1978]1 WLR 1408 which emphasised the important principle that had already been recognised in a number of cases that contractual differences between a claimant and a comparator are not relevant differences “unless it results in an actual difference in what is done in practice. It is by comparing their observed activities not their notional paper obligations that the relevant differences are to be ascertained.” (@E-F 1427). The tribunal concluded that Shields was of “very little relevance to at least the majority of the factual disputes before us here” [13] for five reasons. Firstly because Shields was a like work, not an equal value case; secondly because the respondent did not rely on any substantive provisions in the contracts of employment as explanation for the higher rate of pay for the comparators; thirdly because the tribunal considered it needed to decide “what were the tasks which the sample claimants and their comparators were required to do by their employer” [13.3]; fourthly because if there had been no documentary evidence then only oral evidence could have been put before the tribunal about those tasks; and fifthly because this respondent is a very large employer which:

“13.5.1

operates in a highly regulated environment, which includes strict criminal liability for the sale of food products which (summarising the situation broadly) have exceeded their shelf life or are contaminated,

13.5.2

uses sophisticated digital stock control and related systems, and

13.5.3

has in place an extensive and detailed set of training materials for use in training employees to do the work of the claimants and their comparators.”

The tribunal then concluded that the factors meant that:

“14.1

the respondent had a strong business need for the work of the claimants to be done in the same way throughout the respondent’s stores, and

14.2

the training materials stating how that work was to be done were going to be determinative of what the respondent required the claimants to do unless there was cogent evidence before us to show that one or more aspects of those training materials was not (or was no longer) determinative.”

In relation to the comparators, the tribunal considered that there was even less need for oral evidence because of the respondent’s digital systems [16] and the arm mounted computers (“AMCs”) used by the comparators in the DCs. The tribunal then went back again, for a further time, and discussed the meaning of “work” and some of the authorities previously discussed citing Prest, and Beal with approval describing them as “entirely consistent” with Brunnhofer v Bank der Österreichischen Postsparkasse AG [2001] 3 CMLR 9[19]-[25].

It is explained in the Second Judgment that the conclusions on the factual issues for the purposes of s.65(6) EqA 2010 of the six sample claimants and eight comparators are set out in the eight schedules or appendices to the judgment. Each claimant has her own appendix. The tribunal’s determinations in relation to the eight comparators are in a single appendix, 8. It was not set out in the way that the IEs had requested as the tribunal acknowledged:

“46.

Unfortunately, we were unable to arrange for, or do, what the IEs wanted in that regard. We regret it very much. However, we were by the time of writing these reasons (having made all of our determinations of fact that we thought were required) of the view that there were before us many factual disputes about things which the IEs would not be taking into account, either because they were about things which were outside the scope of the IEs’ inquiry or because they were about something which was trivial. In addition, we thought that the only reliable way in which the IEs could do their job was to work through the tasks which we concluded, in the manner stated in Appendices 1, 2, and 4-8, were required by the respondent to be done, using such assessment method as the IEs thought was right, and then prepare a report by reference to the factual findings that we had made. The reason why we were in that situation was the manner in which the parties had approached the case, and there was, we emphasise, in our view no alternative or easier way for the IEs to proceed.” [46]

The tribunal was aware that the IEs required clarity and certainty so that there were no parts that would be left for the IEs’ interpretation which would lead to further disputes between the parties.

It is not necessary to set out a lengthy summary of the tribunal’s Second Judgment because much of it is not relevant to this appeal and because the parts relevant to each of the grounds of appeal are discussed below under the applicable ground of appeal. But it is necessary to set out a number of their conclusions and general observations about the evidence. All parties were criticised for using the EVJDs “as vehicles for in some cases wide-ranging assertions about things which we conclude were irrelevant” which was described as “a major complication” [45]. The tribunal struggled with the relevance of some of the facts agreed between the parties, also with facts that, although agreed, appeared to be unlikely to be correct, illogical or highly improbable to the tribunal [48].

The tribunal’s solution to what it considered to be the shortcomings in the approach of all parties was to:

“52.1

first analysing (with care and at some length with a view to avoiding doubt and minimising the possibility of future disagreements) what was said in the documents before us which recorded, or stated, the training provided by the respondent to employees doing work of the sort which it was claimed the sample claimants and their comparators did, and then

52.2

going through the parties’ evidence relating to the way in which the relevant employee did his or her work. Only if that evidence added anything relevant to what was in the training materials did we take it into account in deciding what was the work of the employee for the purposes of section 65(6) of the EqA 2010.”

Some disputes between the parties were regarded as trivial and “highly unlikely to affect the assessment by the IEs of the demands of the work of the claimant or comparator in question.” [51]. The tribunal explained that:

“54.

We record here that we did not resolve a dispute even where it was about a relevant matter where the resolution of the dispute was in our view highly unlikely to affect the view of the IEs about the demands and therefore the value of a particular task. That was so where in our view the difference between (1) the thing for which a party was contending and which was not agreed and (2) what was agreed, was of no practical importance because even if we had accepted the contention that the factual assertion about the non-agreed thing was correct, that acceptance would have been highly unlikely to affect the IEs’ view on those things.”

The tribunal noted that it had had to do much of the work that had been expected of the parties in the unusual circumstances of the case which meant that the parties had not had the opportunity to address the tribunal on “the things which we ourselves found by our own research” which the tribunal accepted was not the norm and which they would have avoided if they could have. In the interests of justice, the parties were therefore invited to apply for a reconsideration under Rule 71 ET Rules of Procedure 2013 if any of the parties considered that the tribunal had “made a material misunderstanding of the documentary or oral evidence before” itand the IEs were invited to apply to the tribunal for clarification under rule 6(3) EV Rules of Procedure 2013 [55].

The tribunal then explained how it had approached the evidence that had resulted in the eight appendices:

“56 When we worked through the very long sets of closing submissions for each sample claimant and comparator, we started with the work of Mrs Worthington. In doing so, we tried hard to work with the so-called EVJD for her and to adapt it to suit the approach which we concluded had to be taken. We found that it was possible to use it as a vehicle for determining a number of important factual disputes about the work of Mrs Worthington, but by the end of the process of determining those disputes, we had realised that the respondent’s evidence was in some respects markedly inconsistent with its own training materials.

57 We then worked through the parties’ submissions in relation to the work of Ms Williams. In doing so, we went laboriously through the parties’ contentions and the evidence. We found that we could not accept some key parts of the evidence of Ms Williams, and as a result, we had no alternative to considering each factual dispute with care.

58 We then turned to the evidence about the work of Ms Cannon and the parties’ contentions in relation to that evidence. We describe in Appendix 3

58.1

the process which we initially followed in seeking to determine what was the work of Ms Cannon for the purposes of section 65(6) of the EqA 2010, and

58.2

the reasons why we concluded that we had to stop following that process and take a different approach in the determination of what that work was.

59 In Appendices 4-7 we followed what we considered to be the best approach to the determination of the disputes which the parties maintained in relation to the work of Ms Cannon, Ms Thompson, Ms Oz and Ms Garrod.

60 In Appendix 8, we took a different approach. We did so because the work of the comparators was even more extensively described in training materials than that of the sample claimants, and because we concluded that we could best determine what that work was (1) task by task, and (2) by first considering what were the things which the comparators had to do as part of their work for the purposes of section 65(6) of the EqA 2010, using as our factual source the bundles currently before us, and only then, secondly, considering the parties’ factual assertions in relation to those tasks.

61 In the course of making our factual determinations and stating them in the appendices to these reasons, we arrived at some conclusions which were of general application and which we thought it would be helpful to mention here so that they were collected in one place.”

It then set out a number of “Conclusions to which we came about relevant issues of principle after having made all the factual determinations which we concluded at this stage needed to be made” many of which form the grounds of appeal so are discussed in detail below and are not reproduced here in an attempt to minimise repetition.

After the Second Judgment was sent to the parties,on 11 September 2024, the IEs were supplied with a 19,000 page bundle of the documents that had been referred to by the tribunal in the Second Judgment. The IEs’ understandable first reaction was trepidatious, and they were concerned that undertaking their assessments could be very unwieldy and time consuming as they would normally have received a single job description per job holder which would set out both the agreed and decided facts. But the IEs nonetheless agreed to conduct a trial assessment of three or four job factors and see how they could get on. They explained that they would use the two sources of information – the original EVJDs and the findings of fact embodied in the Second Judgment - but as neither of these documents contained factor by factor breakdowns it would require a considerable degree of analysis and cross-referencing of data.

Four months later on 14 January 2025, after the IEs had tried to complete their trial assessments, they circulated a note to the parties and the tribunal. They explained, with worked examples, all the ways in which their work will be hindered and frustrated by there being no EVJDs incorporating the facts found at Stage 2. If required to work with the information as presented to them, their report would (1) contain many caveats concerning the proper attribution of facts found at the Stage two Hearing to the original EVJDs; (2) contain highly complex strings of references, sub-references and sub-sub-references; and (3) take a very considerable time to produce. They explained that the difficulties would be compounded at the Stage 3 Hearing where the cross examination of the IEs and partisan experts (who would also be called by both sides) navigating between different documents and sources would be highly confusing for all witnesses and the tribunal. The IEs suggested that the difficulties could be resolved if the tribunal reconsidered its position and directed the parties to produce final, definitive EVJDs incorporating the facts found at Stage 2 (SB/1304-1315).

The tribunal refused the IEs’ request and did not ask the parties toproduce final, definitive EVJDs incorporating the facts found at Stage 2. Instead the tribunal decided that “it was in the interests of justice… [for the tribunal] …to prepare new statements of the work of each sample claimant and the comparators.” In a decision communicated in an email from the EJ to the parties on 23 January 2025:

Solicitors for the respondent then requested that the tribunal “pause” its work on EVJD preparation and its reconsideration decision pending this appeal. The Harcus Claimants resisted the respondent’s application,whilst the Leigh Day Claimants resisted the application in relation to determination of the parties’ reconsideration applications, but adopted a neutral stance in relation to pausing the EVJD preparation. The EJ continued corresponding with the parties by email including I noticed on a Saturday morning (22 February 2025). The tribunal sought the parties’ views on whether more work needed to be done in relation to the Stage 2 decision before the IEs could start their work (email of 21 February 2025). It also informed the parties that the tribunal would “make firm decisions on the relevance or otherwise of all the factual issues that the Harcus claimants had asked for.”

On 24 March 2025 the EJ again wrote to the parties explaining that the tribunal was “a long way down the road of a careful reconsideration of the factual determinations” in the Second Judgment, in the course of which it said it was “creating new, simplified and concise, statements of our determinations of the work of the sample claimants and their comparators”. The EJ was, once again, critical of the EVJDs that the parties had produced which he described as “confusing, repetitive and argumentative” and effectively criticised the parties and the IEs for framing the template EVJDs without input from the tribunal with a view to making the work of the IEs simpler. He described their structure as buckled wheels which were so unhelpful that they required replacement. He also explained that the tribunal had decided to reconsider its findings of fact in the Second Judgment in a “comprehensive way” and in doing so had found errors in what they had decided previously and he accepted that there were errors requiring correction. They were reviewing the work of all 14 claimants and comparators. The EJ said that he refused the respondent’s request that the tribunal’s work be paused. He considered that the appeal tribunal hearing this appeal would benefit from the reconsideration judgment. In addition he speculated that the issue of the meaning of “work” might need to be heard by the Court of Appeal and even the Supreme Court since he considered that the point of law that had arisen was not the subject of any binding appellate authority, only the High Court and the EAT ([13]).

The parties had taken the tribunal up on its invitation in the Second Judgment to apply for a reconsideration under Rule 71 ET Rules of Procedure 2013. The respondent raised 150 matters for reconsideration in an 88 page document whilst the claimants’ applications were more modest (29 pages in aggregate: 15 pages for the Leigh Day application and 14 pages for Harcus). The tribunal kept the parties informed of its progress as set out in the chronology above and spent a further 54 days without the parties present between 6 January and 23 May 2025. It promulgated a Reconsideration Judgment on 30 May 2025 (“the Reconsideration Judgment”) of 306 pages: 88 pages in the body of the judgment and 14 appendices. The final paragraph of the body of the Judgment explained:

“334.

We are hoping that we have now, in appendices 1-14 below, stated all of the relevant facts for all of the sample claimants and their comparators. We have, however, again made findings without hearing from the parties on our proposed conclusions. We concluded that it was going to be most effective, and most cost-effective, to make those findings without holding a further hearing and instead to give the parties a further opportunity to tell us if they think that if we have erred in making those findings to such an extent that we should reconsider them.”

The parties were given 42 days, or more if reasonably required, to make any further reconsideration applications to the tribunal about the Reconsideration Judgment. The Reconsideration Judgment was promulgated two days before skeleton arguments were ordered to be exchanged in these appeals. The parties were still digesting the contents of the Reconsideration Judgment and considering their options regarding appeal and/or further reconsideration applications in response to the tribunal’s offer. Beyond being taken to a paragraph here and there, I was not addressed on the Reconsideration Judgment during the hearing before me, and I was not invited by the parties to read it in full.

The IEs were also in the process of considering the Reconsideration Judgment. Since the hearing of this appeal the IEs reported to the tribunal and the parties on 7 July 2025 that providing there are no further changes to the documentation, the IEs expect the report to take 12 months to complete and requested the tribunal to set 31 July 2026 for the submission of the report. All parties were content for this additional information to be included and taken into consideration in this judgment.

The respondent is appealing a number of elements of the First Judgment, the CMO and the Second Judgment and there are thus three case numbers, which have all sensibly been combined and listed together. I will use the numbering adopted by HHJ Tayler and the parties for the grounds of appeal which fall into a number of categories:

There were some agreed amendments to the grounds of appeal and some were withdrawn, but the issue of the scope of these appeals was not agreed and the issue is currently before the Court of Appeal. In their grounds of appeal, the respondent gave a non-exhaustive list of examples of what it said were erroneous factual determinations and other errors by the tribunal in relation to a number of their grounds of appeal. A Preliminary Hearing was heard in the EAT by HHJ Tayler on 27 November 2024 to consider whether any of the grounds were reasonably arguable in which, amongst other things, the claimants objected to the open-ended nature of some, possibly all, of the grounds (Footnote: ⁵). HHJ Tayler gave a reserved judgment on 5 February 2025 (“the 5 February 2025 Judgment”) which concluded that all the grounds put forward were sufficiently arguable to proceed to a full hearing “but limited to the matters specifically raised in the grounds of appeal, absent any successful application to amend” [36]. What the respondent wanted to do was to add further examples and particulars in relation to many of its grounds of appeal at the appeal itself. The ratio of HHJ Tayler ruling challenged by the respondent is as follows:

“21.

Section 3.8 of the EAT Practice Direction requires generally that grounds of appeal should be “short and focussed” and “clearly assert errors of law”. Sections 3.9 and 3.10 require that grounds of appeal which assert “perversity” or “procedural impropriety” must provide “full details”. Section 3.10 includes under the heading “procedural impropriety”, “any material procedural irregularity”. That requires particularity of each dispute rather than the provision of “examples”. In any event, I consider where it is asserted that an error of law resulted in erroneous factual determinations it is important that particularity of the errors are set out. Tesco has the resources to provide the required full details and is to be limited to the specific challenges advanced in the grounds of appeal, absent any successful application to amend.” Tesco Stored Limited v Ms K Element & Ors [2025] EAT 45 at [21]

The respondent has been granted permission by the Court of Appeal to appeal the 5 February 2025 Judgment at an expedited hearing listed for 17 July 2025. There has been no application to stay these appeals pending the Court of Appeal judgment and I have decided that the preferred course of action is to decide the issues before me in this appeal, regardless of the date of the Court of Appeal hearing. I am mindful of the observations by Kerr J in another EAT appeal in this case, Tesco Stores Ltd v Element & Ors [2024] EAT 83 about the need to get on with matters and avoid delay in equal pay cases:

“69.

This appeal shows that, in undertaking case management in equal pay litigation, there is nothing wrong with pressing ahead and forcing the pace; patience is not a virtue in equal pay litigation. The right of access to a tribunal and a remedy for any wrongs done within a reasonable time requires a continuing sense of urgency and momentum.

I have therefore only considered the examples provided by the respondent in their grounds of appeal. I have not considered the additional matters that the respondent hopes that the Court of Appeal will allow them to raise, even though many of them are referred to in Mr Reade’s skeleton argument: I will not go behind HHJ Tayler’s judgment in Tesco Stores Limited v Ms K Element & Ors [2025] EAT 45.

To complete the chronology, following HHJ Tayler’s ruling of 5 February 2025, the respondent applied to the EAT for permission to amend the grounds of appeal to add “further erroneous factual determinations, facts, examples and matters on which the Respondent relies” which were set out in a schedule that listed 146 passages. Their application was refused by the EAT (Tesco Stored Limited v Ms K Element & Ors [2025] EAT 43). There has been no further appeal from that amendment application decision. Mr Reade explained that whilst the respondent disagrees with the judgment, it recognises that it was a case management decision exercised within the judge’s discretionary powers and was not susceptible to appeal.

Before diving into the detail of each of the grounds of appeal it is first necessary to set out some of the ET Rules of Procedure 2013 to understand the scope of the tribunal’s powers. The overriding objective to enable tribunals to deal with cases fairly and justly includes, so far as practicable –

“(a)

ensuring that the parties are on an equal footing;

(b)

dealing with cases in ways which are proportionate to the complexity and importance of the issues;

(c)

avoiding unnecessary formality and seeking flexibility in the proceedings;

(d)

avoiding delay, so far as compatible with proper consideration of the issues; and

(e)

saving expense.

A Tribunal shall seek to give effect to the overriding objective in interpreting, or exercising any power given to it by, these Rules. The parties and their representatives shall assist the Tribunal to further the overriding objective and in particular shall co-operate generally with each other and with the Tribunal.” (Rule 2 ET Rules of Procedure 2013)

Rule 41 provides:

“The Tribunal may regulate its own procedure and shall conduct the hearing in the manner it considers fair, having regard to the principles contained in the overriding objective. The following rules do not restrict that general power. The Tribunal shall seek to avoid undue formality and may itself question the parties or any witnesses so far as appropriate in order to clarify the issues or elicit the evidence. The Tribunal is not bound by any rule of law relating to the admissibility of evidence in proceedings before the courts.”

Since the challenge is partly based on the adequacy of the tribunal’s reasons it may be useful to go back to a few first principles. Reasons must be given for all decisions on disputed issues, proportionate to the significance of the issues and the rules are expressed to be prescriptive as to the specific matters to be covered (Rule 62 (Footnote: ⁶)). However, the case law has consistently held that provided there has been substantial compliance with the rule, a tribunal will not have erred in law by not treating the rule as a straightjacket (see for example Balfour Beatty Power Networks Ltd v Wilcox [2006] EWCA Civ 1240 and the line of cases such as Greenwood v NWF Retail Ltd [2011] ICR 896 (EAT)). The duty is to give 'adequate and intelligible' reasons (Anya v University of Oxford [2001] IRLR 377 at [381], Sedley LJ), or, as it was put in Clark v Clark Construction Initiatives Ltd [2008] EWCA Civ 1446, “candid, intelligible, transparent and coherent” reasons.

English v Emery Reimbold & Strick Ltd [2002] EWCA Civ 605 is as applicable in the tribunal as it is in the civil courts: the judgment must enable the appellate court to understand why the judge reached his decision, but not every factor which weighed in the judge’s appraisal of the evidence needs to be identified and explained ([19]-[21]). Furthermore, since the “reasons of an Employment Tribunal are directed towards parties who know in detail the arguments and issues in the case.”, they do not need to be spelt out in the detail required, “were they to be directed towards a stranger to the dispute.” Derby Specialist Fabrication Ltd v Burton [2001] IRLR 69.

The case law has repeatedly stressed the importance of not taking an unduly legalistic and analytical approach to employment tribunal reasons when considering whether they disclose an error of law: Union of Construction, Allied Trades and Technicians v Brain [1981] IRLR 22, [1981] ICR 542, per Donaldson LJ:

Twenty years later Sedley LJ made the same point in Anya:

“26.

… The courts have repeatedly told appellants that it is not acceptable to comb through a set of reasons for hints of error and fragments of mistake, and to try to assemble these into a case for oversetting the decision. No more is it acceptable to comb through a patently deficient decision for signs of the missing elements, and to try to amplify these by argument into an adequate set of reasons. Just as the courts will not interfere with a decision, whatever its incidental flaws, which has covered the correct ground and answered the right questions, so they should not uphold a decision which has failed in this basic task, whatever its other virtues.”

More recently, Eady J in Godfrey v Natwest Market [2024] EAT 81, [16] held “…where an ET has fallen into error, the EAT’s role is not to strive to uphold a decision where the reasoning reveals a fundamental error of approach”.

The first tier tribunal is a fact finding tribunal, its role being to find the facts in dispute relevant to a claim. This is especially so at a Stage 2 Hearing: rule 6(1)(a) states that its only purposes are to make a determination of facts in dispute and fix a date for the Stage 3 Hearing. Appellate bodies are rightly slow to interfere with a tribunal’s findings of fact, absent perversity or an error of law. Perversity is a high bar and “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” (Crofton v Yeboah [2002] IRLR 634 (EAT) Mummery LJ).

Happily the parties are all agreed as to the applicable law that the tribunal should have applied.

Section 64(1) EqA 2010 provides that the equal pay provisions apply when: “a person (A) is employed on work that is equal to the work that a comparator of the opposite sex (B) does” and the definition of work of equal value in s.65(6) EqA 2010 is set out in paragraph 1 above. At this stage of the case we are not concerned with the value of the work of either a claimant or comparator, but the role of the tribunal is just to decide the facts of the work. The evaluation of the work comes later. “Work” is not defined in that section and the parties now agree that the claimants’ joint note provided to the tribunal in October 2023 is accurate:

“CLAIMANTS’ JOINT NOTE ON THE LAW

1.

Principles which emerge from the legislation, rules and case law on the task of the Employment Tribunal at a stage 2 equal value hearing are:

a.

At stage 2, the Tribunal will determine disputes (of fact and relevance) to establish the “facts relating to the question” of whether the claimants’ “work” is of equal value to that of the comparators. (Footnote: ⁷)

b.

The facts that need to be determined in order to go on, at stage 3, to answer “the question”, are those that relate to what “work” the individual claimants and comparators did. (Footnote: ⁸)

c.

The starting point is “what is done in practice” by the claimants and comparators (the jobholders) – i.e. the jobholders’ “observed activities not their notional paper obligations”. (Footnote: ⁹) The Tribunal must “look at what the employee actually did, not simply at the documents (such as […] work manuals)”. (Footnote: ¹⁰)

d.

All three of the following different categories of things constitute (part of) the “work” of a jobholder: (Footnote: ¹¹)

i.

things that the jobholder is instructed, requested or encouraged to do (even if those things are only done very infrequently or did not actually occur during the Evaluation Period);

ii.

things that the jobholder does as a matter of practice which are simply a way of doing something which is part of their work (regardless of whether their managers know what they are doing); and

iii.

things which are done with the knowledge and explicit or implicit acquiescence of their managers.

e.

Whether or not something is a “requirement or expectation” of a particular role is a relevant consideration but is not a necessary condition in order for that thing to form part of the jobholder’s “work”. (Footnote: ¹²)

f.

Documents such as “work manuals” are “relevant, but not necessarily determinative” of what constitutes the “work” of a jobholder. (Footnote: ¹³)

g.

Training requirements and working conditions are also potentially relevant matters regarding which the Tribunal may make factual findings. (Footnote: ¹⁴)

h.

In relation to the matters at (a) to (g) above, the Tribunal is entitled to determine the level of detail which is relevant to “the question” and to establish only such facts as it considers to be relevant to “the question”. (Footnote: ¹⁵)

2.

These principles are not intended to be exhaustive.”

I agree with the parties that the claimants’ note is an accurate summary. Mr Jones boiled down the definition of “work” still further into the single proposition of it being “what the employer requires in practice”. It is thus not a binary question of only what happens in practice, or what the employer requires, but a combination of the two. Mr Bryant accepted Mr Jones’ proposition, which I agree was an uncontroversial and accurate distillation of the case law. In light of their agreement it is not necessary to dwell in any detail on the law and little more need be said about it. Suffice to say that the parties were also agreed that Shields remains good law. The parties are also all agreed that if the tribunal approached its fact-finding task as requiring consideration of generic jobs, rather than consideration of the 14 job holders’ work, it would have erred and A1G2 must succeed. The footnote in Prest which was a case on amendment does not mean that the work of specific claimants and comparators need not be considered and determined. It was also common ground that if the tribunal had approached the evidence on the basis that the training materials were conclusive evidence of the claimants and comparators’ work, thus raising an irrebuttable presumption, that too would have been an error of law (A1G2). The tribunal should take into account relevant evidence on the issues before it, whether oral or documentary and relevant evidence must not be disregarded in principle. The claimants’ response to these two grounds of appeal was that the tribunal fell into neither error.

Dealing firstly with A2G2, the tribunal’s approach to the issue of what constitutes work, the questions in A2G2 are firstly, is it possible to discern how the tribunal directed itself and if so, what definition of work did the tribunal actually adopt. On these questions, the parties did not agree.

As set out in the chronology, the tribunal repeatedly set out its views on the meaning of “work” in a multiplicity of places: in the First Judgment, the CMO, the Second Judgment and in email correspondence on 27 October 2023 and 9 November 2023 and the Reconsideration Judgment ([51]), and on each occasion using various iterations and sometimes different forms of words and went back again and again to discuss the case law. It was not helpful for it to return repeatedly to the same issue when it had already given its judgment. It makes it hard for both the parties and an appeal body to know which iteration is intended to be the definitive one and how to deal with any inconsistencies as between one document and another. As could easily have been predicted, all sides have quoted selectively from the range of options open to them to support their respective submissions.

For the respondent, Mr Reade submitted that the tribunal erred in proceeding on the basis that “work” means the activities that the employer generally expected or required persons holding a given position to do. Its error is contained in a number of places he submitted. It is implicit in the tribunal’s dismissal of the respondent’s submission that the Stage 2 Hearing “…had to be about what the sample claimants and their comparators in fact did on a day-to-day basis, not what the respondent required them to do as their jobs” [3] and as explained at greater length in paragraphs [9]-[26]. The error was repeated in [11] of the CMO when the tribunal implicitly acknowledges its unorthodox interpretation of the law, stating that its “conclusion was based on its own interpretation of section 65(6) EqA 2010, reading it against the background of the equal pay legislation as we understood it.”

In paragraphs 11-13 of the Second Judgment the tribunal was wrong to hold that Shields could be distinguished, and did not apply in equal value claims which in any event contradicted what the tribunal had said about Shields in its First Judgment at [12]. The tribunal then wrongly concluded that its approach was entirely consistent with EU law. It quoted paragraph 48 of Brunnhofer however, it (a) selectively quoted from that paragraph (and, in doing so, substantially altered the meaning of the text) and (b) omitted to quote or apply the passages from paragraphs 42 and 43 of Brunnhofer which referred to Macarthys Ltd v Smith [1980] ICR 672, ECJ which held that the EU case law was “clear” and equal pay comparisons are “exclusively concerned with the nature of the work actually performed”.

The tribunal itself considered that its interpretation of the law was different to that of the parties which explained why the tribunal had wanted the whole case reformulated by the parties. The tribunal acknowledged as much in [51] Reconsideration Judgment when it announced that if the tribunal’s understanding of the law was correct and that:

“…the work had to be determined by reference to what the employer in fact (i.e. as a matter of fact, found by us) required of the employees, and not simply what they actually did, then the parties’ representatives had asked the parties the wrong question. If you ask the wrong question then you are likely to get the wrong answer.”

In summary the respondent submits that the tribunal’s interpretation of “work” for the purposes of equal value was profoundly wrong, contrary to the submissions of all the parties, a heterodox position unsupported by the well-established case law and this error has infected the tribunal’s entire approach to all its findings of fact, such as they are. Having misunderstood its task, the tribunal then wrongly elevated the status and importance of training documents as a basis for its factual findings and looked at generic work, not the specific work of the claimants and their comparators.

The claimants accept that they initially pushed back on the tribunal’s apparent interpretation of the law set out in the First Judgment and had tried to assist the tribunal with their Joint Note set out above that was prepared after the First Judgment was promulgated. Actually however, although the tribunal had expressed itself a little differently, and even though it itself may have considered that it had come up with a novel approach, in fact it had done no such thing. This could be seen from all of the tribunal’s utterances when read fairly as a whole, but the respondent had been very selective in the extracts it had put before the EAT and the agreed formulation in the Joint Note and Mr Jones’ further distillation of the principle is the law that the tribunal had in fact applied.

A preliminary question I have considered is if it is possible to understand the direction that the tribunal has given itself as to the meaning of work from all its many pronouncements. I am satisfied that it is possible to gain a clear sense of what the tribunal is saying, and, if read in a fair minded way as a whole, there is a coherent and consistent approach, which is not so riven with internal contradictions as to lack discernible meaning.

The next question is whether it has accurately articulated what is in fact an uncontroversial area of law and is the parties’ agreed understanding of what the law is. I conclude that although, in places it may appear that the tribunal’s focus on the case of Beal has elevated the importance of what an employer requires an employee to do at the expense of what an employee does in practice, this is not what the tribunal has actually done, as a careful reading shows. Mr Reade is right that the startling and clumsily expressed first pronouncement in the First Judgment at [3] stating that the respondent was “wrong” to submit that “the stage 2 hearing had to be about what the sample claimants and their comparators in fact did on a day to day basis, not what the respondent required them to do as their jobs” was an unpromising start. But on a reading of the entire First Judgment it was not throwing out 50 years of equal pay orthodoxy. Mr Reade also points to [11] of the CMO set out in paragraph 33 above and says it is here where the tribunal’s error is clearest: as it has drawn a hard line and distinction between the question of what are the employer’s requirements of the job-holder in the comparator’s job and how the job holder in practice did that job. The tribunal considered the former to be the correct approach and the latter the wrong one, on the authority of Prest. It inevitably meant that the factual conclusions could not be relied on as the tribunal had asked itself the wrong question. One can see why the parties were alarmed when they first read the First Judgment and the claimants prepared their Joint Note to assist the tribunal. But my reading of the tribunal’s conclusions from its discussion of the case law is that it has clearly read and cited the leading authorities, understands the importance of Brunnhofer in its entirety and expressly stated that it was following Brunnhofer. It set out the law in laborious detail. The tribunal’s focus on Beal, quoted at some length by the tribunal, explored the boundaries of what constituted work within the confines of established case law, and memorably coined the “loafing” test as an example used to describe:

“….something [that] may not be part of an employee’s work if they have not been instructed, requested or encouraged to do it, their doing it has not been approved by their employer and it does not simply constitute a way of doing something which forms part of their work.”[32(4)]

The point being that if an employee is in practice loafing about all the time and that is not behaviour condoned, known of or encouraged by the employer, it does not become part of the employee’s work, just because that is what, in practice, he is doing. But in its concentration on Beal, the tribunal did not inadvertently lose sight of the basic principles. Thus when the tribunal stated in the CMO that it was important “to focus on what work the employer in practice required persons who were in the role of the comparator to do, rather than the way in which the comparator actually worked.”[11] the tribunal is doing no more than referring to what was memorably described during the course of our hearing as the “loafing limit”. The “loafing limit” marks the boundary between “work” and what might in other contexts be described as a frolic of one’s own, or doing something that might elicit a response from an employer when they find out what the employee is up to of: “I don’t pay you to [e.g. loaf around] all day, but to get on with the work in the way that I have told you how to do it. Now get back to work.” The tribunal has just used a slightly melodramatic way of expressing itself but was following the orthodoxy.

Nor do any of the other documents, after the First Judgment, when the tribunal repeatedly went back over the same ground demonstrate that the tribunal had misunderstood the meaning of “work”, such as the CMO, the Second Judgment, the Reconsideration Judgment at [51] and all the emails whether read independently from each other or when taken as a whole. The tribunal wording is in places infelicitous, for example, in the Second Judgment under the heading: “Our further reasons for our conclusions on the correct approach to take” in paragraph 9, set out in full at paragraph 48 above, but the tribunal was merely summarising Beal [32]- [33] which is worth setting out in full beyond the selective quotation in paragraph 91 above:

“32.

Of course, where an employee is contractually required to do something (and that requirement has not fallen into desuetude or otherwise been varied), then that activity will form part of their work (even if, in practice, they neglect or refuse to perform it). But most of the issues in the present case concerned activities where the contractual position was not so clear-cut. On the whole, the dispute was not as to what the employee did, but as to whether it formed part of their work. I will deal with the individual issues later, but it may be helpful to set out in general terms what seems to me to be the appropriate approach. In general terms, therefore:

(1)

Where an employee is instructed by their manager to do something, then, if they do it, that is surely part of their work. Moreover, that is so, even if they might have been entitled to say, “But that is not something I am obliged to do.”

(2)

The same is likely to be the case where the manager does not instruct, but requests or encourages, the employee to perform the activity in question. On the other hand, in such a case, it may be relevant to note for the expert’s benefit (if it is the case) that the employee could not be required to perform that activity.

(3)

Where an employee does something which they have not been instructed, requested or encouraged to do, it may still constitute work if, for instance:

(a)

it is simply a way of doing something which forms part of their work; and/or

(b)

their manager knows that they are doing it, but does not object and thereby tacitly approves of their doing it.

(4)

On the other hand, something may not be part of an employee’s work if they have not been instructed, requested or encouraged to do it, their doing it has not been approved by their employer and it does not simply constitute a way of doing something which forms part of their work.

33.

I stress that these are merely general considerations, which are not intended to place a gloss on the Act and that each disputed issue has to be considered on the basis of its own particular facts.”

It is worth noting that the tribunal quoted both these paragraphs from Beal in full in its First Judgment

What Mr Reade is right to be puzzled by is why the tribunal considered its understanding of the law

Once again there was no dispute as to the law, which is that it is the work of a specific claimant and a specific comparator that is to be considered, not a generic employee. The claimants accepted that if the tribunal had approached the matter by reference to a generic employee it would have erred in law. The question was whether they had done that. Once again it is necessary to keep focussed on the ratio of the tribunal’s findings and not become distracted by what are obiter comments and musings.

Mr Reade relies on a comment in the Second Judgment that the claims were made by thousands of “customer assistants” and it would be “very surprising” if the value of their work differed according to which employee performed the job [18]. He also relied on the tribunal’s emphasis and reliance on Prest v Mouchel Business Services Limited[2011] ICR 1345, (in particular endnote 4 in the judgment of Underhill LJ (see [26] First Judgment)) in various places. Mr Reade drew particular attention to [19] of the Second Judgment where the EJ notes that none of the parties had referred him to Prest and nor had the editors of the five volume loose leaf leading textbook Harvey on Industrial Relations and Employment Law. With tremendous confidence the tribunal considered that “rather than seeing [the omission] as a reason to doubt the correctness of our understanding about the manner in which the parties should have approached the case, we regarded it as an omission on the part of the editors of that work” [19].The confidence was misplaced. Prest was, as Mr Reade described, concerned with a technical pleading point about a change of comparator part way through proceedings and whether the substitution of one comparator for another constituted a new claim for the purposes of determining the “arrears date” under what was then s.3ZB(3) Equal Pay Act 1970 and whether the arrears of remuneration were to be calculated from the date of the claim lodged with the original comparator or the new one. In Prest Underhill P wondered in an endnote whether the naming of a comparator might not always be necessary in a collective claim, but it has no relevance to the issues in the Stage 2 Hearing in this case because comparators had been named. It was a red herring.

There were slim pickings for Mr Reade to construct an argument that from the tribunal’s enthusiasm for Prest and other essentially throw away comments it could be deduced that the tribunal had not considered the individual claimant and their comparators’ work. The tribunal was focussed on what the individual claimants and their comparators did in the 619 pages of the Second Judgment which made reference to the particular evidence and facts of each of them, albeit some were arrived at by considering the training materials. As the tribunal acknowledged, how the equal value conclusions on the sample claimants and comparators in the first of the three tranches would roll out to others amongst the 55,000 claimants was for a later stage. It was unhelpful of the tribunal to speculate that its eventual conclusions at a Stage 3 Hearing for the sample claimants in this part of the case and their comparators might have implications for other of the claimants, because it was merely speculation, and formed no part of the tribunal’s task at the Stage 2 Hearing for the first tranche of sample claimants. The tribunal’s job was only to find facts on which the question of the relative value of the jobs could then be based. But it was not more than a whimsical aside and digression by the tribunal. It did not demonstrate a fundamental misunderstanding of its role. As to the criticism of [18] Second Judgment this was also misplaced: here the tribunal was reminding itself of a long-established equal pay principle that the fact that some employees are better at the job than others is not part of the equation.

Ground A1G1 is dismissed.

Once again the parties were agreed that if the tribunal had said that it would close its mind to any evidence about the work of the claimants and the comparators, apart from the training materials for the various roles, the tribunal would have erred and it would have infected all the findings of fact about the work done by them. The question, once again, is whether that was how the tribunal had approached the evidence.

There is force in the Leigh Day claimants’ argument that, faced with an overwhelming amount of material, the tribunal was trying to find the best place to start when analysing the evidence: should it be the EVJDs, or the witness statements of the claimants, or the comparators, or the respondent manager witnesses or the documents and if so, which ones? It was a fair question given the volume of material.

The tribunal had identified early on that they were interested in the documents of in-person training given to the claimants and the comparators and also the documents that they were required to read and apply (see paragraph 18 above).

As explained in the First and especially the Second Judgment the tribunal found the EVJDs and RoD hard to navigate and untangle. They found the EVJDs repetitive; “the words of each of them differed in a number of ways even where they purported to describe what we regarded as the same work” ([43] Second Judgment); the tribunal was frustrated by the witness and EVJD evidence including matters that were outside the scope of the IEs’ inquiry ([46] Second Judgment); irrelevant material ([46]); evaluative material that was wrong to be included; and other matters that were trivial ([51]). In the First Judgment the tribunal quoted extensively from the Leigh Day submissions on the generic flaws in the respondent’s EVJDs for comparators and their alternative wording for the disputed parts of the claimants’ EVJDs, seemingly with approval.

The tribunal’s solution to the best way into the evidence, given that it frankly considered some of the parties’ witness evidence to be unreliable in places, was to start with the training materials required to be followed and applied by the claimants and the comparators. The tribunal concluded that on the specific facts of this case of this employer and their claimant and comparator employees, the very comprehensive and prescriptive training materials that dictated every part of the work of all 14 employees whose work they were considering, was a good starting point. It was open to it to do so, especially given its findings of fact at [13.5 and 14 Second Judgment] set out above at paragraphs 49 and 50. It was a rebuttable, not an irrebuttable presumption, from which to start. Where the evidence from any of the witnesses, EVJDs or other materials demonstrated that the training materials did not give an accurate picture, the oral or contradictory documentary evidence prevailed.

At no point did the tribunal say that its’ exclusive focus was on training materials and it explained why it had approached the evidence in the way that it had, see for example [37] and [39] CMO: training materials would be considered first, oral evidence secondly, or primarily if the oral (by which I understand the tribunal to mean witness, rather than documentary) evidence concerned a matter which was not covered by any of the training materials. Given the shortcomings identified in the respondent’s evidence in particular it was a perfectly sensible way to proceed, well within the tribunal’s powers and discretion.

Everyone is agreed that the tribunal was wrong to say that Shields could be distinguished from the present case because Shields was a case which concerned like work, not work of equal value ([13.1] Second Judgment) and in doing so contradicted its conclusion in the First Judgment at [12]. But the respondent cannot profit from this error. There are insuperable difficulties with Mr Reade’s argument. Firstly the different form of equal work under consideration in Shields was not the only or even the principal reason why the tribunal considered that Shields was of limited assistance. Its reasoning goes back to its analysis of Beal and the loafing limit and the factual differences in the two cases, which led it to the conclusion that the training documents were very useful in understanding what work the claimants and the comparators were required by the employer to do in practice in this case. The tribunal engaged with and analysed the evidence of the very highly regulated and controlled environment, and the desire of the very large respondent employer to achieve uniformity of service and consistency by requiring employees to follow the detailed and extensive training materials across all its stores and DCs.

It is classically for a fact finding tribunal such as this one to decide what weight to attach to the various forms of evidence it had before it, which is what this tribunal did. In stating that the training materials “were likely to be the best evidence of what the job in each case entailed” ([75] First Judgment) it then looked carefully at all the other evidence. It took over a week to go through the evidence and all the submissions of just one of the claimants, Mrs Worthington ([89] First Judgment) it concluded that the training materials were relevant to the job. They had not closed their mind to the value of witness evidence and the EVJDs.

Mr Reade prays in aid Volpi v Volpi [2022] 4 WLR 48 and he is right to do so as it is a key authority. However, from the tribunal’s judgments it is apparent that it followed all the key principles identified in [2] of Volpi including 2(iv) and considered the material evidence and chose what weight to give it. As stated in principle 2(iv) the tribunal did not need to discuss all the material evidence in the judgment itself.

Ground A1G2 is dismissed.

The serious procedural irregularity set out in this ground of appeal is said to be the tribunal’s decision after the Stage 2 Hearing to make determinations primarily or exclusively by reference to training materials and other similar documents, unless there was cogent evidence to show that the training materials were not determinative. This is a separate, but related point to A1G1 and it does not necessarily fail simply because there was no error in the tribunal’s decision to see the training materials as important evidence as explained above. The criticism is that the parties had prepared the case in accordance with the earlier made directions and the tribunal’s view of the importance of the training materials wrong-footed the respondent: the effect was that they did not present their evidence, cross-examine the other side’s witnesses, or make submissions on the basis on which the tribunal (after the event) has now decided the case. The parties had prepared the case on a wholly different basis to the way in which the tribunal decided to approach the evidence in the case.

The respondent argued that the tribunal acknowledged that the parties had prepared the case on a different basis to the one on which the tribunal would have preferred them to have done (see [4] First Judgment). Their initial decision to ask the parties to prepare the case differently demonstrated that it had appreciated the disadvantage and unfairness to the parties from their evidence and submissions not having addressed the matters that the tribunal thought ought to have been addressed. The tribunal also expressly noted that the parties had not addressed it on some of the documents that the tribunal considered to be important (for example [85] First Judgment, [10] Second Judgment). When the tribunal then changed its mind and decided in its words “to get on with the job” the unfairness crystallised. It was pervasive and fundamental to the decision-making process, incapable of being cured, Mr Reade submitted.

The claimants argued that the issue of the importance of the training materials had been flagged up early on. The claimants had always made clear that they relied on the training materials as the best evidence of what the sample claimant and comparators’ work was, whilst the respondent’s position on the evidence was that the ways of working set out in the training materials did not necessarily reflect what happened in the real world and that much of the training and policy material merely set out objectives and aspirations. The IEs had also identified the importance of training materials. The respondent lost the argument before the tribunal, but that was not on account of any legal error, but on the tribunal’s entitlement to consider the evidence in the way that it did and make the findings of fact it was entitled to reach. There was no procedural irregularity in this regard, serious or otherwise: the tribunal had simply preferred the evidence and submissions of the claimants. Analysis and conclusions on ground A2G3

There is no doubt that a serious procedural irregularity can amount to an error of law if there is, firstly, a procedural irregularity, which, secondly, is serious and thirdly it must also be established that it “vitiates the judgment” or, in terms of the CPR which has been approved as also being the appropriate test in this tribunal, that it would be unjust not to allow the appeal (see NHS Trust Development Authority v Saiger [2018] ICR 297 at [81]).

There are a number of observations to be made about the tribunal’s case management, but none are relevant to this ground of appeal. It is important to focus precisely on the respondent’s specific ground of appeal.

It was apparent from an early stage in the litigation that the training materials as evidence of the claimants and comparators’ work would be highly relevant and there were issues between the parties about them. For example, at a PH on 19 December 2022 the relevance of training materials was considered and in the case management summary that followed the hearing the tribunal thanked Mr Jones for pointing out that it might be possible to deduce what work the claimants did and the skills required to do that work from the training that the claimants had received.

It is also apparent from the extracts from the transcript of the Stage 2 Hearing in the Supplementary Bundle. On 23 May 2023 the EJ raised the possibility that the tribunal might consider training documents to be a useful and accurate source of information about the work that the claimants and comparators did – see for example p.17 line 21-24, and the EJ stated that Mr Epstein KC then appearing for the respondent would have the opportunity to address the tribunal on the point. Mr Epstein then does exactly that (p.136 (internal numbering of the transcript, (SB 127) line 18)) and he refers to the submissions he has already made, with his misgivings about whether the training materials are evidence of the job done, and his view that the training each of the claimants and the comparators actually received was more reliable.

So the argument that the respondent has somehow been taken by surprise by the tribunal considering training materials to be important is hopeless. The other point overlooked by the respondent is the way in which the tribunal expressly gave the parties the opportunity to apply for a reconsideration on any parts of the evidence upon which they had not addressed the tribunal during the hearing and the respondent availed itself of that opportunity. The tribunal has also invited the parties to apply for further reconsideration of its Reconsideration Judgment. Tribunals have a wide power to regulate their proceedings in accordance with the overriding objective (rule 41). I do not consider there to have been a procedural irregularity in the way alleged by the respondent, but even if there had been a procedural irregularity of some sort in this regard, giving the parties a further right to make representations would have remedied any unintended mischief, and the tribunal has acted well within its wide power to regulate its own procedure.

The more subtle argument that the tribunal was wrong to conclude that generic training materials had any evidential value to the work of the individual sample claimants and their comparators is a side wind challenge to the tribunal’s conclusion on the meaning of “work” in s.65(5) EqA 2010 and appeal grounds A1G1, A1G2 and A2G2 dealt with above. What the tribunal did, for the reasons it explained at [13.5] and [14] Second Judgment cited in full above, was to decide that because the generic training materials were so thorough, so prescriptive in places and proscriptive in others and so uniformly given to all employees of the sample claimants’ and their comparators job titles, and they were required to follow them, that they were a reliable guide to the way in which they had done their work. The tribunal’s wording is careful and precise – it is the training documents from in-person training given to the sample claimants and comparators and the documents they were required to read and apply. This is not just any old training document that no-one is ever expected to read or follow. It was not a perverse finding of fact.

The rather unusual way in which the case progressed does not provide a firm launch pad for a procedural irregularity challenge for the respondent because it disagrees with the tribunal’s findings of fact. The tribunal decided in the First Decision (as it was entitled to do) that the training materials were key. It then had the idea that the parties might recast the EVJDs for the reasons it set out in the First Judgment. All parties were agreed that they did not want to do that. Somewhat ungraciously the tribunal backed down and said they would do the best they could on the evidence as it had been presented. The tribunal gave the parties the opportunity to double check that there had been full disclosure of the training materials in making the orders that it did in the First Judgment. The parties confirmed that there had been full disclosure.

The tribunal’s decision that the training materials were, unless contradicted by oral evidence, the best and most reliable evidence of the work of the sample claimants and their comparators, was a decision that was entirely open to the tribunal to make as a finding of fact. They set out in some detail the number of ways in which they found the respondent’s evidence unreliable (as detailed above). It is commonplace for courts and tribunals to consider that contemporaneous documentation not compiled for the purposes of litigation is more reliable evidence than fallible memories, especially where, as here the claimants, comparators and respondent’s managers were being expected to remember quite detailed work processes (which might not be very memorable) over a twelve year period.

Appeal ground A2G3 is dismissed.

Two grounds have been grouped together as being concerned with the tribunal’s approach to the task required of it under r.6 EV Rules of Procedure 2013: A2G1 - not determining facts and not accepting agreed facts and A1G8 – relevance. These linked grounds require detailed consideration of the EV Rules of Procedure 2013.

Rule 6 provides:

“6.

— (1) At a stage 2 equal value hearing the Tribunal shall—

(a)

make a determination of facts on which the parties cannot agree which relate to the question and shall require the independent expert to prepare the report on the basis of facts which have (at any stage of the proceedings) either been agreed between the parties or determined by the Tribunal (referred to as “the facts relating to the question”); and

(b)

fix a date for the final hearing.

(2)

Subject to paragraph (3), the facts relating to the question shall, in relation to the question, be the only facts on which the Tribunal shall rely at the final hearing.

(3)

At any stage of the proceedings the independent expert may make an application to the Tribunal for some or all of the facts relating to the question to be amended, supplemented or omitted.

(4)

The Tribunal shall give the parties reasonable notice of the date of the stage 2 equal value hearing and the notice shall draw the attention of the parties to this rule and give notice of the standard orders in rule 7.”

As noted earlier, “the question” referred to in rule 6 is defined in rule 1(1) and “means whether the claimant’s work is of equal value to that of the comparator.”

The criticism of the tribunal in A2G1 is that it failed to determine all the facts which the parties had not agreed (“the Disputed Facts”) in three ways. The first was in stating that it had determined a Disputed Fact in an appendix, but not then saying what its determination was, instead referring to a number of documents from which the respondent stated it was unclear what the finding might be. The example given was in [83] of Appendix 1 Second Judgment “Mrs Worthington’s job description as determined by the tribunal” in which it is explained that the manner in which the JH [job holder] was required to do the work of replenishment “is stated in the following training documents, with page references if only a particular part was relevant to our determination” and then listed 12 documents but said only a few words about each of them. For example 83.5:

“83.5

C7/234 which contained a number of statements about safe working practices in the warehouse.

83.6

C7/232, which reinforced the importance of the Cold Chain policy when helping to unload a delivery and when replenishing. Pages 1-4 were all relevant to the JH’s work.”

The second alleged error was the tribunal’s decision not to determine some Disputed Facts, even if relevant, if it was considered by the tribunal to be “highly unlikely to affect the view of the IEs about the demands and therefore value of a particular task” ([54] Second Judgment), the alleged error being that at Stage 2 it is only the IEs that can make that assessment. The tribunal set out its reasoning as follows:

“54.

We record here that we did not resolve a dispute even where it was about a relevant matter where the resolution of the dispute was in our view highly unlikely to affect the view of the IEs about the demands and therefore the value of a particular task. That was so where in our view the difference between (1) the thing for which a party was contending and which was not agreed and (2) what was agreed, was of no practical importance because even if we had accepted the contention that the factual assertion about the non-agreed thing was correct, that acceptance would have been highly unlikely to affect the IEs’ view on those things.”

The third strand to this ground of appeal criticises the tribunal for stating in its Second Judgment that a fact was “obvious” but not then setting out what the “obvious” fact was, so it was unclear whether the IEs should take it into account.

“66.

Accordingly, we concluded that any statement in an EVJD to the effect that the jobholder had to exercise skill and care merely stated the obvious and did not need to be there (although the impact of the need to take care had to be taken into account by the IEs). A good example of such a statement was paragraph 6.219 of the EVJD for Mr Hornak (at D6/1.2/81), where this was said (it was originally stated as an obligation to take care, and it was amended to the following words presumably on the basis that the respondent was persuaded that the claimants’ contentions on the need to describe only what the relevant employees actually did, was correct).

“The job holder took care when driving his Loading Truck and was alert to the presence and movement of others in order to avoid collisions with other operatives working in or around the Loading Bay area, and the risk of accidents and injuries.””

The tribunal also concluded: “that if a failure by an employee to do his or her job properly could lead to (1) physical injury, (2) damage to stock, or (3) damage to equipment, then that too was obvious and did not need to be stated specifically either by the parties or us.” [67]. The third instance of this alleged error relied on by the respondent was in Appendix 8, the assembly work of comparator, Mr Jones, where the tribunal said at [322] that it was obvious that there was a need to move the stock with care and for the job holder to ensure that all stock was kept dry and off the ground to prevent damage to packaging and possible contamination.

In the Second Judgment the tribunal set out its findings of fact in narrative form, as it had previously indicated that it would do, rather than tabular form, as has been done in other cases and as the parties had anticipated would be the process adopted in this case in their preparation. The parties have helpfully provided this tribunal with examples of other Stage 2 judgments in retail cases which have all used a Scott Schedule type format which are shown in Appendix A to this judgment. It would have been more helpful if the tribunal in this case had followed a similar format to make it easier for the parties, the IEs and the tribunal conducting the Stage 3 Hearing because their findings would have been more easily discernible, as would their reasons for them. That however, does not amount to an error of law. The question is whether the parties and this tribunal can understand what decisions the tribunal had reached and understand why it reached them (see [74] – [78] above). It is so easy to pick holes in judgments such as these that are necessarily lengthy when so much is in dispute between the parties but hints of errors and fragments of mistakes are not enough (see Anya).

In the first limb of the first part of this ground I have tried to work through one example and considered the training materials in the supplementary bundle, at C7/248 “Know Your Stuff for…Fresh Food Replenishment” which was found by the tribunal as describing part of the work of Mrs Worthington at [83.9] Appendix 1, alongside section B of her EVJD “Replenishing to display stock for sale.” Whilst it is a very laborious process, it looks to me as if it is possible to marry the two up. In the RoDs the claimants’ and the respondent’s proposed wording is placed side by side and by reading the training materials that the tribunal has said form part of Mrs Worthington’s work against the RoD, although difficult, it is possible to divine the findings.

There are a number of difficulties however: the first is that the tribunal itself has acknowledged that its decision does not set out all the facts it has found in a self-contained way and that further co-operation between the parties is required:

“..the parties can now reasonably be expected (despite the possibility of an appeal succeeding in relation to our conclusions on that overall approach, which for this purpose we believe we should ignore) to co-operate by (1) looking individually and together at the training materials which we have concluded are the primary and in many cases the determinative evidence of what was the work of the relevant employees for the purposes of section 65(6) of the EqA 2010 and then (2) agreeing, in the light of our judgments (of 12 July 2023 and 4 July 2024, including for the avoidance of doubt the appendices to the latter) as much as is reasonably possible about whatever else, factually, might reasonably (in the light of our judgments) be thought to affect the IEs' assessment of the demands of the work in question.” (email EJ Hyams to parties 23 July 2024)

It is a little unclear what exactly was expected of the parties by this paragraph.

The second difficulty is that, as identified by the IEs, the process of doing whatever it was that was expected by the parties is unlikely to result in agreement between them, which is therefore likely to involve the need for a further adjudication by the tribunal.

Another problem is that I have looked at just one instance in detail and I am mindful that for the IEs the task will have to be completed for all the disputed facts that the tribunal purported to find. The IEs graphically described the complexity involved and that it may be necessary to have strings of references and sub-sub-references that might make the finding of fact too unwieldy and complicated to be useful.

But at present it is only the respondent that considers the facts are not discernible and both sets of claimants’ express confidence that the task is do-able. This is a borderline ground of appeal: on the one hand until the parties and IEs have tried to compile the comprehensive document and without a really clear specific and precise example of the impossibility contended for by the respondent, the first strand of the challenge could be said to be academic and speculative. On the other hand I anticipate difficulties ahead and the Second Judgment reads as incomplete. Whilst it is not wrong in principle for a fact contained in a separate document to become a finding of fact by incorporation into an employment tribunal judgment, I had real doubts that the tribunal’s findings were sufficiently clear for the parties to understand them and thus failed the intelligibility and transparency requirement set out in Clark (see paragraph 75 above).

As to the second strand, whilst rule 6 sets out what is required of the tribunal at Stage 2, it does not mean that it must slavishly cede all control to the parties and exercise no judgment at all. A difficulty in this case is the way in which the litigation has been conducted by the parties so that thousands and thousands of disputes have been identified, some of which were on matters that appeared pointless, trivial and insignificant to the tribunal. The tribunal was entitled to say so and use its wide power to regulate its own procedure to decline to indulge the parties in the way they requested in every respect. The tribunal was perfectly entitled to take sensible measures to manage and control the case as it explained it would do as a general principle in [54] Second Judgment. No instances have been cited in this ground of appeal of where the application of the general proposition was said to have resulted in a legal error. In the absence of a specific instance of error this ground of appeal is academic and dismissed.

In the third strand under this head the respondent risks appearing obtuse when it submits that it cannot understand whether the tribunal has found a fact which it declares is obvious: it is self-evident that is has as the tribunal has set out the thing that it states is obvious. That is sufficient for the parties and the IEs to understand what the tribunal has found – a fact that is obvious is still a fact. But where the respondent has a legitimate concern is in respect of parts of the EVJDs and RoDs that the tribunal considers too obvious to mention and has therefore not referred to at all and left out. The consequence may be that an important, but obvious matter is not then contained in the facts on which the IEs and tribunal will subsequently conduct their analysis of the value of the work performed. An obvious relevant fact still need to be stated and since the tribunal has announced that it does not intend to state the obvious the parties do not know what other facts (if any) the tribunal thought were too obvious to mention. This part of this ground is therefore allowed.

The second limb of this ground is the tribunal’s failure to accept agreed facts. The criticism here is rejection by the tribunal of some of the facts that the parties had agreed, and had been required to agree, by rule 4. The error was said to have manifested itself in the tribunal’s discussion about the job description of Mr Jones, one of the comparators and his changing the battery of a LLOP (an assembly truck which was an electric powered piece of equipment he drove as part of his work). In 12 paragraphs ([283]-[295] Appendix 8 Second Judgment) the tribunal discussed the evidence and the parties’ submissions in detail and explained exactly why it concluded Mr Jones’ EVJD was fundamentally flawed and parts of the respondent’s case were “ a nonsense.” Having appeared to have considered the evidence in very great detail it found that there would “never be a need to tilt a cage to get it onto the forks” of the LLOP, which was what the parties had eventually agreed should be in Mr Jones’ EVJD. The tribunal was entitled to analyse the evidence, particularly since this was apparently a live issue between the parties until very late in the day. It was also significant as it impacted on the credibility of other aspects of the respondent’s case. The finding does not appear perverse and the tribunal gave its explanation for its conclusion (see [293]).

The heart of the criticism is that the parties did not address the tribunal on the facts that were agreed and which they assumed that the tribunal would accept. There are two things to say about this: firstly, Mr Jones and his LLOP was not agreed until the closing submissions, which were delivered some weeks after the close of evidence and after the tribunal had been considering the evidence together, and secondly any difficulty is cured by the opportunity the tribunal gave to the parties to apply for reconsideration of any matters on which they had not addressed the tribunal.

The next specific criticism is of [898] Appendix 8, EVJD of the comparator, Mr Young:

“898.

While it could be said that even what we have just said was itself the obvious resolution of the dispute to which it related, in order to understand the background to the dispute, we had to read the whole of the passage of which it formed a part, i.e. paragraphs 6.260-6.264 of the EVJD for Mr Young. Having done that, we decided that it was necessary to say something about the content of those paragraphs, despite the fact that they were otherwise in substance, if not actually, agreed. That something was that we concluded that the only accurate way to describe the work of an assembler in the circumstances described in paragraphs 6.260-6.264 of the EVJD for Mr Young was either in conjunction with the content of the AMC guide for picking by line and any other relevant training materials, or simply by referring to the relevant parts of those documents.”

Properly read, this paragraph does not replace agreed facts with a different determination, it merely attempts to illuminate the task of paying attention when checking the content of a multi-product pallet, which reveals no discernible error. The third criticism is of [501] of Appendix 8, the EVJD of comparator Mr Hornak, when the tribunal explained why it found parts of the agreed facts about scanning and unloading UODs that formed part of Mr Hornak’s work were “unreliable”. The tribunal went on to explain why it had done so. It was entitled to look carefully at the agreed as well as the disputed facts and exercise its judgment. If (and it is very much only if) there was any unfairness detected by the respondent from not having addressed these issues in the hearing, it was corrected by the tribunal’s reconsideration offer. There is no right to address the tribunal on every matter that might form part of its judgment, especially when so much has been thrown at the tribunal.

Whilst it could have been expressed more clearly and succinctly, the tribunal’s conclusions are perfectly discernible from the paragraphs. It also contains an excellent example and illustration of the type of dispute that the parties were expecting the tribunal to resolve for them in [506] –[507]: whether the respondent’s wording of the hand movements required for UODs of “pushed, pulled and manoeuvred” or the claimants’ contention that “pushed and pulled” were otiose and were incorporated in the word “manoeuvre” and so should be deleted. It was a frustrating exercise for the tribunal. The tribunal stated that it agreed with the claimants and the additional words were unnecessary and it was a dispute that should not have been maintained. There can be no quibble with that.

This ground of appeal is therefore upheld in part only.

The respondent challenges the tribunal’s clear ruling in its First Judgment that, at Stage 2, it had to decide what was relevant, not, as contended for by the respondent what “might reasonably be relevant for the purposes of the later assessment of value by the IEs or party experts” [66]. In doing so it relied on the interests of justice, the legislative framework and the wording of r.6(3). The tribunal considered that it had to decide what was relevant and if the IEs thought that it had failed to make a finding on, or in relation to something relevant, then they could ask for a finding on, or in relation to that thing. The tribunal followed through with this interpretation and applied it to the evidence in its findings.

Mr Reade’s interpretation of the EV Rules of Procedure 2013 in support of this ground of appeal was different. Here it is necessary to set out the precise wording of parts of Rules 4 and 6 to understand the arguments. The agreed written statement to be provided to the tribunal about the work of the claimant(s) and comparator(s) must specify:

“4.—(1) At a stage 1 equal value hearing a Tribunal shall, unless it considers it inappropriate to do so, order that—

…..

(d)

the parties shall before the end of the period of 56 days present to the Tribunal an agreed written statement specifying—

(i)

job descriptions for the claimant and any comparator;

(ii)

the facts which both parties consider are relevant to the question;

R.6 then refers to the tribunal making a determination of the “facts relating to the [equal value] question” (Footnote: ¹⁶). (emphasis mine).

Mr Reade submits that the use of the words “relating to the question” in r.6 means that “relevant” in r.4 must be interpreted more expansively than would normally be expected, since “related to” connotes a looser connection than “relevant to” and the choice not to repeat “relevant to” in r.6 must have been deliberate. The claimants disagreed. Mr Jones, who led for the claimants in the oral argument on this point argued that the wording of r.6 must be read in the context of what the parties had been directed to do under r.4.

The tribunal identified early on that the respondent’s submission that the tribunal would have to make findings about every disputed fact that might reasonably be relevant, rather than just relevant, assumed that the hearing would have to be adjourned part-heard. It “would be highly undesirable from the point of view of the doing of justice” ([33] First Judgment), since it would significantly expand the areas of dispute between the parties. The tribunal made clear from day 1 of the hearing that it would be determining only relevant factual matters, seemingly with the approval of Mr Epstein who said that the respondent would not be relying on any assertions of fact in the respondent’s witness statements or EVJDs which were not material to the issues to be decided at stage 2 and that things that were relevant only at a final hearing would be immaterial factual assertions. This was with the proviso only that Mr Epstein would not concede that the physical location of a DC was not a material consideration at Stage 2. It was a sensible approach consistent with the requirements of the Rules.

On the second day of the Stage 2 Hearing, 8 March 2023, the tribunal made clear to the parties that the tribunal would:

“…need to confine the scope of our inquiry to those things, and only those things, which were relevant. Albeit that is something factual was relevant only peripherally then we might conclude that it was not in the interests of justice, applying the overriding objective set out in rule 2 of the Employment Tribunals Rules of Procedure 2013 to come to a conclusion on that thing.” [35]

Even with careful case management, it was only by taking the time set aside for the tribunal to deliberate to use for the parties’ advocacy, that the tribunal was able to conclude the evidence and the submissions within the overall time that had been allocated to the case. The tribunal then made further time available for the tribunal panel to deliberate in chambers.

In any event Mr Jones and Mr Bryant’s primary point is correct: the scene is set by r.4 –the parties are required to agree and record what is not agreed as facts which are “relevant” to the question. It would be wholly illogical and perverse for a different and wider test of the strength of connection between the fact or Disputed Fact and the equal value question to then be introduced at the subsequent Stage 2 Hearing, after the parties had prepared the evidence on a narrower test for the purposes of the statements required to be filed under r.4. The respondent’s interpretation risked achieving an absurd result.

As well as failing on principles of interpretation, it would also be very undesirable to add to the volume of evidence and complexity in equal value cases.

As HHJ Auerbach rightly said in his preliminary hearing judgment in this case:

“13.

The equal value rules set out a unique, bespoke, structured procedure for equal value claims, including, in cases where the tribunal decides to appoint an IE or IEs, a stage 1 hearing, then a stage 2 hearing and then a final hearing, and an indicative timetable. Questions of interpretation of particular parts of these rules must be approached within the context of their overall scheme and methodology, and on the basis that individual rules are intended to work in harness with one another.

15.

The submission put to the tribunal by the respondent, and which it rejected, is that what

“relevance” means for stage 2 purposes is that “if a fact might reasonably be relevant” for the purposes of the later assessment of the equal value question by the IE or party experts, it should be included in the job description. Ground 8 asserts that the tribunal erred by holding that it had to make a “final determination” about what facts were relevant to the assessment of the question and should instead have determined that it has to conclude what facts “might reasonably affect the later assessment of value”. The respondent’s main skeleton says: “or, to put it another way, [the fact] might reasonably

be relevant at the later stage.

16.

At the broadest level, there are two ways in which that challenge might be advanced. The first is the contention that, in this particular context, what “relevant” must be understood as meaning is “might reasonably be viewed as relevant by the independent expert or a party expert”. The second is the contention that the reference in rule 6(1)(a) (and elsewhere thereafter) to “facts … which relate to the question” is a significant change in language, because “relate to” is broader in meaning than “relevant to” and embraces “might reasonably be relevant to”.

17.

The respondent’s original skeleton was, respectfully, ambivalent as to how much reliance it placed on the second argument. In one paragraph it was relegated to a footnote indicating that it “may be material” that there was a “loosening of language between rule 4 (‘relevant to’) and rule 6 (‘relate to’)” which “might have been intended to reflect” that at stage 2 the tribunal should “not be concerned with making definitive findings as to overall ‘relevance’”. But a little further on it contended that a fact “relates to the question” if it might reasonably be contended to have an effect on the answer. The further note from counsel for this PH squarely focusses on the proposition that “relate to” is not synonymous with, but broader than, “relevant to”, and was accompanied by some authorities commenting on the meaning of the words “relating to” in other contexts. Mr Epstein KC at the hearing told me that this was now the primary basis on which the challenge was advanced.

18.

I have real doubts as to the arguability of this ground. I think it is strongly arguable that rule 6(1)(a) must be read against the context of what the rules envisage that the parties will ordinarily have been directed to do at the stage 1 hearing and, specifically, the steps that they are enjoined to take at 4(1)(b)(ii) and (iii) and (d)(i)(ii) and (iii). There may be areas of factual agreement or disagreement as to what a claimant or comparator factually does, and/or areas of agreement or disagreement as to which aspects of what they factually do are relevant to the equal value question. The parties are required to tell the tribunal what they agree or disagree about in both respects.

19.

On this reading rule 6(1)(a) then refers back to those different categorisations of the material in the statement that the parties were required to provide to the tribunal, together with the job descriptions, by rule 4(1)(d). Where the parties are in agreement, that something said to be an aspect of the claimant or the comparator’s work is, as such, a fact, and is a relevant fact, there is nothing the tribunal needs to determine. But the tribunal must make a determination in respect of “facts on which the parties cannot agree which relate to the question”. It is arguable that “facts on which the parties cannot agree” is plainly a reference back to “the facts on which the parties disagree” in rule 4(d)(ii),

and, as that provision makes clear, in this context the disagreement may be either as to the factual position on the aspect or topic in question, or as to the relevance of the topic itself, or as to both.

20.

So, in turn, it is arguable, rule 6(1)(a) requires the tribunal to determine disputes about what the factual position on a given topic is, disputes as to whether what the factual position is on that topic is relevant, or both. It is in my view strongly arguable that, against that background, the use of the words “which relate to the question” does not indicate an intention to require the tribunal to apply a different test or approach at the stage two hearing; and that, if Parliament had so intended, it would have spelled out the different test to be applied.

21.

I am also sceptical of the respondent’s arguments that the provisions of rule 6(3) do not, as the tribunal thought, provide an adequate safety net if the IEs consider that they are in difficulty with completing their task on the basis of the facts they have been given. Further, in addition to rule 6(3) there is provision in rule 5 for the tribunal to seek assistance of the IE in establishing the facts on which they may rely, either on its own initiative or on application; and, where appointed at stage 1, the IE can then be on hand or contactable at stage 2, as I am told the IEs were in this case.

22.

The scope of rule 6(3) also appears to me to be broad. The IE can make an application in relation to “some or all” of the facts they have been given, and for some or all of those facts to be “amended, supplemented or omitted”. I am also sceptical of the argument that the respondent’s interpretation must be right, because a party expert would not be able to look to the tribunal for assistance under rule 6(3). An IE is a unique creation of the schedule 3 rules, and has a different and distinct role from a party expert. Even there, the tribunal does not have to appoint one at all, and the rules arguably tightly circumscribe their role, and keep it under the firm control of the tribunal, throughout the process. This includes the strict requirement to base their report only on the facts given to them by the tribunal. The IE does not, it appears to me, have any fact-finding role at all.

23.

Nor do I consider that the other authorities to which I was referred by Mr Epstein KC, on the wide meaning of the expressions “relating to” or “relate to” in other, quite different, legal contexts offer any real assistance on the question of the significance or not of that phrase within rule 6(1)(a), which must be construed within the context of the very specific and unique legal regime, and the other relevant rules and architecture, in schedule 3, of which it forms a part.”

After oral argument the respondent’s case did not improve for all the reasons HHJ Auerbach identified and I adopt his reasoning. This ground of appeal is dismissed.

There were 11 grounds of appeal that challenged various findings of fact made about the work of the comparators which can be dealt with briefly more briefly. In places the grounds of appeal appeared as a classic example of using a nit comb on an elephant, and using the wrong tool for the job. One is looking for errors of law and perverse findings of facts, not nits. I refer again to the general principles of how this tribunal should approach appeals set out above in paragraphs 72 – 79 above.

It was the respondent’s case that the comparators were subject to productivity targets (known as a Performance Index (“PI”) or “PI rates”) when undertaking measured activities including “assembly” and “loading”. The PI rates required the comparators to complete a certain amount of work within a certain time. The measurement took place through the AMCs worn by the comparators. The claimants did not wear AMCs. The respondent’s case was that the performance targets applied to the comparators requiring them to work more quickly and concentrate harder placed demands on their work and imposed mental stress or pressure. The tribunal considered the impact of the use of PI rates in both its First and Second Judgments. In the First Judgment after discussing the evidence in [70]-[71] the tribunal stated: “after careful deliberation, we concluded that those rates were not irrelevant,”[71] since they were relevant to time pressures. Time pressures were relevant because the “demands” of the work had to be to determined in part by reference to (1) physical conditions, such as heat or cold, (2) dangers, albeit as mitigated by the taking of measures to reduce or eliminate those dangers, and (3) pressures of time.

However, it came to a different view in its Second Judgment which it flagged up at [62]-[63]. It explained that after much further consideration of the evidence:

“we could not see how it could properly be decided that an employer could, by imposing unreasonable productivity targets on an employee, increase the value of the work of the employee for the purposes of section 65(6) EqA 2010” [62]

And that having considered the factual material ([111]-[113] Appendix 8):

“That material showed how clearly separate the imposition of a PI target was from the work to which it related, and confirmed for us that a PI target was purely incidental for present purposes.” [63]

In Appendix 8 [684] - [687] the tribunal explained that its view was reliable and it reached “an inescapable conclusion that PI rates had no effect on those demands. That was because the work was what it was and the imposition of productivity targets could not in itself affect that work…[it]… was entirely separate from the task to which it was applied.” [684] It also concluded that the imposition of productivity targets did not affect the relevant conditions in which the comparators worked [685]. It was purely incidental for present purposes. Part of the evidence on which it had relied to reach this conclusion was the comparators’ evidence that Mr Macko and Mr Hornak had rarely hit their PI targets.

The respondent has quite rightly submitted that it should not have formed any part of the tribunal’s reasoning whether or not the productivity targets were unreasonable, which it did at [62]. As Mr Reade bluntly put it: an unreasonable demand is still a demand and must be taken into account when finding facts about the work. The claimants’ response is that the tribunal’s conclusion was one it was entitled to reach on the evidence, which was the basis on which it had reached its findings and the reference to an unreasonable productivity target did not form the basis of their decision.

The difficulty for the claimants is that it is a clear and obvious error of law to introduce the concept of a reasonable employer and what demands a reasonable employer would, might or ought to have placed on its employees and it is not possible to discern from the tribunal’s conclusions what part this misdirection played in its findings of fact. This is no mere “hint of error or fragment of mistake” as per Sedley LJ in Anya but could have infected the rationale of the tribunal’s decision.

This ground of appeal succeeds.

This ground of appeal is linked to A2G4 since it also examines the tribunal’s conclusions about time pressures. In [55] First Judgment the tribunal made comparative findings about time pressures as between the sample claimant Mrs Worthington and her comparators. After four paragraphs of discussion of the evidence of Mrs Worthington ([50]-[54]) the tribunal concluded that:

“55…it was highly unlikely that an employer such as the respondent would impose targets on its distribution centre (or central warehousing) staff but not expect its shop floor staff to work under any kind of time pressures, we came to the firm conclusion that

55.1

the claimants were put under time pressures in the same way that the respondent’s DC staff were put under time pressures, and

55.2

the only material difference in this regard between the two sets of staff was the fact that some of the DC staff wore arm-mounted computers or terminals about which we heard much oral evidence, as a result of which it was possible to monitor more effectively their use of time.”

The respondent’s criticism of the tribunal is two-fold: firstly findings about value ought to have formed no part of the tribunal’s decision at Stage 2. Secondly, the tribunal was perverse in reaching a “firm conclusion” about all the claimants when it had only considered the evidence about one, Mrs Worthington, and without having made clear factual findings about the demands on any comparators.

In addition to the written submissions from counsel for both sets of claimants, Mr Jones made valiant efforts to uphold the tribunal’s finding and reasoning at [55] arguing that the tribunal was not making a finding about value, but merely determining the facts at the level of detail it considered was required. The tribunal did not make a finding as to the relative extent of the time pressures and the respondent had misread the tribunal’s findings.

The difficulty for the claimants is that there was no proper factual basis on which to reach a “firm conclusion” at that point in the Stage 2 procedure. Paragraph 55 begs so many questions, such as what are the findings that the tribunal has made about the “way” that the claimants and the comparators have been put under time pressures? The recitation of the evidence in paragraphs [50]-[54] is insufficient to explain the assertion at [55]. I agree with the respondent’s submission that there were both no findings of fact sufficient to explain the conclusion and the conclusion introduces an evaluation which is not any part of the Stage 2 procedure. This ground of appeal succeeds.

In the tribunal’s first judgment at [87]-[88] at the end of a discussion about training documents and materials, much of which has been discussed already in A1G2 above, it came to the following conclusion:

“87.

…we saw no good reason to make determinations about precisely what training the sample claimants and their comparators had actually received, with one exception.

88.

That exception was that the question whether an employee had received training would be material if the training had led to a determination (by whomever) that the employee was competent to do the thing to which the training related and the respondent would (or could) not permit that thing to be done without such determination. We had in mind in this regard training to be a forklift truck driver. Thus, the question whether a comparator had received that training was material, but we could see no other training in regard to which we would need to make a specific finding of fact of that sort.”

There were two challenges to this conclusion by the tribunal: firstly that it was wrong in principle since training is obviously a relevant fact and secondly wrong in application because there were more roles than just fork lift driver for which successful completion of training was required for the work. This ground has some overlap with A1G2. Read in isolation, [87] does appear to be not only surprising, but wrong: the extent of a job holder’s training is plainly a fact relevant to the equal value question. But the position is more complex given that, for the reasons set out above in A1G2, although not a decision that every tribunal would come to and something of an outlier even in the context of the current crop of equal value cases in the retail and supermarket sector, it is not for this tribunal to interfere with a finding of fact made by a tribunal that has spent as much time considering the evidence – both documentary and oral - in the case and hearing from counsel and the IEs, as this tribunal did and on this point made no identifiable error of law. It is not a finding that I am able to disturb. Considerable leeway should be given to the tribunal that has heard all the evidence. The tribunal concluded that on the particular facts of this case and given the nature of the respondent, the strict requirement for everyone to follow and apply all the training materials to the letter was what this employer required the sample claimants and comparators, in practice to do, in fulfilment of their duties, unless there was evidence to suggest otherwise. It was therefore not important whether or not there is a record of when or whether the individual sample claimant or comparator had him or herself received or undertaken the particular training. The tribunal spent some time explaining how training was delivered and much of it required employees to read it themselves and watch the training video, following the manual and such like and the obligation seemed to be on the employees to make sure they were up to date with it all [87] First Judgment which also has to be read in the context of the preceding paragraphs [75]-[86] to gain a full understanding of how the tribunal arrived at its conclusion. Part of the context to the tribunal’s conclusion was the undisputed evidence that much better records of training were kept at the DCs than the stores which could have inadvertently unfairly disadvantaged the claimants and presented a distorted picture.

The tribunal amplified on its reasons in its Second Judgment Appendix 8 [213] and [8]-[190].

The respondent argued that two training related matters had been conflated – using the generic training materials to make findings of fact about the work carried out by the job holder and the separate issue of what training the individual job holder had received, relevant to the knowledge required by the job holder in order to carry out the tasks. On close analysis it is a distinction without a difference on the particular facts of this case. Since the tribunal effectively found that if the training applied to the job holder’s job, they would have done the training and done the work in the manner set out in the training (absent evidence to the contrary), it could be taken as a given that they had done the training, as a general rule, and the individual training records were not significant.

Seen in that light, A1G5 is no more than a challenge to findings of fact that the tribunal was entitled to make for the reasons it gave. I therefore find that there is no error in principle in [87] given the particular findings in this case.

The exception set out in [88] was where what the tribunal described as “a determination (by whomever) that the employee was competent to do the thing to which the training related and the respondent would (or could) not permit that thing to be done without such determination” [88]. I take that to mean where the job holder needs certification of some sort from having successfully completed the training to do the task, such as is required to operate various pieces of manual handling equipment, such as powered pallet trucks.

If there were tasks other than fork lift driving for which any of the comparators required a licence, registration, certification or confirmation of training of some sort that the tribunal has overlooked, these were exactly the type of matters that the tribunal said it would reconsider. I have not studied it in minute detail but I understood from the parties that this has been addressed in the Reconsideration Judgment at Appendix 7 which has a section on induction training. But if not it is a matter that can perhaps be raised at the disposal hearing

Ground A1G5 is dismissed.

This ground of appeal is a challenge to a sentence in a paragraph in the tribunal’s findings in the First Judgment at [74] that “Contrary to the respondent’s submissions….relationships which the job-holder had with their colleagues would not be relevant at this stage.” It was said to be an error of law since facts relating to interpersonal skills and relationships would be relevant to the assessment of value, to be conducted at Stage 3.

The sentence needs to be placed in the context of the respondent’s submissions referred to in th sentence in [74] which were about the store that Mrs Worthington worked in and how Mrs Worthington’s husband and daughters also worked at the store and that other of the customer assistants were students and younger employees and they would all “laugh and joke while getting the job done and it felt like a family”.

In that context it was open to the tribunal to conclude that the harmonious relationships in Mrs Worthington’s store and all her other family members working there was not apt for inclusion in a factual job description. The tribunal was entitled to reach that conclusion.

Although it is a little ambiguous, since [74] comes under the heading of “Our conclusions on points of principle” I do not read the sentence as saying that communication and interpersonal skills are not relevant to the factual exercise at this stage of the case, but it was intended to relate only to the very specific respondent’s submissions about Mrs Worthington and her store. It is evident in the Second Judgment that the tribunal does indeed take account of the requirement to communicate with line managers as it made a finding to this effect concerning Ms Oz ([49] Appendix 6 Second Judgment) . It also made a finding in relation to the comparators, albeit that it stated it was obvious and therefore did not need to make a finding of fact about it. (Footnote: ¹⁷) As discussed above in A2G1, the tribunal erred in concluding that the obvious did not need stating, but the fact that some of Mrs Worthington’s relations also worked at the store was not relevant information for her EVJD.

The appeal ground is dismissed, insofar as it relates specifically to Mrs Worthington, but for the avoidance of doubt in general terms interpersonal skills and relationships are relevant to the question.

The respondent challenges the tribunal’s conclusion about the comparator Wayne Jones that it did not “see anything material in paragraph 10” in his EVJD concerning “Risks and Hazards” ([333] Appendix 8). It gave three reasons: firstly because hazards were obvious from the tasks of an assembler and the way in which they had to be carried out; secondly the “claimed hazards in the working environment such as poor light, were just as much as was the case in relation to the work done by the sample claimants”; and thirdly because

“…hazards (assuming they did in fact exist) must have arisen because of failures on the part of the respondent….It did not seem right to us to include in our analysis of the working conditions of either the claimants or their comparators hazards which would have arisen (if they did in fact arise) only because of a failure by the respondent to comply with its obligation to take reasonably practicable seps to ensure that the working environment was safe.” ([333] Appendix 8).

The respondent argues that the conclusion was legally flawed since the obvious should still be stated, there was no proper basis for asserting that there may have been any health and safety failures by the respondent and it is not for the Stage 2 Hearing to evaluate relative demands of the work as between the sample claimants and their comparators.

The claimants argued that contrary to the bald statement by the tribunal that there was nothing material in the “Risks and Hazards” section 10 of the EVJDs, elsewhere in the Second Judgment the tribunal had taken account of and recorded risks and hazards and the respondent’s safe system of work, such as about faulty cages. Paragraph [333] Appendix 8 had to be seen in the context of the tribunal’s overarching conclusion that the respondent’s contentions about health and safety were overblown. In any event, the tribunal revisited some of its conclusions on risks and hazards in the Reconsideration Judgment in some respects.

To put this ground in its proper context, the IEs identified the relevance of risks and hazards as a demand factor under the heading “how to ensure personal and operational safety” at 5.2.2., in their sample job descriptions supplied to the tribunal and the parties before the r.4 documents were prepared. In Mr Jones’ EVJD a whole section, section 10, is dedicated to the risks and hazards he faced which the claimants wanted to be deleted or substantially amended on a variety of grounds such as irrelevance, disputed fact, exaggeration, evaluation/analysis not fact and repetition. There are 111 points of dispute between the parties in section 10 of his EVJD.

The difficulty for the claimants is that there are a number of demonstrable errors in [333] Appendix 8. Firstly it is not correct that the obvious does not need stating. It does if it is relevant and proportionate to be included as a fact. Secondly it was also an error to assess hazards and risks by reference to what a legally health and safety compliant employer or reasonable employer would do, rather than what work this respondent required this comparator in practice to do. Furthermore there was no factual finding as to whether there were such hazards or if there were, whether they arose because of any breach of legal duties by the respondent as an employer. Next, it was also not the tribunal’s role at a Stage 2 Hearing to conduct any evaluation between the work of a sample claimant and a comparator. But even if it had been, there do not appear to be any actual findings of fact from which it could do so: for example what were the lighting levels in Mrs Worthington’s store and Mr Jones’ DC, from which the comparison could be made to conclude that they were the same?

It is an error of law for the tribunal to state that there was nothing material in the whole of section 10 of Mr Jones EVJD. This ground of appeal is upheld.

Paragraph [70] Second Judgment comes under the overarching heading: “Conclusions to which we came about relevant issues of principle after having made all of the factual determinations which we concluded at this stage needed to be made”. Paragraph [70] has its own heading: “The impact of a failure to ensure that equipment was in good working order” and cites three sub paragraphs taken from appendix 8 which explain why the tribunal concluded that the value of the work had to be assessed on the assumption that the respondent would keep equipment in reasonably good repair, or replace it, and the tribunal thus excluded the impact of faulty equipment from its findings of fact.

The first reference is the finding at [333] Appendix 8 set out in ground A2G5 above – “it did not seem right to us to include in our analysis…hazards which would have arisen (if they did in fact arise) only because of a failure by the respondent to comply with its obligation to take reasonably practicable steps to ensure that the working environment was safe.” For the reasons set out above in ground A2G5, this is erroneous.

Secondly, citing [522] Appendix 8, the tribunal stated that as a matter of law the value of the work of the loaders could not vary according to the incidence of difficulties with faulty equipment. Its rationale was that it considered it wrong for an employer to make an employee’s work more valuable for the purposes of s.65(6) EqA 2010 by making it harder for the employee to do his or her job by failing to keep equipment used by the employee in good order, or not providing alternative equipment. This statement is also legally flawed for the same reason – the work, and its value, is about what the employer required the employee to do and what the employee in practice did. The tribunal over-thought the question.

The claimants made a number of good points that the tribunal’s conclusion could be justified and explained on other grounds – such as repetition and double counting, inconsistencies in the respondent’s evidence (such as the faulty cages that miraculously repair themselves en route from DC to store discussed above) - but they were not the reasons given by the tribunal for this particular factual conclusion said to be deciding a point of principle. Arguments of this nature may be raised at the disposal hearing. This ground of appeal is upheld.

Paragraphs [79]- [81] of the Second Judgment also come under the overarching heading that it is deciding a point of principle and concludes that the working hours of an employee and therefore overtime are irrelevant and it declared at [80] that the time of day when work was done was not relevant to the determination of what that work was.

I did not understand the claimants’ submissions to be seeking to defend the assertion that the time of day that work was done was irrelevant, but instead they argued that the tribunal had indeed made findings about the times worked by the two sample claimants who worked nights and the tribunal did not object to the inclusion of the parties’ agreed facts that the working hours and shift pattern of the comparator, Mr Pratt, being included within the relevant facts, or the particular hours worked by Mr Jones. They also raised procedural points and argued that the respondent’s grounds amounted to a generalised assertion, not any particular erroneous factual determination.

The difficulty with the claimants’ argument however is that the tribunal has made a clear ruling of law that is unarguably wrong and cannot stand. How exactly it plays out in relation to each of the sample claimants and each comparator will be relevant to disposal, which the parties agreed should be dealt with at a separate hearing.

This ground of appeal succeeds.

The criticism in this ground of appeal is that it is said that in a number of places in its findings the tribunal dismissed the physical effort required of the comparators set out in section 7 of their EVJDs. Unlike the preceding grounds which are described by the tribunal in the Second Judgment to be points of principle and rulings on matters of law, this ground of appeal is not anchored in the Second Judgment principal or generic findings or any ruling on a matter of law or point of principle, but only in Appendix 8 to the Second Judgment specific to a particular comparator.

There is no dispute that “effort” is a factor specifically listed in the non-exhaustive list of factors that are demands made on claimants and comparators in their work set out in s.65(6) EqA 2010 and that physical effort is to be assessed as part of the evaluation process. There is no apparent misdirection and the tribunal had set out s.65(6) EqA 2010 in the early part of its judgment.

I agree with the claimants submission that this ground of appeal can only succeed as a perversity challenge. I have considered and reject the claimants’ submission that there has been a breach of paragraph 3.9 EAT Practice Direction 2024, since this is not a mere generalised allegation that the tribunal did not accept the respondent’s argument, the ground of appeal provides specific incidents and full details of the matters relied on in support of the allegation of perversity in the five instances stated. In order to consider this ground of appeal it is therefore necessary to consider the detail of the five impugned findings in Appendix 8 to see if the high threshold of establishing perversity in a finding of fact is made out.

Section 7 of each of the comparator EVJDs is headed “Key facts including metrics”. For the comparator, Mr Jones, at [329]-[330] of Appendix 8 Second Judgment the tribunal states:

“329 We also could not see how the matters to which reference was made in section 7 of the EVJD for Mr Jones (including, for example, what “Focus and concentration”, or what “Stamina”, was required), even if they were true (and the evidence to support them was scant), were relevant at stage 2, i.e. to the analysis that we were obliged to carry out after the stage 2 hearing that started on 6 March 2023.

330 If the IEs disagree with us in that regard then we will review our conclusion and

consider the parties’ contentions on the disputed things in paragraph 7 of the EVJD

for Mr Jones which the IEs regard as relevant.”

This is an error of law: focus and concentration and stamina are clearly part of effort. They are relevant. It should have been considered. It was not for the IEs to have to apply to the tribunal to have it dealt with properly.

In [544] the tribunal stated that it was baffled by some of the assertions about Mr Hornak in paragraphs 7.4-7.7 of his EVJD to do with loading, which stated, inter alia, that the job holder completed 270 UODs per shift, equivalent of 6 or 7 trailers and he was responsible for the loading of around £140,000 worth of products (perhaps in the Finest range, they did not say), that he pushed pulled and turned by hand up to 45 cages weighing an average weight of 277kg and the total weight handled by him was 74.8 tonnes in 2013-2014 and his approximate distance walked during his shift was set out too. The claimants’ response recorded in the document was to ask for evidence for the figures given and the respondent’s response in the final column was to refer to page references in work studies conducted in 2012 and 2015, the job holder’s PI rates, spreadsheets attached to the witness statement of one of the respondent’s witnesses, Gary Black and some mathematical/ physics calculations. The tribunal did not engage with the respondent’s comments in response to those of the claimants, beyond expressing bafflement by reference to one statistic which was to doubt that Mr Hornak had personally lifted 70 tonnes in a shift, but if that was the case “there was no relevance in the statistic” [544]. It also concluded that the value of the stock was irrelevant, as was his precise step count, although the number of steps walked might be relevant. The tribunal also observed that walking around might be easier than standing still all day [547] and that it was obvious that the job would be physically demanding [546].

The respondent is entitled to quibble with some, but not all of the contents of [544]-[547] of Appendix 8 to the second Judgment. The tribunal was right to criticise the sentence in the EVJD drafted by the respondent that: “The job holder’s work was physically demanding” because it says nothing of substance. The point of Stage 2 was to establish how and in what ways it was physically demanding. So the tribunal had a valid criticism of the sentence, not because it was obvious, but because it was vague and somewhat meaningless.

As I understand it the tribunal’s difficulty with the 70 tonne statistic, was because it did not believe that the job holder had lifted this weight personally during the shift, and if it had been lifted using mechanical handling equipment (MHE) it was irrelevant, since if the machine is doing the lifting it is not the job holder’s physical effort. The respondent now say that 70 tonnes is what was in the cages personally manoeuvred by the job holder. I was not familiar enough with the evidence to know if this was a perverse finding of fact by the tribunal or the respondent’s fault for not making it clear, in which case it cannot bite the cherry for a second time (subject to the tribunal’s approach to reconsideration). Since part of this ground of appeal succeeds, whether, and if so, how far [544] and the 70 tonne finding can be re-opened can be discussed at the disposal hearing.

Whether standing or walking is more tiring was a quite unnecessary observation by the tribunal since, as it acknowledges at the end of [547], that will be something for the IEs to assess. The tribunal has therefore demonstrated that it understands its role at this stage is to set out the facts and the role of the IEs to do their evaluation assessment, although it would have been better if it had not first expressed its own opinion and set the hare running. There is nothing in this part of this sub-ground of appeal.

The next perversity criticism is of the tribunal’s findings for the comparator Mr Pratt [732]-[734]. There is the same criticism about whether the weight lifted during the shift was done by manual handling or mechanical handling and the same comments apply as for Mr Hornak. It can be dealt with at the disposal hearing.

Ground A2G8 is partially upheld.

The respondent asserts that the tribunal perversely declined to make findings of fact on disputed facts relating to the monitoring of the comparators at work as they were not relevant to equal value. In comparator Mr Jones’ case the tribunal declined to make a finding as it was not considered relevant at Stage 2 ([331] Appendix 8 Second Judgment) but the tribunal said that it would review its conclusion if the IEs disagreed (see above). For the comparator Mr Todd, after 15 pages of findings from [736]-[791] of Appendix 8 the tribunal breezily stated that it did not believe that the other disputes in his EVJD needed to be determined for one or more of the reasons in paragraphs [9]- [78] Second Judgment, but did not say which or how ([791] Appendix 8). For the comparator Mr Young in the last paragraph of Appendix 8 to the second Judgment the tribunal declined to make any findings on any part of sections 8 (Performance and accountability), 9 (Physical conditions inside the warehouse) and 10 (risks and Hazards) “on the similar basis that what we say above or in our judgment of 12 July 2023 meant that it was disproportionate to address any of the many disputes” in those sections. The fatigue that the tribunal had reached by the end of its Second Judgment is entirely understandable, but more was required by way of making findings, or explanation why none of these disputed facts required resolution. It appears to have just run out of steam. This ground of appeal is upheld.

Appeal ground 10 of the second appeal is, like ground 9, founded solely on perversity and relies on the same paragraphs as for ground A2G9. For the same reasons that the tribunal erred in not making findings or by not explaining satisfactorily why it was not making findings on monitoring, management and surveillance, it was also wrong to exclude (or not explain satisfactorily why it had excluded) aspects of the physical working environment. On the face of it, findings on the physical working environment would be relevant, and more explanation required of why it was not necessary on the facts in this case. This ground of appeal is also upheld.

The tribunal found that it would take no account of the requirement of a comparator to do mental arithmetic, since a reasonable employer would allow an employee to use an electronic calculator or the function on a mobile phone ([71] Second Judgment). It was a reference to its detailed findings on the EVJD for one of the comparators, Mr Todd, (at [761]-[765] of Appendix 8) in relation to the task of checking goods which was set out in 44 paragraphs of the EVJD that the respondent had prepared for him. In it, the respondent asserted that Mr Todd would use mental arithmetic to check that the correct number of units were on a pallet by multiplying the number of layers on the pallet by the number of units on each layer, but with complicated or more uncommon sums, as a checker, he would use pencil and paper to work it out.

The tribunal was sceptical of the evidence and “doubted” a checker would use pen and paper to do the calculation as it would be less reliable than checking on a calculator function on a mobile phone. It did not however make a clear finding of fact either rejecting or accepting the evidence, but went off on something of a tangent to find that a reasonable employer would allow an employee to use their mobile phone to do this sort of calculation. It went on to say:

“If the respondent failed to provide a checker with a calculator, then that was in our view incapable of adding to the demands of the work of a checker. That is because in our judgment, it would mean it was open to an employer to argue that work was more demanding than it would have been if the employer had provided what a reasonably efficient business would have provided.” [762]

The tribunal’s conclusion was self-evidently problematic. As noted above in relation to other grounds, what a reasonable employer would do is not the question, it is what work this employee was required to do and did in practice do for this employer. What a reasonable employer might do could be relevant in assessing the credibility of an employer’s account of the employee’s work and used as a tool for evaluating the evidence, but in these paragraphs the tribunal declined to make a finding of fact. This ground of appeal succeeds.

In light of my upholding some of these specific challenges, I have revisited my conclusions on the first category of grounds of appeal – approach to work, in light of the specific errors identified in some of the later grounds, such as A2G5, A2G6 and A2G7 where I have upheld the appeals, to see if it demonstrates a more fundamental error in the tribunal’s approach to the meaning of “work”. Even on a close re-reading, I am satisfied that they are random swallows that do not a summer make.

This is the most sprawling, unwieldy litigation that has been exceptionally hard for the tribunal to manage. The documentation for only these appeals, which concerns just one small part of the case, was very lengthy and much of the contents of the supplementary bundle was required reading. The respondent’s skeleton argument and appendices was over 100 pages long and the Harcus claimants’ skeleton 59 pages. Even though the skeleton arguments were well signposted and carefully indexed, the parties had thought it necessary to attempt to agree a table for navigating the skeleton arguments. But this was nothing compared to the volume of material and number of issues that the first instance tribunal had to deal with. The tribunal was faced with a monumental task. It has spent a total of 163 days recorded in tribunal hearings and full panel deliberation time on the First Judgment, the Second Judgment and the Reconsideration Judgment and that statistic takes no account of judgment writing time and all the time spent by the EJ in drafting emails and considering the parties’ responses. The tribunal has worked extremely hard on this case. The tribunal has done its best to keep on top of the volume of material and the demands of the very well-resourced parties on all sides and their diligence is commendable. The pressure on the tribunal with litigation of this type which is a far cry from the intention for industrial, now employment, tribunals to achieve the aims set out in the Donovan report in 1968 to provide an accessible, speedy, informal and inexpensive way to resolve employment disputes has been immense.

All parties agreed that ideally none of them would be starting from here with the case in its current state. I agree. Although not amounting to grounds for a successful challenge on appeal beyond the matters identified above, procedurally, the case has gone awry and it is necessary to make a few observations which are intended to be helpful.

I am concerned that the EJ has corresponded with the parties directly by email. Judges speak through their judgments and not by way of email running commentary with repeated and further explanations. Similarly there is usually no need in subsequent judgments to add to reasons previously given, or in correspondence intended to deal with separate issues. It leads to confusion and repetition and opens the door for opportunistic appellate litigation, as has happened here. The iterative process of judgment writing stops when a judgment is promulgated. It is unhelpful to give further, or additional reasons for an earlier judgment. I understand the temptation, and appreciate how hard it can be to let go of a draft judgment that has consumed so much time and energy, but it is not a scab to be picked at or a crime scene to be returned to (Footnote: ¹⁸). For a party who disagrees with the outcome, which iteration of the result is to be appealed? How are the parties to know which document was relied on by the tribunal to reach its findings and conclusions? How is an appellate court to deal with any inconsistences between explanations given in different documents? If the explanations are the same, why repeat them and add to the paperwork? What is the status of an email that either contains reasons for an earlier judgment or includes a decision in the email (for example a decision on a postponement request)? Is it to be treated as a decision under the ET Rules of Procedure 2013 or now 2024 requiring an appellant to appeal an email? If it is not a decision what is it and how does a party challenge it? Where the email is not just about case management but appears to contain a ruling apt to be made by a full tribunal, what input have the lay members had?

Procedural rigour is important, especially in well-resourced cases with teams of expert lawyers on all sides where the stakes are high. I have tracked the source of the practice of an EJ corresponding directly with the parties in this case. It started in Covid when it was the only way in which the employment tribunal could keep functioning before digitisation of Employment Tribunal proceedings. EJ Manley carefully set out its rationale and circumscribed its use and it was necessary at that time. But the way in which the practice has developed and continued after Covid has been unfortunate and resulted in a loss of procedural rigour and a lack of clarity as to what has been decided and what has not been decided. It has encouraged the parties to keep revisiting issues that had been previously decided. Tribunals are intended to be more informal than courts, but the rules contain the minimal requirements of formality which need to be adhered to.

The consequence was at times a lack of clarity about when a decision has been made, the basis on which it had been made and the invitation to the parties to make reconsideration applications under r.71 ET Rules of Procedure 2013 on both the Second Judgment and now on the Reconsideration Judgment will add further to the confusion. It makes for something of a moving target and is inimical to the important principle of finality of litigation.

It was a curious, but not unlawful, decision to inform the parties by email on Friday 12 May 2023 towards the end of the evidence in advance of the closing submissions commencing on 23 and 24 May 2023 that the tribunal did not plan to make its findings by reference to the RoDs which had been worked on by the parties for many, many months in accordance with the orders made by EJ Manley in 2020 and as had been done in all the other Stage 2 cases of which the parties were aware. The tribunal acknowledged that the parties had spent two years preparing for the Stage 2 hearing ([32] Second Judgment). It meant that the tribunal was pulling in a different direction to all the parties and the IEs. It made life so much harder for the tribunal by not working with the parties, yet the tribunal noted the tremendous collective experience in the law of equal pay of all counsel appearing before it and paid tribute to their submissions that they said they had found to be of great assistance [28]. There are silks and experienced specialist solicitors on all sides, not the inequality of arms that one sees in some cases. The IEs too had a wealth of experience and were trying to assist the tribunal. Nor was this a case where the parties were saying different things: all sides and the IEs were speaking with one voice. It is perhaps to be regretted that the repeated advice of IEs for a single document EVJD based on their extensive experience, backed up with worked examples, was not followed.

There is a need for tribunals to maintain a firm control on litigation of this type. I am not sure how helpful it is, or even appropriate, to apportion blame for the state of proceedings where it is not directly relevant to the issues in the appeal. This is litigation. Hand-wringing of itself about lack of progress serves little purpose. Fine words butter no parsnips as the saying goes. Robust case management and compliance with the overriding objective is the method by which cases are kept on track. The tribunal has extensive case management powers to make orders and directions and can impose sanctions for non-compliance where it is just and necessary to do so. It may well be in the interests of one side or another to obfuscate and delay but it is the role of the other side to seek appropriate orders and for the tribunal to manage proceedings robustly in order to achieve the overriding objective.

A Stage 2 Hearing is a fact finding hearing and the task of the tribunal is only to find and record the agreed facts ready for the IEs to prepare their report and the Stage 3 Hearing. In this case I have rejected Mr Reade’s trenchant criticism that the tribunal’s judgments were not fit for purpose and were worthless, fit only to be discarded wholesale. He said there needed to be a return to the beginning, to start again on the basis of the evidence that had already been given and transcribed and re-do the exercise. Although I have upheld some of the grounds of appeal, the wholesale challenge is rejected. There is a baby to be saved in the bathwater.

But I also acknowledge that there will need to be very great care and attention paid to the disposal of this appeal to ensure that the IEs have the information they need in the format they require in order to produce the report they have been ordered to prepare which I am confident can be achieved. In the further case management of this appeal all parties and this tribunal will need to keep well in mind Kerr J’s observation in Tesco Stores Ltd v Element & Ors [2024] EAT 83:

“69.

This appeal shows that, in undertaking case management in equal pay litigation, there is nothing wrong with pressing ahead and forcing the pace; patience is not a virtue in equal pay litigation. The right of access to a tribunal and a remedy for any wrongs done within a reasonable time requires a continuing sense of urgency and momentum.”

In summary, grounds A2G2, A1G1 and A1G2, A2G3, A1G8, A1G5 and A1G7 (in so far as it relates to Mrs Worthington) are dismissed. Appeal grounds A2G1 and A2G8 succeed in part and appeal grounds A2G4, A1G6, A2G5, A2G6, A2G7, A2G9, A2G10 and A2G11 succeed.

As discussed and agreed with counsel at the hearing, disposal issues will be dealt with separately. There was insufficient time at the two-day hearing and it was in any event premature. Case management and further directions for the disposal hearing may be necessary to ensure that the tribunal, the parties and the IEs have the findings of fact in the format that will be of greatest assistance for Stage 3.

I wish to thank the parties – counsel, solicitors and their legal teams for all their hard work to date. I wish also to record my thanks for all the tremendous and efficient work of the EAT staff in managing the administrative aspects of this challenging case.

Abdar & Ors v Morrisons1811283/2018, 4 March 2024 table with three columns as follows:

Ms C Green and others vs Co-Operative Group Limited 2404988/2019, 15 January 2025 with six columns as follows:

Mrs A Ahmed & others vs Sainsbury’s Supermarkets Limited and Diamond DCO Two Limited, in voluntary liquidation (formerly Lloyds Pharmacy Limited) 2404988/2019, Stage 2 Judgment 28 February 2025 with six columns as follows:

A reference in this Schedule to a rule, is a reference to a rule in this Schedule unless otherwise

A reference in this Schedule to “these rules” is a reference to the rules in Schedules 1 and 3 unless otherwise provided.

Where the Tribunal has decided to require an independent expert to prepare a report on the

— (1) At a stage 2 equal value hearing] the Tribunal shall—

Subject to paragraph (3), the facts relating to the question shall, in relation to the question, be the only facts on which the Tribunal shall rely at the final hearing.

At any stage of the proceedings the independent expert may make an application to the Tribunal for some or all of the facts relating to the question to be amended, supplemented or omitted.

The Tribunal shall give the parties reasonable notice of the date of the stage 2 equal value hearing and the notice shall draw the attention of the parties to this rule and give notice of the standard orders in rule 7.

The Tribunal may add to, vary or omit any of the standard orders in paragraph (1).

—(1) Where an independent expert has prepared a report, unless the Tribunal determines that

Reasons

The Tribunal shall give reasons for its decision on any disputed issue, whether substantive or procedural (including any decision on an application for reconsideration or for orders for costs, preparation time or wasted costs).

In the case of a decision given in writing the reasons shall also be given in writing. In the case of a decision announced at a hearing the reasons may be given orally at the hearing or reserved to be given in writing later (which may, but need not, be as part of the written record of the decision). Written reasons shall be signed by the Employment Judge.

Where reasons have been given orally, the Employment Judge shall announce that written reasons will not be provided unless they are asked for by any party at the hearing itself or by a written request presented by any party within 14 days of the sending of the written record of the decision. The written record of the decision shall repeat that information. If no such request is received, the Tribunal shall provide written reasons only if requested to do so by the Employment Appeal Tribunal or a court.

The reasons given for any decision shall be proportionate to the significance of the issue and for decisions other than judgments may be very short.

In the case of a judgment the reasons shall: identify the issues which the Tribunal has determined, state the findings of fact made in relation to those issues, concisely identify the relevant law, and state how that law has been applied to those findings in order to decide the issues. Where the judgment includes a financial award the reasons shall identify, by means of a table or otherwise, how the amount to be paid has been calculated.