BAILII [Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]

First-tier Tribunal (Tax)

You are here: BAILII >> Databases >> First-tier Tribunal (Tax) >> Reed Employment Plc & Ors v Revenue & Customs [2012] UKFTT 28 (TC) (06 January 2012)
Cite as: [2012] SFTD 394, [2012] STI 550, [2012] UKFTT 28 (TC)

[New search] [Context] [View without highlighting] [Printable PDF version] [Help]

Reed Employment Plc & Ors v Revenue & Customs [2012] UKFTT 28 (TC) (06 January 2012)
Employment income

[2012] UKFTT 28 (TC)


Appeal numbers: SC/3128-3135/2007




INCOME TAX —and NATIONAL INSURANCE CONTRIBUTIONS — dispensations within ITEPA s 65 — allowances for travel and subsistence costs — character of allowances — found to be Chapter 1 earnings — whether recipients had permanent or temporary workplaces — permanent — travel costs ordinary commuting expenses and not deductible — whether allowances within dispensations — no — whether dispensations could lawfully be granted — no — appeal substantially dismissed
















- and -





Tribunal: Judge Colin Bishopp

Judge John Avery Jones CBE


Sitting in public in London on 14 to 18, 21 to 25, 28, 29 March, 1, 4 to 8 April 2011


Mr Andrew Clarke QC, Mr David Ewart QC and Mr Richard Vallatt, counsel, instructed by Slaughter and May, for the Appellants

Mr Malcolm Gammie QC, Mr Adam Tolley, Miss Abra Bompas and Miss Kate Balmer, counsel, instructed by the General Counsel and Solicitor to HM Revenue and Customs, for the Respondents





Glossary  5

Introduction  7

Background  7

The appeals  9

The facts  10

Reed’s business  11

The 1998 legislative changes  15

Reed and Robson Rhodes  17

The first dispensation  19

The second dispensation  26

The third dispensation  27

The fourth dispensation  31

The fifth dispensation  31

The mechanics of the RTA and RTB schemes  32

The revocation of the fifth dispensation  45

The replacement scheme  49

Reed’s contracts with its employed temps  50

HMRC’s understanding of the operation of the schemes  54

The relevant legislation  55

The issues  60

Issue 1: did the employed temps make an effective salary sacrifice?  63

Reed’s submissions  63

HMRC’s submissions  64

Discussion  67

Issue 2: were the disputed allowances within Chapter 1 or Chapter 3?  72

Reed’s submissions  73

HMRC’s submissions  74

Discussion  77

Issue 3: were the workplaces temporary or permanent?  79

Reed’s submissions  79

HMRC’s submissions  82

Discussion  84

Issue 4: could the inspector lawfully grant the dispensations?  87

Discussion  88

Issue 5: did the dispensations cover the allowances?  89

Issue 6: what is the effect of a dispensation?  89

Issue 7: Reed’s legitimate expectation  90

Issue 8: was reed obliged to make paye deductions in the relevant period?  92

Issue 9: is the outcome for NICs the same as for tax?  92

Conclusions  92





Austin, Jacqueline



The HMRC officer who led the Employer Compliance Review of Reed from 2004 onwards.

Baddeley, Simon

Senior Regional Director of Reed Specialist Limited, a Reed Group company, though not one of the appellants.

Baillie, John

The Contributions Agency inspector who led the Contributions Agency review of Reed from 1998 to 2001.

Beal, Derek

An executive director of all the appellants (save for Reed Health) as well as of all other Reed companies. He became a non-executive director of Reed Health in November 2004.

Brook, David

The HM Inspector of Taxes who granted the Fifth Dispensation.

Chapman, Sylvia

An HMRC officer with particular expertise in the employment status of workers.

Downes, Tim

The HM Inspector of Taxes who negotiated and granted the First Dispensation in 1998.

Hird, Douglas

An HMRC officer part of whose role was to deal with PAYE enquiries.

Kirkham, Diane

An HMRC officer who was involved in the negotiation and grant of all of the dispensations.

Ollerenshaw, Susan

Joined Robson Rhodes as an employment tax adviser in January 1999. After Mr Rayer left Robson Rhodes, Miss Ollerenshaw was Reed’s main point of contact in relation to the Dispensations, RTA and RTB.

Rayer, John

A tax partner at Robson Rhodes. He was Reed’s main point of contact with Robson Rhodes from 1992 until 2001.

Read, Nick

The HM Inspector of Taxes who granted the Second to Fourth Dispensations and was involved in the grant of the Fifth Dispensation.

Robson Rhodes

Reed’s external accountants throughout the relevant period (usually abbreviated within this decision to RR. The firm later became RSM Robson Rhodes and then RSM Robson Rhodes LLP).

Sims, Pat

An employee of RR who assisted Mr Rayer and Miss Ollerenshaw.



Agency worker

A self-employed worker engaged by an employment business under a contract for services to work on assignment for the employment business’s clients on a temporary basis: see para 3.


Reed Travel Allowance and Reed Travel Benefit. The sums for the time being set out in the current dispensation.


First Dispensation granted on 30 November 1998;

Second Dispensation granted on 9 January 2001;

Third Dispensation granted on 7 February 2002;

Fourth Dispensation granted on 7 March 2003; and

Fifth Dispensation granted on 3 February 2004.

Employed temp

An employee of Reed placed on assignment with a client: see para 5.

Employment business

The activity of Reed with which we are concerned: see para 2.


Used as shorthand term to include not only Her Majesty’s Commissioners for Revenue and Customs but also their predecessor bodies, the Inland Revenue, HM Customs & Excise and the Contributions Agency.


Income Tax (Earnings and Pensions) Act 2003.


National Insurance contributions.


Profit-related pay: see para 29.


The appellants collectively.

Reed Health

Reed Health Limited, the third appellant, to which slightly different arrangements applied for part of the relevant period.

Relevant period

6 January 2001 to 5 April 2006.


Robson Rhodes.


Reed Travel Allowance, the scheme introduced in 1998.


Reed Travel Benefit, the amended scheme which succeeded RTA in 2002.


The RTA and RTB schemes.


Social Security Contributions and Benefits Act 1992.


Social Security (Contributions) Regulations 2001.

Temporary employee

Umbrella term to include agency workers and employed temps: see para 3.

Travel-to-work payment

The supplement to temporary employees’ pay Reed paid (subject to the deduction of tax and NICs) while the RTA scheme was in operation.




1.           The appellants are all members of the Reed group of companies, carrying on business as what are generally, though not altogether accurately, known as employment agencies. For part of the period with which we are concerned (6 January 2001 to 5 April 2006, the “relevant period”) one of the appellants, Reed Health Limited (“Reed Health”), was not a member of the group, but the various arrangements we shall describe were adopted by it as well as the other companies in the group, and with limited exceptions we do not need to treat it or any of the other Reed companies separately. Unless there is a reason to distinguish between one of the appellants and the remainder, we shall use the word “Reed” as a term to encompass all the appellants as a group, as well as each individual company within that group.

2.           Throughout the relevant period Reed acted as both an employment agency, properly so called, and as an employment business. It is common ground that the legislation regulating employment bureaux (a portmanteau term to cover both) draws a distinction between “employment agencies”, which place permanent staff, that is workers who become employees of the agency’s client, and thereafter have no continuing relationship with the agency, and “employment businesses”, which provide “temporary employees”, that is workers who undertake work for the agency’s clients, usually but not always for a fairly short period, but who, despite the use of the term “ temporary employees”, do not become employees of the client. It is also common ground that employment agencies and employment businesses may, depending on the contractual arrangements into which they enter, act either as principals or as agents. We are concerned in this appeal only with Reed’s activities as an employment business.

3.           A temporary employee may be either a self-employed person or an employee of the employment business. Such self-employed persons are generally referred to as “agency workers”, the term we shall use in this decision. Reed acted as agent for its agency workers, and as principal for its employed temps (for the meaning of this expression see para 5 below). Although they do not have contracts of service, agency workers are nevertheless treated for tax, and in particular PAYE, purposes as if they were employees of the agency through which they obtain engagements. This treatment is dictated by s 44 of the Income Tax (Earnings and Pensions) Act 2003 (“ITEPA”), re-enacting without significant amendment earlier provisions of the Income and Corporation Taxes Act 1988 (“ICTA”) to the same effect. We add for completeness that it seems some workers offered their services to Reed through the medium of a limited company, but we are not concerned with such workers in this appeal and leave them out of account in what follows.

4.           As the parties agree that there is no material difference between the ITEPA provisions and those which preceded them, in this and in other contexts, we shall refer only to ITEPA in this decision, unless there is a particular reason to do otherwise. There are similar provisions in respect of National Insurance contributions (“NICs”) in the Social Security (Contributions) Regulations 2001 (SI 2001/1004) (“the SSCR”), which replaced their predecessors, the Social Security (Contributions) Regulations 1979 (SI 1979/591) from 6 April 2001. Again, as the SSCR merely repeat the earlier provisions without amending them we shall refer only to those regulations, unless it is necessary to do otherwise.

5.           Since 1995, the majority of Reed’s temporary employees have been employed by Reed and, apart from Reed Health, Reed has not engaged agency workers at all since 2001. Reed Health ceased to engage agency workers in 2004. Those workers who were Reed’s employees, both before and after 2001, are referred to in this decision as “employed temps”. The original impetus for Reed’s employing its temporary employees was that employees could, while agency workers could not, benefit from the tax provisions relating to profit-related pay (“PRP”). Reed saw its PRP scheme as an attractive incentive and wanted to extend it to as many of its workers as possible. Reed did, of course, have its own permanent staff as well, but with very limited exceptions, which can be ignored for present purposes, the arrangements with which we are concerned did not apply to them at any time.

6.           During the relevant period Reed made (or said that it made) certain payments to its employed temps, described as “allowances”. The arrangements for their payment evolved over time, but there were two basic schemes. The earlier, introduced in 1998, following changes to the treatment of reimbursements of travelling expenses paid to persons whose earnings fell within what was then Schedule E, was Reed Travel Allowance (“RTA”), which was replaced in 2002 by Reed Travel Benefit (“RTB”). Each was designed, Reed says, to reimburse employed temps the cost of travel to their places of work (at Reed’s clients’ premises) and to provide a subsistence payment. Before the allowances were introduced, and shortly before each change in the detail of the arrangements, Reed sought and obtained a dispensation from HMRC in accordance with what is now s 65 of ITEPA. There were five dispensations in all. Reed took the view that the arrangements that it had implemented, coupled with the dispensations, meant that the allowances did not come within the PAYE scheme, and that they were also exempt from NICs. Acting on that view, it paid the allowances gross during the relevant period.

7.           By 2004 HMRC were beginning to have misgivings about the allowances, and after a series of meetings and extensive correspondence the fifth dispensation was revoked with effect from 6 April 2006 (each dispensation revoked its predecessor, thus only the fifth dispensation was extant at this time). Reed thereafter adopted a rather different arrangement. We shall need to describe that arrangement later (see para 169 below), as some reliance was placed on its features by way of contrast to those of the RTA and RTB schemes.

8.           In February 2007 HMRC made a large number of determinations in accordance with reg 80 of the Income Tax (PAYE) Regulations 2003 (SI 2003/2682) (“the PAYE Regulations”) and notices of decision in accordance with reg 67 of the SSCR, both covering the relevant period, that is 6 January 2001 to 5 April 2006. Their combined purpose was to assess both the employer’s NICs for which HMRC say Reed should have accounted, and the sums which HMRC say Reed should have deducted (as tax or NICs) from employed temps’ salaries and paid over to them, during the relevant period. The validity of the determinations and decisions is dependent (putting a series of issues, set out fully at para 189 below, in nutshell form) on the correctness of HMRC’s view that the payments Reed made did not amount to reimbursement of deductible expenses, but were payments of earnings, and thus were not covered by the terms of the dispensations. The overall amount in dispute, including interest, is in the order of £158 million. We were told that about 500,000 workers, in all, had participated in the schemes during the relevant period; it is probably legally and certainly practically impossible for Reed to recover any of the disputed amount from them should it lose these appeals.

9.           Reed’s case, in essence, is that the employed temps made an effective salary sacrifice in exchange for the allowances, the consequence of which was that the allowances were not, and were not to be treated as, earnings; that the allowances (which were contractually due to the employed temps if and only if travelling and subsistence expenses had been incurred by them) properly reflected the employed temps’ tax-deductible expenses; and that they were correctly paid gross. HMRC’s response, in very brief summary, is that Reed’s contracts with its employed temps did not provide for an effective salary sacrifice, and there was no such sacrifice as a matter of fact; that the allowances did not represent the reimbursement of expenses, but were simply part of the employed temps’ salary; that even if they were paid in respect of expenses those expenses were not deductible; and that Reed cannot rely on the dispensations as a means of avoiding its liability.

10.        There are subsidiary disagreements between the parties about the amounts of tax and NICs which are due, the extent of the liability of each of the Reed companies (it may be that tax and NICs due from one company have been assessed against another), and whether the NICs decision is out of time in relation to January 2001, but we are not required, at least at this stage, to determine any of those issues. Although one of the matters which may ultimately need to be decided is whether the outcome for tax purposes and the outcome for NICs purposes are identical or differ, and if they differ in what respect or respects, the arguments we heard focused on the tax consequences of the arrangements, and we shall adopt the same approach in this decision.

The appeals

11.        Reed accepts that it is able to appeal to this tribunal only against the determinations and decisions described at para 8 above. Each of the appellants has made such an appeal (“the tax appeals”) and, as they all raise the same issues, it was directed some time ago that they be heard together. The appellants have, however, also brought judicial review proceedings on the ground that, they say, HMRC’s actions breached their substantive legitimate expectation, based on the dispensations, that tax and NICs would not be due in respect of the allowances. In July 2009 the judicial review proceedings were transferred from the Administrative Court to the Tax and Chancery Chamber of the Upper Tribunal (“UT”) so that they could be case managed jointly with the tax appeals, and in November 2009 the UT and this tribunal, sitting simultaneously, gave directions for the judicial review proceedings and the tax appeals to progress in tandem with a view to a concurrent or consecutive hearing. However, in November 2010, the UT gave new directions that the judicial review proceedings be stayed until after the determination of the tax appeals.

12.        One of the reasons given for adopting this course was the possibility that this tribunal after all has the jurisdiction to determine the legitimate expectation issue itself, as Sales J indicated might be the case in Oxfam v Revenue and Customs Commissioners [2010] STC 686. HMRC maintain that the First-tier Tribunal does not have the jurisdiction to determine issues of that kind, which they say are the exclusive province of the UT and the High Court. Reed advanced no positive case about the extent of this tribunal’s jurisdiction, but it nevertheless put the arguments in favour of the proposition that it does have jurisdiction in order to assist us in coming to a conclusion on the point.

13.        However, the scope of this tribunal’s jurisdiction is now a matter which is before the UT in other appeals. We do not, in those circumstances, think it desirable for us to address the jurisdiction argument; we will instead confine ourselves to making findings of fact which will, we hope, be of assistance in whatever is ultimately determined to be the proper forum.

14.        An incidental consequence of the parties’ disagreement about the extent of our jurisdiction was that they were also not entirely in agreement about the extent to which we should find facts, and about whether such findings of fact as we might make should be binding on the UT should it be required to determine the legitimate expectation issue. The conclusion we have come to in that respect is that we should make all of the findings of fact which the evidence allows us to make, whatever may be the extent of our jurisdiction, and whether or not the findings are necessary for the conclusions we reach. The UT can then decide for itself to what extent it is willing to adopt our findings, and to what extent, if at all, it is willing to make findings of its own.

15.        Before us, Reed was represented by David Ewart QC, Andrew Clarke QC and Richard Vallatt of counsel, and HMRC by Malcolm Gammie QC, Adam Tolley, Abra Bompas and Kate Balmer, all of counsel.

The facts

16.        We take the following facts from the extensive documentary evidence provided to us, supplemented by the statements (which stood as their evidence in chief) and oral evidence of several witnesses, whose positions and roles are described in the dramatis personæ which appears at the beginning of this decision; some of the persons whose names appear there did not give evidence, although they played a part in the relevant events. Those who gave oral evidence, in the order in which they appeared, were Simon Baddeley, John Rayer, Susan Ollerenshaw, Derek Beal, Tim Downes, Nick Read, David Brook, Douglas Hird, Diane Kirkham, John Baillie, Sylvia Chapman and Jacqueline Austin.

17.        Many of the facts were not in issue as such; it was their significance, or the interpretation to be placed on them, which led to contention. In the interests of brevity we have omitted a good deal about which we heard evidence but which does not now seem to us to be of great importance, and for the same reason we have not set out much of the evidence extensively in the narrative which follows, since we have not thought it necessary to identify seriatim the source of many of the facts we have found. On some issues, however, a good deal of detail is required. We should perhaps also add that, with some exceptions, Miss Ollerenshaw, Mr Baddeley and Mr Beal had good recollections of the relevant events, though Mr Rayer less so, while the HMRC officers who gave evidence recalled some matters with relative ease but in other respects could do little more than recite what was in contemporaneous notes.

18.        There are two distinct chronological threads with which we must deal and, even though the two are inter-related, we think it will make this decision easier to read and understand if we deal with the background to the relevant events before proceeding to the grant of the various dispensations, and then the evolution of the RTA and RTB schemes, before bringing those threads together in dealing with the circumstances in which the fifth of the dispensations was revoked. We need, however, to begin with the evidence we heard about Reed’s business—which was largely uncontroversial—and then to deal with the reasoning behind the introduction of the schemes.

Reed’s business

19.        Mr Beal is a chartered accountant and a director of Reed’s holding company, Reed Executive Limited (which is not one of the appellants), a position he also occupied during the relevant period. He has held various senior positions within the group since 1989, and was closely involved in the introduction and implementation of the arrangements in issue in these appeals. He is therefore able to speak with some authority both about Reed and its business, and the arrangements themselves. Although Reed Health was separated from the remaining Reed companies in 2001 (for reasons immaterial to the appeals), and run as a distinct business, it retained some connections with the Reed group (for example sharing some accounting functions). It rejoined the group in 2004, when Mr Beal became a non-executive director. He was, therefore, also able to give some evidence about the relatively minor differences between Reed Health and the other appellants.

20.        Mr Beal told us that the business was founded in 1960 and floated in 1971, but it became a private company again in 2003. It is a substantial concern, with over 280 trading branches located throughout the country. In any given week it might have between eight and sixteen thousand people on its books; but it has, and in the relevant period had, a high turnover of employed temps and, while they were still engaged, agency workers, and in a typical year it might engage as many as 100,000 such workers.

21.        Most of the evidence we heard about the manner in which Reed conducts, and conducted, its day-to-day business of placing temporary employees with its clients came from Mr Baddeley, who has worked for Reed since January 1999. He began as an administrator in a branch, being promoted to “Temps Consultant”, whose role was to supply suitable temporary employees to Reed’s clients, after six months, progressing thereafter to Account Coordinator for a major client of the same branch. He then became a “Perms Consultant”, placing permanent employees with Reed’s clients, after which he underwent steady promotion, becoming successively the branch’s Business Manager, Senior Business Manager, Area Manager, Regional Manager, Senior Regional Manager, Regional Director and finally, in April 2010, Senior Regional Director. We recognise that Mr Baddeley is able to speak from extensive personal knowledge about the manner in which Reed conducts its core business, and accept his evidence about it (which was in any event largely unchallenged).

22.        Reed operated, he said, in a fiercely competitive market. Its policy was to endeavour to fill a vacancy for a temporary employee within no more than half an hour of the client’s request, and ideally while the client was still on the telephone to the temps consultant. The reason was that if Reed could not fill the vacancy within 30 minutes, one of its competitors would; if Reed could place a temporary employee before the end of the client’s phone call, thus eliminating the risk of the business going to a competitor, this was better still. For this reason the 30-minute policy applied regardless of how much notice of the forthcoming vacancy was given. Typically, he said, clients provided one to two weeks’ notice, although this could extend to three weeks or more in the case of business-critical positions, or reduce to a matter of an hour or so if a temporary employee was required to cover someone who had, for example, called in sick. Temporary employees were commonly engaged by a client for a week or two, to provide holiday cover, but they might sometimes remain at a client’s premises for as little as a day, or sometimes for months and occasionally even years.

23.        If the policy of filling vacancies very quickly was to succeed, the temps consultants needed up to date information regarding which of the temporary employees on Reed’s books were available for work at any time, without having to ring them to find out. He had, therefore, to keep in close touch with all the temporary employees who were not on assignment, or whose assignments were coming to an end, and ensure that they were available to fill vacancies as they arose. The temps consultant also had to be able to “sell” the temporary employee he had in mind to the client, but had to take care in doing so since, if the temporary employee was not of the quality expected by the client, Reed would risk losing that client’s future business to a competitor. Similarly, if the position offered did not suit the temporary employee, he (or more often she) would probably not accept the position, and might turn to one of Reed’s competitors to find more suitable vacancies in the future. From time to time, therefore, Reed found itself unable to fill vacancies. It was correspondingly in Reed’s interests to attract a large number of temporary employees, with differing skills and experience, in order to reduce the number of occasions on which this occurred.

24.        A long-standing feature of Reed’s business model, by contrast to most of its competitors, was to site its branches in prime high-street locations at street level, with windows filled with cards advertising positions, and other marketing materials. Many of Reed’s branches opened on Saturdays, to attract potential temporary employees as they went out shopping. It also offered incentives to existing temporary employees who were able to recruit friends and acquaintances.

25.        Despite its need for a large pool of temporary employees, Mr Baddeley explained, it had a rigorous selection policy, by which a candidate’s skills, experience and aptitude were assessed. It did not benefit Reed (or the temporary employee) if Reed had too many of a particular type of worker registered at a branch, since Reed would not be able to find sufficient work and would find itself with disgruntled workers; and it wanted to engage only good quality workers, so as not to disappoint its clients.

26.        When a temps consultant decided to register a candidate, he was required to make various checks (such as verifying the candidate’s identity and taking up references) and to provide the candidate with certain information. There were centrally-produced Reed Handbooks which, until the late 1990s, were supplemented by information packs produced by each branch. Together, they included, in particular, a copy of the relevant terms and conditions of employment; details of how to record time spent on a client’s work on the timesheets provided for the purpose; details of the process for submitting timesheets to Reed and receiving payment; information about claiming statutory sick pay, holiday pay and similar entitlements; and branch-specific information. In about 2001 Reed provided more comprehensive handbooks for temporary employees, which largely eliminated the need for the branches to provide their own information packs. The handbooks contained detailed information about the RTA and, later, RTB schemes; we shall describe that information later in this decision. Both before and after the introduction of the new handbooks, the temps consultants were expected to talk to candidates about the range of benefits available to them (including those available through participation in the schemes), and to explain the eligibility requirements and similar matters.

27.        It was apparent from Mr Baddeley’s evidence that the temps consultants went to considerable trouble to keep in touch with temporary employees, not only in the intervals between engagements when the temp consultants needed to keep themselves informed of the temporary employees’ availability, but while they were on assignment with a client, essentially in order to maintain a close relationship with the temporary employees, to ensure that they were content and that they did not defect to a rival agency. Reed’s temporary employees frequently signed up with more than one employment business at the same time, as a means of ensuring that they were fully employed. It was correspondingly in Reed’s interests to offer as many assignments as possible to its temporary employees, to keep them loyal and to limit defections.

28.        We accept from Mr Baddeley’s evidence, and from what we later heard from Mr Beal, that Reed was keen that its temporary employees should believe that the salary and benefits they received represented competitive remuneration, and that it went to considerable lengths to attract and retain good quality staff.

29.        Mr Beal’s evidence was that until the early 1990s Reed (in common with most, but not all, similar businesses) engaged temporary workers by means of contracts for services, with the result that the workers were regarded as self-employed (and thus agency workers), although as we have said they were treated for PAYE purposes as if they were employees. In 1992 Reed introduced a PRP scheme which it made available to all its staff—permanent employees, employed temps and agency workers. It did so on the strength of Mr Rayer’s view that the same provisions which brought agency workers within an agency’s PAYE scheme also brought them within the ambit of any PRP scheme the agency operated. In February 1995, however, HMRC informed Reed that agency workers were not entitled to participate in PRP schemes, a view which Reed accepted, though evidently with considerable reluctance. It also viewed what it considered to be a change of mind on this point by HMRC with some apprehension, because it feared a retrospective recovery of tax, which in the event did not materialise. Mr Beal said he thought HMRC had acted very properly at this time by implicitly recognising that, whether or not Reed’s agency workers had been entitled as a matter of law to participate in a PRP scheme, Reed had nevertheless operated its scheme in good faith, and after taking and following professional advice. He was evidently somewhat aggrieved that, as he saw it, HMRC were taking a very different approach to the RTA and RTB schemes.

30.        Reed decided, in consequence of its inability to continue to include agency workers in its PRP scheme, to change its contracts so that those who had hitherto been agency workers should become its employees. This step was taken with some trepidation since there were, Mr Beal told us, several disadvantages to Reed in the change, which he described as a reduction in flexibility. On balance, however, Reed decided that it was in its interests to take on its agency workers as employees. About 90% of the agency workers (essentially those working for Reed Staffing Services Limited, then the main supplier within the group of temporary workers) entered into the new contracts, and continued to participate in the PRP scheme. The rest (engaged mainly by Reed Agency Services Limited, whose clients were principally in the financial services sector) remained as agency workers, however, in order that the supplying company could charge VAT to its clients on its commission only, rather than on its entire charge. This was the outcome of litigation in which one of the group companies had been involved (reported as Customs and Excise Commissioners v Reed Personnel Services Ltd [1995] STC 588); the VAT treatment Reed secured in that litigation was important to those of its clients which could not recover VAT.

31.        We interpose, since it is relevant to findings we shall make later, that in our view the disadvantages Reed perceived were less a reduction in flexibility than that employed temps had greater employment rights than agency workers. The evidence showed that it was concerned that the change in their status would expose Reed to claims for unfair dismissal. Indeed, Mr Beal’s witness statement shows that, despite its changing the status of most of its agency workers to employees in order that they might participate in its PRP arrangements, Reed attempted to exclude or restrict any other benefits the workers might have gained from the change. The contracts offered to employed temps differed in many respects from those by which permanent staff were engaged. Of particular importance is the inclusion within the new contracts of a provision that the employment would come to an end at the same time as the worker’s current assignment, the purpose of which, as Mr Beal accepted, was to give Reed at least the opportunity of arguing that it was under no continuing obligation to the employed temp once the assignment had ended. However, despite Reed’s attempts to protect itself as much as possible it recognised, Mr Beal said, that the trend in employment law was such that the employed temps would probably gain employment rights whatever the contracts provided.

The 1998 legislative changes

32.        The PRP scheme continued after the 1995 intervention by HMRC, but now only employed temps were able to participate. Nevertheless, Mr Beal still took the view that, as they were undertaking essentially the same role as employed temps, it was unfair that agency workers could not benefit from the PRP scheme. He tried to find a way of persuading HMRC to allow them to participate, perhaps in a modified scheme. However, HMRC were not to be persuaded and when, for reasons to which we are about to come, it became clear that the PRP scheme would need to be phased out, Mr Beal abandoned the attempt. His efforts thereafter were devoted to the introduction and implementation of the travel and subsistence allowances schemes with which we are concerned. These too he was keen to extend to agency workers but, as he was to learn, HMRC continued to distinguish between the two types of temporary employees and neither of the schemes ever became available to agency workers.

33.        In 1998, as we have mentioned, the tax treatment of travel allowances paid to employees taxed in accordance with Schedule E changed. The forthcoming changes and the phased abolition of PRP were both announced in the 1996 Budget Statement and implemented by the Finance Act 1998. As nothing turns on the phasing out of PRP we shall not deal with it further, save to observe that Reed’s PRP scheme overlapped with the RTA scheme for a short period.

34.        Formerly, the rules relating to the deductibility for tax purposes of travelling expenses were very strict—in essence, only those expenses incurred in actually performing the duties of employment were deductible—and it was the perception that they were excessively strict in relation to the travelling expenses of those workers whose employment required them to work at different places (particularly those working in the construction industry who were expected to attend different sites in the course of their work) which led to the relaxation which in turn prompted the introduction of the allowances. We deal more fully with the law relevant to this issue below, starting at para 177, but a brief summary of the effect of the changes (which is not in dispute) is necessary at this stage in order that what follows may be more easily understood.

35.        ITEPA charges income tax on “earnings”. Certain payments received by employees, which would not otherwise be taxable, are brought into charge as a result of the provisions of the “benefits code” which is defined by ITEPA s 63(1). The code applies to payments in respect of expenses falling within s 72; the aim is to prevent the avoidance of tax by the description of sums as the reimbursement of expenses when they are in truth earnings. The amount of the tax charge will, however, be reduced by any matching allowable deduction so that, first, the reimbursement of wholly deductible expenses will not give rise to any tax charge and, second, a reimbursement that exceeds the deductible expenditure will give rise to a charge only on the excess: see s 72(2).

36.        Sections 338 and 339 of ITEPA provide that a deduction from earnings is allowed for travel expenses incurred by an employee if certain conditions are satisfied. One such condition is that they are not expenses of travel between the employee’s home and a “permanent workplace.” A permanent workplace is defined by s 339(2)(a) and (b) as a place which the employee regularly attends in the performance of the duties of the employment and which is not a temporary workplace. Section 339(5)(a)(ii) adds that a place is not a temporary workplace if (inter alia) the employee’s attendance is in the course of a period of continuous work at that place comprising all or almost all of the period for which the employee is likely to hold the employment.

37.        The effect of these provisions, shortly put, is to allow relief for travel expenses incurred by peripatetic workers, but to exclude ordinary commuting expenses. The essence of the dispute between the parties, in this context, is that Reed says that the employed temps are peripatetic workers being sent to work for many different clients, while HMRC say that they have a separate employment for the duration of the assignment to each client with the result that each client’s premises count as a permanent workplace for the duration of the assignment and the travel expenses are accordingly ordinary commuting expenses.

38.        Class 1 NICs are payable in respect of the earnings of “employed earners” who are defined by s 2(1)(a) of the Social Security Contributions and Benefits Act 1992 (“SSCBA”) as, broadly, employees and office holders. Primary NICs are due (normally by deduction under the PAYE system) from the worker’s earnings (SSCBA s 6). The detailed rules about the rates of contributions and the earnings limits have changed over time, but the changes are not material for present purposes. It is sufficient to mention that primary Class 1 NICs are charged as a percentage of that part of the employee’s earnings which fall within upper and lower limits.

39.        Secondary NICs are due from the employer who pays the employee (SSCBA s 7). They are payable on all of the earnings of the employed earner which exceed what the legislation describes as the “secondary earnings threshold”. Where earnings are paid to an employed earner and there is in consequence a liability for primary and secondary NICs, the secondary contributor is liable to account not only for his own secondary contributions but also for the earner’s primary contributions (see para 3(1) of Sch 1 to SSCBA). As in the case of income tax deducted in accordance with the PAYE scheme, the employer deducts the primary NICs from the earnings before they are paid to the employee; but para 3 of Part VIII of Sch 3 to the SSCR reflects the PAYE provisions by prescribing that payments of, or contributions towards, travelling expenses which qualify under ITEPA ss 338 and 339 are to be disregarded in computing earnings for NICs purposes.

40.        It is convenient at this point to outline the provisions of ITEPA s 65, which is set out in full at para 181 below. It enables HMRC to grant a “dispensation”, a notice issued by an officer of HMRC stating that the officer agrees that, in respect of certain stated payments, benefits or facilities (including travel expenses falling within ITEPA ss 338 to 339), no additional tax is payable by virtue of the relevant provisions of the benefits code. When a dispensation has been granted, the provisions in the benefits code do not have the effect of imposing any additional liability to tax in respect of the relevant payments, benefits or facilities: the benefits code is to that extent “disapplied” and the employer is accordingly not obliged to deduct tax from any payment under the PAYE system. In essence, a dispensation is a means by which an employer can obtain advance clearance for the payment of tax-free expenses to his employees, so long as the expenses would otherwise be taxable under the benefits code; it has no relevance to earnings or to expenses taxable as ordinary earnings. The particular advantage to both employer and employee of a s 65 dispensation is that if the reimbursement of the expense is matched by an allowable deduction, it will also be outside the scope of PAYE—in other words, the employer may disregard such amounts when operating the PAYE scheme.

41.        There is no specific statutory basis for HMRC to issue dispensations in relation to Class 1 NICs. However, employment earnings on which Class 1 NICs are payable are, in general, calculated on the same basis as those earnings are calculated for income tax purposes. If a dispensation is granted in respect of a particular benefit, it has been HMRC’s published practice since 1994 that no NICs will be payable in respect of that benefit. Here, HMRC accept that if the dispensations operated to exclude the allowances from the PAYE system, they also excluded them from liability for NICs. For these reasons, and because we are not required at this stage to determine whether the outcome for NICs purposes differs from the outcome in respect of tax, we shall say very little about NICs, and we will not deal further with the legislation governing them.

42.        We should add for completeness that the RTA and RTB schemes were only one of many topics of discussion between Reed, RR and HMRC during the relevant time. All of the dispensations with which we are concerned covered a number of other payments of no relevance to this decision (for example, overnight subsistence payments to long-distance lorry drivers).

Reed and Robson Rhodes

43.        We understand that Robson Rhodes (“RR”) had been Reed’s tax advisers for some time before 1998. Mr Rayer was the tax partner responsible for Reed’s tax affairs at the time; Miss Ollerenshaw, who played a major role in later events, did not join RR until January 1999. Other RR personnel played a part too, but as the lead was taken by Mr Rayer and, later, by Miss Ollerenshaw, we shall concentrate on their evidence.

44.        Mr Rayer realised—as Mr Beal confirmed—that the phasing out of PRP would be regarded by Reed as an adverse development. Reed thought that its ability to attract good quality staff was enhanced by the benefits of the PRP scheme, and it was keen to find an alternative to replace it once it ceased to be available. Mr Rayer considered that the forthcoming coincidental, though unconnected, relaxation of the hitherto strict treatment of travel expenses paid to employees represented an opportunity which might be used to Reed’s advantage. He was aware of Reed’s policy, following the 1995 review of the PRP scheme, that most of its workforce should be employed by it. As the employed temps had successive assignments to Reed’s clients at different locations, and some were likely to experience gaps between their assignments, he took the view that, under the new rules, expenses incurred by such workers in travelling between their home and their various workplaces, that is the client’s premises, would no longer be non-deductible ordinary commuting expenses, but would become deductible for tax purposes, though initially in the worker’s hands. He therefore approached Mr Beal with a view to Reed’s introducing an appropriate scheme.

45.        Mr Rayer’s suggestion was plainly an attractive one, and it is evident that Mr Beal was very keen to introduce an expenses scheme if it should prove possible. The decision was therefore taken that Reed should endeavour to introduce a scheme by which employed temps who elected to do so and satisfied the eligibility requirements could receive payments in respect of travel and (which the new rules also allowed) subsistence while working away from a permanent workplace. It was clear from the evidence that Mr Beal and Richard Ingram, another member of Reed’s finance team who dealt with much of the detail, considered this an important project and devoted a good deal of their time to discussions with, in particular, Mr Rayer designed to devise and implement such a scheme.

46.        We accept nevertheless that Mr Beal was conscious of the need to avoid the risk, which in the event had not materialised in the course of the review of the PRP scheme in 1995, that HMRC might seek to revisit retrospectively any scheme which was introduced. His evidence, set out in his first statement, was that “There was no question of Reed introducing a travel allowance scheme on a tax free basis without openly and constructively engaging with HMRC, to gain their agreement to the scheme, which after all was being set up following the recent change in the tax regime and would be innovative in its application to the employment business sector (as opposed to, for example, the construction sector).”

47.        Mr Beal’s explanation of the rationale behind the introduction of the RTA scheme was that the pay of all temporary employees has always included an implicit travel allowance, since unless the temporary employee could earn more than the cost of travelling to the assignment and the extra cost of subsistence there would be no incentive to work. Miss Ollerenshaw went further in the course of the correspondence between RR and HMRC: the travel and subsistence element of temporary employees’ hourly rates were, she said, “grossed up” for tax and the increased hourly wage cost passed on to clients. We are bound to say that if that was a feature of Reed’s calculation of the hourly rates it paid, and charged to its clients, there was no evidence that either the temporary employees or the clients were aware of it.

48.        Nevertheless, Mr Beal’s evidence was that the scheme was set up to formalise that element of a temporary employee’s pay package, as Reed perceived it to be, and to allow the temporary employee to benefit from the liberation from income tax and NICs the dispensation brought with it. It was intended, he said, that the allowances should benefit both the temporary employee and HMRC, as each would be saved the time and effort of dealing with individual expenses claims. He made the point that, before the introduction of the scheme, Reed’s temporary employees had suffered deductions of tax and NICs on everything they received because, under the pre-1998 rules, no relief for travelling expenses and subsistence was available to them and, from 1998, few if any would have gone to the trouble to claim the relief. Moreover, they saw some immediate benefit (albeit initially, as we shall explain, they had to wait until £50 accrued) whereas, had they made their own claims, they would have had to wait for several months for a refund, and (because no relief was available in respect of NICs already paid in accordance with the PAYE scheme) would still have paid NICs on the gross pay. The saving in employer’s NICs, of course, all accrued to Reed’s benefit. Nevertheless, as Mr Beal told us, Reed considered the RTA and RTB schemes to be of commercial importance, in that the workers saw some enhancement to their pay, and Reed was able to reduce its charges to its clients.

49.        It is appropriate to observe at this point that we can accept that few of its employed temps would have realised (unless so informed by Reed) that the changes in the tax rules gave them the opportunity of claiming relief for their travelling and subsistence expenses, and that even those who were aware of the consequence of the changes might not have gone to the trouble to claim. We are, however, not persuaded by the premise of Mr Beal’s argument that employed temps of the kind engaged by Reed are in any different position from an ordinary employee, nor that they would see themselves as being in a different position: no-one will be tempted to work if the earnings do not exceed the expenses incurred in undertaking the work. What is conspicuous, as we shall explain below, is that the amounts which an employed temp gained by participating in the scheme were so modest that the only possible conclusion is that they were unlikely to have made any difference to the worker’s deciding whether or not to take an assignment. In other words, although we accept that Reed needed to pay sufficient to attract and retain its employed temps, we are not persuaded that, whatever superficial attraction their presentation may have given them, the RTA and RTB schemes in fact made any material difference to the financial attraction, to a worker, of an engagement by Reed. All of the evidence showed that the schemes were not a means of significantly enhancing the employed temps’ pay, but a means of reducing Reed’s costs.

The first dispensation

50.        Mr Beal told us that, once he had learnt from RR that one might be obtained, he was anxious to secure the protection of a dispensation, incorporating agreed scale rates for the allowances Reed was to pay. It had not been possible to obtain similar protection in respect of the PRP scheme, since such arrangements were not within the scope of s 65 (although Reed had secured dispensations for other staff benefits, such as relocation expenses, in the past); the prospect of obtaining a dispensation for the new schemes, which not only Mr Beal but also Mr Rayer recognised were innovative, and eliminating the risk that Reed would be found later to have implemented them incorrectly, was therefore very important. Though he did not go quite so far as to say that Reed would not have proceeded without a dispensation—the scheme was, we think, sufficiently attractive to Reed to make that a possibility—it was clear that Mr Beal was willing to do everything reasonably possible to secure one.

51.        Mr Beal left the conduct of the negotiations with HMRC to RR, but did say something about the tactics Reed and RR decided to adopt. It was to secure HMRC’s agreement in principle to the payment of travel and subsistence allowances, before engaging in a detailed discussion about the amounts which might be paid, and the manner in which the scheme would be implemented and administered. He said it was to be made clear to HMRC at the first stage that those participating in the scheme would not receive a tax-free supplement on top of their existing pay, but would suffer a reduction of pay (to reflect the fact that what Reed paid already included, at least in Reed’s eyes, some contribution towards travel and subsistence costs) and receive a tax-free allowance in addition to the reduced pay. It was therefore clear, he said, that the employed temps were giving up some of what they would otherwise have received as earnings in order to obtain the benefit of the allowances.

52.        Despite his continuing desire to include agency workers in the scheme which Reed eventually introduced, Mr Beal told us that he understood that only employed workers required to work at different sites for periods of no more than 24 consecutive months could be included, 24 months being the boundary between a temporary and a permanent engagement (for the reason for this limit, see para 188 below). His first witness statement shows that he felt some frustration that HMRC did not seem to accept that Reed intended to make the scheme available, at least at the outset, only to those with whom it had a contract of employment. He took the view that a contract which had satisfied HMRC in the context of PRP should be equally satisfactory in the context of RTA, but he realised from what he considered to be obtuseness on HMRC’s part, particularly a reluctance to accept that Reed was not proposing to include its agency workers within the scheme (despite his hope that it would eventually be able to do so), that there remained some doubt whether even what were by this time Reed’s employed temps would be accepted by HMRC as eligible to participate.

53.        Mr Beal told us that employed temps who undertook only one assignment were very rare (Mr Baddeley gave an example of a client which wanted a particular technical task performed for a few weeks each summer, for which purpose he recruited a research student who undertook only that single assignment) and it was, therefore, usual for employed temps to enter into only one contract with Reed covering a series of assignments (though from time to time the employed temp might be asked to enter into a new contract with amended terms) and that branches were instructed to exclude from the RTA and RTB schemes any employed temp taken on for what was expected to be a single assignment. Similarly, Reed took care to exclude those taken on for long-term (that is, more than 24-month) assignments, and those who in fact stayed in one assignment for more than 24 months. We did not discover the detail of how it did so, although Mr Beal made the point that assignments of such length were very unusual as it would normally be more economic for the client to take a permanent employee. As it is not part of HMRC’s case that the 24-month rule was breached, we shall not deal further with this point.

54.        Mr Beal’s first witness statement indicated that he did not have a clear understanding of the requirements HMRC were seeking to impose, beyond his impression that they were anxious to ensure that agency workers were excluded, but he emphasised his lack of any recollection of Reed’s having been advised by HMRC, before June 2006, that a single contract which expressly continued over several distinct assignments was necessary, still less did he recall any suggestion that a commitment on the part of Reed to provide a minimum number of hours of work was required (the significance of these requirements is dealt with at para 169 below). Had there been any such suggestion, he said, he would have put the reconsideration of the contracts in hand immediately, as Reed did after the fifth dispensation was revoked.

55.        We add parenthetically that the evidence indicated that in 1998, not only Mr Beal and Mr Rayer but also the HMRC officers dealing with the matter did not have the need (if there be such a need) for a single contract in mind. Mr Beal’s evidence, as it is set out in his first witness statement, was that while he understood that a contract of service (rather than a contract for services) was essential, that (impliedly rather than expressly) the contract should at least be capable of covering more than one assignment, and that an assignment for a period exceeding 24 months must be excluded, he did not understand that any other formal requirements had to be satisfied. We do not think that his (or, equally, Mr Rayer’s) understanding at that time differed materially from that of the HMRC officers, or that the perceptions, on either side, of the essential requirements changed significantly for some time.

56.        In April 1998 HMRC began a review into Reed’s NICs compliance. It was no more than a routine review, but it was complicated because of the large number of people employed by Reed (or who, for PAYE and NICs purposes, were treated as employed by Reed) in any year, and the fact that many of those people had had more than one employment in the year. For these reasons the review took a long time, and was demanding of Reed’s resources. Mr Beal had little involvement in the review—Mr Ingram was the senior Reed representative—but he was of course aware of it and anxious that the review, the application for a dispensation and the subsequent implementation of the RTA scheme should not become intertwined with the consequence, as he feared, that the application for a dispensation might be compromised.

57.        HMRC wrote to Reed on 17 April 1998 to make arrangements for the review to begin, and in the same letter asked for (among other things) sample contracts of employment, documents relating to the PRP scheme operated by Reed and a copy of the Temp Handbook. The review was, of course, of Reed’s past compliance and had nothing to do with the RTA scheme, then still in the future even if it was later back-dated to 6 April 1998. HMRC’s own records relating to the review show that two contemporaneous sample contracts were provided. Mr Beal took the view that HMRC must have known from the material they received in the context of the review that Reed employed most of its temporary employees, and that the PRP scheme was open only to employed temps.

58.        Mr Rayer recognised that Reed would need to agree with HMRC on scale rates for the amounts to be paid to the employed temps, both to secure the dispensation which he considered to be desirable protection for Reed and in order to reduce the administrative burden on it when operating what, in due course, became the RTA Scheme. A significant amount of work was conducted by Reed, he understood, in calculating appropriate rates for the proposed scale rate payments, a process which required it to undertake an extensive analysis of its employed temps’ work patterns and travelling habits.

59.        It was also apparent from the evidence that Reed recognised that any scheme it introduced would need to meet three essential requirements: that the employed temps agreed to work for a lower wage or salary in order to benefit from the scheme; that the expenses Reed was to reimburse were deductible expenses; and that the amounts paid, taken overall (since round sum payments were to be made) did not carry with them an element of profit in the employed temps’ hands.

60.        The dialogue with HMRC began when, on 15 September 1998, RR wrote to HMRC on behalf of Reed (specifically, Reed Staffing Services Limited) in order to apply for a dispensation in respect of the costs of travel and subsistence incurred by employed temps travelling to work at Reed’s clients’ premises. The request was for a dispensation to take effect retrospectively, from 6 April 1998. RR’s letter described the category of expenses that were to be covered by the requested dispensation as

“travel and subsistence expenses when [employed temps]—

·             have no permanent workplace; and

·             are required to attend at various locations for the purpose of performing tasks of limited duration or for some other temporary purpose.”

61.        The letter went on to make it clear that “limited duration” meant not more than 24 months. In order to anticipate what RR thought would be an inevitable question, it also stated that “our client wishes to have the option to pay ... round sum allowances, which do no more than meet the actual costs incurred.”

62.        We are not entirely convinced the letter was wholly frank. The evidence showed that the analysis which Reed claimed to have undertaken of its employed temps’ travelling habits and costs was by no means as extensive or detailed as Mr Rayer had been led to believe, and that the calculation of the subsistence payments for which approval was sought was somewhat “back of an envelope”. An internal RR email, sent on 17 September 1998 (two days after the letter was sent) included the comment that Helen Riley, a tax partner at RR, “does have some qualms about the issue, in terms of lack of disclosure to the Revenue”.

63.        Mr Rayer’s evidence, however, was that this letter was no more than the first step in exploring the possibility of Reed’s obtaining a dispensation following the change in the rules. He thought, he said, that it was important to secure HMRC’s agreement in principle to a dispensation, before pursuing more detailed discussions with them on the mechanics of how the allowances might be paid. He emphasised that he was very conscious of the innovative character of the proposal (he believed this to be the first such application to be made by an employment business), and was, he told us, at pains to ensure that it was handled properly. Although he considered that HMRC’s published guidance supported the conclusion that Reed’s employed temps might well qualify, he viewed the discussions which took place, after the initial letter was sent, between Reed, RR and HMRC, and the letters RR wrote on Reed’s behalf, as steps in a process of negotiation and exploration. Despite the published guidance he did not know, he said, whether HMRC would accept that the employed temps did indeed qualify: the relaxation had, he was aware, been introduced with peripatetic construction workers in mind. He was conscious too that the legislation provided that a dispensation could be granted only if the inspector was satisfied that no additional tax was payable. Thus he expected that before granting a dispensation HMRC would carry out their own assessment, an assessment which would give him some reassurance that the intended allowances satisfied the statutory requirements.

64.        HMRC did indeed make some enquiries. The evidence indicated that there were three main concerns: whether the allowances were to be payable only to employed temps, evidently because it was still not clear to the officers at that stage that Reed had largely ceased to engage agency workers; that payments would be made only for those days on which the employed temps worked; and that Reed would monitor the scheme sufficiently rigorously to ensure that allowances were not paid to those employed temps who worked, or were expected to work, at one location for 24 months or more.

65.        It is plain from the contemporary correspondence and notes of meetings that the employment status of the workers who were intended to participate in the scheme was the matter of greatest concern on both sides. The HMRC officers took some convincing that they were truly employed by Reed; but rather surprisingly did not ask for sample contracts of employment (which were not volunteered by Reed or by RR on its behalf). Mr Rayer told us that he was of the clear view, after perusing HMRC’s own published guidance and the contrasting examples it provided, that the relevant workers were indeed employed. It emerged that Reed told Mr Rayer that it proposed to take legal advice about the terms of its contracts of employment once a dispensation had been granted; we were told that such advice was taken, but did not discover its result.

66.        After the initial exchange of correspondence with HMRC, Mr Rayer decided it would be sensible to arrange a meeting. In anticipation of that meeting, he discussed his intended approach to HMRC with Mr Ingram. His note of the discussion includes the comment that “the Inland Revenue would be advised that there would be a general substitution of taxable pay with tax free allowances, on the basis that the current hourly rate is calculated on the assumption that travel needs to be incurred but cannot be paid in a tax free environment”.

67.        We should perhaps interpose at this point that that the evidence showed that there was at the time a degree of uncertainty within not only RR and Reed but also within HMRC about how the then new, and untested, rules were to be applied in practice. HMRC published guidance, on which both its officers and RR relied, though it was accepted before us that it was not as clear as one might have wished.

68.        The meeting with HMRC took place on 9 October 1998. Mr Rayer and his assistant, Pat Sims, attended on behalf of Reed; Tim Downes, the inspector who later granted the dispensation and another officer, Diane Kirkham, represented HMRC. The discussion appears to have disposed of most of Mr Downes’ concerns, since he asked Mr Rayer to provide a summary of the proposed arrangements, which he could refer to “higher authority”, apparently a necessary step since he considered a dispensation would set a precedent, while at the same time he made it clear to Mr Rayer that his own view was that the dispensation should be granted. On 23 October Mr Rayer wrote with the summary Mr Downes had requested. He made it clear that Reed recognised that agency workers could not be included within the scheme, that it would have to monitor its operation in order to ensure that those taking engagements exceeding 24 months were excluded, and that the amounts paid would need to be kept under review. He emphasised the research Reed had carried out into the costs actually incurred by its employed temps, and offered a breakdown (of no continuing importance) of the proposed allowances. On 18 November he wrote again with some additional information which had been requested in a telephone conversation earlier that month, and at the same time sent a copy of the Employee Guide—as its name indicates, an explanation aimed at those employed temps who were (or, at this stage, were expected to be) within the scheme, and later recruits intending to join it—which had been prepared by Reed in anticipation of its introduction. The draft of the guide was by then in its near-final form, but it had of course not yet been issued as the scheme had not been implemented.

69.         The need for an Employee Guide had been one of the topics of discussion at the meeting on 9 October, both sides evidently taking the view that the operation of the intended scheme must be clearly spelt out to those affected by it. It described the expenses covered what had by then been named the Reed Travel Allowance, or RTA, scheme as those incurred for “travelling from home to a site (temporary place of work)”. The same guide described those who were eligible to participate as

“All temporaries employed by Reed Staffing Services … unless they have a permanent pace of work defined as:

·             two years actually worked in one location as a principal place of work, or

·             an expectation that two years will be spent in one location as a principal place of work.”

70.        In the two letters, of 23 October and 18 November 1998, written by RR to HMRC, appear respectively the following statements:

“We confirm that the dispensation will apply only to employees of Reed Staffing Services Limited who have no permanent workplace, and who are required to attend various locations for a limited period only.”

“We confirm that the temporary workplace of the employee will vary from one assignment to the other.… The only time similar journeys will be an issue is when the assignment is not for a limited duration or temporary purpose, in which case the employee will not be entitled to any expense allowance for travel and subsistence.”

71.        HMRC say that it is apparent from these statements that, by the time the first dispensation was granted, Reed and RR had concluded for themselves (though whether rightly or wrongly is a matter of dispute) that the employed temps who were to participate in the schemes, if a dispensation were granted, had no permanent workplaces, because the clients’ premises were to be regarded as temporary workplaces. In our view that must be right; these passages have no other possible interpretation.

72.        Both of the letters also made several references to Reed’s paying the allowances to the employed temps. The letter of 23 October referred to “reimbursement of travel and subsistence costs”, and stated (in support of Reed’s request for permission to make scale payments) that “Our client wishes to make Round Sum Allowances”; it also referred to the “Round Sum Allowances our client will pay”. Similar expressions appeared in the later letter.

73.        The First Dispensation, which as requested had retrospective effect to 6 April 1998, was granted by a letter written by Mr Downes on 30 November 1998. His letter specifically referred to the two passages quoted at para 70 above and, say HMRC, expressly incorporated them within the terms of the dispensation. Mr Downes’ letter set out various maximum allowable amounts. The permitted amounts are not of importance in themselves; we need merely observe that it is not suggested that Reed ever exceeded the amounts from time to agreed with HMRC. The letter began by stating that its effect was to relieve Reed of the obligation of reporting the various payments to which it applied in its annual returns—a paragraph which, with minor variations, appeared in every dispensation. The second and third paragraphs, too, were common (with minor, insignificant, differences) to all the dispensations; they said that:

“I am giving you this dispensation because I am satisfied, on the basis of what you have told me, that no additional tax would be payable by the employees concerned on these expenses payments and benefits. I am authorised to do this by [s 65 of ITEPA].

This dispensation applies only to the expenses payments and benefits, set out below, in the circumstances there set out. If the expenses payments or benefits are paid or provided in circumstances which give rise to additional tax, this dispensation will need to be revoked. Where necessary, the revocation may apply to expenses payments and benefits already provided. In that case additional tax will be due. So it is important that you let me know if you alter your system for controlling expenses payments and benefits, or increase their amounts, or change their nature or make any other changes which may affect their taxability.”

74.        The “circumstances there set out” were those detailed in the two passages from RR’s letters to which Mr Downes had referred, with minor supplements of no continuing significance. The warning within the dispensation itself was reinforced by the statement in Mr Downes’ covering letter that “I would stress that this dispensation applies only to such expenses incurred in the circumstances detailed in your letters … all other expenses and benefits … should continue to be reported on forms P11D” (the form by which an employer is required annually to provide to HMRC a list of expenses payments made to employees).

75.        Reed began to operate the scheme once the first dispensation had been granted. It is more convenient to deal with the mechanics of the RTA scheme and its successor, the Reed Travel Benefit or RTB, scheme after concluding the chronology of the dispensations.

76.        We should perhaps also make the point that although we have dealt with the granting of the first dispensation in some detail, since the circumstances in which it was granted set the scene for what follows (its successors led to no more than evolutionary change to it), it was revoked and replaced by the second dispensation only three days into the relevant period.

The second dispensation

77.        As we have said, Miss Ollerenshaw joined RR in January 1999, shortly after the grant of the first dispensation. She has, she told us, been an employment tax specialist for most of her working life, spending some time with HMRC before moving into private practice. She joined RR in order to lead the PAYE section of RR’s tax investigation team, and was initially engaged on other matters for Reed (and other RR clients). On 10 April 2000, however, she drafted a letter to HMRC, which Mr Beal was asked to approve before it was sent to HMRC, in which she requested on Reed’s behalf an increase in the scale rate payments recorded in the dispensation then in force, to take effect from 6 April 2000. The increase requested was one which would at least match the rise in the Retail Price Index since the (back-dated) inception of the RTA scheme in April 1998. By April 2000, as the letter (which Mr Beal duly approved, and which was then sent) pointed out, almost 18 months had elapsed since the first dispensation was granted.

78.        After some initial doubts had been dispelled, HMRC’s response to the request for an increase in the rates was favourable (it was implicit in the first dispensation that the rates would be reviewed on a regular, probably annual, basis), but the negotiations nevertheless proved to be more protracted than those for the first dispensation, almost certainly because Reed, and RR on its behalf, endeavoured to persuade HMRC to agree not only to an increase in the allowances, but also to the extension of the scheme to agency workers. Lengthy correspondence led to a meeting on 29 September 2000, attended by Mr Rayer and Miss Ollerenshaw for RR, and Nick Read and Mrs Kirkham for HMRC. By this time, Mr Read had replaced Mr Downes as the inspector responsible for considering Reed’s application. Mr Rayer’s recollection of the meeting was that some old ground was covered, no doubt because Mr Read was unfamiliar with the details. A clear distinction was again drawn between the employed temps and the agency workers, who by this time had become entitled to certain benefits under European Community regulations. Mr Rayer told us he believed both sides understood that the existing dispensation applied only to employed temps. Although an increasing majority of Reed’s temporary employees (save those registered with Reed Health) fell into that category, it still had some agency workers.

79.        On 5 December 2000, after it had been approved by Mr Beal, Miss Ollerenshaw sent a formal letter to HMRC, requesting an increase in the monetary limits set out in the first dispensation, and including arguments (to which we shall return shortly) in support of an application to extend the dispensation to include agency workers, essentially on the ground that, despite their different employment status, the essential features of their relationship with Reed were the same as those of the employed temps. Mr Read replied on 9 January 2001. He asked for further information about the agency workers, to enable him to understand better the difference between them and the employed temps (a fact which encouraged Miss Ollerenshaw to think they might be included), although the observation in his letter that “the main difference is that the latter category (i.e. temporary employees [by which he clearly meant employed temps]) are engaged under a contract of employment” suggested to Mr Rayer that he already understood the distinction.

80.        By a second letter of the same date, Mr Read granted the second dispensation, revoking the first dispensation. This dispensation, unlike the first, extended to all the Reed group of companies, and it covered a number of expenses not included with the first, though none of the added categories is of present relevance. Some of the monetary amounts were increased, as Reed had requested. The letter referred to payments to “employees”, and made no mention at all of agency workers, though a covering letter made it clear that the latter were after all excluded from its scope. It contained the same warning about revocation as the first dispensation (see para 73 above). We should mention for completeness that Reed Health demerged from the rest of the Reed group during the currency of this dispensation, but the allowances it paid remained covered by it.

 The third dispensation

81.        The second dispensation was evidently regarded, at first, by Reed and RR as a staging post on the way to having agency workers included within the RTA scheme. Mr Beal and Miss Ollerenshaw, who by this time was increasingly heavily involved in developments, remained keen to pursue the application for that extension. Her letter of 5 December 2000 to HMRC, which led to the increases in the monetary amounts effected by the second dispensation, set out the reasoning behind the request:

“We consider that if an agency worker is engaged to work at an agency’s client’s premises, then provided they are not engaged merely to work on a particular contract and are available for work on contracts with other clients of that agency, tax relief should be due on their costs of home to the client’s premises and any associated subsistence on the same basis as for employees.”

82.        Miss Ollerenshaw’s response to Mr Read’s letter of 9 January 2001 (see para 79 above), seeking further information about the agency workers, was sent by fax on 24 January. It contained the following passage:

“You are correct in your understanding that ‘employed temps’ are engaged under contracts of employment but they are not full time contracts and only apply when the employed temps are carrying out assignments on behalf of Reed. At these times they have all of the benefits and rights afforded by their contracts. The ‘temporary temps’ [ie agency workers] are workers engaged under agency contracts and would, in the absence of the deeming legislation under s 134 ICTA 1988 [now ITEPA s 44], be classed as self employed. Consequently the temporary temps would only have any rights afforded to ‘workers’ rather than employees under certain law, for example, the Working Time Directive.”

83.        In her first witness statement she added “I intended to reflect what I understood from my conversations with Reed, that the Temporary Employees [ie employed temps] were only entitled to be paid for the hours which they worked or for holidays (or statutory payments).” We observe at this juncture that that understanding was correct; no payment, other than those required by statute, was made when the temporary employee did not work.

84.        Miss Ollerenshaw’s statement in her fax that the employed temps did not have full-time contracts is a point on which HMRC place considerable reliance. It was, she said, made without giving the matter a great deal of thought. At that time, she said, both she and (as she believed) HMRC considered that only two conditions, employment and an expectation that the employee would work on more than one assignment at different locations, needed to be satisfied if the travel and subsistence expenses were to be allowable.

85.        It is also necessary to record Mr Beal’s evidence about Miss Ollerenshaw’s fax. At para 110 of his first statement he said:

“I should make clear, with regard to this letter, that it is Reed’s case that the employment contracts with Reed’s Temporary Employees [again, meaning employed temps] in fact continued across assignments (including across any gaps between assignments). However, Sue Ollerenshaw is telling HMRC in this letter that our contracts of employment are not expressed in that way, and that the contracts of employment only apply when the Temporary Employees are on assignment. Had Nick Read thought that the nature of the Temporary Employees’ employment contract with Reed was relevant to the grant or maintenance of the Dispensation, and in particular had he thought that there had to be a single contract that continued across assignments, he would presumably have said so upon receipt of this letter. He did not do so. Equally and in light of the way that HMRC puts its case in these proceedings, if Nick Read had thought it essential that Temporary Employees be paid in any gaps between assignments, it should have been clear to him from Sue Ollerenshaw’s letter that Reed did not pay its Temporary Employees during gaps.”

86.        The nature of the relationship between Reed and its employed temps, and the effect of the contracts by which they were engaged, is one of the issues we must decide, and we shall return to this point later (see para 157 below170 below).

87.        In March 2001 Mr Read asked for copies of Reed’s contracts with its permanent employees, employed temps and agency workers (which, quite surprisingly in view of the importance both sides attached to the matter, had hitherto been neither requested nor volunteered) and enquired about the criteria which determined whether a worker was taken on in one capacity rather than another. Miss Ollerenshaw’s reaction to that request is dealt with at para 35 of her first witness statement:

“… I only considered HMRC’s request to have been made in the context of Reed’s proposal that the (then) Second Dispensation be extended to cover Agency Workers. The Second Dispensation covering Reed’s Temporary Employees had already been issued and so I understood that HMRC had already satisfied itself that the travel expenses paid to Reed’s Temporary and Permanent Employees were properly deductible.”

88.        She did, however, provide copies of the contracts (one example each of Reed’s contracts with its permanent employees, its temporary employees and its agency workers) under cover of a letter of 23 March 2001 to Mr Read, in which she also offered an answer—if we may say so somewhat disjointed and uninformative—to Mr Read’s question about the criteria which dictated whether temporary employees were taken on as employed temps or agency workers.

89.        We did not discover whether Mr Read was satisfied by the explanation since the evidence showed that by this time Mr Beal had become somewhat nervous about the possible ramifications of the extension of the RTA scheme to agency workers, and in particular the effect it might have on the VAT treatment of Reed’s charges to those of its clients with which it placed agency workers. Many of those clients were unable to recover VAT (which was the reason why Reed, and Reed Health in particular, continued to engage agency workers) and Mr Beal was concerned that Reed might be compelled to add VAT to the entirety of its charge for the supply of agency workers to its clients, rather than to the commission element alone. We record this point as a reservation entertained by Mr Beal, but it is not appropriate for us to say any more about it, as Reed is engaged in other litigation about the VAT treatment of its charges, and it is not an issue before us.

90.        However, the reaction to Mr Read’s request is material to this decision. It led to a meeting between Mr Rayer and Mr Beal on 9 May 2001, at which Mr Beal voiced his concerns. The meeting is the subject of para 67 of Mr Rayer’s first witness statement, which describes the course of action on which they decided, and contains a further observation to whose significance, as we see it, we shall return:

“During the meeting, Mr Beal and I discussed Reed’s application to extend the Dispensation to cover their Agency Workers and HMRC’s request for copies of the contracts to which I referred above. I recall mentioning in the discussion that whilst providing to HMRC copies of the contracts applying to ‘temporary temps’ [ie agency workers] was relevant to Reed’s application for an extension, the contracts relating to ‘employed temps’ did not appear to be relevant to the application. The Dispensations relating to ‘employed temps’ had been granted after two detailed reviews by HMRC and I was alive to the possibility that if all the contracts were to be reviewed this could result in an unnecessary review being carried out with all the attendant cost and inconvenience to Reed. We discussed this and the potential impact of the impending draft Regulations (latterly the Conduct of Employment Agencies and Employment Businesses Regulations 2003), which would result in Reed losing the benefit of paying VAT only on the commission element of hiring out staff and that the Agency Workers would be employees in any event. This would negate the need to seek to extend the Dispensation to Agency Workers. In the light of this discussion, Mr Beal decided that the application to extend the Dispensation to include the Agency Workers should be withdrawn.”

91.        Mr Rayer, who was to leave RR later in 2001, played no further part in the relevant events; Miss Ollerenshaw, who had hitherto been assisting him, took over the main responsibility for advising Reed about its employment tax issues, including the negotiation of the later dispensations. Her principal contact at Reed remained Mr Beal; she continued Mr Rayer’s and her own practice of obtaining his approval of letters to HMRC before they were sent. The HMRC officers with whom she dealt were, initially, Mr Read and Mrs Kirkham; she had had extensive dealings with them in the course of the negotiation of the increases in the allowances as they were incorporated in the second dispensation. Mr David Brook replaced Mr Read at a later stage.

92.        Shortly after the 9 May meeting between Mr Rayer and Mr Beal, Miss Ollerenshaw telephoned Mr Read to inform him that the request for an extension of the current dispensation to agency workers was not to be pursued, giving as the reason the forthcoming regulations and their assumed effect of compelling Reed to employ all of its temporary employees in the future. She does not appear to have mentioned VAT as a cause. We should, however, add that Miss Ollerenshaw is not a VAT practitioner, and we accept that this consideration would not have been at the forefront of her mind. Reed Health, which continued to use agency workers until 2004, did not abandon the request but it was not, in fact actively pursued and the dispensation was never extended to include Reed Health’s agency workers.

93.        One immediate consequence of the withdrawal of the application for the extension was that the request for copies of the contracts was not pursued, and none were supplied at that time. In fact, the request was not renewed until July 2004 when, as Mr Beal understood, HMRC had become concerned that the RTB scheme, which was then in place, was leading to “tax leakage”. We shall return to this topic (see para 170 below).

94.        During the remainder of 2001 discussions continued between Reed and RR about the impact of the regulations. These discussions led to the decision that all the Reed companies save Reed Health should cease employing agency workers altogether. Reed Health, which was at that time not part of the Reed Group (though RR remained its tax advisers), and to which special considerations applied, continued to do so, but only until 2004. The decision appears to have been taken at, or in the light of, a meeting on 8 October 2001 attended by Mr Beal, Miss Ollerenshaw and Ms Lorraine Parkin, a VAT specialist employed by RR. The principal focus of the meeting, according to Miss Ollerenshaw’s recollection, was the VAT consequence of changing the status of agency workers to employees, and as that was Ms Parkin’s field she prepared the meeting note. In it appears a record of Mr Beal’s having said that “the permanent contracts are from assignment to assignment. There is no umbrella contract and no mutuality of interest.” Miss Ollerenshaw’s evidence was that she did not recall the remark.

95.        At the same time further discussions with HMRC continued. They focussed not only on Reed’s wish to secure an increase in the allowances, but also on HMRC’s concern that the payslips were poorly laid out, leading to confusion in the minds of the employed temps, and a continuing large volume of enquiries to HMRC’s helpline (see para 121 below). On 29 January 2002 Miss Ollerenshaw wrote to Mr Read formally asking for increases in the permitted amounts, giving an explanation of the route by which she had arrived at those amounts. The layout of the payslips was a continuing concern to Mr Read, and Reed went to some lengths to produced revised payslips to his satisfaction. Eventually, in the course of further meetings, he accepted that the revised payslips were a sufficient improvement on the old version.

96.        The third dispensation was granted, by Mr Read, in a letter dated 7 February 2002. The amounts now permitted in respect of Reed’s employed temps were daily subsistence of £5.50, plus a daily travel allowance, in London equivalent to the average cost of a zones 1 to 6 travel card (£9.50), elsewhere on the average cost of public transport (£3.95), in each case “as detailed in your letter of 29 January 2002”: the amounts allowed were exactly as Miss Ollerenshaw had requested. She was clearly elated by the grant of this dispensation, and she wrote immediately to Mr Beal to give him the good news. Her letter contained the following passage:

“I also enclose a copy of my letter to the Inland Revenue dated 29 January 2002 and the attached pay-slip layout which clearly shows a ‘reward adjustment’ of £14.75 for the week which I explained to the Inland Revenue could cover salary sacrifices in respect of pension contributions, other benefits under the Reed Benefits Scheme and effectively sharing the benefit of the travel arrangement. I made it clear that the level of the salary sacrifice would be agreed in advance with the employees so that:

it would be effective for tax purposes; and

the employees would understand their pay-slips and would not therefore need to contact the Inland Revenue at Bradford Valley View.

The above is excellent news for the Group from several points of view. Firstly, it would now be very difficult for the Inland Revenue to seek any tax and NIC from the company retrospectively. This is on the basis that our recent meetings and my letter clearly demonstrate to the Inland Revenue that, under the current arrangement, salary sacrifices are calculated individually based upon the grossed up equivalent of the expenses payable per the P11D Dispensation. Whilst the Inland Revenue indicated that they were not happy for this practice to continue, they have not tried to recover any tax or NIC for the past.

Secondly the Inland Revenue are aware that for the future the arrangement is being operated on quite an ‘aggressive’ basis as the company is sharing the benefits by way of salary sacrifices.…”

The fourth dispensation

97.        The grant of the fourth dispensation, which again increased the permitted amounts (to subsistence of £5.66, and travel to £11.45 in London and £4.47 elsewhere), was relatively uneventful. There was little more than a meeting between Mr Read and Mrs Kirkham for HMRC, and Miss Ollerenshaw and a colleague for RR, on 19 February 2003, followed by a letter from Miss Ollerenshaw of 5 March, leading to Mr Read’s letter granting the revised dispensation on 7 March. We need only record the observation in Mr Read’s own note of the 19 February meeting that “Apparently one or two employees had complained they had no entitlement to SSP because [of] inadequate NI contributions. In Read’s opinion employees could not enjoy the advantages of the scheme, ie less tax/NIC and then expect same SSP.”

98.        A separate, but for present purposes identical, dispensation was granted to Reed Health at the same time.

The fifth dispensation

99.        The fifth dispensation, too, was relatively uncontroversial. It was granted by a letter of 3 February 2004, written by David Brook, who had by this time succeeded Mr Read as the responsible inspector. There was a modest increase in the daily subsistence allowance, to £6.09, but the travelling allowances were unchanged. By a second letter of the same date Mr Brook granted an identical dispensation to Reed Health.

The mechanics of the RTA and RTB schemes

100.    Mr Beal’s first description of the schemes appears at para 13 of his first witness statement (in which, as elsewhere, he refers to “Temporary Employees” although it is clear he means, in the terminology we have adopted, employed temps):

“The essence of the arrangements was that in return for receiving the scale rate allowance, the Temporary Employees ‘sacrificed’ an amount of pay based on the benefits that could be paid free of tax under the Dispensations and adjusted so that the Temporary Employee shared in the tax saving. Under the original version of the arrangements (known as Reed Travel Allowance (‘RTA’)), there were two adjustments. One was a deduction intended to ensure that the inclusion of the tax free allowance did not affect the Temporary Employee’s net pay (which was an adjustment wholly in Reed’s favour) and the other was the payment of a taxable ‘travel allowance’ (which represented the Temporary Employee’s share of the benefit). This travel allowance was initially paid in £50 tranches and later on a weekly basis. Under the later ‘Reed Travel Benefit’ Scheme (‘RTB’), there was a single (negative) adjustment pursuant to a matrix which, effectively represented Reed’s share of the tax benefit. The essential element of the arrangements, that the Temporary Employees sacrificed an amount of salary based on the amount of expenses that could be paid free of tax under the Dispensations, in return for receipt of the scale rate allowances, remained (and remains) the same.”

101.    One might be forgiven for finding that description somewhat opaque. Indeed, as what follows demonstrates, a lack of clarity in the descriptions of the schemes was a recurrent theme in much of the documentation produced over the years they were in operation.

102.    The Staff Handbook issued by Reed was the principal means by which the workings of the schemes were communicated to participants and prospective participants, although an oral explanation was also offered to new recruits. We shall examine the contracts between Reed and its employed temps, and the provisions relevant to the RTA and RTB schemes, at para 157 below. There were in addition two letters addressed to employed temps during 2001, though nothing turns on the letters beyond the fact that they were sent.

103.    Reed’s position with respect to communication of the details of the schemes to employed temps—which appears to have remained constant throughout the relevant period—was set out in the letter from Mr Rayer to Mr Downes of 18 November 1998, written in the course of negotiating the first dispensation (see para 68 above) in which he responded to Mr Downes’ request for a copy of what was at that time termed the “Employee Guide”. Mr Rayer’s letter contained the passage

“This [the Employee Guide] aims to provide an outline only of the scheme, as there is concern that publishing detailed rules would seek to confuse [sic]. The branch managers will have the details of the payments to be made under the dispensation, and will also be able to advise the employee of the statutory relief available, if the employee wishes to make their own claim. In practice it is anticipated that few employees will wish to opt out of the dispensation, as the amount of work involved in making a statutory claim is likely to be off-putting to most.”

104.    The draft guide attached to the letter included the following:

“Other agencies expect temporaries to meet daily expenses out of the hourly rate which they are paid ie from their net pay. The purpose of this scheme from Reed is to identify separately a level of expenses which would normally be incurred on average by temporaries, and with the agreement of the Inland Revenue pay that round sum as a non-taxable allowance.

Payment of a non-taxable allowance reduces the overall tax paid by Reed and temporaries thus benefiting both parties.…

The allowances are daily rates, paid for each qualifying day of work. The actual rates of allowance will vary from time to time, but will be based on the actual costs incurred by Reed’s temporary workers as a whole. More details may be obtained from your normal contact at your local Reed branch.…

We shall be paying your expenses to you as a non-taxable allowance, so your gross pay will be reduced accordingly. SMP, pension or any other benefit derived from gross pay, taxable pay will all be reduced.”

105.    Mr Baddeley told us that it was indeed part of a temps consultant’s function to explain the workings of the RTA and RTB schemes to new recruits and, when the scheme was introduced, to existing employed temps who were to be included within it. It was, however, clear that the explanation was of the impact of the scheme for the time being in effect on the individual employed temp; there was no evidence that the terms of the dispensation were explained, or even that the temps consultants knew the details of the dispensations themselves.

106.    The following example of the description Reed offered to employed temps is taken from what appears to be a Staff Handbook (which replaced and incorporated what had previously been the Employee Guide) issued in the second half of 2001, and it relates to the later version of the RTA scheme. Although a bundle of handbooks was produced for the hearing, there was some uncertainty about their exact chronological order, since Reed had not maintained a comprehensive library, and some assumptions had to be made. This description is slightly clearer than that which appears in Mr Beal’s witness statement, but in our view is still less than transparent. We have set it out in full because of its importance in relation to issue 1 (see para 189 below).


Participation in this scheme means that for each day that you work in a booking for Reed you can benefit from an amount additional to your normal hourly rate.


If you are eligible (see below) you will receive a Travel Allowance for each day that you work as a Reed Temporary: The value of your Travel Allowance will show on your payslip that week. The scheme is designed to be a tax efficient benefit agreed with the Inland Revenue.


The current rate is an extra £1.50 a day for each day you work over 5 hours with the same client or, if you work less than 5 hours, the rate is 75p per day. These rates may be revised from time to time. So, if you have worked more than 5 hours a day for us every weekday for a year you will receive £378 over the course of the year. Tax and National Insurance is taken off the Travel Allowance Scheme amount at your normal rate.


All temporaries working tor Reed Staffing Services Limited are eligible except those who submit a claim for travel expenses. Unfortunately, if you trade with Reed as a Limited Company, you will not be able to be included in the scheme. Also, if you work for Reed Agency Services, then you will not be able to participate. Your Reed branch will be able to inform you if you are working in a Reed Agency Services booking.


No, if you are eligible, you need do nothing. Please note that if you are in the scheme, you must not include your actual travel and subsistence costs incurred whilst working through Reed as an expense on your tax return. If for any reason you wish to opt out of the scheme, you may do so by letting us know in writing.


·             Your total travel and subsistence allowance is shown as ‘Travel Allowance’ beneath your timesheet pay on the left of your payslip. This value is a gross value, ie, Tax and National Insurance will be deducted from it.

·             Beneath this a figure appears next to the phrase ‘Exp Adj’. This represents the adjustment to your gross pay to allow for the reduction in the total amount of Income Tax and National Insurance due under the scheme.

·             The agreement with the Inland Revenue means that the Tax and National Insurance deductions on your total pay (shown on the right of your payslip) are lower than they would have been without the scheme.

·             The end result is that you get more pay in your pocket than you would have without the scheme.

·             You can work out approximately how much more you get by taking your normal rate of tax off the Travel Allowance sum.


All you need do, as an Inland Revenue requirement of the scheme, is to tell us your daily mileage and if you use public transport to travel to work. You can do this by filling in the boxes on your timesheet each week. Please make sure that you do this, if appropriate, as failure to do so may result in your exclusion from this benefit.”

107.    We interpose at this point the observation that the wording of this version of the Handbook is rather less clear than that of the draft attached to Mr Rayer’s letter to HMRC of 18 November 1998 and of which an extract appears at para 104 above. It is also conspicuous that this version makes no reference to the salary sacrifice mentioned in the final sentence of that extract, unless one regards the text of the second bullet point, obscure though it is, as a reference to a salary sacrifice. Rather to the contrary is the statement in the opening paragraph that “you can benefit from an amount additional to your normal hourly rate”, which does not seem to us to be consistent with the notion that the employed temps were required to give anything up.

108.    As the Handbook indicated, it was possible to opt out of the scheme, though we understand few employed temps did so. Mr Beal told us that employed temps who were not liable to pay income tax or NICs were automatically opted out of the scheme: the obvious reason was that Reed would otherwise be required to pay the £1.50 or 75p per day, but would not be able to make any saving of tax or NICs in return. Once in the scheme, the employed temp was required to do no more than tick boxes on the timesheet which he or she submitted each week, to indicate whether they had worked in central London or elsewhere, and whether they had travelled to work by public transport or by car, in the latter case also stating the miles covered. As the schemes evolved there were some minor changes to the manner in which travel by car was treated, but the changes are of no present importance and we shall not deal with them. No receipts were required and, because of the flat-rate amounts of the allowances, the actual cost of a journey by public transport was irrelevant. It will be observed that the handbook indicated that employed temps participating in the scheme must not make a claim for an expenses deduction by means of their tax returns.

109.    Before the RTA scheme was introduced, a typical temporary employee (whether an employed temp or an agency worker) was placed with a Reed client and paid an agreed sum per hour worked—weekly or monthly earnings consisted of the agreed hourly rate multiplied by the number of hours worked during the week or month (for simplicity we will leave the PRP arrangements out of account, so far as possible, for present purposes). We understand most temporary employees were in fact paid weekly, and we assume weekly payments in the descriptions which follow. There was no retainer for times when no assignment was available; Reed’s contractual obligation amounted, to put it at its highest, to no more than to use its best endeavours to find a placement (and that remained the case until 2006, as explained at para 169 below). The temporary employee, too, was under no obligation to accept any assignment which was offered. Of course, repeated refusals of assignments would lead Reed to cease using that temporary employee, just as Reed’s failure to find placements would lead the temporary employee to look elsewhere. Reed was, however, obliged by law to pay holiday pay, statutory sick pay and maternity pay to its employed temps, and it duly did so. Although those obligations complicated the calculations, they had no impact in themselves on the RTA (or, later, the RTB) scheme since an employed temp receiving any payment of that kind was, necessarily, not working and therefore not entitled to receive the allowances.

110.    The starting point for the operation of the RTA scheme remained the hourly rate multiplied by the hours worked. The total so determined was then adjusted, as the scheme was explained to us, first by the deduction of an amount which was equal to the allowance (for travel expenses, subsistence or both) permitted in the case of that temporary employee by the application of the scale rates set out in the then current dispensation. The tax and NICs for which the employed temp was liable were then calculated by reference to the net amount. The amount previously deducted was then added back, as a non-taxable payment. The taxable pay was then reduced again, by such an amount (the “Exp. Adj.” figure) that the net sum the employed temp received was the same as he or she would have received in the absence of the RTA scheme. The “Exp Adj” figure was simply the difference between the tax and NICs the employed temp would have paid had he or she not participated in the scheme, and the reduced tax and NICs which resulted from that participation. Finally, the taxable pay was increased (taking the figures applying from 2001) by £1.50 or 75p per day, depending on the number of hours worked, a sum which on the payslips was misleadingly called “travel allowance”, though elsewhere it was described as a “travel-to-work payment” (the term we use in this decision) The benefit to the employee of being in the scheme was the after-tax and –NICs amount of this payment.

111.    Each employed temp was required to complete a time sheet in every week he or she worked in an assignment with a Reed client, and arrange for it to be countersigned by the client before it was submitted to Reed for the amount due to be calculated and paid. At first the time sheets were paper-based, but Reed later moved to an electronic system, though this did not affect the principles of the schemes. As we have said, those participating were in addition required to tick boxes indicating where they had worked (differentiating only between central London and elsewhere), whether they had used public transport and, when the current scheme arrangements so required, the miles covered if they used their own cars.

112.    Miss Ollerenshaw had a meeting with Reed’s finance director, Malcolm Paget on 31 January 2001 (shortly after the grant of the second dispensation, and when the earlier version of the RTA scheme was still in use). She produced a “worked example” in order that she could explain the mechanics of the scheme and the benefit to Reed of operating it; it was as follows:

No scheme With scheme Payslip

Gross Pay 100.00 100.00 100.00

Less Travel Allowance (47.25)

Taxable & NICable 52.75

Tax @ 22% (22.00) (11.61) (11.61)

Employee NIC @ 10% (10.00) (5.28) (5.28)

Net pay 35.86

Plus Travel Allowance 47.26


Less Travel adjustment (15.11) (15.11)

Total net 68.00 68.00 68.00

Employer NIC @ 12.2% 12.20 6.43

Cost to Reed 112.20 91.32

Saving to Reed

ie £47.25 x 44.2% (tax plus NIC) 20.88

113.    The “Travel adjustment” (on the payslips described as “exp adj”) is the same as the difference between the total of the tax and NICs figures in the “No scheme” column, that is £32, and the total of the tax and NICs in the “With scheme” column, £16.89. As we have explained, its purpose was to reduce the net pay to the amount it would have been if the employed temp did not participate in the scheme. Although described in the handbook as an adjustment to gross pay it is more accurately an adjustment to net pay, because the adjustment to gross pay (and with it the tax and NICs liability) had already been made before this adjustment was applied.

114.    The example shows that Reed reaped the entire benefit of the tax and NICs saving. It was in order to ensure that the employed temp saw some advantage from participation in the scheme that the travel-to-work payment was added to his or her pay. It amounted, when the RTA scheme was first introduced, to £1 per day. As Mr Beal’s description rather obliquely indicates, that sum was at first accrued until it reached £50, when it was paid, after deduction of tax and NICs. From April 2001 the travel-to-work payment was altered, to £1.50 per day if the employed temp worked for more than 5 hours, or 75p if less than 5 hours were worked. The resulting amount was no longer accrued, but paid weekly, again after deduction of tax and NICs.

115.    The note of the meeting produced to us shows that the worked example did not eliminate Mr Paget’s concerns. It reads:

“After a brief discussion it was agreed that the following are the areas of risk:

• The lack of clarity of the calculations on the payslip and whether therefore Reed was complying with its obligations to the temps in respect of the format of payslips.

• Whether the deduction of the tax saving is legal

• Whether the Inland Revenue requirements are being adhered to …

• Whether if you take out the travel allowances [here meaning the amounts permitted by the dispensation] the National Minimum Wage requirements would be met.”

116.    There were several sample payslips within the documents produced to us. One typical of the pre-April 2001 system was described in some detail by Mr Beal. It shows on the left hand side that the employed temp, who appears to have had several assignments during the week, earned total gross pay of £523.26. From that sum was deducted an item identified as “PRP/EXP ADJ” of £50.60. At this time (March 1999) Reed was still operating its PRP scheme. Although the PRP and RTA schemes were distinct, no attempt was made, at least on the payslip, to segregate the portions of the £50.60 which were attributable to each of them. The purpose of the deduction, as Mr Beal also explained, was to bring the net pay back to what it would have been without participation in the scheme. In this case, the gross pay after the adjustment was £472.66. On the right hand side of the payslip appeared the income tax (£55.47) and employee’s NICs (£36.48) deductions, leaving net pay of £380.71. In a box at the foot of the payslip appear the words “This is what your payslip would have shown if you were not included in the PRP and expenses scheme this period”, followed by other figures leading to a final net sum of £380.71, the same as the amount actually paid. However, in another box was shown the aggregate of the travel-to-work payments accrued to date, in this case £9. There was no explanation on the payslip of the calculation of that amount, or even a statement of the amount which had accrued during the current week.

117.    At para 271 of his witness statement Mr Beal set out the result of that presentation of the RTA scheme on the payslips, as the employed temps were intended to perceive it:

“The payslip therefore enabled the Temporary Employee to see that he or she was better off as a result of participation in RTA as the travel allowance represents an extra payment which the employee would not otherwise have received. It is important to distinguish between the travel and subsistence allowances that Reed was entitled to pay free of tax and NICs under the Dispensation … and this £1 a day ‘travel allowance’ which was the mechanism for passing part of the benefit of the Dispensation to the Temporary Employees.”

118.    We observe at this point that, while the employed temp might have been able to see that participation in the scheme led to some increase in his or her net pay, it was not possible to discover from examination of the payslip how the adjustment had been determined, nor was any information provided to him or her, in the payslip, the handbook or otherwise, which would have revealed the amounts set out in the dispensation current at the time. When the payslips discussed above were produced, the first dispensation was in effect. It allowed Reed to pay travel expenses to those employed temps using public transport of £5.00 per day in central London, and £1.75 elsewhere, plus a daily subsistence allowance of £3.15 in London and £2.35 elsewhere. As Mr Beal’s explanation reveals, Reed actually offered £1 per day for the travel-to-work payment at this time, regardless of area or distance, and nothing for subsistence.

119.    We were also provided with an example of a typical post-April 2001 payslip, that is, one produced during the currency of the revised RTA scheme, and after the second dispensation had replaced the first. By this time the PRP scheme had come to an end. In this case the employed temp earned a gross amount of £354. To that were added holiday pay of £208.60, “travel allowance” of £6 and “expenses non-taxable” of £89.04, followed by the deduction of “exp adj” of £40, leaving £617.64. Tax and employee’s NICs deductions reduced the net sum payable to £518.49. As before, there is a box at the foot of the payslip in which comparative figures are provided. They show that if the employed temp had not participated in the scheme, the net pay would have been exactly the same amount. However, there is no longer any figure for the accrued travel-to-work payment and, if one can take the payslip at face value, the employed temp gained nothing at all from participation in the scheme. We were told that there might have been a computer problem at the time which resulted in the production of incorrect figures. Mr Beal explained that because of a programming error the travel-to-work payment was included in the hypothetical gross pay, whereas it should have been excluded with the result that the advantage should have been the after-tax amount of the travel-to work payment. As this payslip conflicts with every description of the operation of the scheme we are willing to accept that that may be the explanation.

120.    However, even those later payslips which showed that participation in the scheme conferred some benefit on the employed temp also showed that the benefit was very modest. A payslip from late 2001, after implementation of the revised RTA scheme (in which the payments were made immediately, rather than accrued), and when, it seems, the computer problem had been resolved, showed that the worker earned total gross pay of £455 which, after adjustments and deductions, resulted in net pay of £342.67. The comparative calculation indicated that the net pay, without participation in the scheme, would have been £341.58, a difference of £1.09.

121.    It is not altogether surprising that an employee help-line was necessary as the operation of the scheme was, even on Reed’s own case, difficult to understand. There is a revealing comment in an email sent by Miss Ollerenshaw to a colleague within RR on 22 August 2001, which goes even further: “… the current payslips are misleading to say the least.” Indeed, many of the participants were so confused, and in some cases so seriously concerned, about the impact on them of the schemes that they sought further help not only from Reed’s helpline staff, but also from HMRC. The number of enquiries made of HMRC by employed temps who could not understand their payslips was one of the factors which contributed to the misgivings they eventually harboured about the schemes.

122.    Douglas Hird was the HMRC officer in charge of a unit which, among other things, dealt with PAYE enquiries from employees. His evidence was that the unit received numerous telephone calls from Reed employees querying their payslips; many were worried that they were not paying enough tax or NICs. He gave an example of one Reed employed temp who had been receiving refunds of tax every week and who had paid NICs of only 82 pence on approximately £800 of earnings over an 8 week period. Mr Hird had found it impossible to work out how her net pay had been determined, even after a member of his staff had contacted Reed’s payroll department for further information. At para 32 of his witness statement he made the point that

“ … she appears to be no better off in terms of her net pay by being in the expenses scheme. She might in fact be at a disadvantage if she were to have subsequent periods of unemployment and no (or reduced) National Insurance Contributions in terms of acquiring entitlement to some statutory benefits. Some statutory entitlements such as SSP and SMP require … a certain level of NI contribution in order to qualify…. She might also lose out in that her entitlement to claim a tax refund in a later period of unemployment would be reduced.”

123.    Miss Ollerenshaw gave an example of an employed temp who suffered in exactly this manner, during the course of the correspondence about the payslips, which we set out at para 129 below.

124.    The number of enquiries, and Mr Hird’s inability to reconcile the figures shown on the payslips, led to a chain of correspondence and other communications, beginning in 2001 (shortly after responsibility within HMRC for Reed’s staff was transferred to Mr Hird’s office), between HMRC and Reed’s payroll staff; Mr Read and Miss Ollerenshaw were involved to some extent. Initially, Mr Hird’s requests prompted Reed to produce copies of letters sent by it to two employed temps in which, as he put it, the writer referred to “an ‘arrangement’ between Reed and the Inland Revenue that [was] claimed to enable Reed to pay the employee a proportion of gross pay free of tax as a ‘travel to work allowance’”. The letters went on to state that the net pay received by the employed temp was the same as it would have been without participation in the scheme, but that the employed temp would receive a “travel allowance”. Neither letter made any mention of the amounts permitted by the dispensations, nor did they offer any explanation of the manner in which the adjustment was calculated.

125.    This—in Mr Hird’s eyes rather unsatisfactory—response led him to refer the matter to Mr Read (who had granted the dispensation then in effect). Mr Read spoke to Miss Ollerenshaw in terms which evidently left her rather concerned. On 16 November 2001 she sent an email to a large number of RR colleagues which included this passage:

“[The Inland Revenue] have been inundated with calls from employees failing to understand their payslips.… One of the employee’s husband from Reed Health [sic] is a financial expert and thinks that deductions to buy shares are being used to pay travel and subsistence expenses. He thinks that large amounts of tax and NIC are dropping down a black hole. He referred to the fact that the payslip refers to an amount that would have been shown on the payslip if not in the expense scheme. He said that he would fax copies of the payslip to me and could I explain by the end of next week what is happening.

I have warned both Derek [Beal] and Malcolm [Paget] that the scheme could be revoked retrospectively (potential exposure could be c£10m which I did not mention), could be revoked from today or the least likely outcome would be that a severely toned down version of the scheme could be agreed for the future.”

126.    Mr Read and Mrs Kirkham arranged a meeting with RR (Miss Ollerenshaw) which took place on 30 November 2001. They were offered the explanation that the travel and subsistence payments were included in the hourly rate paid to the employed temp, and accordingly reflected in the gross pay for the week. The allowances were then deducted from the salary, in order to reduce the amount subject to income tax and NICs, and then added back as an amount which was payable without deduction of tax or NICs. The further adjustment, designed to reduce the amount actually paid (disregarding the £1.50 or 75p per day) to the amount which would have been received if the employed temp had not participated in the scheme, was mentioned. In fact, as we have said, it represented the aggregate saving in tax and employee NICs which resulted from the employed temp’s participation in the scheme.

127.    It is apparent from contemporaneous records, as well as their evidence, that Mr Read and Mrs Kirkham found the explanation they were given to be both surprising and somewhat baffling; our view is that they probably did not understand it. However, Reed emphasises the fact that, then and for some time thereafter, and despite their bafflement, HMRC did not say to Reed that the manner in which it was operating the schemes, or perhaps more accurately applying the dispensations, was incorrect or otherwise unacceptable. Instead, on this occasion, Mr Read asked that the payslips be laid out in a clearer fashion, in order to reduce the number of calls by employed temps to HMRC. The contemporaneous note of the meeting indicates that part of the blame for the lack of clarity was placed upon a computer programme which Reed had purchased but which did not do quite what was intended.

128.    We cannot say we are altogether satisfied by that explanation. Miss Ollerenshaw’s note of the meeting at which she had explained the calculations to Mr Paget in January 2001 (see para 115 above) includes what are in our view two telling passages. The first is that “SO showed MP a worked example of the current calculation … and explained which calculations are transparent and which are calculated by the computer and not shown on the payslip.” The second is a comment about “The lack of clarity of the calculations on the payslip and whether therefore Reed was complying with its obligations to the temps in respect of the format of payslips.” We will come to the conclusions we draw, from the manner in which the schemes were presented to the participants and to HMRC, and from the way the evidence was given, at para 172 below.

129.    In April 2002 there was a further meeting between HMRC and Miss Ollerenshaw, following which she wrote to Mr Read on 15 April 2002. Her letter included the following passage:

 “As discussed, the employees who state that they are worse off by having participated in the arrangement, as they would otherwise have been entitled to a higher tax refund are clearly mistaken. They would only be entitled to a higher refund if they had paid the tax in the first place.”

130.    We find that remark somewhat surprising, and perhaps an indication of what, as we are bound to say, we considered a blind spot on Miss Ollerenshaw’s part. It is perfectly true that the employees had not paid the tax; but they had suffered a deduction from their pay of the same amount as a consequence of the operation of the scheme (see the worked example at para 112 above, and the commentary which follows). An employed temp who had not participated in the scheme would, ordinarily, be entitled to a tax rebate following a period in which he or she had not worked; an employed temp who had participated and who was later not working would not receive a rebate, even though the latter would have received only the addition of the taxable travel-to-work payment, invariably less than the tax saving of which the benefit accrued exclusively to Reed. It did not seem to us that Miss Ollerenshaw recognised how poor a bargain the schemes were from the employed temp’s perspective. However, her letter seems to have satisfied HMRC for the time being, and there was a period of relatively little contact between RR and HMRC about the dispensations and the schemes.

131.    In April 2002, while the third dispensation was in operation, and in part motivated by its concerns about the RTA scheme, Reed replaced it by the Reed Travel Benefit, or RTB, scheme. The major effect of the change was that there was now to be a single adjustment to an employed temp’s pay, which (it was claimed) he or she would know in advance. Ms Ollerenshaw’s evidence was that following the introduction of the RTB scheme she “became more comfortable with the basis on which the Temporary Employees could agree to permanently forgo an element of their pay. They were aware of the formula on which the pay to be forgone was calculated, as they could see from the matrix [see para 137 below] what the deduction could be.” Those remarks, say HMRC, reveal that she had her own misgivings about the RTA, and was conscious of the shortcomings in its operation, and in Reed’s communication of that operation to the temporary employees and to HMRC.

132.    A leaflet was produced by Reed for use by their temp consultants when explaining the new scheme to existing and newly recruited employed temps. It consisted of a list of questions and answers about the scheme, outlining how it would work:

“How does the RTB differ from the Reed Travel Allowance Scheme?

In short, participating in the Reed Travel Allowance Scheme meant that an individual would benefit by receiving an additional 75p or £1.50 per day, depending on the number of hours they worked. The RTB however works differently, in that the benefit to each Temporary/Contractor will depend on their individual Tax and NI circumstances.

What do the Temporaries/Contractors need to do?

As before, there will be a box on the timesheet for the Temporary/Contractor to indicate if they travel to work by Public Transport as well as a box to complete the number of miles they travel to work if they use their own transport. However an additional box will now be included on the timesheets. This box will need to be completed by the Temporary/Contractor with the number of days in which their day covers a meal break.

What do the figures on the matrix mean?

This table shows the daily amount by which the Temporary/Contractor is agreeing for their gross pay to be reduced by in order that they can receive the net benefit of participating in the RTB.”

133.    The material provided to employed temps made it clear that they would ordinarily be included, but could opt out of the scheme if they wished, and could opt back in again, at any time. The only restrictions were that those who chose to claim relief themselves for their travel expenses could not participate, and that those who did not pay tax or NICs could derive no benefit from it (in truth, it was Reed which could derive no benefit) and would be excluded.

134.    A new timesheet was introduced, which included two additional boxes for the employed temps to complete: “I wish to claim subsistence allowance for [ ] days” and “My round trip each day averages [ ] whole miles”. Employed temps could also indicate that they had travelled to work by bicycle or on foot; in these cases they received no travel-to-work payment, and their gross earnings were paid without adjustment in accordance with the new scheme.

135.    The handbook, too, was revised. The February 2006 version included the following:

“As a Temporary Worker, Reed offers you the opportunity to increase your take-home pay through the Reed Travel Benefit (‘RTB’).

The travel benefit has been negotiated with HM Revenue and Customs on your behalf and provides you with a tax- and NI-free travel and subsistence allowance as part of your pay rate. This reduces your taxable and NI-able income and therefore increases your take-home pay.

Will I ever receive less pay by being in the RTB?

No. On your payslip each week, the Tax, National Insurance and Net Pay that you would have received had you not participated in the RTB will be shown. This will demonstrate that you do not receive less net pay through the RTB and in the majority of cases you will receive more.

How do I claim my Travel Benefit?

We require you to complete your Timesheet with the information listed below, which enables Reed to calculate your Tax and National Insurance free expense value.

To allow Reed to apply the RTB, you will need to make a salary sacrifice reduction to your gross pay. The amount of this reduction will depend on your Tax and National Insurance position.”

136.    The “matrix” referred to by Miss Ollerenshaw and in the leaflet given to the temps consultants was further explained by internal guidance provided for Reed’s payroll department. The guidance explained the working of the scheme in this way:

“(1) The Temporary Employee’s gross weekly pay is calculated on the basis of the number of hours worked and their agreed hourly rate as if they were not in the scheme. This figure is used to determine which tax/national insurance rate they would pay, and therefore determine their tax bracket on the RTB matrix.

(2) Using the tax bracket, and information from the timesheet to ascertain a) whether the Temporary Employee travelled by public or private transport, and b) whether the booking branch was in ‘inner London’ or elsewhere, the daily ‘sacrifice’ is worked out from the RTB matrix. This is then multiplied by the number of days worked, and is the ‘RTB Adj figure’ (i.e. the salary sacrifice);

(3) The RTB Adj is deducted from the gross pay. This gives the Total payments;

(4) The subsistence and travel expenses which Reed are ‘reimbursing’ the Temporary Employees for the week is calculated from the information on the payslip. Subsistence is payable where a Temporary Employee has worked at least 4 hours in one day. For each working day public transport is a flat rate expense (either London or elsewhere); private mileage expenses are calculated on the number of miles recorded by the Temporary Employee on their timesheet. The total figure is the ‘RTB Expenses TP’;

(5) The RTB Expenses TP is deducted from the Total Payments to give the taxable pay. The tax and NIC due on the taxable pay is calculated.”

137.    The matrix was provided to the participants, as an annex to a circular letter of 22 March 2002, announcing the replacement of the RTA by the RTB scheme. It was divided into lettered columns and numbered rows, the columns reflecting various possible combinations of tax and NICs liability, the rows the different travel bands—for those working in Inner London, for those using public transport outside London, and for those using their cars. It became possible for an employed temp to determine the gross deduction from his or her pay which would be made. An example was given of an employed temp paying standard-rate tax of 22% and standard NICs of 10%, travelling by public transport outside London, who would fall in box E2 of the matrix and suffer a daily gross deduction of £1.49. An employed temp earning at a steady rate would be able to see in advance by this means what the deduction would be; another, earning at a fluctuating rate, or returning to work after an interval without work, would almost certainly not be able to do so. It is clear from an examination of the matrix (and would have been clear to any employed temp who took the trouble to examine it) that the deductions were entirely driven by the employed temp’s tax and NICs liability. Indeed, the fact that those who did not pay tax or NICs were excluded should have made it clear that the scheme was primarily a device for saving tax and NICs, and not one whose essential purpose was the payment of expenses in a tax-efficient manner.

138.    We did not have a “worked example” of the RTB scheme in the form set out at para 112 above, but did have some sample payslips. One was for a worker who fell within box E2, and it showed gross pay for the week of £225, from which a deduction, described as “RTB ADJ”, of £7.45 (5 days at £1.49) was made, leaving “Total payments” of £217.55. That figure was also recorded as “Gross pay to date”—the payslip assumed for simplicity that it was the first week of the tax year. The tax and NICs deductions were shown as £13.48 and £8.13 respectively. They were deducted from the “Total payments” to arrive at net pay of £195.94. As before, there was a box in which was shown what the net payment would have been without participation in the scheme: £225 less tax of £25.58 and NICs of £13.60, leaving £185.82. If this sample was typical, the employed temp derived significantly more from the RTB scheme than from its predecessor. Mr Beal’s evidence was that, overall, Reed would take 40% and the employed temp 60% of the benefit. We did not discover whether this ratio was achieved in practice. It appeared that Reed took all of the benefit of the reduction in employer’s NICs.

139.    The sample payslip also shows that the “Taxable pay to date” was £170.30, and that the “RTB NON TAXABLE EXP TP” was £47.25. That was explained in the circular letter in this way:

“The net value of the RTB plan depends on which travel and subsistence rates apply to you and on your individual tax position.

The benefit to you comes from the Tax and National Insurance savings that are made because your taxable income is reduced by these tax free amounts. This will be shown on your payslip as RTB NON TAXABLE EXP TP. Reed can confirm categorically that you will not become liable for these Tax and National Insurance savings. If you do not pay Tax and National Insurance, there will be no benefit.”

140.    Although the fact of HMRC’s agreement to the making of certain payments free of tax and NICs was mentioned, the amounts set out in the dispensation were not disclosed, and in our view the clear impression given by the letter was that they were the amounts set out on the matrix. How, therefore, Reed arrived at taxable pay of £170.30 from gross earnings of £225 would not have been apparent to the recipient of the payslip. It can, however, be seen, once one understands the scheme, that it is £225 less the RTB NON TAXABLE EXP TP of £47.25 and the RTB ADJ of £7.45.

141.    The internal guidance was not made available to the employed temps (or to HMRC, at the time). As can be seen from the foregoing description, the amount of part of the “sacrifice”, if that is what it was, was not a fixed daily or weekly sum, but one which varied according to the temporary employee’s tax and NICs liability. It is true that the liability would, as a general rule, vary in line with earnings, but there might be cases of temporary employees with particularly high or low tax allowances, for quite separate reasons, whose “sacrifice” was more or less than that of another temporary employee earning the same amount and entitled, or supposedly entitled, to the same dispensation allowances.

142.    Miss Ollerenshaw’s claim that the temporary employees could work out for themselves in advance what the adjustment would be (see para 131 above) is correct, but with some limitations or exceptions. An employed temp with an unchanging tax code, earning the same amount and becoming entitled to the same RTB allowances (the travel-to-work payment) each week, would be able to work out, from the matrix, what to expect by way of deduction. It would in our view be much more difficult, if not impossible, for an employed temp with varying earnings or a varying entitlement to RTB allowances (or whose tax code had changed) to work out what the deduction might be from the information Reed provided. In short, prediction was possible only for those certain of their position within the matrix.

The revocation of the fifth dispensation

143.    On 19 July 2004 Mr Read telephoned Miss Ollerenshaw to inform her that HMRC were looking into the RTB scheme. He requested copies of Reed’s current contracts of employment and confirmation that Reed did not engage either agency workers or employed temps on fixed term contracts. Miss Ollerenshaw supplied a sample of a Reed Health contract, making at the same time various observations about Reed’s relations with its temporary employees which were no more than repetition of what she had said before. Further exchanges revealed that HMRC’s principal concern now was, as Mr Read’s call suggested, whether Reed engaged workers on fixed-term contracts; the sample contract did not satisfy them and it was for that reason that Mr Read asked for copies of all the current contracts. Miss Ollerenshaw provided them in November.

144.    There was then a further period of silence until April 2005, when a meeting between Reed, RR and HMRC took place. Its principal purpose was to conclude the employer compliance review which had been started as long ago as 2003. Although a number of relatively minor errors had been identified, and were dealt with at the meeting, Miss Ollerenshaw’s recollection was that nothing was said about the dispensation or the RTB scheme, and none of the corrections required as a result of the review had any relevance to them. However, despite the fact that some identified errors were dealt with, HMRC did not consider the review was at an end.

145.    In June 2005 Reed, through RR, asked HMRC to agree to a further increase in the scale payments set out in the fifth dispensation. HMRC made it clear they were not willing to consider any change in the dispensation until the employer compliance review had been completed to their satisfaction. In fact, by this time they entertained considerable disquiet about the manner in which Reed was applying the dispensation, and about its appropriateness. Nevertheless, it was not until 3 November 2005 that Mrs Austin wrote to Reed, rather than RR, expressing her concern about the employment status of those to whom the current dispensation was applied, a concern prompted by what, as her letter showed, she thought was a discrepancy between what had been said to Mr Read in January 2004 and what Miss Ollerenshaw had said to one of her colleagues at a meeting which had taken place in September 2005. It seems that she may, in fact, merely have misinterpreted Miss Ollerenshaw’s continuing wish to have the dispensation extended to agency workers, but she was by this stage anxious to enquire further into the matter and she asked yet again for copies of Reed’s current contracts with all of its employees, temporary or permanent, and with its agency workers. Her letter went on to say, in relation to the current dispensation,

“Having carefully considered the implications of each type of worker, and their hourly rate, please will you confirm to whom Reed have applied the Dispensation to/for in relation to each category of worker; and that the payments are only made on occasions when qualifying journeys have actually been made. In addition, could the company confirm that subsistence is only paid/reimbursed for days where for [sic] 5 hours or more is actually worked (excluding any travelling time), and whether the company has controls in place to verify that any expenditure has actually been incurred? Several queries have arisen which would suggest that certain individuals sacrifice a proportion of their salary which in theory reduces their hourly rate by associated travel and subsistence payments. I refer you to EIM42774 Salary Sacrifice. The guidance can be found on the Inland Revenue internet website. I consider the company to have a duty of care to any worker to consider carefully the effect, or the potential effect, that a reduction in their pay may have on:-

• Their future right to the original (higher cash salary).

• Any pension scheme being contributed to.

• Entitlement to Working Tax Credit, or Child Tax Credit.

• Entitlement to State Pension or other benefits such as Statutory Maternity Pay.

I would also ask the company to comment on how this salary sacrifice scheme interacts with the national minimum wage legislation, as it appears the salary sacrifice scheme may reduce the cash pay to below the national minimum wage. The national minimum wage provides a legally binding minimum hourly rate of pay to workers aged 18 years or over, with few exceptions. It is with this in mind that the Reed Travel Benefit Plan, which utilises the rates as per Dispensation, is called into question.

You may wish to research this issue and comment accordingly.

The current scale rate payments within the Dispensation appear to be excessive in view of current technical advice, and the scale rate payments for subsistence appear to be generous having previously stated that expenditure needs to be ‘actually incurred’. An important point to bear in mind is that these rates should be a reasonable reflection of the amounts that your employees actually spend on qualifying expenses.”

146.    Miss Ollerenshaw’s reply, dated 9 December 2005, addressed these points in some detail. She emphasised the fact that she had had annual meetings with Mr Read while he was the responsible inspector, at which they had discussed the circumstances in which the allowances were payable, their rates, and (although Mrs Austin had not mentioned them) the layout of the payslips. She challenged Mrs Austin’s belief that the rates were excessive, pointing out that they were lower than those payable to civil servants (a contention which HMRC does not accept, though it is unimportant for present purposes), and also pointed out that the adverse effect on participants in respect of state and similar benefits was made clear to them in the explanatory material. She added that “The current arrangements were agreed to reduce both my client’s and HM Revenue & Customs’ time spent on administering hundreds of thousands of expense claims”, a comment which, in the light of RR’s view that very few temporary employees would trouble to make their own claims, can only be regarded as disingenuous.

147.    We should also add that although a comment about the loss of certain contributory benefits appeared in the draft guidance supplied to HMRC (see para 104 above) we were unable to find any equivalent information in the material actually provided to the temporary employees.

148.    Miss Ollerenshaw’s reply did not satisfy HMRC and, as we have said, the fifth dispensation was revoked in March 2006, the revocation to take effect from 6 April. No reasons were given at that time, despite Reed’s request for an explanation and despite, Mr Beal said, an intimation that the dispensation might continue albeit with lower scale rates. Internal HMRC emails show, however, that those involved in the decision were of the view that Reed’s contracts with its employed temps did not provide for continuity between assignments but only for employment for the duration of each assignment (thus the workplaces were not “temporary”), that the allowances were excessive in amount and that Reed was not, in fact, making the scale payments at all.

149.    During 2005, Reed Health was advised, not by RR, but by BDO Stoy Hayward LLP (“BDO”), although there was evidently close liaison between BDO and RR. On 3 August 2005 BDO wrote to HMRC to explain Reed Health’s operation of the RTB scheme, in response to an earlier request from HMRC: the content of the letter indicates that the focus of the enquiry was VAT rather than income tax or NICs. HMRC nevertheless place some emphasis on the following comments, in the context of these appeals. After some opening remarks of no present relevance the writer provided answers to questions which had been raised:

“1. Reed Health Group does not charge an administrative fee in respect of the RTB.

2. Employees participating in the RTB scheme sacrifice a proportion of their gross salary. A tax-free element in respect of travel and subsistence is then added to the net salary figure.

3. Employees have the option before commencing employment of whether they wish to take advantage of the RTB saving. An employee who decides not to take advantage of the scheme will receive a higher gross salary than an employee who accepts employment on the RTB terms, but the RTB employee will receive a higher net salary due to the tax and NI saving.

4. The salary sacrifice is greater than the tax free element added back to the employees net pay. This is not brought to the attention of the employee by Reed Health Group.”

150.     We do not know what was HMRC’s immediate reaction to that information, though it cannot have done much to allay their concerns.

151.    Reed ceased to operate the RTB scheme on 5 April 2006, when the revocation of the fifth dispensation took effect, so far as it related to the RTB scheme. Mr Beal’s evidence was that the ending of the scheme caused Reed some commercial damage, particularly because it could not explain adequately to its temporary employees why the scheme had been terminated when it did not itself understand HMRC’s reasons, but that it was not prepared to continue paying the allowances without the assurance which the dispensation had given it that the payments were not subject to tax and NICs.

152.    The dialogue between Reed, RR and HMRC continued. Other parts of the dispensation survived (as we have mentioned, the RTA and RTB schemes represented only part of what each dispensation covered) and Reed remained hopeful that the RTA and RTB schemes could be restored to it although, as we shall shortly explain, a rather different scheme was introduced instead. Relations between Reed and HMRC were evidently somewhat strained since by this time it was apparent that HMRC might seek to recover tax and NICs for which Reed had not accounted in the past as, of course, they did, although not until February 2007. It is plain from further email exchanges that there was also some disquiet within HMRC about the history of the dispensations, and the manner in which HMRC had approached them.

153.    Reed places particular reliance on an internal email of 20 October 2006 whose author, David Stephens of the Central Policy Unit, set out the basic facts as he understood them, and then remarked that

“It looks to me like we have cocked-up here. Reed applied for a dispensation and contended that there was an overriding contract of employment. We met with Reed’s tax advisers to discuss the position and raised our concerns as to the employment status of the workers concerned. Inexplicably, we did not ask to see the written contract.

It seems to me that there is at the very least an arguable case to be made by Reed that we gave representations (a ruling) to the effect that we too considered that the workers were employees: (what other construction can be put on our agreement to grant the dispensation?!). Under administrative law Reed could have a viable claim against us if, having put all their cards face upwards on the table, they acted on our ruling.

NB. Employment status is one of the 5 categories covered by COP 10 in which we will give guidance and will be bound by it (even it turns out later to have been wrong) where all the relevant facts were provided in the sense that the taxpayer put his cards face upwards on the table.

We may think it necessary to consult lawyers but I think that Reed may well have a strong case under administrative law that they were entitled to rely on our representation (ruling) that the workers were employees under an overriding contract of employment.”

154.    We will return to this email and its significance when we deal with our findings of fact relevant to Reed’s claimed legitimate expectation (see para 294 below).

The replacement scheme

155.    On 5 April 2006, as we have said, Reed was forced to cease using the RTB scheme, but it was reluctant to abandon what it perceived as a beneficial arrangement, and tried to identify a replacement. In June 2006, Mr Beal said, Reed received an approach from the accountants KPMG, who had successfully negotiated with HMRC, on behalf of a rival of Reed, for the implementation of a similar scheme. KPMG explained that if allowances for travel and subsistence were to be payable tax-free, the employment business needed to use a contract of employment of an overarching nature, one which guaranteed a minimum number of hours of work in a 12-month period. The minimum number which appeared to be acceptable to HMRC was 336, equivalent to seven hours a week for 48 weeks each year. Mr Beal said he could not understand why such a provision should make any difference to the tax treatment of allowances such as Reed had been paying, but it became clear that HMRC would not agree to a new dispensation without such a provision and he reluctantly agreed to the amendment of Reed’s contracts in that way. He said that the amendment was on a “without prejudice” basis, meaning that it was done in order to secure HMRC’s agreement to the reinstatement of the RTB scheme, but without any concession on Reed’s part that it was a necessary condition for the implementation of the scheme.

156.    A replacement scheme was accordingly introduced, once Reed had amended its contracts with employed temps to include the 336-hour guarantee. We are not concerned in this appeal with the question whether the replacement scheme achieves Reed’s purposes (HMRC say it does not) though a brief description of it, and of its differences from and similarities to its predecessors appears at para 169 below.

Reed’s contracts with its employed temps

157.    The form of contract Reed introduced in 1995, when the decision was taken that the majority of its workforce should become its employees, provided that the employment was to begin on the later of 3 July 1995 and the start of the employed temp’s next (or first) assignment. The choice of dates was necessary to cope with the situation of those who were currently on assignment when the new contract was introduced, and to avoid a change of status part-way through an assignment. The example of the conditions in use before April 1999 produced to us contained these clauses:

“1. The Temporary Employee’s employment and continuous employment begins on the date of the commencement of the current assignment.

2. Reed will endeavour to find the Temporary Employee the opportunity to work in the capacity specified on the Temporary Employee’s copy of the time sheet where there is a suitable assignment with a Client for the supply of such work. Reed reserves the right to offer any assignment to such temporary employees as it may elect where that assignment is. suitable for several workers.

3. The duration of the Temporary Employee’s employment will be for so long as Reed offers work to the Temporary Employee. It is anticipated that this will be for the duration of the assignment with the Client provided that the Temporary Employee satisfies the Client’s requirements. Reed may instruct the Temporary Employee to end the assignment at any time without specifying reasons.”

158.    The conditions went on to provided that wages, “a proportion of which may be Profit Related Pay and Travel Expenses”, were payable only in respect of the hours worked. Reed was obliged to endeavour to find the employed temp work, but it could elect to which of its employed temps it offered any particular assignment. The employed temp was under no obligation to accept any particular assignment Reed offered. Reed and the employed temp were each obliged to give the other notice “in accordance with statutory requirements”. The relevant statutory minimum notice periods applicable to all employees therefore applied, but the contract provided in addition that there was no obligation on Reed to provide work (or, since the employed temp was entitled to wages only for hours worked, any pay) during such notice periods. There was at this time no entitlement to holiday pay, but statutory sick pay was provided for.

159.    At this time, the statutory notice requirements were prescribed by the Employment Rights Act 1996, s 86, re-enacting without further amendment the (several times amended) s 49 of the Employment Protection (Consolidation) Act 1978. Section 86(5) excluded from the notice provisions “a contract made in contemplation of the performance of a specific task which is not expected to last for more than three months unless the employee has been continuously employed for a period of more than three months.” That subsection was repealed, without replacement, on 1 October 2002 by the Fixed Term Employees (Prevention of Less Favourable Treatment) Regulations 2002.

160.    In October 1998 a new contract was introduced, in order to effect the changes required by the Working Time Regulations 1998. In particular, the employed temp now had a right to paid holidays, an entitlement which accrued over the holiday year. However, the Staff Handbook at this time stated that in order to qualify for holiday pay

“You must work for 13 weeks (the ‘qualifying period’) without a ‘break’ in your relationship with Reed. For example, if you don’t work for Reed for 2 weeks or more this would count as a break. If you return to work for Reed at a later date you would have to start your qualifying period again.”

161.    Reed could not make a payment in lieu of holiday to the employed temp when any particular assignment ended since, unless a contract of employment is terminated altogether, it is unlawful to replace a right to paid holidays with a payment in lieu (reg 13(9)(b) of the 1998 Regulations). Employed temps were required to give 2 weeks’ notice of an intended period of leave. In all other respects the contract was materially the same as its predecessor.

162.    The nature of the contract was one of the topics of discussion at a meeting between Mr Rayer, Mr Beal and Ms Sims on 4 November 1998 (while negotiations with HMRC for the grant of the first dispensation were in progress). The note of the meeting records that

“DB pointed out that all temporary workers work a 13 week contract of employment, although they only get paid for the hours they work. The balance of the time is effectively a zero contract arrangement. The main reason for this being introduced was at the instigation of the Inland Revenue some years ago. By treating temporary workers as in employment for 13 weeks, there is no need to issue a form P45 every time a temporary assignment ceases; a P45 is issued only if the employee is not in work at the end of a 13 week period.”

163.    As we understood the evidence, there was little significant change to Reed’s contracts with its employed temps between the introduction of the RTA scheme in 1998, and 2004, save for the variation dictated by the move from the RTA to the replacement RTB scheme, which we described at para 131 above. We had, and should record, some evidence about Reed’s own perception of their character. On 24 January 2001 Miss Ollerenshaw wrote to Mr Read, responding to his concern about the employment status of those participating in the scheme as it then was, “You are correct in your understanding that ‘employed temps’ are engaged under contracts of employment but they are not full time contracts and only apply when the employed temps are carrying out assignments on behalf of Reed.” That was not merely Miss Ollerenshaw’s understanding: as we have mentioned, the minutes of a meeting between Mr Beal and various RR employees, including Miss Ollerenshaw, on 8 October 2001 record that Mr Beal “advised that the permanent contracts [that is, those with its employed temps] are from assignment to assignment. There is no umbrella contract and no mutuality of interest.”

164.    There was a significant change to Reed’s conditions of employment in April 2004, it appears as a result, in part, of the coming into force of the Conduct of Employment Agencies and Employment Businesses Regulations 2003. It introduced two forms of engagement, on assignment and on secondment. Relevant conditions included the following:

“3. The Temporary Employee’s employment and continuous employment begins on the date of the commencement of the current assignment or secondment.

4. Reed will endeavour to find the Temporary Employee the opportunity to work in the capacity as agreed at registration and specified on the Temporary Employee’s copy of the time sheet where suitable work with a Client is available. Where the Temporary Employee is offered work with a Client, his/her copy of the time sheet will indicate whether this will be on an ASSIGNMENT or a SECONDMENT basis.

5. If the assignment basis applies, the Temporary’ Employee’s services will be supplied to the Client for the duration of the assignment and the common terms of this agreement (paragraphs 1 to 22 inclusive) will apply together with the assignment only terms (paragraphs 23 and 24). If the secondment basis applies Reed will second the Temporary Employee to work under the Client’s direction and control for the duration of the secondment and the common terms of this agreement will apply together with the secondment only terms (paragraphs 25 and 26).…

7. The duration of the Temporary Employee’s employment will be for the duration or likely duration of the assignment or secondment with the Client as notified prior to the commencement of the assignment or secondment provided that the Temporary Employee satisfies the Client’s requirements. Reed may instruct the Temporary Employee to end the assignment or secondment at any time without specifying reasons.…”

165.    The essential difference between the assignment terms in paras 23 and 24 and the secondment terms in paras 25 and 26 was that in the former case, it was Reed which was responsible for paying the employed temp’s salary, whereas in the latter it was the client’s responsibility, albeit Reed itself which undertook the calculations. The contract was changed again in October 2004. The only amendment of significance on that occasion was that the contract was expressed to begin at the start of the temporary employee’s first (rather than, as hitherto, current) assignment or secondment. Despite the different arrangements for the payment of salary to those on secondment and those on assignment, it was not suggested to us that there was any material difference relevant to the RTB scheme then being used.

166.    Mr Beal made the point, which we accept has some force, that constant changes in the law relating to employment dictated frequent revisions of Reed’s contracts, and that their evolution was driven mainly by employment rather than tax considerations. His view was that, put to the test in an Employment Tribunal, the existing contracts would be interpreted as providing for continuity of employment across assignments, notwithstanding gaps between them, and despite the wording used in them which stated that the employment lasted only until the end of the current assignment. That wording was retained, he said, in order to give Reed an argument it might deploy if necessary.

167.    However, by mid-2004 Reed recognised that a provision which might assist it in an Employment Tribunal was putting its dispensation in jeopardy, because of HMRC’s increasing focus on the question whether there was continuity of employment. He learnt from Miss Ollerenshaw that Reed Health (at the time still a separate company, although it too was advised by RR) had recently amended its contracts (this amendment may have coincided with Reed Health’s decision not to engage agency workers in future, though that is not certain) and that the wording used in the amended contracts was clearer than the wording the other Reed companies were then using. It is clear from his first witness statement that by this time Mr Beal was concerned about HMRC’s enquiries into the nature of Reed’s contracts, and he agreed that the wording should be changed even though, he said, he did not think the change would make any practical difference to Reed’s legal relationship with its employed temps. Against that background it is rather surprising that he accepted Miss Ollerenshaw’s suggestion that HMRC be sent a copy of Reed Health’s contract, rather than Reed’s contract, following its amendment. When the amendment was made, it removed the statement that the term of the contract was “for the duration or likely duration of the assignment or secondment”, but did not replace it with anything else—in other words, although the contract identified its start date, it made no provision for expiry.

168.    We should add that Mr Beal’s evidence was that none of the Reed companies, including Reed Health (of which he was to become a director later that year), had ever engaged either temporary employees or agency workers on fixed-term contracts, by which he meant contracts which lasted for a set period. When Miss Ollerenshaw sent a copy of the Reed Health contract to HMRC, she included in her letter a statement that fixed-term contracts were not used. It is apparent that she too was using “fixed term” in the same sense as Mr Beal, since by this time she was well aware of HMRC’s view that each assignment represented a separate engagement, and of the consequences for Reed should that view prevail.

169.    After the revocation of the fifth dispensation and the discussions which followed the contracts were changed again, as we have said, in a manner which, as Reed understood, made a fundamental difference to HMRC’s perception of the arrangements; the revised terms came into effect in July 2006. The new form of agreement introduced the guarantee of a minimum of 336 hours’ paid work per complete 12 month period to which we have already referred. However, many of the remaining provisions of the earlier contracts were materially unchanged. There was still no guaranteed duration for any assignment; an assignment could be terminated by Reed immediately and without notice; and the employed temp was paid only for the hours worked on the assignment. As before, the contract began at the start of the employed temp’s first assignment, and continued until terminated by either party giving notice (that is, there was still silence about expiry). The notice periods were again the minimum periods required by statute, plus an extra week (to take advantage of an Employment Appeal Tribunal decision which has otherwise no relevance to the present case). Reed was, as hitherto, required to endeavour to allocate suitable assignments to the employed temp (though with no obligation to offer an assignment to either one of two suitable employed temps); in so far as it was unable to do so it guaranteed to pay for a minimum of 336 hours work in each 12 month period, paid at least at the amount of the National Minimum Wage. Now, in the absence of good cause, the employed temp was required to accept offers of suitable assignments (as the quid pro quo for the guarantee of the minimum amount of pay).

HMRC’s understanding of the operation of the schemes

170.    As we have mentioned, HMRC did not pursue the request for copies of Reed’s contracts with its employed temps and agency workers until July 2004, in the context of a re-examination of the RTB scheme. We are bound to say we found it remarkable that HMRC, despite their concerns about the employment status of those participating in the schemes, left it so long before repeating a request which had first been made several years before, and when they did raise it confined the enquiry to whether Reed engaged agency workers or employed temps on fixed-term contracts. It may well be this concern which deflected HMRC also from enquiring about what, as is now clear, is a critical issue, namely whether the contracts extended over multiple assignments, or there was a separate engagement for each assignment, although it is apparent to us that HMRC did not entirely realise the significance of the point until about 2004.

171.    We have dealt above with the considerable concerns HMRC harboured about the layout of the payslips, the confusion they engendered in the employed temps, and the burden of the enquiries directed at HMRC officers. As the evidence relating to these issues showed, several HMRC officers saw the payslips, had them explained and were also provided with details of the information given by Reed to the employed temps. Again, it is surprising, at least at first sight, that HMRC did not fully understand until quite a late stage, it seems at some point in 2005, precisely how it was that Reed was applying the dispensations. If they did not feel confident earlier of their understanding, they could have demanded a more detailed explanation. Mr Read made the comment in his second statement that he could “not understand how a salary sacrifice could be geared to an employee’s tax and NIC rates rather than to the salary itself”. As it is, the officers’ concerns seem to have been about peripheral matters rather than the fundamental issue of whether the dispensations could properly be applied in the manner in which Reed was doing so. In due course, HMRC recognised themselves that their approach might have been better (see the email set out at para 153 above).

172.    What we think it appropriate to observe at this stage is that the evidence made it clear to us that Reed, and RR on its behalf, were throughout at pains to say as little as they could to HMRC of the manner in which Reed was applying the dispensations. It was apparent from the correspondence, notes of meetings and their evidence before us that Miss Ollerenshaw, in particular, and to a lesser extent Mr Beal and Mr Rayer all seemed to find it difficult to speak of the schemes in a way which treated them candidly for what they were: a device by which Reed exploited the potential for its employed temps to obtain tax relief for their travelling and subsistence expenses, not in order to enhance their earnings, but for its own benefit. As we have said, the uplift to the employed temps’ earnings achieved by the RTA scheme was, at best, modest; and, even leaving aside the possibility that (if Reed is right) the employed temps could have claimed the relief themselves, receiving in some cases significantly more than the travel allowance Reed in fact paid, while others were potentially worse off. We should, however, repeat the observation we have already made that the RTB scheme appeared to confer rather more of the benefit on the employed temps.

173.    Before coming to the law and the issues we need to record some further findings of fact. Despite the preceding comments, we are satisfied that Reed’s statement of the cases and circumstances (including the information given in the course of meetings and in the Employee Guide) which led to the grant of the first dispensation was given in good faith and contained all the facts that it and RR considered relevant at the time. In particular, Reed informed the inspector that 16,000 persons were employed at any time, that they were all employees and not agency workers, and that (factually) they had no permanent workplace but were required to attend various locations for a limited period (not exceeding 24 months). We do not consider that this is to be read as a representation that the workplaces were temporary workplaces in law. That is not a matter to be included in a statement of the cases and circumstances in which payments are made or benefits provided, namely the facts, but an application of the law to the facts which is for the inspector to make in order to be satisfied that a dispensation must be granted.

174.    We have concluded that neither Reed nor any of the inspectors concerned appreciated, at least until 2004, that the distinction between a job-by-job employment and a continuing employment was relevant. Accordingly, neither considered that the terms of the contracts with the employed temps were relevant so long as the contracts were employment contracts and not contracts with agency workers, which they both understood (correctly) that they were. Although, as we have said, it is surprising that HMRC did not ask for copies of the contracts sooner than they did, we accept it is unlikely that before 2004 they would have considered them in order to determine the duration of the engagement.

175.    The inspectors who granted the dispensations must accordingly be taken to have understood that the applications were made in respect of a large number of employees, that they were all employees, and that the nature of their duties was that (factually) they had no permanent workplace but attended various locations for a limited period (not exceeding 24 months). In relation to each of the dispensations the inspectors were satisfied that no additional tax was payable, though “on the basis of what you have told me.” These facts included the information given during the application for the first dispensation as well in as correspondence and meetings relating to later applications.

176.    It is a necessary inference from the fact that the dispensations were granted that the inspectors decided (or assumed without considering the matter) that the facts stated resulted in a potential liability to tax under the listed provisions (Chapter 3) but that the employed temps had temporary workplaces as a matter of law with the result that the disputed payments were deductible, and thus no additional tax was payable in respect of them.

The relevant legislation

177.    It is, we think, convenient to set out all the relevant provisions of the legislation with some comments about their significance in the context of this appeal before embarking on an examination of the issues and the parties’ submissions in relation to them. For the reasons given at para 41 above, it is not necessary to deal with the NICs legislation, and we accordingly confine ourselves to the tax provisions.

178.    Part 3 of ITEPA (which is one of its “income Parts”) is divided into several Chapters, of which only the first three are relevant in this case. Chapter 1, which consists only of s 62, defines “earnings”; Chapter 2, entitled “Taxable Benefits: The Benefits Code”, includes ss 63 to 69; and Chapter 3, entitled “Taxable Benefits: Expenses Payments”, comprises ss 70 to 72.

179.    Section 62 reads as follows:

“(1) This section explains what is meant by ‘earnings’ in the employment income Parts.

(2) In those Parts ‘earnings’, in relation to an employment, means—

(a) any salary, wages or fee,

(b) any gratuity or other profit or incidental benefit of any kind obtained by the employee if it is money or money’s worth, or

(c) anything else that constitutes an emolument of the employment.

(3) For the purposes of subsection (2) ‘money’s worth’ means something that is—

(a) of direct monetary value to the employee, or

(b) capable of being converted into money or something of direct monetary value to the employee.

(4) Subsection (1) does not affect the operation of statutory provisions that provide for amounts to be treated as earnings….”

180.    This section encapsulates the dispute between the parties in its simplest form: HMRC say that everything the employed temps received constituted either salary or wages within s 62(2)(a), or emoluments within s 62(2)(c), and, if so, that is determinative of the matter: there is no need to consider anything else. Reed argues that the disputed allowances cannot be “earnings” since they fall squarely within Chapter 3 as “expenses payments” and that, if so, the combined effect of Chapters 2 and 3 is to exclude them from liability for tax and NICs deductions.

181.    We do not need to set out ss 63 and 64. Section 63, as is uncontroversial, brings within the benefits code expenses payments falling within Chapter 3; s 64 contains some miscellaneous provisions designed, for example, to exclude double counting. Section 65 is entitled “Dispensations relating to benefits within provisions not applicable to lower-paid employment”. It is, as we have said, the section pursuant to which the dispensations in this case were granted (or purportedly granted) and pursuant to which the fifth was revoked, and it provides as follows:

“(l) This section applies for the purposes of the listed provisions where a person (‘P’) supplies the Inland Revenue with a statement of the cases and circumstances in which—

(a) payments of a particular character are made to or for any employees, or

(b) benefits or facilities of a particular kind are provided for any employees, whether they are employees of P or some other person.

(2) The ‘listed provisions’ are the provisions listed in section 216(4) (provisions of the benefits code which do not apply to lower-paid employments).

(3) If the Inland Revenue are satisfied that no additional tax is payable by virtue of the listed provisions by reference to the payments, benefits or facilities mentioned in the statement, they must give P a dispensation under this section.

(4) A ‘dispensation’ is a notice stating that the Inland Revenue agree that no additional tax is payable by virtue of the listed provisions by reference to the payments, benefits or facilities mentioned in the statement supplied by P.

(5) If a dispensation is given under this section, nothing in the listed provisions applies to the payments, or the provision of the benefits or facilities, covered by the dispensation or otherwise has the effect of imposing any additional liability to tax in respect of them.

(6) If in their opinion there is reason to do so, the Inland Revenue may revoke a dispensation by giving a further notice to P.

(7) That notice may revoke the dispensation from—

(a) the date when the dispensation was given, or

(b) a later date specified in the notice.

(8) If the notice revokes the dispensation from the date when the dispensation was given—

(a) any liability to tax that would have arisen if the dispensation had never been given is to be treated as having arisen, and

(b) P and the employees in question must make all the returns which they would have had to make if the dispensation had never been given.

(9) If the notice revokes the dispensation from a later date—

(a) any liability to tax that would have arisen if the dispensation had ceased to have effect on that date is to be treated as having arisen, and

(b) P and the employees in question must make all the returns which they would have had to make if the dispensation had ceased to have effect on that date.”

182.    The provisions listed in section 216(4) include payments which fall within ITEPA Part 3, Chapter 3. “Lower paid employments” are those in which the gross annual emoluments do not exceed £8,500, a limit which has been unchanged for many years. We understand that few of Reed’s temporary employees earned less than that amount. That, however, is by the way (even though HMRC believe that the allowances may have been paid to low paid employees, to whom the provisions of the benefit code do not apply by virtue of section 216 ITEPA, a possibility we leave out of account for present purposes). The purpose of s 216 is to exclude various benefits from tax if they are paid to those in lower paid employment; s 65 extends the exclusion to payments covered by a dispensation granted in accordance with its terms, whether or not the recipient is in lower paid employment.

183.    It will be observed that s 65 enables HMRC to revoke a dispensation retrospectively. That course has not been adopted in this case; the revocation of the fifth dispensation took effect on 6 April 2006 and only prospectively. HMRC’s position is that the payments Reed made did not come within that dispensation or its predecessors, and there is accordingly no need for retrospective revocation.

184.    The remaining sections of Chapter 2 are of no present relevance.

185.    Section 70 is entitled “Sums in respect of expenses”. Only sub-s (1) is material in this case:

“This Chapter applies to a sum paid to an employee in a tax year if the sum—

(a) is paid to the employee in respect of expenses, and

(b) is so paid by reason of the employment.”

186.    It is common ground that if the disputed allowances are payments in respect of expenses at all, they fall within this provision. Section 71 contains explanatory provisions of no application to this case, and it is necessary to pass on to s 72:

“(1) If this Chapter applies to a sum, the sum is to be treated as earnings from the employment for the tax year in which it is paid or paid away.

(2) Subsection (1) does not prevent the making of a deduction allowed under any of the provisions listed in subsection (3).

(3) The provisions are—

section 336 (deductions for expenses: the general rule);

section 337 (travel in performance of duties);

section 338 (travel for necessary attendance) ….”

187.    Thus by the combined operation of ss 70(1) and 72(1) payments made by an employer to an employee in respect of expenses are “to be treated as earnings from the employment” (and are known as “Chapter 3 earnings”) and taxed accordingly, subject to relief for deductible expenditure. Sections 336 to 338, which need to be read with s 339, deal with what is deductible expenditure, and are of some importance. Excised of the irrelevant they read:

337 Travel in performance of duties

(1) A deduction from earnings is allowed for travel expenses if—

(a) the employee is obliged to incur and pay them as holder of the employment, and

(b) the expenses are necessarily incurred on travelling in the performance of the duties of the employment …

338 Travel for necessary attendance

(1) A deduction from earnings is allowed for travel expenses if—

(a) the employee is obliged to incur and pay them as holder of the employment, and

(b) the expenses are attributable to the employee’s necessary attendance at any place in the performance of the duties of the employment.

(2) Subsection (1) does not apply to the expenses of ordinary commuting or travel between any two places that is for practical purposes substantially ordinary commuting …

339 Meaning of ‘workplace’ and ‘permanent workplace’

(1) In this Part ‘workplace’, in relation to an employment, means a place at which the employee’s attendance is necessary in the performance of the duties of the employment.

(2) In this Part ‘permanent workplace’, in relation to an employment, means a place which

(a) the employee regularly attends in the performance of the duties of the employment, and

(b) is not a temporary workplace.

This is subject to subsections (4) and (8).

(3) In subsection (2) ‘temporary workplace’, in relation to an employment, means a place which the employee attends in the performance of the duties of the employment—

(a) for the purpose of performing a task of limited duration, or

(b) for some other temporary purpose.

This is subject to subsections (4) and (5).

(4) A place which the employee regularly attends in the performance of the duties of the employment is treated as a permanent workplace and not a temporary workplace if—

(a) it forms the base from which those duties are performed, or

(b) the tasks to be carried out in the performance of those duties are allocated there.

(5) A place is not regarded as a temporary workplace if the employee’s attendance is—

(a) in the course of a period of continuous work at that place—

(i) lasting more than 24 months, or

(ii) comprising all or almost all of the period for which the employee is likely to hold the employment, or

(b) at a time when it is reasonable to assume that it will be in the course of such a period.

(6) For the purposes of subsection (5), a period is a period of continuous work at a place if over the period the duties of the employment are performed to a significant extent at the place ….”

188.    The parties agree that while this section defines the distinction between permanent and temporary workplaces, and in particular is the source of the 24-month time limit referred to in the dispensations, and is therefore of some significance, it does not itself provide an answer to the question whether the employed temps had single, continuing contracts, or engagements lasting only for the duration of each assignment.

The issues

189.    Before the hearing began the parties identified a number of issues, and provided us with a list. As the hearing proceeded it became apparent that some modest reformulation of the issues would make them clearer, and also that it is neither desirable nor, we think, possible to approach the issues as presented, in the form of discrete questions. There is considerable overlap between them, and some cannot be effectively answered without considering one or more of the others at the same time. As will become clear, we have broken down the matters we must decide in a rather different manner, essentially as factors leading to an overall conclusion on the principal issue, that is whether (leaving legitimate expectation to one side) HMRC are right in their view that the allowances were not covered by the dispensations, and may recover the tax and (assuming the outcome is the same) the NICs for which Reed had not accounted during the relevant period.

190.    Nevertheless, we recognise that we should provide answers to all the questions raised by the parties which do not become redundant by reason of the answer to another question (save that, as we have mentioned, we do not propose to decide whether we have the jurisdiction to determine questions of legitimate expectation and are not required at this stage to decide whether the outcome for NICs is the same as for tax, the subject of issue 9). Even though we have decided upon a different approach the issues identified by the parties are a convenient starting point. As reformulated by us, and with some added explanation, the questions we must answer are as follows:

A: Basis of charge

1. Did those of the employed temps who received payments under RTA or RTB:

(a) enter into an effective salary sacrifice, with the consequence that Reed paid them a reduced salary plus the allowances; or

(b) receive only a salary?

It is common ground that if the answer to question 1 is (b), everything Reed paid to its temporary employees was earnings within ITEPA Part 3 Chapter 1 and (subject to the answer to question 7) we must decide these appeals in favour of HMRC, leaving Reed to pursue such other remedies as it may have elsewhere.

2. If the answer to question 1 is (a), were the allowances:

(a) nevertheless earnings under ITEPA Chapter 1 (and in particular section 62); or

(b) sums to be treated as earnings under ITEPA 2003 Chapter 3 (and in particular section 72) as reimbursement of expenses?

In fact, HMRC put this question a little differently, by arguing that what was paid by way of allowances was nothing more than Chapter 1 earnings, within ITEPA s 62, dressed up as expenses; Reed counters that reimbursements of home to work travel expenses are not Chapter 1 emoluments as a matter of general principle (whether the workplace is temporary or permanent), and cannot be brought within the charge to tax by Chapter 1 but only as a reimbursement of expenses, falling within Chapter 3, and in particular s 72.

B: Availability of deductions

3. Did the employed temps travel from their homes to:

(a) temporary or

(b) permanent workplaces

for the purposes of ss 338 and 339 of ITEPA?

It is agreed that if the workplaces were temporary, a deduction from earnings is allowed in respect of the travel and subsistence payments, pursuant to section 338(1). If they were permanent, it is agreed that no such deduction from earnings is allowed, under section 338 or otherwise. We add for clarification that even if a deduction is allowable, it does not necessarily follow that the allowances fell within Chapter 3 or, if they did, that they were covered by the dispensations.

To answer question 3 it is necessary to address two subsidiary issues:

(i) were the employed temps engaged under what may be termed an “overarching” contract of employment with Reed (that is, a contract that continued beyond the conclusion of any particular assignment until terminated by notice), or under a succession of “job-by-job” contracts?

(ii) If the employed temps were engaged under a succession of “job-by-job” contracts, did each employed temp have only a single “employment” with Reed within the meaning of ss 338 and 339 of ITEPA?

C: Scope and effect of the dispensations

Issues 4, 5 and 6 arise only if the allowances were earnings within Chapter 3 (that is, if the answer to question 1 is (a) and the answer to question 2 is (b)). If the answer to question 1 is (a) but the answer to question 2 is also (a), with the consequence that the allowances were earnings under Chapter 1, the outcome of the appeal will be determined by the answer to question 7 below.

4. Can an inspector lawfully grant a dispensation in relation to payments which are chargeable to tax and, if so, what conditions, if any, must be satisfied in order for him to do so?

5. If the answer to question 4 above is yes, did the dispensations cover the allowances? That is to say, were HMRC “satisfied”, in the manner required by s 65, in respect of the allowances that no additional tax was payable by virtue of the “listed provisions” of the benefits code (as defined in ITEPA s 63)?

6. If the answer to both question 4 and question 5 is yes, what is the effect of a dispensation as a matter of law? In particular:

(a) Does a dispensation relieve the employer of any obligation to deduct tax under PAYE that might otherwise arise (and if so in what circumstances)?; or

(b) Does a dispensation remove only any obligation that would otherwise arise under the PAYE regime to return details on form P11D of certain expenses and benefits paid to employees (and if so in what circumstances)?; or

(c) Does a dispensation remove any income tax charge (including any liability to deduct under PAYE) that would otherwise arise under the listed provisions (and if so in what circumstances)?

D: Legitimate expectation

7. Did Reed have a substantive legitimate expectation that

(a) the allowances would not be subject to income tax or NICs under any provisions;

(b) the dispensations would not be revoked retrospectively in the absence of serious and material misrepresentations; and

(c) if the dispensations were not revoked retrospectively, HMRC would not seek tax or NICs retrospectively (from Reed or its employees);


(a) if so, to what extent?

(b) if and to the extent that Reed had any substantive legitimate expectation, what are the consequences for the statutory appeals before this tribunal?

For the reasons we have given (see para 13 above) we shall confine ourselves to making findings of fact on this issue.

E: Practical consequences

8. In the light of the answers to the preceding issues, were Reed, as employers, under an obligation to make a PAYE deduction in respect of the allowances during the relevant period?

9. Does the outcome for NICs in these appeals follow the outcome for tax?

The answer to issue 8 is wholly dependent on the answers to preceding issues, and does not demand separate examination. As we have said, we are not required at this stage to deal with issue 9, which we set out only for completeness.

191.    We add, for clarity and again to avoid unnecessary repetition, that Mr Clarke QC and Mr Tolley addressed us on issues of employment law for Reed and HMRC respectively, while Mr Ewart QC and Mr Gammie QC dealt with the taxation issues.

Issue 1: did the employed temps make an effective salary sacrifice?

192.    As the comments we have added to the list of issues indicates, this is the core issue and, if we decide it in favour of HMRC, it is determinative of the appeals. An important preliminary point, not separately identified in the list of issues although it was a matter of controversy, is whether the RTA and RTB scheme adjustments, whether or not they constituted an effective salary sacrifice, and whatever the correct view of their consequences for tax and NICs purposes, were incorporated into the employed temps’ contracts at all.

Reed’s submissions

193.    Reed argues that the resolution of this issue requires no more than an answer to a single, simple question: did the contract between Reed and the employed temps who participated in the RTA and RTB schemes provide for payment only of a salary, albeit part of it was described as an allowance for travel and subsistence, or, instead, for payment of a salary plus an allowance of an amount agreed with HMRC? It concedes that the contracts themselves did not include any provisions about the salary sacrifice which it says the employed temps made, but maintains that the handbooks (and other communications with the employed temps, such as the information they were given on recruitment) made it clear that they would be paid on the basis of the current scheme unless they opted out. Each scheme had as its core feature the computation of the employed temp’s salary by the application of an agreed process to the “headline” wage (the hourly rate multiplied by hours worked) and the provision of a travel and subsistence allowance in amounts agreed with HMRC. Crucially, says Mr Clarke, unless they opted out the employed temps were never contractually entitled to the “headline” wage, but only ever to the wage found by the operation of the current scheme, plus a reimbursement of expenses.

194.    It does not matter, he says, that the contracts themselves did not spell out the detail of the schemes; the material set out in the handbooks, in letters sent to the employed temps and on the timesheets together with, importantly, what they were told on recruitment were all terms apt for incorporation into individual contracts of employment, and were so incorporated, both by the fact of their communication to the employed temps and by the parties’ conduct. That conclusion was reflected in the fact that, consistently over a period of several years, the employed temps’ payslips showed that they were paid in accordance with the schemes. There is, moreover, ample judicial authority supporting the proposition that material outside a contract of employment may be incorporated within it: see, for example, Aspden v Webbs Poultry & Meat (Holdings) Ltd [1996] IRLR 521 and Carmichael v National Power plc [1999] ICR 1226, in which Lord Hoffman observed, at p 1233, in relation to the rule about the construction of documents, that

“It applies in cases in which the parties intend all the terms of their contract (apart from any implied by law) to be contained in a document or documents. On the other hand, it does not apply when the intention of the parties, objectively ascertained, has to be gathered partly from documents but also from oral exchanges and conduct. In the latter case, the terms of the contract are a question of fact. And of course the question of whether the parties intended a document or documents to be the exclusive record of the terms of their agreement is also a question of fact.”

195.    In Malone v British Airways plc [2011] IRLR 32 the Court of Appeal decided that the touchstone of incorporation is whether the relevant provision impacts on the working conditions of the employee. Since there is nothing more fundamental to an employee than his pay, terms which impact upon his pay on a regular basis must satisfy Lord Hoffman’s test as it was explained in Malone. That the informal incorporation of a term into a contract is possible is demonstrated by Petrie v Mac Fisheries Ltd [1940] 1 KB 258, in which the court found that a notice about sick pay posted on a factory notice board had contractual effect. Custom, or long practice, had a similar result, as in Harlow v Artemis International Corporation Ltd [2008] IRLR 629, in which the making of enhanced redundancy payments was found by the court to have become a term of the contracts of all employees, because of both prior conduct and custom and practice.

196.    Once it was accepted that those employed temps who did not opt out had contracted to be paid in accordance with the terms of the current scheme, it necessarily followed, since it was the essential feature of the schemes, that they had agreed to forego part of what would otherwise have been salary, in order to receive a different payment in a different form. That, says Mr Clarke, is all that is necessary for a salary sacrifice to be effective: the question is not one of form, but one of substance. If the effect of the contractual arrangements in which the employed temps acquiesced over several years was as Reed maintained, nothing more was required.

HMRC’s submissions

197.    HMRC’s case is that Reed’s employed temps did not enter into a salary sacrifice, still less an effective salary sacrifice, at any time while the RTA and RTB schemes were in operation. The disputed allowances, whatever Reed may have called them, and however they might have been described in the handbooks, were simply a component of the employed temps’ contractual wage or salary, falling within and to be taxed as Chapter 1 earnings, and not as Chapter 3 expenses. Mr Gammie referred us to some authorities on the meaning of the words “wage” and “salary”, but as Mr Ewart did not disagree with the proposition that what was paid to the employed temps (leaving aside the travel-to-work payment which, as both parties agree, fell within Chapter 3 and was always subject to tax and NICs) was Chapter 1 salary unless Reed could demonstrate that it was something else, we need not deal with those authorities.

198.    HMRC do not deny that it is open to an employer and employee to agree to employment terms under which the employee agrees to work in future for a reduced salary or wages (by making a salary sacrifice) in return for some benefit, whether or not monetary. However, if a salary sacrifice is to be effective for tax purposes (thus becoming an “effective salary sacrifice”) the employee must actually agree to work in future for the reduced salary or wage, and the employer must provide some other benefit in a form which is not readily convertible into money. In other words, there is, they say, no effective salary sacrifice if all the employer does (with or without the employee’s consent or agreement) is to pay part of the employee’s contractual salary or wage in some other form, in particular by meeting, or purporting to meet, some or other of the employee’s expenses.

199.    Here, Mr Tolley argues, the purported salary sacrifice was ineffective because it was not incorporated in the temporary employees’ contracts with Reed; alternatively, if it was incorporated, it was not implemented. As a further alternative he says that, should we decide in Reed’s favour in respect of issue 3 (that is, we should conclude that there were continuing rather than job-by-job contracts) there was no effective variation of the contracts. That argument is, we think, little more than a supplement to the first, and it would in any event not apply to those who joined (or re-joined) Reed after the schemes had been introduced.

200.    HMRC first point to the fact that the contracts between Reed and the employed temps did not contain any reference to a salary sacrifice or similar arrangement, at least until the comprehensive change of the scheme which took place in 2006 (and HMRC reserve their position in respect of this change, which of course is not the subject of these appeals). They recognise that something was said in the handbooks and (as we have found as a fact) that the temps consultants offered an explanation, even if, HMRC say, an inadequate (because it was intentionally incomplete) explanation, as new employed temps were recruited, but argue that if conditions are to be incorporated it must at the least be a reasonable inference from the circumstances that the parties intended the contents of the extraneous documents in which they were set out to have contractual effect. And they must be capable of having contractual effect: that is not possible if the documents are written in vague terms, or merely express an aspiration.

201.    For the latter proposition HMRC rely on the observation of Auld LJ in Keeley v Fosroc International Ltd [2006] IRLR 961 at p 966:

“it does not necessarily follow that all the provisions are apt to be terms of the contract as some provisions, read in their context, may be declarations of an aspiration or policy falling short of a contractual undertaking ... It is necessary to consider in their respective contexts the incorporating words and the provision in question incorporated by them.”

202.    Here, not only was there insufficient evidence from which one might infer an intention to incorporate provisions found outside the four walls of the contracts, the available evidence pointed to the conclusion that there was no such intention. Until the 2004 revision, Reed’s contracts with its employed temps were expressly said to set out their entire terms. They were entitled “Conditions of Employment” and made no reference at all to the handbooks or any other communications. The 1999 version of the contract (as it was printed on the reverse of the timesheets then in use) was quite clear: it stated that “the conditions below, together with the details of your assignment on the front of this copy, contain full details of the Terms and Conditions of your assignment”. The 2004 version, by contrast, included the words “the conditions below together with the details of your assignment or secondment on the front of this copy and the Reed Temporary Workers’ Handbook contain full details of your terms and conditions.”

203.    From those circumstances it must follow that, at least until the 2004 revision, it was not open to Reed to rely upon what was said by the House of Lords in Carmichael v National Power; the terms on which the parties had agreed excluded the incorporation of other terms by reference. But even after the 2004 revision, incorporation of the handbook by reference achieved nothing, since it was replete with vague and aspirational statements, such as “Reed’s aim is to find you a regular supply of suitable, interesting and rewarding temporary assignments”, statements plainly not intended to have contractual effect, nor capable of having any such effect.

204.    Moreover, the passages in Reed’s handbook relating to the schemes were not apt for incorporation because they conflicted with the express provisions relating to salary in the employed temps’ contracts. Those provisions made no mention at all of any salary sacrifice: on the contrary, they stated that the employed temps would be paid their full salary or wage entitlement. Again, and contrary to what was suggested by Reed, custom and practice required much more than inference. Mr Tolley referred us to what was said by Peter Pain J in Bond v CAV Ltd [1983] IRLR 360, at para 54 (relying on earlier authorities to similar effect), “A custom of the trade must be reasonable, certain and notorious.” The example of the passage dealing with the schemes which appeared in the handbook which we have set out above (see para 106) was none of those things.

205.    Even if the provisions of the handbook, and any other information communicated to the employed temps were incorporated, HMRC say, they did not give rise to an effective salary sacrifice. All the employed temps agreed to do was to receive their salary in a particular way. In addition, the provisions allowed employed temps to opt out of the schemes; thus nothing was ever truly sacrificed. In order for there to be an effective salary sacrifice the employee making the sacrifice must give up the contractual right to that part of his or her future remuneration, and should generally be unable to undo the sacrifice and revert to the original cash salary. If an employee can elect, at will, to give up the benefit and receive the salary again, the benefit has “money’s worth”, falls within ITEPA s 62 and is taxable as earnings. Since the employed temps could opt out of these schemes at any time, there could not have been any genuine (or effective) salary sacrifice.

206.    In any event, says Mr Gammie, the schemes as implemented (if they were implemented at all) did not lead to a true sacrifice of any part of the salary. The contractual earnings entitlement remained essentially the same—the hourly rate agreed for the assignment multiplied by the number of hours worked. Moreover, the contracts continued to provide that employed temps were engaged at an hourly rate notified in advance, that hourly rate including (since there was nothing to indicate otherwise) whatever amount of travel benefit was eventually paid. The gross pay calculated was shown, just as it had been before implementation of the schemes, on the top line of the payslips. It was true that by ticking or completing the relevant boxes on the timesheets the employed temps elected to take part of their gross pay tax and NICs free, but the size of the tax benefit (and as what Reed referred to as the matrix figures made clear) depended upon the employed temp’s individual tax and NI rate, rather than on the magnitude of the expenses actually incurred. Thus there was no true link between the allowance so calculated and the expense actually incurred by the employed temp.

207.    HMRC also make the point that, in order to be effective, a salary sacrifice scheme cannot lawfully reduce an employee’s cash pay below the national minimum wage. If and insofar as the purported salary sacrifice scheme put in place by Reed did so in any particular case, it would have been ineffective on this ground. This was a point made by HMRC in the course of the correspondence (see para 145 above), but as we had no evidence, one way or the other, whether any employed temp received less than the national minimum wage as a consequence of participation in either scheme we leave this possible factor out of account.

208.    Secondly, HMRC say, even if the salary sacrifice could have been effective as a matter of law, it was not implemented. Reed did not actually operate either the RTB or RTA to take effect as a salary sacrifice so as to reduce the contractual wage or salary in their calculation of the employed temp’s pay. Instead they simply deducted an amount calculated in accordance with the current dispensation from the gross pay in order to add the same sum back and pay it tax free (or supposedly tax free).


209.    We agree that this issue demands answers to the two discrete questions: whether the terms of the RTA and, later, RTB schemes were incorporated into the employed temps’ contracts; and, if so, whether they did result in an effective salary sacrifice. We do not accept Reed’s implicit argument that mere incorporation of the current scheme is enough since it results, ipso facto, in a salary sacrifice.

210.    We think it appropriate to begin with the employed temps’ perception of the RTA scheme, as Reed intended it to be, which is to be derived from the passage in Mr Beal’s first witness statement we have set out at para 117 above. It is quite clear from that extract and from the evidence about their operation in practice that, although the presentation of the scheme to the employed temps was designed to put it in the most favourable light possible, the participants in fact derived very little benefit from the RTA scheme; almost all the benefits accrued to Reed, as was plainly Reed’s intention from the outset. Although, as we have recorded, there was some evidence that the employed temps derived a greater share of the RTB savings, and there was a little more transparency in the presentation, it is an inescapable conclusion that, during the currency of both the RTA and the RTB schemes, Reed concealed not merely the detail of the current dispensation, in particular the amounts of the allowances set out in it, but also the manner in which Reed was operating it, from its employed temps. None of the information given to the employed temps, on their payslips, on recruitment or otherwise, would have made it possible for them to work out for themselves what was the underlying structure of either scheme.

211.    Mr Beal’s evidence was that one could deduce the amount of the RTA scheme allowance, determined by the combination of the amounts permitted by the current dispensation with the boxes ticked and mileage entered on the employed temp’s timesheet (the sum which is also the amount claimed to be the salary sacrifice), from the payslip by a process of elimination. It is, he said, the difference between the headline salary plus “travel allowance” minus Exp Adj (the total of which is shown on the payslip as “Total Payments”), and the amount of “Taxable Pay This Period”. We think he was mistaken, in as much as the Exp Adj sum should not be taken into account in this calculation. But, whether or not he is right, anyone unaware of the manner in which Reed was applying the dispensation could arrive at that conclusion only because there are no other possible sources of the difference between these two figures. It is in our view unlikely that any of the employed temps would, or even could, have worked this out. The working of the RTB scheme was, as we have said, rather less opaque but we are satisfied that an understanding of the underlying structure of that scheme too was beyond the reach of the participants.

212.    We find it difficult to imagine that any of the employed temps would have made an informed decision to participate in either the RTA or the RTB scheme had they understood that underlying structure. We accept that the actual benefit to the employed temps, even under the RTA scheme, was in many cases not much less than it would have been had they sought relief for their travelling and subsistence expenses in the conventional way, assuming such relief was available at all, and that in order to obtain relief they would have been required to keep records and make an annual return. But we have little doubt that, if they had been provided with clear information, many would have been put off by the potential of participation in the scheme to diminish their entitlement to various contributory benefits, and that they would have been surprised, to put it at its lowest, to see that under the RTA scheme Reed could achieve a saving of as much as £20.88 (see the worked example at para 112 above) on a weekly salary of £100, while the employed temp gained only to the extent of (at that time) £5 less tax and NICs.

213.    However, while those conclusions are (as we shall indicate) relevant to the question whether there was an effective salary sacrifice, they do not in our view undermine Reed’s case that the RTA and RTB schemes were incorporated into the employed temps’ terms of employment. We recognise the force of the point Mr Tolley made, that for part at least of the relevant period the terms and conditions themselves provided that they constituted the entire contract, but we do not regard that fact as determinative. We were left in no doubt from Mr Baddeley’s (on this point unchallenged) evidence that an explanation (whether or not adequate) was given to the employed temps on recruitment, and we are satisfied that the handbook indicated a change in the method of payment for those who did not opt out, that some letters were sent and, perhaps most importantly, that the conduct of the parties was consistent with their mutual acceptance that the current scheme had been incorporated into the contract: the employed temps completed the relevant boxes on the timesheets, Reed paid the travel-to-work allowance which their doing so entitled them to receive, and (as far as the evidence indicated) the employed temps consistently accepted payment on that basis. With limited exceptions of no relevance here, a contract is not vitiated because it is a bad bargain, or because one party does not fully understand all its implications.

214.    In our judgment the combination of the information they were given, the opportunity to opt out which was not taken, the active completion of the timesheet boxes and the passive acceptance of the determination of their overall pay in the manner for which the schemes provided, must lead to the conclusion that the participating employed temps agreed to be paid in accordance with the current scheme, either from the start of their relationship with Reed if they joined after implementation of the RTA scheme, or by variation of their existing terms if they were already in such a relationship.

215.    That answer leads to the question: what were the relevant terms of the schemes? The description of the RTA scheme set out in the handbook issued to employed temps in the latter part of 2001 (which is set out at para 106 above) makes no mention whatever of any salary sacrifice, not merely in the sense that the phrase does not appear (the words used are, of course, not in themselves of any significance provided the meaning is clear) but in that there is nothing anywhere to suggest that the employed temps were giving up anything at all. They were offered an additional benefit, what in that description was referred to as the “Travel Allowance” (for which we have used the term travel-to-work payment), but one searches in vain for any indication that part of the salary had to be given up, even if it was nevertheless paid in a different guise. The nearest one comes to any such indication is in the paragraph reading

“Beneath this a figure appears next to the phrase ‘Exp Adj’. This represents the adjustment to your gross pay to allow for the reduction in the total amount of Income Tax and National Insurance due under the scheme.”

216.    This statement does not assist Reed because, as we have said, the “Exp Adj” is, despite its description, a deduction from net pay and it is not claimed to reduce taxable salary, which is necessary to its constituting a salary sacrifice. It is also the wrong amount as the sacrifice should be the amount of the allowance, calculated in accordance with the current dispensation, whereas “Exp Adj” equals the aggregate of the tax and NICs on the amount of the allowance. In any case it is, at best, doubtful whether any employed temp reading that passage would understand it to mean that he or she was giving up any salary; moreover, the numerous enquiries by concerned employed temps, to Reed and to HMRC, are in our view clear evidence that they did not. Although, as we have said, it may not be necessary for the validity of a contract that both parties fully understand it, it seems to us to be a bare minimum (in this context) that the employed temp should know he or she was required to give up x in order to receive y (even if x and y happen to have the same monetary value). We do not understand how anyone can be said to have agreed to sacrifice anything in complete ignorance. But even if we are wrong in that conclusion, it does not seem to us that the employed temps made a salary sacrifice as a matter of fact.

217.    There was no evidence, from the material given to the employed temps, from what Mr Baddeley told us, or otherwise, that participants in the RTA scheme were told that they were to be paid anything other than a salary derived from multiplying the agreed hourly rate by the number of hours worked. As Mr Gammie correctly said, the product of that calculation appeared on the payslip. Moreover, it is not Reed’s case that the employed temps agreed to accept (say) £9 per hour rather than £10, in exchange for something else. There was a complete absence of any variation of the hourly rate: it remained as it had been before the scheme was introduced (and, moreover, was the same whether or not the employed temp participated in the scheme). Further, no figure for the amount of the sacrifice is ever given to the employee; we do not consider that the remote possibility of his being able to deduce it from the other figures in the payslip is sufficient. In other words, the salary was paid in full, even if there was a later manipulation. Accordingly, in our judgment, no part of the salary itself could be said to have been sacrificed.

218.    Can it, instead, be argued (despite the other reservations, particularly about incorporation, we have mentioned above) that the later manipulation led to a salary sacrifice? The process of manipulation, it will be recalled, was to deduct from the gross salary a sum calculated by reference to the amounts specified in the current dispensation, to the extent justified by the entries on the timesheet and by the individual employed temp’s tax and NICs position, and then to add exactly the same sum back again. That (“step 1”) was done, of course, in order to segregate that part of the employed temp’s earnings which was subject to tax and NICs from the part which (if Reed is right) was not; but it is difficult to see how the employed temp could be said to have made any sacrifice by a step in the calculation which resulted in no change to the gross receipt. And it does not, in our judgment, help Reed to argue that the deduction represented the sacrifice and the adding back the benefit received in return for it. However one views the contracts, the employed temps were not offered the benefit of the allowances (as we have said, not merely their amount but even their existence was concealed by Reed), but of the travel-to-work payment, and it is only that payment which might properly be regarded as the benefit. The supposed sacrifice was simply cancelled out, resulting in no reduction of the gross salary.

219.    The next step (“step 2”) was to deduct the “Exp Adj”. This, at first sight, does amount to a deduction from the salary (albeit the net salary), with the consequence that it might amount to a sacrifice. However, as the short extract from the handbook set out at para 215 above correctly shows, it reflects the reduction in tax and NICs which was achieved by step 1—thus, again, a deduction is matched by a corresponding, and equal, addition, albeit in this case in the form of a saving.

220.    For those reasons we have concluded that the employed temps participating in the RTA scheme did not give up anything which was not exactly matched by a corresponding enhancement, and that no part of the salary was sacrificed. If that conclusion is right, the question whether there was an “effective salary sacrifice”, in the sense of reducing the amount on which the employee is taxed, does not arise. Even if it did we consider that any sacrifice would have been ineffective because of the possibility of the employee opting out of the scheme at any time and thereby restoring the headline salary (see the handbook description set out at para 106 above, under the heading “Do I Need To Do Anything To Be Included In The Scheme?”). Mr Beal’s evidence was also that an employee could opt out at any time.

221.    We have given several examples of comments made by Reed or by RR in relation to their own misgivings about the RTA scheme. They, and in particular Miss Ollerenshaw, took the view that the misgivings were addressed by the changes effected by the introduction of its replacement RTB scheme, and that, if there had been legitimate doubt whether the employed temps made a salary sacrifice before, there could be no such doubt now. Moreover, Reed says, the description of the scheme offered to employed temps in the revised handbook (see para 135 above) made it clear that participants had to make a sacrifice in order to benefit.

222.    Before dealing with that point we think we should address Mr Gammie’s argument that even if there was a sacrifice it was ineffective because of the employed temp’s right to opt out. This is the same argument as that addressed at para 220 above, and it perhaps has even greater force than in relation to the RTA scheme since, whereas under that scheme there were some restrictions on opting back in once one had opted out, that was not so in the RTB scheme: as we explained in para 133 above employed temps could opt in or out at will. We agree with Mr Gammie that this feature alone defeats the efficacy of the scheme; a “sacrifice” which can be turned on or off in that fashion cannot truly be regarded as a sacrifice. However, in case we are wrong in that conclusion we have examined the scheme to see whether, leaving the opt-out facility to one side, it did provide for a salary sacrifice.

223.    At first sight it did: first, a sum, described as RTB Adj, determined from the matrix, was deducted from the gross pay, and secondly, the balance was attributed as to part to the amount payable within the dispensation, described as “RTB non-taxable exp TP” [TP means this period] (or “RTB Expenses TP”) and the remainder to “Taxable Pay TP”. That the RTB Adj might vary for reasons related to the participant’s tax and NICs position rather than to the expense actually incurred is, we think, of no consequence in itself, though it does highlight the fact that the scheme had, in truth, only a tenuous connection with travel and subsistence costs. The amount to be deducted was indeed described in the material given to participants as a sacrifice; but the label used is not, in itself, of importance. The corresponding benefit, that is what the employed temps gained in exchange for the sacrifice or supposed sacrifice, cannot be so readily identified from the same sources. It was described variously as “the net benefit of participating in the RTB”, and “the opportunity to increase your take-home pay through the” RTB. The new handbook, rather more explicitly, indicated that the scheme “provides you with a tax- and NI-free travel and subsistence allowance as part of your pay rate”, though did not indicate what it was.

224.    As before, the employed temps earned a sum calculated by multiplying the agreed hourly rate by the number of hours worked. There was no separate identification of the travel and subsistence allowance claimed to be included, and no differentiation in the calculation of the “headline” pay between those who did and those who did not incur travelling and subsistence expenses—and one who used public transport one week and walked or cycled to work the next earned the same gross amount in each week (assuming the same hourly rate and the same number of hours). Similarly, two employed temps with identical expenses but different tax and NICs liabilities were treated differently: as we have said, the scheme had only a tenuous link with actual travel and subsistence costs. The revised RTB payslips identified a sum as “RTB non-taxable exp TP” (or “RTB Expenses TP”), a payment which one might deduce had been made in respect of expenses, but its make-up was not revealed, nor was there anything on the payslip from which it might be worked out: to do that it was necessary to know the amounts set out in the current dispensation, but they were not revealed. It is apparent that the aggregate of that sum and what is recorded as “taxable pay TP” equals the “total payments” (that is, the gross pay less RTB Adj).

225.    The supposed sacrifice under the RTA scheme was matched by a corresponding gain, so that there was no true sacrifice. The RTB scheme was different; the sample payslips produced to us show that the amount sacrificed was not the same as the amount gained, but in some cases was higher and in others lower. But we do not think that matters. In our view a salary sacrifice implies reciprocity: the employee gives up a portion of his or her earnings, even if the portion is variable, in exchange for an identified benefit provided by the employer. Reed, however, did not provide any benefit at all; it merely applied the dispensation in order to enable it to attribute part of the pay, entirely notionally, to the reimbursement of expenses, so that the tax and NICs burden could be reduced. Far from providing a benefit to the employed temp, it appropriated a significant part of the saving to itself; and the supposed sacrifice, however it was presented, was no more than an arithmetical adjustment whose purpose was to ensure that Reed secured the intended share of the benefit. It was not, in our view, a sacrifice in the true sense of that word.

226.    Accordingly, although by different routes, we reach the same conclusion in relation to both of the schemes.

Issue 2: Were the disputed allowances within Chapter 1 or Chapter 3?

227.    We necessarily approach this issue upon the footing that our answer to the first question is wrong, and that the employed temps in fact made an effective salary sacrifice. It does not, of course, follow automatically that if they did, what they received in return fell within Chapter 3; it is quite possible to give up one form of Chapter 1 earnings for another. In essence, the question we must answer is whether the allowances were deducted from the employed temps’ earnings (or never formed part of them) by reason of the assumed salary sacrifice, and were then paid to the employed temps as Chapter 3 emoluments (whether or not they represented the reimbursement of deductible expenditure); or, instead, were simply paid to the employed temps as part of their remuneration, even if a separate part, in which case they could only be Chapter 1 earnings. If they were Chapter 3 earnings they had at least the potential to be the subject of a dispensation, whereas Chapter 1 earnings can never fall within a dispensation.

228.    A considerable part of the hearing was devoted to arguments relevant to this question which would fall for our determination in some circumstances, but which fall by the wayside (whatever the answer to issue 1) in view of our conclusion on issue 3 (see para 280 below). Those arguments related to the interpretation of what we take to be the three leading cases on the deductibility of travelling expenses against employment income, namely Pook v Owen [1970] AC 244 (House of Lords), as developed and explained in Taylor v Provan [1974] STC 168 (House of Lords) and Donnelly v Williamson [1982] STC 88 (Walton J); we had, in particular, a very detailed analysis from both parties of the speeches in Pook v Owen. While we will need to touch on those arguments, our other conclusions make it unnecessary for us to add to the length of an already long decision by descending to great detail.

Reed’s submissions

229.    Reed’s starting point on this issue was its analysis of HMRC’s argument which, it says, is based on the premise that reimbursements of “ordinary commuting expenses” are, by their nature, emoluments. There is, Reed says, no link between the deductibility of a payment and its status as an emolument within s 62(2)(c) (and therefore within Chapter 1). In particular, even if the allowances were not matched by deductible expenditure, they were brought into charge as reimbursed expenses by ss 70 and 72 (and therefore fell within Chapter 3), and accordingly can be within the scope of a dispensation. A genuine reimbursement of expenses which have a sufficient link to an employment is not an emolument at all, whether or not the expense is deductible, and even if the reimbursement involves the payment of allowances on a scale rate. This, Reed says, is the conclusion to be drawn from the line of authorities which began with Pook v Owen.

230.    Once one accepted that the employed temps were obliged to incur and pay what Reed described as “the reimbursed expenses” as the holders of their employments, and those expenses were attributable to their “necessary attendance at any place in the performance of the duties of their employment” (to use the words of ITEPA s 338), it immediately followed that any reimbursement of them must come within Chapter 3, regardless of the deductibility of the expense. This was the conclusion to be derived from the authorities, particularly Pook v Owen itself. There, a doctor (for most of his time a general practitioner) was paid a travel allowance to defray the cost of travel from Fishguard, where he lived and where his surgery, accepted to be his ordinary place of work, was situated, to a hospital in Pembroke, about 15 miles distant, at which he held a part-time appointment which required him to be on call to attend the hospital and act as obstetrician and anæsthetist. His duties (and responsibility for the patient) began as soon as he received a telephone call from the hospital, when he gave initial instructions before setting off for the hospital in his car; occasionally he was able to deal with the matter by telephone alone. He received certain payments in partial reimbursement of the cost of his journeys.

231.    The principal issue, with which we do not need to deal, was whether the expenses Dr Owen incurred were deductible from his Schedule E earnings, under the pre-1998 régime. What is relevant for present purposes is that the House of Lords held, by a majority (Lords Guest, Pearce and Donovan), that the payments were not emoluments. The majority concluded that the payments could not constitute an emolument because they had no element of profit: they merely repaid expenses Dr Owen had incurred because of his employment and were in no sense a reward for his services. Reed points out that the majority rejected the view of the minority (Lords Wilberforce and Pearson), and now advanced by HMRC, that if a payment is made to reimburse the cost of travel to a permanent workplace (what is now called “ordinary commuting”) it falls into general earnings for the same reason as it is not a deductible expense under what is now ITEPA s 338. In other words, the reimbursement of employment expenses, as long as there is no profit element, is not an emolument as a general principle, and deductibility is not a relevant factor. The same conclusion can be drawn from Taylor v Provan and Donnelly v Williamson. Thus even if we were to find that the employed temps incurred only ordinary commuting expenses the allowances would be within Chapter 3.

232.    The fact that the allowances were paid on a statistically-based fixed rate scale does not imply any element of profit in them for these purposes. Walton J addressed this point in Donnelly v Williamson, holding that it was irrelevant to the question whether allowances were emoluments that the teachers’ travel allowances which were in issue in that case were paid by reference to a fixed scale, so that a teacher who managed to find, for example, unusually cheap petrol could make a personal profit from them.

233.    The argument apparently advanced by HMRC, that a reimbursement of expenses must be an emolument unless the expense is deductible, is not merely contrary to the authorities, but difficult to reconcile with the legislation. If HMRC are right, the only reimbursed expenses brought into charge by s 72(1) (that is, the only ones not caught as emoluments in any event) would be those matched by a deduction in accordance with s 72(2). If that were so, s 72 could not have any practical effect as it could never result in a charge to tax. Parliament cannot have intended to enact empty legislation.

HMRC’s submissions

234.    HMRC’s case is that if its primary position, that the allowances were simply salary or wages, is rejected, they would nevertheless remain Chapter 1 earnings if they led to a “profit” within s 62(2)(b) or, more probably, amounted to “anything else that constitutes an emolument of the employment” within s 62(2)(c). It is, they say, obvious that an employee’s ordinary contractual cash remuneration (even if not described as salary or wages) is a profit or emolument and fully taxable as Chapter 1 earnings because it represents what is paid to him “in return for acting as or being an employee” (see Hochstrasser v Mayes (1959) 38 TC 673 per Lord Radcliffe at 707). This remains the case even though the payment is designed to reimburse or meet a particular expense that the employee incurs (or is expected to incur) in the course of his employment.

235.    Mr Gammie provided us with several examples of payments in respect of expenses which nevertheless remain Chapter 1 earnings. They include cases where the reimbursement represented no more than the actual expense: (Westall v McDonald (1985) 58 TC 642); where the employer meets expense directly rather than by reimbursement (Hartland v Diggines (1926) 10 TC 247); or where the employee is contractually required to incur particular expenditure in the course of performing, or for the purpose of performing, his duties, as in Fergusson v Noble (1919) 7 TC 176, in which an allowance for a detective constable’s “plain clothes” was found to constitute part of his earnings.

236.    The conclusion to be drawn from those authorities is that even if there was an effective salary sacrifice, all Reed achieved by agreeing with the employed temps that they would be paid a lower hourly salary and receive the allowances instead was that one Chapter 1 cash payment replaced another: all an effective salary sacrifice would have achieved was to divide Chapter 1 earnings into two parts, without taking either part out of Chapter 1. An arrangement of this kind, no more than an agreement between employer and employee for the payment of a lower salary, supplemented by a contribution to the employee’s personal expenditure, is ineffective to reduce the employee’s Chapter 1 earnings.

237.    It is not disputed that expenses may be deductible even though a payment made by the employer in respect of them comes within Chapter 1. Expenses may be incurred, paid and reimbursed in a variety of circumstances; what is important is to identify the circumstances in which the payment is made, and the character of the resulting receipt. In Fergusson v Noble, to take only one of the examples given above, the payment was made as part of the ordinary terms of employment, and amounted to a round sum allowance that the police officer was free to spend as he thought fit. Similarly here, even if the allowances represented the reimbursement of particular expenditure incurred by employed temps, separate and distinct from the salary that Reed agreed to pay for their work, the only conclusion to be drawn from the authorities is that a cash allowance of this kind is Chapter 1 earnings unless it can be shown to be something else. The difficulty facing Reed was that, just as in Fergusson v Noble, the recipients could spend the money as they liked.

238.    HMRC do not argue that every specific cash allowance, even when paid in accordance with the terms of an employment contract, must be classed as Chapter 1 earnings: in some circumstances a specific cash allowance may well fall within Chapter 3. This, however, is not such a case. The inescapable conclusion is that, whether or not there was an effective salary sacrifice, the employed temps merely agreed to work in the future for a reduced salary plus the replacement benefit of the allowances. The replacement benefit was consequently part of the employee’s ordinary remuneration for his or her work, and thus constituted Chapter 1 earnings at the equivalent of its money’s worth. Where, as here, the replacement benefit consists of a cash payment its money’s worth is self-evident.

239.    The fundamental problem with Reed’s proposition that, even if the allowances reimbursed the employed temps’ ordinary commuting expenses, they could only be chargeable to tax under section 72 ITEPA as Chapter 3 expenses, Mr Gammie says, is that ordinary commuting expenses are not “employment expenses” but personal (based on the employee’s choice of where to live) and the reimbursement of such personal expenditure is necessarily wholly profit even if all that is reimbursed is the actual expense incurred and no more. This is the case even if the allowances were paid, not as part of the employed temps’ salary or wage, but as a separate cash allowance. HMRC accept that the allowances were intended to do no more than reimburse the costs expended in the journey to work, but the payment was nevertheless a profit or emolument, which is properly charged as Chapter 1 earnings.

240.    There is nothing in Reed’s argument that HMRC’s case cannot be reconciled with the legislation. The qualification to section 338(1), contained in section 338(2), namely that the deduction permitted by sub-s (1) for travel expenses “does not apply to the expenses of ordinary commuting …”, has the consequence that where an employer reimburses an employee the costs he or she incurs in travelling to work each day, the reimbursement can be brought into charge only as a Chapter 3 expense since it necessarily falls within s 70, being “paid to the employee in respect of expenses”.

241.    It is not, as Reed suggested in argument, common ground that the employed temps were obliged to incur and pay reimbursed expenses as holders of their employment, or that the allowances were attributable to the employed temps’ necessary attendance at the clients’ premises in performance of the duties of their employment (a proposition which, if correct, would go some way to support its case that the expenses were deductible). HMRC’s case is the simple one that the allowances reimbursed (if they truly reimbursed anything) the costs incurred by employed temps in travelling to permanent workplaces, and that they were therefore ordinary commuting expenses. That they were dependent on the employed temps’ choice not only of where to live but of mode of transport, rather than as a necessary expense of their employment, can be seen from Reed’s decision, when developing the RTB, that if a temp recorded that he or she had travelled to work by bicycle or by foot no travel to work payment was made; instead, the hourly wage was paid without adjustment (see para 134 above).

242.    The allowances were not paid as a specific cash allowance, separate and distinct from the employed temps’ ordinary remuneration, and were accordingly not of the same character as those in issue in Pook v Owen and the other authorities relied on by Reed. There was no question in those cases of the separation, by salary sacrifice or in some similar way, of part of the employee’s earnings in order that it might be treated as the reimbursement of expenses; the payments there were only ever an addition to earnings, and the level of earnings was unaffected by the fact or the magnitude of the payment in respect of expenses. Thus none of those cases was authority for the proposition that the allowances could not come within Chapter 1.

243.     What the House of Lords decided in Pook v Owen was that the reimbursement of expenses incurred in the performance of an employee’s duties cannot amount to Chapter 1 earnings; save in respect of lower-paid employees, such reimbursements are treated as earnings, but only because they count as Chapter 3 expenses This is not so in the case of reimbursements of ordinary commuting expenses, because such expenses are never incurred “in the performance of the employment duties” but only to put a person in a position to perform the duties, and their reimbursement therefore confers a personal benefit (see the leading case on this point, Ricketts v Colquhoun [1926] AC 1, which has been followed on many occasions).

244.    In Reed’s case, the allowances (if they amounted to reimbursement at all) did not reimburse expenses incurred in the course of the employed temps’ performing their duties, but were paid to put them in a position to perform those duties. It is self-evident, says Mr Gammie, that the reimbursement of such expenses confers a personal benefit and that the reimbursement constitutes Chapter 1 earnings. That proposition was put beyond doubt by Walton J’s observation in Donnelly v Williamson [1982] STC 88 at 94 that

“... if an employer pays the expenses of the employee’s travel to work … there cannot be any dubiety as to the status of the cost of such provisions as an emolument.”


245.    We repeat, for clarity, that we deal with this issue on a hypothetical basis, that is on the assumptions that the employed temps made an effective salary sacrifice and received the allowances in exchange for that sacrifice, and that they had only temporary workplaces, so that the allowances were at least capable of being deductible. We then consider whether the result would be different if the allowances were not deductible because the workplace was a permanent one. If the conclusion we have reached on the first issue is right it is perfectly clear (and Reed would not dispute) that the allowances were straightforward earnings falling within ITEPA s 62(2)(a).

246.    If there was an effective salary sacrifice, it is, we think, clear from the authorities that whether the allowances fall within Chapter 1 or Chapter 3 does depend, as the first step in the analysis, on whether they represented reimbursement (assuming they were to be treated as reimbursements at all) of ordinary commuting expenses or of expenses incurred in the course of the employed temps’ employment. We agree with Mr Gammie that if it was the former, what Walton J said in Donnelly v Williamson is determinative: they remained emoluments within s 62. It is not the fact that the expense is not deductible (true though it is) which leads to this conclusion, but that the payment defrays what has to be regarded as a personal (getting to work) rather than an employment (doing the work) expense. There is nothing in Pook v Owen, or any of the other authorities, which casts doubt on the fundamental distinction, as it was drawn in Ricketts v Colquhoun, between expense incurred in putting oneself in a position to do the work, and expense incurred in doing the work itself.

247.    The answer is not quite so easily found in the latter case, that is if the employed temps had a series of temporary workplaces. Reed’s argument, in summary and with a little paraphrasing, is that the sums fell squarely within s 70 as they were “paid to the employee in respect of expenses”, and were “so paid by reason of the employment”. Although they were also “getting to work” expenses, they were of a different character since the employed temps were travelling, not to a permanent place of employment and thus putting themselves in a position to do the work (as in Ricketts v Colquhoun), but fulfilling their contracts of employment by travelling to the premises of the employer’s client, in order that they might discharge the employer’s duty to the client, thus satisfying the second part of the s 70 test.

248.    HMRC’s argument, in similar summary and also with some paraphrasing, is that the description of the allowances as the reimbursement of expenses amounted to nothing more than a label since, just as in Fergusson v Noble, the employee could spend the money on whatever he chose. It is not that there was some element of bounty (Mr Gammie did not contend that the use of round-sum payments which might in some cases have exceeded the actual expenditure was important) but that the entire payment was bounty. If that is correct the label attached by the parties is of no consequence: the payment can only be an emolument.

249.    The conclusion we have reached is that on this issue Reed’s argument is to be preferred. If one assumes there was an effective salary sacrifice, that the allowances are to be treated as having been paid to the employed temps and that the employed temps were required to, and did, attend temporary workplaces, it seems to us that the allowances must be regarded as payments in respect of expenses. It is important in this context that it was only those employed temps who completed the relevant boxes on their timesheets (to indicate that they had in fact incurred expenses, whether by using public transport or in travelling by car, and had worked for a sufficiently long time to become entitled to a subsistence payment) who received them. These were not, therefore, payments made prospectively which the employed temp could spend as he or she chose, as in the case of the police officer’s plain clothes allowance, but reimbursements of expenses already incurred. In our judgment that is what is contemplated by s 70: quantifiable expenses, rather than allowances paid without regard to the actual expense. If the various assumptions we have made were correct, therefore, the payments would fall within Chapter 3.

250.    If instead we assume that there was an effective salary sacrifice but the workplaces were permanent, as we decide below, we consider that Mr Gammie is right and the allowances would be taxable under Chapter 1. Pook v Owen in our view depends on Dr Owen (or any other holder of the employment) having two workplaces. He was appointed by a hospital in Haverfordwest to perform stand-by duties on the basis that he was in practice in Fishguard. It was not a question of his deciding to live there but of the hospital appointing him when he was already in practice (and lived) there, and when anyone they appointed would be in practice elsewhere and would need to incur travelling expenses. The travel expenses that were reimbursed (the first 10 miles of the journey) were not emoluments because the journeys were undertaken in the performance of the duties. We agree with Mr Gammie that Lords Guest, Pearce, Donovan and Wilberforce decided that the reimbursement was not an emolument on this basis, regardless of whether this emerges from what Lords Guest and Pearce said on the issue concerning the deductibility of the cost of the additional five miles of the journey. If reimbursed expenses are ordinary commuting expenses they are not incurred in the performance of the duties of the employment and the reimbursement represents a personal benefit taxable under Chapter 1, as in Walton J’s observation in Donnelly v Williamson quoted above. That is the case here.

Issue 3: were the workplaces temporary or permanent?

251.    As we have said, it is common ground that if the employed temps travelled to temporary workplaces, the relevant expenditure was deductible whereas if they were travelling to permanent workplaces it was not. In other words, the parties agree that all the conditions for deduction are met, save for those relating to the nature of the workplace (and there may be some dispute about the amount eligible for deduction). The core of the dispute on this issue therefore is whether, as Reed says, the employed temps were engaged by a single continuing, or “overarching”, contract to work on several assignments, with the consequence that the clients’ premises are to be treated as temporary workplaces and, in principle, the cost of travel and subsistence is an allowable deduction for tax purposes; or, as HMRC argue, each assignment was, or is to be treated as, a separate engagement with a single workplace, with the consequence that the workers incurred only ordinary commuting expenses for which no deduction is permitted.

252.    The parties also agree that this issue turns on whether the employed temps had a single “employment” with Reed that continued beyond the end of a particular assignment with a client (sub-issue (i) as we have identified it in the list of issues at para 190 above). Reed submits that they did, for two reasons. First, the single, continuing contract of employment for which it argued remained a contract of employment between assignments. Secondly, even if there was no contract of employment in the gaps, there was nonetheless one single “employment” with Reed throughout (that is, the answer to sub-issue (ii) is “yes”) and this is sufficient to satisfy the statutory test. HMRC’s pleaded case is that if the contracts are correctly analysed, it is apparent that there were distinct contracts in place in respect of each assignment, and that each such contract (following the July 1995 changes) was a contract of employment.

253.    An appropriate starting point, before coming to the parties’ detailed submissions, is the definition of employment in s 4(1) of ITEPA, which is:

“In the employment income Parts ‘employment’ includes in particular—

(a) any employment under a contract of service,

(b) any employment under a contract of apprenticeship, and

(c) any employment in the service of the Crown.”

254.    The term “includes” leaves open the possibility that there are other types of employment falling within the definition although it is difficult to think what these might be, and we do not think it necessary to speculate.

Reed’s submissions

255.    Mr Clarke’s principal point is that HMRC’s pleaded case did not reflect the correct analysis, since it did not properly distinguish between an engagement that gave rise to a contract of employment during the currency of a single assignment, and a contract of employment which persisted beyond the end of the assignment in question. Whether or not a worker has the status of employee will “need to be resolved as a question of fact according to the particular circumstances of each case”: see McMeechan v Secretary of State for Employment [1997] IRLR 353 at para 35. Nevertheless, says Mr Clarke, while the fact-finding exercise is important it cannot displace the usual approach to the construction of contracts. He emphasises that the contracts in use here described themselves as contracts of service, a term treated as synonymous with a contract of employment. He referred us to an observation of Lord Denning MR in Massey v Crown Life Insurance Co [1978] IRLR 31 at para 13 (echoed by Lawton LJ at para 27) that “If the true relationship of the parties is that of master and servant under a contract of service, the parties cannot alter the truth of that relationship by putting a different label upon it.” However, he says, the label adopted by the parties must be a relevant factor when determining the true status of their relationship, and one which may be decisive if the nature of the relationship is otherwise ambiguous.

256.    He accepts that the label chosen by the parties to a contract of the kind in use by Reed says nothing about whether the employment for which it provides is coterminous with the first assignment, or extends over several assignments. However, where, as here, the parties have enshrined their whole agreement in a document it is not the role of any tribunal to seek to re-write the contract under the guise of a fact finding exercise: see Autoclenz Ltd v Belcher [2010] IRLR 70. Thus the express terms of an agreement may only be disregarded if they do not set out the true agreement between the parties. There was no evidence to suggest, and no other reason to think, that that was the case here. The parties had clearly agreed that the employed temps should be Reed’s employees; the only remaining question, in relation to the contracts, was whether, as Reed says, they established what the Court of Appeal, in McMeechan, described as a “general engagement”, meaning one which would not terminate as the assignment to which it related came to an end, but could cover work on a number of separate assignments.

257.    Mr Clarke argues that two of the contractual terms were of particular importance: the term providing for notice, and the term which linked the duration of the employment to the duration of the assignment, the latter being within the standard form only until the October 2004 revision. Although, Mr Clarke acknowledges, at first sight those provisions conflicted, as HMRC maintained, there was on closer examination no such conflict. The form of contract used until November 2001 stipulated that

“The duration of the Temporary Employee’s employment will be for so long as Reed offers work to the Temporary Employee. It is anticipated that this will be for the duration of the assignment … Reed may instruct the Temporary Employee to end the assignment at any time without specifying reasons.”

258.    From November 2001 the corresponding provision was

“The duration of the Temporary Employee’s employment will be for the duration of the assignment … Reed may instruct the Temporary Employee to end the assignment or secondment at any time without specifying reasons.”

259.    In October 2004 the provision changed again, to become

Reed will inform the Temporary Employee of the likely duration of each assignment or secondment … Reed may instruct the Temporary Employee to end the assignment.”

260.    The need to distinguish between the terms of a contract of the kind used by Reed which provided for times when the employed temp was working in an assignment, and the terms which provided for the intervals between assignments was identified by the Court of Appeal in McMeechan, at [36]:

“… following what appears to be a common (though potentially confusing) practice, the agency and the temporary worker have committed themselves to standard terms and conditions which are intended to apply both to the general engagement and to the individual stints worked under it. The only result of that fusion is that the same conditions will have to be interpreted from a different perspective, according to whether they are being considered in the context of the general engagement or in the context of a single assignment. That does not make the task of the tribunals any easier, and is liable to lead to the unsatisfactory consequence that the same condition may need to be given a different significance in the one context from that accorded to it in the other. Those disadvantages do not, however, supply any valid reason for denying the temporary worker or the contractor the right to have the issue of contractual status judged separately in the two contexts.”

261.    The Reed contracts, Mr Clarke argues, were less difficult to interpret than those in issue in McMeechan, since neither of the two relevant terms (the term as to notice and the assignment term) required two interpretations, one for use in each context. Each term operated in only one context. The true analysis is that the rights which related to the assignment (in particular the right to the payment of wages) ceased at the end of that assignment, while other rights and obligations, including those providing for notice, continued after its end. This analysis defeats HMRC’s position: there was no conflict between the provisions relating to the assignment and the general provision as to notice. The latter recognised that, although the assignment had ended, the overall employment relationship was to remain in existence. Other rights and obligations continued, such as Reed’s obligation to endeavour to find work for the employed temp, the obligation imposed on the employed temp to keep in touch with the temps consultant and “remain on the books”, the employed temp’s right to holiday pay, and the provisions relating to the giving and receiving of notice. Indeed, the requirement on both parties to give notice of termination in accordance with statutory requirements is inconsistent with the proposition that the employed temp had no more than a series of engagements, each ending at the same time as the assignment to which it related.

262.    It is inconsistent too with the position which pertained if notice was not given, and further assignments were offered and accepted. The contract continued, as did the parties’ rights and obligations. An obvious example was that the employed temp continued to accrue paid holiday entitlement; there was no discrete entitlement for each assignment. It was only when Reed did not wish to offer further assignments, or the employed temp did not wish to accept any further offers, that the contract was brought to an end.

263.    Reed also submits that the definition of employment in s 4 of ITEPA is an inclusive one and therefore does not necessarily imply a single contract of service. A purposive interpretation for determining whether a workplace is permanent or temporary within the meaning of s 339(2) and (3) of ITEPA does not depend on whether there are technically one or more contracts of employment. Section 212 of the Employment Rights Act 1996 recognises that an employee’s period of employment can include periods while he is not working, whether or not there is a contract of employment in being. Even if, here, there was no contract of employment there were nevertheless the continuing contractual terms mentioned above. Alternatively a contract of employment and an employment relationship need not be coterminous, as in the recognition of the period of “stable employment relationship” in connection with bringing claims under the Equal Pay Act 1970.

264.    There is no possible room for doubt that during an assignment there was a contract of employment. There is equally nothing to negate the proposition that the contract continued unless and until it was terminated by notice, and its status as a contract of employment did not change. In Wiltshire County Council v National Association of Teachers in Further and Higher Education [1980] IRC 455 the Court of Appeal held that a teacher’s fixed term contract, obliging her to work for the period of the academic session if so required, continued in effect even where the teaching work and, in consequence, the right to payment had come to an end before the end of the fixed session. The position here, Mr Clarke argues, was similar.

HMRC’s submissions

265.    HMRC start from the proposition that “employment” in the relevant tax provisions necessarily means a single contract of employment, rather than a succession of discrete contracts, even if they are between the same parties. They do not deny that the employed temps were Reed’s employees while working in an assignment; their case is that the employed temps had no contract of employment in any gaps between assignments. They do not argue that the contracts lose the status of contracts of employment once work on an assignment ceases, and object to Reed’s characterisation of their case in this way. Their case is the simpler one that the contracts were, and were always intended to be, contracts subsisting only for the duration of each assignment (and, moreover, there was no guarantee that an assignment would last for any particular period, since Reed could bring an assignment to an end at any time without giving reasons). HMRC add that Reed’s case, that there was a continuing contract, is at odds with its publicly stated position at the relevant time, not only in its dealings with its employed temps, but also in arguments presented to Employment Tribunals. We heard submissions on this point, and there was a significant amount of documentary evidence supporting the proposition that Reed had earlier, and in different contexts, advanced arguments wholly contrary to those advanced before us, but (as HMRC themselves concede) such a change of position is not determinative and we think little is to be gained by our exploring it.

266.    It is important, HMRC say, to distinguish between contractual and statutory rights. Statutory rights apply to a “worker” (defined in the Working Time Regulations 1998) who may be either employed or self-employed: a self-employed agency worker is a “worker” within this definition. Tax law, however, draws a distinction between these two categories. The concept of “continuous employment” is used solely in relation to statutory rights to determine such matters as whether the employee has a right to seek a remedy for unfair dismissal, or is entitled to holiday pay or to a redundancy payment.

267.    Contrary to Reed’s case, the terms of the contracts relating to the duration of the engagements (see paras 257ff above) are of particular importance. They stipulated, until October 2004, that the employment began and ended with the assignment; it was impossible in those circumstances to see how the contracts could be construed, contrary to their clear wording, as continuing contracts across assignments. Although, from October 2004, that express stipulation was absent, there was nothing in the amended contract which extended the length of the engagement beyond the end of the assignment. Whatever the relationship between Reed and the employed temps in the intervals between assignments, it was not that of a contract of service. In support of that proposition Mr Tolley referred us to a number of authorities but the point is an essentially straightforward one which we have found it more convenient to deal with in our discussion of the arguments (see para 275 below); we see no need to explore all the authorities.

268.    It is not merely the contracts which are inconsistent with Reed’s argument that there was a continuing contract. The handbooks, too, made it clear that Reed had no obligation even to look for work for its employed temps (there was no more than an understanding that it would offer suitable assignments when they were available). The obligation to provide holiday pay arose not from the contracts but from the law (the Working Time Regulations 1998), which conferred a right to holiday pay on all workers, whether or not they were engaged under a contract of employment. Indeed, Reed’s contracts specifically circumscribed the employed temps’ entitlement so that it was limited to the rights conferred by the Regulations, and excluded the accrual of paid holiday entitlement in the periods during which they were not working.

269.    The employed temp, too, assumed no contractual commitment at all between assignments, but did no more than indicate a willingness to be invited to perform suitable assignments as and when they became available. There was no obligation on any employed temp to accept any particular assignment, suitable or not: there was not even any requirement to consider an offer in good faith. In addition, there was no exclusivity agreement between Reed and the employed temps, preventing them from working for any other agency or employer during the gaps between assignments. Although we had no direct evidence on this point, it is an obvious inference from Mr Baddeley’s evidence that this occurred, and that Reed attempted to prevent it, not by any form of contractual bar, but by making its terms of engagement more attractive than those of its rivals. Even the expectation that the employed temps would “keep in touch” with Reed between assignments was, says Mr Tolley, no more than an aspiration, which Reed could not enforce.

270.    It follows from those characteristics of such relationship as Reed had with its employed temps (using that term loosely) between assignments that there was no mutuality of obligation—on the employer, to provide work and pay, on the employee to do the work—an essential feature of any true contract of service.

271.    Contrary to Reed’s position, the contractual terms relating to the giving of notice, whether before taking paid holiday or in order to terminate the contract, could have no bearing on the employed temps’ employment status during gaps between assignments. The fact that such terms might be effective and enforceable during an assignment says nothing about their effect and enforceability between assignments. Even assuming there was a contract of employment between assignments, in practical terms the employed temp would suffer no loss, and would have no remedy, if Reed terminated the contract without notice at such times, since there was no right to work and no right to pay; and Reed would suffer no loss, and would have no remedy, if the employed temp terminated the contract since he had no obligation to accept any assignment it might have offered. In addition, Reed’s contracts specifically provided that if Reed gave notice of termination, it was not required to offer any work, and was accordingly under no obligation to provide pay, during the notice period. In any event, Mr Tolley adds, as Mr Baddeley’s evidence made clear, the notice provisions were clearly routinely ignored in practice.

272.    HMRC do not accept that the new form of contract introduced in 2006, which includes an obligation on Reed to provide at least 336 hours of work, or pay in lieu if work cannot be found, in any 12-month period, necessarily amounts to a continuing contract of employment but, as we have mentioned, that is not an issue before us. What they do say in the context of this appeal is that the obligation on Reed to provide something, even when there is no current assignment, is wholly lacking in the earlier contracts. Thus if the 2006 contract satisfies the bare necessity of mutuality of obligation, its predecessors manifestly do not.


273.    In our view the only question we have to answer in respect of this issue (since all else falls into place once the answer is found) is whether the employed temps were engaged under a series of job-by-job contracts, as HMRC contend, or under a contract that continued (as an employment contract) following the end of an assignment (we do not say “between assignments” because there may be no further assignments), as Reed contends.

274.    We do not consider that we should read into the ITEPA s 4 definition anything based on the statutory concept of continuity of employment, under which various absences (including being “absent from work in circumstances such that, by arrangement or custom, he is regarded as continuing in the employment of his employer for any purpose”: s 212(3)(c) Employment Rights Act 1996) count in computing the employee’s period of employment. That concept serves an entirely different purpose, of determining whether, for example, the employee can bring a claim for unfair dismissal, which is a purely statutory right for which the employee needs a stipulated minimum period of continuous employment; as a matter of contract law the only issue would be whether the termination was in accordance with the contract. While for employment law purposes there may be times where there is no employment contract in being but nevertheless the time counts towards the employee’s period of employment, for tax purposes we are concerned with the question whether there is “any employment under a contract of service.” To answer this question one looks at the contractual position. For tax law purposes the issue is therefore whether, when the employed temp is not on an assignment, he is employed under a contract of service.

275.    This breaks down into the questions whether, when he is not on an assignment, there is a contract at all and, if there is, whether it is a contract of employment, or there is some other relationship between Reed and the employed temp. The answer to the first depends on the contractual terms applying when there is no assignment. The answer to the second depends, as HMRC say, on the classic test of employment, namely whether the “irreducible minimum of obligation” exists between the parties during the period following an assignment. We agree with Mr Tolley that the essential requirements were succinctly put by Stephenson LJ in Nethermere (St Neots) Ltd v Gardiner [1984] ICR 612, at 623F-G, in what we take to be the universally accepted exposition, one quoted by many other authorities:

“A contract of service exists if these three conditions are fulfilled. (i) The servant agrees that, in consideration of a wage or other remuneration, he will provide his own work and skill in the performance of some service for his master. (ii) He agrees, expressly or impliedly, that in the performance of that service he will be subject to the other’s control in a sufficient degree to make that other master. (iii) The other provisions of the contract are consistent with its being a contract of service.”

276.    We summarise the terms of Reed’s employment contracts that we consider to be relevant to this issue:

·             The contract was expressed to last so long as Reed offered the employed temp work, which was anticipated to be until the end of the assignment (which could be brought to an end at any time). The post-October 2004 contracts did not contain this provision, but were silent on the point.

·             Payment was for hours worked only.

·             Reed “will endeavour to find the Temporary Employee the opportunity to work in the capacity [agreed at registration and] specified on the Temporary Employee’s copy of the time sheet where there is a suitable assignment with a client for the supply of such work.” (The words in square brackets were added to the October 2004 version.) If several persons were suitable Reed could select which one was offered the assignment. The employed temp was not under an obligation to accept any assignment Reed offered.

·             “The Temporary Employee is entitled to paid annual leave in accordance with the Regulations.” Various conditions applied and the employed temp was obliged to give two weeks’ notice of the intention to take paid leave. The Staff Handbook, 2001 version, stated that if the person did not work for Reed for two weeks or more this would count as a break in the determination of the 13 week qualifying period.

·             Both parties had to give notice of termination “in accordance with the statutory requirements.”

277.    Accordingly, when there was no current assignment there were various contractual terms that still applied. These included the (weak and in practical terms unenforceable) obligation on Reed to endeavour to find the opportunity to work, without any corresponding obligation on the employed temp to accept any work. The employed temp had to give notice to Reed to take paid holiday (which could in practice apply only if the person was on an assignment at the time), and Reed had to meet the statutory requirements about holiday pay. Both parties had to give notice of termination in accordance with the statutory requirements. Reed’s obligations under these last two do not add anything to the statutory requirements. In practice, as we have said, the notice requirement seems to have been largely ignored; there was no standard form for giving notice and Mr Baddeley said that giving notice was rare. The right to be given notice was of no benefit to the employed temp since Reed were expressly “under no greater obligation than at any other time to provide work during a period of statutory notice” and so the employed temp would suffer no loss if notice were not given. We consider, however, that these terms are just enough to mean that there was a contract of some sort in existence when the employed temp was not on an assignment.

278.    The next question is whether the contract was a contract of employment satisfying the “irreducible minimum of obligation.” Condition (i), as identified by Stephenson LJ in Nethermere (St Neots) Ltd v Gardiner, is that “The servant agrees that, in consideration of a wage or other remuneration, he will provide his own work and skill in the performance of some service for his master.” Here, neither party’s obligation satisfies that requirement: Reed’s obligation was (at most) merely to endeavour to find the opportunity to work, and the employed temp was not under any obligation to accept any offer of work. Nor are conditions (ii) and (iii) satisfied: Reed exercised no control over the employed temp when he was not on an assignment (indeed the employed temp might well be working for someone else who was exercising control over him), and there were no provisions of the contract that were consistent with its being a contract of service.

279.    Accordingly, while we accept that there was a contract of some sort when the employed temp was not on an assignment, it was not a contract of employment. With regard to Reed’s contention that during an assignment there was undoubtedly a contract of employment which must continue until terminated by notice, with the result that its status will not change, we see no reason in principle why a contract should not change from being an employment contract if the current circumstances are such that the “employee” is not under any obligation to perform any services and the “employer” is not under any obligation to provide any work or remuneration, particularly when, as we have already pointed out, the person might well be working for someone else. It does not seem to us that Wiltshire v National Association of Teachers helps Reed since the issue there was the different one of whether the contract was for a fixed term that expired without being renewed. By contrast, the term in Reed’s Staff Handbook that the 13 week qualifying period for holiday pay ceased to accrue if the person did not work for Reed for two weeks or more must be based on the assumption that there was no continuing employment when the person was not on an assignment.

280.    It follows that each assignment represented a separate contract of employment, and hence a separate “employment under a contract of service” for tax purposes. In other words, the answer to both sub-issue (i) and sub-issue (ii) is “no”, and the answer to question 3 is (b). The travel was to a permanent workplace and the expenses were ordinary commuting expenses and non-deductible.

Issue 4: Could the inspector lawfully grant the dispensations?

281.    The formulation of this issue presupposes that the allowances were Chapter 3 earnings paid in respect of non-deductible expenses—in other words the dispensations removed, or purported to remove, the allowances from the tax, and particularly the PAYE, net. It is common ground that the dispensations could never apply to Chapter 1 emoluments (as we have found the allowances were). The answer to this issue, of course, depends on the construction of s 65, which is set out in full at para 181 above.

282.    In brief, Reed contends that the section requires no more than that HMRC should be satisfied that there is no liability to tax on the payments covered by a dispensation; it does not require that there is in fact (or law) no liability. The function and effect of a dispensation is to give an employer certainty that specified payments made to employees will not give rise to a tax charge, and are excluded from the PAYE system. Thus, once granted, a dispensation will prevent a tax charge even if it subsequently emerges that HMRC were mistaken. This, Reed submits, follows from the plain words of the statute. HMRC’s suggestion, that a dispensation can apply only when there is in fact no tax charge, would deprive the dispensation of almost all practical use. The taxpayer would still need to consider the application of the benefits code and would still need to keep all relevant records in case, on reflection and despite the dispensation, HMRC later decided that there was a tax charge. This would be a particular concern in the context of PAYE since the employer would in practice have no means at his disposal for recovering the tax from the employee if the payment later proved to be taxable. The dispensation must therefore offer the employer certainty that the payments can be made without operation of PAYE.

283.    HMRC contend that the section is limited in its effect. Section 65 applies only when the charge to tax is under the “listed provisions” but here the charge is under Chapter 1. Accordingly the dispensations had no effect as the inspector was acting beyond his powers in granting a dispensation for something chargeable under Chapter 1. In relation to the deductibility of the expense the inspector can be satisfied that no additional tax is payable only if this is correct in law. If the inspector were wrongly satisfied that there was no additional tax he would be acting beyond his powers in granting the dispensation. The dispensations were applied for and granted in respect of employees who had no permanent workplace. The disputed allowances were not covered by the dispensations.


284.    Subsection (1) starts “This section applies for the purposes of the listed provisions.” The “listed provisions” are found in ITEPA s 216(4) and include s 70, into which it is agreed the allowances fall, provided they are not Chapter 1 earnings: see para 186 above. Subsection (5) goes on to provide that if a dispensation is given “nothing in the listed provisions applies to the payments, or the provision of the benefits or facilities, covered by the dispensation or otherwise has the effect of imposing any additional liability to tax in respect of them.” It necessarily follows that if the listed provisions do not apply to the payments in question—in other words, they do not fall within s 70 because, as we have found in this case, they are Chapter 1 earnings—the dispensation has no effect on them and, so far as it refers to the allowances, is a nullity. It can be no more than a statement that HMRC are satisfied that there is no liability to tax under the listed provisions; but the liability to tax under some other provision is unaffected.

285.    If, nevertheless, it is assumed that the listed provisions are applicable, by sub-s (1) the payer (“P”) can give a statement to HMRC of the cases and circumstances in which payments, benefits or facilities are provided to employees. Subsections (3) and (4), which for ease of reference we set out again, then provide that:

“(3) If the Inland Revenue are satisfied that no additional tax is payable by virtue of the listed provisions by reference to the payments, benefits or facilities mentioned in the statement, they must give P a dispensation under this section.

(4) A ‘dispensation’ is a notice stating that the Inland Revenue agree that no additional tax is payable by virtue of the listed provisions by reference to the payments, benefits or facilities mentioned in the statement supplied by P.”

286.    This by its terms depends on HMRC’s being satisfied that no additional tax is payable (necessarily by virtue of the listed provisions); a dispensation is no more than an acknowledgment by HMRC that no additional tax is payable by reference to the facts in the statement furnished by P. There is no requirement that HMRC must be correct in being so satisfied. They might be wrong because the statement of the cases and circumstances wrongly describes the facts, or (as here) a fact that is critical to their deductibility—whether the employment was per assignment or continuing—was not included in the statement because neither the taxpayer nor the inspector considered it to be relevant, or because the inspector was wrong in law in considering that the employees had a temporary workplace so that travel expenses were deductible and no additional tax was payable. What a dispensation cannot do, for the reasons we have given, is apply any one of the listed provisions to the payments, when they are in truth taxable under some other provision. Whether HMRC are right or wrong in agreeing “that no additional tax is payable by virtue of the listed provisions”, they cannot be taken, by reason of the grant of a dispensation, to have agreed that no tax is payable in accordance with the correct provision.

287.    If HMRC subsequently discover that they were wrong for whatever reason they are adequately protected by their power in sub-ss (6) and (7) to revoke the dispensation either prospectively or (particularly) retrospectively, in which case sub-ss (8) and (9) provide that the liability to tax removed by the dispensation arises again (the date from which this arises depending on whether the dispensation is revoked prospectively or retrospectively), and the employer and employees must make all returns that the dispensation relieved them from making (again depending on whether the dispensation is revoked prospectively or retrospectively). It follows expressly from sub-s (5) and by implication from the effect of the revocation of the dispensation in sub-ss (8) and (9) that while the dispensation applies it removes both the liability to tax and the duty to make returns that would otherwise apply.

288.    We therefore agree with HMRC that liability to tax under the listed provisions is an initial requirement, and that if there is no such liability a dispensation has no effect, or (put another way) it was beyond the powers of HMRC to have granted it. But we agree with Reed that, on the basis that the listed provisions are applicable, a dispensation merely requires that HMRC be satisfied that no additional tax is payable, regardless of whether they are right or wrong. HMRC can therefore lawfully grant a dispensation on being so satisfied even though they are wrong (whether in fact or in law, unless they are wrong about the applicability of the listed provisions). Such a disposition is validly given and has full effect to remove the liability to tax and the duty to make returns unless or until it is revoked.

Issue 5: did the dispensations cover the allowances?

289.    The short answer to this question is “no” since, as we have determined, the allowances fell within Chapter 1 and could therefore never be covered by a dispensation. We therefore consider what would be the position should we have decided the allowances fell within Chapter 3, and were subject to the benefits code, albeit (as we have also concluded) they were not paid in reimbursement of deductible expenses.

290.    The answer to this question, in these hypothetical circumstances, follows straightforwardly from what we have said in respect of issue 4. The answer to that issue is not a straight yes or no in all circumstances: if HMRC were wrong in considering that the listed provisions apply then, for the reasons already given, the dispensation had no effect; but if they were wrong for any other reason the dispensations had effect (though subject to HMRC’s right of revocation).

Issue 6: what is the effect of a dispensation?

291.    It will be recalled that this issue was broken down into three questions, with which we can deal fairly briefly. We repeat, to avoid confusion, that we have concluded that the listed provisions did not apply to the allowances and the dispensations therefore had no effect. What follows, again, is based on the hypothesis that the dispensations were effective.

292.    The first question is whether a dispensation relieves the employer of any obligation to deduct tax under the PAYE system that might otherwise arise—and if so in what circumstances? It seems to us clear that a dispensation does have that effect, provided only that the listed provisions are applicable (that is, by its own terms, s 65 cannot relieve the employer of an obligation to deduct tax which is due for other reasons).

293.    The second and third questions can conveniently be taken together. They are whether a dispensation removes only any obligation that would otherwise arise under the PAYE regime to return details on form P11D of certain expenses and benefits paid to employees; or whether it removes the underlying income tax charge (including any liability to deduct the tax in accordance with the PAYE scheme) that would otherwise arise under the listed provisions. It follows from what we have already said that the answer to the second question is that does remove the P11D obligation, but that is not its only effect, and that the answer to the third question is “yes”. In other words, an effective dispensation, unless and until revoked, removes the payments in question from tax altogether. This is so even though the inspectors were wrong in agreeing that the disputed payments were deductible on the basis that the employed temps had temporary workplaces. Since the dispensations were not revoked with retrospective effect this would have remained the position while each one was in force, assuming they had been effective.

Issue 7: Reed’s legitimate expectation

294.    As we have already indicated, we do not intend to consider whether this tribunal has the jurisdiction to determine questions of legitimate expectation, but to limit ourselves to findings of fact which may be useful elsewhere. Many of the relevant findings have already been set out above, and there is nothing to be gained by our repeating them. A short summary may, however, be of assistance. We add, as background, that the question which will eventually fall for determination is whether HMRC can be required, for this reason, to apply a dispensation they had no power to grant.

295.    Although there is inevitably an element of hindsight in this conclusion, it seems to us that HMRC could have been rather more vigorous in seeking information, and more searching in their enquiries before granting the first dispensation, or before replacing it with its successors. HMRC themselves recognised this—see para 153 above. Mr Read told us, too, that he would not have accepted that the employed temps were truly making a salary sacrifice had he realised that Reed was basing the supposed sacrifice on the employed temp’s tax and NICs position, rather than on either the expense actually incurred or the amounts set out in the current dispensation. We merely comment that we found it surprising that HMRC did not discover how Reed was utilising the dispensations much sooner than they did.

296.    However, that observation has to be considered against the background of what s 65 requires, and the consequence that the dispensations were granted on the strength of Reed’s representations, and limited to those circumstances which were described in those representations. We set out relevant passages from the first dispensation, making it clear that the inspector relied on what Reed told him, and was granting the dispensation in respect of payments made in the manner Reed had specified, at para 73 above; they reflect s 65, in requiring HMRC to grant a dispensation if they are satisfied, from the statement furnished, that no additional tax is payable. It follows that an applicant for a dispensation bears the burden of determining the relevant facts and conveying them to HMRC, and that he bears the consequences of any error he makes.

297.    We have already indicated (see para 173 above) that the statement which led to the grant of the first dispensation, and was effectively repeated in relation to its successors, reflected what Reed and RR believed to be relevant and correct at the time, namely that those to whom the dispensation would be applied were employees, and that they had temporary workplaces. We have concluded that the latter belief was wrong, as a matter of law, but we can accept that at the time Reed believed the contrary. We repeat, however, the point we have already made that it was not only Reed and RR, but HMRC too, who had not appreciated the significance of this point at that time, both sides thinking that what mattered in this context was only that the participants should be Reed’s employees, and not agency workers.

298.    The evidence showed quite clearly that Reed and RR knew that the schemes were risky and, as Miss Ollerenshaw put it, “aggressive”. It will also be apparent from our narrative of events that we have set out above that they were not forthcoming about the manner in which the dispensations were being applied, and in particular that the employed temps were not themselves reaping more than a very modest part of the benefit, although it is true that it and RR did not actively conceal what was being done, and did disclose to HMRC that Reed was taking part of the benefit for itself. It also seems to us that, had they enquired rather more deeply, in particular in connection with the queries about the layout of the payslips, HMRC might have discovered sooner than they did that Reed was only nominally paying the allowances to its employed temps. Miss Ollerenshaw, as we have said, was uncomfortable about the operation of the RTA scheme (though less so about the RTB scheme); Reed itself shared some at least of her concerns (see, for example, para 115 above) and, as we have recorded (see para 117 above), Mr Beal fully recognised that Reed was not paying the dispensation allowances to its employed temps.

299.    HMRC’s position is that Reed could have no legitimate expectation because it did not fully disclose all relevant matters. We are satisfied, and find as a fact, that Reed did not volunteer to HMRC all of the details of the schemes. We have identified the more significant of those details already, and we have commented on Reed’s own recognition that the schemes were innovative. However, although Reed and RR were less forthcoming than they might have been, we do not find that there was any deliberate concealment. We also repeat the observation we made earlier that neither Reed nor HMRC recognised the relevance of some matters until a late stage.

Issue 8: was Reed obliged to make PAYE deductions in the relevant period?

300.    It necessarily follows from what we have already said that there is only one possible answer to this question: yes. The allowances were Chapter 1 earnings; but even if they were not (and instead fell within Chapter 3) they did not meet expenditure for which the employed temps could properly claim relief. In either case, therefore, they were subject to deduction of tax (and NICs).

Issue 9: is the outcome for NICs the same as for tax?

301.    Although we are not asked to provide an answer at this stage, it seems to us that this question is as easily answered as its predecessor, and it would be appropriate for us to record a conclusion. Once it is accepted that a s 65 dispensation applies (in practice) to NICs as it applies to tax, it inevitably follows that a dispensation which is ineffective for tax must be equally ineffective for NICs. Whether the allowances amount to Chapter 1 earnings or Chapter 3 earnings which are not covered by an effective dispensation, they plainly attract NICs.


302.    We are satisfied that the allowances, although purportedly covered by a dispensation, were Chapter 1 earnings; that even if that conclusion is wrong they were Chapter 3 earnings which did not attract relief because they were paid (if they were truly paid at all) to reimburse ordinary commuting expenses (because the employed temps had series of job-by-job contracts, and not continuing contracts of employment), that Reed should have accounted for tax and NICs on the allowances throughout the relevant period and that the assessments and determinations were, in principle, correctly made.

303.    Subject to the resolution of the subsidiary issues identified at para 10 above, which we do not decide at this stage, the appeals are dismissed.



Colin Bishopp

Tribunal Judge



John Avery Jones

Tribunal Judge

Release Date: 6 January 2012


Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII