Mainpay Ltd v The Commissioners For HMRC

In considering both the 2010 Contract and the 2013 Contract, the FTT began by setting out the well-known principles described by McKenna J in Ready Mixed Concrete (South East)Limited v Minister of Pensions and National Insurance [1968] 2 QB 497 (“RMC”) at page 515:

These three criteria are generally referred to as mutuality of obligation, control and the “third stage”. In this case, the battleground was mutuality of obligation.

In relation to the 2013 Contract, the FTT explained, at FTT[42]:

The FTT considered what features were required in the gaps between employments for the 2013 Contract to be a continuing or overarching contract of employment. The parties disagreed on whether control must exist, but the FTT found it unnecessary to determine that question, as it decided that the necessary mutuality of obligation did not exist between assignments.

In terms of the obligation on Mainpay under the 2013 Contract, the FTT decided that there was an obligation on Mainpay to provide work, stating as follows, at FTT[126]-[128]:

HMRC seek to challenge that conclusion; we return to this below.

At FTT[130]-[136], the FTT then turned to mutuality of obligation from the perspective of the workers, after having discussed the respective submissions of the parties. It stated that the only relevant provision in the 2013 Contract was Clause 3.1, which requires the worker to “consider any suitable Assignments obtained by the Company”. It rejected Mr Firth’s “tentative” suggestion that an obligation to give good faith consideration to a suitable assignment should be implied into this provision. The FTT construed Clause 3.1 against the factual background known to the parties when they entered into the contract.

Mr Firth argued before the FTT that the position in this case was analogous to that in ABC News Intercontinental Inc v Gizbert [2006] All ER (D) 98 (“Gizbert”), in which mutuality was found to exist. The FTT rejected that argument at FTT[133]:

The FTT’s conclusion was set out at FTT[135]:

Mr Firth argued that the FTT erred in law in concluding that (1) workers had an unfettered right to refuse any assignment, and (2) there was no obligation on Mainpay to pay for a minimum number of hours of work if that work was not offered by Mainpay. These errors, said Mr Firth, led to the FTT erroneously concluding that there was no mutuality of obligation for employment purposes during the gaps between assignments.

In relation to the obligation on a worker to consider a proposed assignment, Mr Firth argued that the FTT was wrong to conclude that this obligation was not one that had to be performed in good faith. He relied in this respect on Brogden v Investec Bank plc [2014] EWHC 2785 (Comm) (“Brogden”) and Braganza v BP Shipping Limited [2015] UKSC 17 (“Braganza”). The principle that a contractual obligation must be exercised in good faith, said Mr Firth, applies to all contractual discretions unless there is clear language to the contrary.

In relation to the obligation on Mainpay, Mr Firth argued that there was in fact an obligation to pay for the stated minimum number of hours of work, because if the obligation to offer the stated minimum of hours was breached, then Mainpay would have to pay damages for that breach, and that does not need to be spelt out in the contract.

Mr Firth said that the implied requirement for employees to consider proposed assignments in good faith combined with the effective obligation on Mainpay to pay for a minimum number of hours of work meant that the position under the 2013 Contract was the same as that in Gizbert. The necessary mutuality of obligation therefore existed and the FTT was wrong to conclude otherwise.

The Upper Tribunal has recently considered mutuality of obligation in the context of an overarching contract in Exchequer Solutions Ltd v HMRC [2024] UKUT 00025 (TCC) (“Exchequer”). We agree with the following principles summarised at [11] of that decision:

Regardless of whether it “locates the contract in the employment field”, there must be a mutuality of obligation. In Cotswold Developments Construction Limited v Williams [2006] IRLR 181, Langstaff J said this, at [54]-[55] (emphasis added to original):

We consider that the FTT properly directed itself as to the law relating to mutuality of obligation, both in relation to these and other applicable principles. The FTT also referred (at FTT[78], [79] and [81]) to the guidance regarding mutuality given recently by the Court of Appeal in HMRC v Atholl House Productions Limited [2022] EWCA Civ 501. (Footnote: ³) The FTT found that there was a sufficient obligation on Mainpay in the gaps between assignments, but not on the workers. In order to succeed under Ground 1, Mr Firth must establish that, although the FTT correctly directed itself as to the applicable law, it erred in law in reaching this conclusion by erroneously finding that there was no obligation on workers to accept work, and that there was no obligation on Mainpay to pay workers in respect of the minimum number of hours of work which Mainpay was obliged to offer them.

Mr Firth accepted that there was no explicit obligation on workers to accept work under the 2013 Contract. However, he argued, a sufficient obligation arose under the 2013 Contract by virtue of the requirement for workers to consider proposed assignments in good faith considered together with Mainpay’s obligation to pay for the stated minimum number of hours worked. Those features, he said, meant that the necessary mutuality existed on the part of workers, as shown in Gizbert.

Gizbert is an unreported decision of the Employment Appeal Tribunal (the “EAT”) (UKEAT/0160/06/DM) concerning a war reporter who worked for the broadcaster ABC. One of the issues was whether during a particular period Mr Gizbert was continuously employed by ABC under an overarching contract of employment. The Employment Tribunal concluded that “there was no mutuality of obligation in respect of the assignments in that [Mr Gizbert] was not bound to accept assignments”. The EAT disagreed, describing the facts as differing materially from earlier decisions. It stated (at [21]) that under the relevant contracts the position was as follows:

The EAT concluded on this basis that “as a pure matter of construction of those written contracts the necessary mutuality was present”: [22].

It is not clear from the transcript of the EAT’s decision whether or not Mr Gizbert’s obligation to act in good faith in deciding whether to accept an assignment was explicitly stated in the relevant contracts. At [6] the decision records that under one of the contracts he was “[not] bound to accept any assignment offered”, but the terms of [21] and the reference at [22] to the construction of the written contracts might suggest otherwise. The decision in Gizbert was considered by the Upper Tribunal in Reed Employment plc v HMRC [2014] UKUT 160 (TCC) (“Reed”). The Tribunal said this, at [319]-[320] (emphasis added to original):

We fully agree with the analysis of Gizbert in Reed. Whether or not the obligation on Mr Gizbert to act in good faith which was found to exist by the EAT was an explicit contractual obligation or was implied, it was critical to the EAT’s conclusion on the facts in that appeal that ABC was found to have a contractual obligation to pay Mr Gizbert for 100 days work a year at $1,000 a day: [6] and [21] of Gizbert. As the Upper Tribunal put it in Reed, a contractual obligation to pay regardless of whether work is accepted or done implies an obligation on the employee to act in good faith in considering a proposed assignment as otherwise the employee “would receive the pay for nothing”.

We have no hesitation in rejecting Mr Firth’s argument that Mainpay was in the same position as ABC in Gizbert simply because it would be potentially liable in damages if it failed to offer the specified minimum amount of work. There is a material difference between such a potential liability for breach (which would be measured in terms of damage by conventional principles including mitigation) and what is in commercial terms a quantified and guaranteed retainer. In Exchequer the Upper Tribunal correctly referred in this context to Clark v Oxfordshire HealthAuthority as an example of when “an obligation to do the work if offered and an obligation to pay a retainer if no work was offered would be sufficient [to create mutuality]”. There was no obligation amounting or equivalent to a retainer under the 2013 Contract, meaning that the factual position is materially different to that in Gizbert. Thus, in our view, the implied obligation on a worker in considering a proposed assignment under the 2013 Contract was materially less onerous.

As we have explained, it is not clear whether the good faith obligation in Gizbert was express, or implied in light of the guaranteed payment obligation. In this appeal, we know with certainty that there was no express good faith obligation. In relation to Mr Firth’s argument that such a term must in any event always be implied into any contractual power, we consider that proposition is much too broad, and in any event has not been shown to apply in relation to the 2013 Contract.

We note that the argument was not put by Mr Firth to the FTT in the same terms. The FTT stated at FTT[130]:

The reference to business efficacy shows that the argument before the FTT was considered on the basis of conventional principles as to implied terms in a contract, as summarised in Lord Neuberger’s judgment in Marks & Spencer plc v BNP Paribas SecuritiesServices Trust Co (Jersey) Ltd [2015] UKSC 72, at [18] to [21].

Since, however, it raises a pure point of law, and Ms Choudhury confirmed that she took no procedural objection on this basis, we have considered the argument.

Mr Firth relied in particular on Brogden and Braganza. We gratefully adopt the following summary of the Braganza principle provided recently by Falk LJ in HMRC vHFFX LLP [2024] EWCA Civ 813, at [35]-[39]:

In Brogden, another case concerning a discretion regarding employee bonuses, which preceded the Supreme Court’s decision in Braganza, Leggatt J referred (at [91]) to “the now well established principle that in the absence of very clear language to the contrary, a contractual discretion must be exercised in good faith for the purpose for which it was conferred, and must not be exercised arbitrarily, capriciously or unreasonably (in the sense of irrationally)…”. Leggatt J described the rationale for that principle in relation to a discretion given to an employer.

We do not accept Mr Firth’s argument that, in effect, the Braganza principle applies to any contractual discretion unless excluded by specific wording. That is clear from Falk LJ’s summary set out above. Whether or not a contractual power must be exercised in good faith or rationally is fact-specific: as Baroness Hale said in Braganza any implied restriction on the exercise of a contractual power “may vary according to the terms of the contract and the context in which the decision-making power is given.” A good faith obligation is more likely to apply in relation to a discretion exercisable by an employer (because of what was described in Braganza as the “imbalance of power” between the parties).

In this case, particularly given the absence of any contractual obligation on Mainpay to pay a worker a minimum fee, we are clear that the obligation in Clause 3.1 of the 2013 Contract for a worker to “consider any suitable Assignments obtained by the Company” does not create the necessary mutuality of obligation on the part of a worker.

The FTT was therefore fully entitled to reach the decision which it did on this issue.

In light of our conclusion, we do not need to consider the relevance of control during the gaps in assignments, and it would be appropriate for that to be considered in a case where it makes a difference to the outcome.

Nor do we need to consider HMRC’s additional argument challenging the FTT’s decision that there was the necessary obligation on Mainpay in relation to mutuality of obligation.

The appeal under Ground 1 is dismissed.

Mainpay argued in the alternative that even if the 2010 Contract and the 2013 Contract (the “Contracts”) were not overarching contracts of employment, they nevertheless gave rise in each case to a single employment relationship. As a result, workplaces where workers carried out their assignments could not be “permanent workplaces” because they would not be workplaces for that single employment. This was the case even though the single employment was not continuous, because there were gaps between assignments.

The FTT considered this issue at FTT[138]-[150]. It set out Mr Firth’s arguments, noting that Mr Firth accepted that there was no authority to support his proposition. It stated that Ms Choudhury described the submission as “surprising”, and that in her submission “the contract between the agency and the worker is best understood as a framework agreement under the terms of which separate contracts of employment are entered into”. The FTT referred to the brief dismissal of the argument by the Upper Tribunal in Reed, and considered that the Tribunal had reached the right conclusion. The FTT noted that ITEPA linked the concept of “employment” to particular types of contract, and that on a natural reading of the statutory provisions each assignment under the Contracts, governed by a separate contract of service, was a separate employment.

The FTT concluded as follows, at FTT[149]-[150]:

The term “employment” is defined in s 4 ITEPA as follows:

Mr Firth argued as follows:

While Mr Firth deserves credit for his ingenuity and tenacity in devising and pursuing this argument, to a large extent his submissions were made to and properly dealt with by the FTT.

As a preliminary observation, we agree with Ms Choudhury’s statement in HMRC’s skeleton argument, as follows:

Put another way, we do not accept that it is inherently surprising or illogical that an umbrella contract may be capable, depending on its terms, of being an overarching employment contract, or an agency contract, or neither. Nor is it surprising or illogical that the applicable categorisation could bring with it different tax and employment law consequences.

We consider that the FTT made no error of law in determining this issue, for the following reasons.

First, while the absence of any authority supporting Mr Firth’s analysis does not mean it must be wrong, it would nevertheless be surprising if it was correct, or arguably correct, but had not been the subject of argument in any of the numerous decisions (up to and including PGMOL) relating to contracts argued by one party or another to be overarching contracts.

Second, as the FTT noted at FTT[146], it appears that the only authority which does deal with the argument, albeit in a somewhat elliptical fashion, is Reed, in which the argument was rejected. At [321]-[323] of Reed, the Upper Tribunal said this:

Third, the definition of “employment” in section 4 ITEPA, while not exclusive (“includes in particular”) focusses on particular types of contract, in this case a “contract of service”, and we agree with the FTT (at FTT[148]) that section 4 provides no support for the suggestion that a single employment may encompass more than one contract of service. In agreement with the FTT, we consider that the more natural reading of section 4 in relation to an umbrella contract such as the Contracts is that a contract of service could arise either under the contract itself, if it were an overarching contract of employment, or under separate contracts of service arising pursuant to the umbrella contract. It is not evident that the definition encompasses Mr Firth’s suggestion of a contract which gives rise to a single employment but is not itself an overarching contract of service.

Fourth, we reject Mr Firth’s assertion that the “natural” analysis of the Contracts is that they give rise to a single continuous employment, and the FTT’s and HMRC’s contrary analysis is “artificial”. The FTT correctly stated, at FTT[144]:

In relation to the Contracts, we consider that the FTT was entitled to reach the conclusion, for the reasons which it gave, that each contract was properly construed (as HMRC suggested) as a framework agreement which provided the basis on which consecutive contracts of service for individual assignments could arise.

Fifth, we do not agree with Mr Firth’s suggestion that it is somehow anomalous or unfair that Mainpay is in a worse position than if the Contracts had fallen within the definition of an “agency contract”. ITEPA provides a specific regime for certain agency contracts, and deems them to be employment contracts. That does not support Mr Firth’s submission.

Sixth, the authorities make it clear that it is perfectly possible to have a succession of short employments, so the position in relation to the issue of a P45 in relation to individual assignments is simply a function of the distinction between a contract of service and a contract for services.

Finally, we do not accept Mr Firth’s submission that it is not possible to discern any legislative purpose in distinguishing between the Contracts and an overarching employment contract. Nor do we accept his assertion that “the critical point which HMRC are still unable to explain…is why each assignment (which may be on different terms as to pay, duration, location) is part of the same employment if there happens to be mutuality in the gaps between employment, but not if there is not mutuality in the gaps” (Footnote: ⁴). The short answer is that, as we have set out above, that is what the law says, and, in our view, a dividing line which takes into account the presence or absence of mutuality in the gaps between assignments when considering an umbrella contract is entirely logical; it is the status of the contract during those gaps which determines whether there is a single overarching contract of employment.

The FTT reached the right conclusion on this issue and the appeal under Ground 2 is dismissed.

To recap, section 338 ITEPA prevents a deduction for expenses of “ordinary commuting”, which is, broadly, travel to and from an employee’s “permanent workplace”. A permanent workplace is defined by section 339(2) as “a place which (a) the employee regularly attends in the performance of the duties of the employment, and (b) is not a temporary workplace”. “Temporary workplace” is defined by section 339(3) and, by virtue of section 339(5)(a)(ii), a place is not regarded as a temporary workplace if 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.

The FTT decided that there was no overarching contract of employment and that each assignment was a separate employment. We have rejected Mainpay’s appeals against those decisions. As a consequence, each workplace at which a worker carries out an assignment is prevented from being a temporary workplace by s 339(5)(a)(ii) on the basis that the period of “the employment” is the period of the assignment and that the individual will attend the workplace for all or almost all of that period.

One might be forgiven for thinking that if each workplace was not a temporary workplace, it must necessarily be a permanent workplace, so triggering the disallowance of deductibility of travel expenses. Before the FTT, however, Mr Firth argued that this did not follow. His submission was helpfully summarised at FTT[152]-[153] as follows:

The FTT rejected this argument. It decided that (1) as a matter of statutory construction, a workplace which was prevented from being a temporary workplace was necessarily a permanent workplace, and (2) even if this was wrong, an employee who attended a workplace every day during which an employment subsisted must be taken to “regularly attend” that workplace.

By Ground 3 Mainpay appeals against the FTT’s decision on both of these issues.

Before we discuss this ground of appeal, we record that we have some concern that it might be academic, or very largely academic, on the facts of this appeal. Even if one accepts the submission that there is a “third category” of workplace, Mr Firth’s argument then relies on an employee not being in “regular attendance” for a “short assignment”. However, as the FTT noted, at FTT[168], Mainpay’s own evidence appeared to show that the average assignment lasted eight weeks, so there would be “few instances” where the workplace would not have been regularly attended. It is not part of the appeal process, including the time and resource of this Tribunal, to determine issues which are not relevant on the facts. In our view it would have been preferable if Mainpay’s application for permission to appeal on this ground had been considered against evidence of specific assignments (and corresponding expense reimbursements) said by Mainpay to be so short as not to have resulted in regular attendance.

Mainpay appeals against the FTT’s decision that, as a matter of statutory construction, there was no third category of workplace, so that a workplace which was not a temporary workplace must be a permanent workplace. However, if we were to agree with the FTT’s interpretation of the phrase “regularly attends”, that would be sufficient to dispose of this ground of appeal. Therefore, we begin by considering that question.

We proceed, for this purpose only, on the assumption that in order for a workplace to be a “permanent workplace”, section 339(2) requires that, in addition to not being a temporary workplace, the workplace must be one which “the employee regularly attends in the performance of the duties of the employment”.

Before the FTT, Mr Firth referred to the alternative meanings of “regularly” in relation to statutory obligations regarding school attendance by a child suggested by the Supreme Court in Isle of Wight Council v Platt [2017] UKSC 28. He suggested that the relevant alternative meaning in section 339(2) must be attendance “sufficiently often”, and an employee could not be said to attend a workplace sufficiently often if “only there for a day or for a week”. He suggested that attendance for a month might be “sufficiently often”. Mr Firth acknowledged that this would lead to the odd result that an employee who expected to attend a workplace for several months would nevertheless be able to claim travel and subsistence expenses for the first month of the assignment.

The FTT considered that such an anomaly demonstrated why Mr Firth’s construction could not be correct, and also referred to the problem that his construction “requires an arbitrary line to be drawn in order to determine the point at which attendance qualifies as regular”. The FTT stated as follows, at FTT[165]:

The FTT rejected Mr Firth’s argument that this could not be correct for an assignment of one day because attendance on only one day could not be “regular”.

On the meaning of “regularly attends”, Mr Firth argued as follows:

We do not agree with Mr Firth’s arguments, and we are clear that the FTT reached the right decision on the meaning of “regularly attends”.

The starting point is that, as with any statutory language, section 339(2) must be construed purposively, by reference to the words chosen by Parliament, and in context. Mr Firth’s submissions focus on the words “regularly attends” in isolation, but the wording which requires construction is “regularly attends in the performance of the duties of the employment”. It is critical to consider the wording in its entirety, and in context, because the words “regularly attends” looked at in isolation inevitably beg the question: regularly attends for what purpose?

In section 339(2), we are told by the drafter that the regular attendance must be “in the performance of the duties of the employment”. Here, in light of our dismissal of Grounds 1 and 2, “the employment” is that arising under the separate contract of service for a particular assignment. That assignment might last hours, days, weeks or months. Therefore, it follows that the FTT was correct to construe the language as referring to attendance at a workplace “every day during which the employment subsists”, even if that is only for one day. We accept that for a one-day assignment, “regular” is not the most obvious choice of word, but in that context it simply means regular attendance during that day in the performance of the employment duties.

Mr Firth’s interpretation seeks to measure and define “regularly attends” by reference either to the length of time of an assignment and/or the number of days of attendance. There is no warrant in the language (construed in its entirety) for such an approach. The measurement and definition of regular attendance in section 339(2) is by reference to the duties of the employment in question. It is entirely sensible that what it requires therefore varies with the length of that employment and the attendance necessary and is sufficiently flexible to accommodate any assignment giving rise to a contract of service.

The FTT’s interpretation, which we consider to be entirely consistent with the language in section 339(2) looked at in its entirety, avoids the anomalies and uncertainties generated by Mr Firth’s interpretation.

For our part, we do not find the discussion in Isle of Wight Council v Platt to offer any great assistance in interpreting section 339(2), since it concerns the phrase “regular attendance” in the context of completely different legislation. In any event, if one accepts Mr Firth’s suggestion that that discussion supports the interpretation of the wording in this case as meaning “sufficiently often”, that is consistent with the interpretation of the FTT: attendance is regular if it is attendance which occurs sufficiently often to perform the duties of the particular employment.

Mr Firth’s assertion that the FTT considered “regularly attends” to mean “attendance when required to do so” is not made out. The FTT refers to that formulation of “regularly attends” only at FTT[166] and only for the purpose of addressing an argument of Mr Firth regarding attendance on an assignment of a single day.

Our conclusion as to the meaning of “regularly attends” means that the appeal under Ground 3 fails. It is not necessary for us to determine the question of whether section 339 contemplates that a third category of workplace may exist which is neither temporary nor permanent as defined, and we express no view on that question.

The appeal under Ground 3 is dismissed.

In 2009, HMRC introduced a system of benchmark scale rates which employers could use to make subsistence payments to employees who had incurred such expenses. Subject to various conditions, the rates could be reimbursed without a tax liability. It is necessary under that system for employers to apply for and obtain from HMRC a dispensation.

Mainpay did not apply for or obtain such a dispensation, and the determinations and decision notices which HMRC issued to Mainpay were issued for the additional reason that scale rates could not be utilised without a statutory dispensation. Before the FTT, Mainpay argued that such a dispensation was unnecessary and it was nevertheless open to Mainpay to reimburse expenses to workers using the benchmark scale rates as long as the workers had in fact incurred some expense. The FTT rejected that argument and by Ground 4 Mainpay appeals against that decision.

Given the conclusions which we have reached on Grounds 1 to 3, it is not necessary for us to address this issue, but as it was fully argued and we can set out our conclusions fairly briefly, we will do so.

We were asked by the parties to determine this issue as matter of principle, such that if we were to agree with Mainpay’s submission, our decision would relate only to reimbursements by Mainpay to workers who did in fact incur some subsistence expenses. We have proceeded on that basis, but we note that the FTT found at FTT[193] that even if they had accepted Mainpay’s argument as to the position in principle, “we do not consider that the arrangements operated by Mainpay were a genuine attempt to reimburse subsistence expenses actually incurred”. We return to this finding below.

The normal requirement in obtaining a deduction for an expense that there is evidence that the expense was incurred was modified for the years relevant in this appeal (Footnote: ⁵), in cases where a dispensation agreement is reached with HMRC under section 65 ITEPA. So far as relevant this stated as follows:

Mr Firth argued that the ability of an employer to set up an arrangement which is intended to provide a genuine reimbursement of expenses which are then deductible was supported by section 65 itself and by the decision of the Court of Appeal in Cheshire Employer and SkillsDevelopment Limited v HMRC [2012] EWCA Civ 1429 (“Cheshire”). The FTT did not accept those arguments. It concluded as follows, at FTT[183]-[189]:

In this appeal, Mr Firth argued as follows:

In short, we consider that the FTT reached the right decision on this issue, for the reasons which it gave. We respectfully agree with the FTT’s reasoning and conclusions at FTT[183]-[189] set out above, and the FTT made no error of law in reaching those conclusions.

Mr Firth’s submissions in this appeal reformulated and, to a significant degree, simply repeated the arguments put to the FTT, rather than identifying what are said to be errors of law in the FTT’s decision. The relevant passages of Mr Firth’s skeleton argument do not discuss the FTT’s decision or reasoning at all.

Indeed, it is not clear that the two arguments formulated by Mr Firth would serve to identify any error of law in the FTT’s decision on this issue. Mr Firth’s first proposition is that it is possible to make a reasonable estimate of expenditure and still claim a deduction for it. That may be so, but the specific argument addressed and decided by the FTT was that “it is open to an employer to come up with an arrangement which is intended to provide a genuine reimbursement of expenses to an employee and that, as long as the arrangement does not contain any profit element, the amount is not taxable (or, more accurately, is able to be deducted) irrespective of whether it precisely equals the expenses actually incurred”: FTT[175]. That is a much narrower question than the broad statement of principle formulated before us. The second proposition is that section 65 only permits HMRC to confirm that a payment which is not taxable is, in fact, not taxable. However, that proposition was accepted by the FTT in the course of its reasoning, at FTT[186].

Further, although we have not taken this into account in reaching our decision, while it may be acceptable to determine an issue as a matter of principle, with the consequences for quantum then agreed between the parties or remitted to the FTT, in this case there is no challenge to the FTT’s finding of fact that that “the arrangements operated by Mainpay were [not] a genuine attempt to reimburse subsistence expenses actually incurred”. It is not apparent what purpose would be served by a decision in principle on Mr Firth’s two propositions in this appeal in light of this finding.

The appeal under Ground 4 is dismissed.

The assessments and notices issued by HMRC to Mainpay for the tax years ended 5 April 2011 and 5 April 2012 were issued more than 4 years after the end of the relevant tax year. In those circumstances, they would have been out of time unless the extended 6-year time limit in section 36 TMA applied. Section 36(1) TMA provides as follows:

Section 118(5) TMA states as follows:

This issue related only to the 2010 Contract, since that was the contract which was relevant to the assessments for the tax years ended 5 April 2011 and 5 April 2012.

Before the FTT, Mr Firth argued that there was no failure to take reasonable care on the part of Mainpay in view of the advice which it took. He also criticised HMRC’s statement of case as being vague on this issue and making no pleading at all in relation to causation (ie how HMRC alleged any carelessness by Mainpay “brought about” the loss of tax): FTT[198].

Having considered the position and the advice taken by Mainpay, the FTT rejected the complaint as to pleading and concluded that Mainpay did fail to take reasonable care and that this resulted in a loss of tax: FTT[207]. As a result, the six-year time limit in section 36 TMA for making the assessments and determinations applied, and they were as a result validly issued.

In this appeal, Mr Firth made the following arguments in relation to Ground 5:

HMRC’s pleading of carelessness

Before the FTT, this complaint was dealt with at FTT[198]-[200]:

Mr Firth argued in this appeal that HMRC’s pleading was “totally inadequate”. He said that a taxpayer faced with an allegation of carelessness is entitled to know what exactly it is alleged that it should have done and why it is said that it failed to do so, because only in those circumstances can a taxpayer prepare to meet the case against it. That is the whole purpose of a statement of case. He described HMRC’s pleading as “so vague as to be meaningless”. Further, it contains no explanation of the causation issue ie how the careless behaviour brought about the loss of tax. He submitted that, in any event, the FTT’s conclusion on the pleading issue was wholly unreasonable.

Ms Choudhury pointed out that the question of whether HMRC ought to have been permitted to rely on their pleading in relation to carelessness was essentially a case management decision, and, as such, this Tribunal should be very slow to interfere with it. She also noted that at no stage prior to the hearing had Mainpay sought further particulars on this issue. As to the pleading regarding causation, Ms Choudhury said the FTT was correct to hold that it was implicit that it was the failure to take reasonable care to ensure that the 2010 Contract was an overarching contract of employment which caused the loss of tax. That was also clear, she argued, from the wording of section 118(5) TMA.

Consistently with the general principle that a party must be able to understand the case against it, Rule 25(2) of the Tribunal Procedure (First-tier Tribunal ) (Tax Chamber) Rules 2009 provides as follows:

Sub-paragraph (b) of Rule 25(2) means that the statement of case should set out the case which the appellant is expected to meet. In HMRC v Ritchie [2019] UKUT 71 (TCC), the Upper Tribunal said that it should set out HMRC’s arguments “clearly and unequivocally with sufficient detail” (Footnote: ⁶). In relation to the validity of an assessment under section 36 TMA, HMRC bear the burden of proof: Burgess v HMRC [2015] UKUT 578 (TCC) (“Burgess”). The practical effect of this is that HMRC should address the relevant validity issues in their statement of case regardless of whether the appellant has expressly challenged them.

Where, as in this case, the burden was on HMRC to establish that Mainpay had brought about the loss of tax by a failure to take reasonable care, the evaluation of whether or not HMRC should be permitted to rely on their pleadings in relation to carelessness involved an exercise of judicial discretion by the FTT. This Tribunal should be slow to interfere with such a decision unless the FTT applied the wrong principles, took account of the wrong factors, or otherwise reached a decision so plainly wrong that it must be regarded as outside the generous ambit of the discretion available to the FTT.

HMRC’s consolidated Statement of Case relevantly stated as follows:

Mr Firth challenged as contradictory in this passage the reference to the contract reflecting what the workers would be doing. We think that this can readily be understood in context by the earlier statement of HMRC’s position at paragraphs 85 and 86 of the Statement of Case:

In relation to the question of whether HMRC’s pleading identified the relevant carelessness sufficiently for Mainpay to know what case it had to answer, we consider that the FTT reached a decision which was within the range of reasonable decisions available to it. As to the applicable test, the second sentence of paragraph 91 of the Statement of Case summarised in general terms what HMRC would seek to establish, namely looking at what Mainpay did or failed to do and asking whether a prudent and reasonable person would have done that or failed to do that in those circumstances. If nothing more had been said, then the pleading would in our view have been deficient. However, paragraph 91 went on to identify ways in which or reasons why HMRC asserted that Mainpay had failed to take reasonable care not to bring about the loss of tax. It stated that Mainpay (1) had failed to take reasonable care to ensure that the 2010 Contract was an overarching contract of employment, (2) had failed to take steps to ensure that the contact was a reflection of what the workers would be doing, and (3) had failed to confirm the tax treatment of the reimbursements with qualified personnel.

Read in the context of the Statement of Case in its entirety, we consider that this did justify the FTT’s decision, set out above, to permit HMRC to rely on its pleading. Mr Firth was critical of the factors identified by HMRC in its Statement of Case and argued that some issues were directly contradictory to others, but that is beside the point. The question for the FTT on this point was whether Mainpay was sufficiently aware of the case which it had to answer, not whether that case was strong or weak.

We agree with the FTT that there was some force in Mr Firth’s criticisms before the FTT. However, we also see some force in Ms Choudhury’s observation that at any stage prior to the hearing Mainpay could have made an application to the FTT for HMRC to be directed to provide further and better particulars, but did not do so.

Mr Firth said that the FTT did not in fact limit the carelessness issue to the overarching contract question, because at FTT[209] it criticised Mr Hugo in relation to information regarding particular expenses and their reimbursement. However, that paragraph in its entirety is responding to Mr Firth’s argument that one of the aspects demonstrating reasonable care taken by Mainpay was the advice taken from Mr Hugo. In context, it is part of the FTT’s discussion of the overarching contract issue. At FTT[206], to take one example, the FTT dismissed an argument raised by HMRC in relation to the use of scale rates as irrelevant because it was not part of the case put forward by HMRC in relation to carelessness.

Therefore, we do not accept Mr Firth’s argument on the pleadings issue. The complaint that HMRC did not particularise the issue of causation is best addressed in the context of the argument that the FTT erred in finding that Mainpay’s carelessness brought about the loss of tax. Therefore, we deal with this under issue (3) below.

Failure to take reasonable care

At FTT[196]-[197], the FTT stated as follows:

Mainpay had also taken advice from Mr Hugo, who the FTT described as “a consultant who advised Mainpay in respect of accounting and finance, including the operation of PAYE”: FTT[20]. The FTT summarised Mr Hugo’s discussions with Mishcon beginning in November 2007, noting that the individuals he was dealing with were members of Mishcon’s employment group. The FTT set out the parties’ arguments as to whether there had been a failure to take reasonable care on the part of Mainpay. Its conclusion, at FTT[207], was that “Mainpay did fail to take reasonable care and that, as a result of this, there was a loss of tax”.

The FTT explained its conclusions in relation to the advice taken by Mainpay at FTT[208]-[211]:

Stepping back from the detail, Mainpay’s primary argument was that it had taken reasonable care because it had taken appropriate advice. What the FTT essentially did was to test that proposition, by scrutinising what advice was given, by whom, with what qualifications, and on the basis of what information, assumptions and draft documentation. Having carried out that exercise, the FTT effectively concluded that the advice was deficient, because none of Mishcon, Mr Hugo or Dr O’Brien had been asked the right questions, with sufficient clarity and provided with the necessary draft documents, to advise Mainpay whether the 2010 Contract and the proposals were effective to create a single employment, which was the essence of the arrangement.

Mr Firth challenged virtually every aspect of the FTT’s reasoning and conclusions on this issue, as set out above. However, we do not agree that the FTT made any error of law in reaching its decision. The extent to which taking advice provides a reasonable care defence depends on the facts, in particular whether the relevant questions were the subject of advice by someone with appropriate expertise who had knowledge of all the relevant facts, including the relevant documentation.

What are described as “errors of law” by Mr Firth in relation to the FTT’s reasoning are on examination not errors of law but disagreement with that reasoning or the weight afforded by the FTT to different factors. In particular, Mr Firth raised the following objections:

In conclusion, we consider that the FTT was entitled to find that Mainpay had failed to take reasonable care.

Causation

As we have mentioned, Mr Firth argued that HMRC had failed adequately to plead its position on the causation issue. He also argued under Ground 5 that the FTT had erred in law in finding that Mainpay’s failure to take reasonable care had brought about the loss of tax. Both of these arguments relate to what is necessary for HMRC to meet the burden of proof that the relevant failure to take reasonable care has “brought about” the loss of tax.

The FTT set out its conclusions in relation to the causation issue at FTT[212]-[213]:

Mr Firth argued in this appeal that HMRC failed to explain at all or in sufficient detail:

Mr Firth said that the FTT “had to create its own case” on these questions, and HMRC’s pleading on the issues was wholly inadequate. He said that the FTT’s decision that Mainpay’s carelessness brought about the loss of tax also contained the following errors of law:

Ms Choudhury challenged all of these arguments, and advanced an additional “knockout” argument. This was that because section 118(5) TMA provides that “a loss of tax or a situation is brought about carelessly by a person if the person fails to take reasonable care to avoid bringing about that loss or situation”; thus, once a failure to take reasonable care is established, it inevitably follows that the taxpayer brought about the loss or situation. Ms Choudhury relied in this respect on the Upper Tribunal’s decision in Atherton v HMRC [2019] UKUT 41 (TCC) (“Atherton”). As she put it in her skeleton argument, given section 118(5) “there is no need to establish a separate causal link between the alleged carelessness and the loss of tax”. In her oral submissions, Ms Choudhury argued that there was support for this interpretation in the changes made to the relevant statutory provisions in 2008. She said that as result of these changes, causation could not be approached as a “but for” test.

In Atherton, one of the issues was whether the insufficiency in Mr Atherton’s tax return was brought about by the carelessness of Mr Atherton or a person acting on his behalf. Mr Atherton argued that the insufficiency in his return was attributable only to the relevant numerical entry in a box on the form, and that the lack of explanation of that entry had no causal effect, because even if a full explanation of the use of the relevant box had been provided in the white space on the tax return, the insufficiency would still have arisen by virtue of the presence of the figure in that box: Atherton at [57]. The Upper Tribunal considered that “as a matter of legal causation…there appears to be force in Mr Atherton’s argument”, but that section 118(5), which had not been referred to by the FTT, had made specific provision for this issue. The Upper Tribunal then said this:

Bella Figura was concerned with a pension scheme which had made a loan to a company called Falken Ltd. HMRC had decided that the loan was an “unauthorised payment” for pension scheme purposes and a “scheme chargeable payment” and assessed Bella Figura to a scheme sanction charge, an unauthorised payments charge and an unauthorised payments surcharge. Bella Figura's response was that it was not careless since it had been advised by a firm of pension administrators (PPCL) that the loan was not an unauthorised employer payment or a scheme chargeable payment. The FTT held that HMRC could use its extended time limits since Bella Figura had not discharged its burden of showing that supportive advice had actually been received from PPCL. The Upper Tribunal stated, at [61(2)]:

  After considering another issue, the Upper Tribunal went on to allow the appeal, in part because, as explained at [85(4)]:

There is no escaping the fact that Atherton and Bella Figura are very difficult to reconcile in their approach to the causation issue. Bella Figura was decided over a year after Atherton, but does not refer to that decision. We note that in Strachan v HMRC [2023] UKFTT 617 (TC), the FTT directly addressed the two conflicting decisions, and decided that because the burden of proof did lie on HMRC to establish causation^, this meant that HMRC effectively needed to show what the taxpayer’s advice would have been if reasonable care had been taken (Footnote: ⁷). Contrastingly, in Delphi Derivatives Ltd v HMRC [2023] UKFTT 722 (TC), the FTT addressed the need for a “causal link”, and the difference which a second opinion to the taxpayer would have made in this context. The FTT decided that a “but for” test was inappropriate, and refused to make a finding as to what such an opinion might have said, as that would have been “pure speculation”: [248] of that decision.

In determining the position, and addressing these conflicting decisions, we start from the proposition that HMRC bear the burden of proof in relation to establishing that the careless behaviour brought about the loss of tax. That much is clear from Burgess, and the decision in HMRC v Household Agents [2007] EWHC 1684 (Ch) which was relied on in Burgess. For our part, we do not read Atherton as disputing that proposition; when the Upper Tribunal said at [12] of that decision that “the burden of proof is on HMRC to establish on the balance of probabilities that the discovery assessment was validly made”, it was doing so by reference to its citation in the preceding paragraph of the statement in Burgess regarding the “relevant conditions” for assessments outside the normal time limits.

The more difficult question is what HMRC need to do to discharge that burden, and whether they did enough in this case.

We accept that it is possible to read the passages we have referred to in Atherton as meaning that once HMRC have established careless behaviour and a loss of tax, they need do nothing more to establish causation, simply because of the wording of section 118(5). If that is what was being said, then we respectfully disagree. Nor do we accept Ms Choudhury’s argument that section 118(5) can be said to have such an effect by reference to the changes in statutory terminology in 2008. Such a reading would have the result that in practice there was no additional burden of proof on HMRC in relation to causation, and that cannot in our opinion be right.

What the Upper Tribunal was pointing out in Atherton was that the important word to focus on in section 118(5) was “avoid”. If a taxpayer failed to take reasonable care to avoid the loss or situation, then they would have brought about the loss of tax. The point being emphasised was that that was not the same as a common law test of causation. However, the Upper Tribunal was not, in our view, saying that HMRC did not bear the burden of establishing that different test of causation. As we have said, if it was, then we would disagree.

As to what precisely HMRC must show in order to establish that a taxpayer has failed to take reasonable care to avoid the loss of tax, that depends entirely on the facts. To the extent that the carelessness which has been found relates to a deficiency in advice taken by a taxpayer, we consider that HMRC do need to establish that the relevant deficiency could have been avoided by the taxpayer. How it could have been avoided will depend on the facts and on the nature of the deficiency. If, for instance, no advice was taken on a material issue, then the carelessness could have been avoided by taking advice on that issue. If the advice was from an adviser on whom it was unreasonable for that taxpayer to rely, then it could have been avoided by taking advice from a qualified adviser. If the relevant advice did relate to a material issue but was given on the basis of incomplete information or documentation, or of assumptions found not to be applicable in practice, then it could have been avoided by supplying the information or documentation or clarifying the position in relation to the assumptions.

To that extent, HMRC does in our view need to show what the taxpayer “should have done differently”. Insofar as Bella Figura endorses the position we have just set out, we agree with its approach.

However, we do not consider that the burden on HMRC carries with it an obligation to prove a particular counter-factual outcome. In particular, in relation to carelessness said to result from deficient advice, it does not require HMRC to establish to the balance of probabilities what the result of remedying the deficiency would have been. For example, it does not require HMRC to establish that if the deficiency had been remedied, a failure in the arrangement to achieve its intended tax purpose, which led to the loss of tax, would have been remedied. Not all problems (or arrangements or schemes) can be “fixed”, and the FTT should not be placed in the position of having to speculate as to a taxpayer’s precise response to advice to a standard of reasonable care. If the Upper Tribunal in Bella Figura was indicating to the contrary, then we respectfully disagree.

We should make clear that nothing in section 118(5) removes the need for a connection (to use a more neutral term) between the carelessness and the loss of tax. It is not enough for HMRC to show that the taxpayer was careless; that carelessness must have been a failure to avoid bringing about the loss of tax. So, for instance, a general lack of reasonable care by a taxpayer in completing a tax return or in keeping records which did not contribute to the loss of tax is not enough.

In this case, we consider that the FTT was entitled to find as it did at FTT[212]-[213] in relation to the causation issue. Some of Mr Firth’s specific criticisms are adequately dealt with by the FTT at FTT[213]-[215]. More generally, the FTT was entitled to conclude that it was Mainpay’s failure to take reasonable care in ensuring that the 2010 Contract was an overarching contract of employment which caused the loss of tax. On the facts found, what Mainpay “should have done differently” was to have asked an appropriately qualified adviser whether the 2010 Contract as drafted (including a provision stating that it was not a contract of employment) was an overarching contract of employment, or otherwise effective to achieve the aim of the arrangements in relation to reimbursed expenses. In our view, the FTT was not required, by Bella Figura or otherwise, to put HMRC to proof of establishing what course of action Mainpay would have taken if that had been done, namely whether it would have amended the contract, introduced a retainer in an effort to create mutuality in the gaps, or decided not to claim the deductions which gave rise to the loss of tax.

It follows from what we have said that we do not accept Mr Firth’s argument that HMRC’s pleading was inadequate. The FTT’s conclusion at FTT[212] on this issue was reasonably open to it.

Reliance on Mr Hugo

Mr Firth argued that “the FTT wholly failed to deal with the Appellant’s submission that Mr Hugo was an independent contractor and it was reasonable for Mainpay to rely on him to obtain the appropriate advice”. The FTT described Mr Hugo as “an accountant with significant experience in the operation of umbrella companies…well aware of the complexities of ensuring that expenses could be reimbursed on a tax free basis”: FTT[208]. It followed, he said, that it was reasonable for Mainpay to rely on Mr Hugo to ask the right questions of Mishcon and obtain the right form of contract.

We do not agree. The FTT clearly understood that Mr Hugo was not an employee of Mainpay but an accountant who acted as a consultant. Like most advisers (including Mishcon) he was independent of Mainpay. It described him, at FTT[20], as “a consultant who advised Mainpay in respect of accounting and finance, including the operation of PAYE”. The mere fact that Mr Hugo was a consultant does not, however, establish that Mainpay took reasonable care in relying on his advice, any more than it did from the mere fact that Mishcon was a respected law firm. We agree with HMRC that it is not sufficient for Mainpay simply to say that it was relying on an adviser when it did not establish that Mr Hugo had himself given advice of his own as to the employment status of the 2010 Contract, including the provision in the 2010 Contract stating that it was not an employment contract. The FTT found that notwithstanding the absence of such advice Mainpay had failed to ask Mr Hugo, Mishcon or anyone else for advice which was crucial to the effectiveness of the proposed arrangements. The FTT was entitled to find that that was a failure to take reasonable care to avoid bringing about the loss of tax.

The appeal under Ground 5 is dismissed, and the FTT’s decision that the determinations and decisions for the relevant tax years were made within the relevant statutory time limit in section 36 TMA is upheld.

The appeal is dismissed.