This is the appeal of the claimant in the employment tribunal from the decision of the tribunal at Watford, Employment Judge R Lewis, arising from a hearing conducted by CVP over an initial day in December 2021 and then a further day in January 2022. At a rule 3(10) hearing, at which the claimant had an ELAAS representative, a single amended ground was permitted to proceed. It relates solely to the tribunal’s decision to refuse applications to amend made by the claimant in respect of claims for wages by way of the national minimum wage (NMW) and for holiday pay.

In the tribunal the claimant was in person and the respondent was represented by Mr Feeny of counsel. Before me, the claimant has been represented by Mr Findley of counsel, under the auspices of Advocate. Mr Feeny has appeared again for the respondent.

The tribunal described the respondent as providing a service to estate agents and letting agents, applying the Uber business model to the function of property viewing. Individuals who conduct viewings are referred to by the respondent as “viewbers”. On a given day of the week there were three time slots: morning, afternoon and evening. When an opportunity became available to show a particular property at a particular time, it would be offered to all viewbers who had offered availability to show in that postcode in that time slot, on a first come, first served basis. The respondent was not obliged to offer any opportunities, and viewbers were not obliged to accept any that were offered.

The respondent’s procedures also catered for viewbers holding keys to some properties, to enable them conveniently to carry out showings, as and when offered and accepted by them.

The claimant did work for the respondent as a viewber in a period between August 2018 and July 2020. When registering with the respondent, he ticked availability for all slots apart from Sunday morning and Sunday evening. The tribunal had evidence of how much he had earned, and reckoned that this equated to him having carried out around 150 or more viewings, over that period.

After the relationship came to an end, the claimant presented a claim form to the employment tribunal on 21 October 2020, acting as a litigant in person. He brought claims under the Equality Act 2010. He also ticked the box to indicate that he was claiming notice pay, and the particulars of claim also set out other alleged breaches of contract.

The appeal before me does not now relate to the Equality Act claims and I do not need to consider them in any detail. But I should note that they related to a viewing opportunity, which was then later extended to embrace a second viewing of the same property, on 22 July 2020, to the parties’ communications, or attempts to communicate, in relation to that, and to the respondent’s handling of the aftermath. The complaints were later analysed as being of direct discrimination, harassment and victimisation. As the judge did, I will refer to them compendiously as the discrimination claims.

A response defending the claims was submitted during December 2020. The respondent was represented by solicitors. On 2 March 2021 the tribunal directed that there be a public preliminary hearing on 30 July 2021 to consider whether any of the claims should be struck out, or made the subject of a deposit order, as having either no, or little, reasonable prospect of success, or neither.

On 23 March 2021 the claimant sent an email which included an application to amend to add a complaint that the respondent had failed to pay him the national minimum wage (NMW) in respect of the period August 2018 to July 2020 for what he described as “key-holding jobs”. On 24 March the claimant emailed the tribunal again applying to withdraw that claim. But, as we shall see, in any event, at the hearing with which I am concerned, the tribunal treated that potential complaint as first having been raised by him in the 23 March 2021 email.

On 12 June 2021 the claimant served an amended schedule of loss referring to claims for the NMW and, for the first time, for holiday pay. Under the heading “unlawful deduction from wages”, he advanced a case that, in addition to carrying out viewing assignments, he had done further work in relation to the holding and management of keys, amounting to 24 hours, 7 days per week, in the period 29 August 2018 to 30 July 2020. Applying the applicable NMW hourly rates from time to time, his schedule calculated that he was due £68,921.16 for day work during that period, and the same amount again for night work. The schedule went on to assert that the claimant had accrued annual leave entitlement throughout the second of two annual leave years covered by that overall period, which was accrued but unpaid on termination. Applying NMW rates on that basis, to a twelve-hour day throughout that holiday year, the claimant calculated that the 28 days’ leave that he had accrued was valued at £2292.22.

The hearing listed for 20 July 2021 did not go ahead on that date. I am told that this was because no judge was available. It was relisted for 6 December 2021. However, during the course of that day, the first of what proved to be the two days of the hearing, the judge decided that insufficient time had been allocated, further steps were needed, and that a second day should be added. The case-management minute records that, at the start of the hearing, the claimant withdrew the claim for breach of contract. There was discussion as to how he was putting the NMW claim, and of the significance of the decision in Royal Mencap Society v Tomlinson-Blake [2021] UKSC 8; [2021] ICR 758. The particular matters covered by the discrimination claims were identified.

The hearing was adjourned to 14 January 2022. The respondent was directed to send the claimant skeleton submissions in between, to assist him to understand how the respondent was putting its arguments. The claimant was permitted, but not required, to put in, sequentially, a skeleton in response. In the event, each of them tabled skeleton arguments. The claimant also tabled a further revised schedule of loss. In that revised schedule the claimant reframed his calculation of the NMW claim, taking account of the Royal Mencap Society decision, on the footing that he was claiming for the whole period from August 2018 to July 2020, in respect of 12-hour days, the claim totalling, as before, £68,921.16 for the days, but revising his claim in respect of nights, on the footing that he was now claiming for a sleep-in night shift, with a value calculated as being £15,712.05.

At the resumed hearing on 14 January the tribunal heard oral argument from both parties. At the end of the day a short oral decision was given. Then, as requested, written reasons were subsequently issued with the written judgment. The breach of contract claim was dismissed upon withdrawal. The discrimination claims were struck out as having no reasonable prospect of success. The amendment applications were refused. The notice money claim was dismissed on the basis that, as a breach of contract claim, it had been withdrawn, and as a wages claim it had no reasonable prospect of success. The overall proceedings were therefore dismissed in full.

The written reasons identify that at the hearing the parties had disagreed as to whether the judge should decide the strike-out issue first, on the footing that, if all the live complaints were struck out, the application to amend would fall away, or whether the judge should consider the application to amend first. In his decision the judge stated that the first approach “seemed to me correct logically and chronologically”; but, in the event, he decided all points on both matters. He stated for the avoidance of doubt that he would have refused the amendment application, even had he not entirely struck out the discrimination claims.

A substantial part of the reasons were devoted to the application to strike out the discrimination claims. There is then a heading “Discussion of amendments”. I will set out the section of the reasons that follows it, in full.

There is then a final short section in which the judge disposes of the notice money claim, whether as a contract or a wages claim; and a final observation, that the judge had made no adjudication of whether the claimant’s relationship with the respondent met any of the potentially applicable definitions of employee or worker, which the judge accepted may be fact-sensitive.

There is one amended ground of appeal that is live before me, in the following terms.

There was no dispute before me as to the relevant guidance in the authorities relating to applications to amend. Both counsel referred of course to Selkent Bus Co Ltd v Moore [1996] ICR 836, and, in particular, to the statement at point (4) of the guidance given by the EAT in that case at 843D: “Whenever the discretion to grant an amendment is invoked, the Tribunal should take into account all the circumstances and should balance the injustice and hardship of allowing the amendment against the injustice and hardship of refusing it.” The overriding importance of balancing hardship against hardship in this way has recently been emphasised again by the EAT in Vaughan v Modality Partnership [2021] ICR 535.

The essential overarching issue raised by this appeal is whether, on a fair reading, the tribunal did properly consider, and weigh, what the hardship to the claimant would be, were his amendment applications to be refused, against what the hardship to the respondent would be, were they to be granted. Mr Feeny submitted that, in this regard, the tribunal’s decision must be read fairly as a whole, and in context; and he cited the well-known observations in Fuller v London Borough of Brent [2011] EWCA Civ 267; [2011] ICR 806, warning against an appellate court engaging in a fussy, over-pernickety critique or being hypercritical of the way that the decision was written.

Mr Findley accepted that the tribunal’s decision should be read fairly and as a whole, but submitted that Fuller was not really in point, as his challenge was that there had been a simple failure by the tribunal to address matters that it should have addressed. He submitted that the words of Sedley LJ in Anya v University of Oxford [2001] EWCA Civ 405; [2001] ICR 847 were more in point, where he said, at [26], that it is not 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.”

A further point debated in the skeleton arguments and oral argument today concerned the fact that the tribunal did not in its decision cite Selkent, Vaughan, or any other authority, nor state the guiding principles emerging from the authorities, in relation to applications to amend, even briefly. The position in this regard is well-established. Tribunals ought in their decisions to give some statement of the relevant law, and indeed rule 62(5) enjoins them to do so. However, a failure to do so does not necessarily mean that the tribunal has made an error of law such that an appeal against its decision should succeed. What matters is whether the EAT can be satisfied that the tribunal has in fact understood and applied the law correctly when reaching its decision. If the tribunal has stated the law correctly, then the EAT may be slow to infer that it has not applied it correctly, although merely referring to the relevant authorities may not be sufficient to assuage any such doubts.

If the tribunal has, as here, not stated the law, then the EAT may be left more uncertain, from the language of its dispositive reasoning alone, as to whether it has in fact adopted the legally correct approach. Having said that, it was part of Mr Feeny’s case that I should also read what this tribunal said in its decision in the context of the fact that, between the two days of the hearing, the parties had tabled skeleton arguments, and these had referred to the law and set out their respective cases very fully. So, he said, it could be inferred that the tribunal had all of this material very much in mind. I also myself observe that the decision was addressed first and foremost to the parties themselves, who would have been familiar with those arguments and how they had been deployed.

However, Mr Findley submitted that this was not sufficient to make good the deficiencies in the tribunal’s decision. He accepted that the tribunal did not need to use any particular designated form of words to indicate, for example, that it had weighed up the balance of hardship to the two parties of either refusing or granting a particular amendment. But it did still need to demonstrate that it had considered relevant matters falling on both sides of the ledger, and that it had then weighed them up in order to come down on one side or the other.

Further, submitted Mr Findley, the tribunal had not only failed to address the balancing exercise, but had, in fact, further erred by treating two factors as knock-out or determinative points, being what it called the “extensive recasting” of the claim that the amendments would entail, and the points it made about the amendments having been advanced at a point in time when, as independent claims, they would have been significantly out of time. It was an error, he submitted, to treat such points as determinative, rather than weighing them in the scales – a point well-established by authorities such as TGWU v Safeway Stores Ltd, UKEAT/0092/07 and Abercrombie v AGA Rangemaster Ltd [2013] EWCA Civ 1148; [2014] ICR 209. He acknowledged that this was not a perversity challenge. But, he submitted, the tribunal’s reasoning was fundamentally deficient in these respects, and the skeleton arguments could not be resorted to, in order to fill the gaps.

Mr Findley submitted that there were five respects in which the tribunal’s reasoning was deficient and showed that it had failed to consider matters that it ought to have considered, although the second and third formed an overlapping pair. These were: (1) the impact on the claimant of a refusal of permission to amend, in particular the fact that he would lose the chance to pursue potentially extremely valuable claims; (2) the evidential consequences of granting the amendments; relatedly (3) the overlap between the factual issues raised by the existing complaint and the proposed new complaints; (4) failing to take account of the impact of the second and third UK lockdowns on the timing of when the claimant advanced his applications to amend; (5) failing to take account of, or address, differences between the proposed holiday-pay claim and NMW wages claims.

As to whether I can be confident that the tribunal applied the correct approach applications to amend, I note that the essential authorities, not only Selkent, but authorities such as TGWU v Safeway and Abercrombie, are extremely familiar to employment tribunals, regularly cited and regularly applied by them. In particular the overriding requirement to consider the overall balance of hardship has recently been restated by the EAT in Vaughan and in Chaudhry v Cerberus Security and Monitoring Services Limited [2022] EAT 172. Furthermore, both Selkentand Vaughan were specifically cited by the respondent in its skeleton argument, and it fairly highlighted the overriding test being balance of hardship. All of that being so, and though the judge ought to have said something, however brief, about the law, in his decision, I think I can proceed on the basis that he understood and had in mind the general principles that he ought to be applying.

However, Mr Findley contends that I cannot be confident that the judge properly considered the specific features that he highlighted in his submission, given the brevity and lack of coverage of the discussion in the dispositive part of the reasoning at [52] to [59]. I turn to that aspect.

I start by observing that it seems to me that [52] is about both the NMW and proposed holiday pay claims, [53] – [58] are then about the NMW claim, and [59] is then about the holiday pay claim.

As to the first matter that Mr Findley says the judge neglected to address, being the potential high value in particular of the NMW claim that the claimant was seeking to advance, it seems to me that the tribunal plainly had this in mind. The judge had plainly seen the schedule of loss and the calculations of how the claimant had arrived at the figures claimed. He had noted earlier in the decision that the claimant had reckoned that he had had an income of about £2600 over the 22-month or so period, from the time spent on actual viewings. At [53] he identified that proposed wages claim was not for further payment in respect of assignments carried out, but for payment in respect of what the claimant claimed should count as working time in the course of the relationship outside of time spent carrying out actual viewings. At [54] he identified that the claim was now being put on the basis of a claim for every day of the year during the whole period of the relationship; and he reckoned, whether or not accurately, that this equated to 100 years of the actual earnings. The broad point is that the judge plainly had in mind the very substantial sum that the claimant was seeking to claim.

The point was made to me in submissions by Mr Feeny that this potentially cut both ways. The claimant, putting his proposed claims at their highest, potentially stood to lose the opportunity to pursue a very valuable pair of claims. But the respondent was correspondingly at risk of further significant financial exposure, if the applications were granted, putting matters at their potentially worst for it. I do not think that the judge needed to spell this out. It was plain and obvious, and he plainly understood, on the claimant’s case at its highest, what was potentially financially at stake.

As to whether the judge considered whether allowing the amendment to introduce the NMW claim would significantly alter the evidential canvas, the tribunal described the proposed amendments as involving an “extensive recasting” of the claim. Mr Feeny submitted that I could infer that the judge had in mind the respondent’s submission that this would considerably widen the evidential canvas, because, for example, consideration would be needed of further evidence about the arrangements for keyholding and what this involved, and in order to identify during which particular periods the claimant was holding one or more sets of keys, and so forth.

Mr Findley submitted that the expression “extensive recasting” was at best ambiguous. I could not be sure that the tribunal was thereby referring to the respondent’s submissions about the impact of the amendments on the evidential canvas, with which, in any event, he disagreed. He submitted that the tribunal might be referring here to no more than the fact that the amendments would introduce different legal types of complaint, or, at best, different legal types of complaint with a significant additional value. But a failure specifically to consider whether allowing them to be added would alter in substance the evidential territory that would need to be covered would be an error. See the discussion in Abercrombie. Once again, although he disagreed with the tribunal’s assessment, Mr Findley accepted that this was not a perversity challenge. But this was an essential consideration that the tribunal needed to address, and I could not infer that it had had it in mind.

On this point, it is correct that the tribunal did not expressly explain what it meant by “extensive recasting”, nor did the tribunal specifically discuss anywhere whether allowing the NMW complaint amendment would result in the respondent having to gather and adduce further significant evidence. This strand of the challenge has given me some pause, but in the end it does not succeed.

First, I think it is clear that the reference to “extensive recasting” is not to the fact that NMW/wages complaints would be different legal types of complaint from the current live complaints, although indeed they would. The passage which this phrase introduces, beginning at [53], highlights, correctly, that the NMW complaint was not a complaint about what the claimant had been paid for viewing jobs that he had carried out. The judge’s point is that the proposed NMW complaint was different in its subject matter and went back over the whole course of the 22-month relationship. The point being made here was not therefore about it being an NMW complaint about working time, as such, but about the factual scope of it.

While the judge does not specifically refer to it, it is also a plain fact that the discrimination claims in this case were very narrow in their factual compass, being specifically about the events or alleged events relating to specific bookings on the 22 July 2020, and the aftermath of that. The breach of contract claims had gone, and the notice money claim, prior to withdrawal, focussed on the very end of the relationship, including the import of a letter or email that the claimant had sent in August.

The tribunal also had identified and plainly understood, that the basis on which the potential NMW complaint was being advanced was that what the claimant called keyholding time or keyholding work, should, as a matter of fact and law, be treated as working time for which he was entitled to be paid throughout the relationship. The tribunal highlights it being about keyholding time at [54] and goes on to consider the merits of the claim, put in that factual way, at [55].

Whilst the tribunal also referred in this passage to the monetary value being placed on the claim, that came in the course of a discussion over the course of [53] and [55] about the underlying nature and factual basis on which the proposed complaint was being advanced. It was plainly also one of the considerations that the tribunal took into account, as it was entitled to do, that it considered the prospects of success of such a proposed complaint to be weak. It indicated this not only at [55] but, earlier in its decision, at [9], when it indicated that, had it allowed it to proceed, it would have made a deposit order in respect of the claim, which it there succinctly described as being for entitlement to the minimum wage for keyholding duties.

It therefore seems to me that the tribunal has shown that it had in mind the substance of the proposed NMW complaint, in terms of the very different legal and factual issues that it raised from the existing claims. It was not about whether the claimant was a worker or an employee when carrying out actual assignments, but raised a distinct issue about whether what the claimant called key-holder time was working time, and, with that, inherently, the factual issue as to what period or periods of time during the course of the 22 months such a claim would cover. The tribunal therefore it seems to me did consider this and was entitled to place it on the respondent’s side of the scales.

The third point, as I have indicated, and I think Mr Findley accepted, essentially overlaps with the second point, and I have addressed it, save that there may be a distinct point in relation to one way of looking at a sub-element of the proposed holiday-pay claim, to which I will return.

As to the fourth point, the tribunal in terms addressed the claimant’s submission about lockdown at [56] as well as other factors that the claimant relied upon as having an impact on his ability to include these complaints as part of his original claim. Mr Findley submitted that, while it had addressed the period covering when the claim itself was presented, the tribunal had not addressed the further period during which such a claim would potentially have been in time, extending into early in the new year, when further lockdowns or local lockdowns were in force. However, given what the tribunal said at [56] I do not think it needed to address the claimant’s arguments in greater detail than it did. It clearly was not persuaded that the impact of a lockdown or lockdowns provided a compelling explanation for why the claimant could not still have done the research that he needed to do, or got the advice that he may have needed, to enable him to cover this ground in a timely claim.

I turn then to the final point, relating to the holiday pay claim. It is apparent from [59] that the judge considered that his reasoning in relation to the NMW claim applied equally to the holiday pay claim; but that additional points on the respondent’s side of the scales were that the claim form has a specific box relating to a holiday pay claim, but which the claimant had not ticked, and that there was what the judge called the additional passage of time, which I take to be a reference to the difference between March, when he first signalled that he might wish to bring an NMW claim, and June, when he first raised the possibility of a holiday pay claim.

Mr Feeny submitted that it was not an error for the tribunal to regard the points applying to the NMW claim as applying equally to the holiday pay claim, given that the holiday pay claim was framed as covering the entire period of the relationship, at a rate of twelve hours a day. It seems to me that, at its highest, it was indeed predicated on the same underlying proposition: that the claimant was, throughout that period, doing what he described as keyholding work, that this counted as working time, in respect of which he was not only, therefore, on his case, entitled to be remunerated with wages, but also entitled to be treated as having accrued holiday entitlement.

Mr Findley however submitted that, even if that was right, as such, which I think it was, there was a potential further point which the tribunal overlooked and did not address, being that the claimant had, within his proposed holiday pay claim, a potential narrower claim, which did not depend on the keyholder-working-time argument. The claimant had, more simply, a potential narrower holiday pay claim for payment in lieu, in respect of holiday entitlement claimed to have accrued during the period of viewing assignments which he in fact carried out. Mr Findley noted that the tribunal had not said that it would have ordered a deposit in respect of the holiday pay claim, nor had it determined whether the claimant was an employee or at least a worker when carrying out assignments. It had not needed to do that, given the basis on which the discrimination claims had in any event been struck out as having no reasonable prospect of success.

As to this, I do observe that the claimant argued for his proposed holiday pay claim purely by reference to the key-holding point. He did not himself expressly advance in the alternative a narrower claim to be entitled at least to pay in lieu of holidays accrued during the periods when he was actually carrying out viewing assignments. That being so, I am conscious that it might be said to be a little harsh to criticise the tribunal for not picking up on that embedded narrower scenario; but Mr Feeny did acknowledge in the course of argument, that this was a point that might potentially have been picked up on, and addressed. On balance I have come to the conclusion that the tribunal ought to have considered it, once the claimant had put a holiday pay claim into the mix, as something that he was seeking to introduce by way of amendment, covering the entire period of the relationship, and, by his calculations, therefore including periods during which he carried out actual assignments.

That said, the tribunal’s other points about why the claimant faced some additional difficulties, in relation to his application to amend to introduce a holiday pay claim, having regard in particular to the stage and time at which it was advanced, were, as such, points that the tribunal was entitled to raise and weigh in the balance.

As against this, I see force in Mr Findley’s submission, that the tribunal’s point in relation to the NMW claim involving an extensive recasting of the existing complaints, would not carry across to the narrow holiday pay claim, which would not depend on the assertion that key holding time was remunerative working time. I do not think it can be assumed that the evidential or legal implications of allowing that narrower claim to proceed would have been the same. It would not have involved the tribunal in considering whether keyholding was work, attracting a right to wages and to accrue holiday, nor the implications of Royal Mencap Society, or anything of that sort. It would also not have widened the evidential canvas in the same way, although further consideration might have needed to be given to the question of how precisely the evidence that the tribunal already had, of the amounts that the claimant had in fact invoiced and been paid, could or could not be translated back into a determination of the number of viewing jobs he had done, or hours he had put in to those jobs.

Pausing there, for the reasons I have given so far, I do not accept that, in relation to the NMW claim, the tribunal failed sufficiently to address in its decision the different factors that Mr Findley said ought to have been addressed and weighed in the scales. Nor did it so fail in relation to the related holiday-pay claim, that is to say, the wider holiday pay claim that was dependent on the keyholding argument. But I do accept that there was a failure by the tribunal to address the potential distinct, narrower and different embedded strand of the proposed holiday-pay claim, by reference solely to the time spent in fact carrying out actual viewing assignments.

Reverting to the NMW pay amendment application, Mr Findley, as I have indicated, argued that nevertheless the tribunal had also erred by not completing the exercise of weighing the balance of hardship, by explaining how it had come down on one side or the other. Rather, he argued, it had simply, on his case, treated the points about “extensive recasting”, and the claim being out of time, and the “not reasonably practicable” test for an extension of time, as being determinative.

As to this, it is correct of course to say that it would be an error for the tribunal to treat a point such as the fact that, if this proposed claim had been raised as a stand-alone claim, it would have been out of time, as automatically fatal. However, that does not mean that a tribunal cannot treat such a factor as highly significant and, potentially as determinative, in the sense of carrying such weight that it decisively tips the scales down on the side of the respondent. As Underhill J (as he then was) put it in TGWU v Safeway Stores at [10]: “Thus the reason why it is ‘essential’ that a tribunal consider whether the fresh claim in question is in time is simply that that is a factor – albeit an important and potentially decisive one – in the exercise of the discretion.”

In the present case, I cannot see from the language used that the tribunal made the mistake of treating either of these factors as determinative automatically in and of themselves. There is nothing to suggest, for example, that the tribunal considered that, because, had the amendment been raised as a freestanding new claim, it would have been out of time, and time would not have been extended, that simply and automatically led to the conclusion that it must be refused. The tribunal refers to a combination of features at [57]. Similarly, the point about “extensive recasting” was not, it seems to me, treated by the tribunal as necessarily or automatically determinative. Rather, the sense of this part of the reasoning is that these two factors weighed very heavily against the claimant and in favour of the respondent and so were effectively determinative of the balancing exercise in that sense.

It is unfortunate that the tribunal did not put in a concluding sentence or paragraph saying just that, but reading this part of the decision as a whole, I do not think that the reader is left really in any doubt that this was what the tribunal thought. It considered that the claimant’s case in favour of the amendment being permitted was a very weak one, having regard to the extensive recasting, the poor view that the tribunal took of the underlying merits and the time point. The tribunal, as I have said, plainly had in mind that, in so deciding, it was depriving the claimant of the opportunity to advance a claim which, at its highest, had potentially a very high value.

For all of these reasons this appeal does not succeed in so far as it relates to the refusal of the amendment to introduce the NMW complaint, nor in relation to the part of the proposed holiday pay claim that went hand in hand with it, being also based on the keyholder-working-time premise.

I return to the discrete basis for an embedded, narrower holiday pay claim. The tribunal plainly took the view that, so far as the time point was concerned, the claimant’s position was no better than it was in relation to the proposed NMW claim. Indeed it permissibly took the view that his position was worse. But I do accept that the tribunal erred by assuming uncritically that the point about there being a significant recasting of the original claims carried across. I cannot be sure that, if the tribunal had reflected on that aspect, it would, for certain, have considered that, even in respect of that narrower and discrete sub-claim embedded within the holiday pay claim, the balance still tipped against the claimant. Though realistically it is perhaps not very likely that the judge would have come to a different view, given the tenor of the overall reasoning, I cannot entirely rule it out.

So I must turn, in this respect, to the respondent’s second line of defence to this appeal. It argues that, as the judge struck out all of the existing complaints, there was, in any event, nothing left to amend. Therefore, in any event, he did not err in declining to permit this claim to be added. As I have described, there was argument as to whether the judge should therefore determine the application to strike out the existing complaints first. But, although the judge stated that, as he put it, that seemed to be correct logically and chronologically, he did not, in the event, take that approach.

In the course of argument I was referred to one authority, Sakyi-Opare v Albert Kennedy Trust, UKEAT/0086/20, in which it was held, on the facts of that case, that it would be an error to deal with a strike-out application without at the same time considering amendment applications. It is clear, however, that, on the facts of that particular case, that was because the two applications were closely related. The gist of the argument was that the proposed amendments, if allowed, might have made a meaningful difference to the complaints which were the subject of the strike-out application, and hence the outcome of that application.

In the present case, says Mr Feeny, the position is factually quite different, because the discrimination complaints in particular were factually completely different in their subject matter from the proposed amendments with which I am concerned, and there were no potential ramifications of one for the other. That said, Mr Feeny ultimately accepted in argument this morning that, in a case like this, where the tribunal is seized of both matters at the same hearing, and regardless of which application was made, in point of time, first, there can be no hard and fast rule for every case, as to whether the tribunal should consider the applications in a certain order or, indeed, hand-in-hand.

I agree. In a case where the tribunal is seized of two such applications on the same occasion, neither of which it has yet determined, it is, in principle, as a starting point, a matter for the exercise of case management discretion as to what approach to take. That discretion must of course be exercise judicially and in a principled way, and having regard to relevant considerations arising including the particular nature and subject-matter of the amendment and strike-out applications in the given case. On the facts of the present case, I do not think I can go so far as to say that the tribunal would have erred had it not determined the strike-out application first, and, if, as happened, the discrimination claims were struck out, then concluded that necessarily the amendment applications must fall away.

What remains is whether I can say that there could only be one right answer to the question of whether to permit the narrow holiday-pay claim amendment, or whether it needs to be remitted.

I have now heard further argument from both counsel about next steps.

In light of my decision the appeal will be dismissed in relation to the refusal of the amendment in relation to the NMW claim and the wider holiday pay claim put on the basis of the keyholder-working-time argument. In relation to the proposed claim that holiday accrued during the time spent on actual viewings during the final year, counsel agree, as do I, that I do not have sufficient material before me to be able to redetermine whether the amendment should be permitted. It must therefore return to the tribunal to consider. It should do so on the basis that, while the findings relating to other matters relating to the balancing exercise are a given, the judge will need to consider, and weigh into the balance, whether this proposed claim would, of itself, have involved a significant recasting of the evidential canvas of the original complaints, and the significance of EJ Lewis’ decision that the other complaints had no reasonable prospect of success, on the footing that this amendment application is still to be treated as if it were being considered on the same occasion as the strike-out applications.

During the course of argument I was told that the claimant had raised a complaint about EJ Lewis, though nothing more about it. That is wholly irrelevant to my decision. The fact that a party has raised a complaint does not, as such, of itself entitle them to insist that the judge should be recused from further involvement in the matter. I put that entirely to one side.

I bear in mind Mr Feeny’s submission that the matter being remitted was not considered by the judge hitherto, as a discrete sub-strand, either way. But I also bear in mind that it is important that the judge is able to look at the application with a fresh eye. I think it would be a big ask to expect EJ Lewis to put entirely out of his mind the previous conclusions that he reached on the wider basis of the proposed holiday-pay claim, which effectively led him to refuse to allow it to be added entirely. The issue – in terms of whether to allow the narrow claim to proceed by amendment as such – is also now a very narrow one, and another judge will have the benefit of the broad background work done by EJ Lewis in his decision. The hearing to deal with this point should be relatively short.

I will therefore direct that this issue be remitted for consideration by a different judge.