Syed Asim Raza Gillani & Ors v Veezu Limited & Ors

This is a claimants’ appeal from a case management decision of Employment Judge Jenkins sitting in Cardiff on 4 October 2024. The judge is managing claims brought by (currently) 509 drivers of private hire cars against the respondents, which are “Veezu” companies that provide private hire transport to the public in various parts of Wales and England. The current “multiple” is not closed. I am told claimants have been added since the judge’s decision and others may be added in future.

The claims are for holiday pay, failure to pay the national minimum wage and failure to provide written statements of particulars of employment. The claims rest on the proposition that the claimants are workers or agency workers under the relevant legislation. A preliminary issues hearing is to take place over 20 days in about a year from now, in June and July 2026. At the hearing, the tribunal will determine whether certain lead claimants are or were “workers” and if so, at what times.

The case management model is the now familiar one involving the selection of lead cases and lead claimants under rule 36 of the former 2013 Employment Tribunal Rules of Procedure (Footnote: ¹). The other claims are stayed behind the lead claims. The order appealed against is for the provision of information by the claimants. It is part of a suite of directions to facilitate the preliminary issues hearing just mentioned.

The actual words of the order appealed against, which I will call the contested order, are:

“By 4 June 2025, the Claimants shall give further particulars of:

(a)

The dates and times that they were logged onto any other apps for the purposes of driving, delivery or related private hire or courier services;

(b)

The dates on which they made themselves available to other businesses and/or clients for the purposes of driving, delivery or related private hire, chauffeur or courier services; …”

The deadline of 4 June 2025 could not be met because of this appeal. The oral hearing before me was held on 10 June 2025. The judge below has therefore stayed the contested order pending the outcome of this appeal and it is agreed that, if the appeal fails and the contested order stands, a new deadline will need to be set and that any revised timetable should not, if at all possible, imperil the hearing dates for the preliminary issues hearing.

The claimants contend in three overlapping grounds of appeal that the judge disregarded relevant matters or took account of irrelevant ones; that the decision to make the contested order was irrational and perverse; that he failed to apply the overriding objective properly or at all; and that he failed to provide adequate reasons for his decision. The respondents contend that the decision was properly made in the exercise of wide case management powers; that it was proportionate and in accordance with the overriding objective; and that it was properly reasoned.

In the judge’s words, accepted by both sides (Footnote: ²), the claimants (RPH [2]):

“work … as private hire drivers. They make themselves available for hire via an app operated by one or other of the Respondents. They then provide services to passengers who make requests of the relevant Respondent via an app, with the Respondent, in turn, notifying the Claimants and other drivers logged in to the app, of the potential engagement, which, if accepted, is then undertaken.”

The first claims were brought in 2023. Preliminary decisions were made in 2024. The cohort of claimants has grown from six at the start of the proceedings to over 500. The claims are consolidated and are being managed together in Cardiff by Employment Judge Jenkins. A video directions hearing was fixed for 4 October 2024. Junior counsel for the claimants, Ms Jennifer Danvers, liaised with the respondents’ leading counsel, Mr Richard Leiper KC, in advance of the hearing to explore what could be agreed. There were three areas of disagreement.

The first was about how to select lead claimants and how many should be selected. It was agreed that the parties would seek to reach agreement on the categories of claimants from which lead claimants would be selected. The claimants proposed that after categories were agreed, the claimants should put forward 12 potential lead claimants from which each side would select up to four claimants to be lead claimants at the preliminary issues hearing. The respondents proposed that each side should nominate up to 10 lead claimants in the agreed categories, making 20 lead claimants in all.

Secondly, the respondents sought a direction for early identification of claims that were out of time. The claimants opposed that. Thirdly, the respondents sought a direction in terms similar to what became the contested order, together with a direction for identifying any claims brought by the claimants against other (non-Veezu) respondents relating to the same periods as their claims against Veezu respondents. The next day, the judge heard argument and gave his decision orally.

A few days later he issued the RPH, List of Issues and Orders and Directions. In the RPH at [6], he noted the three areas of disagreement. He referred at [8] to the overriding objective and noted that references “relating to proportionality, flexibility, avoiding delay, and saving expense were particularly relevant in this case”. He noted at [9] that consideration of the disputed matters arose in the context of the forthcoming preliminary hearing to decide the question of worker status.

On the disputed matters, he declined the respondents’ request for a direction enabling early identification of out of time claims. However, he granted the respondents’ request for a direction in the terms of what became the contested order. In relation to lead claimants, he preferred the claimants’ proposal. It is necessary to set out part of his reasoning in the RPH at [13]-[20]:

“Request for information (6.2 above)

13.

With regard to the request for information be provided by the Claimants about their other related activities, I noted that, in relation to the assessment of status, the Tribunal will be required to consider a number of factors, including; the degree of control exercised over the Claimants by the Respondents, and the degree to which the Claimants are integrated in the Respondents’ organisation. The assessment will also involve consideration of whether a Claimant could be said to be operating a business undertaking of which the Respondents were customers or clients.

14.

The fact that Claimants are engaged in relation to other businesses, whether via apps or otherwise, will therefore have a potential bearing on that assessment, and I considered that it would be appropriate to make the orders sought at 6.2.1 and 6.2.2 above.

15.

…

16.

…

Identification of Lead Claimants (6.3 above)

17.

With regard to the identification of lead Claimants, whilst I took the Respondents’ point that all of the Claimants would, in a more limited case, be expected to give evidence, this is a large multiple case, where a narrowing down of the Claimants, in terms of those from whom evidence would be taken, is clearly appropriate.

18.

I could see the potential merits of both parties’ suggestions, but, having directed that the orders to provide further information above should be made, I considered that the Respondents will be in an informed position to consider a list of potential Lead Claimants put forward by the Claimants.

19.

Allowing the Claimants to put forward potential Lead Claimants first will enable the Claimants’ representatives to take into account the particular suitability of each potential Lead Claimant, in terms of matters such as their availability, their location. and their English language ability. That would avoid difficulties in those areas which could arise were the Respondents to be involved in the initial identification of Lead Claimants.

20.

Overall, therefore, I considered that it would be appropriate to make the order in the terms put forward by the Claimants[.]”

In his Orders and Directions at Appendix 2, at [5]-[8] the judge included the following:

“5.

By 4 June 2025, the Claimants shall give further particulars of:

(a)

The dates and times that they were logged onto any other apps for the purposes of driving, delivery or related private hire or courier services;

(b)

The dates on which they made themselves available to other businesses and/or clients for the purposes of driving, delivery or related private hire, chauffeur or courier services;

6.

The parties shall endeavour to agree a list of Lead Claimants to give evidence at the preliminary hearing. To that end:

(a)

By 1 November 2024, the Respondents will identify the categories of case for which Lead Claimants should be identified.

(b)

By 22 November 2024, Leigh Day shall indicate whether those categories are agreed (and, if not, give reasons and identify alternative categories).

(c)

By 13 December 2024, the parties shall seek to agree such categories.

(d)

By 7 February 2025, the Claimants (jointly) to nominate 12 potential Lead Claimants who cover the agreed categories.

(e)

By 7 March 2025, the Claimants (jointly) and Respondents (jointly) to select up to 4 Lead Claimants each from that list.

(f)

By 28 March 2025, the parties shall confirm whether the proposed Lead Claimants are agreed, and if not provide their reasons for objecting to any proposed Lead Claimant. If agreed, the parties shall deliver to the Employment Tribunal a list of the agreed Lead Claimants who will give evidence at the preliminary hearing.

(g)

Where objections are raised, the parties shall, by 11 April 2025, seek to address and resolve any such objections, including as appropriate by agreeing alternative Lead Claimants. If agreed, the parties shall deliver to the Employment Tribunal a list of the agreed Lead Claimants who will give evidence at the preliminary hearing.

(h)

There shall be a preliminary hearing listed on 1 October 2025 for 3 hours for the purpose of resolving any outstanding disputes as to the identification of Lead Claimants. If they are agreed, then the parties are to write jointly to the Tribunal for that hearing to be vacated.

7.

Upon agreement (or determination) of Lead Claimants, the remainder of the claims will be stayed and case managed under rule 36 of the Employment Tribunals Rules of Procedure 2013.”

There was then a disagreement in correspondence in October and November 2024 about whether the parties should jointly apply to vary the direction on “multi-apping”, i.e. the contested order. The claimants suggested its requirements should be confined to lead claimants, not all claimants. I need not, at this stage, go through the detail of the arguments rehearsed in that correspondence. The respondents did not agree to the proposal and said they would oppose any application to vary the contested order.

The claimants then brought this appeal against it. His Honour Judge Auerbach directed that it should proceed to a full hearing. He accepted that the information sought in the contested order could be relevant both to worker status generally and more narrowly to identifying lead claimants. But he was concerned that the width of the order and the burden it placed on the claimants was arguably disproportionate to any evidential benefit at the preliminary issues stage.

Judge Auerbach also noted that the timetable envisaged agreement on lead claimants by late March 2025, when the respondents would not yet have the information, due by 4 June 2025, to be supplied under the contested order. He thought the employment judge may have made some “unintentional error …. in relation to the timetable.” He thought it best to list the appeal for a full hearing but that would not preclude the tribunal, on its own initiative or prompted by the parties, from “revisiting this matter” before the hearing of the appeal.

The claimants’ solicitors took up Judge Auerbach’s suggestion and sent a copy of his written reasons to the tribunal, together with an estimate of the costs and burden of submitting data subject access requests (DSARs) for each claimant (about £300,000 to £450,000 plus VAT for 200 claimants); and proposing a variation to the contested order so as to limit it to lead claimants only, “[r]ather than covering the entire claimant cohort”. They submitted that:

“… the proposed variation would be in line with the overriding objective of dealing with cases in ways which are proportionate to the complexity and importance of the issues, of avoiding unnecessary formality and seeking flexibility in the proceedings, of ensuring the parties are on an equal footing, and of saving expense.”

The respondents’ solicitors wrote to the tribunal on 6 February 2025 opposing the claimants’ proposed variation. They expressed scepticism about the complaint that compliance would be unduly onerous and costly; on the contrary, they said it would be easy for individual claimants to comply as they would either have the information already or easy access to it from the other companies. They accepted that there was an “anomaly in the sequencing” and they did seek a variation in the timetable, so that the dates in February and March 2025 for identifying lead claimants be put back to dates in June to August 2025, after the deadline of 4 June 2025 for complying with the contested order.

The claimants’ solicitors responded in detail on 10 February 2025, objecting to the respondents’ proposed variation and taking issue with the respondents’ position on every point. The appeal proceeded in tandem, with the respondents filing their respondents’ answer on 19 February 2025, defending the judge’s decision to make the contested order, noting that the claimants had not asked him to amplify his reasons but recognising that if the appeal tribunal considered that the dates for compliance with the employment tribunal’s directions needed to be changed, the respondents would invite the appeal tribunal to substitute revised dates.

I interject here that, to recap slightly, the proposed categories of case from which lead claimants were to be identified, were to be agreed by 13 December 2024 (see the timetable at [6](a)-(c) in the Orders and Directions). Correspondence about this (which it is unnecessary to go through) had been passing between the parties since the hearing on 4 October 2024. By 25 February 2025 the parties were quite close to agreeing (and have since agreed) six of the seven categories set out in the respondents’ solicitors’ letter of that date, as follows:

“ …. we suggest that the following categories be agreed with the understanding that a nominated Lead Claimant is able to represent more than one category:

1.

A Claimant representing each of the following operating areas:

a.

Veezu Limited

b.

Veezu Midlands Limited

c.

Veezu North Limited

d.

Northern Taxis Limited

e.

Panther Cambridge Limited

f.

A.B.C. Taxis (EA) Ltd

2.

A Claimant who derives the majority of their income from journeys offered via the Veezu iCabbi system as at the date of their claim;

3.

A Claimant who derives the majority of their income earned from journeys offered via the Veezu iCabbi system in cash;

4.

A Claimant who derives the majority of their income earned from journeys offered via the Veezu iCabbi system from Corporate Account Customers; and

5.

A Claimant who has their own Hackney Carriage licence or is dual licenced.

6.

A Claimant who is licenced as their own PHO

[Note: this category is not agreed because the claimants say there is no claimant with their own private hire operator (PHO) licence.]

7.

A Claimant who derives the majority of their income earned from journeys offered via the Veezu iCabbi system via a pre-registered credit card on the Veezu app.

Based on the above list, it should be possible to nominate two Claimants from each regional operator who also represent at least one of the other 6 categories identified.”

The judge responded to the variation applications by an email from the tribunal dated 7 March 2025. He preferred the respondents’ proposed variations. In explaining why, he addressed Judge Auerbach’s comments on the issue of proportionality of the contested order and the issue of timetabling and the sequencing he had ordered. It is necessary to set out his comments in full:

Later, on 29 April 2025, the judge acceded to the claimants’ suggestion, not objected to by the respondents, that the contested order should be stayed pending the outcome of this appeal, with further directions to follow as necessary, in the light of the outcome. The parties then prepared in the usual way, producing bundles and skeleton arguments. I heard oral argument and written submissions both before and after the hearing. I reserved this judgment.

The starting point is the decision of the Supreme Court in Uber BV v. Aslam [2021] ICR 657, in which Lord Leggatt JSC (the other justices agreeing) answered yes to the two questions at [1]:

“whether an employment tribunal was entitled to find that drivers whose work is arranged through Uber’s smartphone application (‘the Uber app’) work for Uber under workers contracts and so qualify for the national minimum wage, paid annual leave and other workers’ rights; or whether, as Uber contends, the drivers do not have these rights because they work for

Themselves as independent contractors, performing services under contracts made with passengers through Uber as their booking agent. … [and] … whether the employment tribunal was also entitled to find that the drivers who have brought the present claims were working under such contracts whenever they were logged into the Uber app within the territory in which they were licensed to operate and ready and willing to accept trips; or whether, as Uber argues, they were working only when driving passengers to their destinations.”

Other cases have followed in the wake of that seminal decision. As Mr Leiper KC points out, the business models within the industry have moved on; private hire providers have proliferated; and technology has developed. In particular, multi-apping, not a feature of the Uber appeal, is now commonplace. Drivers may offer to ferry passengers for more than one private hire provider, logging on and off their various apps at will; perhaps logging on to more than one app at a time; sometimes accepting a job, sometimes not; and accepting jobs now for one provider, now for another.

Two ongoing multiple claimant employment tribunal decisions cited to me show how the case law is developing to accommodate these developments in working methods and patterns: Bandi v. Bolt OÜ (2206953/2021), at London Central; and Afshar v Addison Lee Ltd (3306435/2020) at Watford. At least one is subject to a pending appeal. It is unnecessary to go through those cases. They establish, beyond question, that evidence about the incidence and frequency of multi-apping by a particular claimant is of obvious potential relevance to whether he (Footnote: ³) is a “worker” or not.

I interject that while the claimants said the information required under the contested order is (to quote from the claimants’ skeleton argument) “not necessary, nor relevant, for the ‘overall picture’”; they accepted later in the same skeleton, correctly, that “a factor that may be relevant to worker status is whether the drivers can and do provide services to others …. .” It is obvious that the incidence and frequency of any multi-apping may shed light on that issue. The separate complaint that disproportionate “granular detail” is required, should not be confused with relevance in principle.

It is common ground that the contested order is a case management decision and as such within the broad discretion of the tribunal below. It is only susceptible to appeal if the tribunal has exceeded the generous ambit within which reasonable disagreement is possible. Applied to this appeal, the issues are whether the judge went outside that generous ambit when making the contested order, i.e. whether in deciding to make that order he acted perversely or irrationally, on flawed or insufficient reasoning; or by imposing an oppressive and disproportionate burden on the claimants.

An order for particulars may be given to avoid a party being taken by surprise at the last minute. Particulars should only be ordered “when necessary to do justice in the case or to prevent adjournment”; the order “should not be oppressive”; and “particulars are for the purpose of identifying the issues, not for the production of evidence”: per Wood J in Byrne v. Financial Times Ltd [1991] IRLR 417, at [18]. Such is the classical exposition of the function of particulars, in the context of litigation in what are now employment tribunals.

A more modern exposition of principles governing orders for provision of documents (disclosure), orders for provision of information (particulars), or a combination of the two, is found in His Honour Judge Auerbach’s account in Bari v. Richmond and Wandsworth Councils [2025] EAT 54. It includes at [33]-[47] a useful review of the modern authorities in the age of the overriding objective. It does not detract from the rationale of Wood J’s classical exposition in Byrne, but the width of the tribunal’s procedural powers has increased since Byrne was decided.

As for disclosure, the authorities cited by Judge Auerbach in his review, which I need not repeat here, show that relevance is a necessary but not sufficient condition for ordering disclosure and that the degree of relevance or potential relevance may be relevant to an assessment of proportionality. He also addressed the making of orders of the kind at issue in this appeal, where the request is for information but to get that information the party providing it has to obtain documents and collate information from within them. As he said at [44]:

“… the general test and principles are broadly the same whether the application is for disclosure or for information. However, when it comes to the application of the test to a request for information there may be real practical differences. If what is being sought is not an existing document or documents, but pure information, the task involved in complying with the order, if made, may be practically very different. For example, it might require considerable work to find and collate the information or to ascertain if it even exists. It may require analysis or processing of raw information in order to answer the specific questions asked. Even if some answers might turn out to be found within existing documents, the substantive nature of the exercise may involve significantly different work. Of course, what would be involved in the given case is case-sensitive.”

The then rule 64(2) of the ET Rules of Procedure 2013 (Footnote: ⁴) provided that the reasons given for any decision “shall be proportionate to the significance of the issue and for decisions other than judgments may be very short”. Where the complaint is that the tribunal’s reasoning is insufficient, a request for amplification should be made before an appeal is pursued: Bansi v. Alpha Flight Services (Note) [2007] ICR 308, per His Honour Judge Serota at [20]-[23]; Royle v. Greater Manchester Police Authority [2007] ICR 281, per His Honour Judge Burke QC at [28]-[35].

For the claimants, Ms Schona Jolly KC submits that the particulars sought are “self-evidently disproportionate, on their own terms”. Proportionality is “a fundamental consideration underlying all aspects of this Appeal”. The claimants have never denied that some of them multi-app. This point is “covered off” by agreeing a category of lead claimant who derives more than, or less than, a certain percentage of his income (say, 25 per cent, 50 per cent or 75 per cent) from doing jobs for the relevant respondent. The respondents have now agreed to a category based on 50 per cent of earnings.

Having agreed that definition, the respondents do not need to know the full details of all multi-apping in the case of every lead claimant; still less, of every one of the 509 claimants. As for the lead claimants, they were supposed to be named before the information was due to be delivered, making the timing wrong. The judge has accepted that this was a mistake. The relevance of multi-apping to the “overall picture” is therefore the only and flimsy basis for making an order that would be very expensive and burdensome to comply with, especially since many drivers have limited English.

Ms Jolly said the contested order is irrational, perverse, oppressive and disproportionate, essentially for those reasons. Particulars should be ordered to define a party’s case and the issues, not to produce advance evidence. The process of collating the required information would involve numerous DSARs made in the name of the individual claimant against other entities for which he does passenger transport work. The information would have to be vetted by the claimants’ solicitors who might need to have privileged conversations about it. It was not correct, Ms Jolly submitted, that the claimants would have access to this information without making DSARs.

The employment judge himself must have recognised how burdensome would be the process of compliance, for he allowed about eight months for it, leading to the preliminary issues trial being fixed for June 2026 instead of October 2025. He should have considered less onerous measures and should not have dismissed the complaint about the cost and burden of complying. There was no basis for “his breezy conclusion on the apparent ease of obtaining the information requested”. There would be difficulties and the respondents would not be slow to assert non-compliance and demand sanctions.

As for the insufficiency of reasons, the claimants did not seek to add anything to the points made in the third ground of appeal. I have looked again at these and they do not add to the claimants’ points in the first ground (irrelevancy, irrationality or perversity) and in the second ground (failure to apply the overriding objective and imposing an unjust oppressive and disproportionate burden). The reasons challenge in ground 3 has to be read in the light of the additional reasoning offered by the judge in his email of 7 March 2025.

In reply, Ms Jolly added that evidence of the frequency and extent of multi-apping by each claimant would only be useful in conjunction with evidence of the reasons why claimants engaged in multi-apping. The information alone would tell the respondents and the tribunal nothing about that. The respondents could obtain their own evidence on that issue, for example by conducting a survey seeking to elicit the motivation of drivers for multi-apping, as well the frequency and extent of it.

Mr Leiper submitted that the RPH, the List of Issues and the Orders and Directions, amplified by the reasoning in the judge’s email of 7 March 2025, demonstrated that he had exercised his case management powers properly and lawfully, weighing the arguments and reaching a conclusion that was well within the generous ambit within which reasonable disagreement is possible. He had made specific reference to the cogent elements of the overriding objective. His reasoning was careful and nuanced. He had found against the respondents on other issues, i.e. selection of lead claimants and dealing with out of time claims. There was no basis for this appellate tribunal to interfere.

The degree of relevance to worker status was high, Mr Leiper submitted. The judge had made the contested order in furtherance of the “overall picture”, not to help identify lead claimants. If the latter had been his purpose, he would not have mistakenly got the sequencing wrong in the way he did; he would have postponed identification of lead claimants until after the contested order had been complied with. The essential part of the reasoning was that multi-apping has a bearing on worker status, a self-evidently sound point the claimants’ solicitors had acknowledged in correspondence.

The claimants, said Mr Leiper, could not complain about absence of proper reasoning because they had failed to ask the judge to expand upon his reasons, as required according to cases such as Bansi and Royle. The inference was that the claimants preferred to present the reasons unadorned by any subsequent expansion of them and then complain that they were inadequate. The reasons were cogent and not required by rule 62(4) of the ET Rules of Procedure to be lengthy. They were then, in any case, by a “sidewind” expanded upon after Judge Auerbach’s comments at the sift stage.

As for the proportionality argument, the claimants had chosen to allege that they were not in business on their own account; that the respondents exerted a significant degree of control over their work activities; that although they can choose when and where they work, there are restrictions on their freedom to accept jobs at will; and that they are integrated into the respondents’ business. It is not denied that a number of them multi-app and drive for other companies. It was a matter for the judge to decide how to address those aspects of the dispute by case management orders.

The fact that the lead claimant case management model is being used does not preclude, Mr Leiper submitted, case management directions that look beyond the cases of the lead claimants and bear on the dispute more widely. The contested order was not unduly onerous on the individual claimants. Those who multi-app would only have to make a small number of DSARs; and should not have to make any as they ought to have the information already in order to fill in their tax returns. Proportionality does not engage the claimants’ position alone; the respondents faced liabilities running to tens of millions of pounds if the claims were to succeed in full.

Any delays should be seen in the context of long drawn out litigation. The first claims were issued in October 2023, not earlier. The need for detailed information about multi-apping arose because the business model had changed. In earlier cases, multi-apping was not possible at all, or only possible for drivers who had their own car, not bearing the livery of the transport provider. The contested order was tailored to the evidential needs in this particular case, where the cars are not liveried and the claimants can work for multiple transport providers. That was not so in earlier cases.

In oral argument, Mr Leiper said that if non-compliance issues arose, the tribunal would be likely to take a lenient view if there were good reasons. He gave no undertaking and accepted that the orders were absolute and not limited backwards in time. While they were not time limited or “best endeavours” orders, that did not make them disproportionate and oppressive. As for the cost, that was not a matter of concern to the claimants themselves but to their solicitors, who would be paying that cost, at least initially. The judge was justified in attaching little weight to that factor.

As is already apparent, I need not deal with the third ground of appeal (reasons) separately from the first two first grounds. I can take the first two grounds together as they overlap substantially. I ask myself whether the judge, when making the contested order, went outside the generous ambit within which reasonable disagreement is possible; whether he acted perversely or irrationally, i.e. whether he acted on flawed and insufficient reasoning; and whether the contested order imposed an oppressive and disproportionate burden on the claimants.

I accept that he had in mind the overriding objective, specifically referred to the pertinent aspects of it and did not misdirect himself in law. The reasoning was in general terms adequate. There is no question of him having unfairly favoured the respondents; he found against them on the two other issues that were not agreed. But that does not help me to decide whether his decision to make the contested order was rational and legally sound.

Multi-apping is clearly relevant to the issue of worker status, the subject of the preliminary issues hearing. The claimants’ attempts to argue that it was irrelevant to that issue were unconvincing and contrary to their own solicitors’ acknowledgment in correspondence. An order for particulars of the extent and frequency of multi-apping was justified provided it was not oppressive, i.e. provided the burden of complying with it was not out of all proportion to the case management gain and furtherance of the overriding objective that compliance would bring.

Thus far, the reasoning supports the respondents’ position. But from here on, the problems start. In the RPH at [18], the judge was clearly saying that the information derived from the contested order would help the respondents to pick its four potential lead claimants from the pool of 12 to be proposed by the claimants. The respondents “will be in an informed position to consider a list of potential Lead Claimants”, he said. It is therefore clear that he intended the contested order to have been complied with before the respondents made that selection.

It follows that the case management benefit of the contested order was, in the view of the judge, not just the “macro” one of helping with “the overall assessment of employment status” (as he put it later, in his email of 7 March 2025). It was also, inescapably, part of the intended case management benefit of the contested order that imposing the burden on the claimants of complying with it would assist the respondents in selecting its four candidates for lead claimants from the pool of 12 candidates. At [18], he gave no other reason why the respondents would be well placed to pick from the pool; the only reason he gave was that they would have the benefit of the contested order.

At [19], the judge addressed his reasoning for allowing the claimants to go first in the exercise of selecting potential lead claimants. The claimants would be best placed to nominate as candidates those who were suitable in respects other than those to which multi-apping would be relevant, such as availability, location and command of the English language. Thus, the relevance of multi-apping to selecting lead claimants was expected to be addressed at the second stage, and by the respondents, once the contested order had been complied with. At [20], the judge said that “[o]verall”, he preferred the claimants’ proposal for selecting lead claimants. The word “overall” referred back not just to the reasoning at [19] but also to the reasoning at [18].

I do not criticise the failure of the claimants to seek further reasons from the judge. Ms Jolly aptly described this as something of a jury point. The claimants acted reasonably in seeking to explore a variation of the contested order, which the respondents opposed, while at the same time pursuing an appeal. It might well have caused more delay, in the context of multiple claimant litigation, to hold off from pursuing an appeal until further reasons had been explored. What the respondents would have advocated had further reasons been sought is not known; nor whether the judge would have been willing to provide them until prompted by HHJ Auerbach’s reasoning in his sift decision.

When the appeal came before HHJ Auerbach on the papers, he correctly recognised the relevance of multi-apping to the overall issue of worker status; and also correctly recognised, as the judge below did, the relevance of multi-apping to the narrower issue of selecting lead claimants; see his reasons for allowing the appeal to proceed, at [3]:

“… Although the grounds of appeal make the point that the mechanism provided for by the tribunal’s orders provides for the Claimants to nominate 12 Lead Claimants from which the 8 will then be jointly selected, prior to that the categories must be agreed, and that process is to start with those categories being proposed by the Respondents. Depending on what categories are agreed and/or decided by the tribunal, information of this type might then be considered to be relevant or needed to be provided, at least in relation to the nominated 12 Lead Claimants, to ensure that they are representative of the cohort.”

HHJ Auerbach was, rightly in my view, concerned about the width of the contested order. His observations indicated quite strongly that it was too wide. He did not just say that was arguable; he said it was “at least arguable” (ibid. at [7]). The judge below was therefore faced with powerful criticism from the claimants with, unusually, some support from the sift judge on appeal. It was in that context that the judge was encouraged to explain his reasoning further.

The respondents in their written correspondence were urging him to double down and defend his reasoning. The judge did so. In the email of 7 March 2025, he sought to play down the significance of the contested order helping the respondents to identify appropriate lead claimants from the pool of 12; and placed the emphasis on the utility of the contested order for illuminating the “overall” picture in relation to worker status. That is apparent from the following extracts:

“It was that ‘macro’ element which the Judge had in mind when summarising his reasons for making the orders at paragraph 5 of Appendix 2, as set out at paragraphs 13 and 14 of the main body of the RPH, rather than the ‘micro’ element of identifying lead Claimants.

….

For completeness, the Judge recognises the inconsistency, as identified by HHJ Auerbach, in paragraph 18 of the RPH …. On reflection, that was clearly a mistake, but the rationale set out at paragraph 19 still stands in relation to that issue.”

There are two difficulties with those passages. First, in reacting to criticism, the judge engaged in advocacy in opposition to the criticism. By aligning himself so closely with the respondents’ position, he was also, necessarily, seeking to undermine the force of the claimants’ criticisms, which were required to be the subject of dispassionate and impartial assessment at the full hearing which Judge Auerbach directed. The judge should have restricted his role to providing information about his reasons, but I think he crossed the line into advocacy in defence of his decision.

The second difficulty is that I do not find the judge’s reasoning “[f]or completeness” consistent with what he said at [18] in the RPH; which, I reiterate, is to the effect that a significant part of the case management benefit from the contested order would be helping the respondents to identify lead claimants. That difficulty with the reasoning is separate from any question of confusion over timetabling and dates. It means that the judge should have struck the proportionality balance with the “micro” as well as the “macro” element in mind. It is far from clear that he did so.

Furthermore, the timetable was such that, whatever the impact of any mistake about dates on the judge’s reasoning, the parties would already be closely involved in the process of agreeing categories of lead claimant long before the information about multi-apping would, on any view, become available to the respondents. The respondents would have to, and have had to, agree categories of potential lead claimants without the benefit of the information about multi-apping which was also relevant in principle to that exercise. That further undermines the judge’s reasoning.

I conclude that the email of 7 March advances an unsatisfactory ex post facto reconstruction of his reasoning. In those circumstances, it would not be safe to assume that the mistake about the timetabling and dates was an immaterial error, as the respondents’ propose. There was at least one serious flaw in the reasoning, amounting to irrationality. Either the reasoning was different from that stated at the time, or the order was made on the false premise that information gained from it would be available at the time of selecting the respondent’s four lead claimants from the claimants’ dozen.

I have therefore concluded that the reasoning is not defensible; the proportionality balance is vitiated by the exclusion from it of the purpose of helping select lead claimants. A proportionality balance for the purpose only of helping with the “overall” picture would be differently struck. A less onerous form of order would be appropriate for that more limited purpose, if any order at all were needed at this stage. The proportionality balance must be revisited.

I also accept the claimants’ submission that the contested order is disproportionate on its own terms, even factoring in its contribution to selection of lead claimants from categories. Taken cumulatively, the impact on the claimants is substantial. It is a burdensome order. To state the obvious, it is the most onerous of the three kinds of order considered in the Bari appeal earlier this year. It is neither an order for provision of information simpliciter, i.e. answers to questions, nor is it an order for provision of documents simpliciter.

It partakes of both: documented information must be obtained from not necessarily friendly non-parties. That means asking the questions of the non-parties, if necessary by serving a DSAR on them; second, receiving information and/or documents in response, or perhaps not; third, dealing with any delay or reluctance to answer the question or DSAR fully; fourth, collating and perhaps analysing the information and/or documents received from the non-parties; fifth, sending it to the respondents; and finally, dealing with any complaints from the respondents of less than full compliance.

That has to be repeated for each of the 509 claimants. A more measured approach would have been, for example, to confine the full exercise to the lead claimants or, perhaps, the pool of 12 potential lead claimants; to require the rest of the 509 claimants – or, perhaps, a random sample of say 25 per cent of them - to answer a questionnaire stating over a defined period (say, starting two years before the date of the claim) which other companies the claimant drove for, which apps were used and how frequently the claimants logged on to each such app.

Such an order would be more obviously proportionate to the case management benefits to be gained, without unfairness to the claimants. A party asks for particulars to know more about what the other side’s case is. Disclosure then informs the party how good the evidence to support that case is. It is unusual to require a party giving particulars to provide verification of them at the same time as giving them. The party receiving particulars or disclosure can approach the source and ask the source for verification or ask for third party disclosure and if necessary a witness order.

It seems to me most unlikely that claimants answering a questionnaire about their use of multi-apping would be so rash as to lie and risk the serious consequences of doing so. Any who were found to have done so would be likely to have their claim struck out, at the very least. The claimants’ solicitors would be astute to warn their clients about the importance of answering questions fully, accurately and truthfully.

If the claimants cannot always give precise answers, I do not see why they should have to do more than answer as far as they remember, or to the extent that they retain records. It is the claimants who have to prove their case and if they cannot do so, that will benefit the respondents. If answers given are scant or unsatisfactory, the respondents need not be slow to invite the drawing of adverse inferences, in the usual way when a witness gives an unsatisfactory answer to a question.

I am uncomfortable with the respondents’ dismissal of the burden of compliance, echoed by that of the judge in his 7 March email. I reject the speculative contention of the respondents that the claimants really ought to have the relevant information already, to prepare their tax returns. I am prepared to take judicial notice of the fact that a tax return requires keeping a record of the amount and timing of earnings, but that is all; it does not normally require production of detailed electronic day to day records. The respondents’ suggestion seems to be at odds with the real world.

The respondents also sought to rely on open source internet material from the website of Transport for London in support of their submission that the claimants could be expected to have good knowledge of English. That evidence was not given below, as far as I am aware. It is as much evidence where it appears in a skeleton argument as if it had been appended to a witness statement signed by Mr Leiper. Of more interest might have been evidence from the respondents themselves about the quality of the claimants’ English, which the respondents may know something about.

I conclude that the contested order is too wide anyway, even if it were not flawed by irrationality. The decision to make it falls outside the generous ambit within which reasonable disagreement is possible. It does not create a “best endeavours” obligation but an absolute obligation. It is not limited by any date range, making it uncertain what the respondents would accept as compliant. It is no answer that the tribunal would take a lenient view because the order is unclear or compliance difficult. The respondents’ interest would be to persuade the tribunal to take a strict view.

Finally, I reject the submission that the burden falls not on the claimants themselves but on their solicitors. That is not accurate. The burden falls on the claimants and it is the solicitors who discharge the burden on their clients’ behalf. The same is true of case management obligations falling on the respondents. Both sides benefit from having their own solicitors and from the other side having solicitors, sparing them the difficulties of dealing directly with the opposing parties. Economies of scale are achieved because there is only one letter to write each time an issue is raised.

For those reasons, the contested order cannot stand and I will set it aside and allow the appeal. As for the remedy, my powers are set out in section 35(1) of the Employment Tribunals Act 1996:

“35.— Powers of Appeal Tribunal.

(1)

For the purpose of disposing of an appeal, the Appeal Tribunal may—

(a)

exercise any of the powers of the body or officer from whom the appeal was brought, or

(b)

remit the case to that body or officer.”

The issue that arises from case law interpreting that provision is whether, in a successful appeal against a case management decision, this appeal tribunal may substitute its own case management decision, or whether it is bound by authority to remit the matter to the employment tribunal unless the parties consent to it not doing so or there is only one possible outcome. If I am able to exercise the power to make a substituted decision further issues arise: whether I should do so and if so in what terms.

These points were in part debated at the hearing and then became the subject of two rounds of supplemental written argument, at my invitation. For the claimants, Ms Jolly was inclined to treat the case management issues as issues of fact rather than evaluation of competing suggested orders:

“[i]n the absence of the employment tribunal having heard any application, submissions or evidence for or against the Appeal Tribunal’s potential alternative solutions to the order made by EJ Jenkins (save for that set out …. below) (Footnote: ⁵), it is submitted that the Employment Appeals Tribunal does not have the power to make those Orders pursuant to s.35 ETA 1996 since there are no proper factual findings, assessment or analysis, or indeed any proposed draft questions, upon which to base them”.

She cited Jafri v. Lincoln College [2015] QB 781, CA, in support that that position. The appeal tribunal had no properly evidenced application before it, as the judge below had not. Ms Jolly warned me not to descend into the arena, referring to a line of authorities emphasising the adversarial nature of the proceedings. The proceedings are not inquisitorial, so it is not for the appeal tribunal to propose its own solution, she submitted. The only option to which the claimants consented and the only option open to me apart from remission back, therefore, was to make the order sought by the claimants as an alternative to the judge’s contested order.

In further written submissions, Ms Jolly confirmed her position that the reasoning in Jafri applied to case management decisions and is not confined to substantive factual decisions and conclusions, such as that a person has been unfairly dismissed. A case management decision is either lawfully made or it is not; if it is not, “it is not for the EAT to make any factual assessment itself (which an evaluative case management decision such as this one in this case will usually, and does, involve): that would usurp the role of the employment tribunal”.

Ms Jolly pointed out that Jafri had proceeded from well established previous authority and its application to case management decisions had been confirmed by Elias LJ in Kuznetsov v. Royal Bank of Scotland plc [2017] IRLR 350, at [34]. She accepted that the appellate judges had emphasised their encouragement to parties to give their consent to the appeal tribunal substituting its own case management decision, but clarified that the claimants were only willing to consent to the alternative order, i.e. that the contested order’s terms should be narrowed so that it applied to the selected lead claimants only.

She submitted that the issue here was not binary; at least four possible substituted outcomes had been canvassed in oral argument and it was open to a party to give its consent on terms, i.e. provided that the appeal tribunal was willing to make the order that party wished it to make and not otherwise. The claimants would not consent to the appeal tribunal substituting its own case management decision without knowing in advance what that decision would be. All candidates except her preferred one had not been adequately aired in evidence or argument, Ms Jolly submitted.

For the respondents, Mr Leiper took issue with:

“the proposition that the Appeal Tribunal does have the power to consider the Claimants’ proposed alternative (particulars limited to the Lead Claimants) but not some middle ground (particulars limited to a statistically material number of randomly selected Lead Claimants). The material before the Tribunal and the Appeal Tribunal is identical for each. The Claimants did not advance their now proposed alternative at the original hearing, and only when they sought a variation to the orders originally made.”

On a remission, said Mr Leiper, the employment tribunal would be able to make a “middle ground” order of the kind discussed during the oral argument. There was no jurisdictional bar to the appeal tribunal making its own decision on the appropriate order; it had the power to do so under section 35(1)(a) of the 1996 Act, as Underhill LJ had pointed out in Jafri. He submitted that the reasoning and decision in Jafri did not extend to case management decisions. They are not an “industrial jury” question; the employment judge sits alone.

Mr Leiper submitted that case management decisions are about the exercise of discretion and do not demand deference to the “industrial jury”, as the Court of Appeal had noted in Burrell v. Micheldever Tyre Services Ltd [2014] ICR 935 (per Maurice Kay LJ at [17]). The Court of Appeal had earlier, in Canadian Imperial Bank of Commerce v. Beck [2009] IRLR 740, upheld the exercise of discretion by HHJ McMullen QC, sitting in the EAT and reversing on appeal a refusal by an employment judge to make an order for disclosure; see Wall LJ’s judgment of the court, at [42].

Mr Leiper submitted that Elias LJ’s observation in Kuznetsov at [34] that the reasoning in Jafri applied to case management decisions, was obiter because the parties’ account of the law was “undisputed” (per Elias LJ at [18] and because the ratio of the Court of Appeal’s dismissal of the appeal was that the employment judge would have made the same decision even if he had not erred in his approach to it (ibid. at [32]). Mr Leiper reminded me that whether an error of law had been committed by an employment judge in exercising a discretion was a different and prior question from whether remission must follow once an error of law has been detected.

Where an error of law is detected on appeal, said Mr Leiper, the appeal tribunal has the right to make necessary consequential orders without sending the case back. He relied on Waite J’s judgment in Medallion Holidays Ltd v Birch [1985] ICR 578). Waite J said at 581A-D:

“We have no right to interfere unless persuaded either that the tribunal’s verdict resulted from a self-direction erroneous in law or (if soundly based in law) that it represented a conclusion untenable by any reasonable tribunal. In cases where the appeal tribunal is so persuaded, and the appeal therefore has to be allowed, then the scheduled powers are available to give the appeal tribunal the necessary ancillary authority to make whatever orders or directions may be required for the future conduct of the claim, without the matter necessarily having to go back to the industrial tribunal for consequential directions. That restrictive interpretation of the scheduled powers—limiting them to a right in the appeal tribunal to step into the industrial tribunal’s shoes solely for the purpose of making such consequential directions or orders as may seem appropriate after their appellate function (properly so called) has been exercised—was approved by the Court of Appeal in O’Kelly’s case [1983] I.C.R. 728, 764, per Sir John Donaldson MR.”

Mr Leiper pointed out that Medallion Holidays was approved and followed by the EAT (Wood J) in Adams v West Sussex County Council [1990] IRLR 215 and by Lady Smith in Amey Services Ltd v Cardigan [2008] IRLR 279. By the same reasoning, he submitted, the appeal tribunal here need not send the case back to the tribunal but can decide for itself what the appropriate order should be in substitution for the contested order of the judge below, if it is set aside.

I come to my reasoning and conclusions. I have not found this part of the appeal easy. I will start with the more straightforward parts of the exercise. First, under section 35(1) of the 1996 Act, I “may” either remit the matter under (b) or, under (a), “exercise any of the powers of the body or officer from whom the appeal was brought”. I therefore do not agree with the claimants that the appeal tribunal “does not have the power” (in Ms Jolly’s words) to make an alternative order. That is not accurate. The submission should be that it would be a wrong use of my power to do so. Underhill LJ made the same point in Jafri at [47].

Second, I recognise on the authority of Jafri that being in as good a position as the tribunal below to make the decision in question does not justify exercising a decision, if it is a decision which it is the exclusive province of the tribunal below to make. That flows from the reasoning of Laws LJ that the tribunal is the industrial jury exclusively charged with finding the facts and deciding what conclusion flows from them, with an appeal to the appeal tribunal on a point of law only.

Third, I reject the proposition that I would be descending into the arena if I were to make any order other than one endorsed by the claimants. The arena is the underlying appeal, in which the claimants have succeeded (without me descending into it). The remedy issue is ancillary. I was careful to run past the parties in oral argument not just their own suggestions, but any I might have in mind if I were persuaded not to remit the case. I mentioned the analogy of “trial by sample” in the Technology and Construction Court, in that regard. I was careful to ensure the parties had every opportunity to address all the points I had in mind, orally and in writing.

Next, I do not accept that a case management decision is primarily a fact finding exercise. The facts are taken as those pleaded. The parties make submissions about what order is appropriate on the basis of the pleaded facts and the issues they raise. The judge then decides. It is an evaluative decision, to use the current terminology. The parties may proffer factual evidence about the burden of compliance; but assessing such evidence is not a forensic fact finding exercise comparable to assessing a witness’s credibility or analysing email exchanges during someone’s employment. Nor is the evaluative decision to make an order for particulars or disclosure comparable to a conclusion that on the facts found the claimant was, or was not, unfairly dismissed or discriminated against.

I do not accept, aside from authority, that substituting a different case management decision on appeal would usurp the role of the employment tribunal. I agree with Mr Leiper that the single tribunal judge making a case management decision is not sitting as an industrial jury in any real sense. The judge’s evaluative decision is subject to appeal on a point of law only; and if, as here, the decision is flawed in point of law, the powers of the appeal tribunal are those in section 35(1) of the 1996 Act. To exercise the first of those powers having set aside a case management decision does not, of itself and aside from authority, usurp the tribunal’s role, in my judgment.

I come next to the question of consent. If the parties consent to the appeal tribunal substituting its own decision, it is common ground that any difficulty about usurping the tribunal’s role falls away. This alone shows that the power to do so exists; if it did not, the parties could not confer it by consent. It is also common ground that the parties can be encouraged to consent and “in an appropriate case …. be strongly encouraged to do so” (per Underhill LJ in Jafri at [47]). But here, there is a difficulty: while the respondents’ consent is unconstrained, the claimants’ is only given on terms that the order made will be the one acceptable to them.

I do not think it is a principled position to offer consent on terms. A party should either consent or not; it should not seek to negotiate with the appeal tribunal. I am not comfortable with the argument that I will be misusing my power unless I exercise it to the claimants’ liking. Consent should either be unqualified or withheld (unless, perhaps, any envisaged qualification is also a matter of agreement between the parties). That means not knowing in advance what the outcome will be. I decline to decide between remission back and making the order the claimants seek; and, as Ms Jolly accepts, I cannot do so anyway because the respondents do not agree to that course.

There is therefore no consent to the appeal tribunal substituting its own decision. And this is certainly not a case with only one possible lawful outcome. Several possible lawful outcomes, short of the oppressive and wrongly reasoned contested order, have been discussed. It follows that I can only avoid remission back if it would be a proper exercise of my power under section 35(1)(a) of the 1996 Act to do so without consent and with several possible lawful outcomes. The claimants say that road is barred by binding Court of Appeal authority, but the respondents beg to differ.

In examining the authorities, it is worth saying that the weight of judicial opinion, including among judges of the Court of Appeal who have expressed views on the subject, is to deprecate the application of the Jafri analysis to case management decisions. Hence the attempts to find ways round it by taking a robust view on the “only possible outcome” issue or putting pressure on the parties to consent to a substituted decision. There is broad judicial agreement that the Jafri dogma undermines the overriding objective by causing delays and increasing costs.

I am duty bound both to further the overriding objective and to follow binding Court of Appeal authority. I have found it difficult to do both at once but, doing the best I can, I turn to the authorities. The only way I can think of that reconciles these two potentially conflicting duties is to accept the submission of Mr Leiper that I am not bound by Jafri, or Kuznetsov, or any other binding authority, to apply Laws LJ’s reasoning in Jafri to a case management decision of the kind at issue in this appeal.

I do so with diffidence because Underhill LJ said in Jafri at [47] that it is “fact-finding” that is the exclusive territory of the employment tribunal and he added, albeit in parentheses, that fact finding “includes fact-assessing and discretion-exercising”. That echoes Ms Jolly’s submission in this case but I have not accepted that submission because, as I have explained, I believe the facts are those taken from the pleadings. Those facts are not found to be true or untrue by the tribunal; the tribunal evaluates them as allegations and crafts an appropriate case management order accordingly.

My diffidence is redoubled when I look at what Elias LJ said in Kuznetsov at [34]. It was, as Mr Leiper submitted, obiter because Elias LJ had already decided at [33] to dismiss the appeal on the ground that the employment judge’s error had been immaterial. But he did say that only legislation or the Supreme Court could change the law that “remission is currently required absent the consent of both parties”; and he said it in the context of an appeal which was about whether to allow amendments to the pleaded case, i.e. par excellence a case management decision.

Despite the Olympian stature of those observations, I am persuaded by Mr Leiper that neither is strictly binding authority requiring me to remit the matter here. There is enough traction in the cases he cites – Canadian Imperial Bank of Commerce v. Beck, Medallion Holidays Ltd v. Birch, Adams v West Sussex County Council and Amey Services Ltd v Cardigan – to provide a foundation for a proper exercise of my section 35(1)(a) power to determine the appropriate order myself. I reach that conclusion with relief, for it avoids having to undermine the overriding objective.

I therefore decide, albeit with trepidation, that I would not necessarily be misusing my power under section 35(1)(a) of the 1996 Act if I decline to remit the matter back and decide for myself what is the appropriate order in substitution for the contested order. There are sound reasons here for not referring the matter back. The obvious ones are that to do so would build in further cost and delay, as recognised in the authorities. More specifically, this is multi-handed litigation managed by a single assigned employment judge, to whom the matter would normally be referred back if it went back at all. But I have already explained why I regard him as parti pris on the issues in this appeal.

For that reason if I referred the matter back, it would be to a different employment judge, who would not have the knowledge and experience of managing the case comparable to that of the managing employment judge. I do not have deep knowledge and experience of the history of the litigation to compare with that of the assigned judge, but I have become familiar with the issues in this appeal and do feel qualified to make a sensible and proportionate decision on the matter. To do so would help to preserve the viability of the preliminary issues hearing dates in June and July 2026. Not to do so could put those dates at risk, when it may be years yet until liability is determined.

I think the right order to make is as follows. I will require the pool of 12 potential lead claimants, from which the final eight will be selected, to comply in full with the requirements set out in the contested order. In addition, I will require 125 randomly selected claimants (which may include members of the pool of 12) to answer a questionnaire verified by a statement of truth in the same form as used in civil proceedings under the CPR, i.e. as set out in quotation marks in PD 22, paragraph 2.1, and signed by each of the 125 sample claimants personally.

The particulars and questions will both relate to the period starting two years before the date of the claim and ending on the date the questionnaire is submitted by each claimant to the claimants’ solicitors. In the questionnaire, the sample claimants will be asked, in respect of that period, to give their best estimate, having made reasonable enquiries and considered any physical or electronic records in their possession, of the dates and times that they were logged onto any other apps for the purposes of driving, delivery or related private hire or courier services; and of the dates on which they made themselves available to other businesses and / or clients for the purposes of driving, delivery or related private hire, chauffeur or courier services.

I will give the parties a short period from the date this judgment is handed down to agree the timetable for that exercise, constructed so as to preserve the viability of the 2026 preliminary issues hearing dates, and other consequential matters. The timetable for the nomination process for selecting lead claimants will need to be revisited and should dovetail with the questionnaire process and provision of particulars by the 12 potential lead claimants. To the extent that matters are not agreed, I will decide any outstanding disagreements on the papers before finalising the appeal tribunal’s order.