England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE UPPER TRIBUNAL
(TAX AND CHANCERY CHAMBER)
[2018] UKUT 236 (TCC)

		Royal Courts of Justice Strand, London, WC2A 2LL
		22/01/2020

Lord Justice Henderson:

Introduction

1. In 1988, Parliament legislated to unify the rates of tax on income and capital gains: see section 98 of the Finance Act 1988. In principle, that then remained the position for 20 years until 2008, when a single rate of capital gains tax ("CGT") of 18% was introduced with effect from the tax year 2008/09.
2. This case concerns a relief from the higher rates of income tax, usually known as corresponding deficiency relief ("CDR"), in the tax years 2006/07 and 2007/08. At that time, there was no upper limit to the amount of CDR which a taxpayer could use, in the year in which it arose, to set against his income for that year which would otherwise have been chargeable to higher rate income tax. The effect of the relief was that the income in question was instead charged to income tax at the basic rate. The issue, in short, is whether the taxpayer was also entitled to a similar reduction in the amount of his chargeable gains for the same year which would otherwise have been liable to CGT at the higher rate of 40%, with the result that those gains were instead charged to CGT at the lower rate of 20%.
3. The taxpayer is Mr Andrew Scott. In 2006/07 he had chargeable gains of approximately £8.84 million, and in 2007/08 of approximately £14.71 million. On the assumption that his liability to CGT at the higher rate could not be reduced by CDR in the same way as his liability to higher rate income tax was admittedly reduced, the Commissioners for Her Majesty's Revenue and Customs ("HMRC") issued closure notices on 13 February 2015 requiring Mr Scott to pay CGT of approximately £9.42 million for the two years.
4. Mr Scott's appeals to the First-tier Tribunal ("the FTT", Judge Ashley Greenbank, [2017] UKFTT 385 (TC), [2018] SFTD 42) and thence to the Upper Tribunal ("the UT", Nugee J and Judge Nicholas Aleksander, [2018] UKUT 236 (TCC), [2018] STC 1589), were dismissed. Mr Scott now appeals to this court, with permission granted by Lewison LJ.
5. For the reasons which follow, I consider that the FTT and the UT came to the right conclusion. Accordingly, if the other members of the court agree, Mr Scott's appeal will be dismissed.

Background

(1) The statutory regime for taxing gains on life assurance policies

6. Gains arising from policies of life assurance are charged to income tax under Chapter 9 of Part 4 of the Income Tax (Trading and Other Income) Act 2005 ("ITTOIA 2005"), which runs from sections 461 to 546. For the purposes of Chapter 9, a gain is treated as arising on the occurrence of specified "chargeable events", including the surrender or part surrender of the policy: sections 462 and 484. The amount of the gain is calculated in accordance with a detailed and rigid set of rules, contained in sections 491 and following. It has often been observed that those rules are capable of operating in a way which produces apparently arbitrary results contrary to commercial reality: see, for example, Mayes v Revenue and Customs Commissioners [2011] EWCA Civ 407, [2011] STC 1269, at [38] and [49] (Mummery LJ) and [102] to [104] (Toulson LJ), referring to the legislation as it stood before its restatement with minor changes in ITTOIA 2005.
7. In particular, as Mummery LJ noted in Mayes at [49]:

"The legislation operates in a way that encourages the retention of life policies as long-term investments. If a large part of the value of a policy is surrendered in the early years, disproportionately large gains will be attributed."

8. An individual is liable for income tax under Chapter 9 if he is the beneficial owner of the rights under the policy, and UK resident in the year when the gain arises: section 465(1), (2). The charge to tax is at the higher rate only, because section 530(1) confers a credit for basic rate tax on the chargeable amount. This credit reflects the fact that the insurance company is in principle liable for tax at the basic rate on the investment gain attributable to the policy which accrues while it remains in force.
9. Section 539 of ITTOIA 2005 is headed "Relief for deficiencies". Its purpose (broadly stated) is to confer relief from higher rate income tax in circumstances where a final chargeable event gives rise, not to a gain, but (by application of the same computational rules) to an arithmetical loss, or "deficiency". This is likely to happen, for example, on the surrender or termination of a policy, after an earlier part surrender has generated a heavily "front-loaded" gain of the type described by Mummery LJ in Mayes. The earlier gain is allowed as a deduction in the computation of the deficiency under section 541, and if the amount realised on termination is comparatively small (because the part surrender exhausted much of the value of the policy), the deficiency may be correspondingly large.
10. Since the provisions of section 539 are of central importance to this appeal, I will set out the section as it stood in 2006/07. Ignoring the complication of the charge to income tax at the lower rate on income from savings and distributions, which in this context can be ignored, the section reads:

"539 Relief for deficiencies

(1) A deficiency from a policy or contract arising on a chargeable event is allowable as a deduction from an individual's total income for a tax year if, had a gain arisen instead on that event –

(a) the individual would have been liable to income tax on the gain for that year, or

(b) the individual would have been so liable apart from the requirement in section 465(1) that the individual must be UK resident in the tax year in which the gain arises.

(2) See section 540 for the cases in which such a deficiency is treated as arising, section 541 for how the deficiency is calculated and section 469(5) for the apportionment of deficiencies in cases where two or persons are interested in a policy or contract.

(3) Subsection (1) only applies for the purpose of determining the individual's extra liability.

(4) For this purpose, an individual's extra liability is the amount by which the individual's liability to income tax exceeds the amount it would be on the assumptions specified in subsections (5) and (6).

(5) It is assumed that income charged to tax at the higher rate is charged –

…

(b) … at the basic rate.

(6) It is assumed that income charged to tax at the dividend upper rate is charged at the dividend ordinary rate."

11. Although section 539(1) says that CDR is "allowable as a deduction from an individual's total income", it should be noted that, by virtue of subsection (3), the relief "only applies for the purpose of determining the individual's extra liability". The effect of the definition of "extra liability" in subsections (5) and (6) is that the relief applies only from income tax at the higher or dividend upper rates, just as the charge to tax under Chapter 9 is confined to a charge at the higher rate. Accordingly, it is only for the limited purpose of providing relief from the higher rate of income tax (an expression which I will use to include the dividend upper rate) that the taxpayer's total income is reduced by the amount of the deficiency under section 539(1).
12. It should also be noted that CDR is available only in the year of assessment in which it arises. There is no facility to carry the relief either forwards or backwards into other years.

(2) To what extent can CDR reduce the rate of CGT payable on chargeable gains in the same year of assessment?

13. CDR is a relief from the higher rate of income tax. Without special provision, it would not afford any relief from CGT. In the years with which we are concerned, the starting point (and default position) was that chargeable gains were taxable at a rate equivalent to the lower rate (not the basic rate) of income tax. This was the effect of section 4(1) of the Taxation of Chargeable Gains Act 1992 ("TCGA 1992"), which provided that:

"(1) Subject to the provisions of this section…, the rate of capital gains tax in respect of gains accruing to a person in a year of assessment shall be equivalent to the lower rate of income tax for the year."

14. Section 4(2) then provided:

"(2) If income tax is chargeable at the higher rate or the dividend upper rate in respect of any part of the income of an individual for a year of assessment, the rate of capital gains tax in respect of gains accruing to him in the year shall be equivalent to the higher rate."

Accordingly, if the taxpayer was subject to higher rate income tax in respect of any part of his income for the year, he would be liable to pay CGT on the whole of his chargeable gains for the year at a rate equivalent to the higher rate of income tax.

15. Subsections (3) and (4) then dealt with the position where the taxpayer was not liable to income tax at the higher rate on any of his income for the year, and part of his "basic rate band" was unused. The broad effect of the subsections was that an amount equivalent to the unused portion of the taxpayer's basic rate band was set against his chargeable gains and charged to CGT at the lower rate, while the higher rate was chargeable only to the extent that his total chargeable gains exceeded the unused portion of the basic rate band. The subsections provided as follows:

"(3) If no income tax is chargeable at the higher rate or the dividend upper rate in respect of the income of an individual for a year of assessment, but the amount on which he is chargeable to capital gains tax exceeds the unused part of his basic rate band, the rate of capital gains tax on the excess shall be equivalent to the higher rate of income tax for the year.

(4) The reference in subsection (3) above to the unused part of an individual's basic rate band is a reference to the amount by which… the basic rate limit exceeds his total income (as reduced by any deductions made in accordance with the Income Tax Acts)."

16. The expression "basic rate band" is not defined, but the "basic rate limit" was defined by section 1(3) of the Income and Corporation Taxes Act 1988 ("ICTA 1988"), which provided that:

"(3) The amount up to which an individual's income is by virtue of subsection (2) above chargeable for any year at the starting rate or the basic rate shall be known as the basic rate limit: …"

See too the definition of "basic rate" in section 832(1).

17. The basic rate limit for 2006/07 was £33,300. For 2007/08, it was £34,600. It is common ground that, if section 4(4) of TCGA 1992 stood alone, there could be no question of the basic rate limit exceeding Mr Scott's total income for either year. The reason for this is that CDR does not operate for the purposes of section 4 as a deduction from total income. Despite the apparently unqualified wording of section 539(1) of ITTOIA 2005, that subsection applies only for the purpose of determining Mr Scott's "extra liability" to income tax in each year: see section 539(3). "Total income" is defined in section 835(1) of ICTA 1988, in relation to any person, as meaning "the total income of that person from all sources estimated in accordance with the provisions of the Income Tax Acts". In 2006/07, Mr Scott's total income before deductions exceeded £44 million, including £41,300,247 in the form of gains on life insurance contracts. In 2007/08, his total income was over £8 million, including £13,863 in the form of gains on life insurance contracts. Accordingly, while CDR operated as a relief from the higher rate of income tax charged to Mr Scott in each year, it could not in the absence of further provision afford him relief from CGT at the higher rate on his very substantial chargeable gains of £8.84 million in 2006/07, and £14.71 million in 2007/08.
18. Mr Scott's argument depends on a special provision contained in section 6(2) of TCGA 1992. As amended and in force for 2006/07, section 6(2) provided that:

"Where for any year of assessment –

(a) by virtue of section 539 of ITTOIA 2005 (gains from contracts for life insurance etc) a deduction of an amount is made from a person's total income for the purposes of extra liability, or

…

section 4(4) shall have effect as if his income for the year were reduced by that amount."

Mr Scott's argument, in a nutshell, is that section 6(2) mandates a purely arithmetical calculation, whereby the full amount of the taxpayer's CDR may be deducted from income, producing a negative "total income" figure for the purposes of section 4 of TCGA 1992. This negative figure is in turn deducted from the basic rate limit as required by section 4, having the effect of extending the basic rate band beyond the basic rate limit. In this way, Mr Scott says that he is only liable to pay CGT at the lower rate, since his chargeable gains fall comfortably within the extended basic rate band to which he would be entitled under section 6(2).

Facts

19. There is little which needs to be added to the basic facts which I have already mentioned. The substantive issue which divides the parties is one of construction of the relevant legislation. At all stages, it has been mainly argued by reference to hypothetical examples and an assumed basic rate limit of £40,000. There was a statement of agreed facts, and the FTT heard no oral evidence from Mr Scott or his accountants.
20. Mr Scott's personal circumstances do, however, raise the issue with particular clarity, because of the very substantial scale of his investment in life insurance policies, the very large amounts of CDR available to him in the two tax years under appeal, and the substantial amounts of his chargeable gains which, if he is right, would fall to be taxed at the lower rate of 20%. We are not concerned with the underlying transactions which gave rise to these figures, and we do not know whether they formed part of a tax avoidance scheme. If they did form part of a scheme, there was no longer any dispute by the time when the closure notices were issued in February 2015 that Mr Scott was entitled to CDR in amounts which greatly reduced his exposure to higher rate income tax.
21. According to the skeleton argument of Michael Furness QC and Michael Firth, appearing for Mr Scott before us as they did before the UT, the amounts of CDR which Mr Scott claimed, and to which he was entitled, were £20,212,917 for 2006/07 and £22,613,114 for 2007/08. These amounts of relief are shown on the revised tax calculations which were provided by HMRC to Mr Scott with the closure notices.
22. Before the FTT, Mr Scott was represented by Mr Firth alone. HMRC have throughout been represented by Simon Pritchard.
23. It is also convenient to record at this stage that the relevant legislation which I have set out above was replaced in the second year under appeal, 2007/08, with modified versions pursuant to the tax law rewrite project. It is unnecessary to set out those later versions, because the parties are now agreed that there is no relevant difference in the meaning and effect of the legislation in force in the two years.

Discussion

24. The crux of the matter, as it seems to me, is whether counsel for Mr Scott are correct in their submission that section 6(2) of TCGA 1992 has the effect that "total income", for the purposes of section 4(4), can be reduced to a negative amount of less than zero. If that proposition holds good, it is then said that the very large amounts of CDR to which Mr Scott was admittedly entitled in each tax year (some £20.2 million in 2006/07, and £22.6 million in the following year) operate to provide relief from CGT as if his basic rate band had been extended by the same amount, with the consequence that all gains falling within the extended band are chargeable to tax at the lower rate of 20% instead of the higher rate of 40%.
25. Mr Furness submits that this is indeed the correct construction of section 6(2). The first step in the argument focuses on the opening words of the subsection: "[w]here for any year of assessment… by virtue of section 539 of ITTOIA 2005… a deduction of an amount is made from a person's total income for the purposes of extra liability". That condition is clearly satisfied, says Mr Furness, and the "amount" referred to can only be the amount of CDR actually taken into account in Mr Scott's income tax computations. So far, I agree.
26. The next step in the argument focuses on the mandatory consequence which follows when the opening condition is satisfied:

"section 4(4) shall have effect as if his income for the year were reduced by that amount."

The "amount" must again be the same amount of CDR as in the opening words of the subsection, and section 4(4) is now to take effect on the hypothesis ("as if") Mr Scott's income for the year were reduced by that amount.

27. Mr Furness accepts that, in the normal way, a person's total income cannot be reduced below zero. In the absence of a special provision, a relief which is set against income is exhausted, and ceases to have effect, once the income is reduced to nil. There is no meaningful concept of "negative income", although legislation may of course make express provision for losses (such as trading losses) to be utilised in various ways. In the present context, however, Mr Furness submits that section 6(2), read together with section 4(4), does in principle recognise and give effect to a deduction which exceeds the amount of total income. It does so, unusually, by means of an adjustment to the rate of CGT rather than an income tax relief, but that is because the relevant statutory purpose is to harmonise the rates of the two taxes. Accordingly, when making the deduction required by section 6(2), one must deduct the full amount of the CDR referred to in section 6(2)(a) and, if that produces a negative figure, so be it. Only in this way, says Mr Furness, can effect be given to the parliamentary intention of unifying the rates of income tax and CGT.
28. At this point, I should mention a puzzling feature of the actual figures we have seen for Mr Scott's income tax liability in 2006/07. As I have explained, Mr Scott's total income for 2006/07 was more than £44 million, so reducing that total income by the available CDR of approximately £20.2 million would still leave a figure greatly in excess of Mr Scott's basic rate limit, even after taking into account the other deductions from total income apparently shown on the tax calculation we have seen. On that footing, it is hard to see how any part of Mr Scott's basic rate band could have remained unused for CGT purposes. This point was not taken by HMRC, however, and we heard no argument on it. The explanation may be that we have an incomplete picture of Mr Scott's actual income tax position in 2006/07, precisely because the question of principle has throughout been argued on assumed figures. Furthermore, the apparent difficulty which I have mentioned clearly does not arise in 2007/08, when the CDR of approximately £22.6 million available to Mr Scott was on any view far greater than his total income of about £8 million. I am therefore content to proceed on the basis that, in each year, the CDR available to Mr Scott significantly exceeded his total income.
29. Making that assumption, I am unable to see why the statutory objective of harmonising the rates of the two taxes should lead to the conclusion that a relief from income tax (CDR) should also operate as a corresponding relief from CGT. As Mr Pritchard submitted for HMRC, Parliament has not chosen to assimilate the structure and reliefs of the two taxes, and as a matter of substance they continue to operate in separate spheres. I am therefore satisfied that it would be wrong in principle to approach the construction of section 6(2) on the basis that a way should if possible be found to permit the CDR, which (ex hypothesi) has already been used to reduce the taxpayer's higher rate income tax liability, to bring about a similar reduction in his higher rate CGT liability. That would be to turn a relief from income tax into a corresponding relief from CGT as well. I can find no indication that such an intention should be attributed to Parliament. If that had indeed been the legislative intention, one would expect to find clear and specific provision to that effect in language similar to that used in section 539 of ITTOIA 2005, rather than a special provision ostensibly concerned only with the rates of charge to CGT.
30. If section 6(2) is approached without any preconceptions as to its purpose, it seems reasonably clear to me (as it did to the two tribunals below) that its function is far more limited. The effect of section 6(2) is to modify section 4(4) by the introduction of a hypothesis. Leaving aside section 6(2), the effect of section 4(4) is merely to treat the unused part (if any) of a taxpayer's income tax basic rate band as transferable when computing his liability to CGT. Thus, for example, if his total income (after deductions) were £25,000, and the basic rate limit were £40,000, he would be able to transfer the unused £15,000 of his basic rate band to cover the first £15,000 of his chargeable gains. By virtue of section 4(1), CGT would then be chargeable at the lower rate of 20% on the first £15,000 of gains, while the remainder of his chargeable gains in excess of £15,000 would be charged at the higher rate of 40%: see section 4(3).
31. How, then, does the hypothesis in section 6(2) impact on that basic situation? We are told that section 4(4) is to have effect as if the taxpayer's "income for the year" were reduced by the amount of the CDR for which he had obtained relief from higher rate income tax. Since section 4(4) is concerned with ascertaining the difference between a person's total income and the basic rate limit, I consider that the word "income" in the final limb of section 6(2) must also refer to the taxpayer's total income. Indeed, Mr Furness appeared to accept this. Thus the effect of the hypothesis is that, for the purposes of section 4(4), the CDR is now to be treated as a deduction from the taxpayer's total income. The deeming is needed in order to achieve this purpose, because section 539 of ITTOIA 2005 does not by itself create a deduction from total income, despite the rather misleading appearance to the contrary in the wording of subsection (1). As we have seen, the deduction merely applies for the limited purpose of determining the taxpayer's liability to higher rate income tax: section 539(3).
32. Reverting to section 4(4), if the taxpayer's total income, as (now) reduced by CDR, amounts to less than his basic rate limit, the unused portion of the basic rate band will afford relief from CGT at the higher rate in the way I have described. Returning to the simple example in [30] above, suppose that the taxpayer's total income (after deductions other than CDR) were £525,000, and he had available CDR of £500,000. The effect of the relief, for income tax purposes, would then have been to eliminate his liability to higher rate tax. In fact, he would only have needed to use so much of the CDR as reduced his total income to £40,000, i.e. £485,000. But another way of looking at it would be to say that he used the entire CDR of £500,000, leaving £25,000 to set against his basic rate limit. Either way, the effect of the calculation in section 4(4), as modified by section 6(2), is to ensure that the taxpayer can have the benefit of an amount equal to the unused part of his basic rate band when computing his liability to CGT.
33. Further than that, in my view, section 6(2) does not go. Once the taxpayer's total income has been notionally reduced to zero in the section 6(2) calculation, the relief affordable by the CDR can go no further. There is no longer any income which can be reduced, because negative total income is not a meaningful concept. As Mr Pritchard said in his admirably clear written and oral submissions, the amount of a deduction is not necessarily the same as the amount of available relief, because the deduction cannot exceed the amount needed to extinguish the extra liability that existed before the deduction. Further, as a matter of the ordinary use of language, "income" denotes a positive sum or amount of money or money's worth, whereas "negative income" is not income at all but would be a form of outgoings or loss.
34. Authority is hardly needed for the proposition that statutory language should normally be taken to bear its ordinary meaning in the appropriate context, but Mr Pritchard reminded us of what Lord Nicholls said in R v Environment Secretary, Ex p Spath Holme Limited [2001] 2 AC 349 at 397B:

"In identifying the meaning of the words used, the courts employ accepted principles of interpretation as useful guides. For instance, an appropriate starting point is that language is to be taken to bear its ordinary meaning in the general context of the statute."

See too the observations of Lord Neuberger PSC in Edwards v Kumarasamy [2016] UKSC 40, [2016] AC 1334, at [17]:

"Unless the natural meaning of the words of a statutory provision produces a nonsensical result, or a result which is inconsistent with the intention of the legislation concerned, as gathered from admissible material, the words must be given their ordinary meaning."

35. To my mind, there is no real doubt about the meaning of either section 4(4) or section 6(2), once it is appreciated that CDR does not take effect (under section 539 of ITTOIA 2005) as a true deduction from total income, and once the limited purpose of section 4(4) is properly understood. If Parliament had wished to extend the taxpayer's basic rate band for the purposes of the charge to CGT by the full amount of the CDR which the taxpayer was able to use for income tax purposes, it would have said so clearly and explicitly. Indeed, as Mr Pritchard pointed out, the natural way of achieving such an objective would have been by expressly extending the taxpayer's basic rate limit, as it did in relation to Gift Aid in section 414 of the Income Tax Act 2007. That section states:

"(1) An individual who makes a gift to a charity which is a qualifying donation is entitled to the relief set out in subsection (2).

(2) The Income Tax Acts have effect in their application to the individual for the tax year in which the gift is made as if –

(a) the gift had been made after deduction of income tax at the basic rate, and

(b) the basic rate limit… were increased by an amount equal to the grossed up amount of the gift."

36. The point of statutory construction is ultimately a short one, which does not need further elaboration. Mr Furness advanced various arguments about the assumed statutory purpose, and sought to criticise the UT for failing to distinguish between Parliament's general aim of unifying the rates of income and CGT, which the UT thought could not control the interpretation of section 6(2), on the one hand, and a statutory purpose of rate unification, which he submitted should be applied so as to guide the choice between two possible constructions of section 6(2), on the other hand.
37. In my respectful opinion, these submissions miss the mark. I agree with HMRC that there is no ambiguity in section 6(2), or in the way in which it interacts with section 4(4), so this is not a case where purposive considerations enable the court to choose between two rival constructions the one which is more likely to represent Parliament's intention. Nor do I think it matters whether the unification of tax rates is described as a general aim or a purpose of the relevant legislation. It is always necessary to begin by construing the actual wording which Parliament has used, in its full legislative context. That exercise has led me to conclude, in the same way as the tribunals below, that section 6(2) has only the limited purpose which I have described. On a natural reading, the language of the subsection, taken in conjunction with section 4(4), simply does not admit of the ambitious construction for which Mr Furness contends, and no amount of appeals to the supposed underlying statutory purpose can alter that position.

Conclusion

38. For these reasons, I would dismiss the appeal.

Nicola Davies LJ:

39. I agree.

The Master of the Rolls:

40. I also agree.