Mr T Mushtaq and others v Addison Lee Ltd: 2207566/2021 and others
EMPLOYMENT TRIBUNALS
Case No 2207566/2021
Between
Mr T Mushtaq & Others (see attached schedule)ClaimantAddison Lee LimitedRespondentDate 6 January 2025
JUDGMENT
The claim is struck out.
REASONS
[1]By a letter dated 6 March 2024 the Tribunal gave the claimant an opportunity to make representations or to request a hearing, as to why the claim should not be struck out because it has not been actively pursued.[2]The claimant has failed to make representations in writing, or has failed to make any sufficient representations, why this should not be done or to request a hearing. The claim is therefore struck out.[1]In paragraphs 2-7 below, I state what happened at the hearing which I conducted on 29 October to 18 November 2024 inclusive (although not on all of those dates). In paragraphs 8-12 below I refer to aspects of the procedural history and state the issues which I was at the start of the hearing being asked by the parties to determine. In paragraphs 13-18 below, I describe the evidence which was before me during the hearing. In paragraphs 19-26 below I refer to further aspects of the procedural background to the hearing. In paragraphs 27- 44 below, I make observations on the evidence which gave rise to the claimants’ strike-out application put before me on 29 October 2024 to which I refer in paragraph 4 below. In paragraphs 45-230 below, I state my material findings of fact. I was required by the parties to determine a number of fairly stark conflicts of evidence, and in order to explain how I made such determinations, I found that it was (regrettably) best to set out some lengthy passages from the evidence before me and the parties’ submissions. In paragraphs 231-268 below, I discuss several aspects of the relevant case law and the parties’ submissions on them, and, finally, in paragraphs 269-335 below I set out my conclusions on the issues which, by the time of closing submissions, I was being asked to determine. Because of the length of the part of this document stating the factual background and my findings of fact, by way of an index to paragraphs 45-230 below, I now set out the headings and the sub-headings in that passage, with a statement in brackets of the paragraph numbers which contain the text below those headings. Index to the part of this document where I describe the factual background and state my findings of fact1.1 An overview of the manner of the engagement of the test claimants by the respondent to provide driving services and some further observations on the evidence before me (45-60).1.2 The relationship between the respondent and drivers who hired their vehicles from Eventech (61-68).1.3 The manner in which non-partner drivers worked for the respondent.1.3.1 The determination of the rate of pay for work and the software used by the respondent for allocating the work (69).1.3.2 In what way could a non-partner driver reject a job (70-76).1.3.3 Did the respondent continue to impose economic sanctions for refusing jobs after the Lange judgment was given? (77-114).1.3.3.1 Passenger drivers (77-107).1.3.3.2 Courier drivers (108-113).1.3.3.3 The statistical evidence before me about the proportion of jobs which the respondent’s drivers rejected between 2019 and 2024 (114).1.3.4 The time it might take for a driver’s call to be answered by the respondent’s operational teams (115-120).1.3.5 The overall picture: the effect of what the claimants called the respondent’s “carrot and stick” approach (121).1.3.6 The respondent’s practices in relation to the manner in which drivers dressed (122-144).1.3.7 The respondent’s allegation in paragraph 13(2)(ii)(h) of its amended grounds of resistance at page 172 that ‘There is no longer a “half-hour rule” for logging into the XDA before pick-up time. No Drivers have been specifically required to log on to their XDA 30 minutes before a pre-booked job since September 2017’ (145-149).1.3.8 The respondent’s allegation in paragraph 13(2)(ii)(i) [sic] of its amended grounds of resistance at page 172 that “Drivers no longer have to be in the vehicle and away from home in order to log in. They may log in from anywhere. This change occurred in 2019” (150-152).1.3.9 The respondent’s allegation in paragraph 13(2)(ii)(j) of its amended grounds of resistance at page 172 that ‘The “go home button” is now available at any time. There is no longer a requirement to be logged in for 4 hours before it can be used and no limit to the number of times a day it can be used. This change was introduced on different dates for different types of driver but in all cases by the end of 2018’ (153-157).1.3.10 The respondent’s allegation in paragraph 13(2)(ii)(o) [sic] of its amended grounds of resistance at page 173 that “from June 2021, sub-contracting has been permitted (in accordance with the full recasting of contract terms in June 2021)” (158-163).1.3.11 Relevant aspects of the relationship between the agreement to lease a vehicle and a driver agreement with the respondent (164- 173).1.3.11.1 Insurance (165-166).1.3.11.2 Aspects of the impact of the vehicle hire agreement on the relationship between a driver and the respondent (167-173).1.4 The manner in which courier drivers worked for the respondent in so far as it differed from the manner in which passenger drivers worked for the respondent (174-191).1.5 Additional factual findings concerning the relationship between the respondent and partner drivers (192-205).1.5.1 Sanctions (193-196).1.5.2 Multi-apping (197-205).1.5.2.1 Partner and non-partner courier drivers (198-203).1.5.2.2 Partner and non-partner passenger drivers (204-205).1.6 The periods when the test claimants provided driving services to the respondent (206-230)1.6.1 Introduction (206-215).1.6.2 The position of Mr Nardelli (216).1.6.3 Mr Balog’s holidays (217).1.6.4 Mr Ban’s holidays (219).1.6.5 Mr Da Silva’s holidays (220).1.6.6 Mr Ruiz’s holidays (221-222).1.6.7 Mr Edah-Tally’s holidays (223-224).1.6.8 Mr Payne’s holidays (225).1.6.9 Mr Kidd’s holidays (226).1.6.10 Mr Mahendran’s holidays (227-228).1.6.11 Mr Klepacki’s holidays (229-230). The hearing which I conducted in October and November 2024[2]The hearing which I conducted on the dates stated above was intended to be only about the issues which are the subject of the above judgment. That was on the basis that(1) there were eleven claimants who had been selected by the parties as test claimants, but not on the basis that rule 36 of the Employment Tribunals Rules of Procedure 2013 applied to them so that they were not lead cases within the meaning of that rule, and(2) I would hear oral evidence in relation to those issues from those test claimants and oral evidence from five witnesses for the respondent. The test claimants were all represented by the firm of solicitors by the name of Leigh Day.[3]The hearing was listed to start on 28 October 2024, but I was not available on that day so it started on the following day. In addition, I had a part-heard case on Friday 1 November 2024, so that day was not capable of being used for the hearing in this case.[4]Despite the loss of those two hearing days, I was presented at the start of the hearing on 29 October 2024 with an application made by the claimants for the striking out of part of the respondent’s amended grounds of resistance. That application was the subject of(1) oral evidence from two of the respondent’s five witnesses and(2) submissions, on Tuesday 29 October 2024 and Thursday 31 October 2024. The witnesses gave evidence on the second of those two days. I gave a reserved decision on the application (in fact relatively late) on Friday 1 November 2024. Accordingly, there was no need to record in this document the reasons why I dismissed the application, but the evidence which I heard in order to determine it (and I heard that evidence on the application of the claimants) was relevant to the issues which are the subject of the above judgment. I therefore refer below to that evidence where relevant.[5]I heard oral evidence from the test claimants and the respondent’s witnesses (including further evidence from the two witnesses who had given evidence on Thursday 31 October 2024 and from a sixth witness in the manner described in paragraphs 15-18 below) on Monday 4 November to Friday 8 November and Monday 11 November to Wednesday 13 November 2024. I then spent a day reading some of the applicable case law and re-reading the claimants’ detailed opening skeleton argument in preparation for reading written closing submissions and hearing oral submissions from the parties, while the parties worked on written closing submissions. I did so on the basis that I would hear first from the respondent, on Friday 15 November 2024, and then, on Monday 18 November 2024, from the claimants, after which the respondent would have an opportunity to respond to the claimants’ closing submissions.[6]I received the respondent’s written closing submissions in the morning of Friday 15 November 2024, and I heard oral submissions from Mr Leiper on behalf of the respondent from 12 noon onwards on that day. The hearing ended midafternoon.[7]By agreement between the parties, the claimants’ written closing submissions were sent to the respondent on Sunday 17 November 2024 and copied to me. I then read those closing submissions in the first part of Monday 18 November 2024, and from 11.00am onwards I heard oral submissions from Mr Segal in support of, and supplementing, those written submissions and oral submissions in reply from Mr Leiper. The preceding preliminary hearings[8]The hearing which I conducted was preceded by several preliminary hearings which were conducted by Employment Judge (“EJ”) Tynan, who sits at Cambridge and conducted the hearings by video (CVP, i.e. Cloud Video Platform). I refer below to the content of EJ Tynan’s written records of those hearings only where it is necessary to do so. The agreed list of issues[9]On 20 February 2024, EJ Tynan directed (via his order number 5 of that day, at page 181) the parties to agree a list of issues by 5 March 2024. The agreed list of issues was at pages 184-191 of the main hearing bundle which was put before me for the purposes of the hearing which I conducted. The bundles of documents before me[10]In addition to that main hearing bundle (which contained 8860 pages, although its final page was numbered 8856), three bundles were put before me during the hearing. In what follows a reference to a page without any qualification is a reference to a page of the main hearing bundle. I have referred to the three supplemental bundles below as SB1, SB2 and SB3 respectively. Accordingly, reference to pages of the supplemental bundles are to pages of SB1, SB2, or SB3 as the case may be. For example, a reference to page 1 of SB1 is a reference to page 1 of the first supplemental bundle.[11]The content of the first supplemental bundle led to the application made by the claimants to strike out part of the response. That first supplemental bundle was sent to the claimants on Sunday 27 October 2024 in the manner to which I refer in paragraph 14.1 below. The second supplemental bundle was sent to me (in the manner which I describe in paragraph 15 below) on Tuesday 12 November 2024, and the third one was sent to me (in the manner which I describe in paragraph 18 below) on 13 November 2024 along with a witness statement made by Mr Sanj Gherra in the circumstances which I describe in paragraphs 15-18 The issues which at the start of the hearing I was being asked to determine[12]The parties agreed that I did not need to determine all of the issues in the agreed list of issues. Those which I was at the start of the hearing asked to determine were as follows (and I set them out verbatim with minor clarifications in square brackets). “Worker Status and Working Hours[1]Were any of the Claimants ‘workers’ employed by the Respondent for the purposes of ERA [i.e. the Employment Rights Act 1996; I refer to that Act below as “the ERA 1996”], WTR [i.e. the Working Time Regulations 1998, SI 1998/1833; I refer to them below as “the WTR”] and NMWA [i.e. the National Minimum Wage Act 1998, to which I refer below as “the NMWA”]?[2]If so, which Claimants were ‘workers’ employed by the Respondent and at what dates? For the avoidance of doubt, this issue includes whether an individual who was a worker ceased at any point to be a worker (whether in or about June 2021 or otherwise and if so when).[3]If and in so far as any of the Claimants were at any time ‘workers’ employed by the Respondent, can they (subject to clause [i.e issue] 5 below) in principle qualify as ‘working’ under a ‘worker’ contract at times when their ‘app’ was switched on but they had not yet accepted a passenger journey (Uber v Aslam [2021] ICR 657 SC paras 121-130 refer)? (It may be relevant to this issue to decide whether there was any or sufficient obligation to work in this period).[4]If and in so far as the answer to 3 is ‘yes’, was each Claimant, on the facts of each individual case relating to each period for which a claim is made (or which is relevant to the amount of the claim) ‘ready and willing’ to work so as to qualify in fact as ‘working’ under a worker’s contract at that time? (Uber v Aslam [2021] ICR 657 SC paras 121- 130, esp. para 130 refer).[5]In light of the answers to clauses 3 and 4, were the particular periods for which claims are made (or which are relevant to the amount of claims) when the Respondent’s app was switched on but no passenger journey had been accepted to be treated as working time:a. For the purposes of WTR?b. For the purposes of NMWA and NMWR [i.e. the National Minimum Wage Regulations 2015, SI 2015/621]? Holiday pay The applicability of EU law[6]If and in so far as the Claimants rely on EU law to establish rights in relation to holiday and holiday paya. whether by virtue of Protocol No 30 on the Application of the Charter of Fundamental Rights of the European Union to Poland and the United Kingdom, European law in relation to holiday and holiday pay (including Article 7 WTD [i.e. the Working Time Directive, Directive 2003/88/EC]) does not (or does not in relevant respects) have effect against the Respondent, a private party; alternativelyb. whether any such effect against the Respondent as a private party is by virtue of Section 5 of the European Union (Withdrawal) Act 2018 (and Schedule 8) inapplicable to i. claim forms presented after 31st December 2020 ii. amendments granted after that date. (“the Applicability of EU Law Issue”) Pay for holiday taken or deemed to have been taken[7]To how many days of leave, if any, was each Claimant entitled under WTR in respect of the contract(s) under which he or she was engaged?[8]How many days leave were taken by each such Claimant under each contract in respect of the leave year(s) for which a claim is asserted?[9]Are any of the Claimants to be treated as having taken leave pursuant to reg 13 and/or reg 13A WTR because of the Respondent’s refusal to remunerate them in respect of such leave?[10]Was any leave validly carried over by any Claimant froma. one relevant leave year to anotherb. one contract to another and, if so, how much?[11]For how many days was each Claimant paid holiday pay for the leave years in question and in what amounts?[12]In computing the sums already paid to the Claimants what allowance is to be made in each casea. for credits earned by Claimants used to set off car-rental liabilityb. for payments on account of holiday made since June 2021?[13]Does any failure by the Respondent to make payments due to drivers pursuant to reg 16 WTR give rise a series of unauthorised deductions from the driver’s wages during their time working for the Respondent, contrary to sections 13 and 23 ERA?[14]Subject to the Applicability of EU Law Issue:-a. Is any claim brought outside the primary time limit?b. Do the principles of EU law established by inter alia King v Sash Window Workshop & another [2018] ICR 693, Max-Planck v Shimizu [2019] I CMLR 35, Stadt Wuppertal v Bauer Willmeroth [2019] 1 CMLR 36 and Kreuziger v Land Berlin [2019] 1 CMLR 34, Article 47 of the Charter of Fundamental Rights of the European Union and/or the principles of effectiveness and/or equivalence generally, require any of the applicable limitation periods and/or retrospectivity limits to be interpreted purposively, or alternatively disapplied,? If so, to what extent and in what circumstances?c. Further and in particular, is the limitation to the two year period prior to the institution of proceedings under s23(4A) ERA ineffective, i. as being incompatible with the principles of effectiveness and equivalence under EU law and/or Art 47 of the Charter of Fundamental Rights and/ or ii. on the basis that in providing for the insertion of section 23(4A) into ERA, the Deduction from Wages (Limitation) Regulations 2014 were ultra vires section 2(2) European Communities Act 1972 and/or iii. as being in breach of Article 1, Protocol 1, Schedule 1 of the Human Rights Act 1998? 17 [S]hould reg 30(1)(a) WTR be read so as to mean that a failure or refusal to pay a worker for the leave to which he is entitled under regs 13 and/or 13A WTR amounts to a failure or refusal to permit a worker to exercise the right to paid annual leave under EU law, such that compensation for refusing to permit the worker to exercise the right in full falls to be assessed under regs 30(3)(b) and 30(4) WTR? “Pay in lieu on termination” claims[19]Should any Claimant in any leave year be treated as having been unable or unwilling to take some or all of the leave to which they were entitled under reg 13 and/or reg 13A WTR because of the Respondent’s refusal to remunerate them in respect of such leave?[20]If so, was any Claimant entitled to carry such leave forward to subsequent leave years until the termination of his employment and to receive a payment in lieu of such leave on the termination of his employment?[21]Was any Claimant’s employment terminated during his leave year meaning he is entitled to receive a payment in lieu in respect of accrued but untaken leave on the termination of his employment under regulation 14 of the WTR?[22]For the purposes of reg 14 WTR, when did/does a Claimant’s employment ‘terminate’? In particular:a. Did/does a Claimant’s employment terminate for the purposes of reg 14 WTR when any overarching contract between the Claimant and the Respondent comes to an end?b. Did/does a Claimant’s employment terminate (meaning that a right to payment in lieu crystallises) at the expiry of any relevant contract on which the Claimants were engaged, such that time for claiming runs from the ending of each individual contract?[27]Should the Tribunal:a. Under section 24(2) ERA award a sum to compensate drivers for any financial loss sustained by them which is attributable to any delay in repaying the deductions?b. Under reg 30(4) WTR, in assessing just and equitable compensation, award a sum representing interest, or otherwise as additional compensation, for any financial loss sustained by the drivers, which is attributable to any failure to allow the drivers to exercise their statutory right to paid annual leave and any delay in compensating them?c. Under reg 30(5) WTR, order that any sum which it finds to be due to the drivers should also include a sum of money representing interest and/or reflecting any delay in making payments due under regs 14 or 16? National Minimum Wage For the avoidance of doubt, issues relating to National Minimum Wage do not apply to the Claimants represented by London South Law Chambers.[28]Is each Claimant’s work unmeasured work or output work relevant in respect of the claims being brought?[30]What was the remuneration of each Claimant in respect of each pay reference period? In particular:a. What remuneration was paid by the employer as respects the pay reference period?b. Were there any reductions within the meaning of regulations 11 to 15 of the NMWR 2015/regulations 31 to 31 of the NMWR 1999?c. In considering the remuneration of each Claimant the Tribunal may need to consider (inter alia): i. Whether the Respondent made any deductions or took any payments from Claimants (including any points accumulated that could be exchanged for vehicle hire credits) constituting expenditure in connection with employment and so a reduction within the meaning of reg 13 NMWR 2015 / 32 NMWR 1999? ii. Did any payments made by the Claimants to third parties constitute expenditure in connection with employment and so a reduction within the meaning of reg 13 NMWR 2015 / 32 NMWR 1999?[32]If so, is the additional remuneration to which the relevant driver is entitled under section 17 of the National Minimum Wage Act 1998 the amount described in section 17(2) or the amount described in section 17(4) of that Act?[33]In so far as any Claimant is entitled to additional remuneration under section 17 in respect of any pay reference period, has the Respondent made a series of unlawful deductions from his wages contrary to sections 13 and 23 ERA, comprising the difference between his actual pay and the additional remuneration to which he is entitled (having in particular reference to the criteria indicated in Chief Constable of the Police Service of Northern Ireland v Agnew [2023] UKSC 33 especially at paragraph 127)?[35]Are such claims subject to a two year ‘back stop’ pursuant to section 23(4A) ERA so that the claimants cannot claim amounts which should have been paid to them before the period of two years ending with the date of presentation of the claim? In particular are the Deduction from Wages (Limitation) Regulations 2014, SI 2014/3322 under which section 23(4A) was introduced ultra vires section 2(2) of the European Communities Act 1972?” The evidence before me 13 On 19 July 2023, EJ Tynan ordered (in order 13 on page 147) the parties to exchange witness statements for the purposes of the hearing which, in the event, I conducted. The order was for the exchange of such statements at 4pm on 2 September 2023, but that was clearly intended to be a reference to 2 September 2024. In fact, the parties agreed to a variation of that date and their witness statements were exchanged only in October 2024. The majority of the witness statements then exchanged (all but one of the claimants’) were dated 14 October 2024. That of Mr Gallagher (that is to say, his first witness statement) was dated 15 October 2024. The first witness statements of Mr Kevin Valentine and Mr William Kelly were dated 21 October 2024 (although, for convenience, I record here that Mr Valentine told me when giving evidence that he had approved the contents of his first witness statement on 18 October 2024). 14 In addition to having before me the documentary evidence in the hearing and supplementary bundles to which I refer in paragraphs 10 and 11 above, I heard oral evidence from the following witnesses, on the following days.14.1 I heard first from Mr Patrick Gallagher, who was and had been since March 2020 (as he said in paragraph 1 of his first witness statement) “a statutory director and the Chief Operating Officer” of the respondent. I did so with Mr Gallagher giving evidence via CVP (from Ireland) on Thursday 31 October 2024 (with the hearing otherwise being in person) in connection with the application of the claimants for the striking-out of part of the respondent’s amended grounds of resistance (“the strike-out application”). Mr Gallagher was then cross-examined primarily about the evidence which he gave in what was his second witness statement, which had been sent to the claimants on Sunday 27 October 2024, along with a second witness statement of Mr Kelly, a second witness statement of Mr Valentine, and the first supplemental bundle of documents.14.2 I then heard from Mr Kelly, who was and had been since 1 October 2020 the respondent’s Operations Director, also in connection with the strikeout application. Mr Kelly gave evidence in person for the rest of the morning of Thursday 31 October 2024. He was cross-examined also primarily on the evidence in his second witness statement.14.3 On Monday 4 November 2024, I heard from Mr Khalid Edah-Tally, who gave evidence as a test claimant for the role of standard passenger driver. He had also been an executive driver, and he gave evidence about what had happened when he had been in both roles.14.4 In the afternoon of 4 November 2024, I heard from Mr Hector Ruiz, who gave evidence as a test claimant for the role of owner (or “partner”) driver, but who had previously been a standard passenger driver. Like Mr EdahTally, Mr Ruiz gave evidence about what had happened when he was in both roles.14.5 I heard from Mr Martin Payne in the morning of Tuesday 5 November 2024. He gave evidence as a test claimant in the role of standard passenger driver.14.6 I heard from Mr Thambipillai Mahendran in the afternoon of 5 November 2024. He was a test claimant in the role of executive driver, but he had before he became such a driver been a standard passenger driver and he gave evidence about what had occurred so far as relevant in both roles.14.7 Mr Dennis Nardelli gave evidence in the morning of Wednesday 6 November 2024. He was a test claimant in the role of standard passenger driver.14.8 Mr André Da Silva then gave evidence until lunchtime on that day and, after the evidence of the witness to whom I refer immediately below was interposed, in the afternoon of that day. Mr Da Silva was a test claimant in the role of standard passenger driver.14.9 At 2pm on 6 November 2024, the evidence of Mr Arnold Ban was interposed. Mr Ban gave evidence via CVP from Hungary. He was a test claimant in the role of standard passenger driver.14.10 Mr Rafal Klepacki gave evidence in the morning of Thursday 7 November 2024. He worked for the respondent as a standard passenger driver, then as an executive driver, and finally as an owner driver. He was a test claimant in the role of owner driver but he gave evidence about what had happened when he was in all three roles.14.11 Mr Kelly then gave further evidence. He did so during the afternoon of Thursday 7 November 2024 and for the whole of Friday 8 November 2024. Mr Kelly managed what the respondent called its Car Operations team. That team was called the respondent’s Car Control team until at least 2016, after which it was renamed the Car Operations team.14.12 Mr Adam Balog, a test claimant in the role of courier driver, gave evidence in that regard during the morning of Monday 11 November 2024. Mr Balog had also been a standard passenger driver and an executive driver, and he gave evidence in relation to those roles also.14.13 Mr Gavin White then gave evidence on behalf of the respondent from 12.20 onwards until just after 4pm on Monday 11 November 2024. He was employed by the respondent as its Driver Support Manager from 2018 onwards, and before then he was, from 2009, employed by the respondent as a member of what it at that time called its Driver Liaison team. The name of that team had, by 2018, been changed to Driver Support.14.14 Mr Costas Gavriel gave evidence from 4.05pm on Monday 11 November 2024 until the end of the hearing day (which ended relatively late, at 4.51pm). Mr Gallagher’s evidence was then interposed as I say in the following subparagraph below. Mr Gavriel then resumed giving evidence at 3.40pm on Tuesday 12 November 2024 and his evidence was concluded at the end of the hearing day (which was also, co-incidentally, 4.51pm). Mr Gavriel was employed by the respondent from February 2017 onwards as its Driver Recruitment Manager.14.15 Mr Gallagher’s further evidence was interposed from 10am onwards on Tuesday 12 November 2024. He gave evidence in person until 3.20pm on that day.14.16 I then heard from Mr Valentine, who gave evidence on behalf of the respondent from just after 10am on Wednesday 13 November 2024 until 12:26 on that day. Mr Valentine gave evidence as the respondent’s Head of Delivery Services. In that role he was responsible for the respondent’s courier business. He had by that time been in that role for about 15 years. 15 The respondent then applied to adduce evidence from Mr Sanjeev Gherra in connection with (and only with) the records of the respondent relating to the time taken by the respondent’s Driver Support and Car Operations teams to answer telephone calls. The claimants objected to that evidence being given. A witness statement in Mr Gherra’s name had been put before the claimants on Monday of that week, 11 November 2024, and the documents about which Mr Gherra was intended to give evidence were (Mr Leiper told me, without contradiction by Mr Segal) sent to the claimants more than a week before 13 November 2024. In fact, those documents were in the second supplemental bundle, which was put before me at the time that the respondent made its application to call Mr Gherra as a further witness, i.e. just before the lunchtime adjournment on 13 November 2024. In addition, the respondent had sent to the claimants the night before, i.e. after the end of the hearing day on 12 November 2024, several further documents which were intended to be the subject of evidence (which was not in a witness statement) to be given by Mr Gherra about the length of time it took for telephone calls to the respondent’s Driver Support and Car Operations teams to be answered. 16 Mr Segal objected to the new evidence being admitted in part because it was incomplete and in part because it was not primary evidence since Mr Gherra was not giving evidence about how long he had observed it to take for relevant telephone calls to be answered. In addition, Mr Segal objected to me admitting Mr Gherra’s evidence because it had (necessarily) not been put to the claimants in cross-examination and because if the documents about which Mr Gherra gave evidence had been disclosed, say, six weeks previously, then the claimants could have sought further disclosure and further information, but that was now precluded. 17 In the circumstances that17.1 the claimants had been challenged on their evidence about the time which it took for their telephone calls to be answered by the respondent,17.2 the claimants would have an opportunity to test the evidence of Mr Gherra in cross-examination having seen the documents which were the subject of that evidence, and could then (if appropriate) assert that those documents were of either no material, or at least little, probative value, and17.3 the evidence of Mr Gherra was at least relevant, I concluded that the evidence should be admitted. 18 At the request of the parties, I adjourned the hearing for a short (15-minute) break rather than lunch (it was 1.05pm), and I then heard oral evidence from Mr Gherra. The new documents which were not in the second supplemental bundle were helpfully put into a new bundle by Mr Leiper and emailed to me before Mr Gherra gave evidence. That bundle was the third supplemental bundle. While the evidence of Mr Gherra was relevant to the situation of all of the claimants, since they all said that it sometimes took a number of minutes to get through to the respondent’s operational staff, and that sometimes they were unable to do so at all, I refer to Mr Gherra’s evidence below for the sake of convenience only where I refer specifically to the issue of how long it took to contact the operational staff. That is in paragraphs 118-120 below, where I refer to the evidence of Mr Balog in that regard. The procedural background to these claims 19 The respondent’s witnesses’ evidence was focussed on the period from 2017 onwards. That was (it was clear from the evidence of the respondent’s witnesses) because the respondent had been the subject of an adverse judgment in case numbers 2208029-31/2016 after a hearing at Central London Employment Tribunal conducted by Employment Judge Pearl sitting with non-legal members Mrs Ihnatowicz and Ms Plummer. The tribunal’s judgment was dated 25 September 2017 and was at pages 123-140 of the respondent’s first bundle of authorities. The lead claimant’s name was Mr Lange, and I refer to it below as “the Lange judgment”. That judgment was appealed to the Employment Appeal Tribunal (“EAT”) and, on 14 November 2018, in a judgment which was sealed on 21 November 2018, the EAT dismissed the appeal. 20 The respondent appealed that judgment to the Court of Appeal, and that appeal was determined by Lord Justice Bean (“Bean LJ”) in the judgment at pages 212- 219 of the same authorities bundle. That judgment was dated 22 April 2021. Bean LJ had originally given permission to the respondent to appeal against the decision of the EAT, but, as he described in paragraph 3 on page 213 of that bundle of authorities (to which I refer from now on as “AB1”), he set aside that grant of permission in the light of the judgment of the Supreme Court in Uber v Aslam [2021] ICR 657 (“Uber”) and directed that there be instead an oral hearing before him of the respondent’s application for permission to appeal. Judgments refusing permission to appeal to the Court of Appeal are capable of being cited only if the court (usually consisting of a single Lord Justice of Appeal) gives permission to cite them, and Bean LJ, in paragraph 20 of his judgment (at page 219 of AB1) gave permission to cite his judgment because, as he said, “the case is of some general significance”. 21 In paragraph 14 of his judgment, at page 217 of AB1, Bean LJ said this: “[T]he ET’s finding at paragraph 44 is plainly correct. Each time one of the Claimants logged on there was a contractual agreement in force between him and Addison Lee.” 22 The respondent’s shares were owned in 2017 by a company by (and I take the name from paragraph 3 of the witness statement of Gavriel at WSB page 199, i.e. page 199 of the witness statement bundle) the Carlyle Group. That company sold its shares in March 2020 and Mr Gallagher (who had previously worked for the respondent when it was owned by the original shareholders), as he put it in paragraph 4 of his first witness statement at WSB page 321, “re-joined Addison Lee as part of a new management team alongside Paul Suter, Chief Financial Officer, and Liam Griffin, Chief Executive Officer.” 23 The preliminary hearings conducted by EJ Tynan included one at which it was contended by the claimants that parts of the respondent’s amended grounds of resistance should be struck out on the basis that those parts were an abuse of process in so far as they related to the period from 1 July 2014 to 24 May 2016 inclusive. That was on the basis that those parts were inconsistent with the Lange judgment. EJ Tynan’s rejection of that contention and his reasons for it were at pages 125-144. That document was sent to the parties on 17 May 2023. There were two passages in those reasons which showed why the application was made and which I found of particular importance here. 24 The first of those two passages was written in connection with the claimants’ application for deposit orders, which was made in the alternative to the application to strike out. The passage was at pages 137-138 and is as follows. “(32) ... The Pearl Tribunal’s findings are relatively contained, comprising 22 numbered paragraphs extending over little more than five pages. The Pearl Tribunal observed, “There are few factual disputes between the parties”. (33) Although paragraph 35 of Mr Jeans’ and Ms Belgrove’s skeleton argument [for the respondent] indicates the potential for factual disputes in respect of the Respondent’s business model and its drivers’ working arrangements since 2017 (the fact that these details have emerged somewhat late in the day by way of submissions rather than in the pleadings means that the Claimants have not formally responded to them and indicated the scope of any dispute), it has not been suggested or asserted that this time around there would be significant factual disputes in relation to what I have termed the Lange Arrangements. (34) At paragraph six of its judgment, the Pearl Tribunal noted the dates on which the three Lange Claimants had started driving for the Claimant; otherwise the Tribunal’s findings were entirely focused upon the business model and working arrangements as these operated in relation to approximately 4,000 drivers (with the potential exception of just one driver referred to in the judgment who drove his own vehicle). Having re-read the judgment, it is clear that when the Pearl Tribunal referred to “the drivers” they were referring to the approximately 4,000 individuals who drove for the Respondent rather than to just the three Lange Claimants. Even at paragraphs 22 and 25 – 27 of the judgment, where the Pearl Tribunal addressed the Lange Claimants’ evidence, it is clear that their evidence was relevant to drivers’ experiences in general, including how the business model and working arrangements operated in practice, and was not limited to their individual situations. With the exception of their commencement dates, I cannot identify any findings within the Pearl Tribunal judgment that were specific to any one or more of the Lange Claimants and which evidence working patterns or practices that were particular to them. For example, when Mr Olszeski gave evidence about using his access to the Driver’s Portal through XDA to identify geographical areas where drivers were in demand, this was illustrative of a point of wider general application regarding driver availability for hire. Similarly, Mr Morahan’s evidence regarding his holiday arrangements was illustrative of a point regarding the levying of a service charge against drivers who failed to perform a minimum level of work.” 25 The second passage in EJ Tynan’s reasons of 17 May 2023 which I found to be of particular importance in the context of the evidence before me was at pages 130-131, namely this. “(15) Mr Olszeski, began driving for the Respondent in June 2014, Mr Lange in August 2014 and Mr Morahan in June 2015. Although I cannot see any specific reference to this in the Pearl Tribunal’s judgment, it is common ground in these proceedings that the Lange Claimants ceased working for the Respondent on the same date, namely 24 May 2016. Strictly, therefore, the Pearl Tribunal was only concerned with the Respondent’s business model and the Lange Claimants’ working arrangements over a period of just under two years. Whilst, as I observed in the course of the Hearing, it must be unlikely that the Lange Arrangements were first implemented on the same date that Mr Olszeski commenced working for the Respondent and equally unlikely that they effectively ceased to apply on 24 May 2016, there is no further information available to me in this regard. Mr Jeans endeavoured to provide a summary overview of how the Respondent’s business model had evolved over a period of two decades or more, including the introduction in or around 2010 of technology in the form of a hand held XDA device that was provided to all Addison Lee drivers and through which they were able to log into the Respondent’s operating systems and secure work. However, notwithstanding his best efforts to shed some light on the matter, the reality is that I do not know whether the Lange Arrangements were in place and unchanged from 2010 or were only implemented at some later date. (16) Mr Jeans’ and Ms Belgrove’s skeleton argument focuses upon various changes to the Lange Arrangements which it is said were begun to be implemented almost immediately following the 5 – 11 July 2017 hearing before the Pearl Tribunal. I refer in this regard to paragraph 35 of their skeleton argument and their contention at paragraph 36 that, “… the whole substratum of practices which formed the basis of the Lange cases, has substantially disappeared”. The princip[al] changes that are said to support that submission have emerged in a somewhat unsatisfactory manner in these proceedings; the claimed changes are not referred to in a signed witness statement made by a director or relevant employee of the Respondent, nor are they reflected in the Respondent’s AGoR [i.e. the respondent’s amended grounds of resistance], which were amended as recently as 6 October 2022 (pages 198 – 208). Whilst the Respondent avers at paragraph 13(2)(ii) of the AGoR that there have been material changes in terms and working practices since the dates to which the Pearl Tribunal judgment relates, the only two specific changes pleaded are: the abolition of any possible adverse consequences for drivers who decline journeys, a change said to have been in place since March 2020 (but further qualified in Mr Jeans’ and Ms Belgrove’s skeleton argument and submissions as being the very latest date by which such consequences were abolished); and the introduction of permitted sub-contracting in June 2021. On the basis of the Respondent’s pleaded case, the Leigh Day Claimants contend that any strike out, alternatively any deposit orders, should extend to what they say is the Respondent’s attempt to re-litigate the issues determined by the Pearl Tribunal in respect of the period up to March 2020.” 26 It was the respondent’s position that I needed to make new findings of fact about the period with which the Lange judgment was concerned, namely 1 July 2014 to 24 May 2016. Some observations on the evidence before me 27 It struck me when reading the respondent’s witness statements (of which there were nearly 300 pages; the claimants’ witness statements took up nearly 200 pages of the 517-page witness statement bundle) that there was very little by way of precise evidence about the way in which the respondent operated its business before 2017. Rather, the evidence of for example Mr Gavriel (at pagse 198-250 of the WSB) was of a rather general sort and amounted in part at least to a purported explanation of the impact of documents in the creation of which Mr Gavriel was not involved. In addition, the “explanations” given by him, like those of other witnesses for the respondent, were at times inconsistent with the content of the documents. Having said that, in paragraphs 141 and 177 below, I refer to situations where I concluded that the documentary evidence before me was indeed inconsistent with what I found to be the actual (or factual) situation, so, I concluded, there was a real possibility that the respondent was simply too busy to update its documents to ensure that they always reflected what was happening as a matter of fact. 28 However, because of its significance for the credibility of the respondent’s evidence, I now record that the reason for the strike-out application (i.e., for the avoidance of doubt, the one which was made to me in the first week of the hearing before me) was that the respondent had on Sunday 27 October 2024 informed the claimants that what was a potentially critical document in the bundle was not genuine. That document was at page 8725, and, including the email addresses of the purported sender and recipients (but not some standard text at the bottom, to which I refer further in the next paragraph below) was in these terms. ‘From: “Patrick Gallagher” <Patrick.Gallagher@addisonlee.com> Date: 08/07/2020 at 08:42:29 To: “Victoria Hensley” <Victoria.Hensley@AddisonLee.com>,”Liam Griffin”<Liam@addisonlee.com>,”Glen Davis” <Glen.Davis@addisonlee.com>,”Bill Kelly” <BK@addisonlee.com>,”Leo Wright” <LeoW@addisonlee.com>,”Tony Smith2” <TonyS@addisonlee.com>,”Andrew Gard” <andrewg@addisonlee.com> Subject: RE: Driver sheets Bill and Kevin, a quick follow up to our meeting with “Solicitors” It is imperative that going forward the Operations team do not apply any bans or suspensions to and couriers or passenger car drivers on the Shamrock system Thanks Pat’ 29 That was followed by some standard text starting “This message is intended solely for the addressee and may contain confidential information. If you have received this message in error, please send it back to us, and immediately and permanently delete it.” The existence of that text in that email implied that the text which I have set out in the preceding paragraph above was originally sent as part of a standalone message. 30 However, a careful reading of the email would have revealed that it was stated to be written not only to “Bill”, i.e. “Bill Kelly”, but also to a person called “Kevin” whose name did not appear in the list of addressees. That might well have alerted the claimants to the fact that the email was suspect, but the respondent, to its credit, in the three witness statements which were sent to the claimants on 27 October 2024 in the manner to which I refer in paragraph 14.1 above (i.e. the second witness statements of Mr Kelly, Mr Gallagher and Mr Valentine), informed the claimants that the email was not genuine in the sense that it had not in fact been sent by Mr Gallagher, and that it was part of a sequence of emails of which there was a full copy at pages 12-18 of SB1, i.e. the first supplemental bundle, in the course of which Mr Kelly had on 12 September 2024 created it. 31 The email as it was first created was in the following form (which was sent as a purportedly forwarded email by the email at the top of page 12 of SB1, which was from Mr Kelly to Mr Gallagher and was timed at 14:09 on 12 September 2024; the sequence which I now set out was in the bottom half of that page and the first part of page 13 of that bundle). ‘From: Bill Kelly <bk@addisonlee.com> Sent: 12 September 2024 14:09 To: Bill Kelly Subject: FW: RE: Driver sheets This is an external email. Please treat links and attachments with caution. From: “Patrick Gallagher” <Patrick.Gallagher@addisonlee.com> Date: 08/07/2020 at 08:42:29 To: “Victoria Hensley” <Victoria.Hensley@AddisonLee.com>,”Liam Griffin” <Liam@addisonlee.com>,”Glen Davis” <Glen.Davis@addisonlee.com>,”Bill Kelly” <BK@addisonlee.com>,”Leo Wright” <LeoW@addisonlee.com>,”Tony Smith2” <TonyS@addisonlee.com>,”Andrew Gard” <andrewg@addisonlee.com> Subject: RE: Driver sheets Bill and Kevin, a quick follow up to our meeting with “Solicitors” It is imperative that going forward the Operations team do not apply any bans or suspensions to and couriers or passenger car drivers on the Shamrock system Thanks Pat From: Victoria Hensley <Victoria.Hensley@AddisonLee.com> Sent: 07 July 2020 11:46 To: Patrick Gallagher <Patrick.Gallagher@addisonlee.com>; Liam Griffin <Liam@addisonlee.com>; Glen Davis <Glen.Davis@addisonlee.com>; Bill Kelly <BK@addisonlee.com>; Leo Wright <LeoW@addisonlee.com>; Tony Smith2 <TonyS@addisonlee.com>; Andrew Gard <andrewg@addisonlee.com> Subject: FW: Driver sheets Here you go Vicky Hensley Commercial Finance Manager – Operations’ 32 Mr Kelly’s second witness statement contained the following material paragraphs. “4. I don’t recall the exact times or dates, but I remember Patrick [i.e. Mr Gallagher] informing me on several occasions that he remembered sending an email after joining Addison Lee in March 2020 instructing the operations team to not apply any bans or sanctions to drivers and couriers. I remember he said that he could not locate this email. 5. Patrick asked me to search my email inbox to see if I could locate a copy of the email. He asked me to do this on a number of occasions, although I cannot recall the time and date of all of the occasions. I do, however, remember Patrick repeating this request on or around 11 September 2024. 6. Following this, I did carry out a search of my email inbox. I wouldn’t call it an extensive search. I could not locate a copy of the email Patrick was referring to. 7. I think on the following day, on 12 September 2024, I spoke to Patrick again and he asked me if I had found a copy of the email. I informed him that I had not. 8. I don’t recall exactly why, but something triggered me to go to IT to ask them to help me to locate the email Patrick was referring to. I suspect I felt deep down that I had not done an extensive search and I felt bad about that. Around this time, I remember feeling under a lot of pressure to locate the email Patrick was asking me to find. I felt like a failure because I couldn’t find it. 9. At Addison Lee we have a retention policy which means that historical emails are archived on a Mimecast server. The emails in my outlook inbox did not go back to 2020, so these emails would have been archived on our Mimecast server.” 33 That passage showed beyond any doubt that there was absolutely no point in Mr Gallagher asking Mr Kelly to look for an email from 2020 in his email software without help from the respondent’s information technology (“IT”) team. If Mr Gallagher knew that then he must have known that asking Mr Kelly to find such an email in that way was a nonsense and could not have been done fairly, or, possibly, honestly. The fact that Mr Kelly remembered “feeling under a lot of pressure to locate the email [Mr Gallagher] was asking me to find” was in my view significant. That was because all that Mr Kelly could do was to ask the IT team to help him look for the email in the Mimecast server, and if that team could not find the email then that was the end of the matter, as Mr Gallagher, if he had been giving the matter any kind of thought, must have realised. 34 In fact, Mr Gallagher may well have been somewhat distracted at the time as he was about to undergo surgery for cancer. 35 Nevetheless, as I observed on 31 October 2024, if anyone was likely to have the email in his or her possession, it was Mr Gallagher, since he had sent it. However, the email could in the circumstances have been in Mr Gallagher’s possession only if it had been saved as a document (and therefore kept otherwise than in the respondent’s email software). In thinking the matter through when writing these reasons, it occurred to me that Mr Kelly might have saved the email as a document too. But if neither of them could find the email on their computers or the respondent’s network (ignoring the unlikely possibility that it would have been printed out), then the only place the email could have been was in the Mimecast server, and that was accessible, it was clear to me, only with the assistance of the respondent’s IT team.[36]In any event, Mr Kelly then did do the only further thing he could properly do at that time, which was to ask the IT team to look in the Mimecast server for the email which Mr Gallagher wanted Mr Kelly to find. What happened subsequently was improper, to say the least. The following passage of Mr Kelly’s second witness statement (the rest of it) described those events. ‘10. Isaac Kakyomya in our IT department came over to my desk to help me search for the email. I recall Isaac asking me what I was searching for. I said an email from Patrick referring to “bans” or “sanctions” sometime after he joined Addison Lee.11. Isaac had a laptop with him and he used that to search my emails. I think he was searching my archived emails on the Mimecast server. Isaac located various emails from Patrick going back to when Patrick joined Addison Lee in March 2020, but he was unable to locate the email I understood Patrick was after.12. I do not recall the precise circumstances, I think Patrick was standing in the room typing into his phone at some point. I don’t recall exactly when, but I recall saying something to Patrick along the lines that it was easy to change an old email. I was suggesting that it would be possible to recreate the e-mail he was looking for, if he wanted to do that.13. I don’t recall the precise words he used, but I do clearly recall that he told me in no uncertain terms not to amend an old e-mail.14. One of the emails located by Isaac was an email from Patrick on 8 July 2020 sent to various recipients with the subject line “RE: Driver sheets” [SB/2].15. I amended the text of Patrick’s reply at the top of that chain to the following: “Bill and Kevin, a quick follow up to our meeting with “Solicitors” It is imperative that going forward the Operations team do not apply any bans or suspensions to and couriers or passenger car drivers on the Shamrock system Thanks Pat”16. I did not amend the time, date or recipients of the email, nor did I delete any of the emails lower down in the email chain. I then sent it to Patrick at 2.09pm. Patrick and I did not discuss the e-mail after I had sent it to him, to the best of my recollection. I don’t believe I even thought about it again until 24 October [2024] when I found out it had been included in the bundle.17. The original email was on Isaac’s laptop, because he had located it while searching my archived emails on his laptop. I don’t recall if Isaac was still in the room when I did this, but I forwarded the amended email from Isaac’s laptop to my email [SB/6]. I then immediately forwarded the email to Patrick [SB/12].18. I have tried my best but I have really struggled to recall the chronology of that day. I don’t recall if my conversation with Patrick was before or after I had sent him the e-mail, and I don’t recall whether Isaac was standing with me when I sent it to myself from his laptop, or how long I had been searching for the e-mail for on that day with Isaac. I have looked at my calendar for that day, and my e-mails, but this has not helped to jog my memory about the sequence of events.19. When I forwarded the email to Patrick, I didn’t include anything in the subject line or the contents of the email to indicate it was not genuine. I don’t know why I didn’t – I genuinely can’t recall.20. I would like to make it clear that I never intended the email to be used for the purposes of these proceedings, or any proceedings. I really don’t know why I sent it. I have found the days since I discovered that the e-mail was in the bundle and referred to in my statement to be incredibly difficult and I just can’t remember exactly what was going on on that day or why I sent the e-mail.21. I am fairly sure that I didn’t discuss the email with anyone until Thursday 24 October 2024, as I will explain below. It was only at this stage that I became aware that Patrick forwarded this email to our solicitors, and that it had found its way into the hearing bundle as the document at [HB/8725]. Discovery of the error22. During the afternoon on Thursday 24 October 2024, I was sent a copy of the document at [HB/8725].23. A short time later, I was called into a meeting with senior management, namely Patrick, Paul Suter and Liam Griffin. It is all a bit of a blur, but I think I’d already had a conversation with Patrick by this time and he’d told me the problem. When I was in the meeting with Patrick, Paul and Liam, they showed me the document at [HB/8725] and I immediately admitted that I had amended the e-mail. This was the first point that I became aware that the email was contained in the hearing bundle and specifically the document referred to in paragraph 217 of my first24. The fact that this email is the one I created means that the statement I make at paragraph 217 of my first witness statement is not correct. I believed it to be correct at the time I signed my statement because I had been repeatedly told by Patrick he had sent an email about sanctions shortly after he had joined Addison Lee. I did not realise that the email referred to at paragraph 217 was the email I had amended and forwarded to Patrick as outlined above.25. I fully accept that it was a mistake to amend the email and I’m mortified that this has happened. I apologise unreservedly for this. However, I never intended the email to be used for the purposes of these proceedings.26. I genuinely did not realise that the email had found its way into the hearing bundle as document [HB/8725]. As explained, I only became aware of this on 24 October 2024 and I immediately admitted that I had amended the e-mail.27. My first witness statement was prepared with the assistance of Baker McKenzie. I spent a huge amount of time speaking with solicitors and reviewing my statement. I did not, however, review all of the documents which are referred to in my statement including the document at [HB/8725]. Unfortunately, I reviewed very few.28. That was my mistake and I completely accept responsibility for not checking the documents. In my head I had thought that it was fine because I would have read them all by the time of the hearing. I know I should not have signed off on my statement without having read the documents in the bundle. If I had done so, I would have immediately spotted that the document at [HB/8725] was the version that I had amended and was not genuine, and would have said so.’[37]There was much reason to doubt that Mr Kelly had told the whole truth in that passage. The main reason was that there was no apparent reason for the creation by him of the email at page 8725 unless it was intended by him to be relied on by Mr Gallagher as evidence in support of the proposition that Mr Gallagher had in 2020 given an instruction by email of the sort which Mr Kelly created on 12 September 2024.[38]Another reason for doubting that Mr Kelly had told the whole truth in his second witness statement was that if Mr Gallagher had not colluded with him in regard to the creation of the email which was at page 8725 of the hearing bundle, then Mr Gallagher would have been highly likely to have come and congratulated him (Mr Kelly) for finding that email, or at least sent an email acknowledging its receipt and, probably, saying something positive about the fact that he (Mr Kelly) had found it. It was not said by Mr Gallagher either when giving oral evidence (see paragraph 40 below) or in his second witness statement (see paragraph 12 of that statement, which is in the extract set out immediately below) that he had said anything positive to Mr Kelly about the email after receiving it, and of course it was Mr Kelly’s evidence that Mr Gallagher did not say anything positive about the email. The latter’s second witness statement contained the following passage (which was the whole of the material part of that statement for present purposes). ‘5. On 11 September 2024, I was asked to produce a copy of the e-mail that I believed that I had sent in March or April 2020. I looked for it myself but, as I explained in my first statement, I was unable to find it. 6. On what I believe was 11 September 2024, I asked Bill [i.e. Mr Kelly] if he could find a copy (as I was sure he was one of the people that I would have sent the e-mail to, given his role). I do not recall precisely when we spoke but I believe we did so on a number of brief occasions on 11 and 12 September and that I asked him a number of times if he had managed to locate the e-mail yet. 7. I do recall that Bill made a comment to me to the effect of “do you know it is possible to amend old emails?”. I understood his suggestion to be that he could change the text of an old e-mail from the time to replicate the e-mail that I believed I had sent. 8. I do not recall his precise words but I do recall saying that we could obviously not do that and dismissing the suggestion immediately at that time. I do not recall if this was on 11 or 12 September or precisely where this conversation took place. My recollection is, however, that Bill’s comment was made in passing and that the suggestion was shut down by me immediately. 9. I did not think much of this comment at this time or give it any further thought. It was a foolish suggestion which was dismissed by me immediately. 10. I have checked my calendar for 12 September 2024 and I can see that I had a meeting at 1pm until 2pm, and a Board meeting at 2pm. The meeting at 1pm was my regular weekly driver recruitment and retention meeting and I was attending the Board meeting immediately afterwards and rushing to get to it on time. 11. Whilst I was in my Board meeting, at 2.09pm, Bill forwarded me the email which appears at [HB/8725]. This can be seen from the full version of that e-mail which has been disclosed with this witness statement and is at [SB/12]. I do not recall when I first saw this e-mail but I believe it would have been later in the day, after the board meeting. 12. I forwarded this e-mail to Baker & McKenzie LLP at 17.54. I did not have any suspicion or inkling that this e-mail might have been fabricated by Bill at the time. I did not look at it closely but I did read the text and my assumption was that Bill had located a genuine e-mail that I had sent. He was specifically looking for such an e-mail on my instruction. 13. Before forwarding the e-mail, I removed the e-mails in the chain below the e-mail dated 8 July 2020. I did this as the rest of email [sic] was not relevant and I was highlighting the relevant text showing the last part of the email section I also remember thinking this e-mail was not at all helpful, but I knew that as part of the specific disclosure response which was due on that day, I needed to send it. [Sic; i.e. there was no full stop after “the email section”.] 14. I was absent from work in late September and early October 2024 for a major surgical procedure and a period of post-operative recovery. On my return, I attended a witness statement interview with Baker McKenzie on Friday 11 October 2024. My statement was due to be exchanged on 14 (and subsequently 15) October 2024. 15. When I reviewed the email at that time, I did not recall why the e-mail was sent to the recipients who were listed as having received it in the “to” field. It is not unusual for me to send an email deleting irrelevant recipients on a chain and assumed that I must have intended to have done this and forgotten to, and that this was a mistake on my part. It was also mentioned that it was addressed to both Bill and Kevin Valentine, but Kevin was not copied, which is something I did not notice at the time. 16. I agreed that further investigations could take place to source a copy of the original e-mail (rather than the version that had been forwarded to me by Bill) from our servers directly. At this stage, I had no reason to believe the e-mail was not genuine and I believed that I had likely clicked “reply all” to the wrong e-mail, which explained the issue with the recipients (and the fact that the subject line indicated that the email had been sent as a “reply” to what looked to be an unrelated chain). 17. I became aware that the e-mail may not be genuine on Thursday 24 October in the mid to late afternoon when I was informed by my colleagues Paul Suter, Chief Financial Officer, and Liam Griffin, Chief Executive Officer. 18. I am aware that a copy of an e-mail sent by me at the same time, and to the same recipients (and with the same subject line) as the e-mail at [HB/8725] had been sent to Baker McKenzie on Wednesday 23 October 2024, and that following further investigation, our IT team had confirmed that they had been unable to locate an original version of the e-mail purportedly sent by me in July 2020. 19. At this point I spoke with Bill more or less immediately, and he told me that he had fabricated the email. This was the first point that I became aware that the e-mail was not genuine. We notified Baker McKenzie immediately and I have given this statement as soon as possible to correct the record.’[39]During cross-examination on 31 October 2024, Mr Gallagher said that the only time that he (himself) had sent a document to Baker and McKenzie was on 12 September 2024 when he sent the email at page 8725 of the hearing bundle. Normally, he said, the respondent’s in-house legal counsel liaised with the solicitors.[40]Mr Gallagher did not say even in cross-examination that he had spoken to Mr Kelly to say “well done” for finding the email which was at page 8725 of the hearing bundle, or alternatively that he had sent an email thanking him for finding that email.[41]In all of those circumstances, I concluded that Mr Kelly and Mr Gallagher had both acted improperly: Mr Gallagher by at the least putting unfair pressure on Mr Kelly to “find” an email of the sort that Mr Gallagher wanted to exist. I concluded that Mr Gallagher had then been glad to find that Mr Kelly had “found” such an email, and had been willing at least to be credulous about it. He had certainly with alacrity sent it to the respondent’s solicitors, shorn of the emails above and below it, when if he had been acting correctly (i.e. here I say nothing about his state of mind; only that the right thing to do would have been what I now say) then he would have forwarded the email sent to him by Mr Kelly, preferably as a saved Outlook file, or at worst as an email forwarded without the removal of the documents which followed it in the sequence which were in fact at pages 12-18 of SB1.[42]Despite those findings, I declined to strike out any part of the respondent’s response. That was in part because I concluded that I could do justice by taking into account those findings in any circumstance where the credibility of either Mr Kelly or Mr Gallagher was in issue.[43]The purpose of the second witness statement of Mr Valentine was stated by him in its first paragraph to be “to explain why the document at [HB/8725] is referred to in my first witness statement, exchanged on Monday 14 October 2024 and updated on 21 October 2024.” He referred to that document in paragraph 81 of his first witness statement (at WSB page 371), in the following sentence: “I remember an instruction via email from Patrick Gallagher, Chief Operating Officer, to cease to apply any bans or sanctions to couriers and passengers [HB/8725].”[44]Mr Valentine’s explanation for that erroneous evidence of his included that (as he said in paragraph 4 of his second witness statement) he had not “review[ed] any of the documents which [were] referred to in [his first] statement”. I could not accept that that was a good explanation. A witness who, by asserting in terms that a document in the hearing bundle is what it purports to be, seeks to satisfy the court or tribunal that that document is genuine, is giving very important evidence, especially given the importance (recognised most clearly and cogently in paragraphs 15-22 of the judgment of Leggatt J (as he then was) in Gestmin v Credit Suisse (UK) Limited [2013] EWHC 3560 (Comm) (“Gestmin”) of contemporaneous documents for evidential purposes. In fact, all that a witness can usually do is to give evidence about the manner in which he or she personally was involved in the retrieval or supply of the document for the purposes of the proceedings. That is precisely because of the fallibility of memory to which Leggatt J referred in Gestmin and the transience of memories of documents. Mr Valentine had no personal involvement in the supply or retrieval of the document at page 8725, and all that he could properly do was to say that he had been shown it in the bundle, and (if it was genuinely the case) that he remembered receiving it. Here, he did not even look at it before giving evidence that it was a document which he had received at the time it was sent. That rather dented his credibility, so that, as with the evidence of Mr Kelly and Mr Gallagher, I looked at Mr Valentine’s evidence with even more care than I would normally have done. My findings of fact about the manner in which the claimants were engaged by the respondent to provide driving services An overview of the manner of the engagement of the test claimants by the respondent to provide driving services and some further observations on the evidence before me[45]By the time of closing submissions, it was clear that the claimants were all engaged personally (ignoring the fact that Mr Edah-Tally had nominally provided services via a company whose shares were wholly owned by him; the respondent did not take a point in that regard) under whatever contractual standard terms the respondent used at the time of the commencement of the contract. I understood it to have been agreed also that that contract was the same, no matter whether or not the claimant was a standard passenger driver, an executive driver, a partner driver or a courier driver. Certainly, at least some of the contracts expressly provided for the driver to tick either a box to show that the driver was a “Courier Driver” or a box to show that the driver was a “Passenger Driver”. An example was the contract relating to the test claimant Mr Kidd (who was as a result of a health emergency not well enough to attend to give evidence) dated “2018/04/05” at pages 5439-5448. Mr Kidd was a test claimant for the role of courier driver.[46]The fact that partner drivers were employed under the same contract had the result that their position was treated by the parties as being different only because they (partner drivers) did not hire a vehicle from a company which was part of the group of companies of which the respondent was a part. All of the vehicle hire agreements to which my attention was drawn were with a company called “Eventech Limited”, and in what follows I assume for the sake of simplicity that any vehicle hire agreement between a claimant and a company in the group of which the respondent was part was entered into with Eventech Limited, to which I refer below as “Eventech”.[47]However, by the end of the hearing it was clear that the fact that a partner driver not hire a vehicle from Eventech had consequences for the driver’s capacity to benefit economically from the work that he (all of the test claimants were male, so for the sake of simplicity I refer below to drivers on the assumption that they were all male) did for the respondent. The financial advantages which the drivers who were not partner drivers were able to gain differed from time to time. However, those differences were not material to my conclusions on the issue of the status of the non-partner drivers, although the manner in which non-partner drivers were integrated into the operations of the respondent, including by being given incentives not to reject jobs which they might otherwise have wanted (because of the amount of money that the respondent would otherwise have paid for those jobs) was relevant. Again, though, the precise details of those incentives were in my judgment not decisive, so I did not need to make findings of fact about the manner in which the incentives changed from time to time. I refer further to those incentives in a general way in paragraph 53 below, and in paragraphs 172 and 173 below, I refer to specific examples in the documents disclosed by the respondent.[48]Partner drivers, as Mr Gavriel put it in paragraph 14 of his witness statement, at WSB page 201, “have only been a significant part of Addison Lee’s business since 2016”. He continued immediately after saying that: “and they are the most challenging to recruit, both because of the need for their vehicle to fit Addison Lee’s requirements, and because they will generally be working for multiple taxi companies at any one time. For example, they might be signed up to work for Addison Lee, Uber and Bolt, and will be signed into all three systems at any given time so will accept the best job offer that comes in across the three systems at any given time.”[49]That did not make sense to me: the fact that a partner driver might accept a job from another provider (via its “app”, i.e. through the software which worked via a “smart” mobile telephone) had no apparent connection with the difficulty or otherwise of recruiting a person to be a partner driver. In addition, it was agreed by the parties that there was no requirement for a partner driver to have any of the respondent’s branding or livery visible when carrying a passenger whose carriage the driver had agreed to undertake through the respondent’s app, and therefore with the respondent. That was in my view likely to make it easier to recruit a partner driver. So was the possibility of being able to choose the bestpaying job on offer where jobs were offered simultaneously by two or more providers (such as Uber, Bolt and the respondent) via their apps.[50]I did not see in the many pages of the bundles put before me a reference to a partner driver (i.e. a driver referred to for the purposes of this case as a “Partner Driver”) in any document earlier than 2 March 2017, which was a “Partner Driver Fact Sheet”, which was at pages 690-691 and which, although undated, was said in the hearing bundle index to have been issued on 2 March 2017. In addition, in paragraph 13(2)(ii)(n) of the respondent’s amended grounds of resistance, which was at page 173, this was said: “Hiring the vehicle is no longer universal or near-universal. There was only one owner driver at the time of the Lange and others hearing but ownership of the vehicle now represents a popular alternative to hire. There are over 400 owner drivers”. Although there was no evidence before me from the respondent even to the effect that there was a single partner driver in 2017, let alone 2016, I accepted that there was such a single partner driver in 2016, given what was said in paragraph 6 of the Lange judgment, at page 124 of AB1.[51]I suspected that if the Lange judgment had been in favour of the respondent rather than the claimants in that case, then the respondent would not have started to seek to provide driving and courier services to the public to any significant extent via partner drivers. That was in part because of the extent to which it benefited the respondent to be able to provide services to the public via vehicles which bore what was variously described as its branding, logo or livery. Having said that, executive drivers (and for this purpose I refer only to those drivers who provided driving services as what the respondent called executive drivers via vehicles which were hired from Eventech) provided such services via cars which did not bear the respondent’s livery. Nevertheless, it is at least possible that the buyers of the services of such drivers themselves preferred the cars not to bear such livery. I saw that it was said in paragraph 13(2)(ii)(b) of the respondent’s amended grounds of resistance, at page 171, that ‘It is no longer the case that all cars “have AL livery”. Those classed as “executive cars” and those of “partner drivers” do not. There were some nonliveried cars in 2016 but the number of non-liveried cars has expanded substantially’.[52]It was also possible that the owners of the group of which the respondent and Eventech were in each case a part benefited financially from the contracts which drivers entered into with Eventech for the hire of a vehicle. However, there was no evidence before me about that.[53]Whatever was the reason, the respondent sought to encourage (by way of the provision of a notional carrot) non-partner drivers to accept offers of driving work made by the respondent (for the most part via the respondent’s app, the operation of which I describe immediately below) by agreeing (in ways, or terms, which varied from time to time) to give the drivers financial credits which had the effect of reducing the cost to the driver in question of hiring his vehicle from Eventech. That might be characterised as an incentive or a bonus. I return to the manner in which such incentives or bonuses operated (at least in 2023) in paragraphs 172 and 173 below.[54]Given the existence of the document at page 8725 and the fact that its falsity was relied on as a justification for applying for a partial strike-out of the response of the respondent to the claims, it will be clear that the parties disputed hotly the extent to which the respondent sought to encourage drivers of all sorts to accept jobs that were offered to them by the respondent by the imposition of a sanction of one sort or another if the driver refused the job. Those sanctions were all of a financial nature only. I refer to much of the evidence and make my findings of fact on the imposition of such sanctions in paragraphs 77-114 below, but I record here that(1) it was obviously in the respondent’s interests for drivers to accept whatever jobs were offered to them, and(2) it was contrary to those interests for the drivers to refuse those jobs. That was acknowledged by Mr Gavriel in paragraph 141 of his witness statement, at WSB page 239, in the following way (which I accepted in so far as it referred to the respondent’s interests; I make no finding of fact here about the extent to which it was in practice, as Mr Gavriel said there, “easier and easier for drivers to turn down jobs that [were] offered to them”; I make such a finding in paragraphs 92-93 below, where I return to the following words). “As it has got easier and easier for drivers to turn down jobs that are offered to them, including being able to turn down jobs while they are on their way to them (having accepted the booking), there has been a knock on impact on our customer service. Although the change to stop drivers from being logged off for refusing jobs happened seven years ago [i.e. in 2017], the impact of this on Addison Lee’s business continues to be that they are now faced with an increased number of customers who are stranded. This happens because either no drivers accept the job that customer has asked for, or they have initially accepted the job but have then backed out of it before picking up the customer. This is a real challenge to Addison Lee’s business and particularly its reputation in the market.”[55]I saw that in paragraph 162(b)(iii) of his witness statement (on page 244 of the WSB), Mr Gavriel said this about the reference to a “Service Fee” in what he called (in the opening words of paragraph 162 on page 243 of the WSB) the “sample ‘old style’ (pre-July 2021) driver invoice” at page 4938: “Service Fee: Until in or around July 2016 passenger car drivers who did not earn sufficient points in a given week (as set out in the applicable fact sheets) were charged a fixed payment of £35 to Addison Lee ([HB/630; HB/636]. This applied to both standard passenger car drivers, executives and partner passenger drivers. Couriers continued to pay this (as set out in the applicable fact sheets) until September 2021, when the charge was reduced to £0 [HB/8735] until it was able to be removed from the payments system entirely”.[56]In fact, Mr Gavriel was not responsible for the payment of drivers, so he may have obtained his knowledge about the £35 fee from what was on pages 630 and 636 (although, I should say, I could find nothing about a £35 Service Fee on page 636, but I did find references to such a fee on pages 229, 630, 634, 1041, 1074, 1348, 1464, 1557, 2589, and 2631). In any event, the reference to “partner passenger drivers” in that passage was misleading given what I say in paragraph 50 above (namely that there was in 2016 no more than one such driver). However, there were three documents from 2016 stating that the service fee was no longer going to be charged: the first was at pages 2597-2601 which started with the words “Driver Pay is changing from the 21st May 2016”. The others referred to the change taking place on, respectively, 28 May 2016 (pages 2604- 2608), and 4 June 2016 (pages 2611-2615).[57]I add that there were in the bundle a number of references (such as at page 4172) to “Partner Drivers” which were inconsistent with the way in which that term was used by the time of the hearing before me. That page was part of a set of “FAQs”, which started at page 4167 with the heading “Driver Deal – May 2021; FAQ document prepared by firstlight” and which had just under that heading these words (the italics and underlining being in the original): “Suggested use: Potential to be used proactively with driver communications or as follow up to driver communications”[58]The document was stated in the index to have been dated 10 June 2021, which suggested that it was in fact used, however. In the bottom half of page 4172, and up to the end of the document, which was at the top of the next page, this was said. “Have these changes been made in light of Addison Lee’s loss of the right to appeal the result of the 2017 employment tribunal? We have been working on new driver arrangements since our return to the business in March 2020 and our focus has been on securing the future of our business and of the livelihoods of each and everyone of our Partner Drivers. There is nothing in the current contracts that we would not have implemented, even if we had won our right to appeal and the courts had honoured the contractual arrangements between us and our drivers. Ultimately, a tiny minority of our partner Drivers have been lured by lawyers chasing fees to bring claims against us which aim to negate our longstanding contractual arrangements with Drivers and, by being more explicit in our arrangements, we will both improve our offer to the Driver and secure the self-employed status of our Drivers which provides both the flexibility and earnings that they demand. What is Addison Lee’s view on the recent news that Uber is seeking declaratory relief from the High Court in relation to the Supreme Court’s recent ruling over its worker rights? In many regards, the Supreme Court’s ruling has turned English common law on its head. We have some empathy with Uber on some aspects, but it must be remembered that the Supreme Court decision applied only to a very narrow set of circumstances involving 3 drivers working under convoluted arrangements with multiple corporate entities operating across jurisdictions in order to circumvent tax and drive down pricing and driver pay levels to destroy a long-standing private hire operating model in which Addison Lee thrived by taking care of its drivers and customers. Addison Lee has always paid its tax and has continued to support its drivers and, unlike the ride hailing apps, has maintained its full Operators’ license with TFL for the past 40+ years.”[59]As I read it, that passage was referring to drivers who hired their vehicles from Eventech. The passage starting on the preceding page of the document and continuing to the end of the fourth paragraph on page 4172 (ending with “an approved removable branding solution must replace the standard Addison Lee branding solution”) was also misleading for reasons which will be apparent from what I say in paragraphs 158-163 below about the possibility of sub-contracting and the absence at any time (including up to the time of the hearing before me: see paragraphs 70-74 below) of a button on the respondent’s app entitling nonpartner drivers to reject a job. It was also silent about the impact in practical terms of accepting a job and then calling the respondent to cancel it (to which I refer in paragraph 75 below). The passage was also misleading in that it did not refer the impracticability (to which I return in paragraphs 61 and 62 below) of removing and replacing the respondent’s logo or, as it was said in the passage, the respondent’s “branding solution”. That passage was in these terms (the italics being in the original; for the avoidance of doubt, there were no italics in the passage which I have set out in the preceding paragraph above). “Right to refuse bookings How do I refuse a booking? You can always refuse a booking by contacting Driver Control or alternatively, when the facility is incorporated into the Driver App, by selecting “reject” within XX seconds of the booking allocation. Can someone else drive my vehicle? Your vehicle can only be driven by a qualified PCO licensed driver who has been approved by Addison Lee for regulatory compliance and insurance purposes. Can someone complete bookings on my behalf? You may sub-contract to a qualified PCO licensed driver, but they need to be approved and registered with Addison Lee. We will continue to pay you for the bookings completed and any arrangements with the sub-contractor are between you and the sub-contractor. The vehicle must remain comprehensively insured al all times. Can I drive my own vehicle? Yes, but it must meet all Addison Lee vehicle guidelines. Can I drive with other service providers? Any PCO licensed driver can drive with any service provider. Addison Lee consistently provides the highest quality bookings and the highest and most consistent earnings potential of any provider, so it is unlikely that you would benefit from driving for another provider. If you do choose to provide services to another service provider, however, you must ensure that you are properly insured at all times. If you are hiring an Eventech vehicle and contributing to the Insurance Trust, you will only be insured while driving for Addison Lee or for personal use. Also, it is not permitted that you display your Addison Lee logo, which is mandatory for doing Addison Lee bookings under our Driver Contract, so this must be removed and an approved removable branding solution must replace the standard Addison Lee branding solution.”[60]However, and on the other hand, some of the statements made in the claimants’ witness statements were of a general nature and were also, evidentially, speaking, imprecise. One example is paragraphs 65 and 66 of Mr Balog’s witness statement, at WSB page 13. However, in his next paragraph, Mr Balog gave an example (to which I return in paragraphs 116 and 117 below, where I have set out paragraphs 65-67 of Mr Balog’s witness statement), and in any event, it was difficult to see how Mr Balog could have been more precise than he was in paragraphs 65 and 66 of his witness statement. I nevertheless treated with caution the claimants’ assertions of a general nature, just as I did the respondent’s witnesses’ assertions of a general nature. The relationship between the respondent and drivers who hired their vehicles from Eventech[61]In paragraph 32.1 of their written closing submissions, the claimants said this about standard passenger drivers. “Standard passenger drivers drove a vehicle which they leased from Eventech and were required by clause 6 of the Vehicle Hire Agreement (VHA) to have AL [i.e. Addison Lee’s, i.e. the respondent’s] branding displayed on the vehicle at all times when it was being used for AL bookings - see, for example, clause 6 of the VHA issued to MP on 5 November 2023 at [6183]. This meant in practice, as GW [i.e. Mr White] acknowledged, that passenger drivers could not work for other operators because each time they did so they would have to go to West Drayton to get the branding put back on – an option which he agreed it was not conceivable that drivers would take.”[62]That was in my view an accurate description of what happened. It was accurate in relation to the whole of the period in issue before me, i.e. both before and after 2016.[63]In the next subparagraph of their closing submissions, the claimants said this. “Executive drivers leased a vehicle of an executive standard from Eventech under the terms of a VHA which did not include a requirement to display AL branding - see the VHA issued to TM [i.e. Mr Mahendra] on 25 November 2020 [6798-6828]. There was no material difference between the operating practices which applied to standard passenger drivers and those which applied to executive drivers, other than a stricter dress code, which included the requirement to wear a tie. Neither of the claimants who is a test claimant for executive drivers i.e. TM and KET [i.e. Mr Edah-Tally], nor any of the TCs [ie. test claimants] who have driven as executive drivers, has ever multiapped whilst working for Addison Lee.”[64]I make my findings of fact on the issue of the respondent’s expectations in relation to dress and the manner in which those expectations were followed up by the respondent in paragraphs 122-144 below. I accepted that neither Mr Mahendran nor Mr Edah-Tally “multi-apped” (i.e. took driving jobs from other operators such as Uber or Bolt) while working for the respondent, but the major issue here was whether they were in practice able to do so and, if they were, the legal impact of that. I return to the first of those questions in paragraphs 204-205[65]In paragraph 32.3 of their written closing submissions, the claimants said this about partner drivers. “Partner drivers provide their own vehicle rather than leasing one from Eventech, although the vehicle must be approved by AL - see HR/15 [i.e. paragraph 15 of Mr Ruiz’s witness statement]. The operating arrangements which apply to them are the same as those that apply to standard passenger and executive drivers. Both of the partner driver TCs, HR and RK [i.e. Mr Klepacki] used more than one app during the periods in which they worked as partner drivers, for the reasons explained at HR/43-48 and RK/53 and 88-94.” That too was an accurate summary.[67]In paragraph 32.4 of the claimants’ written closing submissions, they said this about the test claimants who were courier drivers. “The two claimants who are test claimants for courier driver, AB and SK, drove vans hired from Eventech. The operating practices which applied to courier drivers were different in some respects from those that applied to passenger drivers, as discussed below; but not materially for the purposes of the worker status and working time issues.”[68]As I understood the respondent’s position in regard to courier drivers, it was that their position differed because of their greater freedom in practice to reject jobs offered to them by the respondent. I make such additional findings of fact as appeared to me to be necessary about the way in which courier drivers worked for the respondent in paragraphs 174-191 below, after making the following additional findings of fact about all non-partner drivers. The manner in which non-partner drivers worked for the respondent The determination of the rate of pay for work and the software used by the respondent for allocating the work[69]In the period before the Lange judgment was given, so at all material times up to then, all but one (and I say that because of my finding of fact at the end of paragraph 50 above) of the persons who worked as drivers for the respondent hired vehicles from Eventech and were paid to do jobs for the respondent at rates which were determined by the respondent without any input from the drivers. The rates for jobs for all test claimants were determined in the same manner, i.e. at all material times and not just before the Lange judgment. Those jobs were offered via software which was at the start of the period operated (as an app) via a smart mobile telephone which(1) was configured to work solely as a vehicle for the software, and(2) was called in the first part of the period an XDA and then an MDA. That software at some point (it was not necessary to decide when) started to be capable also of being used on a driver’s own smart mobile telephone through being downloaded (as an app) and installed on that telephone. Except in the circumstances to which I refer in the next paragraph below, at no material time (that is to say both before and after the Lange judgment) did the software show the place to which the passenger was to be driven unless the driver had accepted the job. In what way could a non-partner driver reject a job[70]It was agreed by the respondent that at no material time did the respondent provide drivers who hired their vehicles from Eventech an option to reject a job via the app once it was allocated to them, and that the only option for those drivers to reject a job via the app which ever existed arose in the final part of the relevant period, when a driver had a passenger on board. Mr Edah-Tally described the latter process in paragraph 144 of his witness statement (at WSB page 120), in the following manner (and I accepted this): “The only situation where I understood that you could decline the offer of work was when you had a passenger on board. About 5 or 10 minutes before dropping off the passenger, a message would appear on the MDA asking if you wanted to carry on working, to which you could say ‘no’. If you said ‘yes’, you might be offered a choice of jobs one by one if it was busy, showing the pickup time and location and the destination. If you opted for ‘yes’, it meant that you had accepted the job and therefore you had to do it, and if you changed your mind you would have to call the controller ... I do not know when this system was introduced as it was not rolled out universally to all drivers at the same time and they had to update their app for it to take effect, but the process may have started in late 2021. Prior to this update on the MDA, the system operated in the same way, except that the job options did not show the destination so you would be accepting a job without knowing what type of job it was.”[71]In paragraph 146 of his witness statement, at WSB page 120, Mr Edah-Tally said this. “I recall that at some point in 2023, I saw a message on the portal informing drivers that if they wanted to decline work, all they needed to do was to call and the job would be taken off them without them having to provide a reason, but I do not think that this happened in practice at the time or since. In my capacity as IWGB [i.e. the Independent Workers’ of Union of Great Britain] organiser, I heard numerous complaints about drivers feeling unable to decline jobs for fear of negative consequences.”[72]Mr Edah-Tally’s evidence about what happened if he wanted to reject a job after it was allocated to him via the respondent’s app was in paragraphs 140-143 of his witness statement, at pages 119-120 of the WSB. All of the test claimants gave evidence to a similar effect. Mr Payne’s witness statement contained (in paragraphs 62-71 at pages 139-141 of the WSB) a description of what happened which among other things raised the question of how jobs were allocated to drivers: was it only, or at least almost always, done via the app, or was it also done to any material extent through the respondent’s Car Control team, latterly called the Car Operations, team?[73]As I understood the respondent’s evidence on this issue, the work was allocated almost always via the app, but in some limited circumstances the work could be allocated by a member of the respondent’s relevant head office staff, which (as I understood the respondent’s evidence) would at least usually be a member of the Car Operations team. During the course of his second cross-examination, Mr Kelly said something which was the subject of the following passage in the claimants’ written closing submissions (all underlining and bold fonts being in the original; the same is true for all other quotations below, unless otherwise stated). ‘Being reallocated the same job unless logged off[80]It was BK’s evidence, later confirmed by GW [i.e. Mr White] and KV [i.e. Mr Valentine], that until relatively recently (BK could not say when, other than it was after 2020) if a controller did permit a driver to reject a job which had already been allocated to him (and was therefore close to time of pick-up and he was the nearest driver to the pick-up point), that job would automatically be put into the ‘waiting to allocate’ part of the system and then (because that job was now a priority) would be instantly re-allocated to the nearest driver, who would in “all likelihood” be the same driver.[81]That being so, BK said, the ‘only way’ to prevent the driver being reallocated the same job he had just been permitted to reject was to ‘log off’ the driver from the system. Moreover, although it would in theory suffice to log him off for only a few minutes, AL’s policy (not a requirement of the software) was that the standard minimum log off period was one hour: as BK put it, ‘for consistency and fairness [sic]’– although ‘in hindsight’ it would have been better if a more flexible system had been adopted. “[Q: If he didn’t phone up would he have to take it if offered same job again? Mr Kelly, answer the question - if offered the job again what are his options?] Would have to ring in again. [Q: And to break that chain what would happen?] Would be logged off for an hour. … [Q: And you’re saying the only way to prevent it being allocated again is to log driver off?] Sorry, yes.”.[82]This evidence was not contained within the witness statements served by AL and is of the highest relevance. It means that, at least until fairly recently [Footnote 9: AL has provided no evidence and has given no disclosure in respect of what change was made to the software and when, such that the situation described by BK where the rejected job is likely to be automatically re-allocated to the same driver ended. In the circumstances, the ET is invited to find that no such change was made during the material period.], drivers were being routinely logged off for an hour when rejecting jobs (albeit not directly as a punishment, but as a crude ‘work-around’) – constituting a far more draconian response in practice than anything implemented by Uber or Bolt as recorded in the cases involving those operators.[83]This is consistent with the evidence, including over recent years, that even where a controller accepted that the driver was rejecting a job, they would suggest that the driver might as well stop working: see e.g. per ADS [i.e. Mr Da Silva] in x-ex, “… over whole time when I was working, just remember if call them and say don’t want to do job they say okay but you should log off and go home” – presumably the only other way to prevent the system from re-allocating the rejected job to the same driver. Sanctions as punishments[84]It is to be assumed that those ‘standard’ log-off events would not be the subject of entries in the ‘disablement’ tab or the driver log (not least because, if otherwise, those documents would be replete with such entries). What is now being addressed are those occasions where the sanction was applied not as a ‘work-around’, but as a punishment.’ 74 I accepted those submissions, which were in my view a wholly accurate description of what happened when Mr Kelly gave evidence and its implications. I did so not least because I came to the view initially that the jobs were allocated automatically by the respondent’s allocation or operating software (which the respondent called “Shamrock”) unless there was a need to allocate a job otherwise than automatically via that software, when it could be allocated by a member of the Car Control or the Car (or Driver) Operations team, but only via a manual intervention and with the effect that the job was now allocated (after the manual intervention) via the software (to which I refer from now on as “Shamrock”). That evidence of Mr Kelly about the way in which the software then, once a job was taken off a driver, automatically re-allocated the job to the driver if the driver was the nearest one to the pick-up point, confirmed my view that that was the way that the software operated. 75 That, however, led to the conclusion that the respondent had in fact, whether or not knowingly, used that software, by logging off a driver who had refused a job, in an economically punitive way. That was, as the claimants submitted, highly relevant evidence. 76 I add for the sake of completeness that my own note of what Mr Kelly said (and it was said in answer to a proposition that I put to him at the time, which was that a driver could have been logged off for less than an hour) was that logging off for an hour was done for the sake of “consistency” only, and not fairness, but that when I looked at that note, I recalled him saying that it was for the sake of “fairness and consistency”. Even though that evidence was given on 7 November 2024 and I returned to my note of the evidence just over 2 weeks later, so that I might have been mistaken in remembering the use of the word “fairness” as well as “consistency”, the respondent did not say in reply to the claimants’ closing submissions that the words “consistency and fairness” were not used by Mr Kelly in answer to my proposition. Did the respondent continue to impose economic sanctions for refusing jobs after the Lange judgment was given? Passenger drivers 77 Mr Kelly was best-placed to give direct evidence about the imposition of sanctions on passenger drivers who refused jobs. That was because he managed the Car Operations team. The latter was clear principally from what he said in paragraph 7 of his first witness statement, at page 399 of the WSB. What he said there incidentally helped to illustrate the extent to which the respondent sought to avoid the effect of the Lange judgment by painting a picture of the removal of the control of drivers which it evidently thought was a critical part of the Lange judgment. That paragraph was in these terms. “As Director of Operations, the role is similar to my previous role of Head of Car Control although it is a bit broader and includes responsibility for some of our commercial contracts. I am still responsible for the Car Operations team and I will explain what this team does below. The Car Operations team used to be called Car Control. I can’t remember exactly when the name changed, I think it was around 2017. It must have been before 2021 because an email from Patrick Gallagher in April 2021 refers to removing control terminology from our operating system Shamrock [HB/3599]. It was called Car Control for a number of years so even now I do sometimes mistakenly refer to it as Car Control.” 78 There was some evidence before me to the effect that the team which Mr White managed was involved in the imposition of economic sanctions, but there was nothing in Mr White’s 58-page witness statement (ignoring its appendix) about sanctions. (The team was originally called the “Driver Liaison” team and that team also had its named changed. The new name was “Driver Support”. In paragraph 6 of his witness statement, at page 252 of the WSB, Mr White said that he thought that the change may have been made in 2014, but the extract from the email of a member of his team, Ms Chalkley, which I have set out at the end of paragraph 105 below suggests that it was at the earliest in 2018.) That was odd given what Mr Kelly said in paragraph 207 of his first witness statement, which I have set out in paragraph 98 below. In any event, Mr Gavriel’s witness statement contained no direct evidence about the imposition of sanctions, but that was unsurprising since his main role concerned driver recruitment, and, as he said in paragraph 89 of his witness statement, on page 221 of the WSB, he was “not in the Driver Operations or Driver Support team.” 79 Mr Kelly reported to Mr Gallagher. As a result, Mr Kelly was the only witness for the respondent who gave evidence at the hearing before me who was in a position to give direct evidence about the Car (or Driver) Operations team’s dayto-day activities, or practices. However, he was not in a position to give direct evidence about the reasons for entries made by the members of that team in the respondent’s computer records relating to the test case claimants. That was because, it was clear from what he said in each relevant part of his witness statement, he did not know about the entries at the time that they were made and he therefore did not know (personally; i.e. he had no direct evidence to give about) why they were made. That did not stop him purporting to give extensive evidence in his first witness statement about the entries, but only a very small amount of what he said by way of such purported evidence could have been helpful to me in determining what the respondent did when a test claimant refused a job. However, even if what he said that could have been helpful had been accurate (and in a number of respects it was not; by way of example, what he said in paragraph 290(i) on page 479 of the WSB was plainly wrong given the words on page 6943 “driver sent hom[e]”, and what he said in paragraph 292(k) on page 484 of the WSB was misleading for the reasons given in paragraphs 95- 97 below), the facts that the respondent at all material times79.1 did not, except in limited circumstances (see paragraphs 69 and 70 above), tell the non-partner drivers the destination of the job and (I inferred) therefore the remuneration which the respondent was going to pay for doing it until the driver in question had accepted the job, and79.2 required all of the non-partner drivers to telephone the Driver Operations team to say that they were refusing a job, in my view meant that there was a considerable obstacle placed in the way of a driver refusing a job which was a real practical deterrent to such refusal. 80 That obstacle resulted from the facts that80.1 the Driver Operations team would always be likely to seek to persuade the driver to take the job after all if there was no other driver available to do it;80.2 there might well be difficulty in getting through on the telephone, for either, or both of the following reasons (all of which applied irrespective of the accuracy or probative value of Mr Gherra’s evidence):80.2.1 there might be a poor mobile telephone signal in the area where the driver was at the time of trying to call the Driver Operations team;80.2.2 that team might be busy, and would be likely to be so in busy periods, so that the time taken for a member of the team to answer the call would be likely to be longer in busy periods; and80.2.3 the driver would, in the time that he was waiting for his telephone call to be answered and then speaking to the Driver Operations team, not be available for any other job, because the Shamrock software would show him to be allocated to the job which he wanted to reject. 81 In addition, of course, the respondent’s routine practice, revealed by Mr Kelly for the first time as described in paragraphs 73-76 above, was to log a driver off the Shamrock software for an hour when the driver refused a job. That was of course a serious economic sanction for the driver. 82 In fact, while that was a sufficient answer to the respondent’s proposition that no sanctions were in practice imposed after 2017, it was only part of the picture. The rest of it consisted of the following factors. 83 The most important aspect of the rest of the picture was that in 2021, before the imposition of what I will call the new driver deal agreements (with the emails enclosing them and other documents relating to them) at pages 4353-4540, Mr Kelly specifically blocked a proposed communication to drivers in these words (which were nearly at the bottom of page 4274). ‘Right to refuse bookings How do I refuse a booking? You can always refuse a booking by contacting Driver Control [or, when the facility is incorporated into the Driver App, by selecting “reject” within XX seconds of the booking allocation.]’ 84 There were the following comments, in the following order, in the right hand margin of what was evidently a Word document:84.1 “Commented [ja43]: Does this mean that this isn’t in the app yet? If so, suggest we tweak to say something about how Addison Lee is also due to launch this function in the app to make cancellation easier than ever.”84.2 “Commented [AG44R43]: removing this line”84.3 “Commented [PS45]: Bill K to comment”84.4 “Commented [BK46R45]: Hi Gardy, we shouldn’t include this as it will draw attention to it”84.5 “Commented [AG47R45]: removed”.[85]Mr Kelly dealt with that removal of those words in paragraph 222 of his first witness statement, which was at WSB page 450 and was in the following terms. “As part of the roll out of the wholly re-worked driver deal in the summer of 2021, which I understand Patrick Gallagher will discuss in his evidence, as mentioned, my involvement was largely confined to commenting on documents. One such document I commented on in the lead up to the roll out was the FAQ document for drivers [HB/4271]. One of the draft versions of the FAQ document included a FAQ “How do I refuse a booking”. I removed that text [HB/4274]. If you look at the comments down the right hand side of the text I deleted, there is one comment referring to the launch of a new function in the app to make cancelation [sic] “easier than ever” [HB/4274]. I understood this to be a reference to the feature we rolled out to partner drivers in December 2023/January 2024 meaning partner drivers have to actively accept a job if it is one they want to accept (as I explained above). At the time I was worried that the FAQ document would be published before we managed to roll out this new feature, so I didn’t want to draw attention to it. That is why I deleted the comment. As the new feature was only rolled out in December 2023/January 2024 I was proved right.”[86]Even at first sight, that paragraph was at least difficult to believe, if only because the reference to the right to refuse to which the writer of the document at page 4741 onwards referred, plainly applied to all drivers, and “the new feature ... rolled out in December 2023/January 2024” was “rolled out to partner drivers” only.[87]In addition, the proposition that it might be misleading to refer to the ‘facility’ to ‘select “reject” within XX seconds of the booking allocation’ when it was ‘incorporated into the Driver App’, was obviously wrong: the originally proposed words, in full, were in themselves clearly conditional on the new facility being incorporated in the app, so there was no risk of drivers being misled if those words were included in the communication.[88]In my judgment, what Mr Kelly meant by saying that “we shouldn’t include this as it will draw attention to it” was that the respondent should not draw attention to the right to reject in any way because he at least thought that it should not be drawn to the attention of drivers. I concluded that he at least intended drivers not to become aware of the fact that there was a theoretical right to reject jobs.[89]In fact, of course, there was always such a right, as the drivers (including the test claimants) were never obliged to accept an offered job. The only issue was what, if any, was the consequence (over and above the fact that the driver would not do the work and would not get paid for it) for the driver of such a refusal. Nothing was said about such a consequence in the documentation, except in the standard terms, for example in clause 8.1 at page 4360, in the following words. “There will never be any sanction or punitive measure taken against any Driver for rejecting a Booking or for not offering or being available to provide Services at all, at any time, or for any minimum number of hours per day, week or month. In the event that you consider that you have been subjected to a sanction or a punitive measure for doing so, please contact Addison Lee and we will investigate and take appropriate remedial action.”[90]In cross-examination, Mr Kelly accepted that in none of the many communications from the respondent to drivers in the form of a newsletter or a similar document informing drivers of new developments and other relevant things, had there been a communication to the effect that they could (using the words recorded by me in my notes) “reject jobs without adverse consequences”. He was not taken in re-examination to any document which announced that, in whatever way. The bundle supplied to me in digital form was put together digitally in such a way that it was highly resistant to being made searchable. However, I persisted and eventually was able to make it, or at least most of it, searchable. There was no such document that I could see in the bundle (having searched for the words “adverse consequences”, separately, for “punitive measure”, and, separately, for “sanction”, i.e. any word of any length containing that string of letters). There was at page 4257 an email from Mr Paul Suter to Mr Liam Griffin, Mr Gallagher and Mr Kelly and a one-page enclosure. The email enclosed a document with the file name “PS Drivers Comms Note.docx”. It was not entirely clear whether the next page and the one after it were that document, or pages from that document, but on page 4258 there was what looked like a single-page document which was headed “Key Points New Contracts”, in which, in numbered paragraph 4, it was said that “To protect Addison Lee and our drivers there will be a further reinforcement of flexibility and industry leading terms: ... b. Drivers can reject jobs without sanction or penalty”. However, that document was not, it appears, in fact, used. On the following page, i.e. page 4259, there were at the top of the page the words “Communication HEADLINES (TBD)”, and in a red box “TO BE CHECKED”. At the bottom of the page, there were these words. “What not to say: 1) We are doing this to defend against worker status claims (should rather be positioned around maintaining flexibility and independence for the driver while ensuring Addison Lee remains the leading premium transport business in London)”.[91]It was true that after 2021, when drivers had to accept new agreements every 12 weeks if they wanted to continue to receive work from the respondent because the respondent at that time introduced a maximum term of any agreement with a driver of any sort of 12 weeks, the drivers could have seen the words which I have set out at the end of paragraph 89 above. However, I concluded that Mr Kelly guessed, correctly, that most drivers would not read those agreements, so that he was intending them not to find out about the new promise not to impose a “sanction or punitive measure” for refusing a job. I came to that conclusion in part because it was obviously at least sometimes highly problematic for the respondent if a driver refused a job.[92]The difficulty caused to the respondent by such a refusal was acknowledged by the respondent in several places in the witness statements before me which had been made on behalf of the respondent. A particularly clear acknowledgment of that difficulty was in paragraph 141 of the witness statement of Mr Gavriel, which was at WSB page 239, and which I have set out in paragraph 54 above. For convenience, I now repeat the words of paragraph 141 of Mr Gavriel’s witness “As it has got easier and easier for drivers to turn down jobs that are offered to them, including being able to turn down jobs while they are on their way to them (having accepted the booking), there has been a knock on impact on our customer service. Although the change to stop drivers from being logged off for refusing jobs happened seven years ago, the impact of this on Addison Lee’s business continues to be that they are now faced with an increased number of customers who are stranded. This happens because either no drivers accept the job that customer has asked for, or they have initially accepted the job but have then backed out of it before picking up the customer. This is a real challenge to Addison Lee’s business and particularly its reputation in the market.”[93]That paragraph was misleading in so far as it asserted that there was any “change to stop drivers from being logged off for refusing jobs”, whether seven years before, or at any time at all. That was clear in itself, since there was no evidence before me(1) of any such a change having been made or(2) that drivers were precluded from being logged off if they refused a job, but it was also flatly contradicted by the evidence first revealed by Mr Kelly in cross-examination in the manner which I describe in paragraphs 73-76 above.[94]The problems described by Mr Gavriel in paragraph 141 of his witness statement were obviously capable of giving rise to a temptation on the part of the respondent’s staff who managed the allocation of work to impose sanctions on drivers who refused jobs. However, that temptation would have been ignored if the respondent’s relevant managers had made it clear that no sanctions were to be imposed, and then ensured that if a member of the Driver Operations team or the Driver Liaison team had imposed a sanction, that member was subjected if necessary to the respondent’s disciplinary procedure.[95]Mr Kelly accepted in cross-examination that he did not at any time check the respondent’s records in the form of what it called the “driver logs”, even on a random spot-check basis, to see whether there was any record of a sanction being imposed on one or more drivers. In paragraphs 81, 84, 174, 176, 202-214, 216, 220, 223, 229, 283, 284, 285(l), 285(o), 285(q), 285(r), 285(s), 287(h), 288(n), 288(o), 288(p), 288(r), 288(s), 289(n), 289(o), 291(a), 291(bb), and 292(k) of his first witness statement, Mr Kelly referred (in some of those paragraphs or sub-paragraphs more than once) to what had been happening since 2017 in regard to sanctions. For example, in paragraph 292(k), on WSB page 484, which concerned the driver log for Mr Hector Ruiz (which was in part in the document starting at page 6586 and in part at page 4668), this was said. “14 March 2019: While Hector was a passenger driver, he was allocated a virtual rank job which he refused, and then he said he wanted to join the rank again. Gary Morgan, Senior VIP Operator in the operations team, prevented him from logging in, joining the virtual rank, accessing the prebook website and using the going home feature for one hour [HB/4668]. The job in question was going from Terminal 5 to Terminal 4 within the rank area and Gary should not have taken this step. This was not in line with the practice since 2017.”[96]In itself, that was evidence either of the ineffectiveness of Mr Kelly’s management, since he had plainly not got the message through to the team that he managed that they should not impose sanctions, or it was evidence that he himself at least countenanced the imposition of sanctions by members of his team at least in 2019.[97]During Mr Kelly’s cross-examination, I asked Mr Leiper whether there was anything in the documents before me from which the words “The job in question was going from Terminal 5 to Terminal 4 within the rank area” were drawn, or at least on which they were based. He replied that there was not, i.e. that there was nothing which supported those words. As a result, in order to give Mr Kelly an opportunity to clarify the matter, I asked him on what he had based those words, and he said that he was sure that he had based those words on some documentary evidence, and that he would not have put them in his witness statement (or, at least, approved them) unless there had been a basis for doing so. I then checked the spreadsheet containing what Mr Leiper referred to as the native data concerning Mr Ruiz’s work. It was document 1074. At row 8321 of the “Job Data” tab (it was the second of two tabs in the spreadsheet), there was the first of four entries for 14 March 2019. None of those four entries referred to a job of driving a passenger from Heathrow Terminal 5 to Heathrow Terminal 4. The journeys were, rather, taking them in order, as follows:97.1 Coram Street, London WC1N, to Heathrow Terminal 5;97.2 Heathrow Terminal 5 to Stadium Street, London SW10;97.3 Hester Road, London SW11, to Victoria Street, London SW1; and97.4 Heathrow Terminal 5 to Chiswick, London W4.[98]I pause to record (with regret that the passage is so long; I have included it because of its importance) that Mr Kelly said this in his first witness statement (at WSB pages 445-448) . “Sanctions for refusing jobs202. In the past, especially the period up to 2017, if a driver refused an ASAP job or a pre-book job without a reasonable explanation after being sent the job details, the operation team had the option to impose a sanction on the driver. The process that used to apply to deal with refusals is outlined in Car Control Manual dated March 2013 at HB/564. It sets out the process whether the operator is happy with the reason given or not. We generally followed this process up until around 2017 when everything changed as I will explain.203. The actual sanction imposed would depend on the circumstances and the operator would be able to select the type of sanction imposed ranging from preventing the driver from one specific job type only (for example, preventing the driver from the privilege of using the pre-book site for a period of time meaning the driver would still be allocated all other types of jobs such as ASAP jobs or virtual rank jobs), to preventing the driver from logging into the MDA for one hour (which would log off the driver and effectively prevent the driver from being able to work during that time).204. When imposing a sanction, an operator would need to do two things. First, they would need to access the driver browser in Shamrock and click on the ‘disabling’ tab and manually select the type of sanction to impose on the driver by ticking the type of job. An example can be seen if you look at Scott Kidd’s disabling tab which can be found in the bundle at [HB/4667]. As you can see, on the left there is a table which records details of any sanctions imposed. I will explain the table now. The first four columns (‘disable login’, ‘disable rank’, ‘disable pre-book’ and ‘disable going home’) are the type of jobs the driver can be prevented from being allocated. So if the operator ticked ‘disable pre-book’ (as is the case in the first entry on 16 March 2015), that means Scott would have been prevented from the privilege of using the pre-book site for a period of time. This would not have restricted his ability to work generally, or to join a virtual rank. The block would only work so as he could not have the additional benefit of using the pre-book site. He would have been able to be allocated all other job types. The next column (disabled by) indicates the name of the person who took this step; this would typically be someone in Operations or Driver Support. The next column (from) and the column on the far right (till) refers to the dates the measure is effective. The column second on the right (notes) refers to the reason why the step was taken. If a driver (or courier) did not receive any sanction during the periods they provided services to Addison Lee, the disabling tab would be blank. As an example, if you look at the ‘disabling’ tab for Adam Balog [HB/4662] there are no entries and so he would not have received any sanction.205. Second, the operator should manually insert a note on the driver history to record the sanction. This is usually a generic comment without specifying the specific type of sanction imposed. For example, in Scott Kidd’s driver history on 16 March 2015 there is a note that “[Scott] was banned for login or join rank from ‘16/03/2015’ ‘till ‘16/04/2015’” [HB/5600] which could be perceived to ban Scott from logging in for 1 month. In fact, on this occasion Scott was simply prevented from the privilege of using the pre-book site as can be seen by looking at his disabling tab [HB/4667]. Scott would still be allocated all other job types (and I can see that he was allocated and accepted a number of other jobs during this period [HB/5597].206. Unfortunately, the ‘disabling’ tab and the driver history do not speak to each other and so an operator would need to manually update both. I have noticed on some occasions where there is a sanction marked in the driver’s history, a sanction was not actually imposed on the driver (because the operator did not manually select a ban in the ‘disabling’ tab). As an example, if you look at Thambipillai’s driver history there is an entry on 3 November 2011 indicating that Thambipillai received a ban [HB/6941], but if you look at his disabling tab there is no evidence of any sanctions [HB/4669].207. As a result of my role, I didn’t really have any day to day contact with drivers, and so it would have been the operators who made these decisions and not me. If the issue was more serious or a repetitive issue, the issue would be referred to Driver Support. That’s because, as I explained, up until around 2017, operations only ever had the ability to issue sanctions for up to one hour. Anything longer than one hour would have been done by Driver Support and so operations would need to escalate the matter to Driver Support. From 2017 onwards, operators no longer had the ability to impose any sanctions as I will explain below.208. If a matter was passed to Driver Support, they would deal with the matter independently from the operators and they would then decide on next steps. Sometimes they would just speak to the driver about the issue, but where it was a more serious or a repetitive issue, they may have decided to apply a sanction. This could be preventing the driver from accessing the pre-book website for a specific time, or to prevent the driver from logging into the MDA for a period of time. It was rare in practice to prevent a driver from logging into the MDA.209. This all changed as part of the overhaul of practices since 2017 onwards which is when the Addison Lee business philosophy changed in order not to jeopardise the self-employed status of drivers. During or around August 2017, there was an instruction by Catherine Faiers, the then Chief Operating Officer, that drivers were to be able to freely turn down jobs, and not to be logged off or otherwise penalised for doing so. I don’t recall being involved in any discussions leading up to this, we were just told not to sanction for refusing jobs anymore. I don’t recall anything in writing but I do remember the instruction and the change in approach following the instruction.210. I relayed verbally this message to cease to apply sanctions to drivers for rejecting jobs to my direct reports and likely anyone in the operations team. It was quite a topic of discussion around this time so it would have cropped up in various discussions. I think some of the more experienced members of the operation team were quite surprised about the instruction, although most understood that this was the direction of traffic.211. From around this point onwards, in addition to the instruction not to penalise drivers for refusing jobs, we made changes to the sanction process so that an operator no longer had any ability to impose a sanction on a driver. Instead, the matter would need to be escalated to a senior member of the operations team (e.g. a supervisor) or Driver Support. We decided to remove the ability of an operator to apply sanctions to reduce the risk of an operator imposing a sanction in the heat of the moment. Although I am not sure of the exact timing of this change, the software change will have taken time to be developed and implemented.212. I believe the instruction to cease sanctions for a driver refusing a job was generally effective and the practice reduced massively. Although I am aware of the odd occasion when the instruction was breached or the messaging to drivers was not always on point. As an example, on 31 March 2018, Andre Da Silva, one of the test standard passenger drivers exchanged emails with Lorraine Chalkley and she refers to the possibility of Andre getting a ban for an hour if he refuses a job [HB/5125]. [The date of the email exchange was in fact 31 August 2018.] I appreciate that Lorraine’s email does not align with the overhaul of practices since 2017 including the instruction that drivers were to be able to freely turn down jobs as outlined above. Banning a driver for refusing a job was certainly not the policy at the time. I don’t know why she sent this email, but would not have expected drivers to be made inactive for turning down work in August 2018. In any event, I don’t think Andre received a ban on this occasion – there is certainly none recorded on his driver history on this date [HB/5331].”[99]The fact that no sanction was recorded on Mr Da Silva’s “driver history” was not conclusive, as was shown by the opening words of paragraph 206 of that extract which, for convenience, I now repeat: “Unfortunately, the ‘disabling’ tab and the driver history do not speak to each other and so an operator would need to manually update both.”[100]There was no direct evidence before me (in the form of a witness who could speak about the manner in which the Shamrock software was programmed to operate) about the way in which a ban could be imposed electronically, and what Mr Kelly said in paragraph 211 of his first witness statement, which I have set out in paragraph 98 above, about the manner in which that software operated was directly contradicted by what he said in the manner described by me in paragraphs 73-76 above. So, there was a possibility that if a sanction was recorded in one place to have been imposed, then it was in fact imposed, unless there was some evidence to the contrary effect. In some cases, Mr Kelly did refer to such evidence to the contrary effect, and that evidence was, I found, cogent. That evidence to a contrary effect was in the spreadsheets to which Mr Leiper referred as “the native data”, and when I checked what was in the spreadsheets, there was evidence that the driver in question had, during the period of a reported ban, done more work. That evidence was, I was told, drawn from the Shamrock software records, which included the data sent to that software via the driver’s XDA (or, if the driver simply used the respondent’s app on the driver’s own mobile telephone, the driver’s mobile telephone). Mr Segal said that the claimants accepted that the spreadsheet evidence was accurate, and there was no evidence before me to suggest that it was inaccurate. Those factors made me wonder whether the respondent’s need to keep up with a fast-paced and constantly moving situation both in the marketplace and on a day-to-day operational basis had led to wholly explicable contradictions between some of the relevant documentary evidence and what had actually happened. In that regard, I record that I accepted what Mr Gallagher said in paragraph 15 of his first witness statement (at WSB page 322), namely that the respondent’s business is “live and frenetic”, and that “business and management decisions need to [be], and are, made hour by hour”.[101]However, the driver logs for the test claimants showed that sanctions were being recorded by the Driver Operations and Driver Liaison teams (and Mr Kelly was responsible for the management of at least the first of those two teams) long after 2017, and the respondent accepted that at least some of those records of sanctions were accurate. That was because there were entries in the respondent’s “driver logs” for the test case claimants which at the very least were consistent with, and in some cases expressly recorded, the imposition of sanctions, up to and including a sanction imposed on Mr Edah-Tally on 5 April 2021 (to which Mr Kelly referred in paragraph 291(aa) of his witness statement, on page 483 of the WSB, accepting that the sanction in question had in fact been imposed).[102]In fact, the oral evidence of the test case claimants was to the effect that there was no change in practice in regard to the application of sanctions despite what the respondent put in the “new driver deal” of 21 June 2021 at for example pages 4360 (in clause 8.1) and 6017 (in clause 7.1). That evidence was very general, however, so that the claimants did not refer to specific dates when sanctions were imposed. I would have been inclined to accept that evidence of the claimants even if Mr Kelly had not given the evidence in cross-examination to which I refer in paragraphs 73-76 above, if only because of the following factors.[103]In the email exchange at pages 5125-5126, to which Mr Kelly referred in paragraph 212 of his first witness statement, which I have set out in paragraph 98 above, Ms Lorraine Chalkley in answer to Mr Edah-Tally’s question “When I receive a job like this, can I refuse refuse to do it?”, said this: “Unfortuently [sic] not sir, if you do refuse the job you are more likely to get a ban for an hour etc.”[104]In paragraph 45 of his first witness statement, Mr Kelly said this. “When I first became responsible for the Car Operations team in 2010, there were around 100 operators. This has now reduced to around 70 operators.”[105]In cross-examination, Mr Kelly said that Ms Chalkley was a member of the Driver Support team. Mr White managed that team. In paragraph 7 of his witness statement (WSB page 252), Mr White said that its size had “remained approximately the same” over about the last ten years, and that during that period it had had “around 8 to 10 people” in it. In paragraph 118 of his witness statement, at page 273 of the WSB, he said that “From Autumn 2017 onwards I am aware that everyone (across Operations and Driver Support) was instructed not to make drivers inactive for turning down work. I heard this from Bill Kelly, Head of Operations.” That was an odd way of saying (if he was saying it) that Mr Kelly told him to tell his team not to “make drivers inactive for turning down work”, but in any event the reality was that drivers were in fact made inactive for turning down work given what I say in paragraphs 73-76 above. However, if indeed Mr White was told by Mr Kelly to tell his team not to impose sanctions on drivers who refused a job, then Mr White was a singularly ineffective manager if only because of what was said by Ms Chalkley in the email which I have set out at the end of paragraph 103 above. In cross-examination, Mr White initially said that he did not know when Ms Chalkley started to work for the respondent. However, when he was referred to Ms Chalkley’s email at page 2817, he accepted that what she said there was accurate. It was this. “I am 22 years old and have worked for Addison Lee since August 2017. I started work as an Agency Car Jockey and was taken on full time in October 2017. I was then employed as a Driver Liaison Advisor in June 2018. This is how I met you guys!”[106]Accordingly, Ms Chalkley was plainly not told when, in June 2018, she started to work in the Driver Liaison team (now of course called the Driver Support team) that sanctions should not be imposed on drivers who refused jobs. The sanctions were likely to be “a ban for an hour etc”, the word “etc” showing that more than just a ban for an hour was likely.[107]Those factors all pointed firmly towards the conclusion (to which I came) that Mr Kelly’s repeated assertions that the respondent stopped imposing sanctions in practice on drivers in 2017 were not true. That was in part because they were contradicted by, or at least markedly inconsistent with, the documentary evidence to which I refer in the preceding paragraphs above. In any event, I concluded that the key question was not whether or not what could be classified as a “sanction” was ever imposed, but, rather, what economic consequences flowed from a driver’s refusal of an allocated job. That there were such consequences in the circumstances to which I refer in paragraphs 73-76 and (so far as relevant here) 81 above was indubitable. In addition, though, I concluded that either107.1 the respondent’s operation was in the nature of a very leaky sieve in the sense that instructions not to impose sanctions were, as a result of the inefficiency of the respondent’s operations, ignored by the operational staff of the respondent (whose motivation may have been to ensure so far as possible the economic survival and/or prospering of the respondent, and who may have acted out of a misguided sense of loyalty to the respondent, although they may, rather, have been motivated by the fact that they would receive a bonus if the respondent was profitable), or107.2 Mr Kelly and the other senior managers of the respondent, knowing how important it was for the respondent’s drivers not to refuse jobs, made it clear without recording it in writing that the Car/Driver Operations and Driver Liaison/Support teams were free to impose sanctions as and when they saw fit, and there were as a result other situations, which continued up to the time of the hearing before me, at least some, but possibly not all, of which were recorded in the driver logs, in which economic sanctions were in fact imposed (i.e. I found as a fact that they were imposed). Indeed, I could not see how, on the evidence before me, I could rationally have come to any other view about the imposition in practice of sanctions over and above those to which I refer in paragraphs 73-76 and 81 above. Courier drivers[108]Mr Valentine’s first witness statement contained the following material paragraph about the imposition of sanctions on courier drivers (it was paragraph 78, on page 371 of the WSB): “During my time at Addison Lee, I don’t recall any courier receiving a sanction for refusing a job. Even going back to the early 2000’s, couriers were and have always been free to accept or reject any type of job in their absolute discretion and for any reason. There was always a sense of disappointment in the operations room if courier jobs were rejected and that had an adverse impact on service delivery, but I don’t recall that ever turning into sanctions.”[109]Mr Valentine also said this in paragraphs 89 and 90 of his first witness statement (at WSB page 373). “89. Customers understandably want us to deliver our premium service and not showing up for a prebook goes against that. It gives the customer a really bad experience, and it’s an opportunity to earn money another courier could have had. Notwithstanding that, I don’t recall an operator ever giving a courier a sanction for failing to turn up to a pre-book. A short time in advance of the pre-book job, the system would send a notification to the courier’s MDA to remind them of the pre-book job they accepted. The system will tell the operations team if the courier hasn’t viewed a job. If they haven’t viewed the job, an operator will call the courier and if they can’t contact the courier they will simply reallocate the job to someone else. As I mentioned, resources are usually quite thin in the morning so it’s not easy to re-allocate a job early at that time, so it does sometimes mean we can’t get another courier there on time. That’s a bit frustrating for the operations team but I don’t recall them ever handing out a sanction to a courier in these circumstances. 90. In a similar way, we do expect couriers to turn up on time for pre-book courier jobs. The customer would have selected for the package to be picked up at a certain time for a reason. It’s often crucial for the package to be delivered by a specific time. This is the reason why on 26 January 2024 I sent a message on the Driver Portal suggesting that couriers may receive a ban if they don’t turn up to a pre-book job on time [HB/3118]. However, I don’t think we have ever followed through with this because, as outlined above, I don’t recall any sanction applied to a courier for turning up late for any job and that’s the case for both ASAP jobs and pre-book jobs. I believe the couriers would know that this was a hollow threat.”[110]The proposition that drivers would know that a threat of the sort there described “was a hollow threat” did not in my view bear scrutiny. Even if courier drivers knew that their contracts with the respondent provided that no sanction would be imposed if they refused a job, a threat to impose such a sanction would in my view still be likely to be taken seriously. As far as I could see, there was nothing in the driver agreement that related to the imposition of a sanction for turning up late to a job (as opposed to refusing a job), which made the threat of a sanction to impose a ban for turning up late all the more likely to be taken seriously. The terms of the threatened ban were very clear (the ban not being just a temporary one, but, rather, a ban “from the site”), and in my view could not credibly be explained away by Mr Valentine’s witness statement. At page 3118, this was written by him in what was stated to be “A Message From Kevin”: “I would like to remind you that taking a pre-booking carries a responsibility of being on time for it. We have had a number of instances where couriers have run late. This is not acceptable and if it continues we will have no choice but to ban individuals from the site.”[111]That also rather cast doubt on the accuracy of paragraph 84 of Mr Valentine’s first witness statement (at WSB page 372), in which he said this. “If we go all the way back to the 1990’s, I recall some of the operators giving a courier a bit of a grilling for turning up late for a job. That was about the worst it ever got. I don’t remember sanctioning any couriers even back then. But over time the operations have become more and more relaxed and I don’t think any courier will have received a grilling for a number of years, at least not since 2017.”[112]When he was giving oral evidence, I asked Mr Valentine about the difficulties caused to the respondent by a courier arriving to do a job late or refusing a job, by reference to what he said in paragraphs 78 and 84 of his first witness statement. In response, he said this (as noted by me and tidied up for present purposes). “I used to be on the road. I would not want to be shouted at. I would make sure we were as pleasant as possible to the driver. People spend a lot of time and skill putting a job together. If a driver refuses a job it can be a real challenge for the job to be done as well as it can be done; there is a massive knock-on effect if a driver says it cannot be done.”[113]Given the evidence and factors to which I refer in the preceding four paragraphs (108-112) above, I did not believe Mr Valentine’s repeated assertions that no sanctions were in fact imposed in practice on courier drivers who hired their vehicles from Eventech. That was if nothing else because of the statement which he had plainly written in the document at page 3118 set out in paragraph 110 above against the background of the rest of the evidence in paragraphs 108-112 above. Even if I had concluded that it was a bluff to say what I have set out in paragraph 110 above, I would have concluded that that extracted passage was plainly a threat which was intended to ensure that drivers were on time, so it was plainly incapable of being ignored as part of the material factual picture. The statistical evidence before me about the proportion of jobs which the respondent’s drivers rejected between 2019 and 2024[114]The respondent sought to show that the number of rejections of work increased between 2019 and 2024. The claimants dealt with that in paragraph 98 of their written closing submissions. I accepted those submissions, which were to the effect that what the respondent relied on in that regard was only “circumstantial evidence”, which did not in fact “support the proposition that rejecting jobs became easier between 2019 and 2024”. The time it might take for a driver’s call to be answered by the respondent’s operational teams[115]The issue of how long it might take for a driver’s telephone call to the respondent’s operational teams to be answered arose in regard to all of the test claimants. It is convenient to address that issue in relation to the evidence of Mr Balog, as he was the only person who gave oral evidence about his work as a courier driver and as a passenger driver. His evidence about refusing jobs was that, generally, he did not do that. In regard to his period as a courier van driver, he said this in paragraphs 118 and 119 of his witness statement, at page 22 of the WSB: “Refusing jobs[118]As I have set out previously, I generally, I didn’t refuse jobs. However, I would do so if I was going home and had pressed the “Go Home” button if the controllers still allocated a job taking me in the opposite direction from where I lived. Other circumstances in which I would refuse a job would be if I had an appointment and a job was given to me close to the time of that appointment. I used to inform the control team in advance if I had an existing appointment, so that they would not allocate me jobs. However they sometimes ignored this and so I had to refuse these jobs.[119]The process of refusing jobs was the same as when I was an executive driver; I would need to call the controller and explain why I couldn’t do the job. The only difference was that I called different phone numbers to get through to the executive driver controllers. It usually took longer, in my experience, for the executive driver control team to answer my call.” 116 In regard to refusing jobs when he was an executive passenger driver, Mr Balog said this (at page 13 of the WSB). “65. I rarely refused jobs allocated to me when I was working as an executive driver. However, if I did refuse a job, there were always consequences. Jobs were assigned to me through the MDA, and I was expected to accept any job that I was allocated. When a job was allocated, the MDA screen would start flashing and make a beeping noise. On the rare occasion that I wanted to reject a job, I would mute the device, call the controller, and explain why. One of the main reasons I would reject a job was because I had pressed the “Go Home” button and the job that had been allocated to me was taking me in the opposite direction. 66. Phoning the Executive control team was a difficult process. Sometimes my call would be answered immediately, but at other times, it could take between 5 to 15 minutes to get through to a controller. I would call the driver assistance number and type in the extension, then wait for the controller to answer my call. Sometimes, while I was trying the get through to a controller, the passenger would call me through the MDA to ask where I was. This then created a situation where I would have to answer the call and explain that I was unable to take on the job. Sometimes customers would become argumentative telling me that it wasn’t their problem, and I should pick them up regardless. Occasionally, I became so frustrated that I ended my call to the control team, pressed the screen on the MDA indicating that I had accepted the job and went to collect the passenger. 67. I remember on one occasion I argued with one of the controllers called Tony, after I rejected a job. I cannot recall why I rejected the job or when this happened, but I remember Tony telling me to bring my car keys back to the office and that I was “done” with Addison Lee. Subsequently, I logged off and went home. The next day, I was summoned to the office by Paul Regan and asked for an explanation as to why I had rejected a job. For the next two days after this, I was given minimum-priced jobs as a punishment, which meant I earned less.” 117 Mr Balog was cross-examined on that after he said that he had recalled in 2015 or 2016 being on a break in an Ikea or McDonalds car park and being told to do a job. He was then asked where he had referred to that in his witness statement and he said that it was in paragraph 67. When it was put to him that he did not say in that paragraph that he was on a break at the time, he said that while that was true, he remembered exactly where he was. Plainly, that was not correct, as he did not remember the exact car park, but I understood him to have a clear picture in his mind of what had happened. Of course, it was entirely possible, given the factors referred to by Leggatt J in paragraphs 15-22 of his judgment in Gestmin, that Mr Balog had inadvertently created that memory and was genuinely remembering something that had not happened. However, the respondent did not have any evidence from Mr Regan or a controller whose first name was Tony to rebut what Mr Balog said in paragraph 67 of his witness 118 In addition, the evidence of Mr Gherra did not in my view disprove what Mr Balog said in paragraph 66 of his witness statement. That was for the following reasons. 119 As the claimants said in paragraphs 71-72 of their written closing submissions (referring to the evidence of Mr Gherra as that of “Mr Guerra”), it was unclear from Mr Gherra’s witness statement and the table at page 2 of SB2 “whether the data included time spent on calls which were not answered before the caller hung up.” Those submissions continued: “In fact, it became clear, almost fortuitously (because PG [i.e. Mr Gallagher] referred to “weekly [sic] reports” in his oral evidence, following which the Cs asked AL to provide some of those reports overnight), that:(a) the data provided by Mr Guerra did not include the time spent on ‘abandoned’ calls;(b) the latter data was available and included in every daily report provided by Addison Lee’s Management, Information and Reporting team to management;(c) Mr Guerra had not been asked by AL to provide that information to the ET or the Claimants.”[120]Those submissions were developed in the next three paragraphs of the closing submissions. I found those submissions to be apt. As was said in the first part of paragraph 75 of the claimants’ closing submissions, even the figures given by the respondent were consistent with the possibility that “a significant minority of calls took several minutes”. The overall picture: the effect of what the claimnants called the respondent’s “carrot and stick” approach[121]The claimants dealt extensively in paragraphs 99-116 of their written closing submissions with the extent to which drivers were encouraged not to refuse jobs by(1) the use or threat of sanctions, and(2) the provision of a reward for not rejecting jobs. I found all of those submissions (that is to say those made in paragraphs 99-116 inclusive) to be well-made as far as the facts were concerned. That was because of my above findings on the imposition of sanctions or the threat of sanctions and because of what I say in paragraphs 172-173 below, where I refer to the most salient evidence before me about the manner in which the respondent in later years gave drivers economic incentives to do what the respondent wanted of them. I accepted that, as the claimants said in paragraph 111 of those submissions, “the effect of the system was to put pressure on drivers to do more work, more of certain types of work and to refuse fewer jobs”, which was “a type of economic pressure on the driver”. The claimants then said that that pressure had the effect of “adding to the levels of subordination and dependency on [the driver’s] employer.” The respondent’s practices in relation to the manner in which drivers dressed[122]Another thing that the respondent repeatedly asserted was that it had ceased to apply a “dress code” since 2017. In my view the issue was not what sort of dress code there was, but whether or not it was enforced in some way, and, if so, in precisely what way. Even then, I doubted that the issue was of central importance. The actual position, that is to say what the respondent actually did about the manner in which drivers dressed when driving for the respondent, was merely part of the overall picture.[123]The “dress code” which was in operation in 2013 was at page 595 (which was part of the “Driver Operating Guide” dated “August 2013” starting at page 591), where this was said. “Attire In order to maintain the professional image of the company, please, whilst providing your services to Addison Lee, wear a smart collared shirt (white or blue preferred) with tie. Smart trousers must also be worn, no jeans or jeans style trousers or chinos should be worn. No hats, caps or woolly hats should be worn whilst providing your services to Addison Lee, only religious head cover may be worn. Anoraks or any coat with a logo cannot be worn. Drivers may be penalised £30 for non compliance with the above.”[124]The fact that there was a potential penalty of £30 for a failure comply with the “request” showed that the word “please” could sensibly have been replaced with the words “you must”, so that there was not a dress “code” in place at that time but, rather, a dress instruction.[125]There was an undated “Driver Code of Conduct” at page 370 which was signed by “John Griffin (Chairman)”. Mr Gavriel said in paragraph 3 of his witness statement (at page 199 of the WSB) that he had worked for the respondent for over 26 years and that “until 2013, the business was still run by founder John Griffin”, at which time “the Griffin family sold the business to the Carlyle Group”. Accordingly, the Driver Code of Conduct at page 370 was likely to have been issued before 2013. It consisted of a single page. It started in the following way. “DRIVER CODE OF CONDUCT Drivers are expected to adhere to the Code of Conduct irrespective of signing 1. Presentation Codes of Conduct1.1 Drivers must wear a white or pale blue shirt with a tie clearly visible at all times, dark coloured trousers and a suit jacket or blazer. Female drivers must wear a white or pale blue blouse and dark coloured trousers or skirt. All drivers must wear business-like shoes. Sports shoes and sandals are not permitted. Jumpers, fleeces and hats are not permitted. Overcoats must be dark in colour and business-like.1.2 Your car must be kept clean both inside and out at all times. All personal possessions must be stored in the Addison Lee Driver Bag provided. The dashboard must be kept clear and nothing should hang from the rear view mirror.”[126]At the bottom of the page, this was said. “We employ mystery passengers and a team of Quality Assessors to enforce the above Codes of Conduct By signing this Code of Conduct Agreement, you (the driver) will be expected to adhere to all the points herein. Failure to do so will result in you losing your graded status.”[127]The respondent’s closing submissions included (in paragraph 100b) an acceptance that “[u]ntil December 2017, Addison Lee had a ‘dress code’ in place”. The submissions continued: “b. ... but after that time, it ceased to apply (Gavriel ¶ 60, 119 [WB/215, 234]) (White ¶ 83, 87 [WB/266, 267]) (Kelly ¶ 230 [WB/451-2]). Since then, Addison Lee has provided sensible recommendations for drivers, based on the expectations of the passengers. The Employment Tribunal was taken to various documents including, for example, a document entitled ‘Driver guidance’ [HB/1743]. It contains – as the title suggests – guidance. These are recommendations and no more. It is obvious that there is in fact no dress code in place. As Gavin White said, ‘Since Addison Lee stopped setting a dress code in 2017, the way drivers tend to present themselves has definitely become more casual’ (White ¶ 87 [WB/267]). As Martin Payne, one of the Test Claimants, said, ‘From 2021 onwards there was less enforcement of the policy that drivers had to dress smartly. This was common knowledge amongst drivers because we could see that some of the newer drivers did not wear a tie and smart clothes’ (Payne ¶ 91 [WB/144]). It is also obvious, even from Martin Payne’s comment, that some of the drivers wanted to carry on wearing smart clothes themselves, as a matter of personal pride, and did not approve of those who did not. That is reinforced by his email at [HB/6063] where he complains ‘by the way the other day I was at Term 3 and one of the new starters was wearing jeans … he had bare feet and flip flops … and this is where Add Lee is going.’ c. The Claimants rely on the fact that Addison Lee used to use ‘Quality Assessors’ on motorbikes to carry out spot checks of, for example, the dress code. As the dress code ceased to operate from December 2017, spot inspections by Quality Assessors stopped then too. Moreover, the role of the Quality Assessor became very limited (Gavriel ¶ 122, 124 [WB/235]) (White ¶ 84-89 [WB/266]) (Kelly ¶ 231 [WB/452]). Note also Balog ¶ 71 [WB/14] ‘I remember that the inspections seemed to stop at some point while I was an executive driver [ie before 2018], but I am not sure exactly when.’[128]The claimants’ closing submissions on this issue were as follows. ‘The dress code120. With the exception of KET, who transferred to AL from Tristar, all the TCs commenced driving for AL at a time when a driver dress code was prescribed by the Driver Code of Conduct [370] and/or was communicated to them in induction training. All of the TCs testified that they were made aware of a dress code which required them to dress smartly – see AB/30; ADS/10; DN/10; HR/57-58; MP/19; TM/48-49; and RP/100-103.121. If the dress code was relaxed at any subsequent point in time, no communication was sent to existing drivers informing them that it had changed. As MP put it in x-ex, “when I started, and until 2021-2022, we went for an induction with a piece of paper with a drawing of a gentleman with a suit and tie on – they said ‘this is the way you should dress.’ So I just carried on as I had an abundance of suit and ties”.122. In any case, the training given to new drivers demonstrates at all material times up to 2024, AL has informed drivers that they are not just encouraged but expected to dress smartly. This is illustrated by: 122.1. [1085 – 1086] (part of the driver induction presentation dated 4 January 2019). 122.2. [1608] (part of the driver induction presentation dated 2023), to which a different photo had been added and below which a note reminded the trainer to “Discuss what is expected of an Addison Lee Drivers appearance” including that “Exec drivers should wear a tie”. 122.3. The Driver Guidelines [1648-1650] which formed part of the August 2023 Quick Start Guide. They included a section on “Clothing” which cannot simply have been an oversight, because the Guidelines had very recently been updated to include reference to the April 2023 “AL Tiers”. 122.4. [1683] and [1684] (part of the driver induction presentation dated 5 February 2024 to which new photos had been added to illustrate the standard of dress required.123. Further, in evidence which was not challenged in x-ex, Arnold Ban said that when he attended induction training on 8 January 2024, he was told that he always needed to dress smartly for Addison Lee, including a wearing a shirt and trousers – see ABan/20.124. The fact that the dress code continued after December 2017 is also evidenced by: 124.1. an entry made by Freddy Goldman in Radu Rosu’s driver log [8108] dated 29 January 2018 which reads: “QA Spot check: Wearing blue chinos and inappropriate jacket. He knows the dress code and will dress accordingly from now on.” 124.2. The correspondence in April 2021 between Mr Goldman and a driver whose name has been redacted [8565-8569] about the driver’s dress. 124.3. The entry dated made by [sic] Freddy Goldman in KET’s driver log dated 21 January 2021 [7785] to which KET referred at KET/99 which reads: “QA spot check: Above not wearing a tie at pickup [redacted] driver spoken to and reminded of adlee dress code”125. It was repeatedly suggested to the TCs by AL’s Counsel that the dress code was simply advisory, because it was in the drivers’ interests to be well-presented and create a professional impression. But the drivers could only benefit indirectly from repeat business. By contrast, AL had a direct interest in promoting and maintaining the premium nature of the transport service it offered. As GW put it in e-ex “When you market yourself in the industry as a premium cab company, there is a certain expectation from clients. If they want an Uber driver in shorts and a tshirt they book an Uber, and they pay Uber rates. If they pay an increased premium, there is an expectation among clients.”126. One question and answer in the x-ex of RK (referring to the time when he was an executive driver, but in which he confirmed he believed he had to stick to the same standard as an owner driver) highlighted the difference between the parties’ positions: “[Q: Wasn’t a requirementcustomer demand?] No, it was a company standard”.[129]Although I found the whole of that passage to be persuasive, and preferred it to the respondent’s submissions on the point, I found what the claimants said in paragraph 125 of that passage of particular importance and relevance. That was because the persons who drove for the respondent were not able to benefit except in a very general way from maintaining (as a result of the driver in question ensuring as far as possible that the respondent’s image was maintained) the respondent’s claimed premium brand image. That was because the drivers were not able to obtain repeat business themselves from customers, although they might, by providing an excellent service, or merely one that a particular customer liked, be asked for as a repeat driver. The main beneficiary of the claimants acting in the manner recommended by the respondent in regard to dress and other things was the respondent.[130]In addition, while the claimants dealt with the issue of “spot checks” separately, in paragraphs 134-136 of their written closing submissions, I regarded them as being of particular importance in relation to the issue of the relevance of recommendations in regard to dress. While Mr Gavriel said (and the claimants were not in a position to challenge this) in paragraph 124 of his witness statement (at page 235 of the WSB) that “Addison Lee now only employ one Quality Assessor, who will only visit a driver in their vehicle if our customer service team receive a complaint about the cleanliness or safety of that vehicle”, that was not the message which the respondent gave out to inductees. In that regard, I found what the claimants said in paragraphs 135 and 136 of their submissions to be highly persuasive, but in one respect (to which I refer in the subsequent paragraph below) the first of those paragraphs was inaccurate. The passage was in these terms. “135. Whatever the truth about the role of quality assessors, AL consistently informed its drivers that they would be monitored for their compliance with AL standards – see: [1053], a document produced in October 2018 reflecting what drivers including all the TCs were told up to at least that date: “A driver support team of 15 people who continuously monitor driver performance. … Quality Assessors dedicated to conducting driver/car spot checks around London on a daily basis to ensure compliance carried out on a 24/7 basis”; [2757-8] which announced in July 2017 the introduction of “‘mystery journeys’ to ensure that high standards are being maintained at all times”; [1609] (part of the 2023 driver induction), which told drivers that “We have a team of quality assessors that help our drivers by checking we are being compliant and following brand standards”; and [8748] a flyer for the Addison Lee Diploma which referred both to spot checks and positive feedback from “mystery customers”.[136]It is that information which created a level of ‘control’, regardless of whether it was, or became, based on a deception. At the risk of repetition, no such control mechanisms were relied on by the drivers in the Uber and Bolt cases.” 131 The document at page 1053 in fact said (as the first bullet point on the page) that the respondent had “A driver support team of 25 people who continuously monitor driver performance”, not 15 such people. That page was part of the document headed “Addison Lee – Driver Recruitment – Training and Referral Process” which started at page 1046 and was undated but which was stated in the index to have been issued on 23 October 2018. Page 1053 was headed “Further Driver Training and Monitoring”. Two bullet points down from the top of page 1053, this was said, as the third bullet point on the page. “Quality Assessors dedicated to conducting driver/car spot checks around London on a daily basis to ensure compliance carried out on a 24/7 basis”. 132 In paragraph 86 of his witness statement, at page 266 of the WSB, Mr White said that “Soon after the Carlyle Group bought Addison Lee in 2013 they conducted a redundancy exercise and the number of QAs [i.e. Quality Assessors] was reduced to two.” So, by asserting in 2018 that the respondent had a team of 25 people who continuously monitored driver performance and that there was an unspecified number of “Quality Assessors dedicated to conducting driver/car spot checks around London on a daily basis to ensure compliance carried out on a 24/7 basis”, it was at least possible that the respondent was there overstating the extent to which drivers might be monitored. 133 In paragraph 89 of his witness statement (which consisted of two sentences), at page 267 of the WSB, Mr White said that “In or around 2020 one of our QAs sadly passed away, and since then there has only been one.” 134 It was said in the index (and the respondent accepted that it was accurate) that the document at pages 1591-1637 was the material which was used to induct new drivers in 2023. At page 1609, this was said. “Quality Assessors: We have a team of assessors that help our drivers by checking we are being compliant and following Brand Standards”. 135 Accordingly, the claimants’ quotation in that regard was also slightly inaccurate. However, that inaccuracy was immaterial. On any view, the idea that the respondent in 2023 had a team of assessors checking that drivers were being compliant and following “Brand Standards”, was misleading, as there was at that time only one person employed as a Quality Assessor. In fact, paragraph 89 of Mr White’s witness statement continued: “They will now only check a car if there has been a complaint of it being unpleasant for passengers.” Either that was true and the words on page 1609 which I have set out in the preceding paragraph above were untrue, or what Mr White said in the second sentence of paragraph 89 of his witness statement was untrue. 136 I come to a conclusion in that regard in paragraph 141 below, but before doing so I record that Mr Gavriel sought to explain away the impact of the words used by the respondent in the documents which it put before new drivers during their induction. He did so despite saying, in the first sentence of paragraph 33 of his first witness statement, at page 257 of the WSB this: “Driver Support has no involvement in the provision of induction training for drivers.”[137]So, he had no direct evidence to give about what was said to drivers during their inductions. Mr Gavriel also said (in cross-examination) that he was not responsible for the written induction material, including the slides used when inducting new drivers, and that it was his line manager, Mr Tony Smith, the respondent’s “Head of Driver Experience”, who was responsible for what was said in that material and those slides.[138]Mr Gavriel’s evidence was nevertheless that anything said in the written induction material before me which was consistent with the oral evidence of the claimants’ witnesses but which was inconsistent with the oral evidence of the respondent’s witnesses, was simply included in the induction material in error and was not what “would be explained to drivers at the training”. A salient example of that was what he said in paragraph 61 of his witness statement (at WSB pages 215-216), which dealt with two issues, to the first of which I return below (in paragraph 153). ‘Although on slide 37 [HB/1711] this induction training refers to drivers having to be logged in to Addison Lee’s system for at least three hours before they can use the “Going Home” button, this has not been the case since in or around 2018. Slide 38 [HB/1712] also references a maximum “On Break” time of two hours at a time, however this limit does not apply in practice anymore. These two bullet points in the induction training are simply errors that have not yet been corrected in the slides, and do not reflect what would be explained to drivers at the training. I understand that Bill Kelly, Head of Driver Operations, has explained what these “On Break” and “Going Home” options are and so will not repeat him.’[139]That assertion that those bullet points were simply “errors that [had] not yet been corrected in the slides” did not sit well with the opening words of paragraph 66 of Mr Gavriel’s witness statement, at page 216 of the WSB, which were these. “During my time in the Driver Recruitment team this induction training has been continuously tweaked, reflecting changes such as the removal of any dress code.”[140]If the respondent “continuously tweaked” the induction training to reflect changes, then either the continued inclusion of the “two bullet points” to which Mr Gavriel referred in paragraph 61 of his witness statement, which I have set out in paragraph 138 above was a curiously inadvertent or an advertent error. In any event, unless the respondent corrected any error in its induction material in writing, there was at least a real risk that the induction material led to a belief on the part of drivers about the practices of the respondent that was in fact not correct. Returning now to the subject-matter of paragraphs 131-135 above, the respondent did not point to any documentary evidence in the bundle before me which stated that there was at any time only one Quality Assessor employed by the respondent, and that that assessor “will now only check a car if there has been a complaint of it being unpleasant for passengers”. All that was said so far as relevant in the respondent’s written closing submissions was in paragraph 100c, which I have set out in paragraph 127 above.[141]So, I asked myself, did I accept the respondent’s evidence that there was in existence only one “Quality Assessor” after 2020? Especially because it was consistent with the passage in the witness statement of Mr Balog which the respondent set out in paragraph 100c of its written closing submissions, I did accept that evidence. What, I therefore asked myself, did that mean about the induction material? The answer had to be that the material was clearly misleading. I did not need to decide whether or not it was deliberately misleading, although, I concluded when deliberating, it would have helped me to assess the credibility of Mr Gavriel’s assertion that it was not what “what would be explained to drivers at the training” if I had been able fairly to come to a conclusion in that regard.[142]However in the circumstances that the respondent had not called(1) a witness to say what inductees were in fact told, or(2) Mr Tony Smith to tell me why there were errors in the written induction material, I concluded that I could not fairly come to a conclusion on the reasons why the respondent’s induction (or training) materials were misleading. Nevertheless, I was able to conclude that the absence of such witnesses undermined Mr Gavriel’s assertions about what would have been said to inductees. In addition, what a witness says about what would have been done by someone else is not evidence. It is speculation. It may be informed speculation, and therefore, despite not being evidence and because (as stated in rule 41 of the Employment Tribunals Rules of Procedure (2013)), the tribunal “is not bound by any rule of law relating to the admissibility of evidence in proceedings before the courts”, it could be taken into account by an employment tribunal. However, in order to be of some probative value here, Mr Gavriel’s speculation would have had to be based on some direct evidence of his which supported it. Nothing which Mr Gavriel said, either in his witness statement or when giving oral evidence, helped in that regard. He told me that he was not present at, and was not involved in, driver inductions.[143]In those circumstances, I concluded that the respondent had not satisfied me on the balance of probabilities that new drivers even in 2023 were told in effect that they could wear what they liked and that there was no one who could on behalf of the respondent check what they wore and, if it was not satisfactory to the respondent, take any kind of action about it.[144]In addition, the “Driver guidelines” of August 2023 at pages 1648 and 1649 and what was said in February 2024 to new drivers as shown at pages 1683 and 1684 was in my view plainly intended by the respondent to be acted on, even if it was the subject only of peer pressure. Accordingly, even if drivers discovered that there was not a team of quality assessors, but only one assessor, and even if it were the case that that one assessor only assessed a driver and/or his vehicle if a complaint were made about that driver and/or his vehicle, the respondent intended the guidelines to be acted on via peer pressure which, I concluded, if only from the evidence of Mr Payne which the respondent set out in paragraph 100b of its written closing submissions and which I have set out in paragraph 127 above, the respondent counted on being applied. The respondent’s allegation in paragraph 13(2)(ii)(h) of its amended grounds of resistance at page 172 that ‘There is no longer a “half-hour rule” for logging into the XDA before pick-up time. No Drivers have been specifically required to log on to their XDA 30 minutes before a pre-booked job since September 2017’[145]The claimants submitted the following things about this allegation. “The half-hour rule for logging into the XDA before pickup time127. It is common ground that no communication was sent to existing drivers informing them that the rule requiring them to log into the XDA at least 30 minutes before pickup time had been abolished.128. The proposition that it was abolished in September 2017 is impossible to reconcile with the fact that the 30 minute rule was explained to new drivers: 128.1. in the driver induction presentation dated 4 January 2019 [1090] and [1092]; 128.2. in the driver induction presentation dated 2023 [1622]; 128.3. in the Driver Guidelines which formed part of the August 2023 Quick Start Guide [1648]; 128.4. in the driver induction presentation dated 5 February 2024 [1698] and [1702] (although many other slides in this presentation had been updated).”[146]I did not accept that it was impossible to reconcile the proposition that the “halfhour rule for logging into the XDA before pickup time” was abolished in September 2017 with the factors referred to in paragraphs 128.1-128.4 of those submissions. That is because it was possible that the respondent had misled inductees in the same way that it had misled them about the extent to which the respondent’s advice about attire would be enforced, as I have described in paragraphs 122-144 above.[147]The respondent’s submission in this regard relied purely on the evidence of Mr Kelly: the submission was in paragraph 99d, and was this. “The Claimants rely on the fact that, originally, they had to be in their vehicles 30 minutes before a prebook. From September 2017, this changed. From 2017, this has not been a requirement (Kelly ¶ 157 [WB/436]).”[148]In the circumstances that148.1 I had heard no evidence from a witness for the respondent about what instructions, if any, were given to the party which supplied and maintained the Shamrock software in connection with the requirement to log in “30 minutes before pickup time”,148.2 I had heard no evidence from a witness who was employed by, or otherwise acted for, the party which supplied that software, and148.3 the documents to which the claimants referred in paragraph 128 of their written closing submissions all said what the claimants said they said about the need to book in (using the words in the most recent document, at 1698) “no later than 30 mins before your job is due”, I could see no rational conclusion other than that the respondent intended drivers at all material times to log in no later than 30 minutes before the job was due, and that it intended them to believe that they had to do that. I would have come to that conclusion even if Mr Kelly had not created the false document at page 8725.[149]I also concluded from the evidence which I had heard that the respondent would re-allocate a job if the driver had not logged in 30 minutes before the job was due unless there was a discussion with the driver and the respondent was satisfied that it was in the interests of the respondent for the driver nevertheless to do the job. I add for the avoidance of doubt that I do not mean in any way to suggest that it was unreasonable for the respondent to have a rule, or at least a practice, to that effect. The respondent’s allegation in paragraph 13(2)(ii)(i) [sic] of its amended grounds of resistance at page 172 that “Drivers no longer have to be in the vehicle and away from home in order to log in. They may log in from anywhere. This change occurred in 2019”[150]The position in regard to the requirement, which the respondent appeared to me to accept and which I in any event concluded on the evidence before me existed before 2017, for a driver to log in from the vehicle which he was driving, was related to the requirement for drivers to log in at least 30 minutes before a prebooked pickup time. The claimants’ position in paragraphs 129-131 of their written closing submissions was as follows. ‘Being in the car to log in129. No communication was sent to existing drivers, either in 2019 or at any time subsequently, to inform them that they were no longer required to be in their vehicle in order to log in.130. A requirement to be in the car to log on was communicated to new drivers: 130.1. in the driver induction presentation dated 4 January 2019 [1092]; 130.2. in the driver induction presentation dated 5 February 2024 [1702].131. Further, all of the slides presented to new drivers cited above relating to the requirement to log in 30 minutes in advance of a pickup also stated that that: “Drivers must be in the car and On Way to the pre-book pick up address no later than the recommended time when confirmation is given. ...”’[151]The respondent also relied in response to this allegation on the evidence only of Mr Kelly. In paragraph 99c of the respondent’s written closing submissions, this was said. “The Claimants rely on the fact that, originally, they had to be in their vehicles in order to log on to the app. From 2019, this changed. From 2019, it was no longer necessary for a driver to be in his vehicle in order to log in to the app (Kelly ¶ 63(a) third bullet [WB/409]).”[152]For essentially the same reasons as I give in paragraph 148 above, including that I would have come to the same conclusion even if Mr Kelly had not falsified the document at page 8725, I could see no rational conclusion other than that the respondent intended drivers at all material times to log in from their vehicles, and that it intended them to believe that they had to do that. By way of illustration, even in the induction document of which page 1702 was a part (which was dated 5 February 2024), this was said. “Planning Your Journey REMEMBER! Be in your car to log on”. The respondent’s allegation in paragraph 13(2)(ii)(j) of its amended grounds of resistance at page 172 that ‘The “go home button” is now available at any time. There is no longer a requirement to be logged in for 4 hours before it can be used and no limit to the number of times a day it can be used. This change was introduced on different dates for different types of driver but in all cases by the end of 2018’[153]The respondent’s position in this regard was stated in paragraph 99a of its written closing submissions, in the following manner. “The Claimants rely on the fact that they could not press the ‘going home’ button, to attract an auto-allocated job in the direction of home, unless they had been logged on for a set number of hours (either three or four). Whilst there was originally a mechanism of this nature, the ‘going home’ button could be pressed at any time from November 2022 (and without limit to the number of times it could be pressed) (Gavriel ¶ 61 as corrected [WB/215]) (Kelly ¶ 63(h) as corrected [WB/412]). In any case, the Claimants could go home at any time, remaining logged on or logging off as they preferred. The ‘going home’ function was not essential to the drivers’ work; it did not have to exist at all.”[154]The words “November 2022” were a replacement of the words “the end of 2018”, which were originally in the witness statements of Mr Gavriel and Mr Kelly. That claimed correction was made during the hearing before me. Again, though, no documentary or oral evidence was put before me about any instruction given to make the necessary change to the Shamrock software.[155]The claimants’ written closing submissions on this aspect of the matter were as ‘The Going Home Button132. AL’s case on the “Going Home” button has changed, in that BK, closely followed by CG, revised the date on which it is alleged the button could only be exercised after a period logged on from 2018 to 2022.133. It is not in dispute that any such change in the functionality of the Going Home button was never communicated to existing drivers. The contention that this took place in 2022 is not supported by the information provided to new drivers: 133.1. in the driver induction presentation dated 2023 [1625]; 133.2. in the driver induction presentation dated 5 February 2024 [1711].”[156]In fact, the correct reference in paragraph 133.1 of that extract was to page 1624, and not page 1625. At page 1624, this was said. “Must be logged on for at least 3 hours before activating Going Home button”.[157]The same words were used on page 1711. For essentially the same reasons as I give in paragraph 148 above, including that I would have come to the same conclusion even if Mr Kelly had not falsified the document at page 8725, I could see no rational conclusion other than that the respondent intended drivers at all material times not to press the “Going Home” button unless they had been logged on for a period of time (originally it was 4 hours; latterly it was 3 hours), and that the respondent intended the drivers to believe that they could not press the “Going Home” button unless they had been logged on for the relevant number of hours (which was never less than 3). For the avoidance of doubt, Mr Gavriel did not give any direct evidence in regard to the manner in which the “Going Home” button worked (or, as the case may be, did not work). The respondent’s allegation in paragraph 13(2)(ii)(o) [sic] of its amended grounds of resistance at page 173 that “from June 2021, sub-contracting has been permitted (in accordance with the full recasting of contract terms in June 2021)”[158]Before examining the parties’ submissions on the question whether or not the test claimants were able to sub-contract, I record that158.1 the right to sub-contract is commonly relied on by a claimed employer in asserting that a claimant was not an employee, rather than that the claimant was not a worker within the meaning of for example section 230(3)(b) of the ERA 1996 (“a limb b worker”), and158.2 where it is claimed that a person cannot be a limb b worker in part because of the existence of a right to subcontract, the relevance of a genuine right to sub-contract is that it means that the claimed worker has a degree of autonomy.[159]The respondent’s written closing submissions on the question of whether or not the claimants had a genuine right to sub-contract and, if they did, whether the claimants exercised that right in practice, were as follows. “34. Under the new driver deal, introduced in June 2021 for drivers (standard passenger, executive and partner) and in July 2022 for couriers, drivers are entitled to subcontract work to third parties. Clause 1 states as follows: ‘1.4 You have no obligation of personal service under this Agreement. You do not have to undertake any work or perform any services at all. Additionally, you may sub-contract or delegate to one or more individual(s)… for the provision of Services under this Agreement.’35. Although no driver has taken up this right in practice, the right is real and genuine. It has been communicated to drivers and practical arrangements to enable them to exercise the right of substitution have been put in place: see Gallagher ¶ 190-200 [WB/349-352]; Kelly ¶ 264- 270 [WB/463-465].36. Mr Gallagher also explains the different option of delegation (¶ 201-202 [WB/352] and see Kelly ¶ 265 [WB/265]). Here, the driver has a job and, rather than reject it, he asks for the job to be transferred to a specific driver nominated by him. This happens fairly frequently, and has been something that a driver has always been able to do. The Claimants seem to suggest that this is not true delegation / substitution, because the process involves the intervention of Addison Lee. That is wrong: the function of Addison Lee is to move the job from Driver A to Driver B within the system. There has been some suggestion that this is not delegation but nomination of an alternative. That seems to be a distinction without a difference; there is no suggestion that there is any veto exercised by Addison Lee. As Mr Gallagher said in xx, ‘We facilitate that.’ Irrespective of whether the right to sub-contract has proved popular, the right to delegate exists, is exercised, and is effective.37. Equally, there is no suggestion that a driver cannot give out his contact details to a customer. As Mr Kelly said, Addison Lee simply would not know that this had happened.”[160]The relevance of what was said in the final paragraph of that extract escaped me in the context of the regulatory regime. That regime precluded a driver who did not have a licence of the sort described by Lord Leggatt JSC in paragraphs 30- 33 of his judgment in Uber, at [2021] ICR 666-667, namely a private hire vehicle operator’s licence for London, from taking private hire bookings from the public.[161]That which Mr Gallagher “explain[ed]” as described in paragraph 36 of those submissions was not sub-contracting; it was, rather, nomination, as was recognised by the respondent. Then to argue that it was “wrong” to say, as the claimants did, that “this is not true delegation / substitution, because the process involves the intervention of Addison Lee” in my view involved a failure to accept the obvious point that here the only autonomy was to ask the respondent to allocate the job in question to another driver. To the extent that as a matter of fact, there was a right to make such a request, I accepted what the respondent said in paragraph 36 of its closing submissions.[162]The claimants’ written closing submissions in this regard were as follows. “The ‘right to sub-contract’ 40. The so-called right to sub-contract was set out at clauses 1.4 to 1.6 of the June 2021 contract [HB/4356-4357]. On its true construction, this was not a right to sub-contract in the sense in which that term has been used in the authorities – most recently in the decision of the Supreme Court in the Deliveroo case – concerning the right to provide a substitute to perform the work which the worker has agreed will be done. 41. Crucially, the contract did not allow a driver to use a sub-contractor or substitute when he was logged on to the AL app. This was made clear to the drivers in a document entitled “Driver Deal FAQ” which was issued to them in June 2021 [HB/2984-2988]. Under the heading “Can someone complete bookings on my behalf?” [HB/2987], AL stated that a driver could sub-contract his vehicle to another PCO licensed driver, adding: “At no point are you permitted to ask another driver to fulfil jobs allocated directly to you, and you must inform us when any other PCO licensed driver is in possession of your vehicle and available for jobs.” Clause 1.6 of the June 2021 Contract confirmed that, if the driver decided to “sub-contract”, AL would provide separate log-in details and a subaccount for his “sub-contractor”. 42. In reality, therefore, the right that AL introduced was a right for a driver to allow another PCO licensed driver to use his vehicle when he was not using it. The driver could not use a substitute when he himself was logged on. The introduction of a right to share a vehicle does not undermine the Claimants’ contention that, by logging on to the AL app, they undertook to do or perform work or services for AL personally. 43. None of the TCs, including those nominated by AL, has exercised the right to sub-contract. The orders made by EJ Tynan on [1 August 2024] required AL to state, for each of the 12-month periods ending May 2022, 2023 and 2024, the number of drivers in respect of whom sub-contracting arrangements were notified and recorded [205]. AL’s response was that it had no records of any driver ‘sub-contracting’ their vehicle. The LDCs [i.e. the Leigh Day claimants] say that this is consistent with the reasonable expectations of the parties to the contracts. 44. AL’s more recent reliance on what it describes as the right to ‘delegate’ is even more absurd. As the EJ pointed out, and BK agreed, this is a right of one driver(a) (A) to nominate another AL driver(b) (B) to be offered a job that has been/would be offered to A through AL, but he has not accepted because he cannot do it. The proposed ‘delegate’ (nominee) could only be another AL driver, to whom the booking would then be allocated by AL in his own name. Driver A would derive no financial benefit from the allocation to B. Further, AL retains the right to decide whether or not the work should be allocated to Driver B, or to another driver. 45. That is no more a right to ‘sub-contract’ in the relevant sense, than a barrister who is contacted by chambers to do a trial and is not available on the required dates suggesting to chambers that they see whether a named colleague is available for that trial. 46. It was not put to any of the TCs that they had ever exercised the right to ‘delegate’. At BK/268, BK said that it is more common for drivers to delegate a job than it is for them to sub-contract. In x-ex, he admitted that he did not “have visibility” of this happening and could not therefore say how often it had occurred. He accepted that he was not aware of any driver who had entered into a ‘sub-contracting’ arrangement.”[163]In fact, the whole of paragraphs 266-269 of the first witness statement of Mr Kelly (on page 464 of the WSB) was relevant, and not just paragraph 268 of that statement. It is not necessary to set those paragraphs out here, but I record here that, for the avoidance of doubt, I took them fully into account but accepted that what the claimants said in the passage of their written closing submissions which I have just set out was an accurate statement of the reality of the situation as far as the claimants’ theoretical right to sub-contract was concerned. Relevant aspects of the relationship between the agreement to lease a vehicle and a driver agreement with the respondent[164]In addition to the impact of the need to ensure that the respondent’s branding material was displayed when driving a standard passenger vehicle with a customer of the respondent in it, to which I refer in paragraphs 61 and 62 above, the claimants drew my attention in their written closing submissions to the following aspects of the lease agreements entered into by the non-partner driver test claimants. Insurance[165]In paragraph 161 of their written closing submissions, the claimants said this. “At all material times both before and after June 2021, drivers who drove vehicles leased from Eventech and did not opt out from the insurance policy it provided could not use the vehicle for any other business use unless they obtained separate insurance to enable them to do so – see clause 2(c) of Schedule 1 to the VHA [i.e. the vehicle hire agreement] issued to TM on 16 April 2012 [6843] and para 2.4 of Schedule 1 of the VHA issued to TM on 11 August 2023 [6936].”[166]That submission was well-founded, including in relation to the situation concerning courier drivers, to which I return in paragraph 200 below. The agreement of which page 6843 was part was in fact part of an agreement dated 16 April 2021 (so that the numbers “2” and “1” had been mistakenly transposed in the claimants’ submissions). The reference to page 6936 was apt, however. The earliest vehicle hire agreements that I could see for all of the test claimants were issued in 2014. Taking by way of example the document relating to Mr Ruiz at pages 6278-6287, which was dated 10 November 2014, at page 6284 in paragraph 2.c, this was said: “Your Vehicle is only insured for business arranged by or carried out on behalf of Addison Lee or pursuant to your Private Hire Vehicle Driver Contract. No other business use is insured and is strictly prohibited. In the event that you use your Vehicle for other business purposes, you will be fully liable for any claims arising, liable to prosecution under the Road Traffics Acts and liable to have your public carriage office licence revoked”. Aspects of the impact of the vehicle hire agreement on the relationship between a driver and the respondent[167]In paragraphs 266-268 of the respondent’s written closing submissions, this was said, under the heading “F. Breaks”, which was part of the section in those submissions concerning the issue of paid annual leave. “266. As noted above:a. A break is significant both because it can establish a termination (so as to prevent further carry-over) and because it can establish a break in a series of deductions.b. The Claimants’ pleaded case, and evidence, on the breaks that they took is deeply unsatisfactory. All that Addison Lee can do is to identify periods when a claimant stopped working. It has disclosed all of the underlying job data which allows that exercise to be undertaken. All that the Test Claimants have done (with their solicitors) is to review the (also disclosed) pay summaries to make a best estimate of when they might have taken a break. Not one of them suggested that he has reviewed (let alone disclosed) a diary or sought out receipts that would assist in this process (although many of them professed to having provided these documents to their accountants for the purposes of preparing their accounts and tax returns). 267. Of course, a break may be such as to stop a series of deductions without amounting to a termination; it is a matter of degree. 268. The Respondent contends that one or more of the following is sufficient to establish either (and the Tribunal is invited to rule as to whether each or all of these would amount to a termination and/or break in a series of deductions): a. Any protracted break, where (for example) an individual takes time to do different work (whether for Wheely or to undertake construction work) or where they are prevented from working (e.g. because they allow their PCO licence to lapse, are banned from driving, or because of covid). b. Any break where the driver returns the vehicle and MDA, and so cannot undertake any work.c. Any break between contracts.d. Any protracted break where the driver ceases working, for whatever reason.”[168]While those submissions were of importance in relation to the issue of breaks generally, and I return to them in paragraph 206 onwards below, I saw that in paragraph 268.b, the respondent was either saying or implying that there was a direct link between the agreement for the lease of a vehicle (and the use by the driver of an MDA unless the driver used the respondent’s app on his mobile telephone instead) and the relationship between the respondent and the lessee as a driver, so that, it was said, if a vehicle was returned then that automatically ended the relationship between the respondent and the driver, even if the driver was simply going on holiday for a couple of weeks. That submission was strongly disputed by the claimants, as can be seen from what they said in paragraph 326.3 of their written closing submissions, which I have set out in paragraph 210[169]In any event, I saw that it was said by Mr Klepacki that he never returned the vehicle which he rented from Eventech during periods when he took holiday because (as he said in paragraph 58 of his witness statement, at WSB page 181) he believed that if he took a break or a holiday, then “if it was longer than 2 weeks you had to return the car”.[170]When cross-examined on that, he referred to clause 7.3 of the driver agreement which first applied to him, which was at page 5637 and was in these terms. “Company car drivers who wish to leave the company should be aware that they will not be able to re-hire a vehicle for a period of 6 weeks from leaving date.”[171]I saw that page 604, (which was part of the “Driver Operating Guide” dated “August 2013” starting at page 591), this was said. “Hand Backs If you choose to hand back your vehicle it is unlikely that a spare vehicle will become available In less than 6-8 weeks.”[172]One aspect of the factual background which was significant was that after the introduction of 12-week agreements for the hire of vehicles from Eventech, the respondent provided financial incentives to drivers to renew those agreements. That was done by giving drivers discounts for the rental of the vehicle in question at differing rates depending on(1) the extent to which the driver had carried out jobs of the sort which the respondent had incentivised by the award of points which also led to improved chances of getting better-paid work, and(2) how many previous consecutive 12-week agreements the driver had entered into. It was not entirely clear to me precisely when that practice started. That was because the parties did not focus on this issue so I did the best I could with the voluminous documents before me. The first example in the bundle (taking its pages in sequence) of a document in which it was done was at pages 1591-1637, which were described in the index as “Driver Induction 2023” and merely dated “2023”. The rental credits and those points were shown at page 1605. There were several other documents in the bundle before me showing how rental credits could be earned, and for the sake of convenience I refer here in detail only to one of them. That was the document dated in the index as 13 April 2023 at pages 3078-3080. On page 3080 there were stated to be four “driver tiers”, of the following sorts (taking the bottom one first):172.1 “Yellow”, the “Qualifying Criteria” for which were “Entry Level”, and the reward for which was “No Rental Credit(s) Incentive” and “Bronze Collective Benefits for contract period”;172.2 “Bronze”, the qualifying criteria for which were (1) completing “250 Jobs in 12-week period”, (2) “0 Consecutive Agreements Signed”,(3) “Acceptance Rate of 90%, and(4) “Driver Star Rating of >4.50”; the reward for that was (1) “£150 rental credits at point of contractual renewal”, (2) “Bronze Collective Benefits for contract period”, and (3) Prebook Timings (2hr prior access)”;172.3 “Silver”, the qualifying criteria for which were (1) completing “340 Jobs in 12-week period”, (2) “1 Consecutive Agreements Signed”, (3) “Acceptance Rate of 92%”, and (4) “Driver Star Rating of >4.70”; the reward for that was (1) “£250 rental credits at point of contractual renewal”, (2) “Silver Collective Benefits for contract period”, and (3) “Prebook Timings (2hr prior access)”; and172.4 “Gold”, the qualifying criteria for which were (1) completing “400 Jobs in 12-week period”, (2) “2 Consecutive Agreements Signed”, (3) “Acceptance Rate of 95%”, and (4) “Driver Star Rating of >4.75”. The reward for that was (1) “£350 rental credits at point of contractual renewal”, (2) “Gold Collective Benefits for contract period”, and (3) “Prebook Timings (3hr prior access)”.[173]For the sake of convenience, I record here that those were not the only incentives for doing more for the respondent stated in the document at pages 3078-3080. There was also a weekly percentage bonus described on page 3079, which was stated to “increase the more you earn”, but that was, it appeared, not tied to the hire of a vehicle from Eventech, so it applied also to partner drivers. The manner in which courier drivers worked for the respondent in so far as it differed from the manner in which passenger drivers worked for the respondent[174]The manner in which courier drivers worked for the respondent was the subject of direct evidence from Mr Valentine. The flavour of his evidence in that regard was shown by what he said in paragraph 4 of his first witness statement, at page 355 of the WSB, which was this. “The Addison Lee courier business includes(i) pushbike,(ii) motorcycle,(iii) van couriers, and partner couriers (motorbike and van). The operational arrangements across each are quite similar and I will outline these at a high level below and throughout this statement.”[175]Mr Valentine then referred to pushbike and motorcycle couriers, even though neither of the test claimants either were, or had been, such couriers. Mr Valentine then, in paragraphs 25-27 of his first witness statement (at WSB pages 359-360) described in general terms the “van courier” operation of the respondent and then, in paragraphs 28-31 of that statement, described how van courier drivers worked. It appeared to me that what he said in paragraph 31 of that statement was intended to show the freedom of action of van courier drivers in regard to the taking of risks by not wearing “steel toe shoes (or any particular items of clothing)”. Since it has not been claimed that the test claimants were employees, given what was said by the Supreme Court in Uber v Aslam, to which I return in paragraphs 232-236 below, I found that paragraph (31) to be of at best only peripheral relevance.[176]What Mr Valentine said in paragraphs 28-30 of his witness statement was such as to show in my view that van courier drivers who hired their vans from Eventech were not in a materially different position from that of at least standard passenger drivers. Those paragraphs were as follows. “28. Van couriers hire their vehicles from Eventech and pay a rental cost in the same way as motorcycle couriers. 29. Van courier vehicles have always had Addison Lee livery and still do. However, the livery is much more discreet nowadays. 30. Similar to all other couriers, aside from an MDA device, we do not provide any equipment to van couriers nor have we done so since around 2017 or 2018. In the past we used to offer to supply the following equipment to van couriers:a. Ratchet straps and ropes to help them secure their parcels. This stopped around 2017.b. A sweatshirt with Addison Lee branding and other items of clothing. In the past we used to ask van couriers to wear these clothes because they looked smart, but it was never a requirement. We stopped giving them out during or around 2017 or 2018.c. An Addison Lee branded folder to keep dockets and bits and pieces. We stopped giving them out around 2017.d. Addison Lee branded tape. This isn’t around anymore, and I can’t remember when this stopped.”[177]The statement in paragraph 30b that “it was never a requirement” to wear the respondent’s clothing was inconsistent with what was said in the agreement for Mr Balog as part of the agreement dated “2018/01/05” for him at page 4781 (dated in the index as 5 January 2018), where it was said (in paragraph 3) that “Drivers are expected to be suitably dressed and must wear an Addison Lee jerkin, sweatshirt or T-shirt”). The same words were on page 4801, which was dated “2018/12/11”, page 4835, which was dated “2019/10/31”, page 4875, which was dated “2020/10/07”, and page 4918, which was dated “2021/07/03”. The final document chronologically in the series of the same documents relating to Mr Kidd was dated “2021/10/14” and was at page 5571. It too contained the words as part of paragraph 3: “Drivers are expected to be suitably dressed and must wear an Addison Lee jerkin, sweatshirt or T-shirt”. However, paragraph 77 of Mr Balog’s witness statement (at WSB page 15) showed that in practice he and other courier drivers were permitted to wear their own clothes and were only encouraged to wear the respondent’s “branded clothing”.[178]I therefore accepted what Mr Valentine said in paragraph 30 of his first witness statement. It was clear that the changes made after 2017 or 2018 which he described there were, not least because of what was said in paragraph 37 that witness statement, which I set out in the next paragraph below, made with a view to avoiding the impact of the tribunal and court decisions culminating in the Supreme Court’s decision in Uber. Again, though, as far as I could see, they were of at best peripheral relevance only, given that decision.[179]In paragraphs 35-90 of his first witness statement, at WSB pages 361-373, Mr Valentine described the manner in which the respondent provided work to van courier drivers. Paragraphs 35-43 had the heading “Courier Operational Arrangements”. I noted that Mr Valentine said in paragraph 36 that “[t]he operational arrangements for drivers and couriers are largely the same” although there were “some operational differences between drivers and couriers”. In paragraph 37 he said this. “The operation has evolved over time and, in particular from 2017, there was a drive by the business to ensure that we were not acting in a way which might jeopardise the self-employed status of our couriers (and drivers). This message was reinforced when new management joined in March 2020 which I will outline below. We have very actively relaxed the operation so that couriers have complete freedom and independence to choose when, where and how they provide their services to us.”[180]In fact, the self-employed status of drivers has never been in doubt. Only the question whether they were entitled to holiday pay and the national minimum wage has been in doubt. At the time of the hearing before me it was submitted on behalf of the claimants that the only real issue was when the claimants were working for those purposes, i.e. what were their working hours. That was said in paragraph 13 of the claimants’ opening submissions, which was in these terms. “The target of the Tribunal’s inquiry is – or ought to be – not whether the Claimants were ‘workers’ (they obviously were, for instance, to take the most straightforward situations, when they had agreed to perform a job for AL and were driving to that job and then driving the passenger or parcel to the agreed destination); but rather at what times and in what circumstances were they ‘workers’ – which is the same question as, what were their working hours.”[181]However, the position of partner drivers was not so simple (and I examine aspects of that position in paragraphs 193-205 below). For the sake of completeness, I now examine a little more closely the differences between courier drivers and passenger drivers.[182]I saw that Mr Valentine asserted in paragraphs 45.g and 45.h of his first witness statement, at WSB pages 364-365, that “couriers” have never been subject to a maximum break of 2 hours “before losing their empty bonus” and that while “[i]n the past a courier couldn’t use the [Going home] feature until 4pm”, “this changed around mid-2023 so a courier can use the feature at any time of the day and there is no limit to the number of times a day it can be used.” Of course, it was always possible for a courier to log off at any time and go home, since the relationship was not one of slavery. The issue here was what, if any, the impact of doing so would be. That was, said Mr Valentine in paragraph 45.h of his first witness statement, that “they would not receive preferential job allocation (in that they would not necessarily receive a job in the direction of their home) until that time of day.” He then said this: “We made this change to give couriers even more flexibility.”[183]It seemed to me that that was not the whole truth, as it was clear to me that the respondent made the change as part of its attempt to avoid the application of the Uber judgment, but the respondent’s motivation was in fact irrelevant, and so what Mr Valentine said in that sentence was in reality no more than argument.[184]In any event, what Mr Valentine said in paragraph 48 of his first witness statement was in my view inimical to the proposition that courier drivers were any less integrated into the operations of the respondent than passenger drivers. That paragraph was in these terms. “I think around 85-90% of courier jobs are allocated to couriers via Addison Lee’s auto allocation system. The remaining jobs are allocated to couriers by the operators. Operators manually allocate a higher percentage of courier jobs compared to passenger jobs. The main reason is because couriers can and are allocated multiple jobs at once. Whilst the auto allocator is programmed to allocate more than one job to a courier, the auto allocator has a series of rules it won’t break and when you are allocating multiple jobs to one courier it requires more juggling to maximise resource and service delivery.”[185]Paragraphs 50 and 51 of Mr Valentine’s first witness statement were also material, and they were as follows. “50. To manually allocate a courier job, an operator can just allocate it within the system to the courier they think is best placed for the job. The courier will receive an alert on their MDA in the same way they would if the job was auto allocated to them as I will explain below. An operator doesn’t need to call the courier so the courier wouldn’t know if the job had been auto allocated or manually allocated. However, on occasions an operator will call a courier to ask if they want the job before allocating it to them. This may happen if a long-distance job comes in late in the evening and the operator is aware that the courier will be travelling in that direction in the evening. The courier is free to refuse the job and if they do the operator will simply allocate the job to someone else. 51. The two types of jobs available to couriers are ‘pre-books’ and ‘ASAP’ jobs. They are largely the same as pre-book and ASAP jobs carried out by passenger drivers. I understand that Bill will explain what these are in his evidence and so I won’t repeat that.”[186]Mr Kidd’s absence was unavoidable, given what I say in paragraph 45 above, but as a result of it only Mr Balog gave oral evidence on behalf of the claimants about the role of a courier driver. Nothing which he said about that role was such as to justify the conclusion that the role of a courier driver was materially different for present purposes from that of a passenger driver.[187]In paragraphs 53-96 of Mr Valentine’s first witness statement, he described the arrangements under which courier drivers worked. I could not see in those paragraphs anything which justified, or required, a conclusion that courier drivers were in a materially different position from passenger drivers. I saw that in paragraph 32.4 of the claimants’ written closing submissions it was said that “[t]he operating practices which applied to courier drivers were different in some respects from those that applied to passenger drivers, as discussed below; but not materially for the purposes of the worker status and working time issues.” However, I did not see any discussion in those submissions about the differences between courier drivers and passenger drivers. That might well have been because the claimants themselves could see no material differences.[188]So, for the avoidance of doubt, for example, while the time of day when courier services were usually provided was more confined (as stated in paragraph 41 of Mr Valentine’s witness statement at page 363 of the WSB), that was not a material difference.[189]I should say that the factual assertion at the end of paragraph 70 of Mr Valentine’s first witness statement (at page 369 of the WSB) appeared to me to be wrongly made. That assertion was that there was in place “around mid-2024” an “un-allocate” button on the MDA or the respondent’s app which applied to all drivers. That assertion was made via the following two sentences. “Couriers have an ‘un-allocate’ button in the same way as passenger drivers. This feature was introduced in around mid-2024.”[190]The absence of any documentary evidence before me supporting the allegation in the first of those two sentences, especially given(1) the extent to which the respondent’s witnesses had made assertions of fact which were contradicted by documentary evidence before me, and(2) Mr Valentine’s own acceptance that he had given untrue evidence in relation to the document at page 8725, meant that I rather doubted that allegation. In paragraph 104 of their written closing submissions, the claimants said this (and I have already said, in paragraph 121 above, that I accepted that paragraph; I record here that I did not have a note of Mr Leiper asserting that there was a factual error in the paragraph): “PG said in evidence that in November 2021 a reject function was “in the pipeline for us” – see PG/37. When asked in x-ex what he meant by “in the pipeline”, PG corrected his evidence to, it was “being discussed”. However, AL has produced no disclosure reflecting any discussion or exploration of the introduction of a reject button.”[191]If I had had to make a determination on the accuracy of the second of the two sentences from paragraph 70 of Mr Valentine’s first witness statement which I have set out at the end of paragraph 189 above, then I would have concluded that it was not accurate. However, I did not need to make such a determination in relation to the test claimants because the two courier driver test claimants were not working for the respondent in 2024. It was, though, likely to be helpful for the purpose of the resolution of disputes about the status of other claimants to whom the statement would have been relevant to record my current conclusion, which was as stated in the first sentence of this paragraph. Having said that, the impact of the existence or otherwise of an “un-allocate” button was (given what I say in paragraphs 274 and 275 below) in my view minimal.[192]I turn now to the situation of partner drivers generally, and the evidence relating to the manner in which their relationship with the respondent compared with that of the drivers who hired their vehicles from Eventech. Additional factual findings concerning the relationship between the respondent and partner drivers Sanctions[193]The first aspect of that evidence to which it is necessary to refer was of a similarity rather than a difference. That similarity related to the possibility of sanctions being applied to partner drivers. At the time of the start of the first Covid-19 lockdown, and at the end of the period of ownership of the Carlyle Group, the emails at pages 3277-3276 were sent. No reference was made to them in the respondent’s closing submissions. The claimants referred to them in paragraph 200 of their written closing submissions, in the following way. ‘BK’s email exchange with Mr Chadwick and Mr Monk of 12 March 2020 [3277-3276] suggesting that a “3 strike” rule for owners refusing jobs might be replaced by a “one strike” rule, demonstrates clearly that owner drivers continued to be sanctioned for refusing jobs.’[194]The complete sequence of emails was as follows.194.1 On 12 March 2020, Mr Kelly wrote, under the subject “Owners”, to Mr John Chadwick and Mr Gary Monk: ‘Hi lads, Owners refusing jobs, we have discussed this previously and introduced a “3 strike” rule” but in the current climate think it needs to be “one strike”’194.2 Mr Monk replied later on that day: “Chaps I think we need to send out a specific comms to owner drivers, the flavour of which should be: - there is less work available for all our valued drivers - the reality is we will not be able to offer as much work to you as you have recently enjoyed, because understandably people are not travelling as much - we are doing a number of things to stimulate demand wherever we can - we will inevitably be focussing the allocation of the work available to our most loyal and best serving partners - you can help us by continuing to log onto your mdas especially in peak periods - it’s really important that we provide the best service we possibly can during this time and you can do that by arriving on time and not rejecting jobs when they are allocated to you - we will provide you with regular updates as our work volumes become clearer over the coming days and weeks I do think we have to be tough with non performant drivers and those that don’t really do work for us. The reality is we haven’t yet stopped using these drivers, but we have to protect our self employed and most loyal owners. Thoughts?”194.3 Mr Chadwick replied: “Gary Will have something ready for review before lunch on this that’s clear and honest for this population base. I am happy to be tough on the owner drivers not supporting us or refusing jobs with bans. I want to avoid exiting a large number of drivers now and if/when we recover after May we need to start asking these drivers to come back. I see these drivers as similar to on demand staff, we switch off the work but don’t terminate the relationship. Let’s look at wider restrictions for this base which we can 100% support to defend our SE guys.”[195]The reference to “SE guys” was plainly to the non-partner drivers, who were selfemployed, although of course the partner drivers plainly were also selfemployed. In any event, that sequence of emails showed that in March 2020 the respondent had applied sanctions to partner drivers, and was thinking of applying more. The fact that the application of sanctions was likely to be counterproductive will have become more important after the end of the Covid-19 lockdowns, but the possibility that the use of partner drivers made it more difficult to give sufficient work to non-partner drivers in order to be able to make it worth their while to remain non-partner drivers, may have been a countervailing factor.[196]In the circumstances, I found the possibility of sanctions being applied to partner drivers to be of little relevance in deciding whether or not they were limb b workers. Multi-apping[197]However, there was a major difference between the position of a partner driver and a driver who hired his vehicle from Eventech. That was the fact that the respondent did not either discourage or make it impossible in practice for a partner driver to multi-app. I return to the impact of multi-apping in stating my reasons for my conclusions on the status question(s) in paragraph 276 below. I now state my reasons for saying that the position of a partner driver was materially different from that of a non-partner driver in regard to multi-apping. Partner and non-partner courier drivers[198]It is convenient to take what Mr Valentine said about partner courier drivers first. In paragraphs 114-118 of his first witness statement, he said this. “114. I am not aware whether or not any particular courier has provided courier services to any other company. We do not ask this question of our couriers and don’t look to prevent them working for other companies in any way. 115. As touched on above, I am aware that some couriers, especially partner motorbike and partner van couriers, do “multi-app”, meaning that they will be signed up to drive for many operators at the same time. At any given time they will have multiple operator applications open and so they pick and choose the best job across the multiple operator applications. 116 This allows them to be offered a greater selection of jobs at any time and they will pick and choose which operator to work for, generally on a job by job basis depending on which job they consider is best. The best job will almost certainly be the one that pays the most, so a longdistance delivery, but it might be a job in a certain direction if they want to travel home, for example.117. It is also very common for partner motorbike couriers to deliver both packages for Addison Lee but also other operators such as fast food deliveries for Deliveroo. As I mentioned, a lot will deliver for Addison Lee in the day and by night they will deliver fast food packages for Deliveroo or another operator. I suspect some do it side by side (and that’s not an issue for Addison Lee).118. If any courier does any other work, of any kind, we have no way of knowing that and we simply wouldn’t mind. Our contracts are clear that there are no restrictions on the driver’s ability to work for other companies and we reminded drivers of this as part of the roll out of the new courier deal 2022 see for example [HB/3044]. As explained by Bill in his evidence, we understand that couriers value the flexibility of their independent, self-employed status and all our efforts are around helping couriers retain this flexibility while having the best earning opportunity.”[199]It appeared to me that the first two of those paragraphs (114 and 115) were contradictory. In addition, I looked in vain at page 3044 and the next few pages for a statement to the effect asserted in paragraph 118 of that passage. I therefore doubted what Mr Valentine said in paragraphs 114 and 115 about nonpartner drivers multi-apping.[200]A further reason for doubting the accuracy of the assertion that a non-partner courier driver was in practice free to use an Eventech vehicle for carrying other courier services’ parcels was that the insurance contract for Mr Kidd at pages 5585-5594 (dated 14 October 2021) provided in paragraph 2c of Schedule 1 on page 5594 this. “You are only entitled to benefit from the Relevant Insurance and the Trust insofar as you use the Vehicle for business arranged by or carried out on behalf of Addison Lee or pursuant to your Driver Contract, or for private, nonbusiness use. No other business use is permitted and is strictly prohibited. In the event that you use the Vehicle for other business purposes, you will be fully liable for any claims arising, liable to prosecution under the Road Traffic Acts and liable to have your public carriage office licence revoked”.[201]The same was true of the agreements in the name of Mr Balog, the relevant parts of which were at pages 4792, 4826, 4866, and 4909.[202]There was a further factor that was relevant here. That was that the “Better Courier Deal” introduced in July 2022 as shown by the document whose first page was page 3044 made it very much in the financial interests of the driver to do as much work for the respondent as possible. That was clear from the rewards which were to be gained by doing jobs which attracted points, as shown by what was said on pages 3049 and 3050.[203]So, in my view, the only courier drivers who were able in practice, or realistically, to benefit from multi-apping, were partner drivers. Partner and non-partner passenger drivers[204]Similarly, standard passenger drivers who hired their vehicles from Eventech could not use their vehicles to transport passengers who booked via another provider (such as Uber) unless they had(1) obtained their own insurance to do so (see paragraphs 165 and 166 above), and(2) removed the respondent’s livery, or branding (see paragraphs 61 and 61 above), with the result that multiapping was not in reality an option open to them. While the second of those two restrictions did not apply to non-partner executive drivers, i.e. those executive drivers who hired their cars from Eventech, the respondent’s agreements with non-partner executive drivers had the same financial effect as that which I describe in paragraph 202 above.[205]As a result, I concluded that it was in practice only open to a partner driver to multi-app. The periods when the test claimants provided driving services to the respondent[206]In paragraph 269 of its written closing submissions, the respondent made submissions about the factual positions of the test claimants, basing them on the documentary evidence before the tribunal and taking into account what was in the claimants’ solicitors’ schedules at pages 4977-4983, 5378-5382, 5632-5635, 5842-5845, 6267-6269, 6639-6643, 6994-6999, 7828-7834, and 8521-8527.[207]The claimants submitted in paragraph 324 of their written closing submissions that what was said in paragraph 266.b of the respondent’s written submissions (which is set out in paragraph 167 above as part of the whole of paragraphs 266- 268 of the respondent’s written closing submissions) was unfair. That was said on the following basis. ‘The summaries for each individual TC [i.e. test claimant] were included in the agreed bundle without any objection from AL and Leigh Day had no reason to suppose that AL would criticise the summaries as “unsatisfactory” until the commencement of this hearing.’[208]This was then said (in paragraph 325): “Further, in circumstances where AL failed to recognise the claimants’ right to paid holiday for many years, and failed to facilitate arrangements for them to take the leave to which they were entitled, it is unsurprising that individual TCs now find it difficult to recall the individual reasons for taking specific breaks.”[209]That was in my view a fair submission on the facts, but it did not answer sufficiently the proposition that it was for the claimants to prove their cases, and if there was a dearth of evidence on a particular relevant aspect of those cases, then that aspect was likely to be unsuccessful.[210]The claimants’ submissions in paragraph 326 of their written closing submissions were more apt and therefore more helpful. What they said there was as follows. “Turning to the criteria proposed at R/268 for determining whether a break should be regarded as a termination: 326.1. The LDCs suggest that a break of more than 4 weeks in which an individual undertook other work or was prevented from working by loss of their PCO or driving licence would normally amount to a termination. 326.2. The LDCs do not agree that breaks during which the TCs did not work because of the Covid pandemic gave rise to any termination. Where they occurred, such breaks were wholly involuntary and understood by both parties to involve a temporary cessation of work. 326.3. The LDCs do not agree that any break during which the driver returned his vehicle and MDA should be treated as a termination, as proposed at R/268b. The Tribunal has heard that some of the TCs would return their vehicles when they went on holiday in order to avoid incurring rental charges during their absence. It would be absurd if simply taking a holiday were to be regarded as a termination. 326.4. The suggestion at R/268c that “any break between contracts” (by implication, however short) should be regarded as a termination would be wholly incompatible with the principle of effectiveness. 326.5. As to 268d, it is accepted that a “protracted break” may well give rise to a termination, but the reason why a driver ceases working may be a relevant consideration in determining whether this is so.’[211]I found it helpful to consider those submissions against the background of the evidence before me about the manner in which the test claimants actually worked, and the periods when they did not work, as drivers for the respondent. I therefore return to those submissions and the respondent’s submissions on the issues of principle in paragraph 285 below, after considering the evidence before me about the way in which the test claimants worked and had breaks.[212]The claimants, in paragraphs 327-333 of their written closing submissions, made submissions in regard to the breaks in service of the test claimants, responding to the respondent’s submissions in paragraph 269 of its written closing submissions. Mr Leiper made oral submissions in response to the claimants’ submissions. He submitted that the respondent would not know why a driver was returning a vehicle unless the driver gave a reason, but in any event, the respondent could not know when the driver was going to return at that point. It was the respondent’s position (and I have summarised it in my own words, on re-reading my notes of Mr Leiper’s oral submissions) that the contract between the respondent and the driver terminated automatically on the return of the vehicle leased from Eventech because once that vehicle was returned, the driver could not in practice work under the driver agreement. I was asked by Mr Leiper to make findings in relation to the test claimants on the basis that the issue was one of principle.[213]I had considerable difficulty in seeing on what basis I could make a finding on what amounted to a question of mixed fact and law about the breaks between periods of driving for the respondent under a driver’s agreement where the claimants themselves had put no evidence before me about the reasons for those breaks. On the other hand, the respondent argued in relation to holiday periods that the test claimants had taken ample periods of holiday simply because they had had frequent or extended breaks between periods of driving so that they were not entitled to claim that they had not been given an opportunity to take holiday. That argument involved accepting that at least some breaks between periods of driving were to be regarded as holiday periods, and that meant that the respondent could not at the same time argue (as it did in paragraph 268.c of its written closing submissions) that any break between contracts was sufficient to “amount to a termination and/or a break in a series of deductions”.[214]In relation to what the claimants put in paragraph 326.2 of their written closing submissions, which I have set out in paragraph 210 above, Mr Leiper said that the respondent did not know from where (or what) the claimants had got the idea that it was accepted by the respondent that there was an agreed temporary cessation of work during the Covid-19 lockdown periods, or during the period when the government’s response to the pandemic had the effect of inhibiting movement and therefore of affecting the viability of the service provided by the respondent through drivers such as the test claimants. That was because214.1 the periods of inactivity as a driver differed for each test claimant;214.2 some sought and obtained a grant from the government as a selfemployed driver, and others did not: it was a matter for them whether or not they did; and214.3 as was the case generally in regard to absences of drivers, the respondent did not and could not know whether, and if so when, the driver in question might, or would, return to driving for the respondent.[215]As for the impact so far as this case was concerned of the pandemic on any driver who was at the time over the age of 65 (and that was the case for Mr Nardelli), Mr Leiper said that that there was only an understanding that it was not safe for someone to work over the age 65, and that there was nothing stopping a person over the age of 65 from working. The position of Mr Nardelli[216]In fact, the witness statement of Mr Nardelli contained a particularly good illustration of the difficulty of deciding to what extent a claimant should as a matter of mixed fact and law be regarded (i.e. in this instance found by me) to have been in a continuous relationship with the respondent. That was in the following passage (at pages 64-65 of the WSB). “My Working Pattern40. During the time I worked for Addison Lee, I worked only for them. I usually worked at least 6 days per week, taking only 1 day off, but sometimes would work seven days a week. I tended to work around 12 to 14 hours a day, except at weekends when I worked in the region of 16 hours. I tended to work Friday, Saturday and Sunday.41. I sometimes took the odd few days or a week off to visit family or to go away, most typically to Spain, in the summer, but otherwise I worked quite intensively and did not go out very much in terms of socialising; I could not go out drinking, for example, because I would always have to drive.42. I worked in this way mainly so that I could enjoy longer breaks in the winter instead and escape the cold, wet British weather by heading off to Thailand to relax after months of working hard. I usually worked until Christmas and then took a break of about 5 or 6 weeks during January and February. I would usually either book a flight leaving on Boxing Day or wait until after the New Year from mid-January when flights were cheaper. This pattern changed slightly about 7 or 8 years ago after I had reached retirement age, at which point I tended to leave mid-November and return after 12 January, again because flights were cheapest then. Also, in the run-up to Christmas there was often so much traffic on the roads that jobs took too long to do, and I found that it was not really worth working at this time from a financial point of view. On a few occasions I took longer periods off at other times of the year too.43. I cannot recall exactly and have no records of the times when I went away, but my solicitor has shown me copies of a couple of emails I sent to Addison Lee to notify them that I was going away, for example for 3 weeks from around 7 to 28 June and again for 3 weeks from 15 September 2019 (see pages 8374–8375). I have also been shown me [sic] a copy of a schedule which shows periods of time when I was not working, which I believe correspond with the times when I travelled abroad although this only shows details from April 2014 (see page 8521).44. On one occasion, I recall that I decided to go away as I was unable to work due to an issue with my PCO licence. My driver log records that I was banned for 1 month from 18 October 2016 in relation to my PCO licence. The schedule shows that I took a break of 3 months from the week ended 28 October 2016.45. When I first started working for Addison Lee, drivers were expected to inform Addison Lee when they were taking holidays of more than 2 weeks (see page XX), although I used to tell them as a matter of courtesy and I would usually tell them on the day I returned the vehicle. However, if I went away for 2 weeks or so, up to a month, I kept the vehicle and so I did not need to let them know. If it had been more than a month before I handed it back, I had to provide a reason why I could not work otherwise I was not sure of being allocated a vehicle upon my return. If I handed back the vehicle, I also handed back the XDA. In the email examples above, I informed Addison Lee that I would be away because I needed to reschedule the servicing of my vehicle.” Mr Balog’s holidays[217]Mr Balog’s witness statement contained (in paragraphs 128-134 at pages 23-24 of the WSB) a description of his holidays which was to the effect that he took as little holiday as possible and a statement (in paragraph 133) that he took “around four weeks holiday each year”. However, Mr Leiper put it to him in crossexamination that if one looked at document 709, which was a spreadsheet containing data exported from the Shamrock software records drawn from the mobile telephone data relating to the jobs done by Mr Balog at the end of his period of working for the respondent as a driver (in fact as a courier driver), then it appeared that between 15 July 2021 and Mr Balog’s last day of service as a driver, which was 27 August 2021, there was “a total of 17 clear days of not working across 4 separate breaks”. Mr Balog said that he could not remember what had happened during that period and I said that I was not sure how reliable his recollection could be in the absence of contemporaneous records of what he received by way of pay at the time. However, as I record in paragraph 100 above, the claimants accepted through Mr Segal that documents such as the spreadsheet which was document number 709 were accurate. Mr Segal did correctly point out, however, in closing submissions, that it was not surprising that there were as many as 17 days of not working in a six week period, since 12 of the days in that period would have been weekend days.[218]For the sake of completeness, the position of the other test claimants was as Mr Ban’s holidays[219]Mr Ban only worked for the respondent from January to April 2024, and he said nothing in his witness statement about the taking of holidays. Mr Da Silva’s holidays[220]Mr Da Silva’s witness statement contained relatively little about holidays (what was there was in paragraphs 87-91 on pages 40-41 of the WSB), but the table prepared on behalf of the claimants at page 8845 was more precise about the breaks which the test claimants took. While Mr Leiper cross-examined Mr Da Silva fairly closely on the time that Mr Da Silva took off by way of holidays, there was reference in the row of that table relating to Mr Da Silva (row 2) to a fourweek break (which Mr Da Silva said was spent in Brazil) in August to September 2015, and otherwise to shorter holiday breaks. Mr Ruiz’s holidays[221]Mr Ruiz said in paragraph 38 of his witness statement (at WSB page 84) that he “usually took holiday for between 2 and 6 weeks each year, spread out over a couple of periods each year”, and that he “also had three longer breaks.” Those were(1) in December 2014, when he “stopped working for Addison Lee for 6 months to work as a delivery driver”,(2) in 2020, when, he said, due “to the Covid pandemic”, he stopped working for the respondent for 10 months, and (3), after returning to work for the respondent for a short period, he “stopped working for 11 months in 2021.” However, he was able to say this in paragraph 36 of that statement, on the same page, which was (by reference) more specific about the times when he worked and when he was absent. “My solicitor has shown me a spreadsheet which I understand contains the hours that I worked between 2014 and 2024, the pay that I received and any breaks or holidays I had in my employment. I cannot recall the specifics, but the data contained in the spreadsheet seems broadly accurate to me.”[222]In cross-examination Mr Ruiz was able to agree that the spreadsheet in question was at pages 6639 onwards. Mr Edah-Tally’s holidays[223]In paragraph 89 of his witness statement, at page 110 of the WSB, Mr EdahTally said this. “I did not go abroad or take holidays very often, perhaps once a year in July or August, although there were a couple of years when I took two short holidays. My legal representative has shown me a schedule showing that I took 1 week in April 2018, August 2021 and July 2023.”[224]In oral evidence, he clarified that that “schedule” was the document starting at page 7828. Mr Edah-Tally usually returned his vehicle when he went on holiday, as he said in paragraph 85 of his witness statement, at page 110 of the WSB. That was because “[financial] deductions would continue to be made, which [he] did not think made financial sense”. His further evidence about that was in paragraphs 86-88 of his witness statement, which were also on that page. There, he said this. “86. After the Covid pandemic in 2020 or 2021, Addison Lee allowed drivers who were taking a holiday to keep the car and pay a reduced amount, which consisted of paying the insurance only, but I did not do that either. If I went abroad, I usually returned the vehicle to the Addison Lee garage in West Drayton and took a taxi to the airport. When I returned, I went to the garage directly, signed a new contract and drove away in one of the Addison Lee vehicles.87. My understanding was that these deductions for holiday only applied to breaks of up to 2 weeks although I don’t have a recollection of being notified of a limit.88. When I returned my vehicle, I normally told Addison Lee how long I would not be using it, so that I would have a better chance of a vehicle being available upon my return.” Mr Payne’s holidays[225]In paragraph 82 of his witness statement (at page 142 of the WSB), Mr Payne said that he “usually took a week’s holiday once a year” and that he usually did not take any more time off because he “could not afford it”. In cross-examination, he was taken to the document starting at page 6267, which had been prepared by his solicitors. In broad terms, it was consistent with what he said in paragraph 7 of his witness statement about the periods when he worked as a driver for the respondent. Mr Kidd’s holidays[226]Mr Kidd said in paragraph 51 of his witness statement (at page 154 of the WSB) that he “usually went on holiday for around 2 weeks each year” and that during that period he usually kept his van. He said that that was because he did not know whether he would be allocated a van as soon as he returned from holiday and he needed to work “without any interruption”. He continued: “For example, in 2021, I had one week’s holiday in July and one in September (see page 5632).” There was at page 5632 a document of the same sort as for example that which was at page 6267. Mr Mahendran’s holidays[227]In paragraph 24 of his witness statement (at page 164 of the WSB), Mr Mahendran said this. “When I went on holiday for a week or two, I kept my car and continued to pay the insurance. During the weeks in which I took holiday, I had accrued enough points to not have deductions taken by Addison Lee for the cost of the car. I explain the points system at paragraph 26 below. During these periods I usually stayed at home or went to Germany on holiday to visit family. I would usually only take one or two weeks off at a time, between once and four times a year.”[228]Mr Mahendran’s absences over the years were shown in the document starting at page 6994, which the claimant’s solicitors had prepared, and the respondent did not contend that it was not accurate. Mr Klepacki’s holidays[229]Mr Klepacki’s evidence about the holidays which he took was (in addition to paragraph 58, to which I refer in paragraphs 169-171 above) in paragraphs 59- 62 of his witness statement, which bear repeating here. “59. If I was on holiday for just a couple of days or so, I normally kept the vehicle but returned it if longer than a couple of weeks.60. I usually took regular holidays at Christmas, when I would go back to Poland for a couple of weeks to see family, a week or so at Easter and in the summer. I also sometimes took some time off during the school half terms as I found that work was usually very quiet then.61. I have been shown some data by my solicitor which shows periods of time when I was not working according to Addison Lee’s data. It shows that I usually took between 1 and 2 weeks off 4 or 5 times a year and usually took 5 weeks or so off in total. Having seen this data, I agree that the periods shown are likely to have been periods that I would have taken holidays because they are during the summer, Christmas and Easter or half-term periods. However, I cannot remember exactly where I went or what I did.62. I can see from the data that it shows that in August 2017 I took three weeks off and that in August 2018 I took a time off to go to Poland and in September of the same year I went to Spain, for just a few days respectively; however, I believe that during August and September 2018 I worked for Wheely and Blacklane rather than Addison Lee.”[230]In cross-examination Mr Klepacki agreed that the document starting at page 5842 was what he meant by referring to “some data” which he was shown by his solicitor. He could not recall what he was doing230.1 in what looked from rows 10232 and 10233 of the spreadsheet which was document 877 to have been a break between jobs done for the respondent on 27 July and 4 September 2018, and230.2 between 7 September and 16 October 2018, when, it appeared from rows 10235 and 10236, he did no work for the respondent. The relevant statutory provisions and case law[231]The parties’ written submissions referred to the applicable statutory provisions and case law in some detail and comprehensively. Rather than do the same thing here, it was in my view sufficient (and rather better, in an already-long set of reasons) in regard to some of the issues to refer to no law here, and instead to refer to the relevant law when stating my conclusions. In respect of other issues, it was in my view sufficient to refer only relatively briefly here to the statutory provisions and case law, and to refer in more detail to the relevant law when stating my reasons for my conclusions on those issues below, in the next (and final) section of my reasons for this reserved judgment. In the following paragraphs, I first focus on the key passages in the case law concerning worker status. I then refer to the most knotty problems raised by the parties and state my conclusions on the issues of law which they raise. I should make it clear that I was informed by the respondent that it had informed the Attorney General of the claim of the claimants that the introduction of a two-year backstop for claims for unpaid wages was ultra vires. That was done in paragraph 3 of the respondent’s written closing submissions, which was in these terms. “Amongst these issues is the question of the lawfulness of secondary legislation. In relation to this, the Respondent notified the Attorney General of the argument. The Government Legal Department has confirmed that the Department for Business and Trade will not intervene, but may wish to do so in the event that the issue is subject to an appeal.” Status; limb b worker or not?[232]The key words in for example section 230(3)(b) of the ERA 1996 are “undertakes to do or perform personally any work or services for [the claimed employer, using that term loosely]”. As Lord Leggatt JSC, giving what was in effect the judgment of the Supreme Court, said in paragraph 69 of Uber, the primary question here was “one of statutory interpretation, not contractual interpretation”. As he then said in paragraph 70 of the judgment, in the course of that exercise of interpretation, it is necessary to have regard to “the purpose of the statutory provision [in question] and to interpret its language, so far as possible, in the way which best gives effect to that purpose.” In the course of doing that, “if, for example, a fact is of no relevance to the application of the statute construed in the light of its purpose, it can be disregarded.” Then, at the end of paragraph 70, quoting a statement of Ribeiro PJ in Collector of Stamp Revenue v Arrowtown Assets Ltd (2003) 6 ITLR 454, paragraph 35, he said this: “The ultimate question is whether the relevant statutory provisions, construed purposively, were intended to apply to the transaction, viewed realistically.”[233]Here, I found the following statements and passages in Uber to be of particular relevance. Where I refer to a passage, I do not set it out. Where I set out a statement, I give the paragraph number and the place where the statement is in the report at [2021] ICR 657.233.1 “While not necessarily connoting subordination, integration into the business of the person to whom personal services are provided and the inability to market those services to anyone else give rise to dependency on a particular relationship which may also render an individual vulnerable to exploitation.” (Paragraph 74, at [2021 ICR 676-677.)233.2 “The correlative of the subordination and/or dependency of employees and workers in a similar position to employees is control exercised by the employer over their working conditions and remuneration.” (Paragraph 75, at [2021 ICR 677.)233.3 Paragraphs 76-86.233.4 ‘As noted earlier, the vulnerabilities of workers which create the need for statutory protection are subordination to and dependence upon another person in relation to the work done. As also discussed, a touchstone of such subordination and dependence is (as has long been recognised in employment law) the degree of control exercised by the putative employer over the work or services performed by the individual concerned. The greater the extent of such control, the stronger the case for classifying the individual as a “worker” who is employed under a “worker’s contract”.’ (Paragraph 87, quoting what Baroness Hale DPSC said in “the Bates van Winkelhof case [2014] ICR 730.)233.5 Paragraphs 90-91.233.6 “A particularly important consideration is who determines the price charged to the passenger. More generally, it is necessary to consider who is responsible for defining and delivering the service provided to passengers. A further and related factor is the extent to which the arrangements with passengers afford drivers the potential to market their own services and develop their own independent business.” (Paragraph 92, at [2021] ICR 682B.)233.7 “First and of major importance, the remuneration paid to drivers for the work they do is fixed by Uber and the drivers have no say in it (other than by choosing when and how much to work).” (The first sentence of paragraph 94, at [2021] ICR 682. The whole of that paragraph is relevant, but that sentence stood out to me here.”233.8 Paragraphs 95-102. Working time, including when logged onto more than one app at a time[234]The Supreme Court’s decision in Uber was relevant in this regard also. Two passages were relevant here. The first was paragraphs 121-130 at [2021] ICR 689-691. The final paragraph in that passage was of particular importance here: ‘It follows that the employment tribunal was, in my view, entitled to conclude that, by logging onto the Uber app in London, a claimant driver came within the definition of a “worker” by entering into a contract with Uber London whereby he undertook to perform driving services for Uber London. I do not consider that the third condition identified by the tribunal that the driver was in fact ready and willing to accept trips can properly be regarded as essential to the existence of a worker’s contract; nor indeed did the tribunal assert that it was. But it is reasonable to treat it, as the tribunal did, as a further condition which must be satisfied in order to find that a driver is “working” under such a contract.’[235]That passage applied to the situation where the driver was driving exclusively for Uber. Paragraphs 121-130 were, potentially confusingly, entitled “Working time”, and at the end of them, in paragraph 131, Lord Leggatt said: “This brings me to the question of what periods during which a driver is employed under a worker s contract count as working time.”[236]He then, under the heading “The Working Time Regulations”, referred to the WTR 1998’s definition of “working time”, which is in regulation 2(1) and in relation to a “worker” is “any period during which he is working, at his employer’s disposal and carrying out his activity or duties”. The next five paragraphs are of considerable importance here and are as follows. ‘133. There is no difficulty in principle in a finding that time when a driver is “on call” falls within this definition. A number of decisions of the CJEU establish that, for the purpose of the Working Time Directive, to which the UK Regulations aim to give effect and which defines “working time” in the same way, time spent on call counts as “working time” if the worker is required to be at or near his or her place of work. For example, in Ville de Nivelles v Matzak (Case C-518/15) EU:C:2018:82 ; [2018] ICR 869 the CJEU held that time spent by firefighters on stand-by at their homes, which were required to be within eight minutes travelling distance of the fire station, was working time.134. On the facts of the present case, a driver’s place of work is wherever his vehicle is currently located. Subject to the point I consider next, in the light of this case law the tribunal was justified in finding that all time spent by a driver working under a worker’s contract with Uber London, including time spent “on duty” logged onto the Uber app in London available to accept a trip request, is “working time” within the meaning of the Working Time Directive and Regulations.135. The point that – like the majority of the Court of Appeal and Judge Eady QC in the Employment Appeal Tribunal – I have found more difficult is whether a driver logged onto the Uber app in the area in which he is licensed to work can be said to be “working, at his employer’s disposal and carrying out his activity or duties” if during such times the driver is equally ready and willing to accept a trip request received from another PHV operator. It was argued with force by counsel for Uber that a driver cannot reasonably be said to be working for and at the disposal of Uber London if, while logged onto the Uber app, he is also at the same time logged onto another app provided by a competitor of Uber which operates a similar service.136. I have concluded that this question cannot be answered in the abstract. I agree with Judge Eady QC when she said in her judgment dismissing Uber’s appeal to the Employment Appeal Tribunal (at para 126) that it is a matter of fact and degree. Like the majority of the Court of Appeal, I also agree with her that: “If the reality is that Uber’s market share in London is such that its drivers are, in practical terms, unable to hold themselves out as available to any other PHV operator, then, as a matter of fact, [when they have the Uber app switched on] they are working at [Uber London’s] disposal as part of the pool of drivers it requires to be available within the territory at any one time. … if, however, it is genuinely the case that drivers are able to also hold themselves out as at the disposal of other PHV operators when waiting for a trip, the same analysis would not apply.”137. So far as this court has been shown, no evidence was adduced at the hearing in the employment tribunal in 2016 that there was at that time any other app-based PHV transportation service operating in London or that drivers logged into the Uber app were as a matter of practical reality also able to hold themselves out as at the disposal of other PHV operators when waiting for a trip. No finding was made by the tribunal on this subject. In these circumstances I do not consider that the tribunal was wrong to find that periods during which its three conditions were met constituted “working time” for the purpose of the Working Time Regulations 1998.’ Holiday pay issues[237]The calculation of holiday pay here, assuming that the claimants were limb b workers and that they were working for the purposes of the WTR when they were logged on but not driving as well as when they were driving or waiting to pick up a passenger, having accepted a job, was the subject of paragraph 205-338 of the claimants’ written closing submissions. Those paragraphs raised the following questions. The need for effective judicial protection of the claimants’ rights under the Working Time Directive[238]After referring in paragraph 206 to the principle “that national procedural rules should not have the effect of making it impossible in practice or excessively difficult to exercise rights guaranteed by EU law”, and “the principle that national procedural rules should not make the enforcement of EU rights more difficult than the enforcement of analogous national rights i.e. they should not discriminate between claims which derive from national law and those which derive from EU law (the principle of equivalence): Levez v TH Jennings (Harlow Pools) Ltd (Case C-325/96) [1999] ICR 521 at [18] – [19]; Chief Constable of Northern Ireland v Agnew [2023] UKSC 33, [2024] ICR 51 at [50] – [78]”, the claimants, in paragraphs 207 and 208 of their written closing submissions referred to “King” and “Kreuziger”. Those cases were not cited in full in the written closing submissions. The first of those was King v Sash Window Workshop, European Court of Justice (“ECJ”) [2018] ICR 693. The second was Kreuziger v Land Berlin, ECJ, [2019] I CMLR 36.[239]In paragraph 209 of their closing submissions, the claimants referred to “Smith”. Again, that case was not referred to by its full name in the closing submissions. It was the decision of the Court of Appeal in Smith v Pimlico Plumbers Ltd [2022] ICR 818. Paragraphs 207-210 of the claimants’ closing submissions were a helpful summary of the effect of those cases, with which I did not understand the respondents to disagree and which in any event appeared to me to be apt. Those paragraphs were as follows. ‘207. In King the CJEU held that, where a worker has failed to exercise his right to annual leave over several consecutive reference periods because his employer disputes his entitlement to paid holiday, the WTD requires that the worker be permitted to carry over and accumulate paid annual leave rights until the termination of his employment. Having pointed out that the WTD treats the right to take annual leave and to receive payment for such leave as two aspects of a single right, the Court decided that a national rule which required the worker to take his leave before first establishing whether he had a right to be paid in respect of that leave would be incompatible with the right to an effective remedy. The Court observed there was no policy reason why the worker should not be entitled to carry over his untaken leave indefinitely, because(i) not granting paid annual leave is “liable to dissuade the worker from taking annual leave”;(ii) a worker’s failure to take the leave to which he is entitled is to his employer’s benefit; and(iii) it is for the employer to make a proper assessment of its obligations and, if it fails to allow the worker to take the leave to which he is entitled, it has to “bear the consequences”. 208. In Kreuziger the CJEU held that the right to paid annual leave is of such fundamental importance that, even where the right itself is not disputed, a worker must be able to carry over the leave to which he is entitled unless the employer can show that it has given him the opportunity to take that leave, by encouraging him to do so and informing him, accurately and in good time, that if he does not take his leave, it will be lost at the end of the reference period or carry-over period, or upon termination of the employment relationship where the termination occurs during such a period: see paras 52 – 53 of the Court’s judgment. The Court observed that this conclusion was supported by the weaker position of the worker in the employment relationship, which made it necessary to prevent the employer being in a position to impose a restriction on his rights, and by the principle that incentives for a worker not to take leave or discouragement from doing [so] were incompatible with the objective of ensuring that every worker enjoys a period of actual rest. 209. In Smith the Court of Appeal held that the effect of the rulings in King and Kreuziger is that a worker who has been wrongly classified as self-employed is entitled to carry over the right to annual leave from one year to the next until the termination of his employment and to receive a payment in lieu thereof, regardless of whether he has been prevented from taking his annual leave or only permitted to take unpaid leave: see paras 77 – 86 of the judgment of Simler LJ, with whom Lady Justice Laing and Lady Justice King agreed. 210. As noted above, the Court of Appeal held that it was possible to interpret the WTR so that they gave effect to the CJEU’s judgments in King and Kreuziger by reading words into the WTR. The words read in, which were set out in an Appendix to the Court’s judgment, included the following addition to the words of reg 13: “(16) Where in any leave year an employer (i) fails to recognise a worker’s right to paid annual leave and (ii) cannot show that it provides a facility for the taking of such leave, the worker shall be entitled to carry forward any leave which is taken but unpaid, and/or which is not taken, into subsequent leave years.”’[240]The respondent’s response to that passage, in its written closing submissions, was that it accepted (in paragraph 188) that the decision of the Court of Appeal in Smith is binding on an employment tribunal, but (paragraph 189) that240.1 “by virtue of Protocol No 30 on the Application of the Charter of Fundamental Rights of the European Union to Poland and the United Kingdom, European law in relation to holiday and holiday pay (including Article 7 Working Time Directive) does not in any respect relevant to this claim have direct effect against the Respondent, a private party”, and240.2 alternatively any such direct effect against the Respondent (which is a private party and not an emanation of the state) is by virtue of Section 5 of the European Union (Withdrawal) Act 2018 (and Schedule 8 to that Act) inapplicable to(i) claim forms presented after 31 December 2020(ii) amendments grated after that date.”[241]What the respondent then said in its closing submissions about what was said in paragraphs 207-210 of the claimants’ written closing submissions was this. “191. In any event, a worker’s right to ‘carry over’ leave to which he is entitled under regulation 13 WTR does not apply to additional leave to which he is entitled under regulation 13A (see Sood Enterprises Ltd v Healy [2013] ICR 1361 ¶ 47-48). 192. In the present case, as set out above, the Claimants(a) were able to stop working, and did stop working, at will; and(b) from June 2021 (or July 2022 for couriers) were paid holiday pay, in accordance with the express right to holiday pay. From this time, the conditions set out in the rewritten regulation 13(16) WTR were not met and so a claim based upon it must fail. 193. Therefore, there is no basis for any carry over of leave after this time. 194. The Claimants seem to be suggesting that if the contractual right to holiday pay was in any way defective, then this is sufficient to invoke the Sash Window approach. That cannot be right. Where a company recognises that a contractual right for an individual to be paid holiday pay, and pays holiday pay in circumstances where the driver is free to stop working at will (and does, demonstrating that the taking of annual leave has been facilitated), then the right to carry over ceases. If it turns out that the level of pay for annual leave has been miscalculated, that does not mean that the individual has been unable to exercise his right to paid annual leave. The ‘underlying philosophy of the WTD [Smith v Pimlico Plumbers (No 2) (CA) ¶ 16] is not subverted, and the case is not any of a refusal claim, non-payment claim, or a termination claim. [Ibid ¶ 25] This is recognised in the drafting of Reg 13(16) by the Court of Appeal in Smith, which allows a carry-over only of ‘any leave which is taken but unpaid, and/or which is not taken’. This does not extend to leave which is taken but under-paid; whereas the remedy under Reg 30 extends to ‘any part of any amount due to him’. 195. There is one final aspect to this. The basis for the right to carry over leave is within European law; hence the Court of Appeal’s amendments to Reg 13 (but not 13A). ‘Holiday pay’ as paid by Addison Lee was based upon 12.07%, in other words to the entitlement under both Reg 13 and 13A (i.e. 5.6 weeks per annum). If the Claimants wish to seek to carry over a claim under Reg 13, and if (contrary to the above) they show that any underpayment of leave is sufficient to allow a carry over, then they must establish that the 12.07% payment was insufficient to cover their Reg 13 right to 4 weeks’ paid annual leave. See Sood referred to above.”[242]Those submissions were at first sight persuasive: if a worker has in fact taken holiday but not been paid for it, or paid in full for it, then one might have thought the only remedy that the worker requires is the amount of the underpayment and compensation for the fact that the worker has had to take the holiday in the wrong belief that it will be unpaid. Those things are provided for by regulation 30 of the WTR 1998.[243]However, the claimants’ written closing submissions continued in the following way. ‘211. In oral submissions, AL suggested that (i) and (ii) [i.e. of the passage set out at the end of paragraph 240 above] are intended to be disjunctive rather than conjunctive – in other words that a worker will not be entitled to carry forward the leave to which he is entitled in any leave year if his employer either recognised his right to paid annual leave or provided a facility for the taking of leave. The LDCs submit that the two requirements were plainly intended by the Court of Appeal to be conjunctive. That is not only the common sense interpretation, but is the only one consistent with the Court’s analysis of the effect of King and Kreuziger. 212. It is also clearly the interpretation of Parliament. The WTR were amended with effect from 1 January 2024 by the Employment Rights (Amendment, Revocation and Transitional Provision) Regulations 2023 SI 2023/1426 (ER(ART)R). Materially in the context of the instant claims, regulation 13 was amended so as to provide: “(16) Paragraph (17) applies where, in any leave year, an employer fails to—(a) recognise a worker’s right to annual leave under this regulation or to payment for that leave in accordance with regulation 16;(b) give the worker a reasonable opportunity to take the leave to which the worker is entitled under this regulation or encourage them to do so; or(c) inform the worker that any leave not taken by the end of the leave year, which cannot be carried forward, will be lost. (17) Where this paragraph applies and subject to paragraph (18), the worker is entitled to carry forward any leave to which the worker is entitled under this regulation which is untaken in that leave year or has been taken but not paid in accordance with regulation 16. 213. The Explanatory Memorandum to the ER(ATR)R describes these amendments as one part of a “Restatement/codification of interpretive effects” affecting three areas of employment law. The Explanatory Memorandum goes on to state: “7.32 A commitment was made in Parliament during the passage of the 2023 Act not to reduce workers’ rights. The Government is proud of the UK’s record on employment standards. To mitigate risks that the removal of interpretive effects on employment law could lead to a reduction in workers’ rights it is appropriate to use the 2023 Act’s ‘restatement’ powers to maintain existing policy effects which are specifically produced by the application of retained EU-derived principles of interpretation. 7.33 DBT’s assessment is that these rights are largely or wholly dependent on the special features of EU law that are removed by the 2023 Act. 7.34 Therefore, the instrument will restate the following three principles before the end of 2023 to ensure these employment rights continue, notwithstanding the removal of the special features of EU law by the 2023 Act: …• The right to carry over holiday entitlement where the employer has failed to inform the worker of their right to paid annual leave or enable them to take it; and [sic; that quotation ended there, with no closing quotation mark] 214. As the Explanatory Memorandum demonstrates, Parliament considered that the “interpretive effects” of the CJEU’s ruling in King and Kreuziger and the Court of Appeal’s decision in Smith required enactment (‘restatement’) because of the legislative changes introduced by REULA on 1 January 2024.’[244]I saw that in the amended version of regulation 13 of the WTR, i.e. as provided for by SI 2023/1426, there was this final sub-paragraph: “(18) Annual leave that has been carried forward pursuant to paragraph (17) cannot be carried forward beyond the end of the first full leave year in which paragraph (17) does not apply.”[245]That seemed to me to support what the claimants said in paragraphs 211-214 of their written closing submissions, but in any event the submissions made in those paragraphs were in my view to be preferred to those of the respondent which I have set out in paragraph 241 above in relation to this point, and I therefore accepted those submissions of the claimants on this point. Unlawful deductions from wages in relation to holiday pay The ultra vires issue[246]The claimants’ submissions under the heading “Unlawful deduction from wages” were as follows. “216. In addition to being entitled to carry over leave to which he is entitled, whether taken or not taken, until the termination of his employment, a worker who has been wrongly classified as self-employed can contend that the employer’s failure to pay him for any leave which he in fact took constituted a breach of Part II of the ERA – see Revenue and Customs Commissioners v Stringer [2009] UKHL 31, [2009] ICR 985.217. For the avoidance of doubt, the claimants do not contend that Part II of the ERA is applicable to the situation where a worker does not take the leave to which he is entitled and is paid his normal remuneration.218. In Agnew (supra) the Supreme Court considered the test to be applied in deciding whether an employer’s failure to pay holiday pay gave rise to “a series of deductions” within the meaning of s. 23(3) ERA, holding at para 127 that: “127 ... whether a claim in respect of two or more deductions constitutes a claim in respect of a series of deductions is essentially a question of fact, and in answering that question all relevant circumstances must be taken into account, including, in relation to the deductions in issue: their similarities and differences; their frequency, size and impact; how they came to be made and applied; what links them together, and all other relevant circumstances.”219. Overruling the decision of Langstaff J in Bear Scotland Ltd v Fulton [2015] ICR 221, the Supreme Court held that a gap of more than three months between deductions does not prevent an employment tribunal finding that they were part of the same series.’[247]The respondent did not appear to take issue with that passage. In paragraph 213 of the respondent’s written closing submissions, however, the respondent set out the whole of paragraphs 126-129 of the judgment of the Supreme Court in Agnew.[248]The extent to which a claim could be made in respect of past deductions is the subject of section 23(4A) of the ERA 1996 which was inserted by the Deductions from Wages (Limitation) Regulations 2014, SI 2014/3322, with effect from 8 January 2015. Before then, section 23 had no temporal limit apart from the requirement that the claim be made within three months of the deduction in question (subject to the possibility of an extension of time on the basis that, as provided by section 23(4), “it was not reasonably practicable for a complaint under this section to be presented before the end of the relevant period of three months.”)[249]After 1 July 2015, a claim could not be made in respect of a deduction which “was made ... before the period of two years ending with the date of presentation of the complaint”. That was because of section 23(4A), read with regulation 4 of the Deductions from Wages (Limitation) Regulations 2014. Those regulations (to which both parties referred as the “DWLR”; I do the same below) also inserted words into regulation 16(4) of the WTR 1998 so that they were in the following terms (the new words being underlined by me). ‘A right to payment under paragraph (1) does not affect any right of a worker to remuneration under his contract (“contractual remuneration”) (and paragraph (1) does not confer a right under that contract).’[250]As the claimants pointed out in paragraph 222 of their written closing submissions, that change to regulation 16(4)250.1 “prevent[s] a worker from circumventing the limitation on the employment tribunal’s jurisdiction by bringing a claim in the ordinary courts” and250.2 “made it impossible for an employee whose employment had terminated to bring a claim for unpaid or underpaid holiday pay under the employment tribunal’s contractual jurisdiction”.[251]The claimants as a result submit (in paragraphs 224-242 of their closing submissions) that the new two-year limit is251.1 “not compatible with the principle of an effective remedy or with Article 47 of the Charter [of Fundamental Rights of the European Union to Poland and the United Kingdom]”, and251.2 ultra vires in that (in summary) it was purported to have been made under section 2(2) of the European Communities Act 1972 when the changes made were not what that provision enabled.[252]Section 2(2) at the material time enabled ‘Her Majesty may by Order in Council, and any designated Minister or department ... by order, rules, regulations or scheme, [to] make provision–(a) for the purpose of implementing any EU obligation of the United Kingdom, or enabling any such obligation to be implemented, or of enabling any rights enjoyed or to be enjoyed by the United Kingdom under or by virtue of the Treaties to be exercised; or(b) for the purpose of dealing with matters arising out of or related to any such obligation or rights or the coming into force, or the operation from time to time, of subsection (1) above [concerning what was in subsection (1) referred to as an “enforceable EU right”]; and in the exercise of any statutory power or duty, including any power to give directions or to legislate by means of orders, rules, regulations or other subordinate instrument, the person entrusted with the power or duty may have regard to the [objects of the EU] and to any such obligation or rights as aforesaid. In this subsection “designated Minister or department” means such Minister of the Crown or government department as may from time to time be designated by Order in Council in relation to any matter or for any purpose, but subject to such restrictions or conditions (if any) as may be specified by the Order in Council.’[253]In paragraphs 228-231 of their written closing submissions, the claimants said this. ‘228. In considering the approach to be taken to subordinate legislation made under the 1972 Act, the courts have held that it is important to have regard to the purpose of s.2(2). In Oakley Inc v Animal Ltd and others (Secretary of State for Trade and Industry intervening) [2005] EWCA Civ 1191, [2006] CH 337, Waller LJ said: “39 ... [the words of section 2(2)] should be given their natural meaning but as we know context means everything. That context is the bringing into force under section 2 of the laws, which under the Treaties the United Kingdom has agreed to make part of its laws. The whole section is clearly primarily concerned with that obligation and the primary objective of any secondary legislation under section 2(2) must be to do just that. Section (2)(2)(b), and the words “arising out of” and “related to” take their context from that being the primary purpose of section 2. It seems to me that section 2(2)(b) from its position in section 2, from the fact that it adds something to both subsections (1) and (2), and from its very wording is a subsection to enable further measures to be taken which naturally arise from or closely relate to the primary purpose being achieved. I accept that I will be accused of adding the words “naturally” and “closely”, but I believe that describes the context which provides the meaning of the words.’229. The principle that the primary objective of any secondary legislation made under section 2(2) must be to bring into force laws which, under the Treaties, the UK has agreed to make part of its law was approved by Lord Hope of Craighead DSPC in Risk Management Partners Ltd v Brent London Borough Council and another [2011] UKSC 7, [2011] 2 AC 34, at [24], and by Lord Mance JSC in United States of America v Nolan [2015] UKSC 63, [2016] AC 463, who added, at [59]: “59 ... that is the context in which Parliament was prepared to delegate law-making ability to the executive – because the focus of section 2(2) is on obligations to the implementation of which the United Kingdom is already committed (and rights to which it is already entitled) at the European level by virtue of its EU membership. Parliament will itself have had prior opportunities for scrutiny of, and input into the content of, the European measures giving rise to such obligations and rights, through in particular select committee procedures, at the stage when such measures were being developed and proposed by the European Commission and considered in Europe by member states and the European Parliament.”230. Lord Mance went on to consider the distinction between ss. 2(2)(a) and (b) of the 1972 Act, in which context he said: “61 What can in my view be said, from the wording and positioning of these two paragraphs, is that paragraph (a) is the main vehicle for implementation of EU obligations and rights which are not directly enforceable. Paragraph (b) goes further, in authorising provision for different purposes, but those purposes are limited by reference to the United Kingdom’s EU obligations or rights (or the coming into force, or operation, of section 2(1)). The words “arising out of” limit the power to provisions dealing with matters consequential on an EU obligation or right (or the coming into force, etc, of section 2(1)). The further phrase “related to any such obligation or rights”, must, unless redundant, go somewhat further. But the relationship required must exist objectively; and the positioning of the phrase and its conjunction with the earlier wording of section 2(1) suggest to me, as they did to Waller and May LJJ, that by speaking of a “relationship” the legislature envisaged a close link to the relevant obligation or right. A relationship cannot on any view arise from or be created by simple ministerial decision that it would be good policy or convenient to have domestically a scheme paralleling or extending EU obligations in a field outside any covered by the EU obligations. That would be to treat paragraph (b) as authorising a purpose to implement policy decisions not involving the implementation of, not arising out of and unrelated to any EU obligation.”231. In Villiers v Villiers (Secretary of State for Justice intervening) [2020] UKSC, [2021] AC 838, at [143] – [144], Lord Wilson JSC described para 39 of Waller LJ’s judgment in Oakley and para 61 of Lord Mance’s judgment in Nolan as “the most helpful commentaries” on the meaning of s. 2(2). He also pointed out that it must never be forgotten that “the required relationship is to the particular terms of the EU instrument which either already has been, or is being, given the force of law in the UK.”[254]The respondent’s written closing submissions responded directly to those paragraphs, because those paragraphs were in the claimants’ detailed opening skeleton argument. The claimants had in their written closing submissions then responded to the respondent’s written submissions.[255]In relation to the issue of compatibility with EU law, the respondent in paragraphs 225 and 226 of its submissions pointed out these things, respectively.255.1 “There is nothing in the cases cited by the Claimants that suggests that the backstop in respect of a claim for unlawful deductions from wages has the effect of making it impossible in practice or excessively difficult to exercise EU rights in respect of annual leave.”255.2 “A similar restriction in respect of claims under the Equal Pay Act 1970 (at the time, limited to two years) was considered in the case of Levez v TH Jennings (Harlow Pools) Ltd (Case C-235/96) [1999] ICR 521 and it was made clear that: ‘19. …it is compatible with Community law for national rules to prescribe, in the interests of legal certainty, reasonable limitation periods for bringing proceedings. It cannot be said that this makes the exercise of rights conferred by Community law either virtually impossible or excessively difficult, even though the expiry of such limitation periods entails by definition the rejection, wholly or in part, of the action brought: see, in particular, Palmisani [1997] E.C.R. 1-4025, 4046, para. 28; Fantask AIS v. Industriministeriet (Erhvervsministeriet) (Case C-188/95) [1997] E.C.R. 1-6783, 6838, para. 48, and Ansaldo Energia, paras. 17 and 18. 20. Consequently, a national rule under which entitlement to arrears of remuneration is restricted to the two years preceding the date on which the proceedings were instituted is not in itself open to criticism.’[256]Paragraph 217 of the claimants’ written closing submissions (which I have set out in paragraph 246 above) was inserted in response to the respondent’s written closing submissions. For convenience, I repeat that in that paragraph, the claimants said this. “For the avoidance of doubt, the claimants do not contend that Part II of the ERA is applicable to the situation where a worker does not take the leave to which he is entitled and is paid his normal remuneration.”[257]In part for that reason, but also because I understood the challenge to the twoyear limit on recovery of unpaid wages to be advanced primarily on the basis that it was ultra vires, I did not understand that the claimants were relying on the issue of compatibility as their primary argument. In any event, I accepted that in principle there is nothing unlawful in applying a time limit to the enforcement of a right derived from EU legislation, as long as it is not more difficult to enforce that right than it is to enforce an equivalent or comparable right in United Kingdom (“UK”) law.[258]What did concern me was the proposition that it was lawful to introduce the twoyear limit on the recovery of unlawful deductions in question in such a way that the enforcement of UK rights which were not derived from EU legislation was limited in the same way in order to ensure the equivalence of enforcement. That, it seemed to me, was at least dubious. In that regard, the respondent submitted this, in paragraph 249 of its written closing submissions. “The fact that the DWLR also relate to wages other than the entitlement to holiday pay under the WTD does not mean that they are ultra vires. The Claimants contend that to sever the operation of the DWLR so that it only relates to holiday pay would breach the principle of equivalence (C Opening ¶ 157). That is precisely the point. It demonstrates the fact that extending the operation of the DWLR to other wages claim conforms with the necessary purpose under section 2(2)(b) by ensuring equivalence.”[259]In the next paragraph of those submissions, the respondent said this. “If it is accepted that claims for unlawful deductions in respect of holiday pay under the WTD relate to an EU obligation, the alignment of the limitation on holiday pay claims with claims for deductions from other kinds of wages is in accordance with the principle of equivalence. Accordingly, it is right that a procedural limitation on holiday claims is applied similarly to other claims for deductions from wages.”[260]And in conclusion on the issue of whether the restriction by a statutory instrument purportedly made under section 2(2) of the European Communities Act 1972 was ultra vires, or unlawful, the respondent said this in paragraph 252 of its written closing submissions. “The imposition of a restriction on bringing claims for holiday pay with a two year backstop is such a procedural rule; see Levez, above. Amending the procedural limitation on a claim for unlawful deductions in respect of holiday pay is therefore itself for the purpose of dealing with matters arising out of or related to Community obligations or rights under section 2(2)(b).”[261]In paragraph 247 of those submissions, the respondent said this. “The Claimants still further argue that the DWLR were ultra vires because they did not seek to ‘implement’ the WTD: that had already been done and the DWLR were seeking to restrict the enforcement of that right (C Opening ¶ 152). Clearly ECA s 2(2)(a) concerns the implementation of an obligation, or enabling of a right. But s 2(2)(b) is – in its terms – not to confined, extending to “matters arising out of or related to any such obligation or rights”. As Jacob LJ made clear in Oakley ¶ 74, ‘clearly section 2(2)(b) goes further than merely enabling implementation. It allows more to be done by delegated legislation’ and ‘So section 2(2)(b) indeed adds more’ (¶ 80). Each of the judges considered that the narrower approach advocated by Lord Johnson in Addison’s case (quoted by Waller LJ at ¶ 35) was wrong: see Waller LJ ¶ 39, May LJ ¶ 47 and Jacob LJ ¶ 80. There is nothing in Oakley, or indeed any other case, which suggests that s 2(2)(b) can only be deployed in the same instrument as a measure introduced under s 2(2)(a). There is no reason in principle why the two different strands cannot operate distinctly and on different occasions.”[262]In my view, that was not an answer to the question whether it was envisaged by Parliament that section 2(2)(b) of the European Communities Act 1972 could lawfully be used to limit the exercise of a UK statutory right which was not derived from EU law. In my view, Parliament did not envisage that, and in my view it was not lawful to limit the operation of 23 of the ERA 1996 in the manner which I describe in paragraphs 248 and 249 above.[263]I accept that my conclusion in that regard will be challenged and may be wrong. However, in my view the rights conferred by Part II of the ERA 1996 are of such fundamental importance to employees and workers that an amendment which limited their effect in the manner which I describe in paragraphs 248 and 249 above was one which Parliament cannot have envisaged would be the subject of a statutory instrument made purportedly under section 2(2)(b) of the European Communities Act 1972 where the limitation was imposed in order to ensure that the instrument could not be challenged on the basis that a right derived from EU law was more difficult to enforce than a similar UK right. In my view, a change to a primary statutory right, i.e. one conferred by an Act of Parliament, such as those which are conferred by Part II of the ERA 1996, of the sort which the DWLR purported to make, could be made only by a primary enactment, in part because the negative resolution procedure could not be used to cause a statute, as opposed to a statutory instrument, to come into existence. The claimants’ reliance on Article 1 of the FIrst Protocol to the European Convention on Human Rights (“ECHR”)[264]That conclusion was supported by the considerations arising from the claimants’ reliance on Article 1 of the First Protocol to the ECHR, which was incorporated in UK law by Schedule 1 to the Human Rights Act 1998 and provides: “(1) Every natural or legal person is entitled to the peaceful enjoyment of his possessions. No one shall be deprived of his possessions except in the public interest and subject to the conditions provided for by law and by the general principles of international law. (2) The preceding provisions shall not, however, in any way impair the right of a State to enforce such laws as it deems necessary to control the use of property in accordance with the general interest or to secure the payment of taxes or other contributions or penalties.”[265]I could not see how that provision could take effect here except as a support for a conclusion that the DWLR was ultra vires section 2(2)(b) of the European Communities Act 1972. In that regard, I noted what the claimants said in paragraphs 255 and 256 of their written closing submissions, which was as ‘255. The Impact Assessment for the DWLR stated that the Government’s policy objectives in introducing the DWLR were: “(1) To limit the burden on businesses and reduce the scope (and impact) of backdating of liability provisions that currently exist and (2) To provide certainty to employers and workers. Various options are considered that amend the Employment Rights Act 1996 to reduce the scope of liability by limiting how far claims can go back.” 256. The DWLR did not strike a fair balance between these competing interests because: 256.1. although the principal aim of the Regulations was to restrict backdated claims for holiday pay, the Regulations placed a two year limitation on a very wide range of claims, the resolution of which had not placed undue burdens on businesses; 256.2. the limitation arbitrarily restricted the enforcement of important statutory rights which protect some of the most vulnerable members of the workforce, in particular the right to receive the national minimum wage; 256.3. the two year limitation was bound to have a particularly unfair effect on workers who have been mischaracterised as selfemployed contractors and may therefore have been denied the national minimum wage and holiday pay for many years; 256.4. other less intrusive options could reasonably have been adopted (see James v United Kingdom (1986) App no. 8793/79, at [51]) e.g. placing a less restrictive limitation on retrospective claims or limiting any restriction to particular types of claim under the WTR.’[266]The respondent’s closing submissions on the question of whether or not there was here an unlawful interference with a right to a possession within the meaning of Article 1 of the First Protocol to the ECHR (they were in paragraphs 254-265) were cogent at least in so far as they pointed out that the issue was one of legitimate expectation. In UK public law, such an expectation can take effect either as a right to be consulted before that which is legitimately expected to be continued is discontinued, or as a simple, or unqualified, right to have the thing in question continued.[267]I noted the passage from the report of the European Court of Human Rights Béláné Nagy v Hungary (2016) App no 53080/13 set out in paragraph 259 of the respondent’s closing submissions, and in particular the words in bold in paragraph 77 of that passage, which were a quotation from paragraph 52 of the the earlier decision of that court in Kopecký v Slovakia [GC], no. 44912/98, namely: ‘where the proprietary interest is in the nature of a claim it may be regarded as an “asset” only where it has a sufficient basis in national law, for example where there is settled case-law of the domestic courts confirming it’.[268]It seemed to me from that analysis that it had to be open to the UK government to remove by legislation a previously-conferred right, and that the requirement of lawfulness would be met by the requirement that the removal was itself lawful. I therefore concluded that the claimants’ reliance on Article 1 of the First Protocol to the ECHR was well-founded in that (but only in that) it supported the conclusion that the use of a statutory instrument purportedly made under section 2(2)(b) of the European Communities Act 1972 to limit the extent to which claims which were not derived from EU law could be backdated, was ultra vires. My conclusions on the issues which, by the end of the hearing, remained to be determined[269]My conclusions on the issues which remained “live” at the end of the hearing were as follows. Worker status Issue 1: “Were any of the Claimants ‘workers’ employed by the Respondent for the purposes of ERA, WTR and NMWA?”[270]Yes. All of them. The key question is at what times they were such workers: i.e. in each case, when during what was for them a working day they were to be treated as such a worker.[271]However, I concluded without any hesitation that the relationship between the claimants and the respondent at all material times before 2017 when they were logged onto the respondent’s Shamrock software via their mobile devices (whether it was an XDA, an MDA or a mobile telephone with the respondent’s app installed and operating on it) was that of(1) a limb b worker and(2) a worker to whom the WTR and the NMWA applied. That was because271.1 the respondent put before me very little evidence about that relationship and, to the extent that it put before me any such evidence, it did not show that the relationship was any different from that to which the Lange judgment applied; and271.2 I agreed completely with the Lange judgment.[272]In effect, therefore, as far as I could see, the respondent conceded by its conduct that there was no reason for me to come to a conclusion on the facts here which differed from that which was arrived at in the Lange judgment in relation to the period in issue in that case.[273]As for the rest of the period which was in issue in the claims to which the hearing which I conducted related, I came to the following conclusions.[274]The respondent’s attempt to persuade the tribunal (i.e. whichever tribunal heard the claims) that the situation had changed materially since the Lange judgment was given by reason of a material difference in the way in which the respondent operated in relation to drivers who had hired their vehicles from Eventech, was aimed only at the extent to which the respondent imposed financial penalties or sanctions in the event of a failure by a driver to accept a job which was automatically (or, unusually, manually) allocated to the driver via the driver’s mobile device. For the reasons stated in paragraphs 77-114 above, I concluded that those attempts were paid lip service only. While the respondent said in the driver agreement that drivers were free to reject a job, that added nothing material because drivers were always free to reject a job. The issue was the sanctions that would, or might, be applied. If and to the extent that the respondent in practice imposed sanctions on drivers who rejected jobs, despite what was said in clause 7.1 or clause 8.1 (as the case may be) to which I refer in paragraph 102 above, those clauses had to be ignored as being either a partly or a wholly inaccurate reflection of the reality of the situation. Given my conclusions stated in paragraphs 107 and 113 above, I concluded that those clauses were inaccurate and had to be ignored for present purposes.[275]In any event, there was after 2017 (including by the new driver deal of 2021 to which I first refer in paragraph 83 above) an increasing reliance on the part of the respondent on incentives, such as those to which I refer in paragraphs 172-173 above, which in my view had the effect that the claimants were so obviously integrated into the operations of the respondent that275.1 whenever a non-partner driver was logged onto the Shamrock software, and275.2 whenever a partner driver had accepted a job via that software until the job had been ended either by being cancelled or the passenger or package having been delivered to his or her, or its, destination, they were limb b workers and workers for the purposes of the WTR and the NMWA.[276]However, a partner driver, that is to say a driver who did not lease his vehicle from Eventech, and who took work from other providers of work such as Uber or Bolt, was in my view not a limb b worker at that time. That was because in my view that was the inescapable result of the application of the words in section 230(3)(b) “whose status is not by virtue of the contract that of a client or customer of any profession or business undertaking carried on by the individual”. In this regard, I had the misfortune to disagree with what the Employment Tribunal in Bandi & Others v(1) Bolt Operations OÜ and(2) Bolt Services UK Ltd, (Case Nos. 2206953/2021 & Ors) said in paragraph 156 of its reasons, which was that “it would offend language and common sense to attribute to Bolt in its dealings with drivers the status of a client or customer of a business or profession carried on by the drivers. It is very hard to characterise the driver as a business. He was simply a man with a car looking to earn a living from it. It is even harder to characterise Bolt as the driver’s customer. There was no arm’s length transaction. There was never any question of the two parties negotiating or striking a bargain. The business was Bolt’s and the transaction between it and the driver was its purchase of the driver’s labour, strictly on its ‘take it or leave it’ terms.” Issue 2: If so, which Claimants were ‘workers’ employed by the Respondent and at what dates? For the avoidance of doubt, this issue includes whether an individual who was a worker ceased at any point to be a worker (whether in or about June 2021 or otherwise and if so when).[277]There was no need to address this issue separately in the light of the conclusions which I state on issue 1 above. My answer to issue 2 is stated in those conclusions. Issue 3: If and in so far as any of the Claimants were at any time ‘workers’ employed by the Respondent, can they (subject to clause [i.e issue] 5 below) in principle qualify as ‘working’ under a ‘worker’ contract at times when their ‘app’ was switched on but they had not yet accepted a passenger journey (Uber v Aslam [2021] ICR 657 SC paras 121-130 refer)? (It may be relevant to this issue to decide whether there was any or sufficient obligation to work in this period).[278]Again, my answer to this question is stated in my conclusions on issue 1 above. Issue 4: If and in so far as the answer to 3 is ‘yes’, was each Claimant, on the facts of each individual case relating to each period for which a claim is made (or which is relevant to the amount of the claim) ‘ready and willing’ to work so as to qualify in fact as ‘working’ under a worker’s contract at that time? (Uber v Aslam [2021] ICR 657 SC paras 121-130, esp. para 130 refer).[279]My answer to this question is also stated in my conclusions on issue 1 above. Issue 5: In light of the answers to clauses 3 and 4, were the particular periods for which claims are made (or which are relevant to the amount of claims) when the Respondent’s app was switched on but no passenger journey had been accepted to be treated as working time(a) for the purposes of WTR, and/or(b) for the purposes of NMWA and NMWR [i.e. the National Minimum Wage Regulations 2015, SI 2015/621]?[280]Again, my answer to this question is stated in my conclusions on issue 1 above. Holiday pay The applicability of EU law Issue 6: If and in so far as the Claimants rely on EU law to establish rights in relation to holiday and holiday paya. whether by virtue of Protocol No 30 on the Application of the Charter of Fundamental Rights of the European Union to Poland and the United Kingdom, European law in relation to holiday and holiday pay (including Article 7 WTD [i.e. the Working Time Directive, Directive 2003/88/EC]) does not (or does not in relevant respects) have effect against the Respondent, a private party; alternativelyb. whether any such effect against the Respondent as a private party is by virtue of Section 5 of the European Union (Withdrawal) Act 2018 (and Schedule 8) inapplicable to i. claim forms presented after 31st December 2020 ii. amendments granted after that date. “the Applicability of EU Law Issue”)[281]I regarded this issue as an academic issue, which in any event did not require determination as an independent issue (although it was in fact several issues). I saw that the respondent’s proposed answer to the issue was this: “Answer: The Respondent accepts that this issue is not open for argument before the Tribunal, but reserves the right to argue it on any appeal.”[282]That fortified my conclusion that I did not need to answer issue 6. Pay for holiday taken or deemed to have been taken Issue 7: To how many days of leave, if any, was each Claimant entitled under WTR in respect of the contract(s) under which he or she was engaged?[283]The respondent’s proposed answer to this question was: “1. 4 weeks under WTR Reg 13 (as a result of EU law) 2. 1.6 weeks under WTR Reg 13A (as a result of domestic law)”.[284]The claimants did not as far as I could see address this issue specifically, but in any event, I thought that the respondent’s proposed answer was correct. However, it was not something on which I was able to give a judgment since it is not within the jurisdiction of the tribunal to make a declaration of the applicable law. Rather, the tribunal’s jurisdiction is to determine disputed claims on the facts of those claims. I found all of the rest of the issues put before me for determination to be incapable of being the subject of a formal judgment by me for essentially the same reason. I nevertheless turn to those issues and state my conclusions on them, since it appeared to me to be in the interests of justice to do so. Issue 8: How many days of leave were taken by each such Claimant under each contract in respect of the leave year(s) for which a claim is asserted?[285]The respondent’s submission on this issue was that(1) the claimant had “not set out their case on this issue” and(2) “As a matter of principle, annual leave is any period of leave which is not another type of rest period.”[286]This issue was the most difficult one for me to decide on the evidence before me. The hearing which I conducted was focused in large part on the issue of status: that was the main focus of both parties. That was because of the respondent’s response to the claims. In any event, I heard very little by way of submissions on the precise days of leave which I should conclude the claimants had taken. I refer to the claimants’ stance on this issue in paragraphs 207-209 above. Mr Leiper on behalf of the respondent proposed, as I record at the end of paragraph 212 above, that I made findings on the issue of when there were breaks in service as far as the test claimants were concerned on the basis that the issue was one of principle. I concluded that I should, if possible, do that for all of the issues relating to the precise amounts of time that the claimants had taken off by way of holiday and to what, if anything, they were entitled by way of pay for that holiday and that I should adjourn the hearing in order to determine those precise things in the event that the parties were unable to agree the result of the application of my determinations of the principles to be applied. However, I also concluded that I should give the parties indications of my likely conclusions by reference to what I had seen in the evidence. That would assist the parties to come to an agreement on the entitlements of the test claimants, and it would also mean that if I had to resume the hearing because they had been unable to come to such an agreement, then they would know my current thinking on those entitlements.[287]The problem here was that there was little helpful evidence before me to enable me to come to a clear conclusion on the specific amounts of holiday that the claimants had taken. It therefore occurred to me that it was in the interests of justice for me to show by reference to the case of one of the test claimants how I thought that the issue had to be resolved, and then leave it to the parties to seek to determine the factual dispute by reference to the documentary evidence (whether or not it had already been disclosed), allowing for the possibility that the issue was later determined by me at a resumption of the hearing, but with the parties’ efforts aimed specifically at the cases of the test claimants.[288]So, since I set out the evidence relating to Mr Nardelli at some length in paragraph 216 above and it was possible to illustrate my likely conclusions by reference to his case, I did that in relation to Mr Nardelli’s case. His witness statement referred in paragraph 43 to the document starting at page 8521 (in fact it went on to page 8527). His claim form was issued on 27 October 2021. At page 8521 it was said that he had a three week break from 6 August 2021 onwards. The preceding break was for two weeks starting on 27 November 2020. There was a statement that he had taken a break of 4.5 months from 27 March 2020 onwards. That break was likely to be such as to preclude a claim made in relation to the period before 27 March 2020. In any event, the evidence before me showed that Mr Nardelli had taken at least 5 weeks off in the year before he made his claim. It was possible that he had taken other days off during that period, but I could not see whether or not that was so from the document at pages 8521-8527. The answer might be found in the “native” data document concerning Mr Nardelli, which was the spreadsheet with the document number 1533, and that was a question for the parties. If they cannot reach agreement in that regard then I will hear their submissions on the point at the resumption of the hearing which will then need to occur (to which possibility I return in paragraph 322 below, where I refer to the practicalities of the situation). I add that if either party’s resistance to agree a particular point is plainly not well-founded on the documentary evidence then in existence, then it will be open to the other party to argue that the resumption of the hearing was unnecessary, with the obvious potential consequence of a costs order. Issue 9: Are any of the Claimants to be treated as having taken leave pursuant to reg 13 and/or reg 13A WTR because of the Respondent’s refusal to remunerate them in respect of such leave?[289]The respondent’s submission on this issue was as follows. “Answer: 1. Since 2021 the Respondent has recognised a right to take paid annual leave. 2. As a result, leave does not carry forward to the next leave year.”[290]That was in my view not a good answer to the question. That was because the respondent had asserted that it had given a right to holidays by paying a proportion of earnings as holiday pay, but it had paid such a proportion on the basis that only time spent on a job was working time, and I have concluded that in relation to non-partner drivers, time spent logged on was also working time. As a result, I concluded that if a claimant could show in respect of time spent as a non-partner driver that he had not been able to take the amount of leave to which he was entitled under either regulation 13 or regulation 13A of the WTR in any leave year, then as a matter of principle, the claimant was able to carry forward to the following year the right to the leave which had not been taken.[291]I arrived at that conclusion as a matter of principle, applying the principles to which I refer in paragraph 239 above. The claimants’ submissions on the point, in paragraphs 259-261 of the closing submissions, were to the same effect, but they applied the detail in the case law. They were in the following terms. “259. The right to carry over the annual leave they should have received in previous years applies as much to those TCs whose relationship with AL terminated after June 2021 as it does to TCs who ceased driving for AL before June 2021. As made express by what is now regulation 13(16) WTR, expressly intended to codify the purposive interpretion of the Court of Appeal in Smith, regulation 13 imposes requirements that are conjunctive (as discussed above; cf. R’s oral submissions); i.e. the worker is allowed to carry forward the leave to which he is entitled in any leave year until termination of employment unless his employer has, in the relevant leave year, not only recognised his right to paid annual leave but also provided a facility for taking such leave.260. Further, the Court interpreted regulation 14(5) as providing that if on termination the worker remains entitled to leave in respect of any previous leave year which he carried over under regulation 13(16), the worker is entitled to a payment in lieu of such leave.261. It follows (LDCs’ primary case) that the claimants are entitled to be paid in lieu of carried over leave under regulation 14 and regulation 30(1)(b) WTR.”[292]I therefore accepted those submissions. They led to the same answer as that to which I had come as a matter of principle. Issue 10: Was any leave validly carried over by any Claimant from(a) one relevant leave year to another,(b) one contract to another, and, if so, how much?[293]The respondent’s proposed answer to this question was as follows: “1. Not from 2021 onwards. 2. In any event, leave could not carry forward across contracts.”[294]My conclusion on issues 9 (read in the light of my conclusion on issue 1) was a conclusive determination in relation to the first of those proposed answers, and that was that claimants were not precluded after 2021 from carrying leave forwards. The second of those proposed answers required a further analysis. The respondent had in 2021 introduced a maximum term of 12 weeks for vehicle rental agreements: see paragraph 91 above. I could see no legal justification for concluding on the facts before me that the termination of one of those agreements resulted in a break in service for the purposes of the WTR. In any event, it was clear to me that the respondent had introduced those agreements specifically with a view to avoiding the impact of the statutory protections which are in issue here, including, of course the rights of a worker to paid holiday under the WTR. As a result, I concluded that I could not lawfully regard the termination of one of those 12-week agreements as giving rise to a termination of a claimant’s engagement as a worker for the purposes of the WTR, although if the termination co-incided with circumstances which in themselves justified the conclusion that there had been a termination of the engagement then the termination of the contract was a potential reinforcement of that conclusion.[295]For the avoidance of doubt, I therefore accepted the claimants’ submissions on the impact of the 12-week agreements, which were in the following two paragraphs of their closing submissions. “322. The successive fixed-term contracts which AL issued to the drivers did not reflect the reality of the relationship. They were an artificial device, presumably designed to allow AL to argue that limitation had been triggered at the end of an arbitrarily defined period, even though the claimant was continuing to work for the company. The successive fixed-terms were therefore contrary to regulation 30 WTR, which provides that any provision in an agreement is void insofar as it purports to exclude or limit the operation of the Regulations or to preclude a person from bringing proceedings under the Regulations before an employment tribunal (and see Uber para 80, quoted above). 323. Further, contractual terms which purport to introduce a false break into a continuing employment relationship are manifestly incompatible with the principle of effectiveness.” Issue 11: For how many days was each Claimant paid holiday pay for the leave years in question and in what amounts?[296]The respondent’s answer to this question was that it was inapplicable. I found it impossible to answer in the abstract, so I did not answer it. It was a question which had to be answered by reference to the facts of a particular claimant’s case. Issue 12: In computing the sums already paid to the Claimants what allowance is to be made in each case(a) for credits earned by Claimants used to set off car-rental liability, and(b) for payments on account of holiday made since June 2021?[297]The respondent’s proposed answer to this question was predictable: that allowances for both things should be made. The claimants’ submissions on the point were made in paragraphs 264-314 of their written closing submissions. Paragraph 265 summarised the claimants’ position in this regard as follows: “The introduction of the right to holiday pay was accompanied by significant changes in the drivers’ remuneration structure, which were wholly lacking in transparency. AL’s disclosure demonstrates that the intended effect of these changes was that the introduction of holiday pay would not leave the drivers any better off overall and that the total cost of their remuneration to AL would not increase – see, for example, HB/3787, 3795 and 3798.”[298]Reliance was then placed on what was said by the CJEU in Robinson-Steele v RD Retail Services (Case C-131/04) [2006] ICR 932, and by Elias P on behalf of the EAT in Lyddon v Englefield Brickwork Ltd [2008] IRLR 198.[299]The claimants then, applying what was said in those cases, submitted among other things in paragraph 285 of their closing submissions that “It is obvious on the facts of this case that AL has not discharged its burden of proving that there was a “mutual agreement” for a “true addition to the contractual rate of pay”, a “true agreement providing a genuine payment for holidays”. Whether the burden of proving that those requirements are met had been discharged, is a question of substance, not of form. The fact relied on by AL, that in 2021 there were both ‘winners’ and ‘losers’, had nothing to do with the supposed introduction of ‘holiday pay’, but was due solely to the adjustments to the payment terms effected at the same time. It is precisely the intention and the achievement of no additional pay to drivers as the result of the introduction of ‘holiday pay’ [Footnote: In fact, the position for drivers was even worse than ‘cost neutrality’, because the calculations done by AL assumed a 50% take up of the 3% pension employer’s contributions (hence the assumption of 13.5% addition to overall pay: 12% for ‘holiday pay’, 1.5% for pension); but AL knew, and events proved, that the take up for pension was likely to be far lower than 50%.] that negates there having been a ‘mutual agreement for a true addition to the contractual rates of pay’, a ‘genuine payment for holidays’.”[300]The claimants also relied on the propositions that300.1 the introduction of holiday pay was not achieved by agreement but by unilateral variation (paragraph 281 of the closing submissions),300.2 “the introduction of ‘holiday pay’ was anything but “a genuine addition to the remuneration paid” – rather, it was a cynical attempt by AL to circumvent its liability under the WTR by reducing gross earnings and reallocating the reductions to ‘holiday pay’, achieving a cost-neutral result: see [3787, 3795, 3798” (paragraph 282 of the closing submissions), and300.3 there was a lack of the required transparency (paragraph 287 of the closing submissions).[301]In paragraph 310 of their closing submissions, the claimants said this about the rental credits. “Nearly every self-employed worker will have work-related expenses. If that worker is paid a ‘wage’ of £X per week, or month, his employer must afford him paid annual leave based on £X; it is not entitled to make a notional deduction before calculating the sum due for paid annual leave, on the purported basis that such notional deduction will cover some or all of his work-related expenses (whether or not incorporated into the employer’s standard terms by it). That would be a heterodox proposition.”[302]It occurred to me that it was impossible to work out what the claimants were paid unless one took into account the purported holiday pay and any rental credits. On a common sense, or practical, basis, therefore, I could not see how those payments could sensibly be ignored. The only proper objection to taking either of them into account to my mind was that what the respondent called holiday pay was paid only after the introduction of a new payment system which did not show the hourly rate of pay and therefore, if only for that reason, was not transparent.[303]I add that what the claimants said in paragraph 310 of their closing submissions to my mind failed to take into account the fact that a claimant’s hourly rate had to be calculated by deducting expenses from revenue.[304]I was not at all sure that the case law dealt with the question of how it was necessary to approach the question of whether or not a self-employed worker had received paid holiday in accordance with the WTR. Neither party addressed me on that issue. I saw that in paragraph 309 of the claimants’ closing submissions, they relied on what was said in British Airways plc v Williams and ors (Case C-155/10) [2012] ICR 847, but that case concerned employees, i.e. who were employed under contracts of employment.[305]It seemed to me that the only practical and reliable solution in solving the problems thrown up by issue 12 was to do what the claimants urged in paragraphs 311-312 of their written closing submissions, which was to say that the claimants were entitled to the national minimum wage for the hours when they were logged on (unless they were partner drivers: see paragraph 276 above), and then award them holiday pay calculated by reference to what they were entitled at the material time by way of the national minimum wage. In that regard as far as I could see I was obliged to ignore the value of the benefit of the use of a vehicle owned by Eventech. No submissions were made to me on that issue, though, and I therefore may be willing to reconsider that conclusion if the respondent puts submissions on the issue before me pursuant to rule 69 the Employment Tribunals Rules of Procedure 2024 within 28 days of this document being sent to the parties (i.e. I now extend time in that regard from 14 to 28 days), in accordance with what was at the time of writing these reasons rule 70 of those rules. Issue 13: Does any failure by the Respondent to make payments due to drivers pursuant to reg 16 WTR give rise a series of unauthorised deductions from the driver’s wages during their time working for the Respondent, contrary to sections 13 and 23 ERA?[306]The claimants did not make any specific submissions on this issue, dealing with it instead compendiously under the heading above paragraph 205 of their written closing submissions as part of “Issues 7 to 18: Pay for holiday taken or deemed to have been taken”.[307]The respondent’s proposed answer to the question raised in issue 13 was this: “1. No. 2. It is not currently clear how the Claimants put their case on this. 3. In any event, the breaks referred to above create a break in any series of deductions.”[308]The claimants’ submissions on the point were to be found in part in paragraph 218 of their closing submissions, which was in these terms. ‘218. In Agnew (supra) the Supreme Court considered the test to be applied in deciding whether an employer’s failure to pay holiday pay gave rise to “a series of deductions” within the meaning of s. 23(3) ERA, holding at para 127 that: “127 ... whether a claim in respect of two or more deductions constitutes a claim in respect of a series of deductions is essentially a question of fact, and in answering that question all relevant circumstances must be taken into account, including, in relation to the deductions in issue: their similarities and differences; their frequency, size and impact; how they came to be made and applied; what links them together, and all other relevant circumstances.”[309]However, the claimants said in paragraph 217 of their closing submissions that they were not contending “that Part II of the ERA is applicable to the situation where a worker does not take the leave to which he is entitled and is paid his normal remuneration.”[310]I could not see how I could properly conclude otherwise than that a failure to make payments to a worker under regulation 16 of the WTR, for essentially the same reasons although their detail may have differed from time to time, would constitute a series of deductions within the meaning of section 23(3) of the ERA 1996. I therefore concluded that such a failure would as a matter of principle rise to such a series. Issue 14: Subject to the Applicability of EU Law Issue:-(a) Is any claim brought outside the primary time limit?(b) Do the principles of EU law established by inter alia King v Sash Window Workshop & another [2018] ICR 693, Max-Planck v Shimizu [2019] I CMLR 35, Stadt Wuppertal v Bauer Willmeroth [2019] 1 CMLR 36 and Kreuziger v Land Berlin [2019] 1 CMLR 34, Article 47 of the Charter of Fundamental Rights of the European Union and/or the principles of effectiveness and/or equivalence generally, require any of the applicable limitation periods and/or retrospectivity limits to be interpreted purposively, or alternatively disapplied,? If so, to what extent and in what circumstances?(c) Further and in particular, is the limitation to the two year period prior to the institution of proceedings under s23(4A) ERA ineffective, (i) as being incompatible with the principles of effectiveness and equivalence under EU law and/or Art 47 of the Charter of Fundamental Rights and/ or (ii) on the basis that in providing for the insertion of section 23(4A) into ERA, the Deduction from Wages (Limitation) Regulations 2014 were ultra vires section 2(2) European Communities Act 1972 and/or (iii) as being in breach of Article 1, Protocol 1, Schedule 1 of the Human Rights Act 1998?[311]I saw that the respondent’s proposed answers to the questions raised in issue 14 were: “(a)No, not in the case of any of the Test Claimants. (b) Save as to the particular case of (c), this does not seem to be pursued. (c) No, on each basis.”[312]So, the first sub-issue was resolved by agreement.[313]I rather doubted that the second issue was capable of being answered in the abstract, but in any event the claimants did not appear to me to be pressing the point in relation to any of the test claimants.[314]I have already addressed the third issue and come to the conclusions stated in paragraphs 257-268 above. That issue was raised again as issue 35, and those conclusions apply to that issue also. I therefore do not address issue 35 below. Issue 17: Should reg 30(1)(a) WTR be read so as to mean that a failure or refusal to pay a worker for the leave to which he is entitled under regs 13 and/or 13A WTR amounts to a failure or refusal to permit a worker to exercise the right to paid annual leave under EU law, such that compensation for refusing to permit the worker to exercise the right in full falls to be assessed under regs 30(3)(b) and 30(4) WTR?[315]In paragraphs 316 and 318 of their closing submissions, the claimants said respectively that they were not pursuing any claim in relation to regulation 13A of the WTR and that they were not pursuing an alternative claim under regulation 30(1)(a) of the WTR.[316]The respondent’s response in summary to the issue (before being informed that the claimants were not pressing those aspects of issue 17) was this. “1. No under Reg 13A, which is a domestic law claim. 2. No under Reg 13, at least since June 2021 when R recognised a right to paid annual leave. 3. In any event, there was no failure or refusal to pay prior to that date, in circumstances in which R did not know whether any C was at any point seeking to take leave.”[317]I was in those circumstances not sure what the claimants’ position on issue 17 was. It was also something of a general question, in that it was of a sort which was in my view best considered by reference to the particular facts of a particular case. It appeared to me that what the claimants said in paragraph 317 of their closing submissions was, if correct, the answer to the issue. I did not know whether or not the respondent disagreed with what was said in paragraph 317, but it was a proposition of law which was in my view best tested against the facts of a particular case and not an issue which I should address in the abstract. “Pay in lieu on termination” claims Issue 19: Should any Claimant in any leave year be treated as having been unable or unwilling to take some or all of the leave to which they were entitled under reg 13 and/or reg 13A WTR because of the Respondent’s refusal to remunerate them in respect of such leave? Issue 20: If so, was any Claimant entitled to carry such leave forward to subsequent leave years until the termination of his employment and to receive a payment in lieu of such leave on the termination of his employment? Issue 21: Was any Claimant’s employment terminated during his leave year meaning he is entitled to receive a payment in lieu in respect of accrued but untaken leave on the termination of his employment under regulation 14 of the WTR?[318]In paragraph 320 of their closing submissions, the claimants said this. “The principles the Tribunal will be invited to apply in answering Issues 19 to 21, i.e. whether the Claimants were to be treated as having been unable to take paid leave and whether they were entitled to carry such leave forward until the termination of their employment, have already been addressed in the context of issues 7 to 18.”[319]I did not therefore see a need to address issues 19-21. I did see, however, that the respondent’s summary submission in response to issue 21 was that “The Cs have not advanced such a case”. This was therefore another set of issues to which it was not going to be fruitful for me to give an answer here. Issue 22: For the purposes of reg 14 WTR, when did/does a Claimant’s employment ‘terminate’? In particular:(a) Did/does a Claimant’s employment terminate for the purposes of reg 14 WTR when any overarching contract between the Claimant and the Respondent comes to an end?(b) Did/does a Claimant’s employment terminate (meaning that a right to payment in lieu crystallises) at the expiry of any relevant contract on which the Claimants were engaged, such that time for claiming runs from the ending of each individual contract?[320]The respondent’s contentions in summary in response were these. “1. The Cs are not relying on an overarching contract (see opening skeleton). 2. In any event, the fixed term contacts were genuine and there is no basis for disapplying their express terms. 3. As to (b), yes, for the reasons set out above.”[321]I have already concluded in paragraph 294 above that the fixed term contracts were not “genuine” in the sense in which that word was used by the respondent. That meant in my view that the answer to issue 22(b) had to be “no”, and that the issue in each case was when a series of deductions ended, which would be when the claimant ceased to be a limb b worker, and that in turn would depend on the answer to the claimants’ closing submissions set out in paragraph 210 above.[322]While I was inclined to accept the claimants’ submissions which I have set out in paragraph 210 above, I could not see how I could sensibly (i.e. lawfully) decide whether or not those submissions were in fact apt unless they were applied to the facts of a particular case. I had heard no submissions in that regard in relation to the test claimants’ cases. As a result, I concluded that the parties should seek to agree the position in relation to each test claimant in the light of my conclusion stated in paragraph 294 above, namely that the fixed-term contracts (for 12 weeks or less) were in themselves irrelevant to the question of when a driver ceased to be a limb b worker. If the parties are unable to agree the position in regard to any test claimant then there will need to be a hearing by CVP for me to decide what directions to give for and in relation to the resumption of the hearing so that the parties can address the issue in relation to each test claimant’s case. I have not made an order in that regard, but I have in the final paragraph of these reasons below, i.e. paragraph 335, stated my conclusion on the best way forward. Issue 27: Should the Tribunal:(a) Under section 24(2) ERA award a sum to compensate drivers for any financial loss sustained by them which is attributable to any delay in repaying the deductions?(b) Under reg 30(4) WTR, in assessing just and equitable compensation, award a sum representing interest, or otherwise as additional compensation, for any financial loss sustained by the drivers, which is attributable to any failure to allow the drivers to exercise their statutory right to paid annual leave and any delay in compensating them?(c) Under reg 30(5) WTR, order that any sum which it finds to be due to the drivers should also include a sum of money representing interest and/or reflecting any delay in making payments due under regs 14 or 16?[323]The respondent’s summary response to this issue was simply: “No, there is no basis for such awards.” I understood that to mean that there was no factual basis for such awards. I disagreed. If there was a failure to pay sums due to the claimants in respect of holiday entitlements then, if only for the reasons stated by the claimants in paragraphs 334-338 of their closing submissions, which relied on the principles established by the ECJ in Marshall v Southampton and South West Area Health Authority (No 2) C-271/91 [1993] ICR 893, there had to be a power to give compensation for such failure. In fact, I could not see how any other conclusion could reasonably be reached simply by applying the wording as it stood of section 24(2) of the ERA 1996 and regulation 30(4) of the WTR. As for an award under regulation 30(5), that could be classified as an unpaid wages claim to which section 24(2) of the ERA 1996 applied. National Minimum Wage Issue 28: Is each Claimant’s work unmeasured work or output work relevant in respect of the claims being brought?[324]The claimants did not address issue 28 in their closing submissions. Instead, the claimants said this. “339. Issues 28 to 35 are concerned with the Claimants’ entitlement to the national minimum wage. As has been explained, the parties have agreed that issues 29, 31 and 34 should not be decided at this hearing.340. AL complains that the TCs have not provided schedules of loss. They were not directed to do so.341. That is because such schedules will only be relevant to the remedy issues (Issues 29 and 31 in the LoI) – indeed the schedules can only be sensibly written once the determinations of principle have been made by the ET as to: 341.1. which categories of hours ‘count’ as ‘working’ in respect of each TC, for the purpose of NMWR (Issues 3-5); 341.2. what category(ies) of payment ‘count’ as ‘remuneration’ in respect of the TCs, for the purposes of Regs 8-10 NMWR (Issue 30(a)); and 341.3. what categories of ‘reduction’ should be applied to reduce that ‘remuneration’, pursuant to Reg 11 NMWR, and in particular what ‘deductions and payments’ made by the TCs are to be treated as ‘reductions’, pursuant to Reg 13 NMWR (Issues 30(b) and (c)).342. By virtue of s. 28(2) of the NMWA, the burden is on AL to show that the TCs were remunerated at a rate equivalent to the national minimum wage.”[325]The respondent, however, did address issue 28; its answer was that the work of each claimant was “unmeasured work” for the purposes of the national minimum wage legislation. I saw no reason to disagree with that proposed answer in principle. However, I did not think that it was necessary to state a firm conclusion in that regard at this point, given that it would be best to test it against the facts of a particular driver’s case and in the light of specific submissions from the claimants. Issue 30: What was the remuneration of each Claimant in respect of each pay reference period? In particular:(a) What remuneration was paid by the employer as respects the pay reference period?(b) Were there any reductions within the meaning of regulations 11 to 15 of the NMWR 2015/regulations 31 to 31 of the NMWR 1999?(c) In considering the remuneration of each Claimant the Tribunal may need to consider (inter alia): (i) Whether the Respondent made any deductions or took any payments from Claimants (including any points accumulated that could be exchanged for vehicle hire credits) constituting expenditure in connection with employment and so a reduction within the meaning of reg 13 NMWR 2015 / 32 NMWR 1999? (ii) Did any payments made by the Claimants to third parties constitute expenditure in connection with employment and so a reduction within the meaning of reg 13 NMWR 2015 / 32 NMWR 1999?[326]The respondent’s answer to those questions was simply that “The Cs have not advanced evidence to allow the resolution of these issues.” The respondent then, presumably on the same basis, responded that issues 31 and 32 were “N/A”.[327]Issue 31 was not being determined by me at this stage, so that submission had to relate only to issue 32, which I now address. It seemed to me that the resolution of issue 32 would resolve also issue 30. Issue 32: If so, is the additional remuneration to which the relevant driver is entitled under section 17 of the National Minimum Wage Act 1998 the amount described in section 17(2) or the amount described in section 17(4) of that Act?[328]The claimants submitted in paragraph 353 of their closing submissions that “when their remuneration, determined in accordance with Part 4 NMWR, is divided by their working hours, determined in accordance with Part 5 NMWR, it is evident that they have been paid less per hour than the NLW” (which was probably a reference to the national minimum wage, so that the letters should have been “NMW”) but that the parties had “agreed that whether this was so in the case of any individual TC is to be determined at the remedy hearing”. The claimants then, in paragraphs 354-356 of their closing submissions made it clear that the issue here was whether or not (and these words are in paragraph 354 of those submissions) ‘the notional deduction of £0.45 from the ‘mileage rates in the Driver Scheme’ in respect of each ‘Job Mile’ purportedly ‘to cover vehicle related expenses’ should be viewed, for the purposes of Reg 13 NMWR, as “payment(s) paid to the worker by the employer”, such that the drivers’ “expenditure” by way of monies paid to AL or any other person(s) in connection with their employment have to be reduced (or eliminated) to the extent to which that “expenditure” was “met, or intended to be met” by the notional deductions of £0.45.’[329]The claimants’ next paragraph in their closing submissions was to the effect that either the329.1 “the entirety of the “total job earnings” paid to the TCs goes towards satisfying AL’s obligation to pay them the NLW in each pay reference period; but the TCs can set off (inter alia) all of their vehicle-related payments and deductions as ‘reductions’”, or329.2 “the notional deductions of £0.45 do not go towards satisfying AL’s obligation to pay them the N[M]W in each pay reference period, pursuant to Reg 10(l)21 – in other words, only the notional “qualifying earnings” as opposed to the “total job earnings” would constitute ‘remuneration’; but AL would then be able to set off the difference as ‘payments’ falling within the scope of Reg 13(2)”, and that the result was the same, whichever of those two possibilities was correct.[330]It seemed to me that the claimants were correct to say that. Issue 33: In so far as any Claimant is entitled to additional remuneration under section 17 in respect of any pay reference period, has the Respondent made a series of unlawful deductions from his wages contrary to sections 13 and 23 ERA, comprising the difference between his actual pay and the additional remuneration to which he is entitled (having in particular reference to the criteria indicated in Chief Constable of the Police Service of Northern Ireland v Agnew [2023] UKSC 33 especially at paragraph 127)?[331]The claimants referred in their closing submissions to this issue as number 31, although they plainly meant to refer to it as issue 33. They submitted simply (paragraph 371 of their closing submissions) this: “The effect of the Supreme Court’s decision in Agnew has been considered in the preceding paragraphs. The LDCs submit that in failing to pay them the national minimum wage at all material times, AL has made a series of unlawful deductions from their wages.”[332]I agreed. This was another illustration of the raising of an issue which was best determined by reference to the facts of a particular case and was in reality an academic issue unless it was applied to the facts of a particular case. In conclusion[333]The issues put before me for determination were in some respects apt and in others not in a form which enabled or required me to come to a conclusion. I have concluded that if the parties are unable to agree on the outcome of the application of such conclusions as I have stated above, then there will have to be a further hearing to determine the claims of the test claimants. It appears that that was always expected by the claimants themselves, so that conclusion will, I suspect, be welcomed by them.[334]For the avoidance of doubt, in my judgment the holding of a further hearing to determine those claims is in the interests of justice, not least because the focus of the hearing before me which started (in the event) on 29 October 2024 was for the most part on the issue of status, with the respondent putting before me much evidence of a general sort, much of which was of very little evidential weight (that which was not direct evidence) and much of which was of no evidential value at all (such as comments on the evidence of others).[335]In order to avoid the matter drifting, I have concluded that the parties should have three months from the date when this document is sent to them within which to attempt to agree the terms of judgments for the test claimants, and that if they are unable to come to such agreement within that time then they should ask the tribunal to arrange a hearing via CVP to be conducted by me on the soonest date which is convenient to the parties and me. The purpose of that hearing would be for me to(1) discuss with the parties why they had not reached agreement,(2) agree with them the starting day of, and a time estimate for, a further hearing, and(3) give directions for the steps to be taken in preparation for that further hearing. EMPLOYMENT TRIBUNALS Claimants Respondent Mr H Afshar & Others v Addison Lee Limited Heard at: Watford, in person On:18, 19, 20, 23 and 24 February 2026 Before: Employment Judge Hyams, sitting alone Representation: For the claimants: Mr Oliver Segal KC and Ms Melanie Tether of counsel For the respondents: Mr Richard Leiper KC and Ms Sophie Belgrove of counsel CORRECTED SECOND RESERVED JUDGMENT ON A NUMBER OF PRELIMINARY ISSUES[1]An ‘Empty’ period in the respondent’s Execution data for each lead claimant should only be treated as not being working time for the purposes of these claims where the period was in excess of 4 hours after a work shift, which is for this purpose to be regarded as having ended when the lead claimant was last shown as being “active” in the respondent’s Execution data.[2]Mr Klepacki’s working hours(1) between 3 March 2012 and 13 January 2014 are to be calculated by taking the hours worked between the start of the first job on a shift and the start of the last job on that shift and increasing the relevant number of hours by 18% to account for time logged on (ii) before the start of the first job in a shift and (ii) after the start of the last job in a shift;(2) during 2008 were on average 50 hours per week; and(3) between January 2009 and 2 March 2012 were the same as his average working hours from 3 March 2012 to 29 March 2013.[3]The employment relationship between the lead claimants and the respondent ended when, viewed objectively, the parties intended that relationship to end.[4]There was no break in the employment relationship for the purposes of the Working Time Directive between(1) Mr Ruiz and the respondent between 22 March 2020 and 9 February 2021, and(2) Mr Mahendran and the respondent between (i) 16 March 2020 to 28 August 2020 and (ii) 2 January 2021 to 16 April 2021.[5]Any holiday which the lead claimants took in any holiday year of up to 5.6 weeks’ duration must be treated as having been taken pursuant to the Working Time Regulations 1998, SI 1998/1833 (“the WTR”).[6]The change made by regulation 10(3)(b) of the Employment Rights (Employment Particulars and Paid Annual Leave) (Amendment) Regulations 2018/1378, with effect from 6 April 2020, to regulation 16 of the WTR in the form of the insertion of new paragraphs (3)(e) and (f) affected all claims determined on or after that date, irrespective of the date when the claim was made or the period to which the claim related.[7]Tips which were paid by credit card or via a customer’s account with the respondent and processed by the respondent before being paid to the lead claimants formed part of their “remuneration” within the meaning of section 224(3) of the Employment Rights Act 1996 (“ERA 1996”).[8]Vehicle hire “rental credits” must be taken into account when determining what the lead claimants received by way of income over the relevant preceding period (which, given paragraph 6 above was 52 weeks).[9]The respondent’s continuing failure to pay holiday pay for leave taken on an unpaid basis gave rise to a series of deductions which continued throughout the lead claimants’ respective employment relationships.[10]All the earnings of a lead claimant in a relevant week are to be taken into included when determining that claimant’s remuneration for the purposes of the WTR, even if the claimant worked more than five days in that week, with the result that that remuneration should have been paid for a week’s holiday taken by that claimant pursuant to the WTR.[11]The lead claimants’ entitlement to holiday pay must, where material, be calculated by reference to what they were entitled to receive by way of the national minimum wage.[12]The lead claimants’ “employment” with the respondent having ended, the lead claimants have the right to receive a payment in lieu of all leave to which they were entitled under regulation 13 of the WTR during their claim periods, whether they had(1) taken that leave on an unpaid basis or(2) failed to take it.[13]The respondent in that regard made a series of unlawful deductions from the lead claimants’ wages within the meaning of section 23(3) of the ERA 1996, whether the right to take the relevant leave arose under regulation 13 or regulation 13A of the WTR.[14]The lead claimants at all material times had a pay reference period of a week.[15]Where any return of a deposit or part of a deposit made by a lead claimant has been treated by the respondent as a ‘payment’ made by it to the lead claimant, any deduction made from the lead claimant’s pay towards such a deposit must be treated as a deduction within the meaning of regulation 13 of the National Minimum Wage Regulations 2015, SI 2015/621 (“the NMWR 2015”).[16]The lead claimants’ work-related fuel expenses as non-partner drivers are to be calculated as their actual fuel expenditure less their expenditure on fuel for private usage. The latter expenditure must be calculated on the basis of an estimated percentage, or alternatively on the basis of an average weekly mileage. On the evidence before me,(1) Mr Ruiz’s private mileage was on average 40 per week,(2) Mr Mahendran spent on average £16.40 on fuel for private usage per week, and(3) Mr Klepacki’s average expenditure on fuel for private purposes was £9.82 per week.[17]Mr Ruiz’s and Mr Klepacki’s work-related fuel expenses as partner drivers must be calculated on the basis of the percentages of time they spent driving for the respondent (out of the total driving time for all operators) which were(1) 20% and(2) 35% respectively.[18]The lead claimants’ mobile telephone costs (that is, charges for the handset and any network charges, but not additional costs) were expenditure within the meaning of regulation 13(1)(b)(ii) of the NMWR 2015.[19]Motoring fines and penalties (including parking fines and penalties) incurred by the lead claimants were not such expenditure.[20]All of the parking expenses claimed by the lead claimants were such expenditure. 21. 80% of the deductions from the respondent’s “Supplier Payment” made to Mr Ruiz in respect of the cost of self-assessment services provided by Cab Accounting Ltd to him via an arrangement established by the respondent are to be treated as deductions within the meaning of regulation 13 of the NMWR 2015.[22]Payments made by the respondent to the lead claimants of an “Expense Allowance of £0.45 per Job Mile” are to be treated only as money earned by the lead claimants before the deduction of expenditure within the meaning of that regulation.[23]Where the expenses which a lead claimant incurred in a particular pay reference period for the purposes of the NMWR 2015 exceeded his pay, the hourly rate must be calculated as nil (and not as a negative figure).[24]The lead claimants are entitled to interest on any sums awarded to them in respect of their claims for holiday pay at 8% per annum.[34]AL now says that its Execution Data may not constitute an accurate record of the hours during which its drivers were ready and willing to work, because certain technical deficiencies in its software may have had the result that changes in their working status were not correctly recorded. If, however, AL had taken steps at the material times to ensure that its pay system took account of the drivers’ rights to receive the NMW and had complied with its obligation to ensure that “relevant records” were maintained, it would have been required to put in place a system which ensured that comprehensive contemporaneous records of the drivers’ working hours were made and retained.[35]Further, it would have been required to obtain and retain records of all the reductions mandated by regulation 13 NMWR, including sums which the drivers had paid to third parties as respects their expenditure in connection with their employment.[36]AL should not be permitted to take advantage of its own failure to comply with its obligations under the NMW legislation by contending that the records it chose to make of their working hours are deficient or by placing onerous and/or misconceived burdens on the LCs in relation to proof of their expenses. Where an employer has elected to ignore its workers’ rights under the national minimum wage legislation, the effect of the burden of proof in section 28(2) of the NMWA must be that doubt or ambiguity which would not have arisen if the employer had complied with its duty to make relevant records should be resolved in the workers’ favour.” 17 At the start of the February 2026 hearing, I said that I had been looking at chapter 11 of the current (22nd) edition of McGregor on Damages (“McGregor”), and that I would be looking at it again when determining the issues which were going to be determined by me after that hearing. Neither party (treating the claimants as a single party for this purpose) referred to any part of that chapter subsequently, but neither party objected to me referring myself to, and taking into account, what was said in McGregor. 18 There, having referred, in paragraph 11-005, to IRT Oil and Gas Ltd v Fiber Optic Systems Technology (Canada) Inc [2009] EWHC 3041 (QB), this was said (and I have incorporated the content of the footnotes into the text, and taken the footnote numbers out): ‘11-006 Further cases presenting this difficulty of showing the amount of profitable sales that would have been made by the claimant had there been no tort or breach of contract by the defendant have indicated that the claimant is assisted by the principle in the very old case of Armory v Delamirie, (1722) 1 Str. 505, which has today received a new lease of life, the principle being that the court is required to resolve uncertainties by making assumptions generous to the claimant where it is the defendant’s wrongdoing which has created those uncertainties. That principle was endorsed by the Supreme Court in Morris-Garner v One Step (Support) Ltd [2018] UKSC 20; [2019] A.C. 649 at [38], but there remains doubt about the extent of its application. The doubt concerns whether the principle is one of general application or whether it applies only to assist the judge in resolving evidential disputes concerning quantification. The remarks of Lord Reed JSC in Morris-Garner suggest the latter. So too, do the remarks of Longmore LJ (with whom Arden and Aikens LJJ agreed) in Keefe v The Isle of Man Steam Packet Co Ltd, [2010] EWCA Civ 683 at [19], adopted in Fluor v Shanghai Zhenhua Heavy Industry Co Ltd, [2018] EWHC 1 (TCC) at [53], that when the defendant’s breach of duty makes it difficult or impossible for a claimant to adduce relevant evidence as to quantification the defendant “must run the risk of adverse factual findings”. To similar effect are the remarks of Sir Michael Burton in MAD Atelier International BV v Manes [2021] EWHC 3335 (Comm) at [30] that where a defendant’s wrongdoing has created uncertainties, “in the case of a loss whose existence has been proved, the Court will err on the side of generosity in the calculation of it if its exact calculation is difficult if not impossible to arrive at”. 11-007 The basis of the principle as one that resolves evidential disputes as to quantification but does not dispense with a need to adduce reasonably available evidence can be seen in the refusal of Hamblen J to apply the Armory v Delamirie principle to the facts in Porton Capital Technology Funds v 3M UK Holdings Ltd [2011] EWHC 2895 (Comm). In 2007 the defendants in Porton agreed to buy the entire shareholding of a company from the shareholder claimants. The consideration for the purchase was a cash sum together with what was referred to as an earn out payment based on net sales for the year 2009, the defendants being required to continue the business. The earn out payment was the principal return to be made by the claimants from the sale. The defendants in breach of contract closed down the business so that there were no sales in 2009.The claimants, relying on Armory v Delamirie, argued that, since the very actions of the defendants in breaching the contract had made the quantification of damages more difficult, the court should resolve any uncertainties in the claimants’ favour. Hamblen J did not accept the argument. The claim was for lost profits for breach of contract where there were evidence and documentation relating to the claim; the evidential playing field was a level one. The Armory v Delamirie principle should not be extended further than its evidentiary basis permits. 11-008 To the extent that some cases can be read as taking a broader approach or suggesting that the rule is one of substance rather than a manner of evaluating evidence (Fearns v Anglo-Dutch Paint & Chemical Co Ltd [2010] EWHC 1708 Ch at [70] and Double Communications Ltd v News Corp International Ltd [2011] EWHC 961 (QB) at [4]–[5]), that broader approach should not be accepted. On one view, the Armory principle was expanded in this illegitimate way by the Court of Appeal in Gulati v MGN Ltd [2015] EWCA Civ 1291; [2017] Q.B. 149 at [107], where the Court of Appeal upheld sizeable awards for phone hacking despite the absence in some cases of any records detailing the hacking and other wrongful activities. In the judgment with which the other judges agreed, Arden LJ said that the award of damages was: “… an example of the ability of the law to prevent a person responsible for wrongdoing from escaping liability to his victim, without disturbing the general rule as to the conditions of liability.” On an expansive view, Arden LJ might be read as suggesting that a claim can be brought for loss caused by hacking against a proved hacker without proving any of the instances of hacking. This cannot be what Arden LJ meant. Her Ladyship must have been suggesting that in all of the circumstances of that case, inferences could be drawn of the nature and extent of the hacking which occurred based on the extent of related wrongdoing by MGN.’ 19 The most authoritative of the cases referred to in that passage was MorrisGarner v One Step (Support) Ltd. Paragraph 38 of the judgment of Lord Reed JSC was in these terms: ‘Evidential difficulties in establishing the measure of loss are reflected in the degree of certainty with which the law requires damages to be proved. As is stated in Chitty, para 26-015: “Where it is clear that the claimant has suffered substantial loss, but the evidence does not enable it to be precisely quantified, the court will assess damages as best it can on the available evidence.” In so far as the defendant may have destroyed or wrongfully prevented or impeded the claimant from adducing relevant evidence, the court can make presumptions in favour of the claimant. The point is illustrated by Armory v Delamirie (1721) 1 Str 505, where a chimney sweep’s boy found a jewel and took it to the defendant’s shop to find out what it was. The defendant returned only the empty socket, and was held liable to pay damages to the boy. Experts gave evidence about the value of the jewel which the socket could have accommodated, and Pratt CJ directed the jury “that, unless the defendant did produce the jewel, and shew it not to be of the finest water, they should presume the strongest against him, and make the value of the best jewels the measure of their damages: which they accordingly did.”’ 20 I saw too that paragraph 240 of the judgment of Hamblen J in Porton was in these terms: ‘In a different context, the Court was required in Fearns v Anglo-Dutch Paint & Chemical Co Ltd [2010] EWHC 1708 (Ch), to consider the damages to which the claimant was entitled as a result of the defendants’ unlawful conduct in inducing the claimant’s franchisees to buy paint directly from the defendants. George Leggatt QC (sitting as a Deputy Judge) rejected the claimant’s argument that the defendants’ conduct had caused the collapse of his business. He held that it was nevertheless necessary to determine what loss had been suffered, and noted at [70] that while this “necessarily involves a large element of conjecture”, the need for such conjecture “is itself a consequence of the Defendants’ conduct”. The Deputy Judge held that the principle in Armory v Delamirie applied and required the Court “to resolve uncertainties by making assumptions generous to the Claimant where it is the Defendant’s wrongdoing which has created those uncertainties”. He noted that this accorded with the principle applicable to the assessment of damages in a case of patent infringement, that while the object remains to compensate the claimant and not punish the defendant, “damages should be liberally assessed”: see Lord Wilberforce in General Tyre & Rubber Co v Firestone Tyre & Rubber Co Ltd [1975] 1 WLR 819 (cited by the Deputy Judge at [24]).’ Retrospective effect 21 During the February 2026 hearing, I suggested that there might be a problem with retrospective effect only if a statutory provision took away a right of any value. In oral closing submissions, Mr Leiper referred me to paragraph 10-061 of the current (13th) edition of Craies on Legislation. That paragraph, including the very much longer text in the footnote to its second sentence, had to be read along with the text following it, through to paragraph 10-073. One of the more helpful parts of that passage was what was said in paragraph 10-069, together with its footnote’s text, which I have incorporated in the following extract: ‘[R]etrospection may or may not affect vested or accrued rights, and the presumption against retrospective application and the other relevant rules of construction and application will apply in a stronger form depending on whether or not existing rights are affected. See, for example: “In this case, the form of retroactivity contended for is not the strong form where vested or accrued rights are retrospectively taken away; but a weak form involving only the application of an adjusted sanctions regime to conduct of the type which, in a broad sense, was already the subject of a similar though narrower sanctions regime at the time of the conduct complained of. The significance of the distinction between these two forms of retroactivity was emphasised in Wilson v First County trust Limited (No 2) [2004] 1 AC 816, per Lord Hope at paragraph 98, referring to R. v Field [2003] 1 WLR 882.”—McTier v Secretary of State for Education [2017] EWHC 212 (Admin); [2017] A.C.D. 46.’ 22 I also found the following part of paragraph 10-071 to be very helpful: ‘The principal difficulties in the application of the presumption against retrospectivity are expounded in the passage of the speech of Lord Mustill in L’Office Cherifien des Phosphates v Yamashita-Shinnihon Steamship Co Ltd, The Boucraa [1994] 1 All E.R. 20, HL, appended to this work. The essence of that decision is that the question of whether a statute is intended to have retrospective effect should be answered not by the application of any rigid criteria or distinctions between classes of enactment, but by considering, in the light of all the circumstances of each case—• what degree of unfairness (if any) might be thought to be suffered if the provision were applied with retrospective effect; and• that the greater the unfairness the stronger the presumption that Parliament would not have intended it, and therefore the greater the clarity of language required to rebut it. This approach accords with a general trend of the courts (see, in particular, Ch 18) away from the rigid application of formulaic presumptions and the application of common sense in the search for the legislative intention in each context, by having regard to all relevant circumstances. See, for an example of the House of Lords accepting and applying this approach (in the case of considering the retrospectivity of s.3 of the Human Rights Act 1998 (c.42), Wilson v Secretary of State for Trade and Industry, Wilson v First County Trust Ltd (No.2) [2003] 3W.L.R. 568, HL.’ 23 Reference is then (in paragraph 10-072) made to what Lord Mustill referred to in The Boucraa as “the special distinction between procedural and substantive provisions for these purposes”. That distinction is addressed in paragraphs 10- 072 and 10-073 and it was in my view of particular interest that paragraph 10-073 ended with the following words: ‘Not only is there no sign that the strength of the general understanding that procedural matters are an exception to the undesirability of retrospectivity is diminishing, there are even signs that it is turning into a positive presumption in favour of retrospection. In R. v Makanjuola (1995) 2 Cr. App. R. 469 at 472A–B, Lord Taylor of Gosforth CJ said: “The general rule against the retrospective operation of statutes does not apply to procedural provisions … Indeed, a general presumption is that a statutory change in procedure applies to pending as well as future proceedings.” See also R. v Bradley [2005] EWCA Crim. 20.’ The issues which the parties and I agreed should be determined by me after the February 2026 hearing 24 The claimants made it clear in their closing submissions that they were not asking me to calculate the lead claimants’ losses. What they were asking for was decisions on points of principle in almost all respects, but in one or two respects with reference to the facts of the cases of one or more of the lead claimants. I welcomed that, since it made my task much less onerous than it would have been if I had been required by the parties to determine the specific amounts to be awarded to the lead claimants. 25 The parties were clear that a large amount of work had been done between the 2024 hearing and before the hearing of February 2026 and that once I had made my determinations of the issues which I have listed in paragraph 8 above, the parties would be able to calculate and almost certainly agree the outcomes of all of the claims. 26 In the rest of these reasons, I address each of the issues which the parties agreed were for determination by me at this stage in turn. I do so by stating the issue, i.e. the question for determination, and then the claimant’s proposed answer to that question. I then discuss the parties’ submissions in relation to the issue, including where those submissions were supported by evidence, the relevant evidence, stating my conclusions in response to those submissions. In the course of addressing each issue, where necessary I refer to the relevant case law and statutory provisions without setting them out unless it would help to set them out at that point. Issue 1: What were the periods of working time within the meaning of the WTR and the NMWA for each lead claimant? 27 By the time of closing submissions, there were, the claimants said, two particular issues which needed to be determined here. The first related to what the claimants referred in paragraph 1 of the appendix to their closing submissions as “periods of ‘Empty’ as recorded in the Execution Data for non-partner drivers”. The second issue, stated in the same paragraph, related “to [Mr Klepacki’s] working hours for the period before [the respondent] has the relevant data.” The claimants’ proposed answers to those two sub-issues 28 The claimants’ proposed answers to those sub-issues were, respectively:28.1 that I should find that ‘Empty’ periods should only be disallowed where the period is in excess of 4 hours after a work shift”; and28.2 Mr Klepacki’s working hours28.2.1 between 3 March 2012 and 13 January 2014 are to be calculated by taking the hours worked between the start of the first job on a shift and the start of the last job on that shift and increasing the relevant number of hours by 18% to account for time logged on (ii) before the start of the first job in a shift and (ii) after the start of the last job in a shift;28.2.2 during 2008 were on average 50 hours per week; and28.2.3 between January 2009 and 2 March 2012 should be estimated to be the same as his average working hours from 3 March 2012 to 29 March 2013. The claimants’ submissions in relation to sub-issue 1 29 The claimants’ original position was that ‘Empty’ periods should only be disallowed where the period was in excess of 10 hours after a work shift. In paragraph 53 of their closing submissions, they said that the lead claimants’ working hours had been ‘calculated for the purposes of their Schedules of Loss on the basis that ... all Empty periods count as working time other than Empty periods of more than 10 hours, which may reasonably be assumed to be due to have arisen due to some form of error. Each of the LCs’ SOLs [i.e. each of the lead claimants’ schedules of loss] includes a worksheet entitled “C’s empty periods over 6 hours” which identifies the Empty periods which were excluded when calculating the relevant LC’s working hours.’ 30 By the time of closing submissions, the claimants’ position was stated (in paragraph 54 of their written closing submissions) in the following manner (the emphasis being original). “In light of the evidence the Tribunal has heard, the LCs submit that a more nuanced approach would be appropriate. They propose that ‘empty’ periods should only be disallowed where the period is in excess of 4 hours after a work shift (that is, the period within any 24 hours when the driver was clearly working). The latter periods are likely to have arisen due to an error (though this is not certain: the driver could have been driving ‘empty’ looking for work for more than 4 hours at the end of a shift and then just given up and logged off).” 31 In both cases the number of hours was not based on an identifiable evidential factor. However, the claimants submitted (in paragraphs 59-60 of their closing submissions) that “whether or not any period(s) of Empty time should be discounted when assessing the LCs’ working time is a matter for evidence”, and that the evidence before me “demonstrate[d] that all of the LCs experienced periods of Empty time which significantly exceeded the thresholds suggested by [the respondent]”, and referred to paragraphs 19-25 of the witness statement of Mr Ruiz (that is to say, the statement that he made for the February 2026 hearing; any reference below to a witness statement made by someone who gave evidence at the February 2026 hearing is unless otherwise stated a reference to a statement made for the purposes of that hearing), paragraphs 23-28 of the witness statement of Mr Klepacki, and paragraphs 20-39 of the witness statement of Mr Mahendran. I found the evidence of Mr Klepacki to be the most cogent evidence in support of the proposition that drivers might have to wait for up to four hours between jobs, if only because Mr Ruiz and Mr Mahendran tended to seek jobs at airports, and the respondent accepted that time spent in the “virtual queue” at an airport was working time. The specific times to which all three lead claimants referred in their witness statements were derived from the respondent’s digital records in the form of what the respondent called its “Execution Data”. As a result, those specific times were not the subject of a specific memory on the part of the lead claimants. That was not surprising, given the fallibility of memory and the fact that the specific times were long in the past. The respondent’s proposed answer to sub-issue 1 32 The respondent’s proposed approach to this sub-issue was to apply (as it was said in paragraph 27 of the respondent’s opening skeleton argument for the February 2026 hearing) a “statistical model”. The whole of that paragraph bears repeating here. ‘In order to identify how much of the “Empty” time should be excluded, the Respondent has applied a statistical model which demonstrates to a high degree of probability that the Claimants were not ready and willing to work during certain periods. The Respondent has excluded any excess periods of time which are greater than this figure from the calculation of working time. The statistical model is set out and illustrated, using HH as a worked example, at [FHB/4655 ¶ (9.10-9.11)].’ 33 Paragraph 9.10 on page 4655 of the FHB was based on the proposition that “It is empirically accepted in statistics that 95% of data points are expected to fall within two standard deviations of the sample average and the other 5% of data points are considered outliers or unusual due to their distance from the average.” The result of the application of “a standard deviation model which identifies statistically significant excess empty time for drivers during periods in which they were nondriver partners” was that the “Excess Empty Time” for each of the three lead claimants was as follows:33.1 for Mr Klepacki it was 60 minutes;33.2 for Mr Ruiz it was 93 minutes; and33.3 for Mr Mahendran, it was 88 minutes. 34 Accordingly, any empty period which exceeded that time in the case of each lead claimant was ignored by the respondent in its calculation of the working time of that claimant. So, for example, in the case of Mr Ruiz, if he “was recorded as ‘empty’ for more than 93 minutes the total time from the status has been deducted from calculation of working hours as the Respondent contends the presumption should be that the Claimants were not ready and willing to work despite their driver profile remaining logged on to the Respondent’s driver portal in this period”. 35 The “Average Empty time” for each lead claimant was recorded in the table in paragraph 9.10 on page 4655 of the FHB as 15 minutes for Mr Klepacki, 21 minutes for Mr Ruiz, and 19 minutes for Mr Mahendran. In paragraph 27 of the respondent’s written closing submissions, those figures were said to have been explained by Mr Davies’ evidence, but in reality what he did was (not unhelpfully, I add) to state the manner in which the respondent’s digital systems had recorded things from which a conclusion could be derived about the work done by the claimants. Those systems had not been configured with a view to recording that working time. Aspects of the oral evidence of Mr Davies on which the respondent did not rely but to which the claimants responded in their closing submissions 36 The respondent did not, in Mr Leiper and Ms Belgrove’s closing submissions, rely on the evidence of Mr Davies about the possibility of a driver’s status being shown on the respondent’s Shamrock system as “empty” when in fact the driver had switched off his MDA without the switching off resulting in the driver being logged off Shamrock. There was some strong evidence of that as having occurred, and, as the claimants said in paragraph 72 of their written closing submissions, it had “always been accepted that this might happen unusually at the end of a working shift (the sole example given by RD at RD/52 of RK not logging off at the end of his shift 23/24 September 2018, and therefore showing ‘empty’ until 5 October 2018, is obviously of that type).” However, Mr Davies said during crossexamination, and this was not foreshadowed anywhere in the respondent’s evidence and it was not put to the lead claimants in cross-examination, that a driver might have done that in the middle of a working shift rather than put himself on “break”. The claimants submitted (in footnote 3 on page 20 of their closing submissions) this in response. “It is difficult to comprehend why/how AL [i.e. Addison Lee, the respondent] would devise/implement a system whereby if the MDA (and therefore the app) was switched off, that would not register ‘live’ as the driver logging off. Moreover, if this had been an at all common occurrence, it seems likely that AL would have notified drivers that they should not switch off their phone or MDA without first logging off.”[37]In addition, in paragraph 85 of their closing submissions, the claimants pointed out that I had, in paragraphs 73-74 of my 2025 reasons, found (on the basis of the evidence of Mr Kelly to which I referred in those paragraphs) that a driver who refused to accept a job which had been allocated to him would routinely be logged off for an hour. The claimants continued: “The suggestion that drivers could avoid this outcome by switching off their MDA has never been explored with any claimant, including the LCs, and is implausible in relation to specific jobs, because the log off described by Mr Kelly would be triggered by the driver’s failure to view the job (there being no button he could press to reject the job), so that switching off the MDA would be incapable of preventing the driver being logged off.”[38]The respondent also did not rely in closing submissions on a second possibility which was first asserted by Mr Davies in cross-examination. That was (as the claimants said in paragraph 77 of their closing submissions) “where the driver had opened [on his mobile telephone] another app which was running in the foreground rather than the background, it being suggested that this could cause the [respondent’s] App to stop communicating with Shamrock”. The claimants’ closing submissions continued: “77. ... In his witness statement, at RD/52(c), RD gave as an example a partner driver who had another ride app open. He did not appear to suggest that this scenario had any application to non-partner drivers. However, in xex he suggested that a non-partner driver who had Google Maps or Waze operating in the foreground could compromise Shamrock’s ability to communicate with the AL app.78. This too was a scenario which was not put to the LCs in xex. It is also something which might reasonably have been expected to have been the subject of communications from AL to the drivers pointing out that opening another app might compromise Shamrock’s ability to communicate with them. There is no evidence that any such communication or Fact Sheet was sent to them.79. In any event, it is implausible that any driver who had ceased work, whether for the day or because he was taking a break, would have Google Maps or Waze running in the foreground.” My conclusion on sub-issue 1[39]I accepted the claimants’ submissions and their proposed answer to this first subissue, namely that ‘Empty’ periods should only be disallowed where the period was in excess of 4 hours after a work shift. That was for the following reasons.39.1 I accepted the evidence of Mr Klepacki to which I refer in paragraph 31 above and which pointed in that direction.39.2 I could not see how I could lawfully, on any view, without hearing expert evidence from a statistician, accept statistical evidence from the respondent about the likelihood of any of the claimants having to wait any particular time before receiving a job.39.3 There had been no order for the provision of expert evidence here.39.4 The principle stated by Lord Reed JSC in Morris-Garner which I have set out in paragraph 19 above was in my view applicable here and provided strong support for the comments of Judge Barry Clarke in Commissioners for Revenue and Customs v Lees of Scotland Limited [2024] EAT 120 and the submissions of the claimants which I have set out in paragraph 16 above. The claimants’ submissions in relation to sub-issue 2[40]What the claimants said in paragraphs 89-93 of their closing submissions in relation to the working hours of Mr Klepacki was in almost the same terms as the relevant part of their opening skeleton argument, showing that their position had changed in only minor ways. That is not a criticism. It showed that the claimants did not regard the respondent’s evidence or opening submissions as justifying any major departure from their (the claimants’) opening position. I do not need to repeat paragraphs 89-93 of the claimants’ written closing submissions here. I return to them when stating my conclusions on this issue below. The respondent’s submissions in relation to sub-issue 2[41]The respondent’s “primary position [as stated in paragraph 42 of its closing submissions was] that the period for which working time [had] been estimated by the Claimants using job data [was] manifestly out of time”. But, subject to that primary position, as stated in paragraph 9.17 of the respondent’s counterschedule of loss on pages 4656-4657 of the FHB, it was the respondent’s case that41.1 “No waiting time should be assumed for before the first Job of a day as typically the first job would be a pre-booked Job to which no waiting time would apply.”41.2 “The time period between a Job and a subsequent Job within a ‘shift’ should be limited to a maximum of(i) a reasonable estimate of the time required to complete that particular job plus(ii) the EET [i.e. empty extra time] threshold. Time in addition to this should be disregarded consistent to the approach to EET. This avoids assuming that a driver will work a lengthy shift with no break.”41.3 “No uplift of 18% [as contended for by the claimants: see paragraph 45 below] should be applied, on the basis that it is not reasonable to assume that the length of a job is proportionate to the amount of time which a driver spends on ‘shift’ and without any reference to the nature of that Job. (For example, this would result in a consistent material over-counting of time where a driver consistently takes a relatively short job at the end of his shift before going home.) Instead the Respondent contends that the time estimated for the last Job should be a reasonable estimate of the time required to complete that particular Job.”[42]In support of that position, the respondent added, in paragraph 47 of Mr Leiper’s and Ms Belgrove’s closing submissions: “The Job Data shows the length of the last job of the day (pick up and drop off points) so an approximate length of time for that journey can be calculated (rather than simply applying a percentage across all of the data). As Mr Klepacki accepted, if it’s a short journey in central London, it is going to take less time than a long one across central London.” A discussion about sub-issue 2[43]I found paragraphs 89-93 of the claimants’ written closing submissions to be cogent, but I also found what the respondent said in response to be at least at first sight cogent.[44]There was no record of any kind of Mr Klepacki’s working time between 26 June 2008 and 2 March 2012, and only a partial record for the period between 3 March 2012 and 13 January 2014. The partial record was the respondent’s “Jobs Data”. There were no “Execution Data” in respect of that second period.[45]In paragraph 90 of their written closing submissions, the claimants said this. “The way in which RK’s working hours have been calculated in both these periods is explained in RKSOL 9-11. His working hours between 3 March 2012 and 13 January 2014 have been calculated by taking the hours worked between the start of the first job on a shift and the start of the last job on that shift and increasing the relevant number of hours by 18% to account for time logged on (ii) before the start of the first job in a shift and (ii) after the start of the last job in a shift.”[46]The respondent’s answer to that was in part in Mr Leiper’s oral submissions emphasising the need for caution in relation to the evidence of Mr Klepacki about his working hours, and in part was in paragraph 45 of the respondent’s written closing submissions, which was in the following terms (with my inserted text in curly brackets; the text in square brackets was originally there and in those brackets): “The Claimants have suggested an approach which takes an average where there is known data and applies it retrospectively (the precise approach appears to be described differently in his statement [FWB/35 ¶ 30] {i.e. paragraph 30 of Mr Klepacki’s witness statement for the February 2026 hearing, at page 35 of the bundle of witness statements prepared for that hearing} than in the SOL [FHB/4698 ¶ 9]). The Respondent considers that this approach significantly overstates the work undertaken in that period: see the table under ¶ 9.20 of the Respondent’s Principles [FHB/4657]. This demonstrates that in 2012-2013 Mr Klepacki’s working hours (orange line) unsurprisingly mirrored his pay (green line), but that his assumption for the earlier period would produce a figure for hours worked that is considerably higher than his total pay, which indicates that his assumption must be wrong. Indeed, Mr Klepacki accepted that the amount of work he undertook increased in 2012 (which is directly inconsistent with his assumed approach, which takes 2012 and applies it to the earlier period). His position is also that his cash payments had been consistently low, so any change in cash payments does not account for the difference.”[47]The claimants accepted in paragraph 93 of their written closing submissions that the approach taken in paragraph 30 of Mr Klepacki’s witness statement was an error. My conclusion on sub-issue 2[48]The respondent’s table at page 4657 of the FHB assumed that there was the same amount of work available to be done during the period from 2008 to 2012 as there was in 2012-2013. However, there was no evidence before me to support that proposition, and I was aware that there was in 2008 a financial crisis (the “credit crunch” which started in that year), which might have affected the availability of driving work during 2008-2012. Nevertheless, I concluded (without any regret) that I was bound to apply the principle described by Lord Reed JSC in Morris-Garner which I have set out in paragraph 19 above and that what the claimants said in paragraphs 91-93 of their closing submissions supported their proposed answer to the sub-issue sufficiently to justify me in accepted that proposed answer, namely as set out in paragraph 28.2 above. Those submissions were as follows. “91. AL contends at CSOLP [i.e. the document at FHB 4639-4683] 9.17 [which was at FBH 4656-4657] that the 18% uplift is excessive. In order to test whether there is any merit in this criticism, RK’s working hours for the period 18 January 2014 until 23 January 2015 (a period for which Execution Data is available) have been calculated using the applicable Job Data with an 18% uplift and compared with the actual working hours shown by RK’s Execution Data – see the worksheet in RKSOL entitled “Uplift job hours v hours logged on”.92. When the hours shown in columns H and K of this worksheet hours are compared, it is clear that the hours calculated using the Job Data with an 18% uplift are lower than the working hours recorded in the Execution Data in every week of the comparison period but one. This demonstrates that the 18% uplift is not only reasonable but likely to be an underestimate of RK’s total working hours in the period 3 March 2012 to 13 January 2014.93. RK’s instructions are that during 2008 he worked approximately 50 hours per week and that between 2009 and 2014 there was no significant change in his hours, so his average working hours from 3 March 2012 to 29 March 2013 have been used to estimate his working hours between January 2009 and 2 March 2012.” Issue 2: How far back in time is each lead claimant’s claim valid? In answering this question, it will be necessary to decide whether any substantial period (that is to say, over and above what might be expected to have been taken as holiday) during which the lead claimant was not logged onto the respondent’s Shamrock software system gave rise to a termination of the status of worker for the purposes of the WTR and the NMWA.[49]The claimants’ proposed findings on this issue were that49.1 whether a break between individual contracts should be treated as signifying the termination of a driver’s engagement for limitation purposes must depend on whether the break can be regarded as demonstrating that the driver’s ‘employment relationship’ with the respondent had come to an end. The purported existence of 12 week fixed term contracts, is of no relevance to this issue; and49.2 there was no termination for limitation purposes in any of the following circumstances:49.2.1 where a driver took more holiday than 5.6 weeks in a holiday year – whether he retained his hire vehicle during most/all of those periods or on the one occasion Mr Ruiz returned the vehicle (a fortiori telling the respondent why);49.2.2 where there was a continuous break in activity of four weeks or more which was accompanied by one or more of the respondent’s “plus” factors (most of which did not apply to the lead claimants in any event, except in relation to time off during the Covid-19 pandemic; as to which, see paragraph 49.2.4 below);49.2.3 where there was any continuous period of six weeks or more in which a driver did not work for the respondent (again, not relevant to the lead claimants save in relation to the pandemic);49.2.4 where a driver took an extended period off work because of the lack of available work and/or reasonable safety concerns during the pandemic; and49.2.5 where a driver changed from being a non-partner driver to a partner driver, without any break in work. A discussion about those proposed findings[50]As can be seen from the reference in paragraph 49.2.2 above to ‘the respondent’s “plus” factors’, the claimants’ proposed answer to issue 2 involved a number of negative conclusions, namely about when there would not be a termination for limitation purposes.[51]In fact, during oral submissions, Mr Segal pointed out that the claimants were not saying that there was an ongoing employment relationship, in the form of an overarching contract or “umbrella” contract, between the claimants and the respondent. Rather, what was being said on behalf of the claimants was that the lead claimants were “workers” for the purposes of both the NMW and the WTR when they were logged onto the respondent’s Shamrock software. So, what was being contended on behalf of the claimants was that for the purposes of section 23 of the ERA 1996, which applied to both the claims of a failure to pay the NMW and to pay holiday pay, there were unbroken series of deductions (applying the analysis in Chief Constable of the Police Service of Northern Ireland v Agnew [2023] UKSC 33, [2024] ICR 51, especially at paragraph 127), and that for the purposes of the WTR there was an ongoing employment relationship within the meaning of article 7(2) of the Working Time Directive and the case law discussed by the claimants in their closing submissions for the 2024 hearing as set out in paragraph 239 of my 2025 reasons.[52]It is therefore helpful to say here that the claimants’ position was (understandably) in favour of a long period of a continuous employment relationship for the purposes of the WTR and a long series of unlawful deductions, so that for the claimants the presumption was that any events which might be argued by the respondent to have interrupted that relationship and/or that series did not in fact do so. On the other hand, the respondent argued for the shortest conceivable length in the relationship or series, and argued for as many possible situations in which that relationship or series should be regarded as having caused a termination of the relationship or series. It is therefore helpful to refer to the respondent’s submissions first. The respondent’s submissions on issue 2[53]The respondent proposed seven situations in which there would be a termination of the employment relationship. They were set out on pages 4645-4647 of the FHB, and were, respectively:53.1 termination on notice;53.2 termination after the ending of a fixed term contract of the sort which, I had concluded in paragraph 321 of my 2025 reasons, was not “genuine”;53.3 “termination by four weeks inactivity plus” one or more of the following factors:53.3.1 “the driver has not received any payments from the Respondent in respect of Jobs during at least the last week of the Break Period;”53.3.2 “the Break Period includes at least one day in which the driver has declared they were not ready and willing to work due to holiday, but which cannot be statutory holiday under Regulation 13 or 13A WTR;”53.3.3 “the Break Period includes at least one day in which the driver has placed themselves in a position that they cannot work (e.g loss of licence/PCO, being locked out of the Respondent’s systems and/or classified as ‘inactive’);”53.3.4 “the Break Period includes at least one day which is after the end of the Driver Contract Fixed Term;”53.3.5 “for a non-partner driver, they have returned their rental car and/or their vehicle hire agreement has ended;”53.3.6 “for any driver, they have undertaken work under another engagement (including, for partner drivers, work undertaken as multiapping work);” or53.3.7 “any other circumstance ... indicating the intention to discontinue the engagement and/or [which] would otherwise frustrate the Driver Contract” “provided that, for contracts entered into after June 2021, where the Break Period arises during the Fixed Term an Engagement End Date shall only be deemed to have arisen where the driver also undertakes no Jobs for the remainder of the Fixed Term”;53.4 “termination by new contract”;53.5 “termination by inactivity”;53.6 “termination by non-statutory holiday”; and53.7 “termination by change in driver type”. The claimants’ submissions in response[54]The claimants’ closing submissions included, in paragraph 104, that “Issue 2 should be addressed at this hearing in light of the Judge’s conclusion that the fixed terms were not genuine and that their expiry did not automatically bring a claimant’s engagement as a worker to an end. Whether that conclusion involved any error of law is now a matter for the EAT.” I agreed with that submission.[55]The claimants also pointed out in paragraph 106 of their closing submissions that I said this in paragraph 310 of my 2025 reasons: “I could not see how I could properly conclude otherwise than that a failure to make payments to a worker under regulation 16 of the WTR, for essentially the same reasons although their detail may have differed from time to time, would constitute a series of deductions within the meaning of section 23(3) of the ERA 1996. I therefore concluded that such a failure would as a matter of principle rise to such a series.”[56]In paragraphs 109-111 of their closing submissions, the claimants said: ‘109. Whether a break between individual contracts should be treated as signifying the termination of a driver’s engagement must therefore depend on whether the break can be regarded as demonstrating that the driver’s employment relationship with AL had come to an end. Any other conclusion would be incompatible with Article 7(2) and with the principle of effectiveness i.e. the principle that national procedural rules should not have the effect of making it impossible in practice or excessively difficult to exercise rights guaranteed by EU law.110. It follows that the reference to “a worker’s employment” in regulations 14(1) and 14(6) WTR should be construed consistently with the WTD as meaning “a worker’s employment relationship”.111. It is pertinent to add that the effect of section 22(5) of the Retained EU Law (Revocation and Reform) Act 2023 (REULA) is that sections 2, 3 and 4 of REULA “do not apply to anything occurring before the end of 2023”. Section 2 REULA repealed section 4 of the European Union (Withdrawal) Act 2018 (EUWA), which provided that any rights, powers, liabilities, obligations, restrictions, remedies and procedures which, immediately before IP Completion Day, were recognised and available in domestic law by virtue of s. 2(2) of the 1972 Act and were enforced and followed accordingly would continue to be part of domestic law. Section 3 REULA abolished the principle of the supremacy of EU law and section 4 removed general principles of EU law, including the Marleasing principle, from domestic law. By virtue of section 22(5) REULA, these principles remain applicable to the LCs’ claims and continue to require the Tribunal to construe the WTR compatibly with the WTD.’[57]The claimants dealt in paragraphs 112-143 of their closing submissions with the seven scenarios in which the respondent said that there would be a break in the employment relationship. In paragraphs 112-125 the claimants addressed the impact of the taking of non-statutory holiday, and I can summarise their position in that regard by quoting this sentence from paragraph 115: “In circumstances where AL failed to accord the claimants any right to paid leave under the WTR and failed to facilitate arrangements for them to take the leave to which they were in fact entitled by law, it is not open to AL to argue that a driver who took more leave than that to which he was entitled under the WTR thereby brought his engagement with AL to an end.”[58]In paragraph 126 of their closing submissions, the claimants said in relation to “periods of inactivity [of four weeks or more]”, namely where the respondent was relying on a period of four weeks or more of inactivity together with a “plus” factor, that they did not accept that “any of the scenarios postulated should be treated as bringing about the termination of a claimant’s employment relationship.” My conclusions on the respondent’s contentions in relation to issue 2[59]It is convenient to state my conclusions on issue 2 wherever relevant by saying whether or not I accepted the claimants’ response to what the respondent was contending.[60]Before doing that, however, I record that the claimants accepted in paragraph 326.1 of their closing submissions for the 2024 hearing that “a break of more than 4 weeks in which an individual undertook other work or was prevented from working by loss of their PCO or driving licence would normally amount to a termination”. I could see no good reason for disagreeing with that acceptance. Termination on notice (see paragraph 53.1 above)[61]The respondent contended in paragraph 57 of its closing submissions that “Where notice of termination is given by either party, the engagement end date is the date notified by either party to the other. Where no date is specified by the terminating party, the contract would terminate immediately on the date when notice is given unless the driver undertakes work between the date of notice and the end of the fixed term contract, in which case the engagement end date would be the last worked day prior to the end of the fixed term contract.”[62]In the circumstance that the claimants and the respondent were working on the basis that the claimants were not workers, I did not see how I could properly conclude that the formal giving of notice of a contract for the hire of a vehicle or for the provision of driving services gave rise in itself to a break in the relationship of worker. Termination after the driver contract fixed term (2021 onwards) (see paragraph 53.2 above)[63]In paragraph 60 of the respondent’s closing submissions, under the heading “Termination after the driver contract fixed term (2021 onwards)”, this was said: “Where the fixed term has expired and for at least seven days a driver has(i) undertaken no jobs,(ii) received no pay for work from the Respondent, and(iii) returned his car and/or concluded his vehicle hire agreement (in the case of a Non- Partner Driver), the engagement end date would be the last day before that seven day period.”[64]Given what I say in paragraph 54 above, I could not accept that proposition. Termination by four weeks’ inactivity plus (see paragraph 53.3 above) FACTOR 1: “the driver has not received any payments from the Respondent in respect of Jobs during at least the last week of the Break Period”[65]The claimants’ response to this was stated in paragraph 126.1 of their closing submissions, in the following manner. “There may be many reasons why a driver would take a break of four or more weeks which would not support the conclusion that his employment relationship had come to an end e.g. holiday, sickness, family reasons. Indeed, although strictly a matter of evidence in each case, it is counterintuitive to assume that an AL driver who works regularly for a period, then takes say a 5 or 6 week break, then resumes driving regularly for AL, has ‘terminated’ his ‘employment’/his ‘employment relationship’ with AL.”[66]I agreed with that submission. FACTOR 2: “the Break Period includes at least one day in which the driver has declared they were not ready and willing to work due to holiday but which cannot be statutory holiday under regulation 13 or 13A WTR”[67]This factor overlapped with the sixth one, “Termination by statutory holiday”, which I have set out in paragraph 53.6 above and to which I return in paragraph 93 below. It is sufficient to say here that I agreed with the claimants’ submission in paragraph 115 of their closing submissions which I have set out in paragraph 57 above, and that I rejected the respondent’s argument based on this factor 2. FACTOR 3: ‘the Break Period includes at least one day in which the driver has placed themselves in a position that they cannot work (e.g. loss of licence/PCO, being locked out of the Respondent’s systems and/or classified as “inactive”)’[68]In paragraph 126.3 of their closing submissions, the claimants said this: “R’s suggestion that four weeks’ inactivity should be treated as a termination of engagement simply because it includes a single day in which one of these criteria are satisfied is devoid of merit. Further, it would potentially include situations where a driver had been locked out of R’s systems or classified as ‘inactive’ for a single day due to a mistake or by way of AL applying a punitive sanction.”[69]I agreed with that submission also. FACTOR 4: “the Break Period includes at least one day which is after the end of the Driver Contract Fixed Term”[70]The claimants’ response, stated in paragraph 126.4 of its closing submissions was that this was inconsistent with my conclusion that the fixed term contract “was an artificial device” here. Given what I say in paragraph 54 above, I agreed with that submission. FACTOR 5: “for a non-partner driver, they have returned their rental car and/or their vehicle hire agreement has ended”[71]The claimants’ response to this was stated in paragraph 126.5 as follows: “The evidence at PH24 [i.e. the 2024 hearing] showed that non-partner drivers might return their rental car in the event of taking any break of more than a few days, in order to avoid the cost of paying the rental fee for a vehicle they were not using. Similar evidence is before the Judge at RPH [i.e. the February 2026 hearing] – see HH/50, RK/62 and TM/110-111. Return of a vehicle is not therefore evidence that a driver’s employment relationship had come to an end. As to VHAs [i.e. vehicle hire agreements], these were issued to drivers at routine intervals, although the intervals between them sometimes varied, as will be evident from the list of VHAs issued to each of the LCs contained in the index to RB [.ie. the FHB] (section 5B of RB for HH, section 5D for RK and section 5F for TM). AL says that from August 2014 onwards drivers would be asked to sign up to a new VHA every 12 weeks – see AS/7-8 – although it is evident that even after this date longer intervals sometimes occurred. Be that as it may, the fact that a VHA had come to an end would not without more demonstrate that the employment relationship had ceased.”[72]I agree with that submission too. FACTOR 6: “for any driver, they have undertaken work under another engagement (including, for partner drivers, work undertaken as multiapping work)”[73]In paragraph 126.6 of their closing submissions, the claimants said in response to this factor: “Drivers were not prohibited from carrying out other engagements during periods when they were working for AL and AL was fully aware that partner drivers might undertake work via other apps. The mere fact, therefore, that a driver carried out work on another engagement or via a different app during a period of four weeks in which he did not work for AL should not be regarded as demonstrating that his employment relationship with AL had come to an end. The relevant question in any particular case is whether, viewed objectively, the driver’s conduct demonstrated an intention to cease working for AL.”[74]I agreed with that submission, but with the reservation that doing work for another “employer” in a period of four weeks between periods of working for the respondent was in itself capable of being decisive evidence of an intention to cease working for the respondent. Mr Leiper said in oral closing submissions that the claimants’ position here appeared to have shifted from the one which I record in paragraph 60 above. When reviewing the situation when deliberating, I concluded that the two positions were not inconsistent, but that if and to the extent that they were, then that did not matter greatly. What mattered was that I came to the right conclusion on the issue of principle, and I concluded that what I say in the first sentence of this paragraph is an apt statement of the correct approach to take. FACTOR 7: “any other circumstance which [indicates] the intention to discontinue the engagement and/or would otherwise frustrate the Driver Contract”[75]In paragraph 126.7 of their closing submissions, the claimants said that this proposed “plus factor” was “hopelessly vague” and that it was unclear whether the concept of “frustration” was being used in its strict legal sense or in some other way. I saw that the respondent did not say anything further in its closing submissions about that factor, and I agreed with the claimants’ response to it. Termination by new contract (see paragraph 53.4 above)[76]This condition for a break in the employment relationship was stated by the respondent in paragraph 64 of its closing submissions in the following manner. “Where a driver signs a new contract (B) before the end of the fixed term of the original contract (A), and there is at least one clear day between the engagement end date of contract A and the start date of contract B, the engagement end date would be the last date the driver worked under contract A.”[77]In paragraph 143 of their closing submissions, the claimants said that this proposition “wrongly assumes that the Fixed Term is genuine”. Applying what I say in paragraph 54 above, I agreed that there could not be a termination of the employment relationship in these circumstances just because there was a clear day between fixed-term contracts. Termination by inactivity (see paragraph 53.5 above)[78]In paragraph 67 of its closing submissions, the respondent said that where there was “a continuous period of at least six weeks in which the driver [did] not undertake any work for the Respondent and provided that, for contracts entered into on or after June 2021, the driver also [did] not work during the remainder of the fixed term, the engagement end date would be the date of the last job undertaken by the driver prior to the six week period”.[79]In paragraph 127 of their closing submissions, the claimants said in response: “The Respondent also contends, at paragraph 6E of CSOLP, that any continuous period of six weeks or more in which a driver did not work for AL should be treated as bringing that driver’s employment by AL to an end. The LDCs do not agree. There could well be – and there were – circumstances in which a break of 6 weeks or more would not support the conclusion that the driver’s employment relationship had ceased e.g. absences due to holiday (see above), sickness or family reasons, or absences in effect forced by the unavailability of work during the pandemic.”[80]I agreed with that submission. In my view the key question was whether, viewed objectively, the parties (and in this instance it will usually mean the claimant) intended the relationship to end. In that regard, I agreed with what the claimants said in the following two paragraphs (128 and 129) of their closing submissions about the evidence which Mr Smith gave on behalf of the respondent about the backdating of contracts by the respondent after a break in the provision of services by claimants. Whether the claimants were correct to say, in paragraph 130 of their closing submissions, that “[t]he practice of backdating contracts after breaks and storing rental credits demonstrate an acceptance on the part of AL that its employment relationships with its drivers were not brought to an end by periods of inactivity lasting only four or six weeks” was, however, another matter. What those practices demonstrated to my mind was a willingness to welcome back a driver, and to encourage the return of drivers by what I will call “sweeteners”. That, however, was a side-issue. The real issue here was whether a six-week break would in itself mean that the employment relationship within the meaning of the Working Time Directive, Directive 2003/88/EC, had come to an end, and my answer to that question was, no, such a break did not necessarily bring that relationship to an end, or have the effect that the relationship was at an end.[81]One major question here was the effect of periods when the lead claimants were not driving for the respondent during the Covid-19 lockdowns. The over-arching submission for the claimants in this regard was in paragraphs 131-132 of their closing submissions, which were in the following terms. “131. As they submitted at PH24, the LCs contend that the breaks during which they did not work because of the Covid pandemic did not bring their employment relationships with AL to an end. These breaks were caused by factors outside the parties’ control and occurred in circumstances in which it would have been obvious to AL that drivers who were stopping work because of the pandemic did not intend the consequent cessation of work to be permanent.” 132. It is submitted that such breaks would not even constitute a termination in the context of the more strict requirements of a ‘stable work relationship’ in equal pay law: see Preston and ors v Wolverhampton Healthcare NHS Trust and ors (No.3) ... [2004] ICR 993, EAT (‘Preston No. 3’) – same or broadly similar contractual terms; performing essentially the same work; breaks due to the availability of work/the employer’s requirements for employees to perform the work, etc.”[82]The claimants’ closing submissions in paragraph 135 about what happened during the Covid-19 lockdowns were in my view cogent in general terms. Mr Klepacki had by the time of those lockdowns stopped working for the respondent, so there was a need to consider only what happened to the employment relationship between Mr Ruiz and Mr Mahendran. Their evidence was in paragraphs 113-119 and 109-113 of their witness statements respectively.[83]It was accepted on behalf of Mr Ruiz that the break between his claim periods, which was from 16 July 2021 to 17 June 2022, broke the employment relationship between him and the respondent. At the start of the pandemic lockdowns, he stopped driving for the respondent on 22 March 2020, which was the day before (on 23 March 2020) the first lockdown was announced to start on 26 March 2020. He handed his car back on 30 March 2020 and, as he said in paragraph 113 of his witness statement “only rented a new vehicle on 9 February 2021”. In paragraphs 115 and 116 of his witness statement, Mr Ruiz said that when he stopped working for the respondent during that period, he “always intended to return” and that it “just took [him] until February 2021 to do so because [he] wanted to stay away from people as much as possible, and sharing a car with passengers didn’t feel safe.” His evidence continued: “116. ... As described in paragraph 112, the country was going in and out of lockdowns throughout this period, and overall, the recommendation was not to come into contact with more people than necessary. That’s part of the reason why I took the temporary job with Hermes, as I could drop things outside people’s doors, and didn’t have to have any contact with them. 117. I also didn’t think I would be able to earn enough to cover my expenses if I went back to Addison Lee during the Covid period, as people still weren’t travelling very much, and that would have had a big impact on my earnings, particularly as I tended to do mostly airport jobs.”[84]Mr Mahendran’s witness statement evidence on this was in the following terms. “109. When the Covid pandemic struck in 2020, like many other drivers I was forced to stop working. As highlighted above and as shown in my SOL, I stopped working due to the pandemic for the period 16 March 2020 to 28 August 2020 inclusive and again during the period 2 January 2021 to 16 April 2021 inclusive.110. During this time, I returned my Addison Lee vehicle so as to avoid paying the vehicle rent when I did not have the work to pay for it. I also did not want to risk catching Covid-19, especially because my two children suffer from asthma and if passed on to them, they would have been particularly susceptible to becoming ill. During the time when I took these breaks, I concentrated on taking care of my children and helping to home-school them.111. Data from Addison Lee shows that in relation to the first lockdown, I returned the car and XDA/MDA device on 1 April and 4 April 2020 respectively and retrieved both on 28 August 2020. In relation to the second lockdown, I returned my vehicle and XDA/MDA device on 16 January 2021 and retrieved both on 16 April 2021 (see page XX).112. My solicitor has shown me that Addison Lee paid me a points bonus of £165 on 4 February 2022 which carries the reference ‘covid’, which I cannot recall receiving but may have been provided to all drivers as a small gesture of goodwill given the hardship encountered by drivers as a result of the pandemic.113. In terms of surviving financially during the time I was not working, I made use of the grant available from the government for self-employed people who were unable to work at this time. As far as I recall, I received two payments or around £1,000 to £1,500 during 2020 and 2021, although I cannot remember exactly how much or when except that the amounts were based on my previous earnings and also took into account benefits that I was already receiving such as housing benefit, council tax benefit and working tax credits. It was a difficult time as there was no work, and I had to manage the family’s expenses very carefully.”[85]The respondent’s case on the situations of those two lead claimants was advanced in paragraphs 69-77 of its closing submissions. In paragraphs 72-75, this was said in relation to Mr Mahendran. “72. When he returned was entirely his decision (in contrast to Mr Atkins coming off furlough, which was a decision for the Respondent). He returned on 29.08.20 because it suited him – he had spent the summer with his children and started to work again when they returned to school. He returned on 17.04.21, in the middle of the third national lockdown, again a decision entirely his own (as he accepted). The Respondent could not have known on either occasion whether he would return at all and, if so, when.73. There was work available; indeed, the Respondent emailed all ‘exdrivers’ with offers of work (see Mr Atkin’s evidence).74. These facts are not consistent with there being any ongoing relationship between Addison Lee and Mr Mahendran (recalling the Claimants’ acceptance that there was no overarching or umbrella contract – see eg Cs’ Closing for November 2024 ¶ 48).75. One further point: Mr Mahendran’s position is that the covid period led to his holiday year re-starting (see TM SOL calculations – tab ‘C’s holiday dates’). It is unclear how this could be the case if there had also not been a termination and commencement of separate engagements.”[86]In relation to Mr Ruiz, this was said in paragraphs 76 and 77 of the respondent’s closing submissions. “76. The position is similar with Mr Ruiz. He took a break from 22.03.20 – 10.02.21 (some 10½ months). He returned his vehicle from 30.03.20 – 09.02.21 (¶ 113 [FWB/24]). He did not seek work from the Respondent when the first national lockdown stopped (or when people were encouraged to mingle during Eat Out to Help Out from early August 2021). He did not return when Mr Mahendran returned in late August 2021. He did not take up any offers of work. He returned in the middle of the third national lockdown. He accepted that the Respondent did not know whether he would return or if so when. Indeed, he did not himself know that. 77. Again, these facts are not consistent with there being any ongoing relationship between Addison Lee and Mr Ruiz.”[87]In contrast, the claimants submitted (in paragraph 135.3 of their closing submissions) that the respondent “can ... have been in no doubt that its drivers intended to return to work when the pandemic passed”, and, generally (as said in the rest of paragraph 135 of the claimants’ closing submissions), that the drivers’ employment relationships during Covid breaks continued, because87.1 it was clear “beyond doubt” that the respondent “perceived the business interruption caused by Covid to be a temporary phenomenon”;87.2 “Drivers who had stopped work due to Covid were emailed about potential work opportunities to encourage them back to work – see SA/73 and PHB/7278”;87.3 Mr Atkins accepted in cross-examination that the respondent regarded past drivers who were not driving for the respondent during the Covid-19 period to be “on furlough because there wasn’t enough work”;87.4 ‘Directly employed staff such as SA himself were furloughed by AL – in SA’s case for 10 months from August 2020 to May 2021. As SA readily accepted, the drivers were in essentially a “similar” position, save only that because they were treated as self-employed workers they were eligible to receive financial support from the Government under the Self-Employment Income Support Scheme rather than being furloughed.’87.5 “Drivers retained their AL ‘badges’ during breaks throughout the pandemic, for up to two years”; and87.6 “After drivers had returned to work after the pandemic, some, including TM, had their contracts backdated by AL to dates prior to the pandemic”.[88]In relation to Mr Ruiz, the claimants pointed out in paragraph 133 of their closing submissions that Mr Ruiz gave additional oral evidence that he had returned to work in February 2021 because “he felt safer at that point because he had heard from friends that AL had introduced a plastic barrier between the driver and the passenger(s), which provided the driver with some protection.” The claimants continued (and I have incorporated their footnote 6 into the text): “SA [i.e. Mr Atkins] confirmed in xex that it took about 8 months for the installation of a plexiglass screen to be approved by Transport for London. Moreover, the suggestion in xex that HH might have been expected to return to work in August 2020 was contradicted by SA’s evidence that the amount of work available in August 2020, when he himself was first furloughed, [had] not picked up at all.”[89]In reply, in oral submissions, Mr Leiper said that the emails to which I refer in paragraph 87.2 above were sent not only to drivers who had stopped driving during the pandemic but also to drivers who had stopped working for the respondent because of Brexit, i.e. the United Kingdom’s exit from the European Union.[90]The claimants’ final submissions in writing on the impact of covid were in paragraph 136, which was in the following terms. “The one point of distinction AL was left to articulate between AL employees put on furlough and its drivers in a “similar” position, was that AL controlled the date of return to work for the former, but not the latter. That is misconceived. It is based on an implied premise of an umbrella contract, which is not the present case. Whether a non-partner driver went on holiday, was recuperating after illness or surgery, or was keeping safe/not working because of a lack of available jobs during the pandemic (or even taking a weekend off), they were not ‘workers’ and AL never controlled precisely when they would resume ‘work’. There is no conceptual difference between those various situations. Indeed, one imagines that AL are only trying to take this bad point because the pandemic lasted as long as it did; if it had only lasted 4-6 weeks, it is impossible to see what the difference would have been between that time off work and time off work because of a respiratory condition, say. Yet, the fact that the pandemic was a once-a-millenium [sic] event in terms of its duration and effect, tells the tribunal nothing about whether the ‘employment relationship’ was ruptured for the purposes of limitation.”[91]It was in my view necessary for me to decide whether the employment relationship within the meaning of the Working Time Directive between(1) Mr Ruiz and the respondent and(2) Mr Mahendran and the respondent was terminated during the Covid-19 lockdowns by reason of the fact that they were not driving for the respondent during those lockdowns in the situations described in paragraphs 83 and 84 respectively above. That was because in my view the claimants’ proposed finding set out in paragraph 49.2.4 above, namely that there was no termination “where a driver took an extended period off work because of the lack of available work and/or reasonable safety concerns during the pandemic”, was too general. The correct question to ask in my view was whether a particular driver had to be regarded (from an objective point of view, i.e. asking the question asked by the claimants at the end of the extract which I have set out in paragraph 73 above) as having decided not to work for the respondent, and that question was in my view too fact-specific to be capable of being answered by that proposed answer set out in paragraph 49.2.4 above. Nevertheless, I concluded that generally it was correct to say that there was no break in the employment relationship during the Covid19 lockdowns between a driver and the respondent where the “driver took an extended period off work” merely “because of the lack of available work and/or reasonable safety concerns during the pandemic”.[92]So, what were my conclusions about the absences of Mr Ruiz and Mr Mahendran? In my view, both of them had their own good, i.e. valid, Covid-19 and related, reasons for not returning to working for the respondent before they did in fact do so, and in my view, bearing in mind the respondent’s consistent and strong denial that the claimants were “workers” for the relevant purposes, I had to conclude that there was no break in the employment relationship for the purposes of the Working Time Directive between92.1 Mr Ruiz and the respondent between 22 March 2020 and 9 February 2021, and92.2 Mr Mahendran and the respondent between 16 March 2020 to 28 August 2020 and 2 January 2021 to 16 April 2021. Termination by non-statutory holiday (see paragraph 53.6 above)[93]In paragraphs 114-115 of their closing submissions, the claimants said that there was no legal or evidential basis for the respondent’s “suggestion” that “where any driver declares himself to have taken more holiday in a holiday year than than the 5.6 weeks’ holiday to which he was entitled under regulations 13 and 13A, even if only one additional day, the taking of this additional leave brought his engagement by AL to an end and triggered the limitation period for bringing a claim. AL’s reliance on this contention is qualified only by [the] proposition that, for contracts entered into on or after June 2021, there will only be a termination if the driver does no work for the remainder of the Fixed Term.”[94]I agreed with that submission of the claimants. Termination by change in driver type (see paragraph 53.7 above)[95]In paragraph 63 of the respondent’s opening skeleton argument, this factor was justified on the basis that it flowed “necessarily” from my findings in my 2025 judgment about worker status: “the basis upon which a driver operates is so significantly different that there must be a termination of the previous engagement and a new engagement upon the new working basis.” In paragraph 137 of their closing submissions, the claimants said that “Becoming a partner driver did not ... represent a fundamental change in the nature of a driver’s relationship with AL or bring his existing employment relationship with AL to an end.” In paragraph 138 of their closing submissions, the claimants pointed out that this issue affected only Mr Klepacki and that:95.1 the “terms of his Driver Contract remained unchanged and he continued to work under the terms of the contract issued to him on 5 May 2018 [RB/3446- 3455]”; and95.2 “A further Driver Contract involving the same terms was issued on 26 June 2018 [RB/3456-3465].”[96]In fact, in oral submissions, Mr Segal corrected those page references to, respectively, 2993-3002 and 3003-3011, and accepted that the terms of the contracts differed slightly. I concluded that the difference was not material. In any event, I state my conclusion on the impact of a change in driver type in paragraph 100 below. “Alternative approach (performance versus payment)”[97]In paragraph 64 of its opening skeleton argument, under the above heading, the respondent contended that “performance of work is the most appropriate indicator of the existence of a worker relationship and for this reason the engagement end date is identified with respect to the last date worked before a Break Period (with the exception of Break types A and F)”, i.e. the breaks identified in paragraphs 53.1 and 53.6 above). The submissions continued, in paragraph 65: “However, should the ET determine that an engagement cannot be said to have ended until after the last payment for the corresponding jobs (or a similar principle), the Respondent contends the engagement end date should instead be the date on which the last job payment was made to the Claimant following the last date of work prior to the break period.”[98]I could not identify a direct response to this submission from the claimants. I suspect that that was because the issue was one of mechanics rather than principle. In any event, I agreed with the respondent that if there was a break in the employment relationship, then the relationship ended when the claimant last worked for the respondent, rather than the day when the respondent paid the claimant for that last day of work. My conclusions on issue 2[99]I found some of the claimants’ proposed answers, which I have set out in paragraph 49.2 above, to be a little difficult to follow. I regarded the proposed answer which I have set out in paragraph 49.1 above to be insufficient, and that the correct question to ask was as set out at the end of the extract from the claimants’ submissions in paragraph 73 above, namely “whether, viewed objectively, the driver’s conduct demonstrated an intention to cease working for [the respondent].” I also concluded that such an intention could not in itself be decided to have existed merely because99.1 a driver took more holiday in a holiday year than 5.6 weeks, irrespective of whether the driver returned his vehicle to the respondent, thereby ending the hire agreement which related to it;99.2 there was a break in the driver’s work for the respondent of up to four weeks accompanied by any of the 7 factors which I have set out in paragraph 53.3 above;99.3 there was a continuous period of six weeks in which a driver did no work for the respondent; or99.4 a driver took an extended period off work because of the lack of available work and/or reasonable safety concerns during the pandemic.[100]The one situation in relation to which I had some difficulty was “where a driver changed from being a non-partner driver to a partner driver”, even if it was “without any break in work”. While I could see the force of the respondent’s submissions on this issue (including that, as Mr Leiper pointed out in oral closing submissions, for example Mr Klepacki had to put before the respondent “a suite of documents showing” that(1) he had his own vehicle,(2) it was of the right standard,(3) he was entitled to use it for the purpose of driving for the respondent, and(4) it was appropriately insured), I thought that the claimants’ contentions were even more forceful. That was because (1) the claimants were not saying that there were overarching or umbrella contracts between them and the respondent, and (2) as I record in paragraph 96 above, the terms of the contract did not change materially when Mr Klepacki became a partner driver. It was possible to say that the manner in which Mr Klepacki worked was now different, but it was also possible to say that it was another way of working under the same contract. In any event, I concluded that the fact that there was a change from being a non-partner driver to being a partner driver without any break in work did not in itself break the employment relationship within the meaning of the Working Time Directive. Subsequent periods of not working for the respondent might well have broken that relationship, but the mere change in the status of the driver will not have done so. Issue 3: Are any of the lead claimants to be treated as having taken leave pursuant to regulation 13 and/or regulation 13A of the WTR because of the respondent’s refusal to remunerate them in respect of such leave?[101]The claimants’ proposed answer to this issue was that I should find that any holiday which the lead claimants took in any holiday year up to 5.6 weeks’ holiday should be treated as having been taken pursuant to the WTR.[102]The respondent accepted (in paragraph 99 of its closing submissions) that “leave under regulation 13 WTR which had been accrued but had not been taken could in principle be carried forward to the following holiday year, in line with the principles set out in King v Sash Windows C-214/16.” However, the respondent did not accept (for the reasons stated in paragraph 100 of its closing submissions) that “any leave accrued but untaken under regulation 13A WTR in any holiday year was carried forward to the following year.” I agreed what the proposition that leave the right to which was granted by regulation 13A of the WTR could not be carried forward under the principles in King v Sash Windows (and if I had suggested otherwise then, for the reasons which I give in paragraphs 11-12 above, that was an inadvertent error). The respondent did not, however, as far as I could see deal in its written closing submissions with the question of how I should treat the time off that the claimants did in fact take. In any event, I accepted what the claimants said in paragraph 146 of their closing submissions, which was to the effect that it was unarguably correct to say that “any holiday which the LCs took in any holiday year up to the statutory limit of 5.6 weeks holiday per annum should be treated as having been taken pursuant to the WTR even though the LCs did not appreciate at the material time that they had a statutory right to take the relevant leave”. (The actual words of paragraph 146 of the closing submissions might have been read as being to the contrary effect, but their meaning was plainly to the effect which I now state.)[103]Accordingly, I agreed with the claimants’ proposed answer to issue 3. Issue 4: To what holiday pay is each lead claimant entitled? This is a question of the calculation of the rate of pay and the period in respect of which the claimant is entitled to holiday pay. There may be a need to analyse the position before and after the new contractual arrangements were implemented in 2021[104]The claimants’ proposed findings in this regard were as follows.104.1 The lead claimants, their “employment” with the respondent having terminated, have the right to receive a payment in lieu of all leave to which they had been entitled under regulation 13 of the WTR during their claim periods, whether they had taken that leave on an unpaid basis or failed to take it.104.2 The respondent made a series of unlawful deductions from the lead claimants’ wages within the meaning of section 23(3) of the ERA, whether the right to take the relevant leave arose under regulation 13 or regulation 13A of the WTR.104.3 The entitlement to holiday pay must, where material, be calculated by reference to what the lead claimants were entitled to receive by way of the NMW (although the claimants believed this issue now to be agreed).104.4 A 52 week “look back” period should be used for the purpose of calculating the holiday pay to which the lead claimants were entitled.104.5 Tips paid by credit card or via the customer’s account and processed by the respondent formed part of the lead claimants’ “remuneration” within the meaning of section 224(3) of the ERA 1996.104.6 It now appearing to be agreed that “‘rental credits’, ‘points bonuses’, etc., should not be counted towards holiday pay in so far as they were earned for doing certain types, quantities, etc., of jobs”, “no other ‘rental credits’ should be counted towards holiday pay – in particular those awarded between June 2021 and 2023 in circumstances where drivers had inter alia completed and/or signed new 12-week contracts.”[105]The claimants addressed these issues in their closing skeleton argument by stating their case on105.1 the “look back” period;105.2 a week’s pay;105.3 rental credits;105.4 period of entitlement to holiday pay; and105.5 the effect of the June 2021 contracts.[106]I take those issues in turn, with the exception of what was said in paragraphs 173- 177 of the claimants’ closing submissions about the effect of the June 2021 contracts, because the only additional thing said there related to the “look back” period, to which I now turn. The “look back” period[107]In support of the 52-week “look back” period, the claimants submitted: ‘151. The LCs’ SOLs have been prepared using the 52 week ‘look back’ period specified by regulation 16(3)(e). AL, on the other hand, contends that because the look back period was increased from 12 to 52 weeks with effect from 6 April 2020, a 12 week look back period should be used to calculate the LCs’ entitlement to holiday pay in respect of periods prior to 6 April 2020.152. The background to the amendment of the look back period is explained by the editors of Harvey in the following terms in Division C at [191.02]: “One element in the statutory formula for calculating the rate of a week’s pay is that where the basis of the calculation is average normal pay, this is calculated using a 12-week reference period. The Taylor Report (Good Work, July 2017) recommended that the reference period be increased to 52 weeks, to avoid anomalies arising where pay varies because of such things as seasonal fluctuations in workload. The Government in February 2018 accepted this recommendation, and following a consultation exercise, the change has been implemented, with effect from 6 April 2020, by the Employment Rights (Employment Particulars and Paid Annual Leave) (Amendment) Regulations 2018 SI 2018/1378 reg 10. This amends WTR SI 1998/1833 reg 16 to provide for modifications of ERA 1996 ss 221–224, to the effect that the reference period of 12 weeks becomes 52 weeks, or the number of weeks of the worker’s employment prior to the calculation date, if fewer. ...”153. The LCs contend that where, as here, the amount of holiday pay to which a worker is entitled falls to be calculated and determined after 6 April 2020, a 52 week look back period should be used.’[108]In response to this, the respondent’s closing submissions included this passage (at the end of paragraph 116): “The argument appears to be that the statutory change applies whenever the assessment is undertaken after the amendment was introduced. That cannot be right: it is necessary to consider what the reference period was at the time under consideration. To do otherwise would be to give the amendment retrospective effect, which could only be the case if there were express language to that effect or it was a necessary implication. There is no such language and it cannot be implied that this was the legislative intention.”[109]It was only during oral submissions that Mr Leiper referred to any authority for that proposition, and, as I say in paragraph 21 above, he did so by referring to Craies on Legislation. I have analysed the applicable case law in paragraphs 21-23 above. In my judgment, Parliament is to be inferred to have intended the change to regulation 16 of the WTR to be fully retrospective in effect. I came to that view because(1) the change was at least in one sense (if not actually) procedural only,(2) the change did not overtly at least take away any vested right, and(3) I could see nothing amiss in there being a change from 6 April 2020 onwards in the manner of calculation of pay for the purposes of the WTR for all purposes, i.e. including in relation to claims which had already been brought by then. A week’s pay[110]The issue which required my determination here was whether or not tips should be excluded from the calculation of the claimants’ average weekly remuneration. In that regard, the claimants’ submission was in paragraph 156 of their closing submissions, which included the following passage. “Tips which were paid by credit card or via the customer’s account and processed by AL before being paid to the LCs formed part of their “remuneration” within the meaning of section 224(3) ERA - see Nerva v R. L. & G. Ltd [1997] ICR 11 – and have been taken into account in determining the LCs’ average weekly remuneration. This is not only correct as a matter of domestic law but also conforms to the EU principle that holiday pay should reflect the worker’s normal remuneration and that a payment forms part of normal pay if is compensation for performing the duties and tasks of the worker’s job – see British Airways plc v Williams (Case C-155/10) [2012] ICR 847 and Lock v British Gas Trading Ltd (Case C-539/12) [2014] ICR 813.”[111]The respondent’s position on this was stated in paragraph 121 of its closing submissions, which was in the following terms. “As to tips, those paid directly to the driver are excluded. There is no basis for including them. Similarly, those paid via the Respondent are excluded: it did not apply a tronc system (as in Nerva); the sums related to an individual driver and were passed directly to them; the Respondent never asserted title in the money.”[112]While what the respondent said was at first sight persuasive, I was not persuaded by it. It appeared to me that the claimants were not asserting that tips paid directly to drivers were included, so the only difference between the parties here was in relation to tips paid to the respondent and passed onto the claimants.[113]While I accepted that all three of the authorities on which the claimants relied concerned employees rather than workers who were not employees, Williams and Lock in my view required me to find (which I did without regret) that the claimants’ position in regard to tips paid to the respondent and passed onto the claimants was correct, and that those tips were to be included in the relevant claimant’s remuneration for the purposes of the WTR. Rental credits[114]The nature of the parties’ dispute about the manner in which rental credits had to be taken into account for the purpose of determining remuneration for the purposes of the WTR can be seen from what the claimants said in paragraph 158 of their closing submissions, which was as follows. “158. The LCs do not agree (in so far as they understand it) with the suggestion made by R at CSOLP 7.11(c), that holiday pay should be calculated inter alia by “Deducting the total value of (i) Rental Credits ... during the leave year …”. 158.1. First, this seems to assume that drivers did not return their vehicles when taking holiday and that all ‘rental credits’ earned by drivers were applied during periods of holiday. Such assumptions are not correct. 158.2. Second and in any event, rental credits were earned by drivers doing particular amounts of certain types of work, and had nothing to do with the taking of holidays (they would be/were earned at least equally by drivers who took no holiday).”[115]The respondent submitted on this: ‘119. As to rental credits, each time a Non-Partner Driver signs a driver contract, he is given a set amount in rental credits, typically either £150, £250 or £350 that he can apply when he wants. Depending on the vehicle he hires, the weekly hire costs range from around £165 per week (for a standard passenger car) to £191 per week (for an executive passenger car), plus car wash and insurance fees, so the rental credit covers most of the fees for a week. If a driver enters into successive 12-week contracts then in the course of the year this amounts to approximately 4 weeks’ rental credit. See ¶ 33 ff of Mr Townsend’s statement [FWB/99]. These rental credits are intended to be used for the purpose of reducing rental costs in weeks in which drivers they take holiday. This should therefore be left out of the calculation of holiday pay.[120]As the Judgment says, “It occurred to me that it was impossible to work out what the claimants were paid unless one took into account the purported holiday pay and any rental credits. On a common sense, or practical, basis, therefore, I could not see how those payments could sensibly be ignored” [FHB/1870 ¶ (302)].’ 116 I may have misunderstood what the respondent was saying there, but it appeared to me that those two paragraphs contradicted each other. That is because if the rental credits cannot be ignored, then they are not to “left out of the calculation of holiday pay”. Before being reminded of what I said in paragraph 302 of my 2025 judgment, that is to say when looking at the issue of the relevance of rental credits, I came to precisely the same view as the one which I stated in paragraph 302 of that judgment. What I concluded (again) was that the rental credits had to be taken into account in that they reduced the cost to the claimant in question of earning the income which he received, so if nothing else in that sense the credits had to be taken into account when determining what the claimants received by way of income over the relevant preceding period (which as I say in paragraph 109 above was 52 weeks). That was not a direct answer to the question whether the rental credits had to be taken into account in determining holiday pay, but that was because they were not part of a driver’s pay as such but had to be taken into account in deciding what was that driver’s pay for the purposes of both the WTR and the NMW. On that basis, the ruling of the EAT in Lyddon v Englefield Brickwork Ltd [2008] IRLR 198, to which the claimants referred in paragraph 164 of their closing submissions, was not relevant here. Period of entitlement to holiday pay 117 I found what the claimants said about the period of entitlement to holiday pay in paragraphs 166-172 of their closing submissions to be completely apt. I did not understand the respondent to take issue with the statements of principle made in those paragraphs, but in any event, I accepted the submissions made in those paragraphs. Accordingly, I accepted what the claimants said in paragraph 172 of their closing submissions, which was: ‘In light of the principles explained in Agnew and applied in de Mello, AL’s continuing failure to pay holiday pay for leave taken on an unpaid basis gave rise to a series of deductions which continued throughout the LCs’ respective employment relationships. All of the relevant underpayments were linked by a “unifying vice” and formed an unbroken “chain” of deductions from pay.’ The respondent’s submission about “Holiday entitlement (4 weeks or 5.6 weeks)” 118 The claimants did not deal in their written closing submissions with the point made by the respondent in paragraph 109 of its written closing submissions, which was this: “The Claimants suggest that if the worker worked 6 days per week, his entitlement to 5.6 weeks’ holiday must be a right to 33.6 days’ holiday (see ¶ 67 of C’s Opening). As noted above (paragraph 81), this is incorrect; the maximum number of holiday days for any full leave year is 28 (WTR regulation 13A(3)).” 119 I put it to the parties during the hearing (I have a note of having put the point to Mr Leiper specifically) that the right way to look at this dispute was that it was about the amount of pay which a worker should receive per week of holiday. I do not recall either Mr Segal or Mr Leiper saying anything more about the issue, and I could find no record in my notes of the hearing of either of them having done so. My conclusion on this aspect of the matter was that(1) the full earnings in a working week, even if the relevant claimant worked more than five days in that week, were to be taken into account for the purposes of determining that claimant’s remuneration, and(2) that remuneration was to be paid for a week’s holiday taken by that claimant. I came to that conclusion in part because regulation 13A(2) of the WTR refers not to days of holiday but weeks: it gives a right to 1.6 weeks’ additional holiday, although regulation 13A(3) then refers to an “aggregate entitlement” which is “subject to a maximum of 28 days”. In my view, that is a reference to 5.6 weeks’ holiday, not 28 days’ holiday. The calculation of 28 days is arrived at on the assumption that a week’s work is done in five days. In conclusion in relation to issue 4 120 In conclusion,120.1 I agreed with the claimants on the proposed findings set out in subparagraphs 104.1-104.5 above, but for the reasons stated in paragraph 116 above, I disagreed with them in regard to the proposed finding set out in paragraph 104.6 above, and120.2 the impact of a claimant working more than five days in a week is that his remuneration for holiday pay purposes must be calculated by reference to the earnings received for the week which was in fact worked, so if the claimant worked for six days then his pay for the purposes of the WTR must be calculated by reference to his receipts for those six days of work. Issue 5: What pay for the purposes of the NMWA did each lead claimant receive?[121]By the time of the February 2026 hearing, there were only two “live” questions to be answered in this respect. They were(1) what was the pay reference period for the purposes of the NMWR and(2) how deposit payments made by the lead claimants in relation to equipment lent to them by the respondent should be treated. The claimants said in this regard that I should find that121.1 the lead claimants at all material times had a pay reference period of a week, and121.2 in respect of returns of deposits which the respondent has treated as ‘payments’ to the lead claimants, any deductions made from their pay towards such a deposit should be treated as an expense. The pay reference period[122]In paragraph 188 of their closing submissions, the claimants said that the lead claimants’ claims had been calculated on the basis that they had at all material times had a pay reference period of a week and that after oral evidence, the claimants anticipated the point being conceded by the respondent. The respondent did not, however, concede the point. The respondent’s position on the pay reference period after June 2021 was in paragraph 141 of its closing submissions, which was as follows. “The Respondent has applied the appropriate reference period to the NMW calculations in accordance with the applicable driver contract terms, either one week or four weeks [FHB/4641 ¶ (3.3(b))]. The Respondent has applied a four week reference period in the case of Mr Mahendran from 8 July 2021 and for Mr Ruiz from 18 June 2022) [FHB/4653 ¶ (8.2)]. Whilst the Claimants received a payment on a weekly basis, the holiday pay was accrued and paid on a four-weekly basis (see for example TM’s pay statements over a four week period at [FHB/4501, 4505, 4510 & 4515). Thus, overall, pay was calculated on a 4-weekly basis. If the reference period were otherwise, then this would skew the figures in each of the weeks.”[123]I was not referred to any appellate case law on this issue. It appeared to me therefore that it had to be decided as a matter of principle. Section 1(4) of the NMWA provides that ‘For the purposes of this Act a “pay reference period” is such period as the Secretary of State may prescribe for the purpose.’ Regulation 6 of the NMWR 2015 provides that ‘A “pay reference period” is a month, or in the case of a worker who is paid wages by reference to a period shorter than a month, that period.’[124]The June 2021 driver contract at pages 4354-4367 of the bundle for the 2024 hearing contained this provision (clause 7.5): “Subject to Clauses 7.8 and 16.2, we will make settlement against Driver Statements on a four weekly basis by direct payment to your bank account using the BACS payment scheme or make three weekly interim payments with a final payment including any balancing payment or charges that may be determined from time to time should you choose the one-week advance payment option. We shall not be held liable for any error or delay in payment to you caused by a failure of this system or otherwise outside our control.”[125]The claimants referred in paragraph 189 of their closing submissions to a clause to the same effect (numbered clause 6.4, but from what page in the bundle for the 2024 hearing they did not say). The claimants’ submissions on the reference period were as follows. ‘189. R contends that, following the introduction of the June 2021 Driver Contract, the pay reference period for NMW purposes was 4 weeks. This contention is understood to rest on the provisions of clause 6.4 of the contract, which included the following: “6.4. Subject to clauses 6.5 and 15.2, we will make settlement against Driver Statements on a four weekly basis by direct payment to your bank account using the BACS payment scheme or make three weekly interim payments with a final payment including any balancing payment or charges that may be determined from time to time should you choose the one-week advance payment option. ...”190. In practice, however, HH and TM continued to be paid weekly after June 2021 and AL continued to make deductions from their pay on a weekly basis. AL’s attempt to characterise these weekly payments as an “advance payment” was a device designed to create an ... artificial reference period of 4 weeks.191. In a correction to para 165 of his witness statement, RD [i.e. Mr Davies] said that following the introduction of the Driver’s Contract in June 2021 drivers were initially paid at four weekly intervals but that, following complaints from drivers, weekly payments were reinstated.192. The LCs gave evidence that they were always paid weekly. RD accepted that. He also accepted that their weekly pay was for the jobs they had completed in the previous 7 days, as well as any rental credit/point bonus earned under the incentive schemes.193. Under section 17 NMWA, the test to be applied in deciding whether the NMW has been paid is whether the remuneration “received” by the worker for the pay reference period is less than the remuneration he would have received if he had been remunerated at a rate equal to the national minimum wage. The reference in regulation 6 to the “payment of wages” must therefore be given a meaning consistent with Section 17.194. In the instant cases, HH and TM received remuneration and were paid wages weekly at all material times before and after June 2021. The Judge is therefore invited to find that the pay reference period within the meaning of regulation 6 NMWR was one week throughout the period of their claims.’ A discussion[126]I could not tell from the material before me whether or not it would help the claimants if I found that the pay reference period was a week rather than a month, but I could infer that it would do so from the fact that they were contending for a reference period of a week. That, however, was irrelevant. The issue was what, as a matter of law, i.e. statutory interpretation, was the correct period here? That was a difficult question in the circumstances that(1) at all material times neither the claimants nor the respondent treated the claimants as workers, and(2) the amount of work which the claimants did varied according to the time of year when it was done. However, as I said in paragraph 69 of my 2025 judgment, the claimants “were paid to do jobs for the respondent at rates which were determined by the respondent without any input from the drivers”. In addition, as I said in the same paragraph, the jobs were offered via software. Did those things make any difference to the manner in which I applied regulation 6 of the NMWR 2015? I was not at all sure that they did, but if and to the extent that they were relevant, they showed that it was open to the respondent to use software which maximised the chances of the claimants being allocated work on such a basis that they were paid the NMW by reference either to a period of a month or to a period of a week.[127]Putting those considerations aside, however, I asked myself what was the intention of Parliament here? I concluded that that the manner in which an employee was in practice paid was the determining factor, and that it was not what any contract of employment or other contract said. My conclusion on the pay reference period[128]In the circumstances stated by the claimants in paragraph 192 of their written closing submissions, which I have set out in paragraph 125 above, I concluded that the pay reference period here was at all material times a week, and not a month. Deposit payments[129]In paragraph 196 of their closing submissions, the claimants said this: ‘In calculating its “Adjusted Supplier Payment”, AL has included returned deposits as a payment. However, it has not treated deductions towards deposits as a relevant deduction, instead treating them as an exclusion – see CSOLP Appendix B. This is not logical. If the return of a deposit is to be treated as payment, it necessarily follows that a deduction towards a deposit should be treated as an expense.’[130]The respondent did not address this issue specifically in its closing submissions. It seemed to me to be an unavoidable conclusion that the claimants’ submission was correct: if the return of a deposit is to be treated as a payment, then a deduction towards a deposit must be treated as an expense. I have recorded that in my above judgment as a reference to expenditure within the meaning of regulation 13 of the NMWR 2015, but plainly it applied also to the preceding regulations. Issue 6: Was the hourly rate of pay for any lead claimant below the NMW?[131]The claimants invited me to find that for purpose of “calculating relevant ‘deductions’”, I should make the following findings.131.1 “The LCs’ work-related fuel expenses as non-partner drivers are to be calculated as their actual fuel expenditure less their expenditure on fuel for personal driving. The latter figures are to be calculated either on the basis of the percentages, alternatively the fixed weekly amounts of which the LCs have given evidence.”131.2 “HH’s and RK’s work-related fuel expenses as partner drivers are to be calculated on the basis of the percentages of time they spent driving for AL (out of the total driving time for all operators) of which they have given evidence.”131.3 “The costs of the LCs’ mobile phone (handset, contract charges, etc., but not additional spend) are all relevant ‘reductions’.”131.4 “Fines and penalties which were necessarily incurred by a driver in the course of carrying out a job e.g. stopping on a double red or double yellow line in order to pick up a passenger are to be treated as ‘reductions’; fines/penalties relating to other driving infractions are not. If it is not possible to separate out the two categories of fines/penalties, the Tribunal will have to guess a likely proportion/split.”131.5 “All parking expenses claimed by the LCs should be treated as relevant ‘reductions’.”131.6 “Deductions from HH’s Supplier Payment in respect of the cost of selfassessment services provided by Cab Accounting Ltd to HH via an arrangement established by AL are to be treated as relevant ‘reductions’ in the relevant reference periods.”131.7 “Where the expenses which an LC incurred in a particular pay reference period exceeded his pay, the hourly rate must be calculated as a negative figure.” The appropriate approach to take here[132]As the claimants said in paragraph 197 of their closing submissions, “In order to determine whether the hourly rate of any LC was lower than the NMW, it is necessary to decide what reductions have to be made in order to determine their remuneration for the purposes of regulation 8.”[133]The claimants then set out regulations 11 and 13 of the NMWR 2015 and referred to what HHJ Auerbach said in paragraph 36 of the EAT’s judgment in Augustine v Data Cars Ltd EA-2020-000383 (unreported) 21 July 2021. I found the whole of the passage in that judgment up to and including paragraph 38 to be accurately encapsulated by the words taken from the start of paragraph 36 of that judgment set out in paragraph 200 of the claimants’ closing submissions, which were that the expenditure of the claimant in that case (as a driver of a mini cab) “did not, in fact, have to be a requirement of the employment. It neither had to be necessarily incurred, nor wholly or exclusively incurred.”[134]I found the approach of the EAT in paragraphs 51-53 of its judgment in Augustine, read against the findings of fact made by the employment tribunal in that case, albeit in the light of the successful grounds of appeal stated in paragraphs 36-38 of the EAT’s judgment, to be helpful here. Even in the absence of the decision of the EAT in Augustine, I would have been inclined to accept what the claimants said in paragraphs 206-208 of their closing submissions here. Given that decision of the EAT, however, I came to the conclusion that those submissions were plainly right. In fact, I could not see any rational basis for the proposition of the respondent to which the claimants referred at the start of paragraph 207 of their closing submissions. For the sake of clarity this is what the claimants said in paragraphs 206-208 of those submissions: ‘206. Given the long duration of the period over which AL failed to pay the national minimum wage to its drivers, it is not reasonable for AL to suggest that the LCs should identify every expense that was incurred on a different date from that shown on their bank or credit card statements – see CSOLP para 11.4(b). This would be oppressive and inconsistent with the burden of proof in section 28(2) NMWA and the principles outlined above. Further, during the periods in which the LCs were non-partner drivers and working only for AL, all of the relevant expenses were in connection with their employment by AL, other than expenses related to personal use of their vehicles, which are addressed below. 207. By the same token, the Judge should reject the contention at CSOLP 11.6 that there should be “a presumption” that fuel costs are only potentially allowable to the extent that they arise on the same date as a job, the day immediately prior to a job or the day immediately following a job, and a concomitant “presumption” that fuel purchased outside this supposed “Job Window” should not be a potentially allowable expense. The proposed presumptions do not reflect either the LCs’ evidence of the way in which they purchased fuel, or ... a realistic view of how they were likely to have purchased fuel, and are similarly inconsistent with the burden of proof under section 28(2) NMWA. 208. In similar vein, the suggestion at CSOLP 11.8 – 11.11 that there should be “a presumption” that parking charges, including airport parking charges, and road tolls are only a potentially allowable expense when appearing in the SoL on the date a Job was undertaken or the following day is unreasonable to the point of implausibility. It makes no allowance for the facts that(a) it is a priori much more likely that such charges/tolls were incurred while working, and(b) parking charges or road tolls incurred on a particular date would not necessarily appear on a bank or credit card statement as soon as the same day or the following day.’[135]On the question of the burden of proof, the respondent submitted (via paragraphs 135-137 of its closing submissions) that135.1 while it is for the respondent “initially to disprove the presumption that the claimant was underpaid at all, the extent of any such underpayment remains at large; it is submitted that it is for the claimant to prove the amount of any underpayment, and the ET must decide this on the balance of probabilities”; and135.2 (via a quotation from Harvey on Industrial Relations and Employment Law at BI[246.01]) “the onus would still appear to lie on the worker to establish the amount of any underpayment”.[136]It is rare that the burden of proof is determinative. In paragraph 27 of their closing submissions, the claimants said that the respondent’s criticisms of the lead claimants “for not retaining ‘records’ of their expenditure ... are misconceived”. In paragraphs 29-32 of their closing submissions, the claimants said this: “29. ... If a driver purchases fuel for a vehicle that he uses mainly for work but also for personal use, the record of that purchase is available but does not reveal how much of the fuel purchased was used for work. Both at the time and now, all a self-employed person can and is required to do (whether a LC or a barrister) is to provide an honest estimate of the relevant proportion. That is what the LCs did at the time and it is what they have done in evidence at this hearing. 30. As to what the LCs did when they took periods of time off work, they were never required to retain records and generally did not. They cannot be criticised for that. 31. Moreover, there is a critical point to be made in respect of(a) drivers’ working hours;(b) relevant work expenditure; and(c) periods off work: AL not only treated its drivers as not being workers, but (as the Tribunal has found) went to considerable lengths in its attempt to demonstrate that they were not workers and were not being treated as such. The result is that AL did not create and retain the records of the relevant information which they would and should have done had they acted lawfully. 32. In relation to working hours and work expenditure, if AL had accepted its liability to pay the NMW to its drivers, it would have been required to put a system in place which ensured that it made and retained full and accurate records of the drivers’ working hours and details of relevant work expenditure.”[137]I have set out the next, or following, paragraphs of those closing submissions in paragraph 16 above. I agreed with what the claimants said in principle in this regard, but I also concluded that, as the respondent submitted, it remained necessary for the lead claimants to put some evidence before the tribunal from which the tribunal could “establish the amount of any underpayment”. It was, however, in my view, impossible to conclude that it was for the lead claimants to establish the precise amount of any underpayment given that the respondent had not kept “relevant records” within the meaning of section 10(10) of the NMWA.[138]The application of the correct approach as a matter of law to the expenses which were claimed here was not straightforward, however, and I could see that the question of whether an employer had failed to pay the NMW and, if so, by how much was a question of fact. In that regard I agreed with what the claimants said in paragraphs 38 and 39 of their closing submissions, which was as follows. “38. As is set out in detail below, time and again AL has put its case on material issues of fact not by reference to evidence, but on the basis of more or less obviously erroneous ‘presumptions’. That approach is as wrong in principle as it is transparently self-serving; and the Judge is respectfully invited to reject it. The Tribunal is rather required, as a matter of law, to make findings of fact, based on the evidence of what actually happened (even where that evidence is incomplete or necessarily an estimate), on the balance of probabilities. 39. That applies in particular to the need to assess, on the balance of probabilities: 39.1. The proportion of actual expenditure on fuel to be allocated as work expenditure; 39.2. Whether categories of ‘Empty’ time reflected time logged on, or time actually not logged on; 39.3. Whether parking fines were incurred as an incident of working, or negligence.”[139]I therefore now turn to the particular expenses claimed and the basis on which they were claimed here, and state my factual conclusions in regard to those particular expenses. Fuel[140]The respondent’s position on the cost of fuel was stated primarily in paragraphs 158-160 of its closing submissions. I have already indicated (in paragraph 134 above) that I rejected the submission in paragraph 159 that “there should be a presumption that fuel purchased outside of a Job Window should not be a potentially allowable expense”. Given the practical reality that fuel is bought from time to time and that it would be impossible to say precisely what fuel has been used for a particular journey or journeys, in my judgment the only way that any kind of precision could have been achieved in regard to the amount of fuel used for carrying out the work which the claimants did as workers here was by the claimants keeping records of the mileage incurred for private purposes and those miles travelled for private purposes being valued by reference to the typical fuel consumption of the vehicle in question. That would mean that the cost of private mileage calculated in that way would be deducted from the actual costs incurred in buying fuel for the vehicle used to carry out the work. To that extent, I agreed with that the respondent submitted in paragraph 160(a) of its closing submissions in regard to the use of “the applicable HMRC advisory rate for Company Cars”. However, in the absence of evidence of private mileage because the respondent treated the claimants as not being workers,140.1 the claimants could rely only on(1) evidence of the cost of buying fuel for the vehicle driven by a lead claimant as a worker for the respondent and(2) an estimate of that claimant’s private mileage or alternatively the cost of fuel consumed by the vehicle when the claimant drove it for private purposes, and140.2 in my judgment that was a sufficient means by which the claimants could prove that claimant’s fuel expenditure for the purposes of regulation 13(1)(b)(ii) of the NMWR 2015, subject to a reality check of their estimate of the claimant’s private usage of the vehicle.[141]I was therefore forced in the event of a dispute between the claimants and the respondent about the reliability of a lead claimant’s estimate of his private mileage, or alternatively his expenditure on fuel for private purposes, to decide whose estimate was more likely to be accurate. It was possible that I could lawfully come to a conclusion for which neither party contended, as long as it was arrived at rationally and was somewhere between the two parties’ positions. However, the respondent was unlikely to be able to put evidence before me about the claimant’s private usage of the vehicle in question, and was likely only to be able to challenge the evidence of the lead clamant about that private usage.[142]Having said that, having come to the conclusion on the approach stated in paragraph 137 above as a matter of principle, I looked again at the parties’ closing submissions on the question of what fuel costs should be offset against the money paid by the respondent to the leading claimants, and I found that what I had concluded as a matter of principle was what the claimants contended for, and not what the respondent contended for. Accordingly, I accepted what the claimants submitted in paragraphs 209-214 of their closing submissions (although I did not need to apply what was said in paragraph 214 because it was not applicable to the circumstances as I found them to be; but if it had been applicable then I would have agreed with the claimants that if the HMRC advisory rates had been applicable then they would have been applicable throughout the period to which the claims related, and not just when it suited one or other party). Also accordingly, I rejected the approach proposed by the respondent in paragraph 160 of its closing submissions, although I accepted that that approach could be used in submissions as a means by which to seek to undermine the estimate of a lead claimant of the cost of private usage of the vehicle used as a worker for the respondent.[143]So, what proportion of the fuel costs of the lead claimants had to be deducted as having been incurred for personal purposes and not for the purposes of their work done for the respondent as workers? That proportion was dependent on each lead claimant’s usage of the fuel put in the tank of the vehicle which he drove in order to do driving work for the respondent. I could see that if the only vehicle in the possession of the claimant was the one used for that work then that usage might be higher than if the claimant had another vehicle and the one used for that work was liveried. I could also see that a non-liveried upmarket car such as one hired from Eventech for purpose of being used by an executive driver was more likely to be used for private purposes than a liveried vehicle. Mr Ruiz[144]Mr Ruiz’s evidence about his private usage of his car when he was a non-partner driver was in paragraphs 88-99 of his witness statement. I saw that he thought that his fuel costs had been significantly understated because (as he said in paragraph 91):144.1 “I sometimes used my wife’s payment card to refill the tank”, and144.2 “Unfortunately my wife’s bank statements were not disclosed because I did not recall that they were relevant until too late in the process and therefore those expenses have not been included in my SoL.”[145]That may be why he estimated in paragraph 90 of that statement that he “was only spending an average of around £1 a week on personal fuel expenses”. However, what he said in paragraphs 95 and 96 of that statement showed that he would in addition to using his Eventech vehicle, which was at least for part of the time a Volkswagen Sharan, for “only ... short distances, which [he thought] would cost less than £1 in fuel”, he would, “[a]round twice a year ... also go on camping trips”, when he “would usually take [his] Addison Lee car camping because [he] needed the extra space”. In addition, he would typically when using that car in “[his] private life” go “to visit a friend at their house”, or to “Windsor with [his] wife and children”. That meant that his estimate of £1 a week was plainly an understatement.[146]In those circumstances, I was forced to come to an estimate of the fuel usage of Mr Ruiz for private purposes. Assuming that(1) the vehicle had a diesel engine,(2) the fuel consumption for personal usage would have been of about 45 miles to the gallon, and(3) fuel costs were about £1.20 per litre (which was close to the minimum that they were in the relevant period, as far as I could see: I saw that Mr Mahendran referred to them in paragraph 61 of his witness statement as having been “approximately £1.75 to £1.80 at the time”, but that is well above recent and current prices at the pump), that would mean that a gallon would have cost £5.40 and that £1 would have paid for about 8 miles of travel. That sounded like an under-estimate to me if the claimant and his wife went from London WC1E to Windsor and back.[147]After considering the evidence of Mr Mahendran about the extent to which he used his Addison Lee vehicle for private purposes as a comparison, I concluded that a fair estimate of the fuel in Mr Ruiz’s Addison Lee vehicle used for private purposes could be arrived at by concluding that on average Mr Ruiz used that vehicle to drive 40 miles per week, and not 8. Applying a reality check to that conclusion, his average weekly fuel consumption for private purposes when fuel cost about £1.20 per litre will have been £5 and not £1. That reinforced my conclusion that on the balance of probabilities, Mr Ruiz’s private mileage was on average 40 per week.[148]I did not understand what Mr Ruiz said in paragraphs 100-102 about the cost of fuel incurred by him for private purposes when he was a partner driver to have been challenged, but in any event I accepted it. Accordingly, his private mileage was the same when he was a partner driver as when he was a non-partner driver. Mr Mahendran[149]Mr Mahendran’s evidence about his fuel usage was in paragraphs 58-69 of his witness statement. That evidence was in my view more accurate than that of Mr Ruiz in relation to fuel usage, and I saw no reason to disagree with what Mr Mahendran said in paragraph 61 as corrected by him in oral evidence, which was that he spent about £16.40 per week on fuel for personal usage in a normal week. However, I rather doubted that he did not, as he said in paragraph 61 of his witness statement, “use [his] Addison Lee vehicle to go on long trips or holidays.” Especially when he was driving an eClass Mercedes car, I concluded, he was likely to have used the car for going on holiday and on long trips.[150]Mr Mahendran said in paragraphs 58 and 61-62 of his witness statement that he drove his Addison Lee vehicle for only about 25 miles per week for personal purposes and that his usual journeys were short local ones to his children’s schools and local supermarkets. That was a more realistic estimate in my view of the likely usage of the car for weekly purposes, but it failed to take into account the usage, which I have concluded occurred, of that car for long journeys and holidays. The cost of 25 miles per week was less than £16.40 per week, which was Mr Mahendran’s final evidence about his personal fuel expenditure, and therefore I was content to accept that figure of £16.40. Mr Klepacki[151]In paragraph 40 of his witness statement, Mr Klepacki said that he used his Addison Lee vehicle “for personal use about 8% of the time on average” but that “sometimes it could have been as low as 2%”. Given what he said in the rest of paragraphs 40-44 of his witness statement, I saw no good reason to reject that estimate of 8% on average, and I concluded that “the average amount [of money spent on fuel] that ha[d] been apportioned [by his solicitors] for personal use during the period [he] worked as a passenger driver”, which was £9.82 per week and sounded right to Mr Klepacki, was a reliable estimate. I therefore accepted it. That evidence was expanded on but in reality repeated in paragraphs 106-110 of his witness statement.[152]I also saw no good reason to reject Mr Klepacki’s evidence in for example (see paragraph 169 below) paragraphs 20 and 111-112 of his witness statement about the proportion of the working week that he used his non-Addison Lee vehicle for Addison Lee jobs when he was a partner driver. That was “approximately 35%”. Mobile telephone costs[153]The claim of the lead claimants to offset their mobile telephone costs was not straightforward, although on the claimants’ behalf it was treated (in paragraphs 222-224 of their closing submissions) as if it were. The reasons why I did not think that it was straightforward included the relevant evidence of Mr Ruiz. In paragraph 78 of his witness statement, he said: “I remember that when I first started, Addison Lee used XDAs which worked like walkie talkies, and we used them to call the controllers, and they would call us as well to tell us that a passenger was going to be late, for example. I didn’t have to pay anything extra for the XDA I received when I started with Addison Lee, as the device was fixed to the car.”[154]However, as Mr Ruiz’s witness statement continued, the situation then changed and he used a mobile telephone for the purposes of driving for the respondent: “79. At some point, I can’t remember exactly when, Addison Lee replaced the XDAs with MDAs, and you couldn’t use MDAs to call Addison Lee. The controllers could call drivers using the MDAs, but we couldn’t call them back, which was why I had to use my own phone for contacting Addison Lee. The mobile phone was also my main means of contacting passengers: most often, I would have to call the controller and the controller would call the passenger on my behalf. Sometimes the controller would put me through to the passenger, or give me their number so that I could call them.80. The MDA was an app on a Samsung phone, and I had to pay a deposit for the phone.81. The MDA came with a navigation system which we could use when we had accepted a job. However, I found that the MDA navigation system was not very accurate and would sometimes try to send me down roads which had been closed for years. That was why I preferred to use Google Maps or Waze on my own phone for navigation.”[155]That evidence was not challenged by the respondent. Paragraphs 93-97 of the witness statement of Mr Klepacki were to a similar effect, and their substance was not challenged. Paragraphs 80-82 of Mr Mahendran’s witness statement included the statement at the end of paragraph 80 that he considered that his mobile telephone was “indispensable for the purposes of [his] work.” That too was not challenged. The only substantive challenges were to the costs of the mobile telephones themselves.[156]In that regard the respondent understandably was critical of the position of Mr Mahendran, whose monthly mobile telephone charge was (he said in crossexamination) estimated by reference to the contract into which he entered with Vodafone in 2021. In paragraph 13 of the respondent’s closing submissions, this was said: “Mr Mahendran said in his witness statement (¶ 82 [FWB/76-77]) that he does not have records of his Halifax account setting out his mobile phone charges but under cross-examination said that he did have Halifax statements but had not provided them to his solicitors. Similarly, he kept detailed handwritten notes of his personal expenditure in notebooks and accepted that he might have some of those notebooks at home. These are not individuals who are doing what they are required to do in order to evidence their position.”[157]What Mr Mahendran said about the amount which he was claiming for his mobile telephone use was at the end of paragraph 82 of his witness statement, and was this: “In fact, I only paid approximately £44 a month for my own phone which I used for both work and personal purposes. These were not paid from my Barclays Bank account but from a Halifax account which I do not have records of. My solicitors have therefore only included these monthly payments as my regular phone payments in my SOL.”[158]If there were no additional charges for personal use, which I could see was likely at least currently when most mobile telephone contracts allow for unlimited texts and calls, then I could see that it was appropriate to conclude that the whole of the cost of the telephone should be regarded as an expense incurred by a nonpartner driver in driving for the respondent. That was because of the decision of the EAT in Augustine. Here, while I was distinctly unimpressed by the failure by Mr Mahendran to put before the respondent and me any documentary evidence of his claimed Halifax bank account payments for his monthly mobile telephone costs incurred in relation to the work which he did for the respondent as a driver, I concluded that I should accept his evidence that the cost was approximately £44 per month but on the basis that(1) he should disclose to the respondent that documentary evidence and the notebooks containing handwritten records of his personal expenditure and(2) the respondent could then rely, in agreeing with the claimants the income of Mr Mahendran for current purposes, on that further documentary evidence if it showed that any aspect of Mr Mahendran’s evidence on his expenses already given was incorrect. I did not see a need to do more at this stage than to say that if the parties needed me to revisit this aspect of the matter because they were unable to agree it, then I would do so. That was not least because I was not invited by the claimants to do more than rule that “The costs of the LCs’ mobile phone (handset, contract charges, etc., but not additional spend) are all relevant ‘reductions’.” I was happy to do that, although I concluded that I had to do so by reference to the terms of regulation 13(1)(b)(ii) of the NMWR 2015, which I did. Fines and penalties[159]In paragraph 227 of their closing submissions, the claimants said that “after having given further consideration” to the question of what fines and penalties could be offset against income, they accepted “that a distinction should be drawn between fines and penalties which were necessarily incurred by a driver in the course of carrying out a job e.g. stopping on a double red or double yellow line in order to pick up a passenger (which are legitimate ‘deductions’); and fines/penalties relating to driving infractions which could reasonably have been avoided (which are not).” I agreed with that submission if it were the case that the respondent required its drivers to pick passengers up from places where there was a nostopping rule such as where there was a double red line. The respondent’s closing submissions on this issue were in paragraph 167(b) and were as follows. “The presumption is that the following are not potentially allowable expenses: ... Fines and penalties, including (but not limited to) parking fines, TFL Penalty Charges or bus lane enforcement charges incurred by a driver. Such costs are charged back to drivers as related to the conduct of the driver. It is remarkable that the Lead Claimants each think that incurring such fines and penalties are an inevitable part of the job which the Respondent should reimburse. The whole point of such fines and penalties is to stop conduct; they are something that the drivers can and should avoid. To allow these fines to be recoverable as expenses would be impermissibly to condone such driving.”[160]Returning to the claimants’ closing submissions, they continued (in paragraph 228): ‘All the LCs gave evidence about unavoidable parking fines incurred when picking up passengers. As HH explained, “It would depend on where the passenger was waiting, not like I can go round the corner I have to pick them up where they are, with Uber I can but AL was different”. He was not challenged on that and no AL witness gave contrary evidence. If it is not possible to separate out the two categories of fines/penalties, the tribunal will have to guess a likely proportion/split.’[161]I was not comfortable with the idea of me having to “guess a likely proportion/split” unless I had a guess of the same sort from a lead claimant, based on some form of underlying concrete evidence. Otherwise, I concluded, the claimants would have failed to prove their case on a balance of probabilities.[162]On this issue, when deliberating, it occurred to me that the respondent had regarded all fines incurred by drivers as being payable by the driver without reimbursement by the respondent, although the evidence before me was that the respondent might be willing to appeal a parking fine. That was clear from paragraphs 113-115 of the witness statement of Mr Balog, where he said this: “113. While driving for Addison Lee, I estimate that I paid approximately £2,000.00 in parking fines and traffic penalty notifications that were deducted from my weekly combined invoice/statement. In addition to this, I estimate that I may have paid an additional £500.00 in fines, when tickets were issued to me directly by a parking enforcement officer, and not sent to Addison Lee.114. As a courier, I received fines for illegally parking or exceeding the allowed time in a loading bay. This was usually because I was either loading or unloading heavy items from my van, and I needed to park close to where I was delivering or collecting the items, which had parking and waiting restrictions. At other times, collecting or delivering a parcel could take longer, especially if I picked it up from an office in a high-rise building.115. The penalty notices were sent to Addison Lee, who sent them on to me with a letter informing me that Addison Lee would deduct the cost from my weekly earnings. I was sometimes given the option by Addison Lee to appeal the parking penalty but usually I did not have good grounds for appeal.”[163]Similarly, Mr Da Silva said this in his witness statement: “103. During the course of my driving, I sometimes received parking fines or speeding tickets, or other driving related fines. The fines were sent to Addison Lee, who paid them automatically and deducted the amounts from my wages. These deductions can be seen in my invoice statements, for example at page 8810. 104. On occasion Addison Lee sent me the fine beforehand in case I wanted to appeal it, but usually they paid it automatically, so I did not have the chance to appeal. I recall one such incident which I would have challenged involving being caught in the yellow box when turning right (which is not illegal) but was unable to challenge because Addison Lee paid it automatically without informing me until after it had been paid. I do not recall when this was.”[164]Mr Nardelli’s evidence in paragraphs 60 and 61 of his witness statement was to the same effect. So was that of Mr Edah-Tally in paragraphs 28-29 and 150 of his witness statement. Mr Ruiz’s evidence on this in his first witness statement was comparable (it was in paragraph 74 of that statement) but he also said, in paragraph 75 of that statement, that as “an owner driver, I was responsible for dealing with any parking or traffic violation fines”.[165]The other evidence of the claimants about motoring fines was to the same effect. I saw that Mr Klepacki said this in his first witness statement about parking fines and penalties: “116. Sometimes I received parking penalties, especially when I was waiting to pick up a passenger. Addison Lee usually decided whether to pay or appeal the fine without consulting me. If the fine related to parking on a double red or yellow line, they would generally just pay it and just send me a copy of the fine showing a picture of my vehicle and registration number. Once they had paid it, they deducted the fine from my weekly earnings. 117. I usually accepted that it was my fault, but I could not have contested these fines even if I wanted to, because Addison Lee had paid it before sending me the notification.”[166]That evidence rather suggested that there was scope for avoiding parking fines. It also in my view pointed the way towards the right conclusion about any other fines, and that was that it would be contrary to public policy to take them into account in determining what the claimants were paid by the respondent. On this sub-issue, therefore, I was against the claimants. Parking expenses[167]Parking expenses were, however, different. They were simply charges for parking and when they were incurred for a particular job then they were in my view properly deductible, whether or not the respondent reimbursed the claimant in question for the expense in question. I accepted all of what the claimants said in paragraph 229 of their closing submissions in this regard, so that I accepted that it was correct to deduct from the lead claimants’ earnings the parking expenses referred to in that paragraph.[168]In addition, I agreed with what the claimants said in paragraph 230 of those submissions about parking charges incurred otherwise than in connection with a particular job by a non-partner driver, “e.g. if it became necessary to pay for parking when waiting to join an airport rank or to park in Central London in order to go to the toilet.” Those expenses were plainly within the scope of the ruling in Augustine. I understood what the respondent said in paragraphs 161 and 162 of its closing submissions to be an acceptance of that proposition, subject to the “one day leeway” point, on which I disagreed with the respondent (see paragraph 134 above). Apportioning of multi-apping expenses[169]As for a partner driver, I regarded what the claimants said in paragraphs 218-221 of their closing submissions to be largely apt and applicable in relation to expenses properly incurred. My only reservation in this regard is that it would be wrong to say that it was necessary or apt to apply “assumptions” in the apportioning of expenses. Rather, I had to decide whether the estimates of the two lead claimants who were for a period partner drivers about the proportion of their time spent driving for the respondent were reliable. In the absence of any reliable evidence to the contrary, and given that Mr Ruiz (in paragraphs 66, 85 and 101 of his witness statement) and Mr Klepacki (in paragraphs 20, 33, 82 and 111 of his witness statement) stated what looked to me like reasonable estimates, I accepted that they spent, respectively, 20% and 35% of their time when they were partner drivers driving for the respondent. I should say for the avoidance of doubt that in coming to that conclusion, I considered carefully what Mr Davies said in paragraphs 13-22 of his witness statement but I concluded that it was not enough to justify rejecting Mr Ruiz’s and Mr Klepacki’s estimates. Self-assessment expenses[170]In paragraph 231 of their closing submissions, the claimants pointed out that the respondent had “not treated the cost of self-assessment services provided by Cab Accounting to HH via an arrangement established by AL as a reduction within the meaning of regulation 13 NMWR 2015”.[171]When I was deliberating, it occurred to me that the cost of preparing a selfassessment tax return is not regarded as a deductible expense for income tax purposes, but that accountancy costs incurred in connection with calculating the profit of a self-employed person are such a deductible expense. I knew that from my own experience, and I saw that the HMRC website with the address https://www.gov.uk/expenses-if-youre-self-employed/legal-financial was consistent with that experience. Of course, however, I could not take that into account without giving the parties an opportunity to object to me doing so, but I could see that the respondent had applied the principle that the cost of preparing a self-assessment income tax return is not taxable to the expenses which were deductible for the purpose of determining what income a claimant received here. It then occurred to me that the respondent had (it was clear from the questions asked by Mr Leiper in cross-examination) worked on the basis that a lead claimant who used Cab Accounting must have had a statement of his profits prepared by Cab Accounting in order to ascertain what was the income for self-assessment purposes. The only lead claimant who had used Cab Accounting was Mr Ruiz. I therefore looked with care at what Mr Ruiz said about that. It was in paragraphs 41-43 of his witness statement.[172]It was likely that the cost of using Cab Accounting was intended to cover both the time spent in calculating profits and the time spent in completing a selfassessment income tax return. I guessed that the time spent on the latter would be rather less than the time spent on the former, given the relative simplicity of a tax return for a person providing driving services to a supplier of work such as the respondent, and I concluded that I should estimate that 80% of the cost of using Cab Accounting was spent in calculating the profit and that that proportion of the cost of using Cab Accounting was an expense which fell to be deducted from Mr Ruiz’s income for the purpose of determining whether or not he received the NMW. The Expense Allowance in the June 2021 contract[173]In paragraph 234 of their closing submissions, the claimants said that “In relation to the period following introduction of the Driver Contracts in June 2021, AL’s CSOLs ‘assume’ – with no evidential or logical basis - that the Expense Allowance of £0.45 per Job Mile referred to in clause 6 of that contract compensated the drivers for all the expenses they incurred in connection with their employment – see CSOLP 11.19 and 11.20.”[174]Given what I said in paragraphs 329 and 330 of my 2025 judgment, I was not sure why this was an issue before me. However, I was able to conclude now that the claimants’ costs were what they were, and that the payments by the respondent of “the Expense Allowance of £0.45 per Job Mile” had to be regarded simply as part of the money earned by the claimants, from which the claimants’ actual expenses were to be deducted for the purpose of calculating their income and seeing whether or not they were paid the NMW. Negative remuneration[175]That led me to the final sub-issue of issue 6. Only when deliberating did I come to any clear view on it. I was initially sceptical about the claimants’ position, but when I returned to it for the purpose of coming to conclusions on the disputed issues, I thought that their position might well be right: it might well be contrary to the intention of Parliament if(1) a worker who had the right to the NMW incurred deductible work expenses in a week, and(2) those expenses exceeded the payments made by the employer to the worker, so that(3) the worker had to stomach the loss and was to regarded as having received no income rather than what was in effect a negative income. However, on the other hand, if that were the case, then Parliament might well be taken to have concluded also that the calculation of the income of the worker should be made by reference to a period greater than the single week, in order to avoid injustice to an employer whose workers’ incomes fluctuated significantly during the year.[176]I then returned to the provisions of the NMWR 2015, and I could see no alternative to the conclusion that an employer was required to pay in a pay reference period no more than the NMW, and that a worker who earned more than the NMW in another pay reference period could not be required to give credit for the amount earned above the NMW. Therefore, on this issue, I agreed with the following statement in paragraph 185 of the respondent’s closing submissions: “The Respondent contends that the minimum effective hourly rate a driver can be paid is £0 and the maximum hourly underpayment is the statutory NMW rate.”[177]I was by no means sure that a claim for damages for breach of contract, or for a sum due under a contract, could be made for any unrecovered expenses, but I saw that it was contended by the respondent in the next sentence of its submissions that it could. That sentence was: ‘Should there be any expenses which a driver has not recovered at the end of a reference period, the appropriate method by which to recover such expenses is a contractual claim for repayment of that sum and not an “hourly recoupment” amount as suggested by the Lead Claimants.’[178]It seemed to me that there may be such a route for recovery, but that Parliament had not catered for this kind of situation when making the NMW legislation.[179]I saw from paragraph 148 of the respondent’s closing submissions that the parties agreed that a negative income figure for a week could be offset when calculating pay for the purposes of the WTR, but that was because of the long period over which pay for those purposes is calculated (52 weeks now, as discussed in paragraphs 107-109 above).[180]I also saw from paragraph 138 of the respondent’s closing submissions that the respondent was firmly against the proposition that one could import a provision relating to the NMW into the WTR, which, it was said in that paragraph, “would be contrary to general principles of statutory construction.”[181]In all of the circumstances, I concluded that the respondent was right in saying that a negative figure for income in a pay reference period for the purposes of the NMWR 2015 meant simply that the worker was to be treated as having received no income. Given what I say in paragraph 176 above, there was a balancing factor which meant that that outcome was not an absurdity, and that was that no credit needed to be given by a worker for any income earned above the NMW in any other pay reference period. Issue 7: Has the respondent made a series of deductions from wages in not paying the NMW and, if so, when should that series be regarded as having been broken?[182]The claimant’s closing submissions on this issue were the same as their opening submissions on it. I now quote them because I agreed with them. ‘248. AL’s consistent failure to pay the NMW to the LCs plainly gave rise to a series of unlawful deductions from pay within the meaning of section 23(3) ERA.249. The submissions at paragraphs 100 to 103 above [in fact, that was a reference to paragraphs 100-103 of the opening skeleton argument, which were repeated as paragraphs 169-172 of the claimants’ closing submissions] as to the effect of the Supreme Court’s decision in Agnew apply equally here.250. Each of the LCs contends that the series of deductions constituted by AL’s failure to pay them the NMW commenced on the date on which their employment relationship began and continued until the date on which it ended, whether or not there were some pay reference periods in which the relevant LC in fact received the NMW.251. All of the relevant underpayments were linked by the “unifying vice” of AL’s refusal to accept that the LCs were workers who were entitled to the NMW and its failure to calculate their pay in a manner designed to ensure that they received the pay to which they were entitled under the national minimum wage legislation.’[183]My conclusions stated in paragraphs 99-100 above therefore applied here also. Issue 8: What additional compensation should be awarded to each lead claimant? In particular, should interest be awarded or only compensation for such losses as could be the subject of compensation under section 24(2) of the ERA 1996? In answering this question, the impact of the decision of the ECJ in Marshall v Southampton and South West Area Health Authority (No 2) C-271/91 [1993] ICR 893 will need to be considered[184]I had already looked at this issue, which arose only in relation to the WTR: I did so in paragraph 323 of my 2025 judgment, where I came to a conclusion on it. When returning to the issue in my deliberations after the February 2026 hearing, I came to the same conclusion, despite the respondent’s arguments against it in paragraphs 191-198 of its closing submissions. I could not see how the decision of the ECJ in Marshall could be distinguished simply because it was applied to a specific piece of European Community (as it was then) legislation which did not cover the same ground as that which was covered by the WTR. I was fortified in that regard by the decision of Bourne J in Main v SpaDental Ltd [2024] EAT 200, [2025] ICR 1013.[185]What was an appropriate award to make here was plainly a matter of judgment, and in my judgment it was right to do what the claimants sought here in paragraph 262 of their closing submissions, and that was to award the claimants “interest on any sums awarded to them in respect of their claims for holiday pay at 8% per annum”. Consolidation[186]The question of consolidation of all of the new claims which had been made since a stay was put in place was raised by the parties. I could not see any good reason why they should not be consolidated with the previously existing claims, irrespective of any difference in approach taken in them, or in response to them, from the approaches taken to the previous claims. That was not least because consolidation does not require a uniformity of approach, and the fact that the claims are consolidated (the word “consolidate” is not used in the 2024 rules) means only that they are joined to be determined at the same time. In conclusion[187]I have set out in my above judgment my conclusions on the points which it appeared to me to require a formal conclusion. If I have omitted anything which either of the parties believes should be the subject of a judgment, or if it appears that I have erred in any way in any part of what I say above (for example, I am not certain that ruling number 1 aptly states the end of what is referred to by the claimants in paragraph 54 of their closing submissions as “a work shift (that is, the period within any 24 hours when the driver was clearly working)”), then I will be glad to reconsider what I say above. In that regard, time is extended for the purposes of rule 69 of the 2024 rules to 56 days. Approved by CERTIFICATE OF CORRECTION Under rule 67, the judgment sent to the parties on 19 March 2026 is corrected in the following ways.[1]The following words are inserted at the end of paragraph 92 of the reasons: “there was no break in the employment relationship for the purposes of the Working Time Directive between92.1 Mr Ruiz and the respondent between 22 March 2020 and 9 February 2021, and92.2 Mr Mahendran and the respondent between 16 March 2020 to 28 August 2020 and 2 January 2021 to 16 April 2021.”[2]The formatting of paragraph 2(1) of the judgment is corrected by the joining of the text after “first” with the rest of that subparagraph.[3]The numbering of the subparagraphs of paragraph 4 is changed from1.1 and1.2 to(1) and(2) respectively and the numbers (1) and (2) in what is now subparagraph (2) are changed to (i) and (ii). Approved by