CA-2024-002400 - [2025] EWCA Civ 1290

Discussion

Ground 1

Mr Firth agreed, as he had to, that a court’s classification of a contract is a result of an analysis of the legal effect of all of the terms of contract, and does not wholly depend on the parties’ intention (as expressed in the terms which they have agreed in the contract). The effect of his submissions was nevertheless that he insisted that, in this case, all that mattered was that the parties had agreed that there was one contract between them and that that one feature of their express joint intention somehow overrode the court’s independent classification of the contract. The hidden premise of his argument is that the parties can, if they label a contract in a particular way, negate the court’s duty to classify the effect of the contract as a whole. That submission is contrary to decisions like Street v Mountford [1985] AC 809 and the long sequence of employment cases referred to, for example, in this court’s decision in PGMOL. It follows that his agreement, which I recorded in the first sentence of this paragraph, was more apparent than real.

I readily accept that the relationship between the parties to this contract is, at times, to be classified as a contract of employment, and at other times, not. It does not follow from that intermittent classification as a contract of employment, coupled with the fact that the relationship between the parties is governed by the same contract throughout, that the contract is to be or can be classified as a continuing contract of employment throughout its life. It cannot be classified in that way, because in the gaps between assignments, there is no contract of employment at all. The correct analysis is that there is intermittent employment under a contract of employment when a worker is on an assignment, followed by periods when there is no contract of employment, and, therefore, no employment at all, in the gaps between assignments. That analysis is consistent with what the parties have agreed, save that they cannot agree that there is a ‘continuing’ or overarching contract of employment when the correct legal analysis is that there is no contract of employment at all during those gaps between assignments. The intermittent periods of employment are not, for the purposes of the relevant statutory provisions, one employment. They are successive employments. The F-tT and the UT were right so to hold.

Ground 3

Ground 3 raises three issues.

Is there a causal requirement, and if so, what is it?

If there is, did the F-tT err in law in its application of that requirement to the facts it found?

Was the F-tT entitled to hold that HMRC’s pleading was adequate?

Is there a causal requirement?

The first question is what test is imposed by sections 36 and 118(5) of the TMA. There are at least two parts to this question. The first is whether there is a causal requirement at all. The second is, if there is, what it entails in a case like this.

HMRC tried to persuade the UT that the words of section 118(5) of the TMA were a complete answer to Mainpay’s argument on causation. The UT was not persuaded. HMRC did not cross-appeal against that conclusion. It is not, therefore, open to HMRC to revive that argument on this appeal. For the avoidance of doubt, I consider, in any event, that that argument is wrong, as I will now explain.

Section 36 is not the only relevant provision which uses the formula ‘brought about carelessly’. The phrase is also used in section 29(4). The phrase must mean the same in section 36 as it means in section 29(4). Section 29(1) enables the tax authorities to make an assessment when they discover that income which ought to have been assessed to tax has not been so assessed, and to correct that by making an assessment of the amount or further amount which ought to have been assessed to tax. They may not do so if an error or mistake in the return was made on the basis of or in accordance with a generally prevalent practice (section 29(2)). They may only make such an assessment if one of two conditions is met. The first is that the ‘situation’ described in section 29(1) has been ‘brought about carelessly’ (or deliberately) by the taxpayer or his agent. The phrase ‘brought about’ is a synonym for ‘caused’.

The heading of section 36 of the TMA is ‘Loss of tax brought about carelessly…’. Section 36(1) extends the relevant time limit ‘in a case involving a loss of income tax…brought about carelessly by the person…’ The phrase ‘brought about’ is, again, a synonym for ‘caused’. Section 118(5) must be read as a whole. It simply repeats the phrase ‘brought about’. That makes it clear that section 118(5) endorses, and does not qualify, still less contradict, the apparent meaning of the phrase ‘brought about’ in section 36(1). The focus, rather, is explaining what ‘carelessly’ means for the purposes of the TMA. Section 118(5) explains that there is the relevant lack of care if ‘a person fails to take reasonable care to avoid bringing about that loss or situation’.

Paragraph (2) of the headnote of R (PACCAR) v Competition Appeal Tribunal [2023] UKSC 675; [2023] 1 WLR 2594 articulates a general rule of statutory construction, that when a statutory definition is read as a whole, the ordinary meaning of the word of phrase being defined can be used to throw light on the meaning of the phrase as defined. I have reached a clear view about the meaning of sections 29 and 36(1) without resorting to this rule, but if that view requires support, this general rule provides it. If any further support is needed, I rely on Mr Firth’s submission that if section 118(5) has the meaning for which HMRC contended in the UT, it had a profound and unheralded effect on the position under the previous legislative provisions, which clearly required a causal link between the loss and the acts or omissions of the taxpayer. Such a change would have required very clear language. There are no words to show that any such causal link has been removed by section 118(5). On the contrary, the words ‘brought about’, which are used in the heading and in the text of section 36, and in section 118(5) show the very opposite.

What does the causal requirement entail in this case?

There are two broad issues. The first is to identify the ‘loss’ of income tax referred to in the heading of section 36 and in section 36(1). It must be the same ‘loss of tax’ as is referred to in section 29(1). The relevant ‘loss of tax’ occurs for the purposes of both provisions if income ought to have been assessed to tax and has not been.

The second issue is how section 36(1) should be applied. On the arguments in this case, the F-tT had to decide five sub-issues:

whether the taxpayer did something which was careless

if so, what that was;

whether that carelessness brought about the loss of tax;

where burden of proof lies; and

whether HMRC’s pleading was adequate.

Did the F-tT apply the test correctly to the facts it found?

The F-tT’s reasoning was that Mishcon de Reya’s advice made it clear that the 2010 contract was not intended to be a contract of employment. There was little evidence about what prompted the change to the 2013 contract. Mr Hugo’s evidence what that the sector was constantly evolving and Mainpay therefore took advice on ‘an ongoing basis’. Mainpay said initially that it automatically paid the standard subsistence allowance, and told workers that it was ‘able to account for a proportion of’ their income as ‘subsistence Allowance’ which would be ‘tax-free’. They were told this was ‘automatic’ and that they were not asked to send in supporting evidence. After the changes to the amounts of allowance, workers were told that they would get £5 automatically, and that if they wanted to claim £10 all they had to do was log on and change the £5 to £10. They were not required to post any evidence. Subsistence allowances were claimed in about 90% of cases.

The F-tT considered (see paragraphs 36 and 37 above) what basis, if any, there was for the subsistence rates used by Mainpay. The F-tT did not consider that ‘the arrangements operated by Mainpay were a genuine attempt to reimburse subsistence expenses genuinely incurred’ (paragraph 56, above). The F-tT recorded Mainpay’s argument about causation (paragraph 58, above).

The F-tT found, first, that Mainpay had been careless. The point about the advice sought from, and given by, Mishcon de Reya, is that its premise was that the expenses in question were deductible business expenses (see paragraph 66, above). Mr Hugo knew that the question whether expenses can be reimbursed tax-free is itself complicated (see paragraph 66, above). It was not ‘reasonable’ for him to rely on a vague assurance that the form of the contract would not affect Mainpay’s ability to do reimburse expenses. He could not have expected Mishcon de Reya to do that, since they did not have the relevant information (paragraph 67, above). The fact that Mainpay had a different adviser who advised on expenses and who did not appear to have been consulted on that point confirmed the F-tT’s view that it was not reasonable to rely on vague assurances from Mishcon de Reya, when they did not have full background facts (paragraph 69, above).

The F-tT having found that Mainpay were careless then had to consider whether that carelessness brought about the loss of tax. I have italicised, in my summary in paragraphs 57-71, above, some of the passages in decision 1 which show that the F-tT clearly understood that sections 36(1) and 118(5) require a causal link between carelessness of the taxpayer and the loss of tax. In the relevant paragraphs, the F-tT quoted the statutory words four times, and used various synonyms, including ‘cause’ (and cognate words), ‘as a result of’, and ‘led directly to’.

There is nothing in decision 1 which suggests that the F-tT having considered whether, and having decided that, Mainpay failed to take reasonable care to avoid the loss of tax, the F-tT did not then apply a test of causation in deciding that that failure to take reasonable care did bring about, or cause the relevant loss of tax. On the contrary, the F-tT did apply a causal test. Its conclusion was that it was ‘clear that the failure to take reasonable care to ensure that the contract in question was an overarching contract of employment led directly to the loss of tax as a result of Mainpay treating the expenses as deductible when in the absence of an overarching contract, they were not’ (paragraph 71, above).

The loss of tax in this case was that part of the workers’ pay was treated as tax-free when it should have been subject to the deduction of tax. On the F-tT’s findings, it is obvious that had Mainpay taken reasonable care, the contracts would have been overarching contracts of employment. If they had been overarching contracts, the reimbursement of those expenses would not have been liable to tax (subject to the issue raised by ground 2). Mainpay did not take reasonable care to ensure that the contracts were overarching contracts. Mainpay nevertheless reimbursed the expenses free of tax, as if the contracts were overarching contracts, when, in law, those payments were liable to tax. Had Mainpay taken reasonable care, therefore, on the F-tT’s findings, that loss of tax would have been avoided.

On these particular facts, HMRC had, in the words of Mr Firth, ‘done enough’. I do not consider that it was necessary for the F-tT to make any more findings about what would have happened if Mainpay had taken reasonable care. The F-tT nevertheless considered what would have happened if Mainpay had asked for specific advice (paragraph 71, above), even though that was not necessary on these facts. I do not consider that the F-tT was required to speculate about what might have happened if further advice had been sought, all the more so because a taxpayer cannot be required to waive legal advice privilege, so that the F-tT would not necessarily have and in this case did not have all the relevant evidence. I agree with the UT’s analysis in paragraphs 159 and 161 (see paragraphs 87 and 89, above). That is reinforced by what I say in the next paragraph about the burden of proof.

The burden of proof

It is not in dispute that HMRC has the burden of proving that section 36(1) applies (as the UT accepted in paragraph 153 of decision 2). On the facts of this case HMRC had made out a prima facie case that Mainpay had been careless, and that that carelessness had brought about a loss of tax. There was then an evidential burden on Mainpay, if it wished to contradict that prima facie case, to adduce evidence to show, on the balance of probabilities, that it had taken reasonable care, and/or that any lack of care did not bring about the loss of tax. Mainpay did not do that. Mainpay argues, however, that it was not sufficiently informed about HMRC’s case to understand it and therefore appropriately to rebut it. That is the next issue which I must consider.

Was HMRC’s pleading adequate?

The F-tT decided that HMRC’s pleadings, taken as a whole, and properly understood, did raise the issues in a way which enabled Mainpay to know what case it had to meet. The F-tT considered that the statement of case and skeleton argument showed that the carelessness related to ‘the question whether or not the subsistence expenses were deductible at all (on the basis that the contract was not an overarching contract of employment)…’ (paragraphs 58 and 59, above). The F-tT also considered that it was implicit in that way of putting the case that it was the failure to take care to ensure that the contract was an overarching contract of employment caused the loss of tax, given HMRC’s case that the reimbursement of expenses was not deductible if the contract was not an overarching contract of employment (paragraph 70, above).

This issue is a partial exception to my statement in paragraph 16, above. I consider that the UT’s reasons for upholding the F-tT’s decision about the adequacy or otherwise of HMRC’s pleading are relevant to the task of this court. I agree with the UT that the F-tT considered this argument carefully. I also agree with the UT that the F-tT’s view about whether or not Mainpay knew what case it had to meet on its appeal was, in the first instance, similar to a case management decision by the F-tT. The threshold for intervention by the UT (let alone for this court), is commensurately high. I agree with the UT that it is not met.